Biomarker tests for decision-making on chemotherapy for breast cancer: No evidence of transferability

Following a benefit assessment in 2016 and an addendum in 2018, the German Institute for Quality and Efficiency in Health Care (IQWiG) has again examined biomarker-based tests for women with primary breast cancer. These tests aim to identify patients who could omit adjuvant chemotherapy because they have a low risk of recurrence, i.e. they can assume that the cancer will not return after successful initial treatment.

In 2019, the Federal Joint Committee (G-BA) introduced the Oncotype DX test into standard care for certain women without lymph node involvement. The G-BA commissioned the Institute to search for, present and assess the current state of knowledge on a biomarker-based strategy for decision-making on adjuvant systemic chemotherapy in primary breast cancer. If the conclusion on the benefit of Oncotype DX could be transferred to other tests, these tests could also be introduced into standard care.

IQWiG did not find any other randomized controlled trial (RCT) relevant to this question, but identified a few prognosis and concordance studies. On the basis of these studies, the Institute does not consider a transfer of the conclusion on the benefit of Oncotype DX to other tests to be viable – especially because the tests assign different patients to the group “low risk of recurrence”.

No evidence of concordance

Several tests based on biomarkers (such as the expression profiles of different genes) are available; these tests are designed to help certain breast cancer patients decide for or against adjuvant chemotherapy. However, so far only for Oncotype DX is an RCT available that shows a hint of a benefit. This conclusion could be transferred to other tests if they were concordant with Oncotype DX, that is, if they assigned a low, medium or high risk of recurrence to approximately the same women.

This concordance of risk classifications was examined in 7 studies. However, none of the studies applied the same test thresholds as in the RCT on Oncotype DX, which makes it difficult to assess the transferability of the conclusion on the benefit of Oncotype DX. Furthermore, no distinction was made between patients over and under 50 years of age or between post- and premenopausal patients, which would also have been useful in the evaluation.

The agreement of the test results was only between 43 and 74%, which means that they varied greatly in the risk assessment of the tested women. IQWiG’s Director Jürgen Windeler concluded: “The tests therefore assign different patients to the risk groups. This can only mean that they overlook a number of women who could omit chemotherapy without relevantly increasing their risk of recurrence – and in return suggest that many other women could omit chemotherapy, even though it can by no means be ruled out that the cancer will return.”

Prognoses comparable, informative value limited

IQWiG also considered 12 prospectively planned cohort studies with an observation period of at least 5 years. However, in 7 of these studies it is not ensured that the absence of tumour samples was caused by chance. The possibility of a systematic, i.e. disease-related, selection reduces the certainty of results of these studies.

The mortality of low-risk groups after omission of chemotherapy was examined in 4 studies. Similar rates were shown for women without lymph node involvement after Oncotype DX (maximum 7 to 14%) and after 3 other tests (11 to 13%). At 5 to 10% (Oncotype DX) and 6 to 10% (5 other tests), the risk of metastasis (risk of distant recurrence) after omission of chemotherapy, which was investigated in 10 studies, was also comparable.

However, the proportions of patients without lymph node involvement who were assigned to the low-risk groups differed markedly in these prognosis studies. Daniel Fleer from the Non-Drug Interventions Department, who was responsible for the rapid report, notes: “The spectrum ranges from 20 to 86% of women. Together with the sometimes low certainty of results in these studies, this also questions the transferability of a conclusion on benefit from one of these tests to the others.”

Process of report production

The Federal Joint Committee (G-BA) commissioned IQWiG to prepare the report in an accelerated procedure as a so-called rapid report. Interim products were therefore not published or made available for a hearing. This rapid report was sent to the contracting agency, the G-BA, on 18 February 2020.