The clinical effectiveness and cost-effectiveness of different surveillance mammography regimens after the treatment for primary breast cancer: systematic reviews, registry database analyses and economic evaluation

Significance for patients in terms of ill health

Ipsilateral breast tumour recurrence and MCBC have significant adverse effects on the patient. Further treatment is required and this often necessitates mastectomy for the patient who has a previously conserved breast together with the consideration of the use of systemic treatments (chemotherapy, hormone therapy, biological therapy). Disease recurrence has a significant adverse impact psychologically with major implications for the patient, their health and quality of life. There are data to indicate that patients who experience local disease recurrence have a poorer survival than those who do not have local recurrence. 2,7

Significance for the NHS

The significance to the NHS in terms of the provision of clinical and mammographic surveillance, and treatment of IBTR and MCBC, is great. Following the treatment of primary breast cancer, patients are followed up with regular clinical examinations and surveillance mammography carried out at intervals as described below. Subsequent investigations are carried out to confirm disease recurrence or to exclude disease in those incorrectly identified as positive by a prior test or examination (‘false-positive’).

For each annual cohort of approximately 45,000 new cases of breast cancer, 20% (9000) will have developed metastatic disease and die within 5 years, requiring complex, demanding and costly treatment regimens. Taking a mean age of 70 years for primary breast cancer diagnosis, if 1% develop IBTR each year, and accounting for death from other causes, then over a 20-year period approximately 6358 patients will require treatment for this with 20% requiring further treatment for systemic disease.

Management of disease

Women with primary disease are usually treated using a combination of treatment modalities, including surgery, radiotherapy, chemotherapy, hormone therapy and biological therapy (trastuzumab), either alone or in a variety of combinations tailored to the patient and the type and stage of disease. However, most patients undergo surgery initially, which is either mastectomy or BCT, together with axillary surgery (sentinel lymph node biopsy, axillary sample or axillary clearance) and tailored adjuvant therapy afterwards. Patients with large or locally advanced breast cancer may receive either primary (neoadjuvant) chemotherapy or hormone therapy prior to undergoing surgery. A small number of patients who are unfit for surgery but are hormone receptor positive may receive endocrine treatment and/or palliative radiotherapy as their sole treatment.

The management of locoregional recurrence following the treatment of primary breast cancer is variable in the UK, as there is an absence of RCTs to guide management decisions. Following breast conservation surgery it has been reported that, in those patients who experience locoregional recurrence, up to 20% may have distant metastases at the same time. 19,20 Furthermore, other studies have indicated that patients who experience a locoregional relapse have a reduced 5-year survival. 2,3 Where there is a recurrence in the chest wall after mastectomy, patients will frequently have systemic recurrence and therefore all patients presenting with recurrent breast cancer should be restaged [investigations may include combinations of haematological investigations, chest radiography, abdominal ultrasound, computerised tomography (CT) scan of chest and abdomen, isotope bone scan] prior to definitive management. 21

A multidisciplinary approach is required for the management of patients with locoregional recurrence following treatment for primary breast cancer. For patients who have undergone breast conservation surgery, treatment is usually mastectomy (with axillary clearance if not already performed), together with consideration of further systemic therapy (chemotherapy, hormone therapy) to reduce the risk of subsequent metastatic relapse. In patients who have a local chest wall recurrence following mastectomy, local therapy (surgery if possible with radiotherapy afterwards or if surgery is not possible due to the extensive nature of the disease then radiotherapy alone) may be undertaken together with systemic therapy (chemotherapy, hormone therapy) given the high risk of subsequent metastatic relapse in this group of patients.

There are no RCTs on which to base the decision to offer systemic adjuvant therapies. In those patients without detectable systemic metastases, factors taken into account are as for the use of adjuvant therapy, for example tumour size, tumour grade, lymph node status, lymphovascular invasion, hormone receptor and human epidermal growth factor receptor 2 (HER2) status and the time interval from the treatment of the patient’s original primary breast cancer.

Variation in services and/or uncertainty about best practice

There is considerable debate about the optimal organisation of a surveillance service following treatment for primary breast cancer. The number of different guidelines produced by various professional groups and policy-makers reflects this. 21–25 Previous surveys of breast surgeons, oncologists26 and NHS breast-screening units27 indicate that there is wide variation in follow-up practice but, in general terms, surveillance includes regular clinical examination, up to 3–5 years, with annual mammography for up to 10 years, or in some units this is carried out indefinitely. The most recent National Institute for Health and Clinical Excellence (NICE) guidance recommends mammography annually for 5 years and then follow-up through the NHS Breast Screening Programme (NHSBSP). 24 There is concern over whether the 3-yearly interval is sufficient for the group under surveillance, and some believe a stratified approach based on risk of recurrence or contralateral disease is more appropriate. The American Society of Clinical Oncology28 recommended that surveillance should include annual mammography but acknowledges that there is a lack of high-level evidence to support decisions about the frequency and timing of mammographic surveillance. These guidelines have usually been based on consensus approaches and literature reviews, and their key recommendations are shown in Table 1.

For women who have had treatment for breast cancer there is, however, general agreement that there is no survival advantage conferred by an intensive surveillance regimen (comprising chest radiography, liver ultrasound, haematological investigations and isotope bone scans) compared with a simpler follow-up schedule of clinical examination and mammographic surveillance. 29

Given the variation in recommendations, it would be surprising if there were no variation in practice. The results of a survey of practice conducted for this study are described in more detail in Chapter 3.

Current and anticipated costs

As reported above, there are follow-up regimens for women following treatment of primary breast cancer. Illustrative annual costs of alternative surveillance regimens are presented in Table 2. The calculation of these costs is reported in detail in Chapter 7.

Surveillance mammography

Mammography has been in use for > 30 years and is the standard imaging technique for detection of breast cancer. 30 In women previously treated for breast cancer, surveillance mammography is useful for early detection of tumour recurrence or for confirming the absence of recurrent cancer. While tumour recurrence may display similar mammographic features to the previous primary breast cancer,31 interpretation of the surveillance mammogram is hindered by changes in the breast caused by postoperative scarring and changes to breast density caused by primary treatment modalities. For example, following surgery and/or radiotherapy, detectable abnormalities on mammography include haematoma, scar formation, fat necrosis, skin thickening, increased soft tissue density in the ipsilateral breast and microcalcifications, all of which can be misinterpreted as malignancy. Therefore, surveillance mammography after the treatment for breast cancer is associated with the possibility of false-positive results causing further unnecessary investigations (invasive and non-invasive) and reduction in sensitivity for the detection of IBTR. There may also be an effect on MCBC detection with a lack of a comparator side.

Although published figures vary, it has been estimated that approximately 50% of IBTRs will be detected by mammography, with the remainder being detected by clinical examination. 31–33 Patients who have had a mastectomy or those who have undergone breast reconstruction following mastectomy do not have mammographic surveillance of that side. Clinical examination alone is the standard method of follow-up to detect IBTR, although mammography is undertaken of the remaining breast. Patients who have undergone mastectomy may find it easier to detect IBTR clinically than those who have undergone breast conservation surgery or reconstruction. Approximately 10% of breast cancers that can be palpated on clinical examination are not clearly visible on mammography and require the use of additional imaging techniques. Recurrent breast cancers detected by mammography are generally smaller and less aggressive than those found on clinical examination. 31,32 It is presumed, therefore, that mammography, combined with clinical examination, allows the earliest possible diagnosis of IBTR and also allows surveillance of the contralateral breast for the detection of MCBC. Whether such surveillance regimens reduce mortality remains unclear at present.

Mammography involves low-dose X-ray imaging of the breast to create detailed soft tissue, high-contrast, high-resolution images, which are recorded on photographic film. Mammograms are usually produced by a radiographer and interpreted by a radiologist who is trained in breast imaging. Recent developments have led to an increasing use of solid-state detectors rather than X-ray film, termed digital mammography or full-field digitalmammography (FFDM). These convert X-rays into electrical signals. The images produced are displayed on a computer screen but can be printed onto radiographic film that is similar to conventional mammograms. It is possible to manipulate digital images on-screen to enhance visibility of certain areas. Digital mammography is quicker to produce than film mammography, uses lower doses of radiation and digital images require less physical storage space than traditional films. Digital mammography systems are, however, one to four times more costly than film mammography systems. 34 In the screening population, digital mammography has improved performance over film mammography in younger women and in women with dense breasts. Overall, however, the diagnostic accuracy of digital mammography is not significantly greater than that of film mammography. 35

Other relevant new interventions

Identification of important subgroups of patients

It is known that certain groups of patients are at increased risk of IBTR. Those patients tend to be under 50 years of age at the time of diagnosis of their first breast cancer, have tumours classified histologically as being grade 3 cancers, have larger tumours, lymph node involvement and lymphovascular invasion (Table 3 gives an estimate of importance of these factors in IBTR). Pathologists in the UK report this information routinely and much of these data are held by the cancer registries. We focused on these risk factors as it is possible to stratify patients according to these variables and also give some indication on the hazard ratio (HR) of developing IBTR or MCBC when taking these factors into account. These variables have been used to estimate prognosis, as with the Nottingham Prognostic Index, for example, which is based on tumour size, grade and lymph node status. Adjuvant! Online also uses similar criteria to calculate the benefits of different types of treatment for each particular patient.

Summary of existing systematic reviews

Lu5 recently conducted a systematic review to determine the impact of early detection of isolated locoregional and contralateral recurrence on survival. Thirteen studies considered routine follow-up (regular mammography and physical examination) or intensive follow-up (with the inclusion of routine additional tests) aimed at early detection of recurrence. The authors defined early detection as detection by mammography instead of physical examination or in asymptomatic patients, as opposed to those presenting with symptoms or detected via physical examination either by clinician or by the patient. The authors reported better overall survival for recurrences detected by mammography or in asymptomatic patients, with an absolute reduction in mortality of 17–28% if all breast cancer recurrences are detected early. The authors had insufficient data to study the contribution of differing frequency of mammography (e.g. yearly mammography compared with 6-monthly mammography) or to analyse the effect of prognostic factors such as age, tumour stage and surgical treatment. Rojas and colleagues29 focused on the effectiveness of different surveillance policies for the detection of distant metastatic disease and concluded that follow-up programmes of regular physical examination and yearly mammography alone are as effective as more intensive approaches involving laboratory and radiological tests in terms of timeliness of detection of tumour recurrences, overall survival and quality of life. In addition, follow-up of patients performed by a trained general primary care practitioner is comparable to that of hospital-based secondary care specialists in terms of detection of tumour recurrence and quality of life. Collins and colleagues44 reported that patient survival and quality of life were not affected by intensity of follow-up or location of care, although the authors note that data were not sufficiently homogeneous to integrate statistically. Montgomery and colleagues46 systematically reviewed RCT evidence for alternative follow-up methods and concluded that the trials reviewed were not adequately powered to establish the safety of reducing or replacing hospital clinic visits.

