You are here

  • Home
  • Online First
  • Clinicopathological characterisation of MTAP alterations in gastrointestinal cancers

Article Text

Article menu
Original research
Clinicopathological characterisation of MTAP alterations in gastrointestinal cancers
  1. Gianluca Mauri1,2,3,
  2. Giorgio Patelli1,2,3,
  3. Laura Roazzi2,3,
  4. Emanuele Valtorta4,
  5. Alessio Amatu3,
  6. Giovanna Marrapese3,
  7. Erica Bonazzina3,
  8. Federica Tosi3,
  9. Katia Bencardino3,
  10. Gabriele Ciarlo2,3,
  11. Elisa Mariella1,5,
  12. Silvia Marsoni1,
  13. Alberto Bardelli1,5,
  14. Emanuela Bonoldi4,
  15. Andrea Sartore-Bianchi2,3,6,
  16. Salvatore Siena2,3
  1. 1IFOM ETS – The AIRC Institute of Molecular Oncology, Milan, Italy
  2. 2Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy
  3. 3Department of Hematology, Oncology, and Molecular Medicine, Grande Ospedale Metropolitano Niguarda, Milan, Italy
  4. 4Department of Pathology, Grande Ospedale Metropolitano Niguarda, Milan, Italy
  5. 5Department of Oncology, Molecular Biotechnology Center, University of Torino, Turin, Italy
  6. 6Division of Research and Innovation, Department of Hematology, Oncology, and Molecular Medicine, Grande Ospedale Metropolitano Niguarda, Milan, Italy
  1. Correspondence to Dr Gianluca Mauri, IFOM, Milano 20139, Italy; gianluca.mauri{at}unimi.it

Abstract

Background Methylthioadenosine phosphorylase (MTAP) is an essential metabolic enzyme in the purine and methionine salvage pathway. In cancer, MTAP gene copy number loss (MTAP loss) confers a selective dependency on the related protein arginine methyltransferase 5. The impact of MTAP alterations in gastrointestinal (GI) cancers remains unknown although hypothetically druggable. Here, we aim to investigate the prevalence, clinicopathological features and prognosis of MTAP loss GI cancers.

Methods Cases with MTAP alterations were retrieved from The Cancer Genome Atlas (TCGA) and a real-world cohort of GI cancers profiled by next-generation sequencing. If MTAP alterations other than loss were found, immunohistochemistry was performed. Finally, we set a case–control study to assess MTAP loss prognostic impact.

Results Findings across the TCGA dataset (N=1363 patients) and our cohort (N=508) were consistent. Gene loss was the most common MTAP alteration (9.4%), mostly co-occurring with CDKN2A/B loss (97.7%). Biliopancreatic and gastro-oesophageal cancers had the highest prevalence of MTAP loss (20.5% and 12.7%, respectively), being mostly microsatellite stable (99.2%). In colorectal cancer, MTAP loss was rare (1.1%), while most MTAP alterations were mutations (5/7, 71.4%); among the latter, only MTAP-CDKN2B truncation led to protein loss, thus potentially actionable. MTAP loss did not confer worse prognosis.

Conclusions MTAP alterations are found in 5%–10% of GI cancers, most frequently biliopancreatic and gastro-oesophageal. MTAP loss is the most common alteration, identified almost exclusively in MSS, CDKN2A/B loss, upper-GI cancers. Other MTAP alterations were found in colorectal cancer, but unlikely to cause protein loss and drug susceptibility.

  • Gastrointestinal Neoplasms
  • Pathology, Molecular

Data availability statement

Data are available on reasonable request.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/jcp-2023-209341

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Despite being potentially druggable, the impact of methylthioadenosine phosphorylase (MTAP) alterations in gastrointestinal (GI) cancers is still largely unknown.

    WHAT THIS STUDY ADDS

    • Gene loss is the most common MTAP alteration, almost exclusively occurring in upper-GI, microsatellite stable, CDKN2A/B loss cancers. MTAP loss in GI cancers did not impact patients’ prognostic. In colorectal cancer, MTAP alterations other than gene loss were found, but they were not associated to protein loss, thus unlikely druggable in ongoing trials.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • Our study provides the largest clinicopathological and prognostic characterisation of MTAP altered GI cancers. This analysis will be instrumental in refining patients selection for clinical trials harnessing MTAP alterations.

    Introduction

    The mechanism of action of most targeted anticancer therapies is essentially the inhibition of oncogenic stimuli through tyrosine-kinase receptor blockade.1–3 However, oncogene-addicted cancers eventually develop resistance under drug-induced selective pressure, limiting targeted therapy efficacy.4 Thus, alternative or complementary therapeutic approaches are warranted.

