KMT2D/MLL2 inactivation is associated with recurrence in adult-type granulosa cell tumors of the ovary

Adult-type granulosa cell tumors of the ovary (aGCTs) are rare gynecologic malignancies that exhibit a high frequency of somatic FOXL2 c.C402G (p.Cys134Trp) mutation. Treatment of relapsed aGCT remains a significant clinical challenge. Here we show, using whole-exome and cancer gene panel sequencing of 79 aGCTs from two independent cohorts, that truncating mutation of the histone lysine methyltransferase gene KMT2D (also known as MLL2) is a recurrent somatic event in aGCT. Mono-allelic KMT2D-truncating mutations are more frequent in recurrent (10/44, 23%) compared with primary (1/35, 3%) aGCTs (p = 0.02, two-sided Fisher’s exact test). IHC detects additional non-KMT2D-mutated aGCTs with loss of nuclear KMT2D expression, suggesting that non-genetic KMT2D inactivation may occur in this tumor type. These findings identify KMT2D inactivation as a novel driver event in aGCTs and suggest that mutation of this gene may increase the risk of disease recurrence.

1. Loss of KMT2D protein expression is not significantly more common in recurrent aGCT compared to primary aGCT (1/35 vs 5/42). The conclusion that either mutation (mono-allelic inactivation) or loss of expression is more frequent in recurrent than primary aGCT is not supported by the data. They have combined tumors with either loss of expression or mutation; while the tumors with mutation are more common in the recurrent aGCT those with loss of expression are not, and combining them is not justified, in my opinion, given the very imperfect correlation between expression and mutation. 2. In their concluding comments, the authors note "an association between KMT2D inactivation and disease recurrence" and go on to comment that "the lack of complete follow up data for patients with primary included in these cohorts precludes a formal evaluation of KMT2D inactivation as an independent risk factor for recurrence". I worry that this wording could be misinterpreted by readers as indicating that either the presence of KMT2D mutation or loss of expression is of prognostic significance, when they are not. To suggest that multivariable analysis to examine independent prognostic significance is worth doing is not supported by the data, as only a single primary aGCT has a mutation and one show loss of protein expression (a tumor that lacks a mutation). With only 2/35 primary aGCT showing "inactivation" it is not realistic to talk of multivariable analysis. 3. It is unfortunate that there are no paired primary and recurrent samples in the study. Even a single such sample set could have proved informative in addessing the question of whether the mutation is detectable in the primary tumor, and if so at what frequency. To go back and do deep sequencing on multiple samples from the primary tumor in those cases where a mutation had been detected in the recurrence would have been of great interest. 4. I don't understand the last sentence of the main text: "Additional prospective studies will be needed to establish whether all instances of KMT2D inactivation in aGCT are detectable at the time of diagnosis or if such lesions can occur stochastically in a subset of tumors during post-treatment surveillance, portending eventual relapse". As noted above, I believe that the abiliity to detect mutations in tumors that have mutations at the time of recurrence will be most directly addressed retrospectively, with thoroug examination of paired primary tumor samples and I don't see how a prospective approach has merit.

Reviewer #1 (Remarks to the Author):
Hillman and colleagues assess the genomic landscape of primary and recurrent adult-type granulosa cell tumours (aGCTs) using exome and targeted sequencing approaches. They identified mutations and loss of expression of KMT2D in recurrent aGCTs, suggesting convergence on loss of KMT2D as a driver of disease recurrence. This is an interesting and well executed study, that analyses a large number of samples for a rare cancer type. A few minor comments: 1) Please discuss the findings of Pilsworth et al (Modern Pathology (2018)) who performed whole genome sequencing on aGCTs in the introduction.
The important work by Pilsworth et al describing recurrent TERT C228T promoter mutations in adult granulosa cell tumors was published contemporaneously with our initial manuscript submission, and thus was not available at that time for citation in our manuscript. We have now included a discussion of these data in our introduction, and cite the published work [see line 56].
2) In the Methods section please clarify (a) how the mutation rate was calculated for cases without a matched normal, (b) details of the TruSeq Amplicon panel such as which regions of the genes were targeted, and (c) the data availability of the targeted sequencing.
(a) The reviewer raises an important point, since the examination of whole exome tumor sequencing data without the use of a matched normal sample raises unique analytical issues. In the methods section, we discuss how a bespoke "common normal" BAM file was constructed from downsampled paired-end whole exome sequencing reads derived from the peripheral blood of 5 donors. For samples without a matched normal, we used this "common normal" to call single nucleotide variants and small insertion/deletions using the same pipeline used for the other samples. Notably, we did not detect focal or broad copy number variants encompassing KMT2D in any of the tumors containing truncating KMT2D indels. As can be seen in the table, the normalized variant allele frequency for 4 of the 5 KMT2D indels falls in the range 0.29-0.36. For variant allele frequencies in this range, the distinction between clonal and sub-clonal mutations is ambiguous and thus the manuscript does not contain a conclusive statement in this regard. We have added text to the discussion section to specifically address the question of KMT2D indel clonality [see line 93]. 4) Line 195 states the "discovery" cohort, while lines 200 and 205 describe an "exploratory" cohortplease clarify if these are the same cohort or different cohorts.
In the manuscript, the terms "discovery cohort" and "exploratory cohort" referred to the same initial set of 24 cryopreserved adult type granulosa cell tumors analyzed by whole exome sequencing. We have now standardized all references to this cohort to refer only to the "exploratory cohort", in concordance with Figure 2B.

