CATNON interim results: another triumph of upfront chemotherapy in glioma

According to the World Health Organization (WHO) 2016 classification of central nervous system tumors,¹ diffuse gliomas are no longer classified based on morphological features alone but also on molecular markers with strong diagnostic and prognostic value, namely, the presence or absence of isocitrate dehydrogenase (IDH1/2) mutations and 1p/19q codeletion. Non-codeleted 1p/19q anaplastic gliomas represent a subtype of tumors with worse prognosis than their 1p/19q codeleted counterparts and are thought to be less sensitive to chemotherapy.^2,3 Optimal treatment of WHO grade III gliomas and, particularly, the role of adjuvant chemotherapy after irradiation have been the subjects of much debate, leading to high variability in patient management.

Van den Bent et al⁴ recently reported the interim results of the phase III intergroup CATNON trial of radiation therapy with or without adjuvant chemotherapy for 1p/19q non-codeleted anaplastic glioma, the first phase III clinical trial in glioma patients to use molecular criteria for trial entry. Unfortunately, the trial was initiated prior to the identification of mutations of IDH1/2 as diagnostic markers of glioma and their prognostic implications⁵; therefore, the predictive value of IDH mutations will only be analyzed retrospectively. The interim results of the CATNON trial add to a growing body of literature that demonstrates the benefit of sequential use of upfront radiation therapy followed by chemotherapy for gliomas of different grades.^2,3,6,7

The CATNON trial was designed to answer 2 questions about the use of temozolomide for treatment of newly diagnosed 1p/19q non-codeleted anaplastic glioma. First, does temozolomide given during radiation therapy improve survival? Second, does adjuvant temozolomide given after radiation therapy improve survival? The trial has a factorial design, and patients were randomized equally to 4 treatment arms: radiation therapy alone, radiation therapy and concurrent temozolomide, radiation therapy followed by adjuvant temozolomide, and radiation therapy with concurrent and adjuvant temozolomide. Patients were stratified according to institution, performance status, age, 1p loss, presence of oligodendroglial elements at microscopy, and O⁶-DNA methylguanine-methyltransferase (MGMT) promoter methylation status. In total, 748 patients were randomized, and primary intention-to-treat analysis was performed.

At the planned interim analysis when 41% of the required deaths were observed (n = 219), the Independent Data Monitoring Committee recommended early release of the results from the adjuvant treatment data. With a median follow-up duration of only 27 months, the investigators observed a statistically significant benefit of the addition of adjuvant temozolomide following radiation therapy. The hazard ratio for the primary endpoint of overall survival adjusted for stratification factors in the 2 adjuvant temozolomide arms was 0.65 (99.145% CI, 0.45–0.93), mandating immediate release of the results. Moreover, with adjuvant temozolomide use, the median progression-free survival duration increased by nearly 2 years (from 19.0 to 42.8 mo), and the 5-year overall survival rate increased by 12% (from 44% to 56%). Treatment was generally well tolerated, and the toxic effects mirrored those seen in glioblastoma patients.⁶

The results of the CATNON study are remarkable. Adjuvant chemotherapy has not previously been shown to improve overall survival of patients with lower-grade (WHO grade II or III) 1p/19q non-codeleted astrocytoma. The improvements in survival in the CATNON trial were strongly associated with well-known prognostic factors, including young age, good performance status, and MGMT promoter methylation. Despite the fact that most of the patients (73%) received chemotherapy as salvage treatment in the nonadjuvant treatment arms (most of them received temozolomide), recovery of chemotherapy-related benefit did not occur, emphasizing the importance of upfront use of chemotherapy after radiation therapy. Importantly, in some series, almost 30% of patients with grade III astrocytoma could be IDH wild-type.⁵ Thus, a portion of the benefit observed in CATNON could derive from the known efficacy of adjuvant temozolomide in this population.

The impact of radiation therapy and chemotherapy on neurocognitive function and quality of life is a critical consideration in the management of patients with lower-grade glioma with good prognosis. The potential for long-term survival makes these endpoints highly relevant when recommending treatment. Longer follow-up and final analyses comparing concurrent chemoradiation and radiation therapy alone in the CATNON trial may provide more insight into late neurotoxicity of these treatments. Unfortunately, the CATNON trial incorporated the Mini-Mental State Examination (MMSE) as a measure of cognitive function. The MMSE was not designed to evaluate cognitive decline in neurologic disease and is an insensitive measure of neurocognitive function. Fortunately, this trial incorporated a more comprehensive test battery, but only in a smaller subset of patients at selected centers (results remain to be reported). This is a missed opportunity to compare the impact of concurrent chemoradiation versus radiation therapy alone on neurocognitive function. If concurrent temozolomide and radiation therapy proves to be superior to adjuvant temozolomide in the final analysis, the question will remain as to whether a gain in survival may be offset by an unacceptable increase in long-term neurotoxicity. With therapies that significantly prolong survival, understanding the impact of treatment on cognitive function and quality-of-life measures is paramount to inform treatment recommendations.

Many questions remain regarding the optimal use of adjuvant chemotherapy. A recent meta-analysis suggested that the efficacy of 6 and 12 cycles of adjuvant temozolomide was equivalent in glioblastoma,⁸ and it is unknown if 6 cycles of adjuvant temozolomide would produce equivalent outcomes for grade III gliomas. Although not a predefined analysis of the trial, MGMT promoter methylation was an independent predictor of outcome in patients given adjuvant chemotherapy, suggesting that adjuvant chemotherapy should be a priority in these patients. The optimal chemotherapy regimen for grade III gliomas is also unknown where procarbazine, lomustine (CCNU), and vincristine (PCV) has historically been prioritized over temozolomide in previous clinical trials. The value of multi-agent versus single-drug therapy, the potential risk of development of the temozolomide-induced C-to-T hypermutator phenotype and its implications,⁹ the optimal duration of adjuvant treatment, and the impact of temozolomide versus PCV on long-term toxicity are all unanswered questions.

The maturity of outcomes data and molecular subgroup analyses in this trial will be transformative to the field of neuro-oncology. The CATNON trial can be added to several recent clinical trials demonstrating the value of adjuvant chemotherapy for primary brain tumors. Although the mechanisms making sequential radiation and chemotherapy superior to radiation therapy alone at diagnosis and chemotherapy at recurrence is unknown, studies consistently demonstrate the benefit of early adjuvant chemotherapy. Retrospective integration of IDH status with 1p/19q and outcomes data from CATNON will clarify the consensus first-line standard-of-care treatment recommendations for grade III IDH-mutant, 1p/19q-intact astrocytoma.