    Antimetabolic targeted therapies are emerging as new strategies to induce apoptosis and cell death in cancer, given that the pharmacological restraining of certain nutrients or metabolic substrates has shown antitumour activity in preclinical models.5 In this context, polyamine biosynthesis is garnering interest for its role in different cancer types.6

    The MTAP (methylthioadenosine phosphorylase) gene encodes for an enzyme that plays a key role in polyamine metabolism and the salvage pathway of purines and methionine.7 This gene is located in the human chromosome 9p21.3, close to cyclin-dependent kinase 2A and 2B (CDKN2A and CDKN2B), which are two well-known oncosuppressors involved in the regulation of the cell cycle.8 MTAP is often codeleted with CDKN2A/B in a variety of cancers through chromosome 9p21.3 microdeletion events. While MTAP loss was initially alleged as a ‘passenger’ genomic event within the broader context of contiguous tumour suppressor gene loss, recent findings suggest it may also play an independent role in carcinogenesis.9 10 Indeed, preclinical studies showed that MTAP loss promotes tumourigenesis independently from the presence of CDKN2A/B loss.8 11–13 Homozygous loss of MTAP occurs in around 15% of all human cancers and seems to be associated with a more aggressive phenotype with worse prognosis in different malignancies, such as non-small cell lung cancer.14–16 Notably, MTAP loss is highly prevalent in thoracic cancers, reaching approximately 70% in mesothelioma and 60% in lung adenocarcinoma.17 18 Regarding gastrointestinal (GI) cancers, MTAP loss prevalence was reported in heterogenous case series often focusing on individual tumour types. It was reported in 30% of pancreatic cancer (PC),19 20 12%–35% of biliary tract cancer (BTC)21 and approximately 20% in gastro-oesophageal cancer (GEC),22 23 while data are lacking for colorectal cancer (CRC).

    The loss of MTAP has been identified as a potential therapeutic target in cancer due to its multiple cellular effects.7 10 24 In MTAP deficient cells, the accumulation of its substrate methylthioadenosine (MTA)24 can hamper the activity of protein arginine methyltransferase 5 (PRMT5) via direct feedback.10 MTA, as a structural analogue of the methyl donor S-adenosyl-L-methionine (SAM) required for PRMT5 activity, competes with SAM for binding and can effectively inhibit PRMT5 at high concentrations. PRMT5 plays a pivotal role in the methylation of various cellular substrates, including pro-proliferative kinases, histones and critical transcriptional components. Its activity stimulates cellular proliferation and biosynthesis, as showed in different tumour models. From a pharmacological standpoint, exploiting MTAP loss can be approached in at least two distinct ways for two primary reasons: (1) inhibition of de novo purine synthesis and (2) further inhibition of PRMT5, which can be achieved both indirectly, via the accumulation of MTA, and directly through targeted inhibitors. The loss or inhibition of PRMT5 results in altered RNA splicing and an increase in DNA damage, as highlighted in preclinical studies.24 Moreover, effective targeting of polyamine biosynthesis has demonstrated potent antiproliferative and prodifferentiating effects in preclinical models7 10 25 ,26 (figure 1).

    Figure 1
    Figure 1

    Therapeutic vulnerability of MTAP null cells by MAT2A and PRMT5 inhibitors MTAP loss results in intracellular accumulation of MTA, which competes with the activating cofactor SAM, causing a decrease in PRMT5 activity. PRMT5 is a key enzyme for the methylation of proproliferative kinases, thus the reduction in PRMT5 activity leads to activation of oncogenic pathways. Targeting MAT2A or PRMT5 in MTAP null tumours is a potential therapeutic strategy. Created with BioRender.com. Me, methyl group; MTA, methylthioadenosine; MTR-1P, 5-methylthioribose-1-phosphate; MTAP, methylthioadenosine phosphorylase; PRMT5, protein arginine methyltransferase 5; SAH, S-adenosyl homocysteine; SAM, methyl donor S-adenosyl-L-methionine; WDR77, WD repeat domain 77; wt, wild type.

    GI cancers are among the most lethal tumours worldwide, due to a high risk of relapse and metastatisation.27 Among others, cancers arising from the biliopancreatic system retain a particularly poor prognosis owing to the early acquisition of resistance to available anticancer therapies. Identifying new exploitable mechanisms to develop effective targeted treatments is one of the most urgent needs in oncology.

    In this study, we purport to assess the prevalence, clinicopathological features and prognostic impact of MTAP alterations in GI cancers, by leveraging publicly available repositories and integrating data from a patient cohort at our institution.

    Methods

    The Cancer Genome Atlas PanCancer Atlas Analysis

    We conducted a retrospective cohort study using data from The Cancer Genome Atlas (TCGA) accessed via the cBioPortal (https://www.cbioportal.org/, accessed on 25 July 2023).28 29 The scope of our research was confined to GI tumours, specifically GEC, CRC, BTC and PC.30 We queried for patients with unique tumour samples that had been profiled for somatic mutations and copy number alterations (CNA) of MTAP, among other genes.

    Relevant clinical and molecular characteristics were evaluated using the information provided by the cBioPortal.28 29 Further, we narrowed down our investigation to samples exhibiting homozygous MTAP loss, achieved by filtering the downloaded tabular list of cases according to ‘MTAP homodeletion’ status in the ‘R’ software package (V.2023.06.0+421, The R Foundation). A control group for comparative analysis was formed by subjects with diploid wild-type MTAP (MTAP unaltered cohort), excluding those patients with mutations or other MTAP alterations of unknown significance. We then compared clinical and molecular characteristics, progression-free survival (PFS) and overall survival (OS), between these two groups (MTAP loss vs diploid MTAP wild type) using the cBioPortal28 29 and ‘R’ software.

