Methylthioadenosine Phosphorylase Genomic Loss in Advanced Gastrointestinal Cancers

Abstract Background One of the most common sporadic homozygous deletions in cancers is 9p21 loss, which includes the genes methylthioadenosine phosphorylase (MTAP), CDKN2A, and CDKN2B, and has been correlated with worsened outcomes and immunotherapy resistance. MTAP-loss is a developing drug target through synthetic lethality with MAT2A and PMRT5 inhibitors. The purpose of this study is to investigate the prevalence and genomic landscape of MTAP-loss in advanced gastrointestinal (GI) tumors and investigate its role as a prognostic biomarker. Materials and Methods We performed next-generation sequencing and comparative genomic and clinical analysis on an extensive cohort of 64 860 tumors comprising 5 GI cancers. We compared the clinical outcomes of patients with GI cancer harboring MTAP-loss and MTAP-intact tumors in a retrospective study. Results The prevalence of MTAP-loss in GI cancers is 8.30%. MTAP-loss was most prevalent in pancreatic ductal adenocarcinoma (PDAC) at 21.7% and least in colorectal carcinoma (CRC) at 1.1%. MTAP-loss tumors were more prevalent in East Asian patients with PDAC (4.4% vs 3.2%, P = .005) or intrahepatic cholangiocarcinoma (IHCC; 6.4% vs 4.3%, P = .036). Significant differences in the prevalence of potentially targetable genomic alterations (ATM, BRAF, BRCA2, ERBB2, IDH1, PIK3CA, and PTEN) were observed in MTAP-loss tumors and varied according to tumor type. MTAP-loss PDAC, IHCC, and CRC had a lower prevalence of microsatellite instability or elevated tumor mutational burden. Positive PD-L1 tumor cell expression was less frequent among MTAP-loss versus MTAP-intact IHCC tumors (23.2% vs 31.2%, P = .017). Conclusion In GI cancers, MTAP-loss occurs as part of 9p21 loss and has an overall prevalence of 8%. MTAP-loss occurs in 22% of PDAC, 15% of IHCC, 8.7% of gastroesophageal adenocarcinoma, 2.4% of hepatocellular carcinoma, and 1.1% of CRC and is not mutually exclusive with other targetable mutations.


Introduction
][6] Homozygous deletion of tumor suppressor genes plays a key role in oncogenesis, and homozygous 9p21-loss (henceforth 9p21-loss) has been demonstrated to be an early evolutionary event in oncogenesis. 7,8The methylthioadenosine phosphorylase (MTAP) gene is immediately adjacent to CDKN2A within 9p21 and also frequently homozygously codeleted with CDKN2A and CDKN2B during a chromosomal interstitial deletion event-henceforth referred to as MTAP-loss. 9Fig. 1A).From a pan-cancer study of data from The Cancer Genome Atlas, homozygous deletion of MTAP was described in 9.3% of cancers, with loss of heterozygosity at 9p21 due to hemizygous deletion of MTAP observed in another 27.8% of cancers. 10Despite the relatively high prevalence of 9p21/MTAP-loss, investigation in the field has been limited by the challenges associated with targeting loss-of-function mutations.
2][13] MTAP is a rate-limiting enzyme controlling the final step of the methionine salvage pathway, which replenishes intracellular adenine and methionine pools, and plays a crucial role in rapidly proliferating and metabolically stressed tumor cells. 14(Fig. 1B) MTAP-loss causes an accumulation of methylthioadenosine (MTA), which has been associated with aggressive cancer phenotypes. 15,16Although the mechanism of action leading to synthetic lethality between MTAP-loss and MAT2A or PRMT5 inhibitors is yet to be definitively determined, early evidence suggests it could be due to increases in alternative splicing and polyadenylation. 17,18Importantly, proof-of-concept establishing that MTAP-loss can be targeted with PRMT5 inhibitors has been demonstrated in early phase clinical trials with a promising clinical efficacy signal. 19he potential role and impact of MTAP-loss as a therapeutic target in advanced gastrointestinal (GI) cancers have not yet been directly interrogated.Publications examining the genomic landscape and clinical impact of MTAP-loss The Oncologist, 2024, Vol. 29, No. 6