Only Grunfeld and colleagues,47 Montgomery and colleagues46 and Barnsley and colleagues43 specifically considered the role of mammography in surveillance, and Barnsley and colleagues43 focused solely on surveillance mammography of the reconstructed breast, concluding that certain local recurrences are detectable by surveillance mammography but that there is a paucity of evidence.

De Bock and colleagues45 conducted a review of 18 uncontrolled, prospective and retrospective non-randomised studies of the effectiveness of routine follow-up visits and tests. The proportion of isolated locoregional recurrences diagnosed during routine visits or routine tests in asymptomatic patients was compared with the proportion of isolated locoregional recurrences in symptomatic patients. The authors were unable to assess whether recurrences, as defined by the study authors, were detected by physical examination or other tests, including mammography, or whether the detection of asymptomatic isolated recurrences had any influence on the potential for cure or quality of life of the patients.

Montgomery and colleagues,4 in a meta-analysis of 12 studies (11 non-randomised cohorts and one randomised trial), aimed to determine the relevant contributions of clinical examination, patient self-examination and mammography to the detection of potentially treatable locoregional recurrence and contralateral primary cancers. It was reported that 30–40% of treatable recurrences, as defined by the study authors, were detected by the patients self-examining. Prior to 2000, 15% of relapses were mammographically detected compared with 46% detected by routine clinical examination. Post 2000, 40% were mammographically detected and 15% were detected through routine clinical examination. Patients with ipsilateral recurrence detected by clinical examination appeared to do less well in terms of survival than those detected by self-examination or mammography. The authors concluded that there was no evidence that clinical examination confers a survival advantage compared with other methods of detection and thus the need for clinical follow-up in detection of relapse is uncertain. The authors suggest that the temporal trend for increased detection by mammography is due to technical improvements in mammography and better quality assurance. Houssami and colleagues48 recently reported a similar trend for mammographic detection, although they propose that this effect is largely due to increased uptake of surveillance mammography by women rather than increased sensitivity of mammography.

Grunfeld and colleagues47 systematic review to define the effect of routine surveillance mammograms in detecting ipsilateral and contralateral cancer included 15 observational studies (published 1980–99). The 10 studies of ipsilateral recurrence showed that mammography detected the recurrence in between one-quarter and one-half of the women (range 8–50%), with the remainder being found by the women themselves between follow-up or by a hospital practitioner during clinical examination. The majority of studies did not report outcomes. Where this was reported, the method of detection of ipsilateral tumour did not appear to influence survival, except in the study conducted by Voogd. 49 Here it was reported that patients had a better 5-year survival if their tumour recurrence was detected mammographically. The nine studies of MCBC showed similar variation of detection methods. However, only one study reported outcome and this showed there was no difference in survival when comparing mammographic detection of the tumour with other methods. 33 The authors did not conduct a meta-analysis and concluded that further research is needed to better define the optimum surveillance mammography regimen following breast cancer treatment.

Overall, from these reviews the optimal frequency and duration of surveillance mammography is not clear. Furthermore, more recent information is required on whether there is new evidence concerning the effectiveness of surveillance mammography. This is now extremely important because of the development and use of new and increasingly effective treatments for patients with breast cancer since 1990. These may offer women an improvement in survival if there is an early detection of either IBTR and/or MCBC. A further limitation of all these reviews is that they did not consider the costs and cost-effectiveness of surveillance mammography compared with other follow-up regimens despite there being methods for incorporating economic evidence into systematic reviews. 50 This is important because the evidence to date has not been systematically reviewed to assess whether or not surveillance mammography is cost-effective in the follow-up of patients with breast cancer. As health-care resources are limited, they have to be used effectively for the benefit of society. Using limited resources to provide surveillance mammography will mean that we cannot use those resources to provide some other form of potentially beneficial care. For surveillance mammography to be considered efficient the benefits that it provides must be greater than the benefits we could have obtained from providing other care.

Relevance of existing data to the decision problem

The introduction of the NHSBSP in the UK in 1988, coupled with advances in the treatment of primary breast cancer around 1990, has led to improvements in overall survival, with the 5-year relative survival rate now 82% in England and Wales. 1 Although long-term follow-up would be the most useful to inform the decision problem, technological developments in all aspects of diagnosis, treatment and follow-up of women make those women with the longest follow-up data the less relevant and their outcomes less generalisable to current practice. Therefore, in consultation with expert members of our Advisory Group (see Appendix 1 for details of Advisory Group members) we decided to narrow our population of interest to consider data only for women treated for breast cancer from 1990 onwards. In addition, we decided not to include information from 1990 onwards relating to the Breast Screening Programme. This is because the screening population differs greatly to women who have been diagnosed and treated for breast cancer due to changes in breast density following treatment for primary breast cancer. We therefore felt that it would be inappropriate to use data from breast-screening studies of test performance in the systematic reviews (Chapters 4 and 5) to make assumptions regarding test performance for surveillance of the contralateral breast. We used parameter estimates for MRI test performance in the screening population in the economic evaluation in Chapter 7 of this report; however, it was felt that these would provide an indication of the relative value of a more costly but more effective test.

Key issues

The key issues to be addressed are:

• Can surveillance mammography improve overall survival for women treated for primary breast cancer?

• Does surveillance mammography improve detection of IBTR and MCBC?

• What is the incremental cost-effectiveness of surveillance mammography?

Current care pathway

There are a number of different surveillance mammography regimens in place in the UK for women following the treatment of primary breast cancer. In this section, we describe current pathways of care for women who are diagnosed, treated and followed up for breast cancer. In Chapter 3 we describe the potential alternative care pathways that we will attempt to consider and how we derived them.

However, when evaluating different surveillance regimens it is important to understand the sequence of care that a woman might receive after treatment for a primary breast cancer. Consideration can then turn to how different surveillance regimens may alter the care that a woman may receive over time.

Care pathways for current practice

We developed a care pathway for Aberdeen via discussions with experts in Aberdeen (Figures 2–4). This care pathway starts with initial presentation and describes the sequence of events from diagnosis, through treatment and eventual longer-term follow-up. It is useful to consider the whole sequence of events, but of central importance to this project is how different surveillance regimens will potentially alter this care pathway.

FIGURE 2.

Current care pathway: Aberdeen: diagnosis of breast cancer – symptomatic presentation. I-guided core biopsy, image-guided core biopsy; MDT, multidisciplinary team; U/S, ultrasound.

FIGURE 3.

Current care pathway: Aberdeen – clinical management of a woman presenting with a malignancy. CT, chemotherapy; ER, oestrogen receptor; MD, metastatic disease; MDT, multidisciplinary team; NMD, non-metastic disease; PR, progesterone receptor; R/T, radiotherapy.

FIGURE 4.

Current care pathway: Aberdeen: follow-up of non-metastatic and metastatic disease.

Development of alternative surveillance regimens

Taking the care pathways above as a basis, we used the data reported in Chapter 3 to identify potentially relevant pathways for alternative surveillance regimens. We also considered whether or not there are any clinically attractive follow-up regimens that might not be used in practice but that we might consider useful to estimate their effectiveness, cost-effectiveness and feasibility in our subsequent modelling exercise. This consideration was partly informed by our discussions at the last Advisory Group Meeting, the literature and the results of our survey reported in Chapter 3.

When considering what surveillance regimens might be relevant, answers to the following questions were sought.

For mammographic surveillance:

• For which women is the issue of mammographic surveillance relevant?

• What mammography surveillance should be used?

• Does it vary between women and if so why would it vary?

• How often is it performed?

• Where does it take place?

For clinical follow-up:

• What clinical follow-up is used?

• Does it vary between women and if so why?

• Where does it take place?

• How often?

For unstructured primary care follow-up:

• How might a diagnosis be made?

• At what point would these women enter the care pathway described in Figures 2–4?

More specifically, what factors might influence the choices made about mammographic surveillance and clinical follow-up?

Other factors to consider:

• age

• risk factors

• type of primary disease

• type of treatment.

We describe the care pathways developed from this process in Chapter 3. We then used the care pathways to structure data collection in the remainder of the project and to help to define the comparators for the economic evaluation. The structure of the model, which is detailed in Chapter 7, was based upon the current care pathway described above. The structure of the economic model was defined to directly address the aim of the review set out below.

Objectives

• To identify current UK surveillance mammography regimens.

• To inform the feasible alternative surveillance regimens (care pathways) for:

  1. – populating the economic model

  2. – informing the systematic reviews

  3. – providing context for the individual patient data analysis.

• To inform the choice of comparator surveillance regimens (inclusive or not of mammography) for the systematic review components of the wider project.

Population and sample

Our population was all health-care professionals providing surveillance of women following treatment for primary breast cancer. We chose our sample from this population to reflect those most likely to be currently involved with organising and/or undertaking surveillance mammography and to try to ensure UK-wide information. We sampled from the Association of Breast Surgery (ABS) at the BASO and the RCR Breast Group. Both the ABS and RCR Breast Group (to the best of their knowledge) held current and complete e-mail contacts for their members and these lists formed our sample. We sampled all full members (496) and associate members (73) of the ABS, and ordinary members (447 radiologists) and associate members (32 breast physicians) of the RCR Breast Group. We excluded the retired and overseas members of both organisations.

Data collection, management and analysis

Initiation, frequency and duration of surveillance mammography

The large majority of respondents initiate surveillance mammography at 12 months post surgery for women who have had breast-conserving surgery (BCS) (87%) and for women who have had a mastectomy (79%) (Table 5).

Responses ranged from six to 24 months post surgery, with the next most frequent being 24 months (13%) post mastectomy.

Table 6 shows the respondents standard practice frequencies or intervals of surveillance mammography for women after BCS and after mastectomy. Annual surveillance mammography was the most commonly reported standard frequency of surveillance mammography for women after BCS or after mastectomy (72% and 53%, respectively), with biennial mammography the next most frequently reported (12% and 30%, respectively). The ‘other’ responses varied but can generally be described as showing a pattern of increasing mammography surveillance intervals with increasing time since surgery, for example surveillance mammography at 1, 2, 3, 5, 7 and 10 years.