    Niguarda cohort analysis

    From July 2019 to January 2022, we retrospectively collected the results of next-generation sequencing (NGS) analysis obtained through the FoundationOne CDx assay on archival tissue samples of advanced GI tumours at Grande Ospedale Metropolitano Niguarda, Milan, Italy. We included GI tumours only (from oesophagus to anus, liver, pancreas and biliary tract). Samples that failed the NGS analysis due to insufficient or low-quality material were excluded. Data from medical records were annotated on the REDCap platform,31 together with the MTAP status and other molecular results for each subject with MTAP altered tumours, and preserved anonymously with respect to the patients’ privacy. All patients accepted and signed an informed consent for molecular screening through FoundationOne CDx within GO40782/STARTRK-2 trial (NCT02568267).

    We further assessed MTAP protein expression by immunohistochemistry (IHC) using a Rabbit polyclonal anti-MTAP antibody (1:200 dilution, ProteinTech, Tucson, Arizona, USA) in selected tumours harbouring MTAP alterations, other than MTAP loss, to evaluate whether these were associated with loss of protein expression, thus potentially conferring sensitivity to targeted agents currently under clinical investigation in dedicated clinical trials. As previously reported,32 we consider as MTAP deficient only those tumours with complete absence of expression in IHC (score 0). Using the same criteria as a control we also evaluated the expression loss of the protein p16 (the product of CDKN2A) (p16, Clone:JC2, mouse monoclonal antibody, 1:100 dilution, Gennova, Sevillia, Spain), that it is commonly associated with MTAP loss. Immunohistochemistry analyses were performed on formalin-fixed paraffin-embedded tumour sections by an automated staining system Dako Omnis. Normal stromal cells were used as internal positive control.

    Lastly, focusing on MTAP loss, we set a case–control study comparing GI cancers with MTAP loss to MTAP unaltered cases in a 1:2 ratio, matched by primary tumour site. While our primary interest was MTAP gene loss, cases with other types of MTAP alterations were also considered if they exhibited a complete absence of MTAP protein expression, as indicated by an IHC score of 0.

    Statistical analysis

    Due to the potential for an inflated risk of false-positive findings from multiple hypothesis testing in the TCGA cohorts, we set a stringent prespecified level of significance at p<0.001 for all statistical analyses. Results were then validated whenever possible in our independent case–control study according to a standard prespecified p<0.05. We used the χ2 test for categorical data and the Wilcoxon rank-sum test for numerical data to refute the null hypothesis of no difference in variables between the MTAP loss and the MTAP unaltered cohorts (except otherwise specified). Numeric variables were expressed as medians/IQR. Median age was calculated at the time of cancer diagnosis. For survival data (ie, time to progression or censoring for PFS and time to death or censoring for OS), we employed the log-rank test to refute the null hypothesis of no difference in survival rates between the two independent cohorts. PFS and OS data from the TCGA cohorts previously proved reliable and were recommended for use.33 All statistical analyses were executed in ‘R’ (V.2023.06.0+421, The R Foundation), except for those readily computable directly in the cBioPortal.28 29

    Results

    TCGA PanCancer Atlas Analysis

    An initial cohort of 1436 GI cancers was identified through cBioPortal based on the study inclusion criteria. Of these, 73 patients were excluded since the MTAP status had not been profiled, resulting in a selected cohort of 1363 patients. Since MTAP is a tumour suppressor gene reported to promote tumour growth by copy number loss,10 we then focused on 128 cases exhibiting MTAP loss (MTAP loss cohort). For comparison, we also established a control cohort comprising 1224 patients who showed no MTAP alteration (MTAP unaltered cohort) (online supplemental figure 1A).

    Supplemental material

    We first explored the prevalence of MTAP gene alterations in the overall cohort of GI cancers (N=1363), together with clinical and molecular characteristics (online supplemental table 1). This miscellaneous cohort of patients with GI cancer, primarily non-metastatic, exhibited MTAP alterations in 10.3% of cases (N=139), most commonly as copy number loss (9.4%, N=128), followed by mutations (0.6%, N=7) and amplification (0.3%, N=4). Clinical and molecular characteristics of MTAP loss, MTAP mutant and MTAP amplified GI tumours are presented in table 1.

    View this table:
    Table 1

    Clinical and molecular characteristics of GI malignancies exhibiting MTAP alterations, presented as gene loss, mutations and amplification in the TCGA PanCancer Atlas Studies and the Niguarda Cancer Center cohort

    Given that gene deletion was the most common alteration observed for MTAP and that its role in cancer is supported by evidence of pathogenicity,10 we further focused the analysis on MTAP loss cases (N=128). Prevalence of MTAP loss events was higher in PC (22.3%), followed by GEC (12.7%) and BTC (11.1%). Conversely, MTAP loss was quite rare in CRC (1.1%). While addressing a genetic deletion, we investigated CNA for adjacent genes at the same chromosomal location (9p21.3). We observed nearly a complete overlap of CDKN2A/B loss in the MTAP loss cohort (97.7%, N=125), with a statistically significant odds ratio (OR) for co-occurrence (p<0.001, two-sided Fisher’s exact test).

