Age- and sex-based differences in the genomic profiles of patients with gastrointestinal and pancreatic neuroendocrine neoplasms

Abstract

Purpose

Age and sex are known to influence outcomes in neuroendocrine neoplasms (NENs), yet their molecular determinants remain poorly defined. We queried a real-world dataset of gastrointestinal (GI) tract and pancreatic (P) NENs and characterized their clinical, molecular, and immune profiles.

Methods

One thousand nine hundred thirty-five cases (GI: n = 1431, P: n = 504) of NENs were analyzed using Next Generation or Whole Exome Sequencing of DNA, and 1211/1935 cases (GI: n = 917, P: n = 294) underwent Whole Transcriptome Sequencing. We compared the molecular and immune profile with respect to age and sex.

Results

Older age at diagnosis was associated with worse survival in both GI- and P-NENs. In patients with GI-NENs, there was decreased survival in males compared to females. In GI-NENs, TP53, RB1, FAT1, and KMT2D mutations, as well as immune checkpoint gene (ICG) expressions of LAG3, CD80, and HAVCR2 and M1 macrophages increased with increasing age while APC mutations and M2 macrophages decreased with increasing age. In P-NENs, TP53, RB1, KRAS, and SMAD4 mutations and CD4+ T cells increased with increasing age while MUTYH and NTHL1 mutations and M2 macrophages decreased with increasing age. In GI-NENs, TP53, FBXW7, and TERT (promoter) mutations and genomic loss of heterozygosity (gLOH) were increased in males. In P-NENs, PIK3CA mutations and dMMR/MSI-H were increased in females.

Conclusion

Our study queries one of the largest datasets of GI- and P-NENs to date and highlights distinct age- and sex-specific molecular and immune profiles. Given the exploratory nature of these analyses and borderline significance, these results remain hypothesis-generating, providing an initial framework for future validation studies.

Implications for Practice.

Demographic factors such as age and sex are associated with different outcomes in patients with NENs, yet little is known about the molecular basis underlying these differences. This study offers the most comprehensive real-world analysis of GI- and P-NENs to date, revealing age- and sex-associated differences in tumor mutations, immune checkpoint gene expression, and immune cell infiltration. These findings are hypothesis-generating and highlight the need to incorporate demographic stratification in clinical trial design and treatment planning. Ultimately, this work provides a foundation for more tailored treatment approaches for patients with NENs.

Introduction

Neuroendocrine neoplasms (NENs) arise throughout the body and encompass a heterogeneous group of diseases across a full spectrum of potential clinical behavior, from indolent and asymptomatic to more aggressive and highly symptomatic in nature. A small portion of NENs is associated with familial syndromes like multiple endocrine neoplasia type 1 (MEN1). However, in sporadic cases, specific molecular biomarkers have not been fully identified, and this lack of understanding has hindered the ability to tailor treatments and predict clinical behavior. Previous research have identified molecular phenotypes specific to the grade and location of NENs, including MEN1, DAXX/ATRX, and mTOR pathway gene alterations, as being prevalent in P-NENs. It has also been found that P-NENs have less frequent APC mutations than GI-NENs do.^1,2 In a prospective study of NENs that underwent next-generation sequencing (NGS), Vijayvergia et al. discovered that non-P NENs showed a very low incidence of mutations in the classical signaling pathways, which indicate that alternate pathways such as epigenetic modifications are at play.³ In small bowel NENs, there are some noted genomic alterations in the mTOR pathway and CDKN1B pathways, which may allow for deployment of targeted therapies.⁴ Further data and molecular explanation for differential outcomes of patient subgroups can help refine our understanding of these neoplasms. Specifically, demographic variables, such as age and sex, which have long been associated with risk of cancer incidence, therapeutic response, and mortality, may be of specific interest.

