Intensity-modulated radiotherapy for the treatment of prostate cancer: a systematic review and economic evaluation

Notes

Permissions

Chapter 1 Background

Description of health problem

Prostate cancer (PC) is the most common cancer in men in England and Wales and constitutes approximately 25% of the new diagnoses of malignant cancer in men in England and Wales. 1 The incidence appears to be rising. 1 It is also considered to be the most common malignant disease in males in Western Europe and North America. 2 The focus of this systematic review will be to address the question ‘what is the clinical and cost-effectiveness of intensity-modulated radiotherapy (IMRT) for the radical treatment of PC compared with three-dimensional radiotherapy (3DCRT) and radical prostatectomy?’.

Aetiology, pathology and prognosis

The specific causes of PC remain unknown. 2 Hsing and Chokkalingam provided a comprehensive review of PC epidemiology3 and reported that the risk of developing PC was related to: age, genetics, and family history of PC. It was reported that putative risk factors include: obesity, hormones, smoking, dietary factors, physical inactivity, occupation, vasectomy, genetic susceptibility, and sexual factors have also been implicated.

Localised PC (confined within the prostatic capsule) is usually asymptomatic in the early stage. 4 However, with regard to locally advanced PC it is because it is frequently asymptomatic in the early stage that it often presents at a more advanced stage. Lower urinary tract symptoms (LUTS) of frequency, urgency, hesitancy, terminal dribbling and/or overactive bladder can be related to benign prostatic enlargement (BPE) alone, but can also present in early PC and in locally advanced PC cases. 5 The LUTS often present in a similar way to symptoms of benign prostatic hyperplasia within locally advanced PC. 5 However, by the time PC itself causes LUTS, it may have reached an advanced and incurable stage. 4 Garrick suggests that a new onset of impotence should also be recognised. 6 PC frequently develops into bone metastases which causes pain. An abnormal or rising prostate-specific antigen (PSA) level is the most common indication of PC in the UK health-care system. In the UK, a diagnosis of PC is typically prompted by urological symptoms and biochemical information (particularly PSA). 7 Collin et al. conclude in a case-control study nested within the UK population-based Prostate testing for cancer and Treatment (ProtecT) study, that a history of LUTS before PSA testing marginally improves the prediction of an individual’s risk of PC. 8 The PSA test and other ways to measure PC are described in more detail in Subgroups of patients with PC for whom radiotherapy (RT) may be indicated.

Prostate cancer is no longer seen as solely an indolent disease that rarely results in mortality for PC patients. 9 Johansson et al. examined the natural course of early localised PC in a consecutive sample of 223 initially untreated patients observed over 21 years and found that although most PCs diagnosed at an early stage have an indolent course, local tumour progression and aggressive metastatic disease may develop in the long term. 9 The findings support early radical treatment notably among patients with an estimated life expectancy exceeding 15 years. 9 Furthermore, as men’s life expectancy increases, the effect of untreated PC may need to be re-evaluated. A recent study by Widmark et al. 10 showed that in patients with locally advanced or high-risk localised PC, addition of local RT to endocrine treatment halved the 10-year PC-specific mortality (23.9% in the endocrine alone group and 11.9% in the endocrine plus RT group), and substantially decreased overall mortality with fully acceptable risk of side-effects compared with endocrine treatment alone. In the light of these data endocrine treatment plus RT may be regarded as the new standard. 10

Ongoing research may clarify how best to treat men with PC. With regard to early PC, the ProtecT randomised controlled trial (RCT) aims to evaluate the clinical effectiveness, cost-effectiveness and acceptability of active monitoring, radical prostatectomy and radical conformal radiotherapy (CRT) for men with localised PC. 8 In addition, the ProSTART study, a pilot study for an international phase III randomised parallel group trial in patients with favourable risk PC (as defined as clinical stage T1b, T1c, T2a or T2b at time of diagnosis), is comparing active surveillance therapy against radical treatment (prostatectomy or RT). 11 With regard to more advanced PC, the Medical Research Council (MRC) PR07 study,12 is investigating the role of RT in locally advanced PC. In this trial, men with locally advanced non-metastatic PC are randomised between hormone therapy alone and hormone therapy plus RT. However, it is noted that if the subjects are allocated RT [given to the prostate and pelvis (or prostate alone if necessary)] in doses up to 69 Gy this is below the usual IMRT dose level. 12

Incidence and prevalence

The incidence of PC in the UK, in common with many other countries, has been rising. 13 In 2006, there were 35,515 new cases of PC diagnosed in the UK. 14 Despite the large increase in incidence, the mortality rate has been relatively stable. 14

Prostate cancer frequently progresses slowly and men with less aggressive disease rarely die of their cancer, but this is not the case for those men with the most aggressive tumours (poorly differentiated). 2 Table 1 shows the percentage of men in whom PC was detected at autopsy in the USA. 4,15 Burford et al. reported that 93% of PC deaths occur in the 65 and over age group. 4 By the age of 80 years, approximately 80% of men will have some cancer cells in their prostate. Table 1 indicates that prevalence of PC increases with age, but is not insignificant at younger ages. 4,15

TABLE 1 - Percentage of men with prostate cancer at autopsy in the USA4,15

Age (years)	20–29	30–39	40–49	50–59	60–69	70–79
Percentage of men in whom PC was detected at autopsy	8	28	39	53	66	80

However, the extent of age-adjusted PC incidence rates varies considerably throughout the world. Between 1996 and 2006 the age-standardised (to the world standard population) rate in the UK increased by nearly 38%. 16 The lifetime risk of PC diagnosis has been estimated at 10%. 14 Further statistics concerning the incidence rates of PC in the UK during 2006 are reported in Table 2. The majority of cases are in England; however, in terms of the crude rate per 100,000 population, Wales has the highest rate and Scotland has the lowest rate.

TABLE 2 - Number of new cases and rates of prostate cancer in the UK during 200614

	England	Wales	Scotland	Northern Ireland	UK
Cases
Male	30,024	2164	2,506	821	35,515
Crude rate per 100,000 population
Male	120.5	149.8	101.5	96.2	119.6
Age-standardised rate (European) per 100,000 population
Male	98.1	108.2	81.6	92.1	97.1
95% CI	97.0 to 99.2	103.7 to 112.8	78.4 to 84.8	85.8 to 98.4	96.1 to 98.1

Prostate cancer risk is strongly related to age; very few diagnosed cases are registered in men under 50 years and three-quarters of cases occur in men over 65 years. The largest number of cases is diagnosed in those aged 70–74 years. 14 Figure 1 reports the age-specific incidence rates for PC in the UK during 2006.

FIGURE 1.

Numbers of new cases and age-specific incidence rates of male prostate cancer in the UK during 2006. 14

Impact of health problem

Prostate cancer is a major health problem and is a significant burden to patients and the NHS in England and Wales. 13 Sutcliffe et al. 2 reported that PC is a primary reason for consultation with general practitioners (GPs) among men with cancer. It has been estimated that PC in England and Wales costs the health service at least £45M per year. 17 This is likely to be increased by the burden of disease from bone metastasis (and prolonged palliative care) and this has repercussions in terms of cost to society, decreased quality of life and decreased survival. 17 In addition there is a significant patient burden of stress and anxiety, sexual function, urinary function, bowel function18 and potential loss of earnings.

Subgroups of patients with prostate cancer for whom radiotherapy may be indicated

Several patient subgroups exist where RT may be used:

• primary radical treatment of localised cancers

• primary radical treatment of locally advanced cancers (prostate or whole pelvis)

• adjuvant RT treatment for high risk radical prostatectomy patients (prostate or whole pelvis)

• salvage treatment (prostate or whole pelvis)

• palliation of bone metastases.

The latest National Institute for Health and Clinical Excellence (NICE) clinical guideline 20081 states that several factors have been shown to predict the risk of recurrence after treatment of localised PC. These include the Gleason score, the serum PSA level, and the clinical T stage [according to the tumour node metastasis (TNM) staging system]. Table 3 provides a fuller description of the TNM staging system. These predictive factors have been used to classify localised PC into recurrence risk groups, specifically:

• low-risk: PSA < 10 ng/ml and Gleason score ≤ 6 and clinical stage T1–T2a

• intermediate-risk: PSA 10–20 ng/ml or Gleason score 7 or clinical stage T2b–T2c

• high-risk: PSA > 20 ng/ml or Gleason score 8–10 or clinical stage T3–T4. 1

TABLE 3 - Tumour node metastasis staging system (2002)25

Primary tumour, clinical (T)
TX	Primary tumour cannot be assessed
T0	No evidence of primary tumour
T1	Clinically unapparent tumour not palpable or visible by imaging
T1a	Tumour incidental histologic finding in ≤ 5% of tissue resected
T1b	Tumour incidental histologic finding in > 5% of tissue resected
T1c	Tumour identified by needle biopsy (because of elevated PSA level); tumours found in one or both lobes by needle biopsy but not palpable or reliably visible by imaging
T2	Tumour confined within prostate
T2a	Tumour involving less than or equal to half a lobe
T2b	Tumour involving more than half a lobe but not more than one lobe
T2c	Tumour involving both lobes
T3	Tumour extending through the prostatic capsule; no invasion into the prostatic apex or into, but not beyond, the prostatic capsule
T3a	Extracapsular extension (unilateral or bilateral)
T3b	Tumour invading seminal vesicle(s)
T4	Tumour fixed to or invading adjacent structures other than seminal vesicles (e.g. bladder neck, external sphincter, rectum, levator muscles, pelvic wall)
Primary tumour, pathological (PT)
PT2	Organ-confined
PT2a	Tumour involves one half of one lobe, but not both lobes
PT2b	Tumour involves more than one half of one lobe, but not both lobes
PT2c	Tumour involves both lobes
PT3	Extraprostatic extension
PT3a	Extraprostatic extension
PT3b	Seminal vesicle invasion
PT4	Invasion of bladder, rectum
Regional lymph nodes (N)
NX	Regional lymph nodes (cannot be assessed)
N0	No regional lymph node metastasis
N1	Metastasis in regional lymph node or nodes
Distant metastasis (M)
PM1c	More than one site of metastasis present
MX	Distant metastasis cannot be assessed
M0	No distant metastasis
M1	Distant metastasis
M1a	Non-regional lymph node(s)
M1b	Bone(s)
M1c	Other site(s)
Stage grouping
Stage I	T1a	NO	MO	G1 (Gleason Score 2–4)
Stage II	T1a	NO	MO	G2, 3–4(Gleason Score 5–10)
	T1b	NO	MO	Any G
	T1c	NO	MO	Any G
	T1	NO	MO	Any G
	T2	NO	MO	Any G
Stage III	T3	NO	MO	Any G
Stage IV	T4	NO	MO	Any G
	Any T	N1	MO	Any G
	Any T	Any N	M1	Any G

Adjuvant and salvage radiotherapy

Adjuvant RT is an additional treatment given to patients with PC after localised therapy, such as surgery, which ‘assists’ the primary treatment. Adjuvant RT may be offered to some patients in the UK, but until recently was not standard treatment. The NICE PC guideline1 advised that immediate post-operative RT after radical prostatectomy is not routinely recommended, even in men with margin-positive disease, other than in the context of a clinical trial. However, recent trial data appears to supersede the NICE guideline. Cozzarini et al. 19 conducted a phase I to II trial with 50 patients; they found excellent acute and early late toxicity outcomes of a moderately hypofractionated regimen with post-operative early adjuvant RT delivered by helical tomotherapy, which is a form of adaptive IMRT technology. In addition, Wiegel et al. 20 examined 192 men assigned to a wait-and-see policy compared with 193 men assigned to immediate post-operative RT. These authors examined the primary end point of biochemical progression-free survival. Wiegel et al. concluded that adjuvant RT for T3 PC patients with postoperatively undetectable PSA significantly reduces the risk of biochemical progression; however, further follow-up is needed to assess the effect on metastases-free and overall survival. Further recent evidence by Thompson et al. 21 from the South West Oncology Group supports that adjuvant RT after radical prostatectomy for men with PC graded as T3N0M0 using the TNM classification system significantly reduces the risk of metastasis and increases survival.

Salvage RT is performed after surgery to help eliminate any remaining cancerous cells in patients with biochemical relapse. Salvage treatment in the form of radical RT treatment to the prostatic bed should be offered to patients after radical prostatectomy treatment if they have PSA failure, but there is no evidence of metastases. It is uncertain whether adjuvant hormonal therapy should also be given.

Whole pelvis radiotherapy

The role of whole pelvis RT (to include the treatment of pelvic lymph nodes) for primary or salvage treatment is controversial. The treatment may result in an increased risk of adverse effects and these patients have a relatively poor long-term survival (< 10 years). However, the use of IMRT could potentially reduce the toxic effects of treatment due to a more focused delivery of RT. The 2008 PC guideline1 recommends that treatment to the prostate alone is currently the standard approach to radical RT for PC in the UK. In common with other cancer sites (e.g. breast), there may be a benefit from treating regional lymph nodes as well. However NICE also recommends that clinical oncologists should consider pelvic RT in men with locally advanced PC who have a > 15% risk of pelvic lymph node involvement, who are to receive neoadjuvant hormonal therapy and radical RT.

Measurement of disease

Several methods of measurement of PC currently exist. The following section presents the most common measurement methods (which are predictive parameters of disease) seen in studies of IMRT.

Gleason grading system

The initial diagnosis of PC requires confirmation and is performed via histological examination of prostate tissue from trans-rectal ultrasonography (TRUS) biopsy samples. 4 A TRUS biopsy involves taking 10–12 cores of prostatic tissue through the rectum under ultrasound guidance and the result reveals the level of tumour differentiation. 4 The predominant histological system for tumour differentiation is the Gleason grading system via analysing the most common tumour patterns. Each tumour pattern is assigned a grade (1 to 5) and these grades are combined to produce the Gleason score of 2–10. 4 Burford et al. 4 indicate that the lower the score, the more well differentiated the tumour, the less likely the tumour is to progress and the better the prognosis. Tumours can be classified into low grade (≤ 6), intermediate grade (= 7) and high grade (8–10). Berney et al. 22 reassigned Gleason grades to 1789 localised PC patients in the UK between 1991 and 1996 and found that there was significant reassignment in the Gleason score with increases of Gleason score across a wide spectrum of patients (the same data were reclassified as far as it is possible to determine). A drift in Gleason score over the past decade has also been reported by Veldeman et al. 23 This relates to the observation that PCs are now commonly graded higher than in previous decades due to diagnostic improvements, resulting in a greater percentage of higher-grade PCs.

TNM staging system and ABCD system

Developed in France in the 1940s by Pierre Denoix, the TNM classification has become the accepted basis of cancer staging24 and has undergone several revisions. Table 3 shows the 2002 version (Sixth Edition),25 as, where stated, the studies included in this review use either the 1992 or 2002 version. However, a later version (Seventh Edition) was published in 2009. 25 In Europe, the TNM staging system has been most commonly used to establish how far the disease has progressed. 26 The letter T refers to the size of the primary tumour, N describes the extent of lymph node involvement and M refers to the presence or absence of metastases. 27 T1 and T2 are considered to be localised PC and both T3 and T4 are often referred to as locally advanced disease. However, some studies include patients classified as stages T1, T2 and T3 in the study of localised PC. 28,29 There are two key changes to the 2002 TNM classification system compared with the older versions, these being: (1) subdivision of the T2 disease into three clinical substages, and (2) recommendation that the Gleason scoring system is used for grading. 2

The clinical stage is based on information obtained before surgery to remove the tumour and can be limited since the information is obtained by making indirect assessment of the tumour while it is still in the patient. 2 Pathologic staging provides additional information from the microscopic examination of the tumour2 and provides a direct examination of the tumour and its spread.

A second staging method called the ABCD (modified Whitmore-Jewett) system is often used in the USA. It has been indicated that a weakness of the ABCD system is its inability to characterise regional lymph nodes (N+) or distant metastases (M+) relative to the category of the lesion and that N+ or M+ lesions are categorised as stage. 30 Table 4 shows the ABCD grading system.

TABLE 4 - ABCD grading system31

Stage A Very early and without symptoms; cancer cells confined to the prostate A1 Well differentiated and slightly abnormal cancer cells A2 Moderately or poorly differentiated and abnormal cancer cells in several locations within the prostate Stage B Confined to the prostate, but palpable (detectable by DRE) and/or detectable by elevated PSA B0 Confined to the prostate, non-palpable; PSA elevated B1 Single cancerous nodule in one lobe of the prostate B2 Extensive, involvement in one or both prostate lobes Stage C Cancer cells found outside the prostate capsule (membrane covering the prostate); spread confined to surrounding tissues and/or seminal vesicles C1 Extends outside the prostate capsule C2 Bladder or urethral obstruction Stage D Metastasis (spread) to regional lymph nodes, or to distant bones, organs (e.g. liver, lungs), and/or other tissues D0 Metastatic, clinically localised, and showing elevated blood prostatic acid phosphatase levels D1 Regional lymph nodes involved D2 Distant lymph nodes, bones, or organs involve D3 Metastatic disease after treatment

Table 5 is cited as Jewett31 within Sutcliffe. 2 Table 5 shows a comparison between the TNM classification and the Whitmore-Jewett system which was presented in Selley et al. 30

TABLE 5 - Comparison of TNM and Whitmore-Jewett classifications for staging prostate cancer adapted from Selley et al.30

TNM classification	Whitmore-Jewett
T1	A
T1,T2	A1
T2,T3	A2
T2	B
T2a,T2b	B1
T2c	B2
T3	C
T3a,T3b	C1
T3c	C2
M1 and N1	D

Serum prostate-specific antigen level

A further method for measuring and assisting the diagnosis of PC and assessing the risk of disease burden is the PSA level. PSA is a glycoprotein, almost exclusively produced by the epithelium of the prostate gland, responsible for liquefying semen and allowing sperm to swim freely. 4,13 Burford et al. report that due to an alteration in the architecture of the prostate in conditions such as prostatitis and BPE as well as PC, PSA leaks out, leading to increased levels in the bloodstream. 4 Collin et al. describe the test’s limited sensitivity and specificity which may lead to false reassurance or false alarm followed by invasive investigations. 8 The PSA test remains an imperfect predictor of PC, but remains a useful tool that is widely used in clinical practice. 32 However, until a more superior diagnostic blood test becomes widely available, screening for PC will continue to comprise PSA testing in conjunction with digital rectal examination (DRE) and fine-core biopsies of the prostate when indicated by an abnormal PSA and/or DRE. 33 Rosario et al. 34 (the ProtecT trial) indicate that the exact level at which to recommend biopsy is controversial and may start as low as 2.5 ng/ml; however, they conclude that following an initial PSA of 3.0–19.99 ng/ml in men aged 50–70 years, a repeat PSA within 7 weeks allows more accurate risk prediction that may assist in the decision-making as to whether or not to proceed with prostate biopsy.

Risk of recurrence

The latest NICE clinical guideline 20081 discusses predictive factors which have been used to classify localised PC into recurrence risk groups specifically:

• low-risk PSA < 10 ng/ml and Gleason score ≤ 6 and clinical stage T1–T2a

• intermediate-risk PSA 10–20 ng/ml or Gleason score 7 or clinical stage T2b–T2c

• high-risk PSA > 20 ng/ml or Gleason score 8–10 or clinical stage T3–T4.

The above risk categories closely resemble the recurrence risk groups of PC for clinically localised PC stated by the National Comprehensive Cancer Network (NCCN) in their Clinical Practice Guidelines in Oncology. 35 All NCCN recurrence risk groups are presented below:

• Clinically localised:

  1. – low-risk PSA < 10 ng/ml and Gleason score 2–6 and clinical stage T1–T2a

  2. – intermediate-risk PSA 10–20 ng/ml or Gleason score 7 or clinical stage T2b–T2c

  3. – high-risk PSA > 20 ng/ml or Gleason score 8–10 or clinical stage T3a

• locally advanced:

  1. – very high risk T3b–T4

• metastatic:

  1. – any T, N1

  2. – any T, any N, M1

  3. – patients with multiple adverse factors may be shifted into the next highest risk group. 35

Key measurement systems for adverse events

The most common measurement systems of adverse events (AE) is published by the Radiation Therapy Oncology Group (RTOG) based in Philadelphia, PA, USA. The RTOG is a national clinical cooperative group funded by the National Cancer Institute (NCI) since 1968 to increase the survival and improve the quality of life of patients diagnosed with cancer. 36 The RTOG grades are embedded within the system of AE measurement called the NCI–Common Toxicity Criteria, often abbreviated to NCI-CTC. The AEs are graded from 0 to 4 for many different types of cancer. 37 However, from January 2009 the Cancer Therapy Evaluation Program at the NCI recommends the use of the Active Common Terminology Criteria for Adverse Events (CTCAE), Version 4.0 (http://evs.nci.nih.gov/ftp1/CTCAE/About.html). The RTOG grades range from 1 to 5, grade 1 meaning a mild AE to grade 5 meaning death; the CTCAE lists many disorders such as gastrointestinal (GI) disorders and then lists subcategories such as rectal haemorrhage together with the definitions of grades 1–5.

Neoadjuvant/adjuvant hormone therapy

Studies also differentiate between patients who have, or have not, received adjuvant hormone treatment at baseline. As PC is driven in part by male sex hormones, the use of hormonal treatment to reduce the level of circulating male hormones is a potentially very useful method of treating all stages of this disease. 38 This technique is also sometimes called androgen deprivation or suppression. Neoadjuvant hormone therapy in the management of PC translates into administering hormone treatment before the primary therapy to assist in reducing tumour burden. Reviews suggest that adjuvant hormone therapy combined with either prostatectomy or RT can improve disease-free survival in patients with local or locally advanced PC. 38,39 Significant local control may be achieved when given prior to prostatectomy or RT, which may improve the patient’s quality of life. Neoadjuvant hormone therapy is associated with significant clinical benefit when given with RT and improves pathological outcome prior to prostatectomy, but is of minimal value prior to radical prostatectomy. 40 Neoadjuvant hormones also have an important role in reducing tumour volume and therefore potentially reducing dose to the rectum and hence toxicity.