    Comparative analysis between the MTAP loss (N=128) and MTAP unaltered (N=1224), cohorts revealed significant differences among several variables (table 2). Enrichment for MTAP loss was confirmed in a subset of tumour types, that is, PC and GEC (p<0.001), while CRC was confirmed as far less represented in the MTAP loss population as compared with the unaltered one (p<0.001). Interestingly, histology distribution was also different among cases and controls (p<0.001): while adenocarcinoma was similarly represented as the prevalent histology, squamous cell carcinoma cases were fourfold more common in the MTAP loss cohort. As a key issue towards result interpretation, we highlight that almost all the reported squamous-cell cancers were oesophageal cancers in this cohort. From a molecular perspective, MTAP loss cases were almost exclusively classified as MSS as compared with controls (99.2% vs 86.6%, p<0.001), with a small difference in median TMB without clinical relevance (2.5 vs 3.3 mutations/megabase, p<0.001). CDKN2A/B loss was confirmed significantly enriched in the MTAP loss population (97.7% vs 8.1%, p<0.001), and the same result applied to the deletion of several other genes located in proximity to chromosome 9p21.3. Indeed, MIR31HG, IFNA1, IFNA5, IFNA8, IFNE, LINC01239, DMRTA1 and KLHL9 on the chromosome cytoband 9p21.3 were found deleted in ≥50% of MTAP loss cases (compared with <2% in controls, p<0.001). We also observed minor enrichment (deletion in <50% of cases) for other 75 genes spanning from 9p21.1 to 9p24.3 (p<0.001) (online supplemental figure 2). Regarding classical oncogenes and tumour suppressors in GI cancers, we found no difference in TP53 and KRAS mutations, while APC mutations were less common (7.1% vs 35.5%, p<0.001) and ERBB2 amplifications more common (16.4% vs 7.5%, p=0.001) in case of MTAP loss, again likely due to the low prevalence of CRC and high prevalence of GEC in this population. Focusing on individual tumour types, the analysis was not powered to identify differences in subpopulations. As such, we could only observe a trend towards a higher incidence of KRAS mutations in MTAP loss PC as compared with the unaltered counterpart—85.0% (34/40) vs 60.1% (83/138), p=0.003. At the transcriptomic level, there was a significant association between MTAP copy number loss and decreased gene expression (online supplemental figure 3).

    Figure 2
    Figure 2

    No difference in survival analysis of pancreatic cancer patients according to MTAP status in the TCGA PanCancer Atlas cohort (A, B) and the Niguarda Cancer Center cohort (C, D). MTAP, methylthioadenosine phosphorylase; TCGA, The Cancer Genome Atlas.

    Figure 3
    Figure 3

    Immunohistochemistry (IHC) staining of two cases of metastatic colorectal cancer harbouring MTAP alterations other than gene loss. (A, B) Samples were collected from a patient in which an MTAP-CDKN2B truncation was identified by next-generation sequencing (NGS), in which both MTAP and p16 IHC revealed a complete lack of expression for both proteins, respectively. (C, D) Samples were collected from a patient in which MTAP M140V point mutation was identified by NGS, in which both MTAP and p16 protein expression was maintained, respectively. MTAP IHC staining was performed using a rabbit polyclonal anti-MTAP antibody (1:200 dilution, Pro-teinTech, Tucson, AZ). As a control, we also evaluated the expression of the protein p16 (the product of CDKN2A gene) using the p16, Clone:JC2, mouse monoclonal antibody (1:100 dilution, Gennova, Sevillia). Only samples with complete MTAP and/or p16 protein loss were considered MTAP deficient (IHC score 0).

    View this table:
    Table 2

    Clinical and molecular characteristics of GI malignancies exhibiting MTAP loss as compared with MTAP unaltered GI tumours from the TCGA PanCancer Atlas Studies

    We performed survival analysis (online supplemental figure 4). Patients with MTAP loss showed unfavourable PFS (17.06 months, 95% CI 13.15 to 24.33 vs 42.31 months, 95% CI 36.03 to 63.42, p<0.001) and a trend towards worse OS (20.35 months, 95% CI 17.92 to NA vs 49.38 months, 95% CI 44.74 to 58.55, p=0.002). However, after performing a multivariate Cox proportional hazards model to investigate the impact of MTAP loss and different GI cancer types on patient prognosis, this result did not reach statistical significance (HR=0.79, p=0.097). Indeed, when we included the type of cancer in our model, we found significant differences in PFS between different cancer types, with significantly worse PFS for PC (HR=3.13, p<0.001), BTC (HR=3.11, p<0.001) and GEC (HR=1.89, p<0.001). Since we previously observed a different distribution of cancer types according to MTAP status with a higher frequency of more aggressive cancer types in the MTAP loss group, this uneven distribution may explain the observed difference in PFS between the two cohorts, rather than the MTAP classification itself. Indeed, there was no survival difference according to MTAP status when addressing different tumour types separately (figure 2A,B and online supplemental figure 5).