Age and sex have been associated with certain distinct molecular signatures in other cancers. Yao et al. explored age-stratified analysis of biomarkers in those with colorectal cancer and discovered new potential tumor markers for early-onset colorectal cancer, which they found may have a higher incidence of developing into metastatic disease.⁵ In addition to age, an individual’s sex has also long been associated with cancer incidence, aggressiveness, and treatment responses. Yuan et al. have conducted a large-scale pan-cancer analysis and have discovered that for many cancer types, there are distinct sex-based molecular signatures, specifically showing varying mRNA expression of pathways regulating immune response, the cell cycle, metabolism and DNA repair.⁶

The study of demographics and molecular markers has been more limited in neuroendocrine tumors. Abdel-Rahman et al. previously assessed sex-based differences in gastroenteropancreatic-neuroendocrine neoplasms and have shown that female patients tend to have better overall survival compared with male patients, regardless of stage, histology, and differentiation.⁷ A retrospective review has also shown distinct clinical differences in young-onset P-NENs versus older onset and have shown that, in general, younger patients with P-NENs have a better overall survival rate.⁸ However, the molecular determinants associated with age and sex in patients with GI- and P-NENs have been lacking. To delineate the associations of age and sex with outcomes, we queried a large real-world dataset of GI- and P-NENs and characterized their molecular and immune profiles.

Methods

 

NGS of DNA

A total of 1935 GI- and P-NENs were molecularly profiled at Caris Life Sciences (Phoenix, AZ) between 2015 and 2022. Age of diagnosis is defined as the patient’s age at biopsy collection. Most molecular profiling occurred shortly after biopsy collection. Sex was patient self-reported. Here, sex is defined by the biological characteristics of humans such as chromosomes and reproductive organs whereas gender is defined as the socially constructed characteristics humans such as behavior and identities. NGS was performed on genomic DNA isolated from formalin-fixed paraffin-embedded (FFPE) tumor samples using the NextSeq or NovaSeq 6000 platform (Illumina, Inc., San Diego, CA). For the Nextseq sequenced tumors, a custom-designed SureSelect XT assay was used to enrich 592 whole-gene targets (Agilent Technologies, Santa Clara, CA). Further, for the NovaSeq sequenced tumors, 719 clinically relevant genes were enriched at a high coverage and read-depth, along with another panel designed to enrich for >20 000 genes at lower read-depth. All variants were detected with >99% confidence based on allele frequency and amplicon coverage, with an average sequencing depth of coverage of >500 and an analytic sensitivity to identify variants with a variant allele frequency of ≥5%. Genetic variants identified were interpreted by board-certified molecular geneticists and categorized as “pathogenic (P),” “likely pathogenic (LP),” “variant of unknown significance,” “likely benign,” or “benign,” according to the American College of Medical Genetics and Genomics (ACMG) standards. When assessing mutation frequencies of individual genes, “pathogenic” and “likely pathogenic” were counted as mutations. The copy number alteration (CNA) of each exon was determined by calculating the average depth of the sample along with the sequencing depth of each exon and comparing this calculated result to a pre-calibrated value. Alterations were shortlisted based on whether there was at least a trend for statistical significance (P < .05, q > 0.05) and prevalent in at least 2% of cohort (with the exception of sex association in P-NENs, where PIK3CA mutations are at 1.92%).

Tumor mutational burden estimation

After excluding all variants previously described as germline alterations according to dbSNP151, Genome Aggregation Database (gnomAD) databases and those characterized as benign variants by geneticists at Caris Life Sciences, tumor mutational burden (TMB) was calculated by counting non-synonymous, missense, nonsense, inframe insertion/deletion and frameshift mutations present in each tumor. A cutoff point of ≥10 mutations per MB was used to classify TMB-high (TMB-H) tumors, based on findings from the KEYNOTE-158 study.⁹

Microsatellite instability-high/mismatch repair deficient (MSI-H/dMMR) status detection

A combination of multiple test platforms was used to determine the MSI-H/dMMR status of the tumors profiled, including immunohistochemistry (MLH1, M1 antibody; MSH2, G2191129 antibody; MSH6, 44 antibodies; and PMS2, EPR3947 antibody [Ventana Medical Systems, Inc., Tucson, AZ, USA]), NGS (>2800 target microsatellite loci were examined and compared to the reference genome hg19 from the University of California, Santa Cruz Genome Browser database) and fragment analysis (FA, Promega, Madison, WI). The three platforms generated highly concordant results, as previously reported,¹⁰ and in the rare cases of discordant results, the MSI-H or MMR status of the tumor was determined in the order of immunohistochemistry (IHC), FA, and NGS.