Fourteen per cent (26/180) of respondents said that they varied their standard surveillance mammography practice (initiation of surveillance mammography post surgery or frequency of mammography) for women who had BCS. They varied their practice according to the survey-prompted criteria of in situ tumours (n = 14); size of tumour (n = 5); grade of tumour (n = 1); lymphovascular invasion (n = 4); age (n = 9); absence of radiotherapy (n = 3); combinations of these (n = 2); or other criteria (n = 10). Other criteria included ‘close margins’, comorbidities, family history and genes predisposing to breast cancer.

Similarly, 13% (23/180) varied their standard surveillance mammography practice (initiation of or frequency) for women who had a mastectomy, by factors such as age, cancer grade and size, comorbidities, family history, genes predisposing to breast cancer or ‘high-risk’ groups.

In addition, a further 16/183 (9%) commented in text within the questionnaire that they vary their standard initiation and frequency surveillance practices, trying to take into account factors such as age, density of breast tissue, comorbidities, family history, genes predisposing to breast cancer or ‘high-risk’ groups.

When asked about through which service they arrange their surveillance mammography, the majority responded [175/182 (96%)] that it is through their symptomatic breast service, although seven (4%) said through the NHSBSP.

The majority (75%, 136/182) indicated they discharge women from surveillance mammography and they do this most frequently 10 years after surgery (Table 7). The majority (82%, 148/180) do discharge from clinical follow-up and most frequently at 5 years (Table 7). Just over half (55%, 98/179) responded that they discharge women to the NHSBSP (Table 8) if eligible.

However, around 28% (47/167) of those who discharge from follow-up (clinical and/or mammographic follow-up) commented that they vary the duration of surveillance mammography and this is influenced by the age of the women (24%), or by other factors including family history, genes predisposing to breast cancer, and tumour characteristics.

Combining our respondents’ standard initiation, frequency and duration of surveillance mammography resulted in 54 differing surveillance regimens for women after BCS and 56 for women following mastectomy (Appendices 7 and 8, respectively). Fifty-one per cent (79/154) of respondents follow one of four surveillance regimens for women after BCS. The most commonly followed regimens are to initiate surveillance mammography at 12 months after surgery and conduct annual surveillance mammography with indefinite duration (12%, 19/154); discharge from both clinical and mammographic surveillance at 5 years (14%, 22/154) or 10 years (12%, 18/154) after surgery; or discharge from mammographic surveillance at 10 years and clinical follow-up at 5 years (13%, 20/154). Similarly, after mastectomy the most commonly followed regimens are to initiate surveillance mammography at 12 months after surgery and conduct annual surveillance mammography, with indefinite duration (7%, 10/136); or discharge from both clinical and mammographic surveillance at 5 years (10%, 13/136); or 10 years (11%, 15/136) after surgery; or discharge from clinical follow-up at 5 years with continued mammographic surveillance until 10 years (8%, 11/136).

Ideal practice

Twenty-nine per cent (53/180) of respondents suggested that their ideal surveillance mammographic practice differs from their current practice and that this is influenced by the factors listed in Box 1.

The most common ideal frequency of surveillance mammography given was annually for women who had undergone BCS (80%, 85/106) or mastectomy (57%, 61/106) (Table 9). These ideal frequencies of surveillance did not differ from their current practice for the majority of respondents, for women after BCS (80%, 84/106) or for women after mastectomy (69%, 73/106). However, three respondents suggested that their ideal practice would be to arrange surveillance mammography through the screening units, as they are set up to manage the appointment and recall system.

Development of alternative surveillance regimens

Taking the care pathways described in Table 1 (see Chapter 1) and Figures 1–3 (see Chapter 2) as a basis, we used the results of the surveys to identify if there were any clinically attractive follow-up regimens that might be used in practice or are currently not used in practice. This consideration was partly informed by our discussions during project Advisory Group Meetings (which were informed by the literature and the results of the survey described above).

When considering what surveillance regimens might be relevant, answers to the following questions were sought, which can be briefly summarised as: who would be under surveillance/follow-up; what technology would be used (e.g. mammography, clinical examination, etc.); where would the surveillance be performed; who would perform the surveillance; and what would be the frequency of surveillance/follow-up (the questions used are described in more detail in Appendix 9).

Surveillance regimens

Figure 5 describes the potential alternative care pathways developed from this process. For example, individuals can be followed up using surveillance mammography at different intervals, for example once yearly, every 18 months, every 24 months or every 36 months. Alternatively, individuals could present to a GP with a problem, i.e. discover a lump. Individuals who present to their GP with a lump would be given a clinical examination by the GP. Current practice in the economic model is assumed to be once-yearly mammograms.

FIGURE 5.

Potential alternative care pathways. F/U, follow-up.

The alternative surveillance regimens in the economic model vary by screening interval and/or screening technology. For example, alternative mammographic surveillance regimens to the standard regimen would be for mammography to take place at less or more frequent intervals, for example every 18 months or every 24 months. Although not explicitly noted, one important option to consider as an alternative would be surveillance mammography organised through the NHSBSP.

Alternative primary care regimens would be for an individual to attend a GP surgery and receive a clinical examination followed by a mammogram if there was a suspicious finding on the clinical examination. Other potential surveillance regimens include the use of alternative technologies, i.e. MRI or ultrasound in replacement of mammography. For all regimens other than GP opportunistic finding, individuals are invited to attend screening at different intervals, for example once yearly. An individual can either choose to attend or not attend the screening programme. Given that this is a higher-risk group (women who have previously had breast cancer), and, also for simplicity of modelling, we are assuming that all individuals who are invited for screening do attend. In the intervals in which screening does not occur, we assumed that individuals could still be diagnosed with breast cancer through their GP.

Following further discussion within the project Advisory Group, these options were further reduced to three regimens that we felt broadly represented the most relevant comparators. This decision was also informed by knowledge of the preliminary findings of the research reported in Chapters 4–6. These regimens were: mammographic surveillance with and without clinical follow-up organised either through secondary care or through the screening service (this option embraces regimens 1, 3 and 6 in Figure 5) and the identification of cancer following referral from primary care following the identification of a suspicious lump on self-examination (regimen 7).

Inclusion and exclusion criteria

Search methods for identification of studies

We conducted an extensive electronic search to identify reports of relevant published and ongoing studies, as well as any grey literature. The search strategies were designed to be highly sensitive, including both appropriate subject heading and text word terms to capture the concepts of surveillance mammography or other follow-up strategies and the study designs meeting the inclusion criteria for this review. The searches were restricted to full text papers published from 1990 onwards without language restriction. We searched the following databases for primary studies: MEDLINE, MEDLINE In-Process, EMBASE, BIOSIS, Science Citation Index (SCI), CANCERLIT and Cochrane Central Register of Controlled Trials (CENTRAL). We also searched the Cochrane Database of Systematic Reviews (CDSR), Database of Abstracts of Reviews of Effects (DARE) and the HTA Database for reports of evidence syntheses. Reports of ongoing and recently completed trials were sought from the Current Controlled Trials (CCT), Clinical Trials, WHO International Clinical Trials Registry Platform (ICTRP), NCI Clinical Trials Database, National Research Register (NRR) Archive, and NIHR Portfolio Database. Appendix 10 gives full details of the search strategies used.

In addition, we searched relevant websites, namely those of the National Cancer Institute, National Comprehensive Cancer Network, CancerWEB, Breast Cancer Surveillance Consortium, and the National Library for Health, as well as relevant professional organisations including the American Society of Clinical Oncology, the American Society of Breast Disease, the American College of Radiology, and the European Society for Medical Oncology. We scanned reference lists of all included studies for additional reports.

Data extraction strategy

One reviewer (from GM, CR, RT and SZ) screened the titles and available abstracts of all reports identified by the search strategy for relevance to the inclusion criteria. One reviewer independently assessed full text copies of all potentially relevant studies to assess them for inclusion (from CB, CR and SZ). An economist reviewer examined reports relating to an economic evaluation or cost analysis.

We conducted a 10% check of inclusion assessment for all potentially relevant studies (RT). We resolved any disagreements by consensus or arbitration by a third party. A list of the included and excluded studies is given in Appendices 11 and 12, respectively.

One reviewer (from CR and SZ) independently extracted details of study design, participant characteristics, description of the intervention and outcome data (see Appendix 13 for data extraction form). A second reviewer independently validated the data extraction (from CR and SZ). In the event of any uncertainty, a third reviewer advised on and validated the data extraction (CB).

Quality assessment strategy

We assessed the methodological quality of non-randomised studies using a quality assessment tool (Appendix 14) adapted from the Review Body for Interventional Procedures (ReBIP) checklist for quality assessment of non-randomised studies (comparative studies and case series). We included additional items (questions 18 and 19) to assess whether study authors attempt to correct for lead and length time bias in their analyses. Each of the items was checked as ‘yes’, ‘no’ or ‘unclear’. Each item was worded so that a rating of ‘yes’ was the optimal rating of methodological quality, except item 14 regarding differential dropout rate/participants lost to follow-up. We planned to use an adapted version of the Cochrane Collaboration’s tool for assessing risk of bias51 for assessing the methodological quality of individual RCTs. For the quality assessment of any economic evaluations, we planned to use the NHS Economic Evaluation Database Handbook. 52

Data analysis

We planned statistical synthesis of results (using meta-analysis) of included studies directly comparing different surveillance mammography regimens or comparing surveillance mammography with an alternative follow-up regimen, for RCTs and non-randomised comparative studies, favouring intention to treat over per-protocol results for our analysis. We planned to derive a pooled HR for time-to-event outcomes (e.g. recurrence and survival). For data on harms of mammography, adverse events and quality of life we planned to use standardised mean difference to combine quality-of-life scores depending on the suitability of study data. We did not plan quantitative synthesis of economic outcomes data.

Number and type of studies included and excluded

Figure 6 shows the number of potentially relevant studies identified by the search strategy with details of the number meeting the inclusion criteria and the number that were ineligible by exclusion criteria.

FIGURE 6.

Flow chart of the number of potentially relevant reports of identified studies and the number subsequently included and excluded from the effectiveness review.