    Niguarda cohort analysis

    We analysed NGS reports of 558 GI tumours. Sequencing failed for 50 samples, leaving an initial cohort of 508 GI cancer patients to be screened for MTAP alterations (329 CRC, 80 PC, 47 GEC, 36 BTC, 16 small bowel or other tumours).

    MTAP alterations were found in 27 (5.3%) (online supplemental figure 1B). Characteristics of patients are reported in table 1. Overall, most patients (14/27, 51.8%) were metastatic at diagnosis. Adenocarcinoma was the most common histology (25/27, 92.6%). MTAP alterations were found in PC (12/80, 15.0%), GEC (4/47, 8.5%) and BTC (2/36, 5.5%), CRC (7/329, 2.1%), and others including small bowels (2/16, 12.5%) (online supplemental table 2). Homozygous MTAP loss was the main deleterious alteration retrieved (22/27, 81.5%), followed by mutations (5/27, 18.5%), although differences were noticed according to primary tumour histology.

    All tumour types but CRC were characterised by gene loss as the only MTAP genetic alteration. In fact, all MTAP alterations other than gene loss were retrieved in CRC, that was indeed enriched with 5/7 alterations being mutations (three reported as pathogenic: splice site 34–1G>A, A191fs*6, MTAP-CDKN2B truncation; two as variants of unknown significance). One of these cases was hypermutated due to microsatellite instability. In all 5 MTAP mutant samples from CRC patients, we performed IHC analysis to investigate the genomic effect on protein expression. Only MTAP-CDKN2B truncation as assessed by NGS led to complete lack of MTAP expression (IHC 0), coupled with absence of p16 proteins expression (IHC 0) despite no CNA for CDKN2A/B was reported; this annotation suggests that both MTAP and CDKN2B were involved in a truncation event leading to absence of translation to the protein level, differently from other somatic variants with no or minor impact on the protein level. In all other cases, MTAP protein expression was retained (figure 3). We performed IHC in a case of CRC harbouring MTAP loss: although there was minimal heterogeneity in staining, 90%–98% of cancer cells did not express MTAP protein, as compared with intense and homogeneous protein expression in proficient cases.

    Focusing on MTAP loss, all cases were microsatellite stable (100%) and CDKN2A/B deleted (100%). Given that PC was most represented in our cohort, we built a case–control study with a 1:2 ratio in PC patients (table 3). We found no significant difference in terms of clinical and molecular features apart from a higher prevalence of CDKN2A/B loss (100% vs 4.2%, p<0.001). Survival analysis did not show any difference for both PFS to first-line treatment and OS according to MTAP status (figure 2C,D).

    View this table:
    Table 3

    Clinical and molecular characteristics of MTAP loss versus MTAP unaltered pancreatic cancer patients from the Niguarda Cancer Center cohort

    Discussion

    In our study, we found that MTAP alteration prevalence in GI cancers ranges from 5% to 10% across various cancer types. We reported here that PC, GEC and BTC show the highest prevalence of MTAP loss among GI cancers (15%–22%, 9%–13%, 6%–11%, respectively). Conversely, CRC showed a lower incidence of MTAP alterations (2%), which were mostly mutations rather than gene loss. These results are in line with previous reports on the prevalence of MTAP alterations in GI cancers, confirming the enrichment of such molecular feature in upper-GI malignancies.19–23 In addition, our findings bring new knowledge regarding the very low prevalence of MTAP loss in lower-GI tumours (<1% in CRC).

    Similar to other non-GI malignancies, MTAP and CDKN2A/B loss co-occurred in almost all cases. Moreover, in more than 50% of cases MTAP loss co-occurred with the loss of other genes located on chromosomal locus 9p21.3, thus leading to postulate that MTAP loss could be related to a large-scale chromosome deletion. We also investigated the potential clinical impact of MTAP mutations, that is, currently unknown. In our cohort, only one mutation out of five led to IHC protein loss, thereby limiting the potential use of targeted approaches in this setting.

    Regarding prognosis, we found no association between MTAP loss and PFS and OS. Indeed, the higher prevalence of biliopancreatic cancers among MTAP loss cancers emerged as the actual driver of worse prognosis rather than the MTAP loss itself. Accordingly, in patients identified in the Niguarda series we did not observe any survival difference among MTAP loss and proficient cases. These data are in line with another recent study on BTC in which MTAP loss was not prognostic.34

    Even if the overall results from both the cohorts are comparable, there are also some differences. First, the TCGA cohort is larger but less clinically detailed than our institutional cohort. Besides, in the TCGA nearly 75% of patients were affected by non-metastatic tumours, whereas all patients had metastatic cancers in our cohort. Finally, CRC was the most common histology in our cohort due to centre-specific enrichment, while GEC squamous tumours, in which MTAP loss plays an important role in carcinogenesis, accounted for only a minority of cases.14

    Beyond MTAP loss, we found five out of seven MTAP altered CRC cases harbouring alterations other than gene loss. Given the availability of clinical trials targeting cancers with MTAP loss (online supplemental table 3), we performed IHC to test whether protein expression in these five cases was retained, but we found that only the MTAP-CDKN2B truncation led to protein loss, thus potentially having a clinical impact predisposing to sensitivity to MTAP targeted treatment strategies.

