Hypermutated recurrences: Identifying the clinical relevance

See article by Yu et al. pp 1872–1884.

Hypermutated (HM) recurrences of low-grade isocitrate dehydrogenase (IDH) mutated glioma treated with temozolomide chemotherapy have been a major concern since a pivotal report on this phenomenon in 2014.¹ More recently, preclinical and clinical work has unraveled the mechanisms of hypermutation in gliomas, showing two main pathways: A de novo pathway associated with constitutional defects in DNA polymerase and mismatch repair (MMR) genes, and a more common posttreatment pathway, associated with acquired resistance driven by MMR defects in chemotherapy-sensitive gliomas that recur after treatment with temozolomide chemotherapy.^2,3 These findings have left clinicians and patients puzzled. If the relationship between temozolomide and a HM recurrence is indeed causal, does that mean that a standard chemotherapy regimen for gliomas is actually harmful? It has even been suggested that tumors sensitive for temozolomide should be treated with alternative treatments to avoid HM recurrences.⁴

In this issue of Neuro-Oncology, Yu et al put HM recurrences further into clinical context.⁵ The present paper together with the recent publications of Barthel et al³ and Touat et al² allow a much better understanding of the clinical relevance of HM recurrences after previous temozolomide treatment. In a retrospective study including 82 patients with (initial) IDHmt diffuse glioma who underwent second surgery, Yu et al confirm that the HM phenotype is highly prevalent after treatment with temozolomide. No less than 57% of patients in their study were diagnosed with a HM recurrence. The authors address the important clinical question whether patients at risk for the development of HM recurrences could be identified. A previous report from this group showed that the MGMT promotor methylation level is a predictor of HM recurrence,⁴ the current report suggests that tumor volume is associated with shorter time to a higher grade transformation and overall survival. That is intriguing. Assuming linear tumor growth, one may hypothesize that larger tumors were present for a longer period perhaps putting them at a higher risk for more genetic alterations. In contrary to previous reports,² they did not find a dose-dependent effect of temozolomide on the occurrence of HM recurrences. They also report that HM recurrences are associated with distant recurrences and high-grade transformation, and are associated with shorter survival after recurrence. However, this study found no difference in overall survival when assessed from time of diagnosis. This is completely in line with previous findings.³ New here is the observation that HM recurrences do also not occur earlier than non-HM recurrences, when measured in time from diagnosis.

How robust are the data? This is a well-performed study including a substantial number of patients with low-grade glioma, and most of its findings are in line with other reports on this topic. When assessing HM recurrences in a retrospective manner, a selection of patients requiring second surgery is however inevitable: patients that do not relapse do not undergo second surgery. Moreover, heterogeneity of pretreatment (both in modality and in timing) of the patients in this cohort hampers the interpretation of the results.

What do these data learn us? An important take home message from this and other reports is the high percentage of HM recurrences of MGMT promoter-methylated and IDH-mutated tumors after initial treatment with temozolomide. The wrong conclusion would be to assume that because of these HM recurrences, we should stay away from treating patients with temozolomide chemotherapy in the primary setting. The fear of inducing a HM recurrence and thus inducing a treatment-resistant tumor and thereby contributing to a worse outcome is unjustified. The recent data from the CATNON trial show exactly the opposite: patients receiving early treatment with temozolomide actually live longer than those that were initially treated with radiotherapy only, despite cross-over treatment at progression.⁶ A similar survival benefit was observed earlier: patients with MGMT promotor-methylated glioblastoma live longer if radiotherapy is combined with temozolomide.⁷ In the end, the duration of response and overall survival are the clinically relevant endpoint demonstrating patient benefit. It is important to emphasize that Yu et al demonstrate that overall survival from start of treatment is not affected by a potential HM recurrence at the group level, and that HM recurrences do not occur earlier than non-HM tumors.

The fear of a HM recurrence after temozolomide, which appears to occur much less frequent after nitrosourea treatment, can also not settle the ongoing debate on whether procarbazine, CCNU, and vincristine (PCV) chemotherapy is superior to temozolomide in IDHmt astrocytomas. Temozolomide has a better toxicity profile plus is easier to administer than PCV. The prematurely closed RTOG 9813 trial that compared adjuvant temozolomide to adjuvant carmustine or lomustine did not observe a difference in survival between the two treatment arms, although a formal analysis in the subgroup of IDH-mutated tumors was never presented.⁸

Should the presence of hypermutation then guide treatment decisions In the recurrent setting? Here lies an important message of the studies on this topic: HM recurrences will not respond to temozolomide but may respond to lomustine.² Although immunotherapy with PD-1 (programmed cell death protein 1) blockade is possibly effective in glioma with MMR deficiency syndrome,⁹ the value of checkpoint inhibition in temozolomide-induced hypermutation is still unknown. Hypermutation in glioblastomas is not strictly associated with an increased intratumoral T-cell response,¹⁰ indicating that hypermutation per se may not be sufficient for an effective antitumor immunity induced by checkpoint inhibitors. In a retrospective analysis including 6 patients with a confirmed HM recurrence of IDHmt glioma, there was no benefit of PD-1 blockade.² As long as the presence of hypermutation in recurrences of low-grade IDH-mutated glioma has no clear treatment consequences, obtaining tissue only for the objective to determine tumor mutational burden is to be considered experimental, in clinical trials only.

Taken together, Yu et al provide a nuanced message on the clinical implications of the HM recurrences of IDH-mutated glioma. Taking all data together, the conclusion is that it is unnecessary to fear early treatment with temozolomide: the survival benefits of adding temozolomide treatment to radiotherapy simply outweigh the clinical impact of a HM recurrence.