Key health-related quality of life measures

Health-related quality of life (HRQoL) is also measured in relation to PC patients receiving IMRT. HRQoL may play a crucial role in determining treatment modality as it is likely that patients with localised PC will live for a long time. 41 However, despite improvement in survival, many survivors are at risk of post-treatment psychological and physical sequelae which may reduce their quality of life. 42 HRQoL measures for PC in relation to IMRT include the following: Lips et al. 43 used the RAND-36 generic health survey, a cancer-specific QoL measure by the European Organisation for Research and Treatment of Cancer (EORTC) core questionnaire [QLQ-C30(+3)] and the Prostate Tumour-Specific Questionnaire by the EORTC PC module(QLC-PR25). Lips et al. found that IMRT and accurate position verification seem to provide a possibility to increase the radiation dose for PC without deterioration in HRQoL. In addition, within PC the well validated Medical Outcomes Survey Short Form questionnaire-36 items (SF-36) has been used to measure physical and mental components of HRQoL and the 20-item University of California Los Angeles Prostate Cancer Index (UCLA PCI) has been used to measure treatment-specific function and bother (bowel, urinary, sexual). Higher scores represented better function for both SF-36 and the PCI. 42

Definitions of biochemical failure

As an alternative to the clinical measurement of disease-free survival (local disease recurrence, or the development of metastatic disease, or both), PSA testing has largely replaced clinical failure as a measure of treatment efficacy as it is easier to perform on a routine basis than serial bone scans and computed tomographic (CT) scans. 44 The PSA level has been used as a surrogate endpoint in trials of PC. The American Society for Therapeutic Radiology and Oncology (ASTRO) Consensus Panel44 state that ‘biochemical’ failure has come to be widely used in the absence of clinical or histopathology evidence of local persistence or recurrence or demonstrable distant metastasis. An ASTRO consensus panel in 1996 agreed on four guidelines:

1. Biochemical failure is not justification per se to initiate additional treatment. It is not equivalent to clinical failure; it is, however, an appropriate early end point for clinical trials.

2. Three consecutive increases in PSA is a reasonable definition of biochemical failure after RT. For clinical trials, the date of failure should be the midpoint between post-irradiation nadir PSA and the first of the three consecutive rises.

3. No definition of PSA failure has as yet been shown to be a surrogate for clinical progression or survival.

4. Nadir PSA is a strong prognostic factor. 44

However, these guidelines have been criticised as this definition was not linked to clinical progression or survival; it performed poorly in patients undergoing hormonal therapy, and backdating biased the Kaplan–Meier estimates of event-free survival. 45 Consequently, as reported in Roach et al. 45 a second consensus panel convened in 2005 and revised the ASTRO definition. The panel recommended:45

1. A rise by 2 ng/ml or more above the nadir PSA be considered the standard definition for biochemical failure after external beam RT with or without hormonal therapy.

2. The date of failure be determined ‘at call’ (not backdated). They recommended that investigators be allowed to use the ASTRO consensus definition after external beam RT alone (no hormonal therapy) with strict adherence to guidelines as to ‘adequate follow-up’. To avoid the artefacts resulting from short follow-up, the reported date of control should be listed as 2 years short of the median follow-up. For example, if the median follow-up is 5 years, control rates at 3 years should be cited.

Retaining a strict version of the ASTRO definition would allow comparisons with a large existing body of literature. 45 The current review reports, where available and appropriate, the definition of biochemical failure, as used in the studies to allow full comparison.

Kupelian et al. 46 reported that the length of the natural history of localised PC makes the relationship between the biochemical failure and overall survival difficult to establish for patients diagnosed in the PSA era and that biochemical failure after definitive RT for localised PC is not associated with increased mortality within the first 10 years after initial therapy, although a trend toward worse outcome was observed at 10 years. With longer follow-up from initial therapy, significant differences may be observed at 15 or 20 years after therapy and may assist our full understanding of the impact of biochemical failure on overall survival. 46

Chapter 2 Definition of the decision problem

Chapter 3 Assessment of clinical effectiveness

Results

Quantity and quality of research available

Number of studies identified

A flow chart describing the process of identifying relevant literature can be found in Figure 2. Following the removal of duplicates our searches identified 1060 potentially relevant papers. A total of 896 papers not meeting our inclusion criteria were removed at title sift, leaving a total of 164 papers to be screened at abstract sifting stage. Of these, eight studies and five abstracts (in total 13 studies described in 27 publications) were concerned with IMRT compared with 3DCRT and no studies were identified comparing IMRT with radical prostatectomy.

FIGURE 2.

QUOROM (quality of reporting of meta-analyses) flow chart of study selection for the review of clinical effectiveness.

Number and type of studies excluded

In total 83 publications were excluded from those retrieved and inspected. A list of the 83 excluded papers at full paper sift with reasons for specific exclusions are provided in Appendix 5.

Number and type of studies included

There were eight studies meeting inclusion criteria for this review that were published in peer-reviewed journals, six of which were of localised PC, and two on locally advanced PC (see Included studies). In addition, five conference abstracts were considered to meet the inclusion criteria, but are considered separately as localisation of cancer is unclear, as abstracts necessarily provide less information about the studies (see Included conference abstracts).

Included studies

There were eight included comparative studies published as full reports in peer-reviewed journals: Kupelian;28,84–87 Sanguineti et al. ;88 Shu et al. ;29 Vora et al. ;89,90 Yoshimura et al. ;41 Zelefsky;91–99 Ashman et al. 100 (this was part of the Zelefsky study, but is classified here as a different study as it has distinct treatment and population characteristics, suggesting that the patients do not overlap with those from the other Zelefsky publications); and Lips et al. 43

The population for six included studies was patients with localised PC. 28,29,41,88,89,91 For two included studies the population was patients with locally advanced PC. 43,100

Four of the studies were retrospective patient records studies. 28,29,88,89 Three studies were prospective comparisons of case series,41,43,91 and one study100 retrospectively selected patients from one of these prospective studies. 91 Two of the studies compared contemporary series of patients from the same hospital,28,29 four used historical controls from the same centre,43,89,91,100 one study used both contemporary and historical controls from the same centre,41 while the other used historical controls from a different hospital. 88

Four studies had a primary outcome measure of toxicity,29,88,91,100 while for two studies the primary outcome was biochemical relapse-free survival,28,89 and for two studies the primary outcome was HRQoL. 41,43 None of the studies reported overall survival.

Five of the studies were set in the USA,28,29,89,91,100 one was set in both Italy and the USA,88 one was set in Japan41 and one was set in the Netherlands. 43 None of the studies were set in the UK, but patients and treatments in the studies were of relevance to UK practice. There was some difference from UK practice in the Ashman et al. 100 study in that some (30%) patients were given chemotherapy, and in the Lips et al. 43 study for which some (26%) of the 3DCRT group were given hyperthermia treatment (see Appendix 2).

For most studies, patients were assessed at 3 to 6 month intervals,28,29,88,91,100 with one study setting patient assessments at 6 to 12 month intervals. 89 The two studies with primary outcome HRQoL set data collection for 12 months41 or 6 months43 after RT.

Two relevant systematic reviews of IMRT for PC were identified: Pearson et al. ;64 and Mast et al. 101,102 In addition two relevant systematic reviews of IMRT, which included PC, were identified via web searches: Van den Steen et al. ;47 and Veldeman et al. 23 These were searched for relevant comparative studies. No meta-analyses were presented in these reviews.

Study details for the included comparative studies are shown in Table 9.

TABLE 9 - Summary of included published studies

SVs, seminal vesicles.

Study	Population	Intervention IMRT		Comparator 3DCRT		Number of patients	Outcomes	Follow-up
		Whole pelvis or prostate only	Approximate dose	Whole pelvis or prostate only	Approximate dose
Kupelian et al.84 2002 USA	Localised PC	Prostate only or prostate and SVs	70 Gy (hypo-fractionated)	Prostate only or prostate and SVs	78 Gy	282	Biochemical relapse-free survival Toxicity	Median 25 months
Sanguineti et al.88 2006 USA/Italy	Localised PC	Whole pelvis	76 Gy	Prostate only	76 Gy	113	Toxicity	Mean 25.9 months
Shu et al.29 2001 USA	Localised PC	Prostate only (28% patients), whole pelvis (72%)	85 Gy	Prostate only (96% patients), whole pelvis (4%)	84 Gy	44	Toxicity	Median 23.1 months
Vora et al.89 2007 USA	Localised PC	Prostate and SVs	75.6 Gy	Prostate and SVs	68.4 Gy	416	Biochemical relapse-free survival Toxicity	Median 60 months
Yoshimura et al.41 2007 Japan	Localised PC	Prostate only or prostate and SVs	76.5 Gy	Prostate only or prostate and SVs	70 or 73.5 Gy	144	HRQoL	12 months
Zelefsky et al.91 2008 USA	Localised PC	Prostate and SVs97	81 Gy	Prostate and SVs96	75.6 Gy or 66–70.2 Gy	1571	Toxicity	Median 96 months
Ashman et al.100 2005 USA	Locally advanced PC	Whole pelvis	81 Gy	whole pelvis	75.6 Gy	27	Toxicity	Median 30 months
Lips et al.43 2007 The Netherlands	Locally advanced PC	Prostate and SVs103	76 Gy	Prostate and SVs104	70 Gy	170	HRQoL	6 months

Within studies, treatment groups differed in more than just IMRT versus 3DCRT. Dose was higher in the IMRT group than the 3DCRT group for the studies of Vora et al. ,89 Yoshimura et al. ,41 Zelefsky et al. ,91 Ashman et al. 100 and Lips et al. 43 Dose was similar between treatment groups for the studies of Kupelian et al. 28 (doses between two groups were reported as equivalent, 70 Gy in 28 fractions vs 78 Gy in 39 fractions), Sanguineti et al. 88 and Shu et al. 29 In the Sanguineti et al. 88 study, treatments differed in that IMRT involved treating the whole pelvis, whereas 3DCRT was restricted to the prostate. There was higher proportion of the IMRT group that had the whole pelvis treated than in the 3DCRT group of the Shu et al. 29 study. In the Lips et al. 43 study there was more accurate positioning of the prostate for IMRT than 3DCRT. Yoshimura et al. 41 and Zelefsky et al. 91 had two comparator groups with different doses of 3DCRT (see Appendix 2). Some studies had longer follow-up for IMRT than 3DCRT (see Appendix 2).

Between studies treatments differed in terms of PTV, dose, dose constraints, fractionation and patient positioning. More details about RT are in Appendix 2.

Included conference abstracts

Five conference abstracts of comparative studies of IMRT versus 3DCRT in PC were identified (Table 10). Three abstracts were from the same single centre study from the USA, Kirichenko et al. ,105 Sharma et al. ,106 and Morgan et al. ,107 which compared a case series of IMRT with a historical case series of 3DCRT. One of these three abstracts described a subset of patients treated with hormonal therapy,106 and one described a subset of patients with intermediate to high risk disease. 107 One abstract,108 also available as a poster,109 was from a single centre in Germany which evaluated contemporary IMRT and 3DCRT patients in a matched pair analysis. One abstract110 from a single centre in the USA studied IMRT versus 3DCRT, with both treatments utilising adaptive image guided RT. None of the abstracts specified whether the population was localised or locally advanced PC.

TABLE 10 - Summary of included conference abstracts

PNs, pelvic nodes; NR, not reported; SVs, seminal vesicles.

Study	Intervention IMRT		Comparator 3DCRT		Number of patients	Follow-up	Outcome(s)
	Whole pelvis or prostate only	Approximate dose	Whole pelvis or prostate only	Approximate dose
Kirichenko et al.105 2006 USA	Prostate only, or prostate and PNs	Prescription dose 74–78 Gy, 95% PTV received 100% dose	Prostate only, or whole pelvis	Median peripheral CTV dose 72 Gy (range 70–79 Gy) prescribed to 95% isodose line	Total 1417; IMRT 489; 3DCRT 928	Median IMRT 29.9 months; 3DCRT 63.3 months	Toxicity
Sharma et al.106 2007 USA	NR (presume same as Kirichenko et al.)105	Mean 76 Gy (74–76)	NR (presume same as Kirichenko et al.)105	Mean 76 Gy (73–80)	Total 293; IMRT 123; 3DCRT 170	Median IMRT 41 months; 3DCRT 62 months	Toxicity
Morgan et al.107 2007 USA	NR (presume same as Kirichenko et al.)105	Median 81 Gy (range 77–85)	NR (presume same as Kirichenko et al.)105	Median 80 Gy (range 76–82)	Total 376; IMRT 188; 3DCRT 188	Median 35 months	Toxicity, biochemical failure, distant metastasis, cause specific mortality
Boehmer et al.108 2006 Germany	NR	79.7 Gy (78–82), simultaneous integrated boost 2 Gy	NR	72.2 Gy (70.2–73.8)	Total 187; IMRT 96; 3DCRT 91	Median 20 months	Toxicity
Martinez et al.110 2007 USA	Prostate only or prostate and SVs (60%)	Adaptive image-guided, median isocentre dose 79.7 Gy	Prostate only or prostate and SVs (51%)	Adaptive image-guided, median isocentre dose 79.7 Gy	Total 728; IMRT 172; 3DCRT 556	Median IMRT 2.2 years; 3DCRT 4.3 years	Toxicity

Quality of included studies

Quality assessment was undertaken for the eight studies which were fully reported using an adapted Downs and Black checklist83 according to the guidance within the checklist on non-randomised studies. A summary of the quality of the eight studies can be seen in Appendix 3. A table is presented in Appendix 3 for relevant quality headings in the Downs and Black checklist (reporting, internal validity, external validity-bias and external validity-confounding). A narrative summary of the key aspects of the Downs and Black checklist follows.

Reporting

All studies clearly reported an aim or hypothesis. Seven out of eight studies clearly described the main outcomes. Seven studies clearly described patient characteristics, but Ashman et al. 100 did not present population characteristics per group. Seven out of eight studies did not have similar study groups on the basis of population characteristics, but one study did have groups with similar population characteristics. 91 One study88 did not have the same data collection methods owing to the treatments being delivered in different countries, whereas the other seven studies appeared to have the same data collection methods per group. All eight studies clearly described the interventions. Four studies did not clearly describe potential confounders, two studies did clearly describe confounders88,91 and two studies partially described confounders. 43,84 Six studies did not have similar treatment groups displaying differences in treatment years and areas to be irradiated, one study had aspects which were similar but the treatment groups were not identical, and one study91 had similar treatment groups. Findings were clearly reported in six studies, but two studies41,100 did not clearly report their findings. Ashman et al. 100 reported incorrect numbers of patients with grade 2 GU events (of 11 patients with GU events, Ashman et al. 100 state that eight were treated with 3DCRT and four patients were treated with IMRT which is clearly incorrect). Related papers by Ashman et al. were not available to clarify this point. Estimates of random variability were present in all but one study. 29 Seven studies fully addressed adverse events but one study did not88 but focused on late rectal toxicity. Loss to follow-up was not described in seven studies and was unable to be determined in one study. 43 All eight studies reported actual probability values.

External validity

It was not possible to determine from the reporting whether or not those subjects who did participate were representative of the entire population from which they were recruited, although this may be presumed. Seven studies had a location/setting which was representative of the treatment that the majority of patients receive as far as this was possible to determine and for one study it was not possible to determine this aspect. 84

Internal validity-bias

The eight studies were not single- or double-blinded trials and all showed clarity regarding data dredging if this occurred. Five studies adjusted for different follow-up lengths, two studies did not,89,91 and it was not possible to determine this item in one study. All eight studies used appropriate statistical testing. It was not possible to determine if compliance with the intervention(s) were reliable for all eight studies. All eight studies described valid and reliable outcome measure(s).

Internal validity-confounding

Seven studies individually used comparison groups selected from the same institution, apart from one study which had comparison groups from two separate institutions. 88 Only two studies29,91 sampled in the same time period for each group, and the other six studies sampled during different treatment periods/years per group. Kupelian et al. (2005)111 suggests that independent of tumour stage, radiation dose, failure definition and follow-up parameters, the year in which RT is performed is an independent predictor of outcomes. Three studies are worst affected as they have historical controls in the previous decade29,88,89 and these studies may overestimate biochemical control in the IMRT group. 23 None of the comparative studies were randomised, so concealment of randomisation was not an issue. Three studies made adequate adjustment for confounders84,88,91 but five studies did not. It was not possible to determine if the studies accounted for losses of patients to follow-up.

Overall most of the information from all the eight studies has a medium- to high-risk of bias; sufficient to affect the interpretation of results (i.e. weakens confidence in the results). Therefore, the results should be interpreted with caution.

Quality of included conference abstracts

Quality assessment was undertaken for all five105–108,110 included conference abstracts using the adapted Downs and Black checklist. 83 A summary of the quality of the five abstracts can be seen later in Appendix 4. A table is presented in Appendix 4 for key quality headings in the Downs and Black checklist (reporting, internal validity, external validity-bias, external validity-confounding) with regard to the five abstracts. However, it was not possible to perform a full quality assessment using the Downs and Black checklist83 as the information within the abstracts was necessarily limited. None of the studies were identified as having later been published in full.

Within the limits of the abstract reporting level, it was not possible to determine the main outcomes or patient characteristics clearly for all five abstracts. However, the limited findings were largely clear; an attempt to cover the key adverse events was made, but losses to follow-up were not described in all five abstracts. Two abstracts105,110 were marginally more clearly reported than the other abstracts. It was not possible to determine the items regarding external validity for all five abstracts. It was largely not possible to determine items regarding internal validity-bias for all five abstracts; with the exception of two studies106,107 which mentioned that adjustment was made for different lengths of follow-up, one abstract108 gave information about a reliable and valid outcome measure and two abstracts105,107 indicated appropriate statistical testing. With regard to internal validity, only one abstract108 sampled within the same time period for both groups, otherwise it was not possible to determine many of the items for the five abstracts.

Assessment of effectiveness

None of the studies reported overall survival or clinically measured disease-free survival. One conference abstract107 reported, at 4 years, no significant difference between treatment groups in distant metastasis (from n = 88 in both groups, those developing distant metastasis IMRT 4%, 3DCRT 3%, p = 0.36), or in cause-specific mortality (IMRT 1%, 3DCRT 0%, p = 0.32). 107

Biochemical relapse-free survival

The Kupelian et al. 84 study used the ASTRO consensus definition44 of biochemical failure, that is three consecutive rising PSA levels after reaching a nadir, and calculated the time to failure as midway between time of nadir and first PSA rise. No significant difference between treatment groups was found (p = 0.084) (Table 11). When the Kupelian et al. study applied a stricter definition of biochemical relapse-free survival, that is reaching and maintaining a follow-up PSA level at less than or equal to 0.5 ng/ml, there remained no significant difference between treatment groups (p = 0.24). By multivariate Cox proportional hazards analysis, taking into account age, race, stage, pre-treatment PSA, biopsy Gleason score and neoadjuvant androgen deprivation, treatment group showed a non-significant trend favouring IMRT (p = 0.058).

TABLE 11 - Biochemical relapse-free survival

Study	Follow-up	Definition of failure	Number of patients IMRT	Number of patients 3DCRT	Survival IMRT%	Survival 3DCRT%	Comparison
Kupelian et al.84	30 months	ASTRO consensus	166	116	94 (95% CI 91 to 98)	88 (95% CI 82 to 94)	p = 0.084
Kupelian et al. 84	30 months	PSA level reaching and maintaining 0.5 ng/mL or over	166	116	87% (95% CI 81 to 93)	80% (95% CI 72 to 88)	p = 0.24
Vora et al.89	36 months	ASTRO consensus	145	271	95.2	71.1
Vora et al.89	36 months	ASTRO Phoenix	145	271	93.8	89.8
Vora et al.89	60 months	ASTRO consensus	145	271	74.1	60.4	p < 0.0001
Vora et al.89	60 months	ASTRO Phoenix	145	271	84.6	74.4	p < 0.0326

The Kupelian et al. study85 reported 5-year biochemical survival data for the IMRT group only (n = 100) at median 5.5 years follow-up. The five-year biochemical relapse-free survival using the ASTRO consensus definition was 85% [95% confidence interval (CI) 78 to 93%], and using their stricter definition was 88% (95% CI 82 to 95%). Survival data by risk groups were 97% for low-risk disease, 88% for intermediate risk, and 70% for high-risk disease. Using an expanded comparison group treated with 3DCRT (n = 310, median follow-up 71 months) 5-year biochemical relapse-free survival using the ASTRO consensus definition was 78%. A later report of this study,86 when 770 IMRT patients had been recruited with a median follow-up of 45 months, reported that the 5-year biochemical relapse-free survival using the ASTRO consensus definition was 82% (95% CI 79 to 85%).

The Vora et al. study89 using the ASTRO consensus definition, found a significant difference between treatment groups (p < 0.0001) by either univariate or multivariate analysis (Table 11). Multivariate Cox proportional hazards analysis took into account stage, perineural invasion, PSA, Gleason score, hormonal therapy and percentage positive biopsies in addition to treatment group, and found all factors except hormonal therapy and percentage positive biopsies were significantly associated with biochemical failure. The Vora et al. study also used the ASTRO Phoenix definition45 rise in PSA level of 2 ng/ml or more above the nadir with no backdating. There was a significant difference between treatment groups by either univariate (p < 0.0326) or multivariate analysis (p < 0.0359).

The Vora et al. study89 investigated biochemical survival by the NCCN defined risk group, that is risk group decided by PSA ≤ 10 ng/ml, stage T1–T2, Gleason score ≤ 6; if patient meets all three criteria they are considered to have low-risk disease, an increase in one indicator translates as intermediate risk, and an increase in two or three indicators is classed as high risk. 91 Using the ASTRO consensus definition, Vora et al. found 5-year survival for IMRT patients was 87.5% for low-risk patients, 72.6% for intermediate risk, and 60.2% for high-risk patients. For 3DCRT the survival data by risk group were 76.2%, 50.1%, and 35.0% respectively. IMRT significantly improved biochemical survival for intermediate- (p < 0.0001) and high-risk (p = 0.0188) patients, but for low-risk patients there was no significant difference between treatment groups (p = 0.181).

The Zelefsky study does not report comparative data, but reports data from IMRT patients using the ASTRO consensus definition. There was no dose effect on PSA relapse-free survival for IMRT dose of 81 Gy or 86.4 Gy. 92 There was no difference in results according to whether or not patients had neoadjuvant androgen deprivation. 92 For IMRT patients aged ≤ 60 years, a dose < 75.6 Gy was the most important predictor of biochemical relapse in younger patients. 94 The 3-year actuarial PSA relapse-free survival rates by NCCN defined risk group: low risk (n = 275) 92%, intermediate (n = 322) 86%, and high (n = 175) 81%. 92 The 8-year actuarial PSA relapse-free survival rates by risk group for IMRT patients were for low risk (n = 203) 85%, intermediate (n = 255) 76%, and high (n = 103) 69%.

The Morgan et al. 107 abstract reported biochemical failure using the ASTRO Phoenix definition in patients with intermediate to high-risk PC. At 4 years there was no significant difference between treatment groups, with a biochemical failure rate of 18% for the IMRT group (PSA survival 82%), and 19% (PSA survival 81%) for the 3DCRT group (p = 0.675).