    Our study has also limitations. First, no functional assessment of MTAP mutations has been performed. Second, the retrospective nature of the study and the limited sample size of some GI subset in the present cohort represent a further limitation. Third, the analysis was underpowered to distinguish small-effect size differences within individual tumour types, especially given the small numerosity of MTAP loss CRC and BTC. Finally, we acknowledge that all GEC in the Niguarda cohort were adenocarcinomas as a possible limitation, differently from the TCGA cohort. Indeed, since we retrieved molecular profiles data from GO40782/STARTRK-2 clinical trial screening (NCT02568267), according to physicians choice at our institution mostly adenocarcinoma patients rather than squamous cell cancer patients were screened for trial enrolment.

    In conclusion, this is the first study investigating MTAP alterations prevalence in different GI cancers, also focusing on concomitant genetic alterations and addressing MTAP loss also at the transcriptomics and proteomics levels. We found that MTAP loss occurs almost exclusively in MSS tumours and it co-occurs with CDK2NA/B and/or other 9p21.3-located gene loss in almost all cases. MTAP loss was more common in upper-GI cancers (PC, BTC and GEC), compared with lower GI-cancers (ie, CRC). However, MTAP loss did not impact GI cancer prognosis. In CRC, we also identified MTAP alterations other than gene loss but, since protein loss was uncommon in these cases, it is unlikely that they may impact clinical outcomes and therapeutic targetability in GI cancers.

    Data availability statement

    Data are available on reasonable request.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants and all patients accepted and signed an informed consent for molecular screening through FoundationOne CDx within GO40782/STARTRK-2 trial (NCT02568267). Clinical data were obtained by informed consent for publication of anonymised data according to standard internal consent form at Niguarda Cancer Center ('Consent Form for the Acquisition of Personal Data for Clinical Research (ex. Art. 9 GDPR), Epidemiological Research and Epidemiological Study Purposes', Mod02888 Rev. 4, 18 February 2020). Participants gave informed consent to participate in the study before taking part.

    Acknowledgments

    GM and GP are PhD students within the European School of Molecular Medicine (SEMM). EM was supported by the FIRC-AIRC 'Michele e Carlo Ardizzone' fellowship for Italy.