From the literature searches 2849 titles and abstracts were identified, 422 of which were selected for full text assessment. We excluded 414 reports, which did not meet the inclusion criteria for this review, of which we retained 114 reports to assess eligibility for inclusion in the systematic review of test performance (see Chapter 5 of this report). Seven reports were unavailable. We list the bibliographic details of the eight studies that met the inclusion criteria in Appendix 11. We list the bibliographic details of the excluded studies, plus the reasons for exclusion in Appendix 12.

Characteristics of the included studies

Eight studies met our inclusion criteria. 53–60 Appendix 15 provides full details of the characteristics of the included studies. Six studies53–58 were retrospective cohort studies. Two studies59,60 were prospective cohort studies. We did not identify any RCTs or economic evaluation studies meeting our inclusion criteria. Three studies53–55 were conducted in the UK, whereas three58–60 were conducted in the USA. The study by Paszat and colleagues56 was conducted in Canada, and the study by Yau and colleagues57 was conducted in China (Hong Kong). Table 10 provides a summary of overall characteristics for the included studies. Table 11 provides further details of the characteristics of individual studies.

As we lacked RCT studies directly comparing different surveillance mammography regimens we could not conduct a formal meta-analysis of these studies. In addition, none of the included studies compared surveillance regimens; therefore, it was not possible to undertake meta-analysis assessing surveillance regimens, including investigation of subgroup factors. Consequently, we decided to present a narrative synthesis of results for this review.

Four studies reported data for surveillance mammography only56,58–60 and did not report details of any additional follow-up given to participants. Three studies considered surveillance mammography combined with clinical examination. 53–55 The study conducted by Yau and colleagues57 considered surveillance mammography, combined with clinical and ultrasound examination of the breasts, conducted at the clinician’s discretion. Six studies did not include a comparator regimen53–55,57,58,60 Lash and colleagues59 comparatively analysed the number of consecutive years of guideline surveillance (defined as annual history, annual clinical examination and annual surveillance mammography) received by women in their cohort with women who had not received consecutive years of guideline surveillance (i.e. women who had missed one or more annual surveillance appointment for unspecified reasons). Paszat and colleagues56 compared women in receipt of more than one episode of surveillance mammography within their cohort with women who did not receive surveillance mammography.

Overall, the eight studies enrolled 7337 patients. After exclusions, due to eligibility or participant dropout, the studies included 3775 patients in their analyses. The studies included 1626 mastectomy patients and 4864 breast conservation surgery patients all treated for primary breast cancer and without detectable metastatic disease. Five studies53,56,58–60 reported participant age details, with 444 participants aged < 50 years and 4168 participants aged 50 years or older. Two studies54,55 reported mean ages, 56 and 58 years, and age ranges, 24–91 and 28–91 years, respectively. Yau and colleagues57 reported a median age of 46 years and range 25–90 years. The earliest report was published in 200153 and the latest in 2009. 55 The earliest date of primary treatment reported was 199058 and the latest was 2003. 57 Follow-up ranged from 2.4 months to 15 years.

Quality of the included studies

The results of the quality assessment for the individual studies are shown in Appendix 16. Figure 7 summarises the quality assessment of the included studies.

FIGURE 7.

Summary of quality assessment of the included effectiveness studies.

Four studies56,58–60 were considered to include samples that were unrepresentative of those women who we considered eligible for surveillance mammography (i.e. all women treated for primary breast cancer). Three of these studies, conducted by the same lead author,58–60 included only women aged over 65 years. It is unclear whether the cohorts of women included in the studies conducted in 200559 and by Lash and colleagues58 included the same women and hence whether or not the studies had an overlap of patients. This older age group represents only a proportion of what we consider the eligible population. A pragmatic surveillance mammography regimen could include women of all ages, with those over the age of 50 years possibly benefiting from eligibility for inclusion in the national NHSBSP. In addition, we also considered that the sample included in the study conducted by Paszat and colleagues56 was unrepresentative. In this study, the authors randomly selected two samples from their previously identified population of women treated for primary breast cancer. The authors drew one random sample from women without any further breast surgery after their initial primary treatment. A larger second sample was drawn from the same group who had undergone further breast surgery 6 or more months after their initial treatment. The second sampling fraction was larger (0.237 compared with 0.055) to increase the probability of including women with an episode of IBTR and/or MCBC. We therefore considered that this second group of women with subsequent breast surgery were over-represented in comparison to our study population.

It was unclear for five reports53,55–58 whether participants were a consecutively treated series of patients, whereas in three studies54,59,60 patient selection was consecutive. All studies clearly described their inclusion/exclusion criteria and the intervention, and avoided disease progression bias by enrolling participants who were all at a similar point in their condition, as opposed to including patients at mixed levels of advancement in their cancer. All studies used objective outcome measures for ascertaining overall and disease-free survival, and mortality and IBTR/MCBC event rates. All studies included a median follow-up time of at least 5 years, which we considered adequate for detecting important outcome effects. We did not consider lead and length time bias to be applicable to the studies by Montgomery and colleagues55 and Yau and colleagues57 as neither study reported mortality data. The remaining studies did not adjust for lead or length time bias in their analyses.

Four studies54,55,59,60 provided information on non-respondents and dropouts (e.g. incomplete case note data, losses to follow-up, etc.) and all were judged to have avoided attrition bias, defined as bias introduced by high or differential dropout of patients. This information was judged as being unclear in the remaining studies. 53,56–58 All but one study54 identified important prognostic factors for patients’ overall survival or mortality.

Two studies59,60 undertook prospective data collection, whereas the remainder undertook retrospective data review. 53–58

Assessment of effectiveness

Two studies provided data on overall survival, cause-specific survival and the annual hazard rate of ipsilateral locoregional and new contralateral relapse in graph form. 54,55 Those remaining reports eligible for inclusion in this review reported numbers of overall deaths,53,56,58–60 deaths due to breast cancer53,56,58–60 and IBTR and/or MCBC events only. 53,56–58

Introduction

The aim of this review was to determine the diagnostic accuracy of surveillance mammography for detecting IBTR and MCBC in women who were previously treated for primary breast cancer.

• Primary objective To determine the performance of surveillance mammography, alone or in combination with other tests, in detecting IBTR and/or MCBC in women undergoing routine surveillance.

• Secondary objective To determine the performance of surveillance mammography, alone or in combination with other tests, compared with alternative tests, alone or in combination, in detecting IBTR and/or MCBC in women with a prior diagnostic test result indicating suspicion of IBTR and/or MCBC (referred to subsequently as non-routine surveillance).

Inclusion and exclusion criteria

Number and type of studies included and excluded

Nine studies met our inclusion criteria. Figure 8 shows the number of potentially relevant reports of the studies identified, the number included and excluded and a summary of the exclusion criteria. Appendix 20 lists the bibliographic details of the nine studies that were included in the review.

FIGURE 8.

Flow chart of the number of potentially relevant reports of identified studies and the number subsequently included and excluded from the diagnostic accuracy review.

The bibliographic details of the potentially relevant studies identified by the search strategy, for which full text papers were obtained, but which subsequently failed to meet the inclusion criteria, are given in Appendix 21. These studies were excluded because they failed to meet one or more of the inclusion criteria in terms of types of studies, participants, test, reference standard or outcomes (see also Figure 8).

Characteristics of the included studies

Appendix 22 provides details of the characteristics of the included studies. Table 15 provides a summary of the overall characteristics of the included studies. Table 16 provides further details of the characteristics of study design, patient type, considered index/comparator test and follow-up periods for verifying test-negative results for the individual studies arranged alphabetically by author.

Overall, the nine studies enrolled 4002 participants. After exclusions, due to eligibility or participant dropout, the studies included 3724 participants in their analyses. The earliest study took place in 199564 and the latest in 2009. 65 The earliest participant enrolment date given was 199264 and the latest was 2003. 65 Four studies did not give any indication of the enrolment time period. 66–69 One study took place in Sweden,64 two in the UK,67,68 two in Germany,69,70 two in South Korea,65,71 one in Italy66 and one in France. 72 The ages of the participants ranged from 22 to 82 years. 65 Most participants were aged in their fifties. Details of mean, with standard deviations (SDs), and median ages for individual studies are shown, where these were reported, in Appendix 22. Reported follow-up of test negatives ranged from 5 to 32 months.

The studies by Rieber and colleagues69 and Shin and colleagues71 were cohort studies, in which participants received a comparator test (MRI in the study by Rieber and colleagues69 and ultrasound in the study by Shin and colleagues71) and the reference standard. The seven remaining studies64–68,70,72 were direct head-to-head studies, in which participants all received the index test, comparator test and reference standard.

In breast cancer surveillance, the considered diagnostic tests can be used at different stages in the assessment pathway prior to a positive test result receiving verification via the reference standard test. As described earlier, test administration may be as a first routine surveillance test in a patient with no prior suspicion of IBTR/MCBC, or it may be used to evaluate a suspicious test finding on a prior diagnostic test (non-routine surveillance patients). Six studies assessed performance of the diagnostic test used as a routine first surveillance test. 64–67,70,71 The three remaining studies, by Mumtaz and colleagues,68 Rieber and colleagues69 and Ternier and colleagues,72 assessed the performance of the diagnostic test as part of non-routine surveillance to evaluate a suspicious result from a prior diagnostic test.

Three studies reported data on ultrasound. 66,71,72 Six studies reported data on MRI. 64,66–70 Four reported data on specialist-led clinical examination. 64,66,67,72 The studies did not explicitly state whether a hospital-based consultant or an alternative health-care professional conducted the clinical examination. We assumed in all cases that the examination was conducted at a consultant-supervised clinic. Drew and colleagues67 reported test performance for surveillance mammography, combined with clinical examination. Viehweg and colleagues70 reported test performance for combined surveillance mammography, clinical examination and ultrasound (known as conventional methods). This study also reported performance for combined MRI and conventional methods. Kim and colleagues65 reported performance for combined surveillance mammography and ultrasound. None of the studies meeting our inclusion criteria reported data on unstructured primary care follow-up.