495
are currently limited to the pan-cancer setting, with limited individual sample sizes in GI cancers.In a pan-cancer setting, MTAP-loss has been associated with shorter overall survival (OS). 10Importantly, 9p21/MTAP-loss is also an emerging predictive biomarker of an immunogenically "cold" tumor microenvironment and associated with distinct shifts in intra-tumoral immune cell abundance, reduced T-cell receptor repertoire diversity, reduced PD-L1 positivity, and changes in immunomodulatory gene expression. 10,20Across clinical cohorts treated with anti-PD(L)1 immunotherapy and chiefly comprised of patients with melanoma, lung, and urothelial cancer, significantly reduced progression-free (PFS) and disease-specific survival were observed for patients with tumors harboring 9p21-loss versus 9p21-wildtype. 10In advanced GI cancers though, both the prevalence, genomic landscape, and clinical characteristics of MTAP-loss have not been previously examined.The purpose of this study is to describe the prevalence of MTAP-loss in GI cancers and examine differences in coalterations between MTAP-loss and MTAP-intact cancers.We also examined potential differences in immune biomarkers and the prognostic clinical implications of MTAP-loss in common GI cancer types.

Genomics-Only Cohort (N = 64 860)
Approval for this study was obtained from the Western Institutional Review Board (Protocol No. 20152817).For this cross-sectional study, betweenJanuary 1, 2018 and July 15, 2022, patients with any of 5 histologically defined GI cancer types (pancreatic ductal adenocarcinoma [PDAC], intrahepatic cholangiocarcinoma [IHCC], hepatocellular carcinoma [HCC], colorectal carcinoma [CRC], and gastroesophageal adenocarcinoma [GEAC]), who had previously undergone comprehensive genomic profiling at a Clinical Laboratory Improvement Amendments (CLIA)-certified and College of American Pathologists (CAP)-accredited reference molecular laboratory (Foundation Medicine, Inc.) as part of their clinical care, were selected.Tumor profiles were identified based on the histological subtype submitted by physicians and validated by central pathology review.Clinicopathological data including patient age and gender, routine histology and immunohistochemical staining results, and confirmation of the diagnosis were extracted from medical records and pathology reports.

Clinical Outcomes Cohort (N = 102)
To determine the clinical significance of homozygous MTAPloss among patients with GI cancer treated at The University of Texas MD Anderson Cancer Center (MDACC), we identified 102 patients diagnosed with advanced PDAC and IHCC between January 11, 2018 and August 2, 2022 and whose tumors had defined MTAP status (-loss or -intact), determined by copy number status via next-generation sequencing (NGS).Patients in MTAP-intact and homozygous-loss groups were matched by age, gender, and ethnicity.Clinical information, including demographic information, treatment history, and response to treatment, were retrieved from a retrospective medical record review.PFS was defined as the time of initiation of any line of systemic therapy to the date of clinical or radiological disease progression/treatment discontinuation on that line of therapy, as determined by the treating physician.OS was defined as the time of diagnosis to death from any cause.

Comprehensive Genomic Profiling
Comprehensive genomic profiling for the Genomicsonly cohort was performed on United States Food and Drug Administration (U.S. FDA)-approved hybridizationcaptured, adaptor ligation-based libraries using DNA extracted from formalin-fixed paraffin-embedded tumor.All samples forwarded for DNA extraction contained a minimum of 20% tumor cells.The samples were assayed using adaptorligation and hybrid capture NGS for all coding exons from up to 324 cancer-related genes, plus select introns from up to 31 genes frequently rearranged in cancer.Patient samples were sequenced and evaluated for genomic alterations, including base substitutions, insertions, deletions, copy number alterations (amplifications and homozygous deletions), and for select gene fusions/rearrangements, as previously described. 21The bioinformatics processes used in this study included Bayesian algorithms to detect base substitutions, local assembly algorithms to detect short insertions and deletions, a comparison with process-matched normal control samples to detect gene copy number alterations, and an analysis of chimeric read pairs to identify gene fusions as previously described. 22Unless otherwise specified, short variants (single nucleotide variants and short insertion/deletions) were included in the analysis if they were annotated as either "Known Pathogenic" or "Likely Pathogenic."MTAP-intact or loss status was determined using NGS, utilizing a copy number assessment algorithm, which is a component of this assay.All cases reported as MTAP-loss featured an MTAP copy number of zero indicating homozygous loss.