    References

    1. ↵
      1. Sartore-Bianchi A,
      2. Trusolino L,
      3. Martino C, et al
      . Dual-targeted therapy with trastuzumab and lapatinib in treatment-refractory, KRAS codon 12/13 wild-type, HER2-positive metastatic colorectal cancer (HERACLES): a proof-of-concept, multicentre, open-label, phase 2 trial. Lancet Oncol 2016;17:738–46. doi:10.1016/S1470-2045(16)00150-9
      OpenUrlCrossRefPubMed
    2. ↵
      1. Demetri GD,
      2. von Mehren M,
      3. Blanke CD, et al
      . Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 2002;347:472–80. doi:10.1056/NEJMoa020461
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Remon J,
      2. Steuer CE,
      3. Ramalingam SS, et al
      . Osimertinib and other third-generation EGFR TKI in EGFR-mutant NSCLC patients. Ann Oncol 2018;29:i20–7. doi:10.1093/annonc/mdx704
      OpenUrlCrossRef
    4. ↵
      1. Vasan N,
      2. Baselga J,
      3. Hyman DM
      . A view on drug resistance in cancer. Nature 2019;575:299–309. doi:10.1038/s41586-019-1730-1
      OpenUrlCrossRefPubMed
    5. ↵
      1. Stine ZE,
      2. Schug ZT,
      3. Salvino JM, et al
      . Targeting cancer metabolism in the era of precision oncology. Nat Rev Drug Discov 2022;21:141–62. doi:10.1038/s41573-021-00339-6
      OpenUrlCrossRef
    6. ↵
      1. Casero RA,
      2. Murray Stewart T,
      3. Pegg AE
      . Polyamine metabolism and cancer: treatments, challenges and opportunities. Nat Rev Cancer 2018;18:681–95. doi:10.1038/s41568-018-0050-3
      OpenUrlCrossRefPubMed
    7. ↵
      1. Marjon K,
      2. Cameron MJ,
      3. Quang P, et al
      . MTAP deletions in cancer create vulnerability to targeting of the MAT2A/PRMT5/RIOK1 Axis. Cell Rep 2016;15:574–87. doi:10.1016/j.celrep.2016.03.043
      OpenUrlCrossRefPubMed
    8. ↵
      1. Bertino JR,
      2. Waud WR,
      3. Parker WB, et al
      . Targeting tumors that lack methylthioadenosine phosphorylase (MTAP) activity. Cancer Biology & Therapy 2011;11:627–32. doi:10.4161/cbt.11.7.14948
      OpenUrl
    9. ↵
      1. Chen ZH,
      2. Zhang H,
      3. Savarese TM
      . Gene deletion chemoselectivity: codeletion of the genes for p16(INK4), methylthioadenosine phosphorylase, and the alpha- and beta-interferons in human pancreatic cell carcinoma lines and its implications for chemotherapy. Cancer Res 1996;56:1083–90.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Kryukov GV,
      2. Wilson FH,
      3. Ruth JR, et al
      . MTAP deletion confers enhanced dependency on the PRMT5 arginine methyltransferase in cancer cells. Science 2016;351:1214–8. doi:10.1126/science.aad5214
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Fedoriw A,
      2. Rajapurkar SR,
      3. O’Brien S, et al
      . Anti-tumor activity of the type I PRMT inhibitor, GSK3368715, synergizes with PRMT5 inhibition through MTAP loss. Cancer Cell 2019;36:100–114. doi:10.1016/j.ccell.2019.05.014
      OpenUrlCrossRef
    12. ↵
      1. Kadariya Y,
      2. Yin B,
      3. Tang B, et al
      . Mice heterozygous for germ-line mutations in methylthioadenosine phosphorylase (MTAP) die prematurely of T-cell lymphoma. Cancer Res 2009;69:5961–9. doi:10.1158/0008-5472.CAN-09-0145
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Kadariya Y,
      2. Tang B,
      3. Wang L, et al
      . Germline mutations in Mtap cooperate with Myc to accelerate tumorigenesis in mice. PLOS ONE 2013;8:e67635. doi:10.1371/journal.pone.0067635
    14. ↵
      1. Lin X,
      2. Yan C,
      3. Gao Y, et al
      . Genetic variants at 9p21.3 are associated with risk of esophageal squamous cell carcinoma in a Chinese population. Cancer Sci 2017;108:250–5. doi:10.1111/cas.13130
      OpenUrl
    15. ↵
      1. Su C-Y,
      2. Chang Y-C,
      3. Chan Y-C, et al
      . MTAP is an independent prognosis marker and the concordant loss of MTAP and p16 expression predicts short survival in non-small cell lung cancer patients. Eur J Surg Oncol 2014;40:1143–50. doi:10.1016/j.ejso.2014.04.017
      OpenUrl
    16. ↵
      1. Marcé S,
      2. Balagué O,
      3. Colomo L, et al
      . Lack of methylthioadenosine phosphorylase expression in mantle cell lymphoma is associated with shorter survival: implications for a potential targeted therapy. Clin Cancer Res 2006;12:3754–61. doi:10.1158/1078-0432.CCR-05-2780
      OpenUrlAbstract/FREE Full Text
    17. ↵
      1. Illei PB,
      2. Rusch VW,
      3. Zakowski MF, et al
      . Homozygous deletion of CDKN2A and codeletion of the methylthioadenosine phosphorylase gene in the majority of pleural mesotheliomas. Clin Cancer Res 2003;9:2108–13.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Ashok Kumar P,
      2. Graziano SL,
      3. Danziger N, et al
      . Genomic landscape of non-small-cell lung cancer with methylthioadenosine phosphorylase (MTAP) deficiency. Cancer Med 2023;12:1157–66. doi:10.1002/cam4.4971
      OpenUrl
    19. ↵
      1. Yu S,
      2. Doyle LA,
      3. Hornick JL, et al
      . The diagnostic utility of methylthioadenosine phosphorylase immunohistochemistry for pancreatic ductal adenocarcinoma in FNA and small biopsy specimens. Cancer Cytopathol 2023. doi:10.1002/cncy.22777 [Epub ahead of print 6 Dec 2023].
    20. ↵
      1. Hustinx SR,
      2. Hruban RH,
      3. Leoni LM, et al
      . Homozygous deletion of the MTAP gene in invasive adenocarcinoma of the pancreas and in periampullary cancer: a potential new target for therapy. Cancer Biol Ther 2005;4:83–6. doi:10.4161/cbt.4.1.1380
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Karikari CA,
      2. Mullendore M,
      3. Eshleman JR, et al
      . Homozygous deletions of methylthioadenosine phosphorylase in human biliary tract cancers. Mol Cancer Ther 2005;4:1860–6. doi:10.1158/1535-7163.MCT-05-0103
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Powell EL,
      2. Leoni LM,
      3. Canto MI, et al
      . Concordant loss of MTAP and p16/CDKN2A expression in gastroesophageal carcinogenesis: evidence of homozygous deletion in esophageal noninvasive precursor lesions and therapeutic implications. Am J Surg Pathol 2005;29:1497–504. doi:10.1097/01.pas.0000170349.47680.e8
      OpenUrlCrossRefPubMed
    23. ↵
      1. Cheng X-Y,
      2. Liu Z,
      3. Shang L, et al
      . Deletion and downregulation of MTAP contribute to the motility of esophageal squamous carcinoma cells. Onco Targets Ther 2017;10:5855–62. doi:10.2147/OTT.S151953
      OpenUrl
    24. ↵
      1. Kalev P,
      2. Hyer ML,
      3. Gross S, et al
      . MAT2A inhibition blocks the growth of MTAP-deleted cancer cells by reducing PRMT5-dependent mRNA splicing and inducing DNA damage. Cancer Cell 2021;39:209–24. doi:10.1016/j.ccell.2020.12.010
      OpenUrl
    25. ↵
      1. Mavrakis KJ,
      2. McDonald ER,
      3. Schlabach MR, et al
      . Disordered methionine metabolism in MTAP/CDKN2A-deleted cancers leads to dependence on PRMT5. Science 2016;351:1208–13. doi:10.1126/science.aad5944
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. Mariella E, et al
      . Transcriptome-wide gene expression Outlier analysis pinpoints therapeutic Vulnerabilities in colorectal cancer. Accepted for Publication in Molecular Oncology 2024.
    27. ↵
      1. Siegel RL,
      2. Miller KD,
      3. Wagle NS, et al
      . Cancer statistics, 2023. CA Cancer J Clin 2023;73:17–48. doi:10.3322/caac.21763
      OpenUrlCrossRefPubMed
    28. ↵
      1. Cerami E,
      2. Gao J,
      3. Dogrusoz U, et al
      . The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov 2012;2:401–4. doi:10.1158/2159-8290.CD-12-0095
      OpenUrlAbstract/FREE Full Text
    29. ↵
      1. Gao J,
      2. Aksoy BA,
      3. Dogrusoz U, et al
      . Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal 2013;6:pl1. doi:10.1126/scisignal.2004088
    30. ↵
      1. The Cancer Genome Atlas Research Network,
      2. Weinstein JN,
      3. Collisson EA, et al
      . The cancer genome atlas Pan-Cancer analysis project. Nat Genet 2013;45:1113–20. doi:10.1038/ng.2764
      OpenUrlCrossRefPubMed
    31. ↵
      1. Harris PA,
      2. Taylor R,
      3. Minor BL, et al
      . The REDCap consortium: building an international community of software platform partners. J Biomed Inform 2019;95:103208. doi:10.1016/j.jbi.2019.103208
      OpenUrlCrossRefPubMed
    32. ↵
      1. Alhalabi O,
      2. Zhu Y,
      3. Hamza A, et al
      . Integrative Clinical and Genomic Characterization of MTAP-deficient Metastatic Urothelial Cancer. Eur Urol Oncol 2023;6:228–32. doi:10.1016/j.euo.2021.10.006
      OpenUrl
    33. ↵
      1. Liu J,
      2. Lichtenberg T,
      3. Hoadley KA, et al
      . An integrated TCGA Pan-Cancer clinical data resource to drive high-quality survival outcome analytics. Cell 2018;173:400–16. doi:10.1016/j.cell.2018.02.052
      OpenUrlCrossRefPubMed
    34. ↵
      1. Gaspar CF,
      2. Ngoi NY,
      3. Tang T, et al
      . Abstract 963: clinical impact of MTAP status in advanced cholangiocarcinoma: genomic profile and response to treatment. Cancer Res 2023;83:963. doi:10.1158/1538-7445.AM2023-963