The study by Boné and colleagues64 involved mastectomy patients who had all received breast reconstruction with implants. Currently, these patients are offered clinical follow-up for the ipsilateral breast and surveillance mammography for the contralateral breast annually or once every 2 years. Surveillance mammography of the ipsilateral reconstructed breast is performed if there is a clinical concern. A previous systematic review of surveillance mammography following breast reconstruction was published in 2007 by Barnsley and colleagues. 43 This review considered case reports and case series literature published between January 1980 and August 2004. The included studies involved implants and immediate or delayed transverse rectus abdominis muscle (TRAM) flap reconstructions. The authors did not conduct a meta-analysis due to heterogeneity in study design, follow-up and surveillance mammography regimen. Review findings suggested that surveillance mammography is able to detect certain local recurrences, although the authors concluded that, due to the paucity of evidence, further research in this area is required. We therefore believe that this study merits inclusion in our review, as, although these women represent a subset of our considered population, they are an increasingly relevant subgroup who might receive surveillance mammography in the future.

Quality of the included diagnostic accuracy studies

The results of the quality assessment for the individual studies are shown in Appendix 23. Figures 9 and 10 summarise the quality assessment of the included studies.

FIGURE 9.

Summary of quality assessment of included diagnostic accuracy studies for routine surveillance.

FIGURE 10.

Summary of quality assessment of included diagnostic accuracy studies for non-routine surveillance.

None of the studies met all of our quality criteria specified for higher-quality studies, although in five studies65,66,68,69,72 this was due to lack of clarity as to whether reference standard results were interpreted without knowledge of index test results only.

The study conducted by Boné and colleagues64 was considered to be unrepresentative of our considered patient population as a whole; as previously discussed, the participants had all received mastectomy for primary breast cancer with breast reconstruction and implants. As discussed earlier, it is not standard practice to offer routine surveillance of the treated breast to patients receiving either mastectomy alone, or mastectomy with breast reconstruction and implants. We therefore felt that, although this patient group represents a subset of our considered population, they differ from the wider spectrum of women who would receive surveillance in practice.

Only the Shin and colleagues71 study was judged as free of disease progression bias for positive index test results. Disease progression bias occurs when the time delay between the index and reference standard test is such that improvement or progression of the condition may occur in the intervening period. It was unclear whether the remaining eight (89%) studies64–70,72 had avoided disease progression bias for positive index test results, whereas all studies successfully avoided disease progression bias for negative index test results.

Seven (78%) studies64–66,68–70,72 (three routine surveillance and four non-routine surveillance) were free from partial verification bias. It was unclear in the study conducted by Shin and colleagues,71 however, whether test negatives received follow-up. The study by Drew and colleagues67 was considered to be vulnerable to partial verification bias, as only those participants testing positively on MRI received reference standard verification.

In all studies, positive index/comparator test results were verified by the same reference standard (histopathological assessment). In eight (89%) studies (four routine surveillance and four non-routine surveillance), participants with negative index/comparator test results all received follow-up. 64–70,72 In the study conducted by Shin and colleagues71 it was unclear whether all patients with negative test results received follow-up.

While it was unclear for all studies whether reference standard results had been interpreted without knowledge of the index/comparator test result, it was unclear in the study by Viehweg and colleagues70 whether index test results had been interpreted without knowledge of the reference standard. Five (56%) studies (three routine surveillance and two non-routine surveillance) interpreted index and comparator test results independently. It was unclear in the study by Boné and colleagues64 whether index and comparator tests were interpreted independently. In the remaining studies,65,69–72 index and comparator test results were not analysed separately. Clinical examination and mammography were usually performed before MRI or ultrasound. Knowledge of a prior test result could influence the subjective assessment, and hence the diagnostic accuracy performance, of the subsequent test(s).

In the study conducted by Rieber and colleagues69 it was unclear whether the same clinical data were available as would be the case when the test is used in practice. It was also unclear in the studies by Belli and colleagues66 and Kim and colleagues65 whether uninterpretable or intermediate results had been reported, and in the study conducted by Belli and colleagues66 whether the number of or reason for withdrawals had been explained.

Assessment of test performance

Adverse effects, acceptability of the tests, reliability, radiological/operator expertise and interpretability/readability of the tests

None of the included studies reported data concerning these outcomes.

Histology of cancers detected and not detected (true-positives and false-negatives)

The histology of cancers detected and those that were not detected (true-positives and false-negatives), by each diagnostic test, where reported, are detailed in Appendix 24 (see Tables 60–63). We found no discernible pattern for cancers detected and not detected both within and between diagnostic tests.

Methods

In both cohorts of primary tumours, Cox proportional hazards regression models74,75 were used to model four outcomes: time to IBTR, time to MCBC, time to death from all causes, and time to death from breast cancer. All risk factors were explored univariately in a simple Cox regression model and then simultaneously in a multiple Cox regression model. Risk factors modelled were age at diagnosis (≤ 34, 35–49, 50–64, 65–74, 75–79, ≥ 80 years); grade of primary tumour (grade 1, grade 2, grade 3, grade unknown); size of primary tumour (≤ 10 mm, > 10 mm to < 20 mm, ≥ 20 mm, size unknown); nodal status (no nodes involved, one to three nodes involved, four or more nodes involved, nodal status unknown); and vascular invasion (no, yes, unknown). For all risk factors the level of the factor with the best prognosis was used as the reference category, with the exception of age at diagnosis where the screening age group (50–64 years of age) was used as the reference category. If a woman was indicated as having both an IBTR and MCBC then whichever event was detected earliest defined the event for that particular woman. Any IBTR or MCBC that occurred within 6 months of diagnosis was excluded, as this might be identified as part of the management of the primary tumour and therefore would not be identifiable as part of a surveillance regimen.

Ipsilateral breast tumour recurrence (no, yes) and MCBC (no, yes) were entered into the multiple Cox regression for modelling death from all causes and death from breast cancer. HRs and 95% CIs are presented. Complete tables of the univariate Cox regression models are reported in Appendix 25, summary tables of the multiple regression models are below. Kaplan–Meier76 failure curves are presented for selected risk factors and outcomes. Incidence rates were plotted against time for both IBTR and MCBC in both primary tumour cohorts.

Results

There were 32,877 women with primary breast cancer who met the inclusion criteria for analysis (see Table 21). These cohorts were younger on average than the total population of women diagnosed with breast cancer. This is because we excluded all women who were not treated with a surgical option and these tended to be older women. The BCS cohort consisted of 90,171 years of follow-up and the women had a median follow-up of 5 years. There were 73,500 years of follow-up in the mastectomy cohort and median follow-up was 4 years.

Incidence of recurrences

Incidence rates per 1000 persons are plotted against year of follow-up in Figures 12 and 13.

FIGURE 12.

Incidence per 1000 per year of IBTR and MCBC occurrence for BCS cohort.

FIGURE 13.

Incidence per 1000 per year of IBTR and MCBC occurrence for mastectomy cohort.

Breast-conserving surgery cohort

Estimates from Cox proportional hazards regression models for time to outcome (IBTR, MCBC, death from all causes and death from breast cancer) in the BCS cohort are shown in Table 22. Tables with details of the univariate regression models are included in Appendix 25.

All risk factors were univariately associated with an increased hazard of IBTR. In particular, age ≤ 34 years, grade 3 tumour, large tumour (≥ 20 mm), nodal involvement (four or more nodes) and vascular invasion (yes) all had HRs of approximately two or above (Appendix 25, Table 64). In the multiple Cox regression model (Table 22) the estimates were broadly consistent with the univariate models. Grade 3 tumour, age ≤ 34 and nodal involvement (four or more nodes) in particular were associated with elevated risk of IBTR.

Univariately there was little that was associated with an increased risk of MCBC in the BCS group of women (Appendix 25, Table 65). Older women were at a reduced risk of MCBC. Women with a primary tumour ≥ 20 mm were at an increased risk of MCBC, HR 1.60 (95% CI 1.08 to 2.38).

The Kaplan–Meier failure curves in Figure 14 show that women who experienced IBTR were at an increased risk of death from all causes. All risk factors were associated univariately with an increased risk of all-cause death although there was no evidence that the group aged 35–49 years was different in terms of all-cause mortality to the reference group (Appendix 25, Table 66). In the multiple Cox regression model (Table 22) all of the risk factors were associated with an increased risk of death, but the youngest age group (age ≤ 34 years) was no longer different from the reference group. IBTR was a risk factor, which was associated with an increased risk of death univariately, HR 2.29 (95% CI 1.97 to 2.67). In the multiple Cox regression model, adding IBTR was still associated with an increased risk of death, HR 1.76 (95% CI 1.51 to 2.05). There was no evidence that MCBC was related to an elevated risk of death in the BCS cohort, HR 1.05 (95% CI 0.80 to 1.39) (see Figure 14).

FIGURE 14.

Kaplan–Meier failure curves for BCS cohort for time to death from all causes by disease status. BC, breast cancer.

All risk factors were associated univariately with an increased risk of death from breast cancer (Appendix 25, Table 67). In particular, the worst prognostic categories of each risk factor were associated with an increased risk of death from breast cancer. IBTR was also associated with an increased risk of death from breast cancer (Figure 15). MCBC was excluded from the models as a risk factor as there were only three cases out of 260 MCBC cases that were coded as having died from breast cancer. In the multiple Cox regression model (Table 22) there was no evidence that the age groups younger than the reference group were different in terms of risk of death from breast cancer. The remaining risk factors were associated with an increased risk of death from breast cancer and after adjusting for the other risk factors IBTR was an independent predictor, HR 2.13 (95% CI 1.78 to 2.56).

FIGURE 15.

Kaplan–Meier failure curves for BCS cohort for time to death from breast cancer by disease status. BC, breast cancer.

Mastectomy cohort

Estimates from Cox proportional hazards regression models for time to outcome (IBTR, MCBC, death from all causes and death from breast cancer) in the mastectomy cohort are shown in Table 23. Tables with details of the univariate regression models are included in Appendix 25.

Univariately the worst prognosis levels of the risk factors were associated with an increased risk of IBTR in women treated with mastectomy (Appendix 25, Table 68). In the multiple Cox regression model the increased risk was associated with grade 3 primary tumour, nodal involvement (four or more nodes) and vascular invasion (yes).

Univariately the worst levels of risk factors were associated with an increased risk of MCBC in cases treated with mastectomy (Appendix 25, Table 69). In the multiple Cox regression model, older women were at reduced risk of contralateral tumour, adjusting for other risk factors (compared to the reference screening age group). Tumour size (≥ 20 mm) and nodal involvement (four or more nodes) were associated with an increased risk of MCBC.

