When targets miss the markBMJ 2017; 359 doi: https://doi.org/10.1136/bmj.j4934 (Published 26 October 2017) Cite this as: BMJ 2017;359:j4934
All rapid responses
The BMJ published a full-page advertisement promoting the Oncotype DX diagnostic test for early breast cancer in the 28 October 2017 issue that, in our view, contains a misleading claim. The same advertisement was carried in the 19-26 August issue. The advertisement states “Predicts likely benefit of chemotherapy” meaning that the test predicts the sensitivity of individual tumours to chemotherapy. This claim is not supported by robust evidence and as such is likely to mislead its target audience.
Oncotype DX is a complex pathology test which measures the expression of 16 cancer-related genes and 5 reference genes1. It is performed on tumour tissue in a single central laboratory located in California by its vendor, Genomic Health Inc. who promote its use for the common oestrogen receptor positive and HER2 negative subtype of breast cancer. The output of the test is a numerical “Recurrence Score” which ranges from 1 to 100 and which correlates with the risk of future cancer recurrence.
Oncotype DX has been shown to provide superior prognostic information to conventional histopathological assessment in axillary lymph node negative disease2. It is therefore widely used in many health-care jurisdictions to support decision making about chemotherapy use in early breast cancer. The test was evaluated by NICE in the 2013 DG10 guideline3. This states that Oncotype DX “is recommended as an option for guiding adjuvant chemotherapy decisions for people with oestrogen receptor positive (ER+), lymph node negative (LN−) and human epidermal growth factor receptor 2 negative (HER2−) early breast cancer” who are “at intermediate risk”. The test is available and utilised throughout the UK through specialist commissioning on the basis of the NICE guidance.
The claim that the test predicts chemotherapy sensitivity is supported by 2 studies, Paik et al (2006)4 and Albain et al (2010)5, which are referenced in the advertisement. In both studies Oncotype DX testing was performed on a subset of stored tumour blocks from historic trials of chemotherapy vs no chemotherapy. All trial participants were also treated with tamoxifen, the only anti-oestrogen used in early breast cancer at the time, for 5 years. Both studies showed that patients with “high” Recurrence Score tumours (>31) benefitted disproportionally from chemotherapy compared to the entire cohort whilst there was no significant chemotherapy benefit for those whose tumours had “low” scores (<18). This result contrasts with the EBCTCG individual patient meta-analysis of 123 trials of systemic treatment involving over 100,000 women6 which concluded that the relative benefit of chemotherapy was independent of tumour characteristics. The NICE DG10 committee took the view that the EBCTCG analysis was correct.
Paik et al4 retrieved 651 tumour blocks from the NSABP B-20 trial. The parent trial enrolled 2299 women with axillary node negative disease of whom 46% were aged under 50 at enrolment and randomised participants to tamoxifen given alone or combined with either CMF or MF chemotherapy. As the B-20 trial did not show any difference in outcome between the two chemotherapy regimens, these groups were combined in the Oncotype DX study. The SWOG-8814 trial of anthracycline combination chemotherapy in post-menopausal women with axillary node positive disease included 2 chemotherapy arms; in one of these chemotherapy and tamoxifen were started simultaneously which had an inferior outcome. The Albain et al study5 therefore obtained tumour blocks from 367 of the 927 women who were treated with either tamoxifen alone or with sequential chemotherapy and tamoxifen in the parent trial.
Both studies analysed relatively small numbers of patients and consequently suffer from wide confidence intervals. This applies particularly to the Albain et al study. Both parent trials predate the routine use of adjuvant chemotherapy combined with trastuzumab for the high-risk HER2 positive subgroup and therefore included tumours that were discovered to be HER2-positive through Oncotype DX testing. Tumours of this type have a reduced sensitivity to endocrine therapy7 and are likely to have high Recurrence Scores because of the test algorithm1. In the Albain et al study the proportion was 12% which is consistent with the results of community HER2 testing8; the proportion is not disclosed in the Paik et al study. Some tumours from the NSABP B-20 tamoxifen only arm were included in the original discovery and validation studies for Oncotype DX2 which introduces a potential for bias.
The Paik et al and Albain et al studies were reviewed by Ward et al2, who provided the Diagnostics Assessment Report for the NICE DG10 guidelines committee and were also considered in detail by the NICE committee which concluded that the claim that Oncotype DX was predictive of chemotherapy sensitivity was not robust and should be the subject of further research. No further research results are currently available.
The advertisement for Oncotype DX carried by the BMJ therefore presents a claim that is controversial and not adequately supported by the available evidence. The level of regulation for pathology tests is less than for pharmaceuticals so it is possible for manufacturers to make misleading claims with relative impunity. The BMJ should not continue to carry this particular advertisement.
Robert C Stein
Consultant Oncologist and Professor of Breast Oncology, NIHR University College Hospitals Biomedical Research Centre, London and Research Department of Oncology, UCL, London
Iain R Macpherson
Clinical Senior Lecturer in Medical Oncology, University of Glasgow and Beatson West of Scotland Cancer Centre, Glasgow
1. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N Engl J Med 2004;351(27):2817-26. doi: 10.1056/NEJMoa041588 [published Online First: 2004/12/14]
2. Ward S, Scope A, Rafia R, et al. Gene expression profiling and expanded immunohistochemistry tests to guide the use of adjuvant chemotherapy in breast cancer management: a systematic review and cost-effectiveness analysis. Health technology assessment 2013;17(44):1-302. doi: 10.3310/hta17440 [published Online First: 2013/10/04]
3. National Institute for Health and Care Excellence. Gene expression profiling and expanded immunohistochemistry tests for guiding adjuvant chemotherapy decisions in early breast cancer management: MammaPrint, Oncotype DX, IHC4 and Mammostrat [DG10]. London: National Institute for Health and Care Excellence., 2013 http://guidance.nice.org.uk/DG10.
4. Paik S, Tang G, Shak S, et al. Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer. J Clin Oncol 2006;24(23):3726-34.
5. Albain KS, Barlow WE, Shak S, et al. Prognostic and predictive value of the 21-gene recurrence score assay in postmenopausal women with node-positive, oestrogen-receptor-positive breast cancer on chemotherapy: a retrospective analysis of a randomised trial. Lancet Oncol 2010;11(1):55-65. doi: 10.1016/S1470-2045(09)70314-6 [published Online First: 2009/12/17]
6. Early Breast Cancer Trialists' Collaborative Group. Comparisons between different polychemotherapy regimens for early breast cancer: meta-analyses of long-term outcome among 100,000 women in 123 randomised trials. Lancet 2012;379(9814):432-44. doi: 10.1016/S0140-6736(11)61625-5 [published Online First: 2011/12/14]
7. Tovey S, Dunne B, Witton CJ, et al. Can molecular markers predict when to implement treatment with aromatase inhibitors in invasive breast cancer? Clinical cancer research : an official journal of the American Association for Cancer Research 2005;11(13):4835-42. doi: 10.1158/1078-0432.CCR-05-0196 [published Online First: 2005/07/08]
8. Cronin KA, Harlan LC, Dodd KW, et al. Population-based estimate of the prevalence of HER-2 positive breast cancer tumors for early stage patients in the US. Cancer Invest 2010;28(9):963-8. doi: 10.3109/07357907.2010.496759 [published Online First: 2010/08/10]
Competing interests: RCS is the Chief Investigator of the OPTIMA trial, which seeks to determine whether multi-parameter tumour gene expression assays are predictive of chemotherapy sensitivity. IRM is a member of the OPTIMA Trial Management Group and has undertaken paid consultancy work for Genomic Health Inc.