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • AS-B and SS are joint senior authors.

    • GM, GP and LR are joint first authors.

    • Handling editor Yoh Zen.

    • GM, GP and LR contributed equally.

    • Contributors GMauri conceived and wrote the manuscript, collected and analysed data. GP and LR wrote the manuscript, collected and analysed data. EV and EBonoldi retrieved samples and performed the immunohistochemistry analysis. AA, GC and GMarrapese collected the data. EBonazzina, FT and KB analysed the data and critically reviewed the manuscript. SM, AB, AS-B and SS reviewed, edited and supervised manuscript preparation and development. AS-B and SS are joint last authors. SS is the guarantor of the content of this article.

    • Funding This study was supported by Fondazione Oncologia Niguarda Onlus. The research leading to these results received funding from AIRC IG 2017—project ID 20685—P.I. Salvatore Siena (SS); AIRC under IG 2018—ID. 21923 project—P.I. Bardelli Alberto (AB); AIRC under 5 per Mille 2018—ID. 21091 programme—P.I. Bardelli Alberto (AB), G.L. Silvia Marsoni (SM), G.L. Salvatore Siena (SS); European Research Council (ERC) under the European Union‘s Horizon 2020 research and innovation programme (TARGET, grant agreement no 101020342) (AB); International Accelerator Award, ACRCelerate, jointly funded by Cancer Research UK (A26825 and A28223), FC AECC (GEACC18004TAB) and AIRC (22795) (AB); IMI contract no 101007937 PERSIST-SEQ (AB).

    • Competing interests AB reports personal fees from Guardant Health and Inivata during the conduct of the study as well as grants from AstraZeneca, Boehringer-Ingelheim and Neophore outside the submitted work; in addition, AB is a shareholder of Neophore and Kither. SS is an advisory board member for Agenus, AstraZeneca, Bayer, BMS, CheckmAb, Daiichi-Sankyo, Guardant Health,Merck, Novartis, Roche-Genentech and Seagen. AS-B reports personal fees from Amgen, Bayer, Servier, Guardant Health and Novartis during the conduct of the study. GMauri and GP received honoraria from COR2ED.

    • Provenance and peer review Not commissioned; internally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.