Reviewer #2 (Remarks to the Author):
The author's demonstrate a recurrent genetic event (KMT2D truncating mutation) that it more commonly seen in recurrent aGCT, compared to primary tumors. Loss of expression of KMT2D protein (as assessed by immunohistochemistry) is also more common in recurrent aGCT, compared to primary aGCT. Loss of expression corelates imperfectly with mutation status, in that most tumors with a truncating KMT2D mutation have detectable protein, while most tumors with loss of expression do not have a mutation, when a simple positive versus negative classification of the immunostaining is used.
1. Loss of KMT2D protein expression is not significantly more common in recurrent aGCT compared to primary aGCT (1/35 vs 5/42). The conclusion that either mutation (mono-allelic inactivation) or loss of expression is more frequent in recurrent than primary aGCT is not supported by the data. They have combined tumors with either loss of expression or mutation; while the tumors with mutation are more common in the recurrent aGCT those with loss of expression are not, and combining them is not justified, in my opinion, given the very imperfect correlation between expression and mutation.
We agree with the reviewer that the relationship between KMT2D mono-allelic truncating mutation and loss of detectable KMT2D protein expression is admittedly complex. This has been elegantly demonstrated in prior studies of KMT2D mutation in diffuse large B cell lymphoma (DLBCL), published by the laboratory of Laura Pasqualucci [see Zhang et al "Disruption of KMT2D perturbs germinal center B cell development and promotes lymphomagenesis" Nat Med. 2015 Oct;21(10)]. In Figure 1D-E of this paper, the authors provide a comprehensive overview of the observed rates of KMT2D mutation and their relationship to loss of protein expression. In the case of DLBCL, loss of KMT2D protein expression was observed in approximately 11% (2/18 cases) of tumors with monoallelic KMT2D truncating mutations. The authors observed loss of KMT2D protein expression in 12% (9/77 cases) of KMT2D wild type tumors, which they hypothesized may be due to epigenetic mechanisms. We agree with the Reviewer that given the complex relationship between mono-allelic KMT2D truncating mutation and loss of KMT2D protein expression, these categories should not be combined for the purpose of data presentation or statistical tests.
We have therefore made the following changes to the manuscript in order to address the important points raised by the Reviewer: • We no longer combine the rate of mono-allelic KMT2D mutation and the rate of loss of KMT2D protein expression for the purpose of statistical comparison. We have removed reference to these statistical tests from the abstract and manuscript text. • Instead, the emphasis in the abstract and text is now placed on the statistically significant difference in the rate of mono-allelic KMT2D truncating mutation between primary and recurrent tumors. • We have modified Figure 3C in the manuscript so as to no longer combine mono-allelic KMT2D mutation and the rate of loss of KMT2D protein expression. This figure now functions merely as a concise summary of the complex relationship between mono-allelic KMT2D truncating mutation and loss of KMT2D protein expression, in a manner analogous to the presentation of similar data for DLBCL reported by Zhang, et al, as published in Nature Medicine. • We have expanded the Discussion section to explicitly address the low degree of overlap we observe between mono-allelic KMT2D truncating mutation and loss of KMT2D protein expression [see line 191].
2. In their concluding comments, the authors note "an association between KMT2D inactivation and disease recurrence" and go on to comment that "the lack of complete follow up data for patients with primary included in these cohorts precludes a formal evaluation of KMT2D inactivation as an independent risk factor for recurrence". I worry that this wording could be misinterpreted by readers as indicating that either the presence of KMT2D mutation or loss of expression is of prognostic significance, when they are not. To suggest that multivariable analysis to examine independent prognostic significance is worth doing is not supported by the data, as only a single primary aGCT has a mutation and one show loss of protein expression (a tumor that lacks a mutation). With only 2/35 primary aGCT showing "inactivation" it is not realistic to talk of multivariable analysis.
We thank the reviewer for pointing out the imprecision in the manuscript text where we discuss the possible investigation of KMT2D inactivation as a risk factor for recurrence. We agree that even with complete follow up data (which is not available for our cohort), such an analysis would not be possible given the rarity of KMT2D inactivation in primary tumors (2/35 cases). Multivariable analysis to examine the independent prognostic significance of KMT2D inactivation remains an interesting question, but this would require a much larger series of primary tumors. We have amended the text to clarify these points in a manner that is consistent with the Reviewer's comment [see line 198].
3. It is unfortunate that there are no paired primary and recurrent samples in the study. Even a single such sample set could have proved informative in addessing the question of whether the mutation is detectable in the primary tumor, and if so at what frequency. To go back and do deep sequencing on multiple samples from the primary tumor in those cases where a mutation had been detected in the recurrence would have been of great interest.
We wholeheartedly agree with the reviewer that the analysis of paired primary/recurrent samples would be of great interest. Unfortunately, we do not have paired archived tissue of appropriate quality for any adult type granulosa cell tumors. If such paired samples were available then we would absolutely have analyzed them as part of this study. This circumstance is likely due to the nature of M.D. Anderson Cancer Center as a cancer care referral center. Since the large majority of patients with adult type granulosa cell tumors are cured by surgery alone, when treated at our center these patients most often choose to undertake clinical surveillance follow-up closer to home. Conversely, patients seen for management of recurrent disease at our institution almost universally had a primary surgery performed by a local gynecologist many years prior. 4. I don't understand the last sentence of the main text: "Additional prospective studies will be needed to establish whether all instances of KMT2D inactivation in aGCT are detectable at the time of diagnosis or if such lesions can occur stochastically in a subset of tumors during post-treatment surveillance, portending eventual relapse". As noted above, I believe that the abiliity to detect mutations in tumors that have mutations at the time of recurrence will be most directly addressed retrospectively, with thoroug examination of paired primary tumor samples and I don't see how a prospective approach has merit.
The reviewer raises an important point that requires clarification in the manuscript text. This point pertains to the question of whether KMT2D inactivating mutations are always detectable at the time of diagnosis. We agree with the Reviewer that this question will likely be answered most