Reply to Survival analysis and treatment effects in patients with endometrial cancer and <i>POLE</i> mutations

In response to the letter to the editor from Lamothe and Ramia DeCap, we would like to provide a few points of clarification. To begin, we agree with the limitations of our study1 as described by Lamothe and Ramia DeCap in their letter. We have acknowledged these limitations in the introduction and again in the discussion and interpretation of the results in our article. Cancer with pathogenic POLE mutations (POLEmut) account for approximately 10% of endometrial carcinomas. Unlike other patients with endometrial cancer, those with this ultramutated phenotype, despite presenting with unfavorable pathological features, have been described time and again as having excellent outcomes. In our article,1 we analyzed simultaneously the largest collection of POLEmut patient data available, collected from 13 studies from around the world (most of them observational). We identified 365 patient records in which a mutation in the POLE exonuclease domain was reported and in which data about the first line of treatment were available. Of those patients, 294 had pathogenic POLE mutations (POLEmut),2 and only 12 of these patients experienced either a recurrence or a death from the disease. This event rate is much lower than what would be expected in endometrial carcinomas that had similar pathological features but did not have pathogenic POLE mutations. We tried to mitigate shortcomings in the data, to the extent possible, by using a 1-stage individual patient data meta-analysis,3 accounting for between-study heterogeneity via mixed effects models, and minimizing the bias of confounding by indication via propensity scores. We provided various levels of complexity in the data analysis—descriptive (Fig. 2 and Tables 2 and 3) and univariable and multivariable (Table 4)—and described the challenges in interpreting these results in light of the small number of events and the presence of potential unmeasured confounders; indeed, “absence of evidence is not evidence of absence.” Despite these limitations, our analysis consolidated and synthesized all the retrospective evidence available to date on POLEmut tumors. Specifically, we demonstrated 1) the importance of identifying pathogenic POLE mutations versus nonpathogenic ones, 2) the almost uniformly favorable outcomes of patients with POLEmut endometrial cancers despite often unfavorable pathological characteristics, 3) an inability to demonstrate the impact of treatment on outcomes in those POLEmut endometrial cancers, and 4) a high and sustained salvage rate in patients with rare recurrence events. Our study concluded that definitive answers could come only from prospective studies designed for this purpose. We did not make claims about optimal treatment recommendations and encouraged participation in clinical trials. Results from our study are especially important in the dawn of the new World Health Organization endorsement of molecular classification for endometrial cancer4 and new European Society of Gynaecological Oncology/European Society for Radiotherapy and Oncology/European Society of Pathology and other treatment guidelines recommending risk stratification and treatments based on molecular subtype.5 We may still be many years away from definitive answers on the “best treatment” for the reasons that Lamothe and Ramia DeCap have described in their letter: this is a rare cancer subtype with excellent outcomes, resulting in a small number of events which hinders statistical power to provide level I evidence. Nonetheless, patients will continue to be treated on the basis of the best available evidence until such time as those trial results are published. We believe that our study is an important contribution to the current understanding of these cancers on the basis of the size of the cohort and the careful exclusion of those cases with nonpathogenic (passenger) mutations in POLE. No specific funding was disclosed. Talhouk and McAlpine report a patent pending on endometrial cancer classification methods.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.