PD-L1 Immunohistochemistry
PD-L1 immunohistochemistry (IHC) was performed on subsets of tumors in the Foundation Medicine cohort using the Dako 22C3 PharmDx assay in a CLIA-and CAP-accredited reference laboratory, per manufacturer's instructions.Interpretation of PD-L1 IHC was performed by a boardcertified pathologist to determine the tumor proportion score (TPS), which is defined as the number of PD-L1 staining tumor cells with any convincing partial or complete linear membrane staining of viable tumor cells distinct from cytoplasmic staining, divided by the total number of viable tumor cells, multiplied by 100.PD-L1 "low positive" was defined as TPS scores of 1%-49%, while PD-L1 "high positive" was defined as TPS scores > 50%.

Tumor Mutational Burden
Tumor Mutational Burden (TMB) in the Genomics-only cohort was determined on 0.83-1.14megabases (Mb) of sequenced DNA using a mutation burden estimation algorithm that, based on the genomic alterations detected, extrapolates to the exome or the genome as a whole as previously described. 23Assessment of microsatellite instability (MSI) was performed from DNA sequencing across 114 loci, as previously described.Each microsatellite locus had a repeat length of 7-39 base pairs.The NGS-based "MSI score" was translated into categorical MSI-high, MSI-intermediate, or microsatellite stable by unsupervised clustering of specimens for which MSI status was previously assessed via gold standard methods. 24

Genomic Ancestry
Our data in the Genomics-only cohort lacked patient-or physician-reported race, thus patient ancestry was inferred using ancestry-informative markers to classify genomic ancestry.A random forest classifier was used to identify genomic ancestry using genetic variation at single nucleotide polymorphism sites to assign patient samples to one of the ancestral groups-East Asian, European, South Asian, African, Admixed American, and South Asian. 25

Statistical Analyses
Statistical analyses were performed using SPSS 28.1.1 (IBM Corp., Armonk, NY, USA).Categorical variables were compared using Fisher's exact test to determine proportional differences between groups.False discovery rate was corrected using Bonferonni's correction.All statistical tests were 2-sided, and a P-value <.05 was considered significant.Data visualization was performed using R 4.3.1.For survival analysis including OS and PFS, the log-rank test was used to calculate P-values between patient groups, and the Kaplan-Meier method was used to plot survival curves with Prism 10.Cox proportional hazards regression (HR) model was used to conduct multivariate analysis of survival and to calculate the hazard ratio, 95% CI, and associated P-values.
Comparative genomic analysis between MTAP-loss and MTAP-intact mutation profiles in each GI cancer demonstrated near universal concomitant alterations in both CDKN2A and CDKN2B in MTAP-loss tumors (Fig. 2B).All of the 5 GI cancers had statistically different prevalence of CDKN2A and CDKN2B coalterations according    to MTAP status.For example, of the 35 537 CRC tumors that were MTAP-intact, only 1.8% and 0.4% of them demonstrated mutations in CDKN2A and CDKN2B, respectively.The overall prevalence of MTAP-loss in CRC is low at 1.1%, but notably 99.7% and 95% of these cases have coalterations in CDKN2A and CDKN2B.In comparison, MTAP-intact PDAC cases have a high prevalence of CDKN2A and CDKN2B mutations at 43% and 10%, respectively.MTAP-loss PDAC cases occur near universally as part of 9p21 loss, with concomitant CDKN2A and CDKN2B coalterations in 99.8% and 95% of cases, respectively.
A summary table of statistically significant coalterations and immunotherapy markers in MTAP-loss GI cancers is included as Table 2.

