NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Fleeman N, Saborido CM, Payne K, et al. The Clinical Effectiveness and Cost-Effectiveness of Genotyping for CYP2D6 for the Management of Women with Breast Cancer Treated with Tamoxifen: A Systematic Review. Southampton (UK): NIHR Journals Library; 2011 Sep. (Health Technology Assessment, No. 15.33.)

Cover of The Clinical Effectiveness and Cost-Effectiveness of Genotyping for CYP2D6 for the Management of Women with Breast Cancer Treated with Tamoxifen: A Systematic Review

The Clinical Effectiveness and Cost-Effectiveness of Genotyping for CYP2D6 for the Management of Women with Breast Cancer Treated with Tamoxifen: A Systematic Review.

Show details


Our review aimed to answer a number of questions, namely:

  • In patients treated with TAM, do women with breast cancer identified as EMs for CYP2D6 have similar or different clinical outcomes to those identified as PMs, IMs or UMs?
  • Is there a relationship between CYP2D6 status and endoxifen concentrations?
  • Are endoxifen concentrations related to clinical outcomes?

    Do women with breast cancer who are identified as EMs for CYP2D6 have similar or different clinical outcomes with TAM compared with aromatase inhibitors?

  • What is the relative cost-effectiveness of CYP2D6 testing as a management option for women with breast cancer?

This is a relatively new area of research that is evolving rapidly and, although international consortia are collaborating, the data are limited and conflicting, which limited the ability of the review to answer the questions above.

Six individual cohorts83,86,91,96,108,109 suggest that EMs may have different outcomes in terms of relapse/recurrence (but not OS) to PMs and PMs + IMs. However, this evidence is far from conclusive, based on typically small numbers of patients and heterogeneous patient populations, clinical outcomes and alleles tested, not to mention differences in how both phenotypes and clinical outcomes are defined. In addition, three other cohorts have failed to find an association. There also appears to be evidence of a link between endoxifen concentrations and CYP2D6 status, from even smaller cohorts than have been used to assess efficacy, but this seems to suggest a contrary finding – that PMs are different to EMs + IMs − not supported from the evidence to date in terms of clinical outcomes. Unfortunately, we found no studies that measured the association between endoxifen levels and clinical outcomes, and we found no studies that directly assess whether patients who are identified as EMs for CYP2D6 have similar or different clinical outcomes with TAM when compared with aromatase inhibitors. It might be inferred, however, that if patients who are EMs have better outcomes than PMs and if the magnitude of this difference was similar to that identified in studies comparing aromatase inhibitors with TAM, then EMs might be suitable candidates for TAM, whereas aromatase inhibitors might be more suitable for the PMs. However, as just stated, there is no convincing evidence to support this. Finally, given the lack of data available to address the previous clinical questions and given additional uncertainties surrounding costs and pathways of care, it has been impossible to assess the cost-effectiveness of testing for CYP2D6.

Thus, our review has raised more questions than answers, the most pertinent question being ‘what alleles would one test for in clinical practice?’. The evidence base around CYP2D6 testing to date is at too early a stage of development to be able to ascertain which alleles should be genotyped for and how phenotypes should then be derived. In the absence of any clinical utility studies, there are also too many uncertainties about expected future pathways of care, assuming that a CYP2D6 test were to be conducted. Thus, it is impossible to recommend routine CYP2D6 testing in clinical practice based on the evidence so far available.

Implications for service provision

Owing to a lack of relevant data, it has not been possible to ascertain whether testing for CYP2D6 is clinically effective or cost-effective. In particular, it is unclear which alleles would need to be tested and therefore which test, if any, should be used. Consequently, it is not possible to recommend CYP2D6 testing for routine clinical practice.

Suggested research priorities

There are many areas in which there is a need for further data and thus we have identified the following as research questions to be addressed:

  • How many and what type of alleles should be tested for? It is important that studies are able to determine the alleles which appear to be related to clinical outcomes and which would need to be included in any CYP2D6 test. To achieve this, studies need to include adequate numbers of patients, or at least samples that can be genotyped using techniques that can test for a number of different alleles. To date, there is also no evidence surrounding patients with multiple copies of an allele, i.e. UMs, and so testing to identify UMs may also be prudent.
  • What are acceptable levels of sensitivity and specificity for these tests when measuring different outcomes? To date we have found no literature assessing the sensitivity and specificity of CYP2D6 testing, and this is a matter that needs addressing. Although it is accepted that in the absence of a number of standard test alternatives it is difficult to assess sensitivity and specificity, given the high analytical validity of genotype tests these values may be calculated for testing for specific alleles. Acceptable values for sensitivity and specificity should also be agreed a priori.
  • How should phenotypes for the metabolism of TAM be defined? This is another research question that can be adequately addressed only once the relevant alleles have been determined. It may well be that there is no need to define phenotypes and it is enough to use only genotypes, although such classifications would arguably be of greater utility for patients as well as medical professionals who need to interpret tests.
  • What are the important health outcomes for women with breast cancer, and how should these be defined? The most unambiguous outcome to define is OS, but to date there has been no evidence of any difference in this outcome between genotypes or phenotypes, or indeed between TAM and aromatase inhibitors. This is no reason to dismiss it as an important outcome but, arguably, outcomes that measure relapse/recurrence, such as DFS, are more important for a condition such as breast cancer. While DFS is an outcome measure used in many studies, unfortunately, it is not a standardised outcome measure. For example, some definitions include death from any cause, whereas others include only breast cancer mortality. This is not a problem unique to pharmacogenetic studies, however, but occurs in all breast cancer studies.
  • Do different ethnic populations need to be tested using different tests? This is a research question that can really be addressed only once the relevant alleles have been determined. Initially, it might be useful to use standardised tests across different populations, with the possibility of refining the alleles tested for in different ethnic populations once these alleles have been established. The need to carry out such tests would largely be driven by costs of resources (materials, time, etc.) and, ultimately, it may be safer, simpler and no less cost-effective to use the same test in all populations.
  • What pathways of care would a patient follow if pharmacogenetic testing were to be introduced? Would testing be required for women of any menopausal status? To answer these questions, we need more evidence of differences (if they exist) in outcomes by CYP2D6 status in premenopausal women and a better understanding on current pathways of care. While NICE currently does have recommendations for the use of TAM or aromatase inhibitors based largely on risk of disease recurrence, data on the numbers of women using these are currently lacking; an estimate of the current use of TAM and aromatase inhibitors in the UK is required as the basis for calculating resource use data for an economic model.
  • If pharmacogenetic testing were to be introduced, what would be the uptake of pharmacogenetic testing and would uptake be driven mainly by clinicians or patients? A survey of clinicians' intentions may be informative. Evidence from other areas in which pharmacogenetic testing has been introduced may also be useful.

Ideally, clinical studies will constitute companion studies to previously conducted RCTs of TAM. We are aware of such studies being undertaken by the ITPC, as well as of previously conducted trials of aromatase inhibitors versus TAM, and the results from these analyses are eagerly awaited.

Finally, TAM metabolism is complex and CYP2D6 does not appear to account for all variability in endoxifen levels. Studies examining the link between endoxifen levels and clinical outcomes are also needed, as are studies that examine polymorphisms in other TAM metabolic pathway enzymes.

© 2011, Crown Copyright.

Included under terms of UK Non-commercial Government License.

Bookshelf ID: NBK99687


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (988K)
  • Disable Glossary Links

Other titles in this collection

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...