Whole transcriptome sequencing

FFPE specimens (n = 1211 of 1935) underwent pathology review to measure percent tumor content and tumor size; a minimum of 10% of tumor content in the area for microdissection was required to enable enrichment and extraction of tumor-specific RNA. The Qiagen RNA FFPE tissue extraction kit was used for RNA extraction, and the RNA quality and quantity were determined using the Agilent TapeStation. Biotinylated RNA baits were hybridized to the synthesized and purified cDNA targets, and the bait-target complexes were amplified in a post-capture PCR reaction. The Illumina NovaSeq 6500 was used to sequence the whole transcriptome from tumor samples to an average of 60M reads. Raw data were demultiplexed by Illumina Dragen BioIT accelerator, trimmed, counted, excluded PCR-duplicates, and aligned to human reference genome hg19 by STAR aligner. TPM (Transcripts Per Million Molecules) were generated using the Salmon expression pipeline. Estimates of immune cell infiltrates were calculated from bulk RNA sequencing by the QuanTIseq method.¹¹

Immune check point genes analyzed included CTLA4, CD274, PDCD1LG2, LAG3, CD80, CD86, PDCD1, HAVCR2, IDO1, and IFNG.

Immunohistochemistry

IHC was performed on FFPE sections. Slides were stained using automated staining techniques in accordance with the manufacturer’s protocol and were further optimized and validated as per CLIA/CAO and ISO requirements. A board-certified pathologist evaluated all IHC results independently. PD-L1 expression was tested by IHC using SP142 antibody (Spring Biosciences). The staining intensity on the tumor cell membrane was assessed on a semiquantitative scale as follows: 0, no staining; 1+, weak staining; 2+, moderate staining; and 3+, strong staining. Tumors with staining scores of 2+ or 3+ in ≥5% of tumor cells were regarded as PD-L1 positive.

Clinical outcomes

Real-world overall survival (OS) information was obtained from insurance claims data and calculated using Kaplan–Meier estimates from the time of tissue biopsy to last contact. 1396/1431 (97.6%) of patients with GI-NENs and 497/504 (98.6%) of patients with P-NENs had associated outcomes data and were analyzed to identify associations of overall survival with age and gender.

Ethics statement

This study was conducted in accordance with the guidelines of the Declaration of Helsinki, Belmont Report, and U.S. Common Rule. In keeping with 45 CFR 46.101(b)(4), this study was performed using retrospective and deidentified clinical data. This retrospective study was conducted under Caris Life Sciences’ Research Data Banking protocol, which was reviewed and granted IRB exemption by the WCG IRB. The study adhered to the ethical guidelines of the Declaration of Helsinki, the Belmont Report, and the U.S. Common Rule.

Statistics

Chi-square, Fisher’s Exact, Mann–Whitney U, Kruskal–Wallis, ANOVA and t-tests (with Wech’s correction) were used and corrected for multiple hypothesis testing (MHT) using the Benjamini–Hochberg method where appropriate. Statistical significance was considered when q < 0.05. Given the exploratory nature of the study, we highlighted both significant comparisons (solid black lines and asterisks) and insignificant (grey dashed lines and asterisks) after MHT correction.

Results

Demographics

A total of 1935 NEN patients (1431 GI-NEN and 504 P-NEN) had tumors that underwent comprehensive genomic profiling from 2015 to 2022 and were used in this analysis. The demographic and clinical characteristics are delineated in Table 1. Most tumors underwent genomic profiling at time of biopsy collection. These include 1073 (55.5%) male patients and 862 (44.5%) female patients. The same patient population was divided with three age cutoffs: age ≤ 50-year old (group A), 50-year old < age < 70-year old (group B), and ≥ 70-year old (group C) at the time of tissue collection. We were unable to subdivide data into well-differentiated versus poorly differentiated tumor groups due to the lack of information on grade of tumors.