For the studies Kupelian et al. 28 and Vora et al. 89 that reported comparative statistics, Kupelian et al. 28 did not report a difference between treatment groups, whereas Vora et al. 89 found a significant biochemical survival advantage for IMRT. The different results cannot be explained by definition of biochemical failure, as both studies used the ASTRO consensus definition. Across the two studies at 30 or 36 months, biochemical survival data appear similar for IMRT groups (see Table 8); however, for the 3DCRT groups Vora et al. 89 reported lower survival than Kupelian et al. 28 This might be explained by dose. The 3DCRT dose in the Vora et al. 89 study was lower than in the Kupelian et al. 28 study. The Vora et al. 89 study had a lower dose in 3DCRT than in the IMRT group, and the Kupelian et al. 28 study had an equivalent dose between treatment groups. The Morgan et al. 107 abstract, which did not find a treatment group difference in biochemical failure, had similar dose for both treatment groups. The 3DCRT dose in the Vora et al. 89 study was lower than in other studies, and was approximately 68.4 Gy. There is evidence from RCTs that in CRT 68 Gy leads to lower biochemical survival than for CRT at 78 Gy. 59 The lack of dose effect on PSA relapse-free survival for IMRT dose of 81 Gy or 86.4 Gy in the Zelefsky study92 is consistent with the lack of evidence of benefit for increasing dose of 3DCRT above 81 Gy. 64

It is unlikely that the difference between Kupelian et al. 28 and Vora et al. 89 study results could be explained by risk group. Although the Vora et al. 89 study found that low-risk disease did not benefit significantly from IMRT over 3DCRT, they found intermediate and high-risk patients did differ significantly between treatment groups, and the Kupelian et al. 28 study had 70% high-risk tumours. It is unlikely that the difference between study results could be explained by their difference in hormonal therapy between treatment groups, as neither study found an association between hormonal therapy and biochemical survival, Kupelian et al. 28 (p = 0.66) and Vora et al. 89 (p = 0.08), which agrees with the Zelefsky et al. 91 study. However, the focus of this review was not hormonal therapy, and these limited data are not presented as conclusive evidence on the effect of hormonal therapy.

Toxicity

For defining AEs, five of the studies (Kupelian et al. ,28 Sanguineti et al. ,88 Shu et al. ,29 Vora et al. ,89 Ashman et al. 100) used RTOG and the other study (Zelefsky et al. 91) used NCICTC for AEs.

Among all studies, there are no mentions of treatment-related deaths, with the possible exception of one patient from the Kupelian et al. 28 study. However, although one patient treated with IMRT died in this study it was unclear if the death could be attributed to treatment-related late rectal toxicity or if it was due to the patient’s underlying medical condition.

None of the studies report secondary malignancies, probably because the follow-up durations were not long enough. It may be that follow-up in excess of 10 years, and a larger sample size, would be needed to detect group differences in secondary malignancies. 64 There is a lack of comparative data on sexual dysfunction as an AE, but sexual function is measured as part of HRQoL.

The studies concentrate on GI and GU AEs. These are considered separately for acute and late effects.

Acute gastrointestinal toxicity

Kupelian et al. 84 and Zelefsky et al. 91 found a significant advantage for IMRT over 3DCRT (Table 12). Vora et al. 89 did not find a significant difference between treatment groups.

TABLE 12 - Acute gastrointestinal toxicity in localised prostate cancer

NR, not reported.

Study	Follow-up/definition	Definition	Number of patients IMRT	Number of patients 3DCRT	Acute GI toxicity IMRT	Acute GI toxicity 3DCRT	Comparison
Kupelian et al.84	Acute	RTOG rectal toxicity scores	166	116	Score 0 30%, score 1 55%, and score 2 15%	Score 0 12%, score 1 70%, and score 2 18%	p = 0.002
Shu et al.29	Within 6 months	RTOG grades 1–3	18	26	NR	NR	p = 0.003 higher in IMRT group
Vora et al.89	Acute	RTOG grade	145	271	Grade 0 16%, grade 1 34%, grade 2 49%, grade 3 1%	Grade 0 27%, grade 1 20%, grade 2 54%, grade 3 0%	p = 0.83
Zelefsky et al.91	During or within 3 months	NCI-CTC	472	358	3%	1%	p = 0.04

By univariate analysis, the Shu et al. study found an advantage for 3DCRT. 29 However, this might be attributed to the lower proportion of 3DCRT patients treated with whole pelvis (WP) radiation. As well as treatment group effect, the Shu et al. study found that WP radiation was significantly correlated with the incidence of acute GI toxicity (p = 0.001). 29 In the IMRT group, 9 (69%) of 13 patients receiving pelvic irradiation had acute GI toxicity versus 2 (40%) of 5 patients not receiving pelvic irradiation (both cases were grade 1). The Zelefsky et al. study91 did not find any demonstrable influence of age, radiation dose or hormonal therapy on incidence of acute rectal symptoms. For patients treated with IMRT, Zelefsky found no difference in GI toxicity between doses 86.4 Gy and 81 Gy. 92

Across studies, toxicity grades were generally low. Only one (IMRT) patient in the Shu et al. study developed grade 3 toxicity. 29 In the Vora et al. study no grade 4 or 5 toxicity was found. 89

The only study looking at toxicity specifically in locally advanced PC, Ashman et al. (Table 13), found no grade 3 or higher acute GI toxicities developed in any of the patients; however, the sample size was small. 100 As five of the six patients in the 3DCRT group who had grade 2 diarrhoea were additionally treated with chemotherapy, it is difficult to attribute the results to treatment difference between groups.

TABLE 13 - Acute gastrointestinal toxicity in locally advanced prostate cancer

Study	Follow-up/definition	Definition	Number of patients IMRT	Number of patients 3DCRT	Acute GI toxicity IMRT	Acute GI toxicity 3DCRT
Ashman et al.100	During or within 3 months of treatment	RTOG grade 2 diarrhoea	13	14	0%	43% (n = 6)
Ashman et al.100	During or within 3 months of treatment	RTOG grade 2 proctitis	13	14	7% (n = 1)	36% (n = 5)

Three of the studies106,107,110 published as conference abstracts favoured IMRT for acute GI toxicity. The Sharma et al. abstract106 on hormonally treated PC patients found significantly [hazard ratio (HR) = 4.39, p = 0.003] lower acute GI toxicity in the IMRT group than in the 3DCRT group. The Morgan et al. abstract107 on patients with intermediate- to high-risk PC found significantly less (p = 0.015) grade 2 or higher acute GI toxicity in the IMRT group (4%) than the 3DCRT group (10%). The Martinez et al. abstract110 on adaptive IGRT found a significantly (p < 0.01) lower rate of grade 2 or higher rectal pain or tenesmus in the IMRT group (5%) than the 3DCRT group (19%).

The Kirichenko et al. abstract105 did not report a treatment group difference in acute toxicity, but did find that patients with WP treatment involving lymph nodes were more likely to have acute toxicity than patients with smaller treatment fields (HR = 2.1, p < 0.0001).

The Boehmer et al. abstract108 found no treatment group differences were found for rectal pain, bleeding, urgency or incontinence. However, they did report significantly (p = 0.002) more proctitis (common toxicity criteria) in the IMRT group than in the 3DCRT group. Most toxicities were grade 1. 109 Unlike the Shu et al. study,29 the treatment group difference does not appear to be explained by difference between WP or prostate only radiation. There was a higher dose in the IMRT group, but this was also the case for the Vora et al. 89 and Zelefsky et al. 91 studies which did not show the same pattern of treatment group effect.

Acute GU toxicity

There was no significant treatment group effect for acute GU toxicity in the Kupelian et al. ,85 Shu et al. ,29 or Vora et al. 89 studies, although in the Vora et al. study there was a trend for IMRT to have higher acute GU toxicity (Table 14). The Zelefsky et al. study results significantly favoured 3DCRT (p = 0.001). 91 Zelefsky attributes this to the urethral dose not being constrained in the IMRT group. 92 For patients treated with IMRT, Zelefsky found a non-significant trend for higher toxicity at 86.4 Gy than for 81 Gy. 92

TABLE 14 - Acute GU toxicity in localised prostate cancer

NR, not reported.

Study	Follow-up	Definition	Number of patients IMRT	Number of patients 3DCRT	Acute GU toxicity IMRT	Acute GU toxicity 3DCRT	Comparison
Kupelian et al.84	Acute	Urinary toxicity RTOG scores	166	116	Score 0 15%, score 1 62%, score 2 22%, score 3 1%	Score 0 19%, score 1 63%, score 2 17%, score 3 1%	p = 0.64
Shu et al.29	Within 6 months	RTOG grade 1–2	18	26	NR	NR	p = 0.535
Vora et al.89	Acute	RTOG grade	145	271	Grade 0 28%, grade 1 23%, grade 2 46%, grade 3 5%	Grade 0 38%, grade 1 21%, grade 2 40%, grade 3 1%	p = 0.094
Zelefsky et al.91	During or within 3 months of treatment	NCI-CTC grade 1 or higher GU symptoms	472	358	37%	22%	p = 0.001

The Shu et al. study found that WP radiation correlated with the incidence of acute GU toxicity (p = 0.021). 29

Toxicity was generally low grade; in the Shu et al. study no patients developed grade 3 toxicity,29 and in the Vora et al. study no grade 4 or 5 toxicity was found.

The Ashman et al. study of locally advanced PC had slightly higher incidence of acute GU toxicity in the IMRT group than the 3DCRT group [7 or 8 (reporting unclear in publication) out of 13 patients, vs 3 or 4 out of 14 patients]; however, this was based on a small sample size. 100

Three of the conference abstracts106–108 on PC did not find treatment group effects. The Sharma et al. abstract106 on hormonally treated PC patients found no significant treatment group difference in acute GU toxicity. They did find that hormonal treatment of 6 months or longer duration was associated with increased acute GU toxicity (p = 0.045). The authors gave no explanation for this, and no definitive conclusions about duration are drawn from one small study. The Morgan et al. abstract107 on patients with intermediate to high risk PC did not find a treatment group difference (p = 0.116) grade 2 or higher acute GU toxicity between the IMRT (10%) and 3DCRT (5%) groups. The Boehmer et al. abstract108 did not find any treatment group difference for acute GU toxicity.

The Martinez et al. abstract110 on adaptive IGRT found a significantly (p = 0.03) lower rate of grade 2 or higher urinary retention in the IMRT group (2%) than the 3DCRT group (7%). Other acute GU symptoms are not mentioned. This pattern of results is unlike the other studies, and cannot be explained by dose as treatment groups had equivalent doses. This study did differ from the others in that IGRT was used.

Late GI toxicity

The Zelefsky et al. study91 favoured IMRT in terms of late GI toxicity (Table 15), as did the Kupelian et al. study84 when considering only grade 3 toxicity. The Shu et al. study29 did not find significant treatment effects. The Vora et al. study89 did not find significant treatment effect despite the higher dose in the IMRT group. In the Sanguineti et al. study, the difference between groups was non-significant by univariate analysis (Table 15), but multivariate analysis significantly favoured IMRT, with adjusted HR = 0.1 (95% CI 0.0 to 0.06) p = 0.01. 88 The analysis adjusted for dose to most anterior and posterior rectal points along the central axis, age, race, hypertension, diabetes, vascular comorbidity, GI comorbidity, any other comorbidity, transurethral resection of prostate, androgen deprivation, acute toxicity and T-stage.

TABLE 15 - Late GI toxicity in localised prostate cancer

NR, not reported.

Study	Follow-up	Definition	Number of patients IMRT	Number of patients 3DCRT	Late GI toxicity IMRT	Late GI toxicity 3DCRT	Comparison
Kupelian et al.84	30 months	Actuarial combined grade 2 and 3 late rectal toxicity	166	116	5%	12%	p = 0.24
Kupelian et al. 84	30 months	Grade 3 late rectal toxicity	166	116	2%	8%	p = 0.059
Sanguineti et al. 88	Complications developing > 90 days after treatment and those starting prior to and persisting for > 90 days after completion of treatment, 2 years follow-up	Estimated cumulative incidence grade 2 rectal toxicity	45	68	6% ± 4%	21.2% ± 6%	HR (unadjusted) = 0.2, 95% CI 0.1 to 1.1; p = 0.06
Shu et al. 29	Minimum 10 months	RTOG grade 1–3	18	26	NR	NR	p = 0.163
Vora et al. 89	60 months	RTOG grade	145	271	Grade 0 56%, grade 1 20%, grade 2 23%, grade 3 1%	Grade 0 57%, grade 1 26%, grade 2 14%, grade 3 2%	p = 0.24
Zelefsky et al. 91	10 years	NCI-CTC grade 2 or higher	472	358	5%	13%	p < 0.001

Sanguineti et al. dosimetric comparison found that although IMRT had higher average dose to anterior rectum, the percentage of rectum receiving a given dose was significantly (p < 0.05) lower with IMRT than the 3DCRT group. Sanguineti et al. reported that for 22% of 3DCRT patients, and 0% IMRT patients, treatment failed to meet dose–volume constraints. 88 A multivariate analysis of AEs from the Kupelian et al. 28 study found the only factor that significantly predicted grade 2–3 late rectal toxicity was the volume of rectum receiving the prescription dose (p = 0.006). Zelefsky et al. 91 found a higher incidence of late GI toxicity in those patients who experienced acute GI toxicity (p < 0.0001).

For patients with locally advanced PC, Ashman et al. 100 reported that none of the patients developed grade 3 or higher toxicity. Both patients with grade 2 late rectal bleeding were in the 3DCRT group, but sample size was too small to draw conclusions about significance. The two patients who got grade 2 rectal bleeding (Table 16) did not receive chemotherapy.

TABLE 16 - Late GI toxicity in locally advanced prostate cancer

Study	Follow-up/definition	Definition	Number of patients IMRT	Number of patients 3DCRT	Late GI toxicity IMRT	Late GI toxicity 3DCRT
Ashman et al. 100	More than 3 months after treatment (minimum follow-up 10 months)	RTOG grade 2 rectal bleeding	12	13	0%	15% (n = 2)

Four105–107,110 of the five105–108,110 conference abstracts favoured IMRT in terms of late GI toxicity. The Kirichenko et al. abstract105 found a significantly (univariate p = 0.009, multivariate HR = 0.6, p = 0.0499) lower rate of late grade 2 or higher GI toxicity in the IMRT group (6.2%) at 3-year follow-up compared with the 3DRCT group (10.4%). In addition, treatment to the WP or lymph nodes was associated with higher late GI toxicity (HR, p = 0.003). The Sharma et al. abstract106 on hormonally treated PC patients found significantly lower (HR = 2.45, p = 0.02) late grade 2 or higher GI toxicity in the IMRT group (9%) than in the 3DCRT group (22%), based on 5-year estimates. The Morgan et al. abstract107 on patients with intermediate to high-risk PC found a non-significant trend for less (p = 0.061) grade 2 or higher late GI toxicity in the IMRT group (4%) than the 3DCRT group (9%) at 4 years. The Martinez et al. abstract110 on adaptive IGRT found a significantly (p < 0.01) lower rate of late grade 2 or 3 rectal bleeding in the IMRT group (4%) than the 3DCRT group (16%). There was no treatment group difference for late grade 2 or 3 rectal pain, tenesmus or diarrhoea. Median time to rectal bleeding was 11 months for IMRT and 12 months for 3DCRT.

The Boehmer et al. abstract108 did not find any treatment group difference (p = 0.23) for late GI toxicity. This was despite acute GI toxicity results showing more proctitis in the IMRT group than in the 3DCRT group.

Late GU toxicity

The Zelefsky et al. 91 and Shu et al. 31 studies favoured 3DCRT for late GU toxicity (Table 17). There was no significant treatment effect in the Vora et al. 89 study, and the Kupelian et al. 84 study reported they had too few events to determine significance. Shu et al. found late GU toxicity was related to maximal tumour dose (p = 0.019), and WP radiation (p = 0.016). 29 Both Shu et al. 29 and Zelefsky et al. 91 found a higher incidence of late GU toxicity in those patients who had experienced acute GU toxicity (p = 0.025 and p < 0.001 respectively).

TABLE 17 - Late GU toxicity in localised prostate cancer

Study	Follow-up	Definition	Number of patients IMRT	Number of patients 3DCRT	Late GU toxicity IMRT	Late GU toxicity 3DCRT	Comparison
Kupelian et al.84	30 months	RTOG urinary toxicity	166	116	Grade 2 n = 2, grade 3 n = 0	Grade 2 n = 2, Grade 3 n = 0	NA
Shu29	Minimum 10 months	RTOG grade 1–3	18	26	NR	NR	p = 0.025
Vora89	60 months	RTOG grade	145	271	Grade 0 45%, grade 1 27%, grade 2 23%, grade 3 6%	Grade 0 66%, grade 1 13%, grade 2 17%, grade 3 5%	p = 0.33
Zelefsky91	10 years	NCICTC grade 2 or higher	472	358	20%	12%	p = 0.01

For locally advanced PC, Ashman et al. found none of the 12 IMRT patients, but four of 13 3DCRT patients with late GU toxicities (Table 18). 100 All of the patients with late GU toxicity were treated with chemotherapy as well as 3DCRT, making it difficult to attribute the group difference to type of RT delivered.

TABLE 18 - Late GU toxicity in locally advanced prostate cancer

Study	Follow-up/definition	Definition	Number of patients IMRT	Number of patients 3DCRT	Late GU toxicity IMRT	Late GU toxicity 3DCRT
Ashman100	More than 3 months after treatment (minimum follow-up 10 months)	RTOG grade 1–3	12	13	0%	31% (n = 4)

Three105,107,108 conference abstracts reported no treatment group effect on late GU toxicity. The Kirichenko et al. abstract105 did not find a significant treatment group difference in late GU toxicity, although there was a trend (univariate p = 0.06, multivariate p = 0.11) for more toxicity in the IMRT group (8.4%) than the 3DCRT group (5.7%) at 3 years. The Morgan et al. abstract107 on patients with intermediate to high risk PC did not find a treatment group difference (p = 0.661) grade 2 or higher late GU toxicity between the IMRT (2%) and 3DCRT (1%) groups. The Boehmer et al. abstract108 did not find any treatment group difference (p = 0.42) for late GU toxicity.

The Martinez et al. abstract110 on adaptive IGRT found no treatment group difference for late grade 2 or 3 frequency, urgency, haematuria or urethral stricture. However, they did report a significantly (p < 0.05) lower rate of late grade 2 or 3 urinary retention in the IMRT group (0.5%) than the 3DCRT group (3%). This pattern of results is unlike the other studies, and cannot be explained by dose as treatment groups had equivalent doses.

Health-related quality of life

Yoshimura et al. 41 assessed HRQoL in localised PC at three time points; before RT, immediately after RT, and 12 months after RT ended. On the eight domains in SF-36 there were no significant differences between treatment groups. There were also no significant differences across the three time points. On the UCLA PCI there were no significant differences between groups. Although there was a significant interaction between group and time point on the sexual function domain (p < 0.05), this was seemingly due to difference between groups at baseline. There was a non-significantly higher pre-RT sexual function score in the low-dose 3DCRT group, than in the other two groups. The low-dose 3DCRT group deteriorated between pre-RT and immediately post-RT then improved at 12 months, whereas the IMRT and high-dose 3DCRT groups improved between pre-RT and immediately post-RT as well as improving between immediately post-RT and 12 months. All three groups improved significantly between pre-RT and 12 months, though this was greater in the IMRT and high-dose 3DCRT groups (p < 0.001) than the low-dose 3DCRT group (p < 0.01). For all groups there was significant deterioration (p < 0.05) in urinary bother, bowel function and bowel bother domains during RT, but these were restored at 12 months.

Lips et al. 43 assessed HRQoL in locally advanced PC at three time points; before treatment, 1 month after treatment, and 6 months after treatment. Considering the time points baseline and 1 month post-RT, for six of the 29 items there was a significant interaction between treatment group and time, with the 3DCRT group showing more deterioration than the IMRT group. These items were RAND-36 (RAND-36 item health survey, that uses the same items as SF-36 but with different scoring) social functioning (p = 0.006), pain (p = 0.01) and change in health (p < 0.0001); EORTC QLQ-C30(+3) physical functioning (p = 0.006) and role functioning (p = 0.006); EORTC QLQ-PR25 urinary symptoms/function (p < 0.0001). For baseline and 6 months post-RT, there were no significant differences between groups on any of the items measured. For the other 23 items there were no significant differences between groups and no significant interaction between treatment group and time. These included items relating to emotional, cognitive, social and sexual functioning, and GI symptoms.

Kupelian et al. 84 assessed HRQoL in localised PC, but only after treatment, no baseline data were reported. Two years after treatment 77 patients (IMRT n = 38, 3DCRT n = 39) completed the Expanded Prostate Cancer Index Composite (EPIC) questionnaire. The groups did not differ in urinary (p = 0.85), bowel (p = 0.12) or hormonal (p = 0.38) scores, but the IMRT group scored better on the sexual summary score (p = 0.003). There was no difference between treatment groups on the SF-12, either on physical (p = 0.11) or mental (p = 0.81) QoL scores.

Discussion

No comparative studies of IMRT versus prostatectomy were identified. No RCTs were available of IMRT versus 3DCRT in PC. Eight studies29,41,43,84,88,89,91,100 comparing IMRT and 3DCRT were found that were published in full, and an additional five studies105–108,110 published as conference abstracts only were identified. None of the studies investigated overall survival. Comparative evidence on biochemical relapse-free survival was available from two studies84,89 published in full and one abstract. 107 Toxicity data were available from six studies29,84,88,89,91,100 published in full and five abstracts. 105–108,110 Comparative evidence on HRQoL with baseline and follow-up data were available from two studies41,43 published in full. Of the studies published in full, six were studies of clinically localised PC,29,41,84,88,89,91 and two of locally advanced PC. 43,100 The five conference abstracts105–108,110 did not specify whether patients had localised or locally advanced PC. The only study reporting toxicity data for locally advanced PC had a sample size of only 27, precluding significance testing. 100

With regard to overall quality of the eight studies included in the review,29,41,43,84,88,89,91,100 the studies displayed weaknesses which included differences between the treatment and comparator groups on population differences. Two studies involved populations which had higher biopsy Gleason scores in the IMRT group compared with one study which had higher biopsy Gleason scores in the 3DCRT group. Two studies had higher clinical T stages in the 3DCRT group and one study had higher PSA scores in the IMRT group. One study had higher numbers of patients with androgen deprivation therapy in the IMRT group compared with one study which had higher numbers of patients with androgen deprivation therapy in the 3DCRT group. The direction of bias differs between the studies on these population characteristics making it difficult to assess the impact on the treatment effect of the IMRT group. Some studies adjusted for such differences with multivariate analysis, but not all studies did. Overall the disease severity of the IMRT group patients may be higher than in the 3DCRT group patients, which may indicate that the results are biased against IMRT. 23

Half of the studies used historical controls and many of the studies sampled different treatment years per group. With regard to the latter point, many of the studies were conducted within the last decade. However, the studies with historical controls in the previous decade and thus most affected were Vora et al. ,89 Sanguineti et al. 88 and Shu et al. 29 so their results may be potentially biased. 23 Studies had longer follow-up periods for the 3DCRT groups compared with the IMRT groups, as the 3DCRT groups tended to be the historical controls being sampled first. This may have introduced potential bias towards more toxicity, although in some studies the IMRT follow-up may have been long enough to capture late toxicity.