All risk factors were associated univariately with increased risk of all-cause death, although there was evidence that risk of death from all causes in the 35- to 49-year age group of women was slightly reduced (Appendix 25, Table 70). In the multiple Cox regression model all the risk factors were associated with an increased risk of death, but the youngest age group (≤ 34 years) had a similar risk of all-cause death as that of the reference age group. IBTR was a risk factor for death from all causes in a univariate analysis, HR 2.14 (95% CI 1.86 to 2.47), but MCBC was not, HR 1.01 (95% CI 0.83 to 1.24) (Figure 16). The estimates of the HRs for all risk factors from the univariate model were attenuated slightly in the multiple Cox model and IBTR was still associated with increased risk of death. However, a contralateral recurrence was not associated with an elevated risk of all-cause death in the mastectomy cohort.

FIGURE 16.

Kaplan–Meier failure curves for mastectomy cohort for time to death from all causes by disease status. BC, breast cancer.

All risk factors were associated univariately with an increased risk of breast cancer death although there was no evidence that the group aged 35–49 years were different in terms of risk of death from breast cancer compared with the reference group (Appendix 25, Table 71). IBTR was a risk factor which was associated with an increased risk of death from breast cancer, HR 2.78 (95% CI 2.37 to 3.27) (Figure 17). In the multiple Cox regression model (Table 23) all of the risk factors were associated with an increased risk of death, but in the youngest age group (age ≤ 34 years) the risk of death from breast cancer was reduced. There was no evidence that, after correcting for other risk factors, this age group is at a different risk of death from breast cancer compared with the reference group. In the multiple Cox regression model IBTR was still associated with an increased risk of death from breast cancer, HR 2.12 (95% CI 1.80 to 2.50). This indicates that an IBTR was related to an increased risk of death from breast cancer even after adjusting for all the other risk factors. MCBC was excluded from the models as a risk factor, as there were only six out of 262 MCBC cases that were coded as having died from breast cancer.

FIGURE 17.

Kaplan–Meier failure curves for mastectomy cohort for time to death from breast cancer by disease status. BC, breast cancer.

Methods

We modelled the following risk factors relating to characteristics of the primary tumour: age at diagnosis (≤ 34, 35 to 49, 50–64, 65–74, 75 to 79, ≥ 80 years), grade of primary tumour (grade 1, grade 2, grade 3, grade unknown), size of primary tumour (≤ 10 mm, > 10 mm to < 20mm, ≥ 20mm, size unknown), nodal status (no nodes involved, one to three nodes involved, four or more nodes involved, nodal status unknown), vascular invasion (no, yes, unknown), and type of surgery (BCS or mastectomy). Modelled risk factors associated with the second tumour were time to second tumour (< 60 months or ≥ 60 months) and size of second tumour (≤ 10 mm, > 10 mm to < 20 mm, ≥ 20 mm, size unknown). For all categorical risk factors the level of the factor with best prognosis was used as the reference category, with the exception of age at diagnosis where the screening age group (50–64 years of age) was used as the reference category. We undertook a multiple Cox regression model that included the risk factors associated with the primary tumour, time to the second tumour event and the size of second tumour. Outcomes for these models were time to death from all causes and time to death from breast cancer.

Results

There were 1174 women with 3870 years of follow-up in the IBTR cohort, of whom there were 613 deaths from all causes and 442 deaths from breast cancer. Median time from diagnosis of primary tumour to diagnosis of second tumour was 21 months.

Estimates from Cox regression models for time to death from all causes in the IBTR cohort are shown in Table 26. Women with a second tumour of ≥ 20 mm in maximum diameter were at an elevated risk of death compared with the reference group of ≤ 10 mm, HR 1.75 (95% CI 1.29 to 2.37). This was also evident from the Kaplan–Meier failure curves in Figure 20.

FIGURE 20.

Kaplan–Meier failure curves for IBTR cohort for time to death from all causes by size of second tumour.

Estimates from Cox regression models for time to death from breast cancer in the IBTR cohort are shown in Table 26. Women with a second tumour of ≥ 20 mm were at an elevated risk of death compared with the reference group of ≤ 10 mm, HR 1.99 (95% CI 1.37 to 2.89). This was also evident from the Kaplan–Meier failure curves in Figure 21.

FIGURE 21.

Kaplan–Meier failure curves for IBTR cohort for time to death from breast cancer by size of second tumour.

There were 975 women with 4268 years of follow-up in the MCBC tumour cohort, with 358 deaths from all causes and 23 deaths from breast cancer. Median time from diagnosis of primary tumour to diagnosis of second tumour was 52 months.

Estimates from Cox regression models for time to death from all causes in the MCBC cohort are shown in Table 26. Women with a second tumour of ≥ 20 mm in maximum diameter were at an elevated risk of death compared with the reference group of ≤ 10 mm, HR 2.14 (95% CI 1.49 to 3.06). This was also evident from the Kaplan–Meier failure curves shown in Figure 22.

FIGURE 22.

Kaplan–Meier failure curves for MCBC occurrence cohort for time to death from all causes by size of second tumour.

Estimates from Cox proportional hazards regression models for time to death from breast cancer in the MCBC cohort are not included. This was because of the relatively few deaths recorded from breast cancer (see Table 25 and Figure 23). A univariate analysis of the size of second tumour showed that women with a second tumour of ≥ 20 mm in maximum diameter were at an elevated risk of death compared with the reference group of ≤ 10 mm, HR 1.99 (95% CI 1.38 to 2.83) (see also Figure 23).

FIGURE 23.

Kaplan–Meier failure curves for MCBC occurrence cohort for time to death from breast cancer by size of second tumour.

A brief introduction to economic evaluation

The decision to use resources to provide one method of breast cancer surveillance means that the opportunity to use these resources in other desirable ways (either to provide another method of surveillance or to meet an entirely different health need) is given up. The cost of this decision is the benefits (health gains, etc.) that could have been obtained had the resources been used in another desirable way. This is the economic notion of ‘opportunity cost’. Strictly speaking, the opportunity cost of a decision to use resources in one way is equivalent to the benefits that could have been obtained had the resources been used to provide the next best alternative. Economic evaluation is a method of providing decision-makers with information about the opportunity cost of the decisions that could be made. It does this by comparing alternative courses of action in terms of both their costs and consequences. 82

An economic evaluation in this context would involve assessing the relative costs and benefits associated with alternative surveillance regimens for breast cancer. The objective of such an economic evaluation would be to provide information to assist decision-makers in the allocation of available resources so that benefits could be maximised. A cost-effectiveness plane (Figure 24) illustrates how an economic evaluation brings together information on costs and benefits. The vertical axis represents the difference in costs between surveillance regimens (e.g. mammography vs MRI). The horizontal axis represents differences in effectiveness between the two regimens.

FIGURE 24.

Relationship between the difference in costs and effects between a new (experimental) method of surveillance and an alternative (control) method. NE, north-east; NW, north-west; SE, south-east; SW, south-west.

In the north-west and south-east quadrants of Figure 24 a clear decision about which surveillance regimen should be preferred is provided because one or other regimen is less costly but more effective (i.e. it dominates the other treatment). In the north-west quadrant the experimental regimen is more costly and provides less benefit, therefore the control regimen is more efficient (is dominant). In the south-east quadrant the opposite situation occurs and the experimental regimen is more efficient (is dominant), as it is less costly and provides more benefit. The circle in the centre of the figure represents the possibility that no meaningful differences in costs or benefits exist between the regimens and for practical purposes the two regimens are equally efficient.

In the two remaining areas of the figure, the north-east and south-west quadrants, a judgement is required as to whether the more effective regimen is worth the extra cost. To aid these judgements, information can be provided in terms of an incremental cost-effectiveness ratio (ICER). The higher the ICER of one intervention compared with another, the less likely it is that this intervention will be considered efficient.

Economic modelling of alternative surveillance regimens

A surveillance programme needs to be not only effective, but also cost-effective. Using Markov modelling methods, the cost-effectiveness of various surveillance programmes is compared. The economic model describes the pathway of care of individuals from the point where they received treatment for breast cancer and will receive some form of ongoing surveillance. This includes their longer-term (ideally their lifetime) costs and consequences, including those that might arise from any subsequent cancers. Surveillance can be considered as an event undertaken at discrete intervals and repeated over time and hence a Markov model was developed. This can be used to describe the logical and temporal sequence of events following the implementation of alternative surveillance regimens. We used the model to provide the estimated costs and outcomes for a selected period for a cohort of women for different surveillance regimens.

The model

Markov models comprise a set of states and at any point in time an individual will be in one of these states and will stay in that state for a defined period of time (the cycle length) before they are allowed to move to another state. The cycle length must be a period relevant to the condition considered (e.g. 6 months, 1 year, 18 months, etc.) At the end of each cycle, individuals can remain in the state in which they started the cycle or move to a different state. The probabilities of moving from one state to another are called transition probabilities. In each state, the model will assign costs and benefits for each individual according to different interventions and/or time spent in each state. In a Markov model, there must be at least one absorbing state, typically death, from which the person will not be able to leave.

Figure 25 shows a simplified version of the model presented for illustrative purposes (Appendix 27 contains a copy of a section of the full model structure). In this figure, the states are presented as ovals, whereas the arrows show the possible directions in which individuals could move at the end of each cycle. The rate at which an individual moves (makes a transition) between states is governed by the transition probabilities. The states considered in the model are thought to reflect possible paths of individuals. The top line in Figure 25 represents the possible path for individuals who start off after ‘successful’ (the belief being that the woman has been successfully treated for cancer but is at risk of developing subsequent disease) treatment free from cancer but who develop breast cancer over time but remain undiagnosed. The bottom section of Figure 25 represents those individuals who start in the model after ‘successful’ treatment free of cancer, but go on to develop IBTR or MCBC over time but are identified and treated for the disease.

FIGURE 25.

Depiction of a simplified version of the Markov model. Risk profile refers to the mortality risk for a given cancer. In this simplified figure, it is assumed that cancers differ in terms of the risk of death.

If a woman initially has no evidence of IBTR or MCBC then over time she will have the chance of IBTR or MCBC occurring. The natural history of disease and the effectiveness of initial treatment determine the chance of this occurrence. Surveillance will not alter the chance of IBTR or MCBC occurring but may alter the chance of that cancer being detected, the stage at which it is found and hence the treatment and possible final outcome. Within the simplified version of the model shown in Figure 25 only three treatment states are depicted. These treatment states vary according to the risk profile of the breast cancer being treated. Once IBTR or MCBC is identified it is assumed that the cancer is treated and that subsequently individuals may have an altered life expectancy as a result of the recurrence. We also assumed that women who have had a further cancer will be judged as being at ‘moderate’ risk of developing further disease and so will have a more intensive follow-up. The absorbing state in the model is death. Any individual can move into this state from any other state within the model. The chance of moving into this state will be determined by the age of the woman through all-cause mortality and cancer-specific mortality. If a cancer is missed during surveillance then it is assumed that it will remain untreated until it is identified.