Discussion
MTAP-loss is a novel and emerging therapeutic target.Accumulation of MTA in MTAP-loss cancer cells sensitizes them to additional PMRT5 and MAT2A inhibition. 12,28This synthetic lethal relationship is now actively being exploited as a method of targeting MTAP-loss in early phase clinical trials enrolling advanced solid tumors.A Phase I trial of MRTX1719, a PRMT5 inhibitor, has led to RECIST partial responses in MTAP-loss mesothelioma, non-small cell lung cancer, melanoma, gallbladder adenocarcinoma, and malignant peripheral nerve sheath tumor. 19Notably, the early efficacy signal in gallbladder adenocarcinoma, which is frequently an aggressive cancer and resistant to therapies, raises great excitement and interest in investigating MTAP-loss for GI cancers.
As an evolving therapeutic target though, little is known regarding MTAP-loss in GI cancers-including basic epidemiological data such as prevalence and genomic coalterations, which can influence clinical trial feasibility decisions.Here, we undertook a cross-sectional study of >64 000 tumor profiles across the 5 most common GI cancer types to establish the prevalence of MTAP-loss, describe genomic differences between MTAP-loss and MTAP-intact tumors, and delineate any intersection with established immunotherapy markers.To our knowledge, this represents the largest study investigating genomic profiles by MTAP status in GI cancers.
0][31] MTAP-loss is extremely uncommon in HCC and CRC though, making it unlikely that clinical trials of MTAP-loss will specifically seek to enroll patients with these cancers.Regardless of absolute prevalence though, an  The Oncologist, 2024, Vol. 29, No. 6     important finding in our study is that in GI cancers MTAPloss occurs almost exclusively as part of 9p21 loss with codeletions of CDKN2A and CDKN2B.Prior genomic studies and even commercial panels have not always tested specifically for MTAP-loss, making it challenging to perform retrospective studies on clinical outcomes of patients with tumors harboring MTAP-loss. 32,33Testing for CDKN2A and CDKN2B is near ubiquitous though and this finding allows us to extrapolate the presence of MTAP-loss or perform dedicated testing for MTAP-loss in select scenarios.There is also growing literature on the molecular and clinical impact of 9p21-loss, and the recognition that MTAP-loss is synonymous to 9p21loss in GI cancers facilitates a better understanding of evolving literature. 10,20,34In GEAC and PDAC, however, there is a proportion of MTAP-intact tumors with CDKN2A and CDKN2B alterations, making confirmation of MTAP-loss in these tumor types important.No significant differences were observed in the demographic features (age or gender) of MTAP-loss patients with GI cancer.However, our study found that patients with East Asian genomic ancestry were more likely to have MTAP-loss.This could potentially imply an underlying genetic risk with implications on clinical need, such as that seen in EGFR mutations and lung cancers in East Asian patients. 35Additional study is needed though, as there are conflicting observations in the existing literature with other studies supporting both increases and decreases in the prevalence of MTAP-loss in East Asian patients. 36,37ur study does not find mutual exclusivity with other targetable driver mutations in GI cancers with MTAPloss.Although current MTAP-loss clinical trials investigate MAT2A or PRMT5 inhibitors as monotherapy, combination targeted therapy may ultimately prove more efficacious. 38The detailed mutational profiling in our study establishes the prevalence of coalterations that may be amenable to combination therapy.For example, in IHCC, we found that MTAP-loss was associated with a significantly increased prevalence of BRCA2 and BRAF coalterations.This potentially suggests a biological codependency that could be exploited with combination therapeutics.A previous study associated breast cancer risk with defects in methionine metabolism and a methioninedependence phenotype in BRCA1/2 mutation carriers. 39These intriguing links between BRCA2 mutation status with metabolic dependency warrant further preclinical study to establish a potential rational combination therapy of poly(ADP)-ribose polymerase inhibitors and MAT2A or PRMT5 inhibitors in MTAP-loss GI cancers.In PDAC, mutations in PTEN were enriched in MTAP-loss and numerous early phase clinical trials investigating novel PI3K/Akt/mTOR-targeted therapies may provide future candidates for combination therapy. 