There were differences in the details of oncological management between studies and within studies including the use of hormonal therapy (which may affect PSA control, overall survival, acute toxicity, late toxicity and HRQoL). One might assume that use of this therapy suggests more advanced disease, but it might easily just represent change in standard practice. It is unlikely that the difference between study results could be explained by differences in hormonal therapy between treatment groups, as two studies did not find an association between hormonal therapy and biochemical survival,84,89 which also agrees with the Zelefsky et al. study. 91

There were differences between studies use of pelvic nodal RT (which may also affect PSA control, overall survival, acute toxicity, late toxicity and HRQoL) and regarding the use of dose-escalated RT (which may also affect PSA control, overall survival, acute toxicity, late toxicity and HRQoL). The use of dose escalation may indicate a change in standard practice, but it might also represent more advanced disease.

There were inter- and intra-study differences with regard to the definition of biochemical failure; however, studies such as Vora et al. 89 presented treatment group data for both the ASTRO consensus and the ASTRO Phoenix definition of biochemical failure, therefore limiting potential bias.

The reliability of results presented in abstract form may be questionable. Abstracts may present preliminary results of an ongoing study and may differ from those eventually published in full. 2,112

For biochemical relapse-free survival, two studies did not report a difference between treatment groups. One study found a significant biochemical survival advantage for IMRT. This difference between studies is probably explained by dose. The study finding an advantage for IMRT had higher dose than in the 3DCRT group, unlike the other studies that had equivalent doses between treatment groups.

Most of the studies reported an advantage for IMRT in acute and late GI toxicity, either in reducing toxicity or producing similar toxicity to lower dose 3DCRT. GI toxicity was associated with larger treatment field, as WP versus prostate only, or as volume of rectum treated. Late GI toxicity was associated with acute GI toxicity.

Most of the studies found no treatment group effect for acute and late GU toxicity, although one study favoured 3DCRT. Adaptive IGRT favoured IMRT for GU toxicity, however this was only from one study abstract,110 and so does not allow firm conclusions to be drawn. Acute GU toxicity was found to be associated with WP radiation or hormonal treatment of 6 months or longer duration, although this may be due to baseline differences in symptoms, with higher risk disease patients being prescribed these treatments. Late GU toxicity was associated with acute GU toxicity.

In localised PC, for both IMRT and 3DCRT there was a decrease in HRQoL during treatment, but this was restored by 12 months after RT. For patients treated with higher dose RT (IMRT or 3DCRT), HRQoL improved to a greater extent than for lower dose RT. In locally advanced PC, there was no treatment group difference by 6 months after treatment, although 1 month after RT the 3DCRT group showed more deterioration in pain, functioning and urinary symptoms. It may be that follow-up was not long enough in these studies to provide meaningful HRQoL data ‘as late adverse effects frequently do not manifest until after two years of treatment completion’. 62

Although studies had methodological flaws, taken together they seem to support the theory that higher dose up to 81 Gy can improve biochemical survival, and that restricting treatment field, particularly with regard to constraining volume of rectal wall receiving prescription dose, can lessen toxicity.

Chapter 4 Assessment of cost-effectiveness

Independent economic assessment

Methods

Model structure

A discrete event simulation model was developed in simul8^® for patients undergoing radical treatment for PC with either IMRT or 3DCRT. There were no data to model other patient groups. All studies that reported clinical effectiveness did so in terms of PSA failure. The PSA failure hazard was not constant with time (i.e. not exponentially distributed), and the literature indicated that the hazards of clinical failure or death from the time of PSA failure is also not necessarily constant with time, which would make a Markov model formulation complex.

In a discrete event simulation individual patients are followed, with the time of events (such as transition to another state) sampled from a relevant statistical distribution. The simulation time-clock moves immediately to the time of the next event. This contrasts with a state transition model that uses fixed time cycles which is less efficient in periods of no state change, and requires numerous states when patient history affects future transition rates. The time to PSA failure was sampled for each patient and compared with the expected time of all-cause death, taken from life tables. If the expected time of death from other causes was less than the time to PSA failure, then the patient died from other causes, without experiencing PSA progression. Similarly time to clinical progression and PC death are sampled and compared to time to death from other causes, to determine whether the patient dies of other causes or progresses through to more advanced PC disease states. Note the distributions from time from PSA failure to clinical failure and time from PSA failure to PC death were sampled independently, so there is a possibility of patients dying of PC without incurring the costs associated with the more advanced disease states, but they do incur terminal care costs. Distributions for the incidence and timing of late adverse events following RT were also sampled. The model has a lifetime perspective.

Ten thousand individual patients were simulated for each model run. This number was chosen as the mean results had clearly stabilised with this number of patients.

The model structure is shown in Figure 3.

FIGURE 3.

Diagram of model structure.

Patients with localised PC enter the model at the time of treatment (IMRT or 3DCRT). At some time later they may experience PSA failure, then clinical failure and death from PC. Due to the distributions for time from PSA failure to clinical failure and PSA failure and PC death some patients may progress directly from PSA failure to PC death. Death from other causes and late AEs of treatment may occur in any state (with the obvious exception of death). A fixed loss in QALYs is attributed to a patient suffering a late AE at the time that the AE occurs. Acute toxicity is not included, as being by definition of short duration, it has a negligible effect on QALYs, especially as the difference in rates of grade 2 or more adverse effects between IMRT and 3D are small (see Chapter 3, Toxicity).

It is assumed that patients with clinical failure are on hormone therapy and hormone-refractory patients have a course of chemotherapy, as well as palliative care.

Model outputs:

The primary model outputs are:

• total costs for each treatment (IMRT/3DCRT)

• total QALYs for each treatment (IMRT/3DCRT).

From the above the ICERs were calculated for each scenario.

Other model outputs include the proportion of patients who progress to each PC disease state and death from other causes. The perspective of the analysis is England and Wales 2008–9.

Model scenarios

The rationale for IMRT is that higher doses of RT may be given, with the intention of improving PSA survival (and hence disease-specific survival), while limiting additional radiation toxicity to organs close to the prostate. Ideally different scenarios would be modelled with varying doses of IMRT and 3DCRT to investigate the relationship between survival, late toxic effects of RT and treatment costs.

The clinical studies included in the review varied as to whether the IMRT and 3DCRT cohorts received similar total doses of radiation, or whether the dose in the IMRT cohort was higher. In some studies 3DCRT was given at relatively high total doses (≥ 80 Gy). 29,107,110 However, there is considerable heterogeneity between studies in patient disease states, adjuvant hormone therapy and RT techniques. Appendix 9 shows for the included studies a summary of PSA survival and toxicity for IMRT and 3DCRT ordered by RT dose. It can be seen that the results are too heterogeneous for dose effects to be apparent. Thus the studies are too heterogenous both for meta-analysis and to attempt to identify variation in effects by dose.

There is evidence from RCTs of dose escalation for 3DCRT that show that PSA survival improves with increasing dose. 59–62 A recent meta-regression shows the relationship to be linear across a range of doses between 64 Gy and 79.2 Gy, with a 1.8% increase in biochemical control at 5 years for each 1 Gy increase in total RT dose. 65 The results of these studies are relevant to IMRT. There is also evidence from 3DCRT studies of variation in toxicity with increasing dose. An RCT of dose escalation of 3DCRT from 64 Gy to 74 Gy showed an increase in late grade 2+ GI toxicity of 9% at 5 years. 62 The results of a meta-regression of cohort studies show that the incidence of late GI toxicity of grade 2 or more increases by 12% to 16% when the RT dose is increased from 70 Gy to 80 Gy. 117 However, the toxicity results of the 3DCRT studies are not applicable to IMRT as IMRT allows the radiation to be sculpted to the target area of the PC more precisely than 3DCRT, thus reducing toxicity to the surrounding normal tissues at similar doses. Thus there is no data on the effects of dose escalation on adverse effects for IMRT.

The effect of dose escalation on late GI toxicity for IMRT might be ascertained indirectly using the results of an analysis of the correlation of rectal toxicity with the volume of the rectum receiving different RT doses. 118 For example, for every 5% increase in rectal volume receiving 30 Gy the odds ratio of grade 2 rectal bleeding was 1.14 (p = 0.006), and for every 5% increase in dose of 70 Gy the similar odds ratio was 1.41 (p = 0.001). To use these results to compare toxic effects between IMRT and 3DCRT or for dose escalation of IMRT would require a review of dosimetric studies, which is beyond the scope of this review.

Given the limitations described it was deemed most appropriate to treat each study as a different scenario. Only three included studies reported PSA survival,84,89,107 of which one is reported in abstract only. 107 Patients in all of these studies had localised PC and were treated with radiation to the prostate and seminal vesicles (SVs) only. In two studies where similar radiation doses were given to IMRT and 3DCRT patients there was no statistically significant difference in PSA survival,84,107 although Kupelian et al. 85 did report an absolute difference of 7% at 5 years. In both studies relatively high doses of RT were given. Note that although the 3DCRT patients in the Kupelian et al. 84,85 study received 78 Gy compared to 70 Gy for IMRT the doses are considered biologically equivalent as the IMRT was given in fractions of 2.5 Gy (hypofractionated) compared to the more usual schedule of 2 Gy fractions for 3DCRT. In comparison to the total doses reported by other studies both arms in Kupelian et al. 84,85 may be considered to have received RT doses equivalent to 78 Gy. See Appendix 10 for details of the estimation of biologically equivalent doses. Morgan et al. reports a median dose of 80 Gy and 81 Gy for 3DCRT and IMRT respectively. 107

The model scenarios are thus as follows:

1. Morgan et al. 107 IMRT and 3D high dose (80/81 Gy) – no PSA survival difference.

2. Kupelian et al. 84,85 IMRT and 3D high dose (78 Gy) – no PSA survival difference.

3. Kupelian et al. 85 IMRT and 3D high dose (78 Gy) – PSA survival difference as reported.

4. Vora et al. 89 IMRT 75.6 Gy, 3DCRT 68 Gy – PSA survival difference.

Note the total RT dose given to 3DCRT patients in the Vora et al. 89 study is low compared to the current NICE guideline recommendation of at least 74 Gy. The scenario will illustrate however how cost-effective IMRT may be in comparison to 3DCRT if higher doses may be given with IMRT than for 3DCRT.

A baseline age of 70 years is used, the approximate age at which PC incidence peaks. A sensitivity analysis was run on each of the four scenarios with age at 60 and 80 years,

Model distributions

Disease progression: overview

None of the studies included in the clinical effectiveness review report time to progression to the clinical outcomes of clinical progression or PC death. They do, however, report time to PSA failure. In order to estimate the effect of IMRT on clinical outcomes it is therefore necessary to estimate the time from PSA failure to clinical disease occurrence and also to PC death. Death from other causes also needs to be considered. Thus the disease progression model requires three distributions:

• [A] time from treatment (IMRT/3DCRT) to PSA failure

• [B] time from PSA failure to clinical failure

• [C] time from PSA failure to PC death.

The method used to estimate these distributions are described in the following paragraphs.

A Effectiveness of IMRT and 3DCRT: time from treatment (IMRT/3DCRT) to PSA failure

Three studies included in the effectiveness review report time to PSA failure, all showing either a survival or time to failure curve. 84,89,107 The published graphs were scanned and imported into techdig software so that points could be read more efficiently and accurately off the curves. Weibull distributions were then fitted to the data points obtained. Particular issues in fitting curves to the data are discussed for each of the studies below. The two-parameter Weibull distribution was used with shape parameter γ, and scale parameter λ.The form used is shown below, and the Weibull parameters used and resulting mean PSA survival are shown in Table 20.

TABLE 20 - Summary of Weibull parameters and estimated mean survival for different scenarios

Study	Group	Scale	Shape	Mean survival to PSA fail (years)
Vora et al. 200789	IMRT	0.0082	1.755	13.7
	3DCRT	0.0064	2.437	7.1
Kupelian et al. 2002/200584,85	IMRT	0.0171	1.370	17.8
	3DCRT	0.0239	1.370	14.0
Morgan et al. 2007107	IMRT/3DCRT	0.0227	1.463	12.0

The survivor function is:

S ( t ) = exp ( − λ t γ )

The mean of a Weibull (γ,λ) distribution is:

( 1 λ ) 1 γ Γ ( 1 + 1 γ )

where Γ(x) is the gamma function.

Morgan et al.107

The Morgan et al. study is reported in abstract only, but a small figure was published showing actuarial time to PSA failure. At 4 years there was only 1% difference in failure rates between the IMRT and 3DCRT cohorts, which was statistically non-significant. The two failure curves cross, and are in effect indistinguishable from each other. Points were read from both lines and the results averaged at each time point to give a single curve for both IMRT and 3DCRT. The resulting survival curve is shown in Figure 4, together with the fitted Weibull curve.

FIGURE 4.

Actual and fitted PSA survival curves: data from Morgan et al. 107 IMRT/3DCRT.

Kupelian et al.84,85

The length of follow-up in the Kupelian et al. 84 study is relatively short: median 21 and 32 months respectively for IMRT and 3DCRT. The number of patients with PSA failure was small in both cohorts, so the PSA survival curves are relatively flat. Curves fitted to this data reflected this, and gave implausibly long mean survival. Some PSA survival curves in other studies, for example Vora et al. 89 (see below) are also relatively flat (low rate of PSA failure) in the first few months, with the rate of failure increasing with time, suggesting that the length of follow-up in the Kupelian et al. 84 study is insufficiently long to base projections beyond the duration of the study. A later publication85 shows more mature data, albeit for a smaller cohort, but only shows a PSA survival curve for the IMRT group. The study reports an absolute difference in PSA survival of 7% at 5 years, but this was not statistically significant. The IMRT PSA survival curve was used for both 3DCRT and IMRT in the scenario which assumes no difference in PSA survival between the two treatment methods (Figure 5).

FIGURE 5.

Actual and fitted PSA survival curves: data from Kupelian et al. 85 IMRT.

A second scenario was modelled to reflect the difference in survival reported in the study, although it was statistically non-significant. For this scenario the Weibull 3DCRT shape parameter was assumed to be the same as for the IMRT curve, with the scale parameter adjusted to reflect the proportional difference in survival (85% IMRT, 78% 3DCRT) reported by Kupelian et al. 85 [i.e. (78/85) × (shape parameter IMRT)].

Vora et al.89

The Weibull curves were fitted from 20 months onwards for both IMRT and 3DCRT curves to give a better fit to the later data, as the plot of ln(–ln(survival)) with ln(time) was not linear in the first few months. The plots of the original data and fitted curves are shown in Figures 6 and 7 for IMRT and 3DCRT respectively.

FIGURE 6.

Actual and fitted PSA survival curves: data from Vora et al. 89 IMRT.

FIGURE 7.

Actual and fitted PSA survival curves: data from Vora et al. 89 3DCRT.

B/C Time from PSA failure to clinical failure/prostate cancer death

In 2005 a Consensus conference, jointly sponsored by ASTRO and RTOG, was held to consider evidence for a revised definition of PSA failure. At the conference several researchers presented data showing the association between varying definitions of PSA failure and clinical outcomes. This evidence is summarised in a paper by Roach et al. 45 As well as potentially relevant references in Roach et al. a MEDLINE search was undertaken on the names of the researchers named in the Roach et al. study. By this means one study by Kestin et al. 119 was identified which showed survival data from PSA failure to clinical failure and PC death, using relevant definitions of PSA failure. Relevant definitions were defined as either the ASTRO consensus or Phoenix definitions, or other definitions from which outcomes could be inferred for these.

Kestin et al. 119 explored the association of various biochemical failure definitions with clinical outcomes, including clinical failure and PC death. The analysis is based on 727 men at a single institution in the USA with clinically localised (T1–T3, N0, M0) PC treated with radical external-beam RT alone (no hormone therapy prior to clinical failure) between 1987 and 1997, and with at least five post-RT PSA measurements. Thus the data is a good match to the population being considered in the model, although the dates of the collection period might lead to slight underestimation of current PC survival as more recent treatments for advanced PC such as docetaxel chemotherapy would be expected to affect this. 120

Summary results relevant to the model are shown in Table 21. It can be seen that the actual results required (rates of clinical failure and PC death with time) are not reported for the ASTRO Phoenix definition of PSA failure (a rise ≥ 2 ng/ml above nadir PSA). However, they are reported for similar definitions: a rise of ≥ 1ng/ml above nadir PSA and a rise of ≥ 3ng/ml above nadir PSA.

TABLE 21 - Rates of clinical failure and PC death after biochemical failure (from date of failure)

Definition	Years	Rates of clinical failure (%) with time (years)			Rates of PC death (%) with time (years)
		0.5	2	5	0.5	2	5
Rise of ≥ 1 ng/ml above nadir PSA		11	31	70	0.3	4	24
Rise of ≥ 3 ng/ml above nadir PSA		23	54	96	1.1	9	38
Estimated ASTRO Phoenix (rise of ≥ 2ng/ml above nadir PSA) – see text		17	43	83	0.7	6.5	31

The rates of clinical failure and PC death at 10 years after RT for all three definitions (a rise ≥ 1, 2 or 3 ng/ml above nadir PSA) are shown in Table 22. It appears that the relationship is linear, with the results for a rise ≥ 2 ng/ml above nadir PSA midway between those for 1 and 3 ng/ml. It therefore is reasonable to assume that the rate of clinical failure or PC death can be estimated as the mid-points for those for 1 and 3 ng/ml. These estimated values are shown in Table 21.

TABLE 22 - Ten-year clinical failure and PC death from time of RT for various definitions of PSA failure

Definition	Clinical failure (%)	PC death (%)
Rise of ≥ 1 ng/ml above nadir PSA	64	40
Rise of ≥ 2 ng/ml above nadir PSA (ASTRO Phoenix)	73	45
Rise of ≥ 3 ng/ml above nadir PSA	82	50

Weibull curves were fitted to these data points. The data and fitted curves for clinical survival and PC death from the time of PSA failure (ASTRO Phoenix) are shown in Figures 8 and 9 respectively.

FIGURE 8.

Clinical survival from the time of PSA failure (ASTRO Phoenix).

FIGURE 9.

PC survival from the time of PSA failure (ASTRO Phoenix).

The fitted Weibull curves result in mean estimated time from PSA failure to clinical failure and PC death of 3.1 and 7.8 years respectively.

It was originally anticipated that clinical outcomes following PSA failure as defined by both ASTRO consensus and Phoenix PSA definitions would be required as Kupelian et al. 84 reported PSA fail with the former definition. In fact, the results of Kupelian et al. 84 proved to be unusable as the data was insufficiently mature, and those of Kupelian et al. 85 were used in their place, which were reported using the ASTRO Phoenix definition. This had the advantage that all scenarios are equivalent in this respect.

Other clinical parameters

These include duration of hormone-refractory disease, time to all-cause death and adverse effects of treatment.

Duration of hormone-refractory disease

It is assumed that the time with hormone-refractory disease is independent of the time from clinical failure to death from PC. The time in this state was taken from a review of the treatment of hormone-refractory metastatic PC with docetaxel. 120 Median survival from the TAX327 trial was 18.9 months. By fitting a Weibull distribution to the data Collins et al. estimated the mean survival time to be 22.4 months or 1.9 years. 120 Thus in the model it is assumed that a patient has hormone-refractory metastatic PC for the final 22 months prior to PC death.

Time from treatment to all-cause death

Mean survival death from all causes for the different age scenarios modelled was taken from UK National Statistics. 121 It was assumed that all patients survived to this time unless they had previously died from PC.

Adverse effects of treatment

The only adverse effect of treatment operationalised in the model is late GI toxicity. Most studies did not find a difference in GU toxicity between IMRT and 3DCRT (see Toxicity). None of the included studies report changes in sexual function following RT, and therefore it is unknown if there is any difference between the treatment modes.

Studies which showed actuarial curves for late GI toxicity indicate that incident cases with time could be approximated by a straight line, starting at 6 months after RT, and plateauing (no further incident cases) by 5 years. The data used in the model are derived from the clinical effectiveness studies, and the values shown in Table 23. Only grade 2 and 3 effects are considered. Grade 1 effects were seldom reported in the clinical effectiveness studies (see Tables 15, 17) as they are of limited clinical significance, and thus unlikely to have important economic consequences either.

TABLE 23 - Incidence of late GI toxicity with time in modelled studies

Study	Start of late GI toxic effects (years)		Finish time at which no further incidence of late GI toxic effects		Proportion of patients affected		Comment
	IMRT	3D	IMRT	3D	IMRT	3D
Vora et al. 200789	0.5	0.5	5	5	0.24	0.16	Not actuarial
Kupelian et al. 2002/200584,85	0.5	0.5	5	5	0.11	0.264	Estimated – see text
Morgan et al. 2007107	0.5	0.5	3.5	3.5	0.04	0.09

Note that Kupelian et al. 85 only reports adverse effects of treatment for IMRT patients. In order to estimate their incidence for 3DCRT at 5 years the ratio of late GI toxicity for IMRT at 5 years85 to that at 2.5 years84 was applied to that for 3DCRT at 2.5 years. 84

Note it has been assumed that the two different AE reporting systems (RTOG, NCI-CTC) are equivalent for GI incidence.

Duration of gastrointestinal toxicity

Only one included clinical effectiveness study91 reports the duration for GU and GI toxicity. For GI it is reported that 91% of cases were resolved, with a median time to resolution of 26 months. An estimate was made of the lower bound of mean duration assuming:

• 91% of patients have a mean duration of 26 months

• the remaining 9% unresolved have a mean duration estimated as the difference in median follow-up time and median time to incidence, giving 79 months (median time to development 17 months and median follow-up 96 months – difference 79 months).

This results in an overall estimate of 2.6 months. As this is likely to be an underestimate, with some patients’ symptoms unresolved at the end of follow-up, the number was rounded up to 36 months (3 years). Evidence from a 3DCRT dose escalation trial also shows a proportion of patients whose late GI toxicity symptoms take several months to resolve. 62 The RT01 trial showed prevalence of late GI toxicity peaking at 18 months after RT, then declining, with prevalence at 5 years approximately 30–40% of the peak value (estimated from curve). Sensitivity analysis of the duration of late toxicity was undertaken, with values of 2.5 and 4 years.