The model will compare different regimens but, for each regimen, a cohort of women will pass through the different health states. The costs per woman and speed at which they progress through the states will vary between regimens. The intuitive idea behind the model is to identify the regimen that leads to the most effective and cost-effective surveillance regimen.

The regimens considered

We outlined the alternative surveillance regimens in Chapter 3 (see Methods for the survey). The intention was to compare each of these within the economic model. We planned to combine surveillance regimens for hypothetical cohorts of the population defined in terms of the nature of primary disease, treatment and demographic characteristics, etc. These cohorts reflect the prior hypothesised risk of IBTR and MCBC in the population of women previously treated surgically for a primary breast cancer.

As described in Methods for the survey, we identified nine different surveillance regimens. We reduced these to three regimens, which we felt broadly represented the most relevant comparators. Furthermore, as reported in Chapter 5, few data on the diagnostic performance of the alternative methods of identifying a breast cancer were available. Consequently, it was not possible to model all of these options. However, some data were available to facilitate the modelling of mammographic surveillance with and without clinical follow-up organised either through secondary care or through the screening service. The presentation of the woman following referral from primary care following the identification of a suspicious lump on self-examination was also modelled. We used this form of diagnosis in two specific ways within the model. First, we used it to define a situation where no formal surveillance is used. It is also used to model the possibility that a woman presents between surveillance points with symptoms suggestive to a GP of breast cancer, for example if surveillance is performed every 36 months then within this 36-month interval the model will allow a woman to present with clinical symptoms that are suggestive of breast cancer and for this cancer to be identified.

Populating the model with parameter estimates

To provide estimates of relative cost-effectiveness, the model requires estimated values for a range of different types of parameters. Such parameter estimates should be derived in a systematic and reproducible manner to avoid bias caused by the distorted and selective use of data. 50 The assembly of such data need not necessarily be comprehensive; rather, effort should focus on identifying the most relevant data to the decision problem, which in this case was the comparison of alternative surveillance regimens for women after treatment for primary breast cancer.

We assembled the different types of data required for the economic model from analyses of existing data sets, a series of systematic reviews, and focused searches for specific pieces of data. We report the methods and results of the reviews and analyses of existing data sets in detail in Chapters 4–6. In brief, the broad types of data required to populate the economic model relate to:

• the uptake of surveillance and follow-up

• the prevalence, incidence and risk of progression of the disease, i.e. its epidemiology and natural history

• the performance of different regimens (e.g. clinical examinations, mammograms, etc.) in terms of the accuracy of the diagnostic tests

• resource use and unit costs required to estimate the costs of alternative surveillance regimens; the specific parameters and methods used to provide estimates that are relevant to the UK context

• health-state utilities.

Within the model, we based estimates of uptake upon simplifying assumptions and advice from the members of the project Advisory Group. We derived the data on the natural history of women from the analysis of the large data set reported in Chapter 6. Further data relating to the management and outcomes came from the source data used to inform recent NICE guidelines. 24

We derived information on the diagnostic performance of different types of clinical tests, for example the accuracy of mammography, from data reported in Chapter 5.

We derived data on the costs incurred for the different surveillance regimens and their consequences from structured reviews of the published literature, as well as routine data sources such as the NHS Reference Costs. 40 The perspective for costs is the NHS.

Data on the utilities associated with differing severities of cancer and the possible differences in quality of life associated with various surveillance regimens were obtained from the published literature, including the review of economic evaluations, as described above, as well as a search of the Cost-Effectiveness Analysis Registry (CEA Registry: www.cearegistry.org/).

We report how we derived each of these sets of data and the values used in the model in more detail in the sections below.

Uptake of surveillance and follow-up

Within the model, we assumed that, if individuals are invited to attend surveillance, they do in fact attend. This may be too high, as approximately 75–80% of the normal population attend for breast screening. The other variable required for the model is the probability that a woman will present to the GP with symptoms that she thinks are suspicious. Based upon advice from the clinical members of the Advisory Group we assumed that 30% of women with prior treatment for breast cancer might present to the GP per annum. We then converted this percentage into a probability of presenting per 6-month cycle by fitting an exponential curve. The probability used within the model was 0.1393, i.e. in the no surveillance arm of the model, and for during the surveillance interval in the surveillance arms of the model, just under 14% of surviving women who have not been diagnosed with a recurrent cancer will present to a GP every 6 months. The following formula assumes that events occur at a constant rate over time: p = 1–e^–rt, where p = probability, e = base of natural logarithm, r = rate and t = time period.

Epidemiology and natural history of breast cancer

Data relating to the natural history of breast cancer required for the model can be split into four components. These are:

1. recurrence/occurrence rates for women initially treated for breast cancer

2. estimated survival of women without and with IBTR or MCBC

3. estimated proportions of the different types of IBTR or MCBC occurring

4. estimated change in the severity of untreated cancer over time.

Diagnostic performance of tests

As reported in Chapter 5, relatively few data were available on the diagnostic performance of any of the tests. Within the model, we assumed that at the time a woman receives a diagnostic test as part of surveillance she is asymptomatic.

For IBTR, we based data on data reported in Chapter 5, and summarised in Table 20 (Chapter 5), and on discussions with the clinical experts involved in the study. Where relevant published data were available in the absence of pooled data the study judged to be closest to the median of reported results was used to inform the values chosen for the base-case analysis. We used data from other studies to define plausible extremes. Where it was feasible for these tests to be used then they were also used for MCBC, as few additional data were available (Table 34).

For surveillance mammography the values used in the base-case analysis were based upon those derived from Drew and colleagues. 67 We based low and high estimates of sensitivity upon the ranges for these parameters reported in Table 20 (Chapter 5). These data represent extreme values that will be used in the sensitivity analysis.

As reported in Chapter 5, only one study provided data on the sensitivity and specificity of mammography and clinical follow-up. These data did not seem plausible (e.g. the reported sensitivity was 100%). The values reported in Table 34 are assumptions derived following discussions with clinical experts. The consensus of opinion was that the combination of follow-up and mammography would slightly improve the sensitivity and specificity. In a sensitivity analysis we will explore the impact of changing these values between high and low estimates. We will also seek to identify whether there is a threshold in terms of diagnostic performance, which would make the additional cost of clinical follow-up worthwhile.

Within the model, data are also required for the diagnostic performance of a clinical examination when performed by a GP. Again, few data were available and following discussions we assumed that the rates used within the model would be slightly lower than those reported in the systematic review of diagnostic performance (Chapter 5).

We considered the impact of using a higher cost but more effective diagnostic test. As a proxy for such a test data were based upon the performance of MRI. It should be noted that the values identified, especially at the upper level, are where MRI has been used in a higher risk group of women. Hence, the values are not necessarily illustrative of MRI itself but rather of a hypothetical test. The values for the base-case analysis were based upon those reported by Drew and colleagues67 but it was assumed that the sensitivity was slightly less than perfect (i.e. 95% vs the 100% reported by Drew and colleagues). 65 Low values of sensitivity were based upon data from Warner and colleagues90 who conducted a systematic review of prospective studies in which women at very high risk for breast cancer were screened with both MRI and mammography. 90 Hence even these data may not be fully representative of women eligible for surveillance mammography. The specificity values were informed by the estimates of one study91 included in the Warner and colleagues review. 90 This study had the lowest specificity of any of the studies included in Warner and colleagues’ 2008 study. 90

Costing data

The costs of surveillance were broken down into the following cost categories:

• Cost of:

  1. – inviting women for screening

  2. – the surveillance test (e.g. mammogram, MRI, clinical examination)

  3. – health-care professional time (e.g. GP consultation, clinical examination)

  4. – further invasive tests (e.g. core biopsy)

  5. – treatment (e.g. mastectomy, radiotherapy, drug treatment).

Tables 35–37 show the cost estimates used in the economic model. All costs are reported in 2008 pounds sterling. Table 35 shows the current cost of the alternative screening strategies. The cost of inviting women to attend screening was obtained from a recent HTA report. 92 The cost of the alternative surveillance tests were all derived from routine sources. The cost of a mammogram was based on information from the NHSBSP 2009. 93 The NHSBSP estimates the cost of a mammogram in England to be £37.50 per woman invited and £45.50 per woman screened. An alternative costing source was obtained from the Scottish Breast Screening Programme, which estimates the cost of a mammogram to be £77.80. 94 The implications of the variation in costs between Scotland and England were explored in a sensitivity analysis. The cost of an MRI was estimated as being twice the cost of that reported in the NHS Reference Costs40 for an outpatient MRI. This is because an MRI on a breast takes twice as long as a normal MRI and involves the use of a contrast. The lower quartile and upper quartile of the NHS Reference Costs40 for this category are used to inform sensitivity analysis. We derived the costs of a clinical examination from routine data sources. The cost of a GP clinical examination was obtained from the Personal Social Services Research Unit (PSSRU)95 and was based on the average cost of a GP consultation. In addition, we also included the cost of a clinical examination conducted in a secondary care setting by either a consultant or non-consultant. These costs were obtained from NHS Reference Costs. 40 Information on the range of costs (lower and upper quartile) was also available and these were used as upper and lower estimates in sensitivity analysis.

The costs of further invasive tests were obtained from a NICE evidence review group (ERG) report96 and inflated to current prices using the PSSRU inflation index. The cost of a mastectomy was based on the same source, and inflated to 2008 prices. The cost of radiotherapy was based on the cost of complex treatment on a mega-voltage machine,40 assuming that women get on average 20 sessions of radiotherapy. This assumption was based on information from the PRIME trial, which reported that, on average, women receive 20 sessions of radiotherapy. 97 Again, lower quartile and upper quartile estimates of the cost of a single session of radiotherapy will be used in a sensitivity analysis.

The costs of drug treatment, for example the cost of hormone treatment, chemotherapy and combined treatment, were obtained from recent NICE guidance. The cost of hormone treatment was based on information reported in the costing template for technology appraisal guidance 112. 98 This included the costs of tamoxifen for 5 years and the cost of aromatase inhibitors (anastrozole or letrozole) for 5 years. The cost of chemotherapy was based on the costs reported in NICE technology appraisal guidance 109. 99 The cost of chemotherapy is based on the cost of two different regimens (TAC – taxotere, adriamycin and cyclophosphamide; FEC – fluorouracil, epirubicin and cyclophosphamide). This is based on six cycles of treatment.

