Stage I Grade 3 Endometrial Carcinoma: A Close Look at Molecular Subtyping
—These investigators compared criteria from the two commonly used pathologic risk classification systems for endometrial carcinoma, assessed their concordance with molecular subtypes, and evaluated associations with patient outcomes.
Integrating molecular subtyping with established clinicopathologic factors may help clinicians identify those patients with stage I grade 3 endometrioid endometrial carcinoma (EEC) most likely to benefit from adjuvant treatment, according to results of a study conducted at Memorial Sloan Kettering Cancer Center (MSK) in New York.
Noting the current debate over the role of additional therapy in this patient group, the MSK team sought to compare criteria from the two commonly used pathologic risk classification systems, assess their concordance with molecular subtypes, and evaluate associations with patient outcomes.
Some findings were expected, others not so much.
“Our hypothesis was that molecular subtyping would be prognostic in this group of stage I high-grade endometrioid endometrial cancer, and that is what we found,” said William A. Zammarrelli III, MD, lead investigator.
“But we were surprised to find that there was minimal agreement between the 2 established and commonly used pathologic risk classification systems — the Post Operative Radiation Therapy in Endometrial Carcinoma 1 (PORTEC-1) and Gynecologic Oncology Group (GOG-99) classification systems,” added Dr. Zammarrelli,* gynecologic oncology fellow at MSK.
Clinicopathologic and molecular stratification
The investigators identified 75 patients with stage I grade 3 EECs who had undergone surgical staging at MSK: 75% (n=56) had stage IA ECC, 25% (n=19) had stage IB.
At the time of diagnosis, patients in the cohort ranged in age from 39 years to 92 years, with a median age of 61 years. Body mass index (BMI) range was from 18 kg/m2 to 53 kg/m2 (median 28 kg/m2). The cohort was 71% White (n=53), 15% Black (n=11), and 15% (n=11) belonged to other races.
In the 75-patient cohort, 69% (n=52) met criteria for high intermediate risk (HIR) according to PORTEC-1 classification (PORTEC-1 HIR), and 51% (n=38) met high intermediate risk criteria according to GOG-99 classification (GOG-99 HIR). Molecular subtype analysis classified 32% (n=24) as polymerase epsilon ultra- mutated EEC (POLE), 35% (n=26) as microsatellite instability (MSI), 20% (n=15) as copy number-high (CN-H), and 13% (n=10) as copy number-low (CN-L).
PORTEC-1 and GOG-99: discordant and fail to identify high risk
The investigators found minimal agreement between the 2 commonly used clinicopathologic classification systems, PORTEC-1 and GOG-99, in their identification of high intermediate risk patients with stage I, grade 3 EEC; about half of the HIR classifications were not concordant: 68% agreement proportion (95% CI 56.2 to 78.3), kappa value 0.36 (P=.001).
Neither of the 2 classification systems (PORTEC-1 or GOG-99) were able to differentiate the high intermediate risk patients with stage I grade 3 EEC.
For example, progression free survival (PFS) was not significantly different for those with stage I grade 3 EEC classified as HIR compared to non-HIR by either PORTEC-1 (HR:10.9, 95% CI: 0.28 to 4.21) or GOG-99 (HR: 1.22, 95% CI: 0.32 to 4.6).
“In other words,” the investigators write in their paper, “neither of these classification systems was able to identify patients at higher risk for recurrence.”
Further, neither PORTEC-1 HIR nor GOG-99 HIR was concordant with the CN-H molecular subtype, typically associated with the poorest prognosis. Lack of agreement existed between HIR in PORTEC-1 (kappa 0.03, P=.39) or GOG-99 (kappa -0.03, P =.601) and CN-H versus other subtypes.
Molecular subtyping predicts prognosis
In sharp contrast to these findings, the investigators reported that molecular subtyping did predict prognosis.
Statistically significantly worse PFS (HR, 5.67; 95% CI, 1.73 to 18.63) was seen in patients with stage 1 grade 3 EEC tumors of the CN-H subtype compared with those with tumors of the lower-risk molecular subtypes (P=.001).
Three-year PFS was as follows in univariate analysis by molecular subtype:
- CN-H: 60% (95% CI, 31.8 to 79.7)
- CN-L: 80% (95% CI, 40.9 to 94.6)
- MSI: 92% (95% CI, 71.6 to 97.9)
- POLE: 95.8% (95% CI, 73.9 to 99.4)
Overall survival (OS) was also significantly worse for those with CN-H compared with non-CN-H tumors (HR, 5.05; 95% CI, 1.13 to 22.5).
The investigators call the OS findings ‘exploratory’ and in need of further study because of the low number of deaths in the cohort (n=7).
Clinical implications
What do these findings tell clinicians dealing with the lack of clarity regarding which patients with stage I grade 3 EEC are most likely to benefit from adjuvant treatment?
“Our findings demonstrating significantly worse progression free survival for patients with CN-H tumors suggest that molecular subtyping may provide the precision needed to help clinicians identify these patients,” said Dr. Zammarrelli, adding, “We hope that our work and additional studies will shed light on how to best use molecular subtyping in selecting patients at risk of recurrence who will benefit from novel therapeutics and adjuvant therapy.”
That said, in spite of the failure of PORTEC 1 and GOG-99 to differentiate high intermediate risk patients with stage 1, grade 3 ECC in their study, Dr. Zammarrelli still considers these classification systems useful.
“Because we have well established data describing that histologic factors impact prognosis we feel that these findings should complement but not replace current standards of care,” he said.
Stay tuned
While Dr. Zammarrelli notes that next generation sequencing-based tumor profiling is not uniformly performed nationally in the up front setting for endometrial cancer due to cost and availability, he seems to think there may be changes on the horizon.
“I would recommend that clinician readers stay tuned regarding future studies and guidelines to incorporate molecular subtyping as standard practice in the upfront setting for endometrial cancer. This is being adopted by some medical societies, particularly in Europe.”
Noting that, “Clinical trials and further studies are warranted to define exactly which uterine cancer patients should undergo next generation sequencing and at what time point in their treatment,” Dr. Zammarrelli points out that, “The PORTEC-4A trial in Europe is integrating molecular subtyping into their treatment paradigm,” and he believes molecular subtyping should be incorporated into future clinical trials for endometrial cancer.
*Dr. Zammarrelli’s comments were reviewed by co-investigators Jennifer Mueller, MD, Britta Weigelt, MD, and Nadeem Abu-Rustum, MD.
Published:
References
.jpg)