40lternatively, a sequential targeting approach could be envisioned for patients with tumors harboring MTAP-loss as well as targetable genomic coalterations, thus increasing the number of lines of targeted therapy available to such patients.For example, 36% of BRAF-mutated IHCC has previously been described to be coaltered with MTAP-loss, which our current study re-capitulates with an enrichment of BRAF mutations in MTAP-loss IHCC. 41These patients with both BRAF mutations and MTAP-loss would be candidates to sequence BRAF inhibitors, which have tumor-agnostic FDA approval, with a clinical trial of PRMT5 or MAT2A inhibitors on progression, and vice versa. 4221 loss and/or MTAP-loss has previously been associated with poor response to cancer immunotherapy and hypothesized to be related to immune evasion via cell cycle, metabolic, and type I interferon response pathways. 43In the Genomics-only cohort, we found that MTAP-loss CRC, IHCC, and PDAC were significantly less likely to harbor MSI-H.PD-L1 expression was also significantly lower in MTAP-loss IHCC but surprisingly higher in PDAC and CRC.However, MTAP-loss CRC was also associated with a higher frequency of alterations in STK11 and KEAP1, which have been linked to immunotherapy resistance in NSCLC and pan-cancer cohorts. 44,45utational signatures are thought to infer a tumor's mutational fingerprints and elaborate on multiple cancer processes involved in tumor initiation and progression. 27,46In the Genomics-only cohort, we found increased APOBEC and UV mutational signatures among MTAP-loss PDAC and CRC, respectively.This finding may suggest differences in endogenous and exogenous mutational signatures in MTAP-loss GI cancers and further investigation is warranted.
Finally, no clinical data exist regarding the survival outcomes of patients with GI cancer harboring MTAP-loss.In both advanced PDAC and IHCC, patients with tumors harboring MTAP-loss had a shorter median OS compared with those whose tumors were MTAP-intact, although these findings did not meet statistical significance.In IHCC, the presence of CDKN2A alteration, but not MTAP-loss was associated with significantly worse median OS on multivariate analysis.Further prospective data are required to confirm the prognostic clinical significance of MTAP-loss in advanced PDAC and IHCC.
Limitations of our study include a small patient cohort size with clinical outcomes (n = 102) as well as approximately only one-third and <10% of the Genomics-only cohort having PD-L1 IHC and COSMIC trinucleotide mutational signature analysis performed.Additional limitations include those inherent to retrospective studies including risk of selection bias as well as significant heterogeneity in treatment, surveillance, and follow-up among patients included in the Clinical outcomes cohort.Although our study used standardized methods (NGS) for the detection of homozygous MTAP-loss in both cohorts, epigenetic silencing of MTAP via aberrant promoter methylation is another recognized mechanism for MTAP inactivation in melanoma and glioblastoma, which was not explored in the present study. 47,48

Conclusions
In conclusion, MTAP-loss is an emerging therapeutic target with a high prevalence in PDAC, IHCC, and GEAC and unique mutational profiles.MTAP-loss occurs as part of 9p21-loss in GI cancers and our genomic and immunotherapy marker profiling lays the groundwork for future studies in the field.The lack of mutual exclusivity and identification of actionable coalterations in MTAP-loss GI cancers suggests opportunities for sequential or combination therapeutic approaches.

Figure 1 .
Figure 1.Genomic location of 9p21 and consequences of MTAP-loss.

Figure 5 .
Figure 5.Effect of MTAP status on overall survival in the Clinical outcomes cohort.(A) Kaplan-Meier curve for overall survival in patients with advanced pancreatic cancer.(B) Kaplan-Meier curve for overall survival in patients with advanced cholangiocarcinoma.

Table 1 .
Baseline patient characteristics in Genomics-only and Clinical outcomes cohorts.

Table 2 .
Summary of genomic coalterations and immunotherapy markers in MTAP-loss GI cancers.