Each risk profile consists of a series of different types of cancers (defined in terms of ER status, grade, size and number of lymph nodes involved). As described above, an average mortality for each risk profile was estimated by combining information on the expected mortality for each specific cancer within a risk profile with information on the proportion of women in that risk profile that had that specific type of cancer. Adjuvant! Online reports mortality by the type of adjuvant therapy used. The clinical members of the research team determined, based on UK practice, which specific cancer in a profile would receive hormone therapy and/or chemotherapy. Using information on the proportion of cancers in a given risk profile that would be treated with a given adjuvant therapy a proportion of a cost of a course of hormone treatment or radiotherapy was incorporated into the cost assigned to each risk profile.

Table 37 shows the costs of one surveillance regimen for a woman invited to screening and who received a clinical examination and a mammogram. The costs include the costs of screening, the mammogram and clinical examination, conducted by a consultant. On a positive mammogram, the woman would then go on to have further invasive tests to confirm the result (core biopsy). On a true-positive finding, the woman would have a mastectomy followed by radiotherapy, followed by drug treatment (depending on the severity of the IBTR or MCBC). We based the costs of treatment on a number of assumptions:

• It is assumed that all ER+ women will receive hormone treatment. It is assumed that those women who have an excellent prognosis (survival rate at 10 years of 96% or greater) and are postmenopausal will receive tamoxifen for 5 years. Women who are postmenopausal, with a poorer prognosis, will receive an aromatase inhibitor for 5 years.

• All women who are premenopausal and are ER+ will receive tamoxifen.

• All women who have grade 3 tumours will receive chemotherapy.

• Women who are ER+ and have positive lymph nodes will receive combined treatment (hormone + chemotherapy).

• Women who are ER– and have 0 nodes will receive no treatment (exception to this is that 15% might get hormone therapy).

• Women who are ER– and have positive lymph nodes will receive chemotherapy (exception to this is that 15% might receive combined therapy).

Health-state utility values

The primary purpose of the economic model was to inform decision-making in a UK setting, given that treatment for breast cancer affects not only survival, but also quality of life, for example different types and stages of cancer are likely to be associated with differences in quality of life, as would different treatment options. Therefore, we have also sought to assess the impact on quality of life, through the incorporation of health-state utility weights, which have been combined with estimates of survival to estimate QALYs.

Recent guidance suggests that estimates of QALYs should ideally be based on generic health-state valuation methods using UK population tariffs. 100 Therefore, we conducted a focused search of the literature and other relevant sources such as the Harvard cost–utility database. We identified a number of studies reporting health-state utilities. In particular, we found a recent systematic review of breast cancer utility weights. 101 In their systematic review, 59 studies were identified for review and nine studies included. Of the nine studies included, three were based on UK data. 102–104 In addition, the utility values used in the paper by Sorensen and colleagues105 were based on a combination of UK and US data. 105

It is difficult to determine how comprehensive this review is as, being available as a conference poster, the details provided on the literature searching are brief. The authors searched an appropriate selection of databases but the sensitivity of the search strategies used is unclear due to a lack of information. Missing information included whether MeSH terms were ‘exploded’ to include more specific terms, which Emtree terms were used in EMBASE, and how the terms were combined in the final search. From the information reported, one error was noted: ‘breast neoplasms’ was incorrectly described as a non-MeSH term.

Overall, the authors of the systematic review found considerable variability and inconsistency in the reported utility values. A selection of other studies eliciting health-state utilities was further identified. Overall, there was considerable variation in values and in definitions of health states; however, there is a general trend in the values reported in the literature. As would be expected, utilities decrease with increasing breast cancer severity and utilities are also found to be sensitive to treatment. For example, there is a general trend for those receiving chemotherapy to have lower utility values than those receiving hormone therapy, most likely due to the severity of the side effects of the respective treatments.

For the economic model, we have used the results reported in the systematic review of breast cancer utility weights. 101 Using this information, we defined utilities for each of the five risk profiles in the model. For example, risk profile state 1 assumes a utility state with a low value of 0.75 and a high value of 0.85 (based on the distribution of values from the systematic review). We adjusted these utility states to include a decrement for those women who will receive chemotherapy. This decrement is based on the percentage of women in each of the five severity states who would receive chemotherapy. For example, 24% of women in risk profile state 1 would receive chemotherapy. The chemotherapy decrement is based on information on patients’ utilities for cancer treatments. 106 In their study, using the time trade-off method utilities for chemotherapy were estimated to be 0.74 from an actual health state estimated to be 0.94. All health-state utilities after treatment are assumed to be the same as the utilities defined before treatment without the chemotherapy decrement.

Utility values for risk profile states 3 and 4 are based on the health-state values in Tosteson and colleagues. 107 This is based on the value for regional cancer in the age group 50–59 years. The utility value for risk profile state 5 is based on the value provided for distant rather than regional cancer in the age group 50–59 years. Each of these values has also been reduced by the decrement factor for chemotherapy. To achieve the high values reported in Table 38 for risk profile states 4 and 5 an additional 0.05 was added to the low value.

The values used in the base-case analysis are the low values reported in Table 38. Individuals in a ‘no-cancer’ state are assumed to have a health-state utility value of 0.80 in the base-case analysis.

Key assumptions of the economic model

This section provides a brief summary of the key assumptions made when developing the economic model.

Presentation of results

The base-case analysis was run for a cohort of women (starting age in the model 57 years) with surveillance occurring once yearly. The starting age was chosen as this was the mean age of the women contributing to the analysis of data from the WMCIU Breast Cancer Registry, which was reported in Chapter 6. The model was run for different starting ages in further sensitivity analysis. The cycle length of the model is 6 months and cumulative costs and benefits are estimated over a maximum of 100 cycles, which is equivalent to a time horizon of 50 years. This time horizon was taken as a proxy for life expectancy of women treated for primary breast cancer. All costs are reported in 2008 pounds sterling and effectiveness in QALYS. A discount rate of 3.5% for costs and benefits was used following guidelines for NICE. 100 Results are presented as incremental cost per QALY gained. The modelling exercise will use a net benefit framework to combine cost and benefit estimates. The results of the analyses will be presented as point estimates of mean incremental costs, effects, incremental cost per QALY. This measure is a ratio of the difference in costs divided by the difference in effectiveness between two alternative strategies. These data can be interpreted as how much society would have to pay for an extra unit of effectiveness. Whether or not a more costly but more effective regimen is considered worthwhile depends upon society’s willingness to pay for a QALY and, within England, the threshold adopted by NICE lies somewhere between £20,000 and £30,000.

Incremental cost per QALYs is a common way for presenting the results of an economic evaluation. They are, however, difficult to interpret when the choice is between several mutually exclusive options. In this circumstance the judgement can be informed by considering the net benefit statistic. The regimen with the greatest net benefit at a given value for society’s willingness to pay for a QALY is considered to be most cost-effective. The net benefit statistic itself is defined as:

where NB = net benefit, QALY_i = QALYs for intervention i, cost_i = cost for intervention i, and λ = society’s willingness to pay per QALY.

Intervention i would be chosen over intervention j when NB_i > NB_j.

Sensitivity analysis

We did not conduct probabilistic sensitivity analysis. The reason for this is that parameter values used are statistically imprecise and, as data are so limited, the model estimates may be unreliable. Therefore, the results of the economic evaluation should be interpreted cautiously and, at most, indicate situations where a particular method(s) of surveillance may be worthy of further consideration. Nevertheless, we conducted both one-way and multiway sensitivity analysis to assess how results may change as a consequence of plausible changes in parameter values. We also used deterministic sensitivity analysis to identify threshold values for key parameters. The methods used in the sensitivity analysis are described below.

Base-case results

Sensitivity analyses

We conducted a range of different sensitivity analyses, as described above in Presentation of results. As the results for the analyses for women who received BCS for their primary cancer are similar to those obtained when we consider women who received a mastectomy for their primary cancer we present sensitivity analyses solely for the scenario where women received BCS for their primary cancer. However, we also report selected analyses for a model that considers IBTR alone.

Sensitivity analysis around the breast-conserving model

Probability of developing cancer

Figures 26–29 illustrate the impact on incremental cost per QALYs as the incidence of cancer increases. In each figure, three lines are shown:

1. The incremental cost per QALY of mammography alone compared with no surveillance. This line can be used to inform the question: is it worth adopting the more effective but more costly mammography alone follow-up in place of the less costly and less effective no surveillance regimen?

2. The incremental cost per QALY of mammography plus clinical follow-up compared with mammography alone. This line can be used to inform the question: is it worth adopting the more effective but more costly mammography plus clinical follow-up in place of the less costly and less effective mammography alone regimen?

3. The incremental cost per QALY of MRI plus clinical follow-up compared with mammography plus clinical follow-up. This line can be used to inform the question: is it worth adopting the more effective but more costly MRI plus clinical follow-up in place of the less costly and less effective mammography plus clinical follow-up regimen?

FIGURE 26.

Incremental cost per QALYs for the different surveillance regimens at a 12-month surveillance interval.

FIGURE 27.

Incremental cost per QALYs for the different surveillance regimens at an 18-month surveillance interval.

FIGURE 28.

Incremental cost per QALYs for the different surveillance regimens at a 24-month surveillance interval.

FIGURE 29.

Incremental cost per QALYs for the different surveillance regimens at a 36-month surveillance interval.

The results of the analysis shown in these figures suggest that:

• At all screening intervals considered some form of active surveillance might be considered cost-effective.

• Should the incidence of IBTR and MCBC increase towards the upper values of incidence considered, which are typical of those we might expect for higher risk women (e.g. those whose primary cancers were of higher grade, who were younger than 50 years and who had lymph node involvement), a regimen of clinical follow-up and mammography is more likely to be worthwhile. Furthermore, when the surveillance interval is 24 months the incremental cost per QALY compared with mammography alone approaches £30,000. At a surveillance interval of 36 months, it is approximately £25,000.

• As the screening interval and risk of IBTR and MCBC increases towards 36 months, it becomes more likely that a more costly but more effective surveillance intervention (in this analysis typified by MRI plus clinical follow-up) might be worthwhile.