Utility values

For cost-effectiveness analysis the value of health effects are measured in terms of QALYs. QALYs are calculated by weighting life-years with utility values, to reflect patients’ HRQoL. There are different methods of determining utility values. The recommended tool for use in the NICE methods guide is the European Quality of Life-5 Dimensions (EQ-5D),122 a measure based on public preferences.

In order to identify utility values for the patient states in the model a systematic search of the following databases was undertaken: MEDLINE, EMBASE, CDSR, DARE, CCTR, HTA and NHS EED. An example search strategy is shown in Appendix 11.

A total of 101 unique references were identified. Of these 40 were selected on title and abstract for potential inclusion as reporting utility values (ascertained by any method). An iterative method of study selection was planned to identify the best evidence on utility values:

1. values obtained using the EQ-5D

2. values obtained using other public preference-based weights of patient HRQoL scores (e.g. the Health Utilities Index)

3. other studies.

Four studies report EQ-5D utilities for PC. 116,123–125 Of these, two report utilities for states relevant to the model. 116,125 The results of Korfage et al. are of interest only in terms of a global comparison as utility values are reported for all patients who had primary treatment of RT. 123 Pickard et al. was excluded as only a single utility value for PC patients of undefined disease status is reported. 124 As EQ-5D utility values were identified for all model states, utility studies using alternative measurement methods were not considered.

The key characteristics of the three studies reporting relevant EQ-5D PC utilities are shown in Table 24, and their results in Table 25.

TABLE 24 - Studies reporting EQ-5D utility values for prostate cancer

NR, not reported; RP, radical prostatectomy; WW, watchful waiting

Study	Country	Subjects	Mean age (years)	n
Korfage et al. 2005123	The Netherlands	Localised PC RT (or radical prostatectomy), prior to treatment and to 52 months post-treatment	RT patients 68	RT 187
Shimizu et al. 2008116	Japan	Patients receiving RP, RT, brachytherapy, WW, or combination for localised PC, and patients with hormone-refractory PC	NR	323
Sullivan et al. 2007125	Europe, North America and Australia	Symptomatic metastatic cancer	72	280

TABLE 25 - Summary results of EQ-5D studies

Study	Patient group	Utility score	Comments
Korfage et al. 2005123	RT		Decrease with age was probably the result of ageing
	Pre-treatment	0.81
	After 6 months	0.83
	After 12 months	0.82
	After 52 months	0.76
Shimizu et al. 2008116	Sexual function: EPIC score		Apparently some double measurement of urinary symptoms, although both measures (urinary function, LUTS) significant in multivariate analysis of predictors of score, as was bowel and hormone function. Sexual function not significant
	16–85 (best)	0.93
	0 (worst)	0.90
	Urinary function: EPIC score
	100 (best)	0.94
	11–74 (worst)	0.84
	Bowel problem: EPIC score
	100 (best)	0.94
	0–80 (worst)	0.84
	Hormonal function: EPIC score
	100 (best)	0.93
	35–80 (worst)	0.84
	LUTS: IPSS EPIC score
	0–7 (best)	0.93
	20–35 (worst)	0.83
Sullivan et al. 2007125	Symptomatic metastatic hormone-refractory cancer	0.635

It is assumed that following RT for localised PC men not suffering adverse effects of treatment have the same mean utility as similarly aged men. Ara and Brazier show how utility values vary with age in UK adults. 126 For men aged 60, 70 and 80 years their mean utility was 0.850, 0.813 and 0.771 respectively. These values were used for the baseline utility values in the first year of the model for the different age scenarios. These values were also assumed to apply to patients who had PSA failure, but not clinical progression. The baseline utility values for men were assumed to decline with age as the men progress through time in the model according to the Ara and Brazier formula. 126

The utility values reported by Shimizu et al. 116 (see Table 25) are higher in general than those reported by Ara and Brazier,126 even for men suffering adverse effects of RT. They may be from a younger population of PC patients (age not reported), and they are not UK values. In order to estimate the utility of men suffering adverse effects of treatment in a UK population the ratio of utility values for full function and poor function was applied to the unaffected126 utility value. This resulting utility values for men aged 70 years are shown in Table 26.

TABLE 26 - Utility values used in the model, men aged 70 years

State	Utility value	Source
Post-RT, no AE	0.813	Ara and Brazier126
Post-RT, GI toxicity (bowel problem)	0.727	Ara and Brazier,126 Shimizu et al.116 (see text)
Clinical failure (on hormone treatment)	0.734	Ara and Brazier,126 Shimizu et al.116 (see text)
Hormone-refractory cancer	0.641	Ara and Brazier,126 Sullivan et al.125 (see text)

For example, for men who progress to clinical failure, who are assumed to be on hormone treatment, their utility is calculated as the ratio of utility score worst hormone function to utility score best hormone function (from Shimizu et al. 116), applied to the Ara and Brazier126 age-related utility. Thus for a man aged 70 years with clinical failure their utility is (0.84/0.93) × 0.813 = 0.734.

Similarly an adjustment to age-related utility was calculated for hormone-refractory cancer from the utility value reported by Sullivan et al. 125 for men of mean aged 72 years as a ratio of the mean utility for men aged 72 years which is 0.805. 126

Resource use/costs

All unit costs used, together with their source, are shown in Appendix 12. Most costs were sourced from NHS reference costs 2007–8,51 Unit costs of health and social care127 and the British National Formulary. 128 Where costs were not 2008–9 they were inflated by the Hospital and Community Health Sector (HCHS) inflation index. 127

Cost of IMRT/3DCRT

The costs reported in Description of technology under assessment, show an additional cost of £1122 for IMRT compared to 3DCRT. As the objective of the analysis is to determine incremental cost-effectiveness it is sufficient to assume zero cost for 3DCRT and £1122 for IMRT.

Monitoring of patients post-RT (pre-PSA fail)

Initial monitoring of patients for the acute effects of RT is assumed the same for both IMRT and 3DCRT and is included in the costs of treatment shown in Table 8. The NICE guideline1 indicates that patients should have two PSA tests a year for the first 2 years following RT. These normally take place at the oncology centre. It is assumed that this cost is the same for both IMRT and 3DCRT patients as very few patients progress in this time, and is therefore not included in the model.

There is some variation in practice in subsequent monitoring. Some patients may continue to be monitored by an oncology centre up to twice a year. The costs shown below, and used in the model, are based on the recommendation in the NICE guideline that after the first 2 years patients may be monitored by their GP, with one PSA test per year. 1 The costs are shown in Table 27.

TABLE 27 - Annual cost of patient monitoring from 2 years post-RT

Item	Cost 2008–9	Number/year	Annual cost
PSA test	£10.19	1	£10.19
GP attendance	£36.97	1	£36.97
Total	–	–	£47.16

Monitoring of patients post-PSA failure

It is assumed that these patients are monitored as hospital outpatients, with the cost of the PSA tests absorbed into the cost of the outpatient consultations. They have a CT scan and bone scan every 2 years. The costs are shown in Table 28.

TABLE 28 - Annual cost of patient monitoring patients who have experienced PSA failure

Item	Cost 2008–9	Number/year	Annual cost
Oncology outpatient	£88.33	6	£529.96
CT scan (one area)	£112.98	0.5	£56.49
Bone scan	£168.44	0.5	£84.22
Total			£670.67

Monitoring of patients post-clinical failure

It is assumed that these patients are on hormone therapy. While patients may go through a sequence of different hormone treatment strategies it has been assumed for simplicity that these patients are treated with the gonadorelin analogue goserelin, by injection every 3 months. The annual resources and costs for treating patients on hormone therapy are shown in Table 29.

TABLE 29 - Treatment costs for patients on hormone therapy

Item	Cost 2008–9	Number/year	Annual cost
Nurse (GP practice)	£10.27	4	£41.08
Goserelin (Zoladex LA) 10.8-mg syringe	£267.48	4	£1069.92
Oncology outpatient	£88.33	2	£176.65
Dexa scan	£72.92	0.5	£36.46
Total			£1324.12

Hormone refractory/metastatic

The costs of treatment of hormone-refractory metastatic cancer are taken from an analysis of the cost-effectiveness of docetaxel chemotherapy in these patients in comparison to other therapies. 120 Docetaxel with prednisolone is the only chemotherapy regime licensed for use in hormone-refractory PC. 129 While not all men will receive chemotherapy, the majority do, and the costs are likely to be a reasonable representation of the costs of care in the final months of life.

In the economic analysis by Collins et al. ,120 as well as the costs of chemotherapy costs of palliative and terminal care were included, all based on resource use in the TAX327 trial. The total cost of care in 2003–4 (with mean survival 1.9 years) was £15,833, including a cost for terminal care of £3528. These costs, inflated by the HCHS inflation index127 to 2008–9 costs, gives an annual cost of care of £7385 and a terminal care cost of £4007. The latter is implemented in the model at PC death.

Cost of treating late gastrointestinal toxicity

It has been assumed that all patients with grade 2 and 3 toxicities will be monitored in a hospital outpatient setting, with the frequency depending on severity (see Table 30). There is no standard treatment for late GI toxicity. Most are likely to be investigated with flexible sigmoidoscopy, and possibly biopsy. The majority will be treated with low-cost items such as laxatives, the cost of which have not been considered. Some patients with more severe cases may need procedures such as laser treatment.

TABLE 30 - The costs of treating late gastrointestinal toxic effects of RT treatment

Annual monitoring cost	Cost 2008–9	Number/year	Annual cost
Grade 2
Oncology outpatient	£88	3	£265
Grade 3
Oncology outpatient	£88	6	£530
Mean cost (based on 26% grade 3)			£335
Treatment costs per patient	Cost 2008–9	Number/patient	Total cost
Grade 2
Flexible sigmoidoscopy ± colonoscopy, biopsy	£497	1	£497
Laser therapy	£1201	0.25	£300
Enemas (2/day for 2 weeks, community nurse)	£26	28	£728
Grade 3
Flexible sigmoidoscopy ± colonoscopy, biopsy	£497	2	£993
Laser therapy	£1201	2	£2403
Blood transfusion	£462	1	£462
Mean cost (based on 26% grade 3)			£2139

The average monitoring and treatment costs for the treatment of all late GI toxic effects has been calculated by estimating the proportion of patients with grade 3 toxic effects. Table 31 shows the data available from the included studies. The weighted proportion of grade 3 toxic effects of all grade 2 and 3 effects is 26%, as shown in Table 31.

TABLE 31 - Proportion of patients with grade 2 or 3 adverse effects who have grade 3 effects

Study	Treatment group	Number of patients	Proportion grade 3
Kupelian et al. 200284	IMRT	166	0.40
	3DCRT	116	0.67
Vora et al. 200789	IMRT	145	0.04
	3DCRT	271	0.13
Total		698	0.26

All costs and QALYs are discounted at 3.5% per year. 122

Sensitivity analysis

A key parameter in the analysis is the cost difference between IMRT and 3DCRT. The costs were obtained from a single institution only, and therefore subject to some uncertainty as to their generalisability. A one-way sensitivity analysis was undertaken on this key parameter.

The mean duration of late GI toxicity is also an important parameter in some scenarios, where the difference in QALYs between IMRT and 3DCRT depends only on the difference between incidence of late GI toxicity. For these scenarios (one and two) a sensitivity analysis was undertaken, varying the baseline value of 3 years to 2.5 and 4 years.

The effect of the age of men at the time of treatment on the results was tested by running the model with starting ages of 60 and 80 years.

A probabilistic sensitivity analysis was undertaken. The purpose of a probabilistic sensitivity analysis is to show the effect of the uncertainty in individual model parameters on the uncertainty in the model results. In a probabilistic sensitivity analysis, rather than using the expected value of variables, values are sampled from a distribution. The distributions used for each parameter are shown in Appendix 13. For each scenario (e.g. Morgan et al. 107 IMRT) 1000 model runs were made, as this was assumed to adequately reflect the joint uncertainty between parameters within the model.

Results

Baseline scenarios

Table 32 shows the estimated total life-years and proportions of patients who progress to more advanced stages of PC disease for men aged 70 years, for the four modelled scenarios. There is some variation between the scenarios in the proportion of men estimated to die from PC: 42% for Vora et al. 89 (3DCRT) and 21% for Kupelian et al. 84,85 (IMRT). This reflects differences in the study results themselves.

TABLE 32 - Modelled clinical results for men aged 70 years

Scenario number	Scenario	Total life-years	Discounted life-years	PSA failure	Clinical failure	Hormone-refractory metastatic disease	PC death
1	Morgan et al. 2007107 IMRT/3DCRT	11.3	8.9	56%	40%	26%	29%
2	Kupelian et al. 2002/200584,85 IRMT/3DCRT (equal survival scenario)	11.8	9.2	42%	29%	19%	21%
3	Kupelian et al. 2002/200584,85 IRMT	11.8	9.2	42%	29%	19%	21%
	Kupelian et al. 2002/200584,85 3D	11.4	8.9	53%	37%	26%	28%
4	Vora et al. 200789 IMRT	11.8	9.2	49%	34%	22%	24%
	Vora et al. 200789 3D	10.5	8.4	76%	55%	38%	42%

The total discounted costs and QALYs for each scenario for men aged 70 years is shown in Table 33, as well as the resulting ICER. In scenario 1, based on Morgan et al. ,107 IMRT is not cost-effective judged by the maximum NICE threshold of an ICER of £30,000. 122 In this scenario equal doses of RT were given to both IMRT and 3DCRT patients, resulting in the same PSA progression rates for both cohorts. The incidence of late GI toxicity was low in both cohorts compared to other studies (IMRT 4%, 3DCRT 9%), but the difference yields the small 0.01 difference in QALYs (see Table 33) between IMRT and 3DCRT.

TABLE 33 - Total costs, QALYS and ICERs for IMRT and 3DCRT scenarios for men aged 70 years

Scenario number	Scenario	Total discounted costs	Total discounted QALYs	Additional cost IMRT	QALY gain IMRT	Incremental cost/QALY of IMRT
1	Morgan et al. 2007107 IMRT	£6173	6.802	£989	0.010	£104,066
	Morgan et al. 2007107 3DCRT	£5184	6.792
2	Kupelian et al. 2002/200584,85 IMRT	£4946	7.070	£732	0.023	£31,162
	Kupelian et al. 2002/200584,85 3DCRT (survival equal to IMRT)	£4214	7.046
3	Kupelian et al. 2002/200584,85 IMRT	£4946	7.070	£40	0.087	£5295
	Kupelian et al. 2002/200584,85 3DCRT	£4486	6.983
4	Vora et al. 200789 IMRT	£5687	7.015	–£1802	0.613	Dominates
	Vora et al. 200789 3DCRT	£7489	6.402

In scenario 2, based on Kupelian et al. ,84,85 survival is also assumed to be the same for the IMRT and 3DCRT groups, but the estimated difference in late GI toxicity was 15%, giving a total difference in QALYs between the two groups of 0.023. Although IMRT treatment itself is assumed to cost £1122 more than 3DCRT, the additional costs of treating a greater proportion of 3DCRT patients for late GI toxicity reduces the additional total cost of IMRT to £732. The ICER falls at the upper NICE threshold of £30,000. Between £20,000 and £30,000 NICE takes into consideration other factors such as the strength of evidence. 1 In scenario 4 (Vora et al. 89), where a difference between IMRT and 3DCRT in mean survival to PSA failure of 6.6 years has been assumed, IMRT is actually cost-saving, owing to the reduced costs of treating advanced PC. There is also a much greater QALY gain for IMRT compared to 3DCRT, due to the increased survival for IMRT. Thus IMRT yields greater QALYs at lower cost than 3DCRT and is therefore said to dominate 3DCRT.

For scenario 3, where a smaller difference between IMRT and 3DCRT in mean survival to PSA failure of 3.8 years has been assumed, the ICER is small but positive, at £5300.

Sensitivity analyses

Age of men at the time of treatment with RT

Clearly the age of men at the time of RT has an effect on the proportion of men estimated to die of PC as older men are more likely to die of other causes before their PC progresses. Table 34 illustrates the effect for scenario 2 (Kupelian et al. ,84,85 equal survival for IMRT and 3DCRT).

TABLE 34 - Proportion of men who have clinical disease progression and die of PC according to age at time of RT (Kupelian et al.,84,85 equal survival for IMRT and 3DCRT)

Age (years)	Clinical progression	PC death
60	45%	41%
70	29%	21%
80	23%	7%

Table 35 shows the variation in ICER with age. It shows that although there is some variation in the ICER by age, the conclusions are unaffected by the age group modelled. The ICER is generally lower for younger men, as they have more years to benefit.

TABLE 35 - Variation in ICER with age

Scenario number	Scenario	Age at RT (years)
		60	70	80
1	Morgan et al. 2007107	£92,788	£104,066	£133,832
2	Kupelian et al. 2002/200584,85 (equal survival IMRT/3DCRT)	£31,181	£31,162	£38,211
3	Kupelian et al. 2002/200584,85 (greater survival IMRT)	£1762	£5295	£16,914
4	Vora et al. 200789	Dominates	Dominates	Dominates

Relative cost of IMRT compared to 3DCRT

The costs of IMRT and 3DCRT were obtained from a single institution. The relative cost of IMRT compared to 3DCRT may vary according to the equipment used for planning and delivery of RT, and experience of staff with the different techniques. The results of a sensitivity analysis on the additional costs of IMRT compared to 3DCRT are shown in Table 36. The cost difference was varied by plus and minus 20% and 40% of the baseline value at £1122.

TABLE 36 - Sensitivity analysis on the additional costs of IMRT compared to 3DCRT

Scenario number	Scenario	Additional cost IMRT compared to 3DCRT
		Baseline –40%	Baseline –20% (£900)	Baseline (£1120)	Baseline +20% (£1350)	Baseline +40% (£1570)
1	Morgan et al. 2007107	£56,832	£80,449	£104,066	£127,683	£151,301
2	Kupelian et al. 2002/200584,85 (equal survival IMRT/3DCRT)	£12,063	£21,612	£31,162	£40,711	£50,260
3	Kupelian et al. 2002/200584,85 (greater survival IMRT)	£125	£2710	£5295	£5295	£10,464
4	Vora et al. 200789	Dominates	Dominates	Dominates	Dominates	Dominates

Although the ICERs vary quite considerably with the variation in mean cost difference between IMRT and 3DCRT explored in Table 36, in fact the results of the sensitivity analysis show that only scenario 2 is sensitive to the variation. In the first scenario (Morgan et al. 107), where late GI toxicity is low in both cohorts and there is no survival difference the ICER remains above the maximum NICE threshold of £30,000. In scenario 4 IMRT dominates 3DCRT (is less costly and more effective) even if the mean additional cost of IMRT was 40% more (a cost difference of £1570) than the baseline estimate. In scenario 3 the ICER remains well within the £20,000 threshold for the modelled cost differentials. For scenario 2, however, the additional cost of IMRT over 3DCRT is critical to cost-effectiveness. If the baseline additional cost is overestimated the ICER for IMRT in comparison to 3DCRT falls within a threshold of £20,000, but if the additional cost is greater than the baseline (cost difference greater than £1350) it falls beyond the maximum threshold of acceptability to NICE of £30,000.

Duration of late GI toxicity

The duration of late GI toxicity was reported in just one study. 91 Given the variability between studies in other parameters, there is considerable uncertainty around the estimate of 3 years which was used in the model. For scenarios 3 and 4 (Kupelian et al. 84,85 – survival difference, Vora et al. 89) the duration of toxicity is not important, as the far greater difference in QALYs arises principally from the difference in survival. For scenarios 1 and 2 (Morgan et al. ,107 Kupelian et al. 84,85 – equal survival), where no survival difference is assumed between the IMRT and 3DCRT cohorts, the ICERs are sensitive to this parameter, as shown in Table 37.

TABLE 37 - Variation in ICER by mean duration of late GI toxicity

Scenario number	Scenario	Mean duration late GI toxicity
		2.5 years	3 years (baseline)	4 years
1	Morgan et al. 2007107	£131,231	£104,066	£73,107
2	Kupelian et al. 2002/200584,85 (equal survival IMRT/3DCRT)	£42,138	£31,162	£19,842
3	Kupelian et al. 2002/200584,85 (greater survival IMRT)	£5929	£5295	£4244
4	Vora et al. 200789	Dominates	Dominates	Dominates

However, it is only for scenario 2 (equal survival IMRT and 3D, 15% difference in late GI toxicity) that the duration of late GI toxicity is critical to cost-effectiveness.

Probabilistic sensitivity analysis

The mean cost per QALY calculated from a probabilistic sensitivity analysis is preferable to an answer from a deterministic analysis as it explicitly incorporates the uncertainty surrounding each parameter and can take non-linearities and interacting parameters into account. It is of particular use when events which are associated with high costs and high utility losses but occur with low probability exist within the model or skewed distributions such as Weibull are used. 122,130

In fact the results of the probabilistic sensitivity analysis are similar to the results of the deterministic analysis. For scenario 1 (Morgan et al. 107), the expected value of ICER is £104,000 in the deterministic analysis and £120,000 in the stochastic. For this scenario, with no survival difference and little difference in GI toxicity, there is only a 29% probability of IMRT being cost-effective in comparison with 3DCRT with a maximum incremental cost-effectiveness ratio (MAICER) of £30,000 (see Figure 10). Even with a MAICER of £100,000 there is less than 50% probability of IMRT being cost-effective in comparison with 3DCRT.

FIGURE 10.

Cost-effectiveness acceptability curve scenario 1 (Morgan et al. 107).

Given the similarity of the stochastic and deterministic analyses the stochastic analysis was not run for all scenarios. For scenario 2 (Kupelian et al. 84,85 – equal survival IMRT/3DCRT) it is of particular interest as IMRT is borderline cost-effective according to NICE thresholds. The expected value for the ICER of IMRT in comparison to 3DCRT for the probabilistic analysis is £34,781, in comparison to £31,162 from the deterministic analysis. At a MAICER of £20,000 there is only a 20% probability that IMRT is cost-effective, but this rises to 48% for a MAICER of £30,000, as shown in the cost-effectiveness acceptability curve (Figure 11).

FIGURE 11.

Cost-effectiveness acceptability curve scenario 2 (Kupelian et al. 84,85 – equal survival IMRT/3DCRT).

Discussion

A limitation of all economic studies comparing IMRT with 3DCRT, including this one, is the limited clinical data comparing the two treatment modes. There are no RCTs and the comparative studies have weaknesses such as differences between the patient groups. Only three studies (all localised cancer) were identified which reported PSA survival (Morgan et al. ,107 Kupelian et al. ,84 and Vora et al. 89). Of the three studies, one was reported as an abstract only, and this was the only study to report survival (Morgan107).

