Short-duration adjuvant trastuzumab therapy in human epidermal growth factor receptor 2-positive breast cancers: Has its time come?

Gulia S, Kannan S, Badwe R, Gupta S. (Departments of Medical Oncology, Biostatistics and Surgical Oncology, Tata Memorial Centre, Homi Bhabha National Institute, Mumbai, India.) Evaluation of 1-year versus shorter durations of adjuvant trastuzumab among patients with early breast cancer: An individual participant data and trial-level meta-analysis. JAMA Netw Open 2020;3:e2011777.

Summary

In this meta-analysis, the authors used data from six randomized clinical trials (RCTs) in patients with early (non-metastatic) breast cancer that compared short duration (<1 year) with 1 year of trastuzumab as adjuvant treatment. In contrast to the previously published trial-level meta-analysis from the same six studies, this meta-analysis included updated data from the PHARE and PERSEPHONE studies and also extracted individual patient time-to-event data from published Kaplan–Meier survival curves, a unique and more lately reported method to do individual patient data (IPD) meta-analysis. Non-inferiority of shorter duration of trastuzumab in disease-free survival (DFS) compared with 1 year of therapy was assessed in the reconstructed IPD meta-analysis (11 376 patients with 1659 DFS events and 871 deaths) and trial-level meta-analysis (11 603 patients, 1760 DFS events and 930 deaths). The median value (1.3; range 1.15–1.53) of the non-inferiority margins used in the included RCTs was chosen as the non-inferiority margin.

In the IPD meta-analysis, 5-year DFS hazard ratio (HR) for shorter-duration versus 1 year of trastuzumab was 1.14 (95% CI 1.03–1.25 with a p value for non-inferiority=0.004), and HR for overall survival (OS) was 1.17 (95% CI 1.02–1.33). In the trial-level analysis, the HR for DFS was 1.15 (95% CI 1.04–1.26; one-sided p value for non-inferiority=0.002) and for OS, 1.17 (95% CI 1.03–1.33). The risk of congestive heart failure was significantly less with shorter-duration trastuzumab (relative risk 0.53; 95% CI 0.38–0.74). The risk of bias was low, and quality of evidence was graded high for DFS and OS, while moderate for cardiotoxicity. Given these results, the authors concluded that a shorter duration of adjuvant trastuzumab was non-inferior to its 1-year administration and resulted in lower rates of cardiac toxic effects, hence it may be the preferred option for patients with low-risk disease or those with predisposition to cardiac toxic effects.

Comment

Breast cancer is the leading cancer among women and the most common cause of cancer-related mortality among women worldwide. Compared to data from western countries, the prevalence of human epidermal growth factor receptor 2 (HER2)-positive breast cancer is relatively high (ranging from 19% to 31%) among developing countries such as India, but with a majority of patients not able to afford therapy due to financial constraints.^[1] Therefore, an optimum shorter duration of HER2-directed therapy is of utmost therapeutic relevance to this population. Hence, results of the present meta-analysis would be important to oncologists in providing evidence for best practices.

The authors stated that short-duration adjuvant trastuzumab therapy was non-inferior to its 1-year administration and resulted in lower rates of cardiac adverse effects. However, the non-inferiority limit set at 1.3 is same as in the PRESOPHONE trial, which is the only trial included in this meta-analysis that has proven the non-inferiority of shorter duration of trastuzumab and this corresponds to a 4% absolute difference in 5-year DFS. A 4% survival difference in early breast cancer (which corresponds to a hazard of 30%) can be considered too high and was also one of the major criticisms of the PRESOPHONE trial. The PHARE trial with a more reasonable non-inferiority margin of 1.15 was negative for non-inferiority.

Seven meta-analyses were published in 2019–20 comparing short-duration versus standard trastuzumab therapy, with almost all of them except one having used the same RCTs used in the present meta-analysis and all studies reported nearly equivalent HRs for OS and DFS [Table - 1].^{[2],[3],[4],[5],[6],[7]} While all the others concluded the superiority of the standard 1-year trastuzumab therapy, the present meta-analysis was the first to establish non-inferiority of the short-duration therapy. The present meta-analysis was designed to assess the non-inferiority of shorter duration of trastuzumab, an approach different from those used in the previous meta-analysis that showed the superiority of 1-year trastuzumab therapy, findings of which were also replicated in the present meta-analysis. While the authors of the present meta-analysis justified that by using a non-inferiority analysis design, they preserved the non-inferiority interpretation of its constituent trials, the authors of the previous meta-analysis stated that determining treatment superiority via a meta-analysis of non-inferiority studies is statistically acceptable. Hence, whether the result of this meta-analysis is a statistical jigsaw or a shorter trastuzumab therapy is non-inferior to 1-year therapy with regard to survival is uncertain. Moreover, given the use of a new approach to extract IPD from survival curves that have not been previously used in oncology meta-analysis and the lack of power to contrast between the two shorter-duration regimens (9–12 weeks v. 6 months), the results of this meta-analysis should be interpreted with caution and needs to be validated further before recommending a change in clinical practice.

Table 1: Published meta-analysis reports comparing short-term trastuzumab therapy with standard 1-year therapy

Associated cardiotoxicity with trastuzumab is not negligible although it is usually well tolerated with cardiac events being reversible. Short-term trastuzumab therapy is associated with significantly lower cardiotoxicity compared to standard therapy as found in both the previous and present meta-analyses. However, whether the risk of cardiotoxicity differs among anthracycline-free versus anthracycline-based regimens should be further explored.^[8] With the advent of new biomarkers such as high-sensitivity troponin-I and topoisomerase 2β isomerases, we can now risk stratify individuals with greater risk of subclinical cardiotoxicity early into the treatment period. This can help identify target individuals for shorter duration therapies.

Both the present and previous meta-analyses also failed to identify subgroups (hormone positive v. hormone negative, early breast cancer v. locally advanced breast cancer and metastatic breast cancers) for whom shorter durations might suffice. Numerous studies have also shown that trastuzumab-based neoadjuvant therapy has higher rates of pathological complete response (pCR).^[9] Thus, it is also important to identify the optimum duration of trastuzumab among those with and without pCR after neoadjuvant therapy, and evaluate if shorter duration might suffice among those with pCR while longer durations of trastuzumab may benefit those with residual disease. Given the cost of medication being a major barrier for receipt of HER2-directed therapy in low- and middle-income countries, it will be useful if the disease subgroups that benefit from short-duration trastuzumab are identified.

Conflicts of interest. None declared