Based on these three studies four scenarios were modelled. The heterogeneity of the clinical studies is reflected in the results of economic analysis, which, depending on the scenario modelled, range from IMRT dominating 3DCRT (i.e. is both more effective and less costly) to having an ICER of £104,000.

The additional cost of IMRT compared to 3DCRT is relatively modest at £1120, so for scenarios where it is assumed that IMRT can improve freedom from PSA failure and hence survival (likely by dose escalation), IMRT is cost-effective, a result which is robust to variation in other key parameters. The baseline scenarios where equivalent survival is assumed for IMRT and 3DCRT, and QALY differences between the two cohorts are derived from differences in GI toxicity alone, show IMRT to be borderline cost-effective depending on the difference in late GI toxicity, duration of GI toxicity and the cost difference between IMRT and 3DCRT.

At baseline parameter values the scenario with a difference in late GI toxicity of 5% (Morgan et al. 107 – scenario 1) gave an ICER of £104,000, but scenario 2 (Kupelian et al. 84,85 – equal survival) with a difference in GI toxicity of 15% gave an ICER of £31,000. The probabilistic analysis of the latter scenario showed that only with a MAICER of £30,000 or more was it probable that IMRT was cost-effective in comparison to 3DCRT. This is at the upper NICE threshold, where other factors such as the strength of evidence are taken into consideration. The results for this scenario are very sensitive to the incremental cost of IMRT in comparison to 3DCRT, as well as the mean duration of late GI toxicity, both very uncertain parameters. If the incremental cost of IMRT is £860, or around 20% less than the baseline value used, the ICER falls to £20,000. Similarly if the mean duration of late GI toxicity is 4 years rather than 3 years the ICER falls to £20,000. Of course, if these parameters are varied in the opposite direction the ICER of IMRT in comparison with 3DCRT increases, and then clearly falls beyond a threshold of £30,000.

The scenarios modelled were all based on studies where both PSA survival and toxicity were reported. To put the values of incidence of late GI toxicity from the modelled studies in context the results of other studies were considered. Of those where the same RT dose was given to both IMRT and 3DCRT patients with localised cancers (Martinez et al. ,110 Sharma et al. 106), both report a 13% difference in late GI toxicity at 5 years. Sanguineti et al. 88 reports a difference of 15% at 2 years, despite the IMRT group receiving WP RT in comparison to treatment of the prostate only in the comparator group. Both the Martinez et al. 110 and Sharma et al. 106 studies are reported in abstract only, but, with Sanguineti et al. ,88 suggest model scenario 2 (Kupelian et al. 84,85 – equal survival) is more representative than scenario 1 (Morgan et al. ,107 also reported in abstract only). Reduced toxicity from IMRT in comparison to 3DCRT may be greater in patients requiring WP RT, making it more cost-effective than for treatment of the prostate and seminal vesicles alone.

Previous economic analyses of IMRT in comparison with 3DCRT have shown very different results. Pearson et al. 64 assumed no difference in survival, and so in common with some scenarios modelled in this study (1 and 2), relies only on differences in GI toxicity to yield QALY differences between IMRT and 3DCRT treated patients. He reports an ICER of US$706,000. This is considerably greater than the results of our analysis, but the cost differential between IMRT and 3DCRT in the Pearson et al. analysis was estimated to be US$31,500 (approximately £19,000),64 very different from our estimate of £1120.

Konski et al. 131 found IMRT to be cost-effective in comparison to 3DCRT, despite a cost differential almost as great as that of Pearson et al. 64 However, a large difference between IMRT and 3DCRT in PSA survival at 5 years was assumed (16%), derived from PSA survival in two different cohort studies. The greatest PSA survival difference reported in the studies included in this review was 14% (ASTRO consensus definition), but this was reduced to 11% when PSA failure was defined by the ASTRO Phoenix definition. 89 Konski et al. also assumed a long-term higher utility for all IMRT patients compared to those who had had 3DCRT, independent of adverse effects of treatment, the utility values used being derived from two separate studies using different methods of determining utility values. 131

Chapter 5 Assessment of factors relevant to the NHS and other parties

Chapter 6 Discussion

Chapter 7 Conclusions

Acknowledgements

The authors wish to thank Professor David Dearnaley, Professor of Uro-Oncology, Institute of Cancer Research, Royal Marsden Hospital, Surrey; Dr Colin Baker, Lecturer in Radiotherapy Physics, School of Health Sciences,University of Liverpool; Dr Philip Mayles, Head of Physics, Clatterbridge Centre for Oncology, Merseyside; Helen Mayles, Head of Clinical Radiotherapy Physics, Clatterbridge Centre for Oncology, Merseyside; Dr John Staffurth, Clinical Senior Lecturer/Honorary Consultant, Department of Clinical Oncology, Velindre Hospital, Cardiff; and Jason Cashmore, Consultant Physicist, Medical Physics, Queen Elizabeth Hospital, Birmingham.

Ms Gill Rooney (the School of Health and Related Research; ScHARR) organised the retrieval of papers and helped in preparing and formatting the report.

References

1. National Collaborating Centre for Cancer . CG58 NICE Full Guideline, Prostate Cancer: Diagnosis and Treatment 2008.
2. Sutcliffe P, Hummel, S, Simpson E, Young T, Rees A, Wilkinson A, et al. Use of classical and novel biomarkers as prognostic risk factors for localised prostate cancer: a systematic review. Technology Assessment Report commissioned by the NHS R&amp;D HTA programme; 2007.
3. Hsing AW, Chokkalingham AP. Prostate cancer epidemiology. Frontiers in Bioscience 2006;11:1388-413.
4. Burford DC, Kirby M, Austoker J. Prostate cancer risk management programme: information for primary care. PSA testing in asymptomatic men. NHS Cancer Screening Programmes; 2009.
5. Hamilton W, Sharp D. Symptomatic diagnosis of prostate cancer in primary care: a structured review. Br J Gen Pract 2004;54:617-21.
6. Garnick MB. Prostate cancer: screening, diagnosis and management. Ann Intern Med 1993;118:804-18.
7. Royal College of General Practitioners, Prostate cancer: Diagnosis and treatment . EGP Update 2009.
8. Collin SM, Metcalfe C, Donovan J, Lane JA, Davis M, Neal D, et al. Associations of lower urinary tract symptoms with prostate-specific antigen levels, and screen-detected localized and advanced prostate cancer: a case-control study nested within the UK population-based ProtecT (Prostate testing for cancer and Treatment) study. BJU Int 2008;102:1400-6.
9. Johansson JE, Andren O, Andersson SO, Dickman PW, Holmberg L, Magnuson A, et al. Natural history of early localised prostate cancer. JAMA 2004;291:2713-19.
10. Widmark A, Klepp O, Solberg A, Damber JE, Angelsen A, Fransoon P, et al. Scandinavian Prostate Cancer Group Study 7, Swedish Association for Urological Oncology 3, and Endocrine treatment, with or without radiotherapy, in locally advaned prostate cancer (SPcg-7/SFUO-3): an open randomised phase III trial. Lancet 2009;373:301-8.
11. UK Clinical Research Network: Portfolio Database and ProSTART . A Phase III study of active surveillance therapy against radical treatment in patioents diagnosed with favourable risk prostate cancer (START). UKCRN 2009.
12. Mason M, Warde P, Sydes M, Cowan R, James N, Kirkbride P, et al. The National Cancer Institute of Canada Clinical Trials Group PR3/Medical Research Council PR07 Trial Management Group. Defining the need for local therapy in locally advanced prostate cancer: an appraisal of the MRC PR07 Study. Clin Oncol 2005;17:217-18.
13. Melia J. Part 1: The burden of prostate cancer, its natural history, information on the outcome of screening and estimates of ad hoc screening with particular reference to England and Wales. BJU Int 2005;95:4-15.
14. Cancer Research UK . 2008. Prostate Cancer – UK Incidence Statistics n.d. http://info.cancerresearchuk.org/cancerstats/types/prostate/incidence/ (accessed June 2009).
15. Sakr W, Grignon DJ, Haas GP, Heilbrun LK, Pontes JE, Crissman JD. Age and racial distribution of prostatic intraepithilial neoplasia. Eur Urol 1996;30:138-44.
16. Office for National Statistics . Cancer Statistics: Registrations, Registrations of Cancer Diagnosed in 2005, England 2008.
17. Chamberlain J, Melia J, Moss S, Brown J. The diagnosis, management, treatment and costs of prostate cancer in England and Wales. Health Technol Assess 1997;1.
18. Borre M, Nerstrom B, Overgaard J. The dilemma of prostate cancer: a growing human and economic burden irrespective of treatment strategies. Acta Oncol 1997;36:681-7.
19. Cozzarini C, Fiorino C, Di Muzio N, Valdagni R, Salonia A, Alongi F, et al. Hypofractionated adjuvant radiotherapy with helical Tomotherapy after radical prostatectomy: Planning data and toxicity results of a Phase I-II study. Radiother Oncol 2008;88:26-33.
20. Wiegel T, Bottke D, Steiner U, Siegmann A, Golz R, Storkel S, et al. Phase III postoperative adjuvant radiotherapy after radical prostatectomy compared with radical prostatectomy alone in pT3 prostate cancer with postoperative undetectable prostate-specific antigen: ARO 96–02/AUO AP 09/95. J Clin Oncol 2009;27:2924-30.
21. Thompson LM, Tangen CM, Paradelo J, Lucia MS, Miller G, Troyer D, et al. Adjuvant radiotherapy for pathological T3N0M0 prostate cancer significantly reduces risk of metastases and improves survival: long-term follow up of a randomized clinical trial. J Urol. 2009;181:956-62.
22. Berney DM, Fisher G, Kattan MW, Oliver TD, Moller H, Fearn P, et al. Transatlantic Porstate Group. Major shifts in the treatment and prognosis of prostate cancer due to changes in pathological diagnosis and grading. BJU Int 2007;100:1240-4.
23. Veldeman L, Madani I, Hulstaert, F, De Meerleer G, Mareel M, De Neve W. Evidence behind use of intensity-modulated radiotherapy: a systematic review of comparative clinical studies. Lancet Oncol 2008;9:367-75.
24. Brierley J. The evolving TNM cancer staging system: an essential component of cancer care. CMAJ 2006;172:155-6.
25. American Joint Committee on Cancer . AJCC Cancer Staging Manual 2009.
26. Schroder FH, Hermanek P, Denis L, Fair WR, Gospodarowicz MK, Pavone-Macaluso M. The TNM classification of prostate cancer. Prostate Suppl 1992;4:129-38.
27. Hummel S, Paisley S, Morgan A, Currie E, Brewer N. Clinical and cost-effectiveness of new and emerging technologies for early localised prostate cancer: a systematic review. Health Technol Assess 2003;7.
28. Kupelian PA, Reddy CA, Klein EA, Willoughby TR. Short-course intensity-modulated radiotherapy (70 GY at 2.5 GY per fraction) for localized prostate cancer: Preliminary results on late toxicity and quality of life. Int J Radiat Oncol Biol Phys 2001;51:988-93.
29. Shu HK, Lee TT, Vigneauly E, Xia P, Pickett B, Phillips TL, et al. Toxicity following high-dose three-dimensional conformal and intensity-modulated radiation therapy for clinically localized prostate cancer. Urology 2001;57:102-7.
30. Selley S, Donovan J, Faulkner A, Coast J, Gillatt D. Diagnosis, management and screening of early localised prostate cancer. Health Technol Assess 1997;1.
31. Jewett HJ. The present status of radical prostetectomy for stages A and B prostatic cancer. Urol Clin North Am 1975;2:105-24.
32. Fitzpatrick JM, Shulman C, Zlotta AR, Schroder FH. Prostate cancer: a serious disease suitable for prevention. BJU Int 2009;103:864-70.
33. Lamb DS, Slaney D, Nacey JN, Russell G, Scott B, Johnson CA, et al. Prostate cancer: the new evidence base for diagnosis and treatment. Pathol 2007;39:537-44.
34. Rosario DJ, Lane JA, Metcalfe C, Catto JWF, Dedman D, Donovan JL, et al. Contribution of a single repeat PSA test to prostate cancer risk assessment: experience from the ProtecT study. Eur Urol 2008;53:777-84.
35. National Comprehensive Cancer Network . NCCN Clinical Practice Guidelines in Oncology: Prostate Cancer 2009 n.d. www.nccn.org/professionals/physician_gls/f_guidelines.asp (accessed May 2009).
36. Radiation Therapy Oncology Group . A Short History of RTOG 2009. www.rtog.org (accessed May 2009).
37. National Cancer Institute CTCAE v4.0 MedDRA v12.0 . National Cancer Institute, Cancer Therapy Evaluation Program 2009.
38. Kumar S, Shelley M, Harrison C, Coles B, Wilt TJ, Mason MD. Neoadjuvant and adjuvant hormone therapy for localised and locally advanced prostate cancer. Cochrane Database Syst Rev 2006. 10.1012/14651858.CD006019.pub2.
39. Shelley MD, Kumar S, Coles B, Wilt T, Staffurth J, Mason MD. Adjuvant hormone therapy for localised and locally advanced prostate carcinoma: A systematic review and meta-analysis of randomised trials. Cancer Treat Rev 2009;35:540-6.
40. Shelley MD, Kumar S, Wilt T, Staffurth J, Coles B, Mason MD. A systematic review and meta-analysis of randomised trials of neoadjuvant hormone therapy for localised and locally advanced prostate carcinoma. Cancer Treat Rev 2009;35:9-17.
41. Yoshimura K, Kamoto T, Nakamura E, Segawa T, Kamba T, Takahashi T, et al. Health-related quality-of-life after external beam radiation therapy for localized prostate cancer: intensity-modulated radiation therapy versus conformal radiation therapy. Prostate Cancer Prostatic Dis 2006;10:288-92.
42. Bellizzi KM, Latini DM, Cowan JE, DuChane J, Carroll PR. Fear of recurrence, symptom burden, and health-related quality of life in men with prostate cancer. Urology 2008;72:1269-73.
43. Lips I, Dehnad H, Kruger AB, van Moorselaar J, van der Heide U, Battermann J, et al. Health-related quality of life in patients with locally advanced prostate cancer after 76 GY intensity-modulated radiotherapy vs. 70 GY conformal radiotherapy in a prospective and longitudinal study. Int J Radiat Oncol Biol Phys 2007;69:656-61.
44. American Society for Therapeutic Radiology and Oncology Consensus Panel . Consensus statement: guidelines for PSA following radiation therapy. Int J Radiat Oncol Biol Phys 1997;37:1035-41.
45. Roach M, Hanks G, Thames H, Schellhammer P, Shipley WU, Sokol GH, et al. Defining biochemical failure following radiotherapy with or without hormonal therapy in men with clinically localized prostate cancer: recommendations of the RTOG-ASTRO Phoenix Consensus Conference. Int J Radiat Oncol Biol Phys 2006;65:965-74.
46. Kupelian PA, Buchsbaum JC, Patel C, Elshaikh M, Reddy CA, Zippe CD, et al. Impact of biochemical failure on overall survival after radiation therapy for localized prostate cancer in the PSA era. Int J Radiat Oncol Biol Phys 2002;52:704-11.
47. Van den Steen D, Hulstaert F, Camberlin C. Intensity-modulated radiotherapy (IMRT). Health Technology Assessment (HTA). Brussels: Belgian Health Care Knowledge Centre (KCE); 2007.
48. Lips IR, Dehnad H, Van Gils CH, Boeken Kruger AE, van der Heide UA, Van Vuplen M. High-dose intensity-modulated radiotherapy for prostate cancer using daily fiducial; marker-based position verification: acute and late toxicity in 331 patients. Radiat Oncol 2008;3.
49. Chen MJ, Weltman E, Hanriot RM, Luz FP, Cecilio PJ, da Cruz JC, et al. Intensity modulated radiotherapy for localized prostate cancer: rigid compliance to dose-volume constraints as a warranty of acceptable toxicity?. Radiat Oncol 2007;2.
50. National Institute for Health and Clinical Excellence . Prostate Cancer: Diagnosis and Treatment 2008.
51. Department of Health . NHS Reference Costs 2007–8 2009.
52. Cross T, McPhail S. Prostate cancer: diagnosis and treatment: an assessment of need. London: National Institute for Health and Clinical Excellence; 2008.
53. The Royal College of Radiologists . Development and Implementation of Conformal Radiotherapy in the United Kingdom 2002.
54. Advice to the Naitonal Radiotherapy Advisory Group (NRAG) . National Radiotherapy Advisory Group Radiotherapy Technical Development Sub-Group 2007.
55. International Commission on Radiation Units and Measurements . ICRU Report 50: Prescribing, Recording and Reporting Photon Beam Therapy 1993.
56. International Commission on Radiation Units and Measurements . ICRU Report 62: Prescribing, Recording and Reporting Photon Beam Therapy 1999.
57. Brahme A. Optimization of stationary and moving beam raiation therapy techniques. Radiother Oncol 1988;12:129-40.
58. Webb S. Optimisation of conformal radiotherapy dose distributions by simulated annealing. Phys Med Biol 1989;34:1349-70.
59. Peeters ST, Heemsbergen WD, Koper PC, van Putten WL, Slot A, Dielwart MF, et al. Dose–response in radiotherapy for localized prostate cancer: Results of the Dutch multicenter randomized phase III trial comparing 68 Gy of radiotherapy with 78 Gy. J Clin Oncol 2006;24:1990-6.
60. Pollack A, Zagars GK, Starkschall G, Antolak JA, Lee JJ, Huang E, et al. Prostate cancer radiation dose response: results of the M.D. Anderson phase III randomized trial. Int J Radiat Oncol Biol Phys 2002;53:1097-105.
61. Zietman ALD, DeSilvio ML, Slater JD, Rossi CJ, Miller DW, Adams JA, et al. Comparison of conventional-dose vs high-dose conformal radiation therapy in clinically localized adenocarcinoma of the prostate: A randomized controlled trial. JAMA 2005;294:1233-9.
62. Dearnaley DP, Sydes MR, Graham JD, Aird EG, Cowan RA, Huddart RA, et al. Escalated-dose versus standard-dose conformal radiotherapy in prostate cancer: first results from the MRC RT01 randomised controlled trial. Lancet Oncol 2007;8:475-87.
63. Kuban DA, Tucker SL, Dong L, Starkschall G, Huang EH, Cheung MR, et al. Long-Term Results of the MD. Anderson Randomized Dose-Escalation Trial for Prostate Cancer. Int J Radiat Oncol Biol Phys 2008;70:67-74.
64. Pearson SD, Ladapo J, Prosser L. Intensity Modulated Radiation Therapy (IMRT) for Localized Prostate Cancer 2007. www.icer-review.org/index.php/imrt.html (accessed July 2009).
65. Viani GA, Stefano EJ, Afonso SL. Higher-than-conventional radiation doses in localized prostate cancer treatment: a meta-analysis of randomized, controlled trials. Int J Radiat Oncol Biol Phys 2009;74:1405-18.
66. Vargas C, Martinez A, Kestin LL, Yan D, Grills I, Brabbins DS, et al. Dose-volume analysis of predictors for chronic rectal toxicity after treatment of prostate cancer with adaptive image-guided radiotherapy. Int J Radiat Oncol Biol Phys 2005;62:1297-308.
67. Koper PC, Heemsbergen WD, Hoogeman MS, Jansen PP, Hart GA, Wijnmaalen AJ, et al. Impact of volume and location of irradiated rectum wall on rectal blood loss after radiotherapy of prostate cancer. Int J Radiat Oncol Biol Phys 2004;58:1072-82.
68. Bui C, Glassburn J. IMRT for prostate cancer: Improving the therapeutic ratio. Comm Oncol 2006;3:659-61.
69. Donovan E, Bleakley N, Denholm E, Evans P, Gothard L, Hanson J, et al. Randomised trial of standard 2D radiotherapy (RT) versus intensity modulated radiotherapy (IMRT) in patients prescribed breast radiotherapy. Radiother Oncol 2007;82:254-64.
70. Pignol J, Olivotto I, Rakovitch E, Gardner S, Ackerman I, Sixel K, et al. Phase III randomized study of intensity modulated radiation therapy versus standard wedging technique for adjuvant breast radiotherapy. Int J Radiat Oncol Biol Phys 2006;66.
71. Pow EH, Kwong DL, McMillan AS, Wong MC, Sham JS, Leung LH, et al. Xerostomia and quality of life after intensity-modulated radiotherapy vs. conventional radiotherapy for early-stage nasopharyngeal carcinoma: Initial report on a randomized controlled clinical trial. Int J Radiat Oncol Biol Phys 2006;66:981-91.
72. Kam MK, Leung SF, Zee B, Choi PHK, Chau RMC, Cheung KY, et al. Impact of intensity-modulated radiotherapy (IMRT) on salivary gland function in early-stage nasopharyngeal carcinoma (NPC) patients: A prospective randomized study. J Clin Oncol 2005;23.
73. Nutting C. First results of a phase III multicenter randomized controlled trial of intensity modulated (IMRT) versus conventional radiotherapy (RT) in head and neck cancer. J Clin Oncol 2009;27.
74. Cancer Research UK . External Radiotherapy for Prostate Cancer 2007. www.cancerhelp.org.uk/help/default.asp?page=2881 (accessed June 2009).
75. Baker CR, Hardy V. Provision of IMRT in the UK. Part 2. Current levels, planned expansion and obstacles to implementation. J Radiother Pract 2003;3:181-4.
76. Jefferies S, Taylor A, Reznek R. Results of a national survey of radiotherapy planning and delivery in the UK in 2007. Clin Oncol 2009;21:204-17.
77. Marchal C, Lapeyre M, Beckendorf V, Aletti P, Hasle E, Dubois JB, et al. Preliminary results of the assessment of intensity modulated radiotherapy (IMRT) for prostatic and head and neck tumors (STIC 2001). Cancer Radiother 2004;8:121-7.
78. Konski A, Watkins-Bruner D, Feigenberg S, Hanlon A, Kulkarni S, Beck J, et al. Using decision analysis to determine the cost-effectiveness of intensity-modulated radiation therapy in the treatment of intermediate risk prostate cancer. Int J Radiat Oncol Biol Phys 2006;66:408-15.
79. Van de Werf E, Lievens Y, Verstraete J, Pauwels K, Van den Bogaert W. Time and motion study of radiotherapy delivery: Economic burden of increased quality assurance and IMRT. Radiother Oncol 2009;93:137-40.
80. Miles EA, Clark CH, Urbano MT, Bidmead M, Dearnaley DP, Harrington KJ, et al. The impact of introducing intensity modulated radiotherapy into routine clinical practice. Radiother Oncol 2005;77:241-6.
81. Bonastre J, Noel E, Chevalier J, Gerard JP, Lefkopoulos D, Bourhis J, et al. Implications of learning effects for hospital costs of new health technologies: the case of intensity modulated radiation therapy. Int J Technol Assess Health Care 2007;23:248-54.
82. Van der Werf E, Lievens Y, Verstraete J, Pauwels K, Van den Bogaert W. Time and motion study of radiotherapy delivery: Economic burden of increased quality assurance and IMRT. Radiother Oncol 2009;93:137-40.
83. Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health 1998;52:377-84.
84. Kupelian PA, Reddy CA, Carlson TP, Altsman KA, Willoughby TR. Preliminary observations on biochemical relapse-free survival rates after short-course intensity-modulated radiotherapy (70 Gy at 2.5 Gy/fraction) for localized prostate cancer. Int J Radiat Oncol Biol Phys 2002;53:904-12.
85. Kupelian PA, Thakka VV, Khuntia D, Reddy CA, Klein EA, Mahadevan A. Hypofractionated intensity-modulated radiotherapy (70 GY at 2.5 GY per fraction) for localized prostate cancer: Long-term outcomes. Int J Radiat Oncol Biol Phys 2005;63:1463-8.
86. Kupelian PA, Willoughby TR, Reddy CA, Klein EA, Mahadevan A. Hypofractionated intensity-modulated radiotherapy (70 Gy at 2.5 Gy per fraction) for localized prostate cancer: Cleveland Clinic experience. Int J Radiat Oncol Biol Phys 2007;68:1424-30.
87. Reddy C, Mahadevan A, Klein E, Willoughby T, Kupelian P. Hypofractionated intensity modulated radiotherapy (70 Gy at 2.5 Gy per fraction) for localized prostate cancer. Int J Radiat Oncol Biol Phys 2006;66:1053-8.
88. Sanguineti G, Cavey ML, Endres EJ, Franzone P, Barra S, Parker BC, et al. Does treatment of the pelvic nodes with IMRT increase late rectal toxicity over conformal prostate only radiotherapy to 76 Gy?. Strahlenther Onkol 2006;182:543-9.
89. Vora SA, Wong WW, Schild SE, Ezzell GA, Halyard MY. Analysis of biochemical control and prognostic factors in patients treated with either low-dose three-dimensional conformal radiation therapy or high-dose intensity-modulated radiotherapy for localized prostate cancer. Int J Radiat Oncol Biol Phys 2007;68:1053-8.
90. Vora S, Wong W, Schild S, Ezzell G, Halyard M. Analysis of biochemical control and prognostic factors in patients treated with either low dose 3-dimensional conformal radiation therapy or high dose IMRT for localized prostate cancer. Int J Radiat Oncol Biol Phys 2006;66.
91. Zelefsky MJ, Levin EJ, Hunt M, Yamada Y, Shippy AM, Jackson A, et al. Incidence of late rectal and urinary toxicities after three-dimensional conformal radiotherapy and intensity-modulated radiotherapy for localized prostate cancer. Int J Radiat Oncol Biol Phys 2008;70:1124-9.
92. Zelefsky MJ. High-dose intensity modulated radiation therapy for prostate cancer: Early toxicity and biochemical outcome in 772 patients. Int J Radiat Oncol Biol Phys 2002;53:1111-16.
93. Zelefsky MJ, Fuks Z, Happersett L, Lee HJ, Ling CC, Burman CM, et al. Clinical experience with intensity modulated radiation therapy (IMRT) in prostate cancer. Radiother Oncol 2000;55:241-9.
94. Zelefsky MJ, Marion C, Fuks Z, Leibel SA. Improved biochemical disease-free survival of men younger than 60 years with prostate cancer treated with high dose conformal external beam radiotherapy. J Urol 2003;170:1828-32.
95. Zelefsky MJ, Chan H, Hunt M, Yamada Y, Shippy AM, Amols H. Long-term outcome of high dose intensity modulated radiation therapy for patients with clinically localized prostate cancer. J Urol 2006;176:1415-19.
96. Zelefsky MJ, Leibel SA, Gaudin PB, Kutcher GJ, Fleshner NE, Venkatramen ES, et al. Dose escalation with three-dimensional conformal radiation therapy affects the outcome in prostate cancer. Int J Radiat Oncol Biol Phys 1998;41:491-500.
97. Burman C, Chui CS, Kutcher G, Leibel S, Zelefsky M, LoSasso T, et al. Planning, delivery, and quality assurance of intensity-modulated radiotherapy using dynamic multileaf collimator: a strategy for large-scale implementation for the treatment of carcinoma of the prostate. Int J Radiat Oncol Biol Phys 1997;39:863-73.
98. Zelefsky MJ, Yamada Y, Fuks Z, Zhang ZG, Hunt M, Cahlon O, et al. Long-term results of conformal radiotherapy for prostate cancer: Impact of dose escalation on biochemical tumor control and distant metastases-free survival outcomes. Int J Radiat Oncol Biol Phys 2008;71:1028-33.
99. Cahlon O, Zelefsky MJ, Shippy A, Chan H, Fuks Z, Yamada Y, et al. Ultra-high dose (86.4 Gy) IMRT for localized prostate cancer: Toxicity and biochemical outcomes. Int J Radiat Oncol Biol Phys 2008;71:330-7.
100. Ashman JB, Zelefsky MJ, Hunt MS, Leibel SA, Fuks Z. Whole pelvic radiotherapy for prostate cancer using 3D conformal and intensity-modulated radiotherapy. Int J Radiat Oncol Biol Phys 2005;63:765-71.
101. Mast M, Wiggenraad R, Struikmans H. IMRT, what have we learned?. ESTRO 25 Leipzig Duitsland 2006.
102. Mast M, Wiggenraad R, Struikmans H. IMRT, What Have We Learned? A Systematic Review in Prostate Cancer 2009. www.oncozone.net/uploads/pdf/385adbbbcf279060019714f3aa5ad2f7.pdf (accessed July 2009).
103. Dehnad H, Nederveen AJ, van der Heide UA, van Moorselaar RJ, Hofman P, Lagendijk JJ. Clinical feasibility study for the use of implanted gold seeds in the prostate as reliable positioning markers during megavoltage irradiation. Radiother Oncol 2003;67:295-302.
104. Van Vulpen M, De Leeuw JR, Van Gellekom MP, Van Der Hoeven J, De Graeff A, Van Moorselaar RJ, et al. A prospective quality of life study in patients with locally advanced prostate cancer, treated with radiotherapy with or without regional or interstitial hyperthermia. Int J Hyperthermia 2003;19:402-13.
105. Kirichenko AV, Ruth K, Horwitz E, Buyyounouski M, Feigenberg S, Chen D, et al. Intensity modulated radiation therapy for prostate cancer: Preliminary results on treatment morbidity compared to 3-D conformal radiation therapy. Int J Radiat Oncol Biol Phys 2006;66.
106. Sharma N, Li T, Horwitz E, Pollack A, Buyyounouski M. Intensity modulated radiation therapy reduces gastrointestinal morbidity in patients treated with androgen deprivation therapy for prostate cancer. Int J Radiat Oncol Biol Phys 2007;69.
107. Morgan P, Ruth K, Horwitz E, Buyyounouski M, Uzzo R, . A matched pair comparison of intensity modulated radiotherapy and three-dimensional conformal radiotherapy for prostate cancer: Toxicity and outcomes. Int J Radiat Oncol Biol Phys 2007;69.
108. Boehmer D, Herschel R, Badakhshi H, Eichwurzel I, Wernecke K, Budach V. Matched pair analysis of patients with prostate cancer treated with IMRT or 3-D conformal radiotherapy. Int J Radiat Oncol Biol Phys 2006;66.
109. Boehmer D. Matched Pair Analysis of Patients With Prostate Cancer Treated With Imrt or 3-D Conformal Radiotherapy 2007. www.oncozone.net/en/poster/show/id/167 (accessed June 2009).
110. Martinez A, Yan D, Brabbins D, Kestin L, Mitchell C, Wallace M, et al. Acute and chronic toxicity of adaptive image guided radiation therapy (A-IGRT) delivered with 3D-conformal technique (3D-CRT) vs. intensity modulation (IMRT) in 728 prostate cancer patients. Int J Radiat Oncol Biol Phys 2007;69.
111. Kupelian P, Thames H, Levy L, Horwitz E, Martinez A, Michalski J, et al. Year of treatment as an independent predictor of relapse-free survival in patients with localized prostate cancer treated with definitive radiotherapy in the PSA era. Int J Radiat Oncol Biol Phys 2005;63:795-9.
112. Shrerer RW, Langenberg P, von Elm E. Full publication results published in abstracts. Cochrane Database Methodol Rev 2005;2.
113. Konski A, Watkins-Bruner D, Feigenberg S, Hanlon A, Kulkarni S, Beck J, et al. Intensity modulated radiation therapy (IMRT) is a cost-effective treatment for intermediate risk prostate cancer. Int J Radiat Oncol Biol Phys 2004;60.
114. Konski A. Cost-effectiveness of intensity-modulated radiation therapy. Expert Rev Pharmacoecon Outcomes Res 2005;5:137-40.
115. Remonnay R, Pommier P, Lapeyre M, Hay M, Maingon P, Bensadoun R, et al. Cost comparison between standard 3D conformal radiotherapy (CRT) and intensity modulated radiotherapy (IMRT) for prostate cancer: Results of a multicenter prospective medico-economic study in 8 anticancer centers. Int J Radiat Oncol Biol Phys 2005;63:S188-9.
116. Shimizu F, Fujino K, Ito YM, Fukuda T, Kawachi Y, Minowada S, et al. Fujino Factors associated with variation in utility scores among patients with prostate cancer. Value Health 2008;11:1190-3.
117. van Tol-Geerdink JJ, Stalmeier PF, Pasker-de Jong PC, Huizenga H, van Lin EN, Schimmel EC, et al. Systematic review of the effect of radiation dose on tumor control and morbidity in the treatment of prostate cancer by 3D-CRT. Int J Radiat Oncol Biol Phys 2006;64:534-43.
118. Gulliford SL, Foo K, Morgan RC, Aird EG, Bidmead AM, Critchley H, et al. Dose-volume constraints to reduce rectal side effects from prostate radiotherapy: Evidence from MRC RT01 Trial ISRCTN 47772397. Int J Radiat Oncol Biol Phys 2009;76:747-54.
119. Kestin LL, Vicini FA, Martinez AA. Practical application of biochemical failure definitions: what to do and when to do it. Int J Radiat Oncol Biol Phys 2002;53:304-15.
120. Collins R, Fenwick E, Trowman R, Perard R, Norman G, Light K, et al. A systematic review and economic model of the clinical effectiveness and cost-effectiveness of docetaxel in combination with prednisone or prednisolone for the treatment of hormone-refractory metastatic prostate cancer. Health Technol Assess 2007;11.
121. Office for National Statistics . Interim Life Tables. Expectation of Life United Kingdom Males 2009. www.gad.gov.uk/Demography%20Data/Life%20Tables/Interim_life_tables.html (accessed June 2009).
122. NICE . Guide to the Methods of Technology Appraisal 2008. www.nice.org.uk/media/B52/A7/TAMethodsGuideUpdatedJune2008.pdf (accessed May 2009).
123. Korfage IJ, Essink-Bot ML, Borsboom GJ, Madalinska JB, Kirkels WJ, Habbema JD, et al. Five-year follow-up of health-related quality of life after primary treatment of localized prostate cancer. Int J Cancer 2005;116:291-6.
124. Pickard AS, Lin HW, Knight SL, Sharifi R, Wu Z, Hung SY, et al. Proxy assessment of health-related quality of life in african american and white respondents with prostate cancer: perspective matters. Med Care 2009;47:176-83.
125. Sullivan PW, Mulani PM, Fishman M, Sleep D. Quality of life findings from a multicenter, multinational, observational study of patients with metastatic hormone-refractory prostate cancer. Qual Life Res 2007;16:571-5.
126. Ara R, Brazier J. Health related quality of life by age, gender and history of cardiovascular disease: results from the Health Survey for England. The University of Sheffield; 2009.
127. Curtis L. Unit costs of health and social care. Canterbury: PSSRU; 2008.
128. British Medical Association and Royal Pharmaceutical Society of Great Britain . British National Formulary 2007.
129. NICE . Docetaxel for the Treatment of Hormone-Refractory Metastatic Prostate Cancer 2006:1-35.
130. Claxton K, Sculpher M, McCabe C, Briggs A, Akehurst RL, Buxton M. Probabilistic sensitivity analysis for NICE technology assessment: not an optional extra. Health Econ 2005;14:339-47.
131. Konski A, Watkins-Bruner D, Feigenberg S, Hanlon A, Kulkarni S, Beck JR, et al. Using decision analysis to determine the cost-effectiveness of intensity-modulated radiation therapy in the treatment of intermediate risk prostate cancer. Int J Radiat Oncol Biol Phys 2006;66:408-15.
132. National Radiotherapy Advisory Groug (NRAG) . Advice to NRAG Radiotherapy Provision in England. A Report from the National Radiotherapy Advisory Group – Workforce Subgroup 2006.
133. Egger M. How important are comprehensive literature searches and the assessment of trial quality in systematic reviews? Empirical study. Health Technol Assess 2003;7.
134. Moher D, Pham B, Klassen TP, Schulz KF, Berlin JA, Jadad AR, et al. What contributions do languages other than English make on the results of meta-analyses?. J Clin Epidemiol 2000;53:964-72.
135. Pham, B, Klassen TP, Lawson ML, Moher D. Language of publication restrictions in systematic reviews gave different results depending on whether the intervention was conventional or complementary. J Clin Epidemiol 2005;58:769-76.
136. Collette L, Burzykowski T, Schroder FH. Prostate-specific antigen (PSA) alone is not an appropriate surrogate market of long-term therapeutic benefit in prostate cancer trials. Eur J Cancer 2006;42:1344-50.
137. Current Controlled Trials . A Randomised Phase III Multi-Centre Trial of Conventional or Hypofractionated High Dose Intensity Modulated Radiotherapy for Prostate Cancer 2010. www.controlled-trials.com/mrct/trial/487683 (accessed May 2009).
138. Adkison JB, Patel RR, Bentzen SM, Hong TS, McHaffie DR, Ritter MA. Pelvic nodal dose escalation in conjunction with hypofractionated IMRT for high risk prostate cancer. Int J Radiat Oncol Biol Phys 2008;72.
139. Allison RR, Murphy G, Gay H, Arastu H, Frelix G, Sibata C, et al. Daily CT on rails IGRT for post-prostatectomy rising PSA: outcomes and morbidity. Int J Radiat Oncol Biol Phys 2008;72.
140. Amer AM, Mott J, Mackay RI, Williams PC, Livsey J, Logue JP, et al. Prediction of the benefits from dose-escalated hypofractionated intensity-modulated radiotherapy for prostate cancer. Int J Radiat Oncol Biol Phys 2003;56:199-207.
141. Bernard JR, Buskirk SJ, Serago C, Ko SJ, Kim S, Casale HE, et al. Evaluation of genitourinary side effects following image guided intensity modulated radiation therapy to the prostate. Int J Radiat Oncol Biol Phys 2008;72.
142. Bossi A. Modern External-Beam Radiation Therapy for Prostate Cancer: How and When?. Eur Urol 2008;7:S22-8.
143. Bhatnagar A, Heron D, Beriwal S, Flickinger J, Smith R, Deutsch M, et al. Inital outcomes results for large multi-center integrated cancer network for the treatment of prostate cancer with IMRT. Int J Radiat Oncol Biol Phys 2006;66.
144. Buyyounouski M, Li T, Watkins-Bruner D, Horwitz E, Pollack A. A comparison of sexual function following radiotherapy for prostate cancer: Results of a prospective, randomized trial of hypofractionated intensity modulated radiotherapy. Int J Radiat Oncol Biol Phys 2007;69:S7-8.
145. Catton C, Chung P, Haycocks T, Warde P, Alasti H, Bayley A, et al. Hypofractionated intensity modulated radiation therapy for prostate cancer. Int J Radiat Oncol Biol Phys 2002;54.
146. Chan HM, Zelefsky MJ, Fuks Z, Hunt M, Ling C, Leibel SA. Long-term outcome of high dose intensity modulated radiotherapy for clinically localized prostate cancer. Int J Radiat Oncol Biol Phys 2004;60:S169-70.
147. Chao S, Reddy C, Thakkar VV, Kupelian P, Djemil T, Mahadevan A. Impact of high-dose radiation on erectile function in patients treated with intensity-modulated radiation therapy for prostate cancer. Int J Radiat Oncol Biol Phys 2004;60:S270-1.
148. Cheung P, Loblaw D, Morton G, Zhang L, Basran P, Tirona R, et al. Acute toxicity in high risk prostate cancer patients after elective nodal irradiation using IMRT: Comparison to a 4 field technique. Int J Radiat Oncol Biol Phys 2008;72:S291-2.
149. Cohen K, Teh B, Mai W, Paulino A, Miles B, Kadmon D, et al. Erectile function after nerve-sparing prostatectomy and intensity modulated radiotherapy for prostate cancer patients. Int J Radiat Oncol Biol Phys 2006;66.
150. Daly T, Hickey BE, Lehman M, Francis DP. Adjuvant radiotherapy following radical prostatectomy for prostate cancer. Cochrane Database Syst Rev 2008;4.
151. Dearnaley DP, Khoo VS, Norman AR, Meyer L, Nahum A, Yarnold J. Comparison of radiation side-effects of conformal and conventional radiotherapy in prostate cancer; a randomised trial. Lancet 1999;353:267-72.
152. De Meerleer G, Fonteyne V, Meersschout S, Van den Broecke C, Villeirs G, Lumen N, et al. Salvage intensity-modulated radiotherapy for rising PSA after radical prostatectomy. Radiother Oncol 2008;89:205-13.
153. Djemil T, Reddy C, Willoughby T, Kupelian P. Hypofractionated intensity-modulated radiotherapy (70 Gy at 2.5 Gy per fraction) for localized prostate cancer. Int J Radiat Oncol Biol Phys 2003;57:S275-6.
154. Dogan N, Wu Y, Hagan M. Simultaneous-integrated boost (SIB) IMRT for treatment of intermediate-risk prostate cancer with nodal irradiation. Int J Radiat Oncol Biol Phys 2006;66.
155. Dogan N, Siebers JV, Keall P. Clinical comparison of head and neck and prostate IMRT plans using absorbed dose to medium and absorbed dose to water. Phys Med Biol 2006;51:4967-80.
156. Dong L, O’Daniel J, Smith L, Kuban D, Bounds S, Rosen I, et al. Comparison of 3D conformal and intensity-modulated radiation therapy for early-stage prostate cancer. Int J Radiat Oncol Biol Phys 2001;51.
157. Debus J, Zierhut D, Didinger B, Schlegel W, Wannenmacher M. Inverse planning and intensity-modulated radiotherapy in patients with prostate cancer. Basel: S. Karger Publishers Inc.; 2002.
158. Engler MJ, Tsai JS, Vora S, Wazer DE, DiPetrillo T. Application of dynamic intensity-modulated radiation therapy to the treatment of prostate cancer. Int J Radiat Oncol Biol Phys 1997;39.
159. Fonteyne V, Vakaet L, Villeirs G, Oosterlinck W, Denoyette L, De Neve W, et al. Intensity-modulated radiotherapy (IMRT) for prostate cancer at Ghent university hospital (GUH): First report on late toxicity. Eur Urol Suppl 2006;5.
160. Guckenberger M, Goebel J, Polat B, Vordermark D, Flentje M. Acute toxicity after intensity-modulated, image guided radiation therapy of the prostate-carcinoma. Strahlenther Onkol 2008;184:44-5.
161. Guerrero Urbano MT, Nutting CM. Clinical use of intensity-modulated radiotherapy: part II. Br J Radiol 2004;77:177-82.
162. Hsi R, Okura T, Garver E, Parsai H, Cho P, Jacobs D, et al. Acute toxicity results using dynamic intensity modulated radiotherapy (dIMRT) in high-risk prostate cancer. Int J Radiat Oncol Biol Phys 2007;69:S365-6.
163. Johnstone PAS, Williams SR, Riffenburgh RH. The 100-day PSA: usefulness as surrogate end point for biochemical disease-free survival after definitive radiotherapy of prostate cancer. Prostate Cancer Prostatic Dis 2004;7:263-7.
164. NIHR ProtecT . The ProtecT Trial – Evaluation the Effectiveness of Treatment for Clinically Localised Prostate Cancer 2009. www.epi.bris.ac.uk/protect/ (accessed June 2009).
165. Lee R, Chen J, Sause W. Image guided radiation therapy for prostate cancer using implanted fiducial markers with daily portal imaging. Int J Radiat Oncol Biol Phys 2007;69:S373-4.
166. Liauw S, Jani A, Fricano J, Rash C, Pelizzari C, Weichselbaum R. Salvage RT for biochemical failure post-prostatectomy: Do we reduce or increase toxicity of treatment with IMRT?. Int J Radiat Oncol Biol Phys 2007;69:S337-8.
167. Livi L, Paiar F, Banci-Buonamici F, Simontacchi G, Detti B, Gacci M, et al. Localized prostate cancer treated with intensity-modulated radiotherapy. Tumori 2006;92:197-201.
168. Liu YM, Shiau CY, Lee ML, Huang PI, Hsieh CM, Chen PH, et al. The role and strategy of IMRT in radiotherapy of pelvic tumors: dose escalation and critical organ sparing in prostate cancer. Int J Radiat Oncol Biol Phys 2007;67:1113-23.
169. Khoo VSD. Question of Dose, Fractionation and Technique: Ingredients for Testing Hypofractionation in Prostate Cancer – the CHHiP Trial. Clin Oncol 2008;20:12-4.
170. Kosakowski J, Welzel G, bo-Madyan Y, Mai S, Wertz H, Wenz F, et al. Effectiveness and toxicity of the intensity modulated radiotherapy (IMRT) as hypofractionated boost with prostate carcinoma. Strahlenther Onkol 2008;184.
171. Kry S, Vassiliev O, Kuban D, Poenisch F, Salehpour M, Gillin M, et al. Prostate IMRT using unflattened beams. Int J Radiat Oncol Biol Phys 2006;66.
172. Kuczer D, Haidenberger A, Pakisch B. Is the compliance of the ICRU-62-guidelines possible in the high dose therapy of the prostate carcinoma? A comparison of 3D conformal techniques with IMRT. Strahlenther Onkol 2008;184:101-2.
173. Lim TS, Cheung PCF, Loblaw DA, Morton G, Sixel KE, Pang G, et al. Hypofractionated accelerated radiotherapy using concomitant intensity-modulated radiotherapy boost technique for localized high-risk prostate cancer: acute toxicity result. Int J Radiat Oncol Biol Phys 2008;72:85-92.
174. Ma C, Chi Z, Xu Q, Chen L, Paskalev K, Buyyounoski M, et al. Image guided adaptive radiotherapy of prostate cancer: a retrospective study. Int J Radiat Oncol Biol Phys 2008;72.
175. Mahadevan A, Manzoul S, Reddy C. IMRT to pelvic lymph nodes in prostate cancer. Int J Radiat Oncol Biol Phys 2006;66:S270-5.
176. Marchand V, Bourdin S, Charbonnel C, Rio E, Munos C, Campion L, et al. Health-related quality of life after 76 Gy intensity modulated radiotherapy for localized prostate cancer: A prospective and longitudinal study. Int J Radiat Oncol Biol Phys 2008;72.
177. McCammon R, Kavanagh B, Newell S, Raben D. Acute toxicity assessment of a single institution series of prostate cancer patients treated with pelvic intensity modulated radiation therapy with synchronous integrated boost to the prostate or prostate bed. Int J Radiat Oncol Biol Phys 2006;66.
178. McCloskey S, Wollman R, Rose C, Ketting C, Green G, Ellerbroek N, et al. Quality of life (QOL) outcomes after three-dimensional conformal radiation therapy (3D-CRT) and intensity modulated radiation therapy (IMRT) for localized prostate cancer. Int J Radiat Oncol Biol Phys 2005;63.
179. Melcon I, Martin C, Soria P, Marques E, Oviedo A, Martin M. Intensity modulated radiation therapy with ultrasound-based daily target localization for clinically localized prostate cancer: institutional experience and acute toxicity outcomes. EJC Suppl 2007;5.
180. Michalski JM. Radiation therapy in the management of locally advanced prostate cancer. Curr Urol Rep 2006;7:217-24.
181. Molla M, Linero D, Toscas JI, Escude L, Miralbell R. Hypofractionated stereotactic intensity modulated radiotherapy (IMRT) for prostate cancer with low-metastasic potential. EJC Suppl 2007;5:S477-8.
182. Namiki S, Ishidoya S, Tochigi T, Kawamura S, Kuwahara M, Terai A, et al. Health-related quality of life after intensity modulated radiation therapy for localized prostate cancer: Comparison with conventional and conformal radiotherapy. Jpn J Clin Oncol 2006;36:224-30.
183. Perez CA, Michalski JM, Purdy JA, Lockett MA. New trends in prostatic cancer research. Three-dimensional conformal radiation therapy (3-D CRT), brachytherapy, and new therapeutic modalities. Rays 2000;25:331-43.
184. Pervez N, Pearcey R, Parliament M, Yee D, Amanie J, Murtha A, et al. Acute toxicity results using dynamic intensity modulated radiotherapy (dIMRT) in high-risk prostate cancer. Int J Radiat Oncol Biol Phys 2007;69.
185. Pollack A, Hanlon A, Horwitz EM, Feigenberg S, Uzzo RG, Price RA. Radiation therapy dose escalation for prostate cancer: a rationale for IMRT. World J Urol 2003;21:200-8.
186. National Cancer Institute . Comparison of Two Radiation Therapy Regimens in Treating Patients With Stage II or III Prostate Cancer 2009. http://clinicaltrials.gov/ct2/show/NCT00062309 (accessed June 2009).
187. Rembielak A. Intensity Modulated Radiation Therapy (IMRT) for prostate cancer patients. Urol Polska; 2006.
188. Reuther A, Reddy C, Mahadevan A, Kupelian P. Concomitant irradiation of the entire prostate gland to 70 Gy and the peripheral zone to 80 Gy with intensity modulated radiotherapy in 28 fractions: Preliminary results of a phase II protocol. Int J Radiat Oncol Biol Phys 2006;66:S446-7.
189. Ruben JD, Davis S, Evans C, Jones P, Gagliardi F, Haynes M, et al. The effect of intensity-modulated radiotherapy on radiation-induced second malignancies. Int J Radiat Oncol Biol Phys 2008;70:1530-6.
190. Medical Research Council PR07 . A Randomised Trial of Hormone Therapy Plus Radical Radiotherapy Versus Hormone Therapy Alone in Non-Metastatic Prostate Cancer 2009. http://public.ukcrn.org.uk/search/StudyDetail.aspx?StudyID=651 (accessed June 2009).
191. Sanguineti G, Culp L, Pena J, Bayouth J. Acute small bowel and colon toxicity after pelvic IMRT for prostate cancer. EJC Suppl 2003;1.
192. Sanguineti G, John B, Franzone P, Culp L, Sosa M, Cavey M, et al. IMRT to treat the pelvic nodes while escalating the dose to the prostate gland: Acute toxicity data in 45 consecutive patients. Int J Radiat Oncol Biol Phys 2004;60:S466-7.
193. Scarbrough T, Ting J, Feja L, Golden N, Levitt C. Seed marker-based IGRT for prostate cancer: Excellent preliminary toxicity outcomes. Int J Radiat Oncol Biol Phys 2007;69.
194. Singh A, Guion P, Sears-Crouse N, Ullman K, Smith S, Albert P, et al. Whole prostate to 75.6 with concurrent IMRT boost to 95 gray for biopsy proven, MRI defined dominant intraprostatic lesions: Early results of a phase INCI study. Int J Radiat Oncol Biol Phys 2007;69:S369-70.
195. Shahar K, McGary J, Mai W, Caillouet J, Butler E, Teh B. Retrospective analysis of late rectal bleeding in prostate cancer patients treated with intensity modulated radiotherapy (IMRT). Int J Radiat Oncol Biol Phys 2004;60.
196. Soete G, Verellen D, Storme G. Image guided radiotherapy for prostate cancer. Bull Cancer 2008;95:374-80.
197. Su A, Gratzle J, Correa D, Jani A. Comparison of chronic GI and GU toxicity of patients with prostate cancer undergoing IMRT vs conventional radiotherapy using localized fields. Int J Radiat Oncol Biol Phys 2006;66.
198. Teh B, Mai W, Huang E, Carpenter L, Lu H, Chiu J, et al. Late gastrointestinal (GI) and genitourinary (GU) toxicity following intensity modulated radiation therapy (IMRT) for prostate cancer. Int J Radiat Oncol Biol Phys 2001;51:310-11.
199. Teh B, Bastasch MD, Wheeler TM, Mai WY, Frolov A, Uhl BM, et al. IMRT for prostate cancer: defining target volume based on correlated pathologic volume of disease. Int J Radiat Oncol Biol Phys 2003;56:184-91.
200. Teh B, Mai W, McGary J, Smiedala M, Paulino A, Grant W, et al. Long-term outcome of a moderate-hypofractionated intensity modulated radiation therapy (IMRT) (77GY at 2.2GY per fraction) approach utilizing immobilization rectal balloon for localized prostate cancer. Int J Radiat Oncol Biol Phys 2007;69.
201. Teslow T, Ahmad S, Uy N, Thomson W, Grant W, Butler E, et al. Comparison of tumor control probability and normal tissue complication probability between 3D-CRT and IMRT plans in patients with prostate cancer. Eur J Cancer 2001;37:S204-5.
202. Thakkar VV, Khuntia D, Reddy C, Mahadevan E, Klein A, Chehade N, et al. Hypofractionated intensity modulated radiotherapy (70 Gy at 2.5 Gy per fraction) for localized prostate cancer: Long-term outcome results. Int J Radiat Oncol Biol Phys 2005;63:S124-5.
203. Yuen J, Rodrigues G, Trenka K, Coad T, Yartsev S, D’Souza D, et al. Comparing two strategies of dynamic intensity modulated radiation therapy (dIMRT) with 3-dimensional conformal radiation therapy (3DCRT) in the hypofractionated treatment of high-risk prostate cancer. Radiat Oncol 2008;3.
204. Vaarkamp J. Partial boosting of prostate tumours: Forward planned conformal radiotherapy vs. inverse planned intensity modulated radiotherapy?. Radiother Oncol 2002;63.
205. Villa SL, Fantini N, Bedini C, Sangalli R, Valdagni E, Pignoli S, et al. Acute and Bladder Rectal Toxicities of Hypofractionated Intensity-Modulated Radiotherapy (SIB-IMRT) in Prostate Cancer Radical Treatment With Intraprostatic Gold Fiducial Markers. American Society of Clinical Oncology 2008. www.asco.org/ASCOv2/Meetings/Abstracts?%26vmview=abst_detail_view%26confID=54%26abstractID=20224 (accessed March 2009).
206. Wahab S, Michalski J, Pierburg B, Siegel C, Kibel A, Low D, et al. Pelvic intensity modulated radiation therapy for high risk prostate cancer. Int J Radiat Oncol Biol Phys 2005;63:S331-2.
207. Wang L, Movsas B, Jacob R, Fourkal E, Chen L, Price R, et al. Stereotactic IMRT for prostate cancer: dosimetric impact of multileaf collimator leaf width in the treatment of prostate cancer with IMRT. J Appl Clin Med Phys 2004;5:29-41.
208. Wang-Chesebro A, Xia P, Coleman J, Akazawa C, Roach M. Intensity-modulated radiotherapy improves lymph node coverage and dose to critical structures compared with three-dimensional conformal radiation therapy in clinically localized prostate cancer. Int J Radiat Oncol Biol Phys 2006;66:654-62.
209. Wong GW, Palazzi-Churas KL, Jarrard DF, Paolone DR, Graf AK, Hedican SP, et al. Salvage hypofractionated radiotherapy for biochemically recurrent prostate cancer after radical prostatectomy. Int J Radiat Oncol Biol Phys 2008;70:449-55.
210. Wu Q, Arthur D, Benedict S, Tong S, Wu Y, Hagan M. Intensity-modulated radiotherapy for prostate cancer treatment with nodal coverage. Int J Radiat Oncol Biol Phys 2002;54.
211. Zelefsky MJ, Fuks Z, Leibel SA. Intensity-modulated radiation therapy for prostate cancer. Semin Radiat Oncol 2002;12:229-37.

Appendix 1 Literature search strategies

Appendix 2 Data abstraction tables

Appendix 3 Quality assessment of the eight studies fully reported in peer-reviewed publications83

Appendix 4 Quality assessment for the five included abstracts83

Appendix 5 Table of key excluded studies with rationale

The name of the first author, year and reason for exclusion are reported below. The table includes only studies that were examined at full-text sift and were potentially studies of interest, but were not deemed relevant or valid on closer inspection.

Appendix 6 Resource use and cost assumptions for IMRT and 3DCRT

A personal communication with Nuala Close (St Bartholomew’s Hospital, London, UK, 2009) provided resource use and cost assumptions for IMRT and 3DCRT.

Appendix 7 Critical appraisal checklist for the economic evaluations using key components of the British Medical Journal checklist for economic evaluations together with the Eddy checklist on mathematical models employed in technology assessments

Appendix 8 Example search for cost and cost-effectiveness evidence (MEDLINE)

1. 38. Economics/ (25834)

2. 39. “costs and cost analysis”/ (37460)

3. 40. Cost allocation/ (1864)

4. 41. Cost-benefit analysis/ (44373)

5. 42. Cost control/ (18019)

6. 43. Cost savings/ (6144)

7. 44. Cost of illness/ (10938)

8. 45. Cost sharing/ (1430)

9. 46. “deductibles and coinsurance”/ (1204)

10. 47. Medical savings accounts/ (396)

11. 48. Health care costs/ (17205)

12. 49. Direct service costs/ (860)

13. 50. Drug costs/ (8865)

14. 51. Employer health costs/ (995)

15. 52. Hospital costs/ (5709)

16. 53. Health expenditures/ (10360)

17. 54. Capital expenditures/ (1839)

18. 55. Value of life/ (5057)

19. 56. exp economics, hospital/ (15764)

20. 57. exp economics, medical/ (12120)

21. 58. Economics, nursing/ (3849)

22. 59. Economics, pharmaceutical/ (1958)

23. 60. exp “fees and charges”/ (24129)

24. 61. exp budgets/ (9937)

25. 62. (low adj cost).mp. (12442)

26. 63. (high adj cost).mp. (5477)

27. 64. (health?care adj cost$).mp. (2047)

28. 65. (fiscal or funding or financial or finance$).tw. (51383)

29. 66. (cost adj estimate$).mp. (955)

30. 67. (cost adj variable).mp. (25)

31. 68. (unit adj cost$).mp. (976)

32. 69. (economic$ or pharmacoeconomic$ or price$ or pricing).tw. (108532)

33. 70. budget$.tw. (12665)

34. 71. (cost$ adj2 (effective$ or utilit$ or benefit$ or minimi$)).tw. (59198)

35. 72. (fee or fees).tw. (9053)

36. 73. (value adj2 (money or monetary)).tw. (787)

37. 74. or/1-36 (371937)

38. 75. Radiotherapy, Intensity-Modulated/ (896)

39. 76. intensity modulated radiotherap*.tw. (1254)

40. 77. intensity-modulated radiotherap*.tw. (1254)

41. 78. intensity modulated radiation therap*.tw. (1364)

42. 79. intensity-modulated radiation therap*.tw. (1364)

43. 80. imrt.tw. (2402)

44. 81. image guided radiotherap*.tw. (193)

45. 82. igrt.tw. (123)

46. 83. dose compensation.tw. (30)

47. 84. electronic compensation.tw. (28)

48. 85. e compensation.tw. (0)

49. 86. forward planning.tw. (100)

50. 87. inverse planning.tw. (334)

51. 88. field in field.tw. (18)

52. 89. physical compensation.tw. (4)

53. 90. 38 or 39 or 40 or 41 or 42 or 43 or 44 or 45 or 46 or 47 or 48 or 49 or 50 or 51 or 52 (3718)

54. 91. prostatic neoplasms/ (63810)

55. 92. (prostat$ adj5 (cancer$ or carcin$ or tumor$ or tumour$ or neoplasm$)).tw. (63945)

56. 93. ((carcinoma or neoplasia or neoplasm$ or adencarcinoma or cancer$ or tumor$ or tumour$ or malignan$) adj3 prostat$).tw. (61853)

57. 94. 54 or 55 or 56 (78190)

58. 95. 53 and 57 and 37 (19)

59. 96. from 58 keep 1-19 (19)

Appendix 9 Summary study survival and toxicity data by dose

Appendix 10 Biological equivalent radiotherapy doses

If N is the total number of fractions of RT given in a treatment course, and d is the dose per fraction, the biological equivalent dose (BED) is calculated as:

The α/ß ratio is a measure of the effect of radiation on different tissues. Kupelian et al. assumed an (α/ß) ratio of 3. 85

For the 3DCRT group given a total dose of 78 Gy by 39 fractions of 2 Gy each the BED is:

The IMRT group was given a hypofractionated regime of 28 fractions of 2.5 Gy each to give a total dose of 70 Gy.

Thus Kupelian assumed that although the IMRT group received a total dose of 70 Gy compared to a dose of 78 Gy for the 3DCRT group, the hypofractionated regime of the IMRT group meant that the BED doses of the two regimes were comparable. 58

Appendix 11 Example search for utility values in prostate cancer (MEDLINE)

1. prostatic neplasms/

2. (prostat* adj5 (cancer* or carcinoma* or tumor* or tumour* or neoplasm*)).tw.

3. ((carcinoma* or neoplasia or neoplasm* or adenocarcinoma* or cancer* or tumor* or tumour* or malignan*) adj3 prostat*).tw.

4. 1 or 2 or 3

5. (utillity or utilities or eq5d or eq-5d or europol or qwb or hui2 or hui3 or 15d or sf-6d or sf6d or aqol).mp. [mp=title, original title, abstract, name of substance word, subject heading word]

6. 4 and 5

Appendix 12 Unit cost

Appendix 13 Prostate-specific antigen variable distribution parameters

Glossary

Adjuvant radiotherapy

Radiotherapy supplementary to the main treatment (usually surgery, chemotherapy), given after main treatment.

Androgen deprivation

A form of hormonal therapy, given to suppress or block the production or action of male sex hormones (androgens).

Chemotherapy

Treatment with cytotoxic drugs that kill cancer cells, or prevent or slow their growth.

Clinical target volume

Clinically defined target volume, containing tumour, unless surgically excised, and microscopic invisible tumour; to be treated with the prescribed radiation dose.

Compensators/dose compensators

Method of dose modulation.

Computed tomography

An X-ray technique using a scanner to take a series of images across the body.

Conformal radiotherapy

Radiotherapy delivered by non-modulated beams, which can be shaped geometrically to avoid irradiating normal surrounding tissue.

Distant recurrence

Recurrence of cancer at distant sites.

Dose escalation

Increasing the total radiotherapy dose.

Dose–volume histogram

Histogram showing the dose distribution within an outlined structure, often presented as cumulative plot-volume of organ plotted against dose.

Dosimetrist

Specialist radiotherapy planning staff, also known as clinical technologist.

Forward planned intensity-modulated radiotherapy (IMRT)

The planner modifies a provisional plan (based on the treatment beam arrangements that are likely to be used) until the dose distribution is improved.

Fraction

A unit of radiotherapy treatment.

Fractionation

Schedule of treatment sessions required for a course of radiotherapy treatment.

Gross tumour volume

The gross palpable or visible/demonstrable extent and location of the malignant growth.

Gy

A unit of radiation (Gray) used to measure radiation dose.

Histological grade

Measure of the malignancy of a tumour.

Hypofractionated

The radiotherapy dose per fraction is greater than standard care, requiring fewer total fractions to achieve the same biological equivalent dose (BED) as for standard fractionation (see Appendix 9).

Immobilisation

Techniques designed to reduce patient movement during radiotherapy.

Incremental cost-effectiveness ratio (ICER)

The ratio of the incremental cost of the new treatment compared with the existing treatment to the incremental effectiveness, the latter usually being expressed in quality-adjusted life-years (QALYs).

Intensity-modulated radiotherapy

The beam of radiation is not uniform across the field to be irradiated, but consists of beamlets of varying intensity. Combinations of several intensity-modulated fields coming from different beam directions are used to create a highly conformal dose distribution.

Inverse planned IMRT

Utilises computers, manipulating many hundreds of treatment beamlets, in order to produce highly complex treatment plans. The computer iteratively attempts to fulfil the planner’s defined dose target and normal tissue constraints.

Linear accelerator

Machine that generates and delivers radiation.

Locoregional recurrence

Recurrence of cancer at local or regional sites, e.g. within lymph nodes.

Lymph nodes

Small structures that act as filters in the lymphatic system.

Metastases/metastatic cancer

Cancer which has spread to distant sites from the primary tumour.

Multileaf collimator

Device attached to or inherent within to linear accelerator to allow beam modulation, uses a number of leaves to create an irregular shaped radiation beam.

Neoadjuvant

Treatment given prior to surgery, chemotherapy or radiotherapy.

Overall survival

Outcome measure defined as the hazard of death from any cause after a given follow-up period, or time from randomisation to death from any cause.

Planned dose inhomogeneity

Planning higher doses to regions of high risk and lower doses to regions of low risk, usually termed with treatment delivered with a single treatment plan.

Planning target volume

Geometrical concept defined to ensure that appropriate beam sizes and arrangements are chosen so that the prescribed dose is directed to the clinical target volume, taking into account setup and delivery errors and physiological changes such as motion and changes in tumour volume.

Progression-free survival

Outcome measure defined as the hazard of disease progression or death from any cause after a given follow-up period, or time from randomisation to first of these events.

Prostatectomy

Surgery to remove all of the prostate gland and some of the tissue around it, to treat prostate cancer.

Quality-adjusted life-years (QALYs)

The number of life-years adjusted for population preferences for different health states. A life-year in perfect health is 1 QALY.

Quality assurance

Procedures that ensure consistency of the prescription and safe delivery of prescription dose to the target volume.

Radiotherapy/radiation therapy

Radiation delivered locally to affected site to kill cancer cells, or to stop cancer cells from dividing and growing.

Short Form questionnaire-36 items

A health-related quality of life scale.

Staging/stage of cancer

An internationally recognised system for defining a tumour in terms of its size and degree of spread through the body.

Toxicity grade

A measure of the severity of adverse events, either during radiotherapy (acute) or long-term toxicity (chronic).

Wedges

Method of dose modulation.

List of abbreviations

3DCRT

three-dimensional conformal radiotherapy

AE

adverse event

ASCO

American Society of Clinical Oncology

ASTRO

American Society of Therapeutic Radiology and Oncology

BED

biological equivalent dose

BPE

benign prostatic enlargement

CHHiP

Conventional or Hypofractionated High dose Intensity-modulated radiotherapy for Prostate cancer

CI

confidence interval

CRT

conformal radiotherapy

CT

computed tomographic

CTV

clinical target volume

DRE

digital rectal examination

EORTC

European Organisation for Research and Treatment of Cancer

EORTC QLQ-PR25

European Organisation for Research and Treatment of Cancer Prostate Cancer-specific Questionnaire

EPIC

Expanded Prostate Cancer Index Composite

EQ-5D

European Quality of Life-5 Dimensions

ESTRO

European Society for Therapeutic Radiology and Oncology

GI

gastrointestinal

GP

general practitioner

GTV

gross tumour volume

GU

genitourinary

HR

hazard ratio

HRQoL

health-related quality of life

ICER

incremental cost-effectiveness ratio

ICRU

International Commission on Radiation Units and Measurements

IGRT

image-guided radiotherapy

IMRT

intensity-modulated radiotherapy

IPEM

Institute of Physics and Engineering in Medicine

LUTS

lower urinary tract symptoms

MAICER

maximum incremental cost-effectiveness ratio

MLC

multileaf collimator

MRC

Medical Research Council

MRI

magnetic resonance imaging

NCCN

National Comprehensive Cancer Network

NCI

National Cancer Institute

NCI-CTC

National Cancer Institute–Common Toxicity Criteria

NICE

National Institute for Health and Clinical Excellence

NRAG

National Radiotherapy Advisory Group

PC

prostate cancer

PSA

prostate-specific antigen

PTV

planning target volume

QA

quality assurance

QALY

quality-adjusted life-year

RCR

Royal College of Radiologists

RCT

randomised controlled trial

RT

radiotherapy

RTOG

Radiation Therapy Oncology Group

SCOR

Society and College of Radiographers

SD

standard deviation

SF-36

Short Form questionnaire-36 items

SV

seminal vesicle

TNM

tumour node metastasis

TRUS

trans-rectal ultrasonography

UCLA PCI

University of California Los Angeles Prostate Cancer Index

WP

whole pelvis

Notes