Table 1.

Demographics and clinical characteristics GI- and P-NENs by age group.

	≤50 yo (Group A)	50 <age < 70yo (Group B)	≥70yo (Group C)
Group	No. (%)	No. (%)	No. (%)	P	q
GI-NENs	n = 237	n = 737	n = 457
Median age [range]	45 [10-50]	61 [51-69]	75 [70->89]	0	0
Sex				.05	0.05
Male	113 (48)	401 (54)	263 (58)
Female	124 (52)	336 (46)	194 (42)
P-NENs	n = 116	n = 264	n = 124
Median age [range]	42 [20-50]	61 [51-69]	74 [70->89]	0	0
Sex				.43	0.43
Male	66 (57)	151 (57)	79 (64)
Female	50 (43)	113 (43)	45 (36)

Abbreviations: GI, gastrointestinal; P, pancreatic; NEN, neuroendocrine neoplasm; yo, years old.

With regards to association of age of diagnosis with sex, in GI-NENs, the median age for females (n = 654) was 63-year old and the median age for males (n = 777) was 65-year old (P & q < .05), suggesting earlier age of onset in females than males. In P-NENs, there was no significant differences in median age in females (n = 208, median age: 60-year old) versus males (n = 296, median age: 61-year old).

Overall survival

Overall survival was significantly associated with age at diagnosis in GI-NENs and P-NENs and with sex in GI-NENs (P < .05; Figure 1). For the association of age with overall survival in GI-NENs, the median overall survival for Group A, B and C was 27.96, 23.13, and 16.69 months, respectively (P < .0001). For P-NENs, the median overall survival for Group A, B and C was 57.8, 31.7, and 26.2 months, respectively. Similarly, for GI-NENs, median overall survival for males and females was 17.5 and 25.5 months, respectively.

Figure 1.

Overall survival according to Kaplan–Meier estimates among patients with GI-NENs by age (A) and sex (C) and P-NENs by age (B) and sex (D).

Molecular landscapes associated with age in GI- or P-NENs

Key alterations associated with age in patients with GI- or P-NENs are illustrated in Figure 2A and B. In GI-NENs, TP53 and RB1 mutations were more prevalent and APC mutations were less prevalent with increasing age (TP53 at 27.35% vs 37.26% vs 44.24%, P < .05, q < 0.05, RB1 at 13.44% vs 20.71% vs 23.23%, P < .05, q > 0.05, and APC at 14.77% vs 13.31% vs 9.47%, P < .05, q > 0.05 in groups A, B and C, respectively; Figure 2A).

Figure 2.

Molecular alterations in patients with GI-NENs associated with age (A) and sex (C) and in patients with P-NENs associated with age (B) and sex (D). Significance is represented as follows: Dashed lines: statistically trending with asterisk(s) reflective of the corresponding P-value; Solid line: statistically significant with asterisk reflective of the corresponding q-value; For asterisks above dashed lines: *: P < .05; **: P < .01; *** P < .001; ****: P < .0001; For asterisks above solid lines: *: q < 0.05.

In P-NENs, TP53, RB1, KRAS, and SMAD4 mutations were more prevalent with increasing age while MUTYH and NTHL1 mutations were less prevalent with increasing age (TP53 at 16.38% vs 27.95% vs 28.69%, P < .05, q > 0.05, RB1 at 4.6% vs 10.13% vs 18.56%, P < .05, q > 0.05, KRAS at 4.31% vs 12.50% vs 16.26%, P < .05, q > 0.05, SMAD4 at 0.86% vs 5.77% vs 6.56%, P < .05, q > 0.05, MUTYH at 6.96% vs 1.15% vs 0%, P < .05, q > 0.05, and NTHL1 at 6.78% vs 1.72% vs 0%, P < .05, q > 0.05 in groups A, B, and C, ­respectively; Figure 2B).

Molecular landscapes associated with sex in GI- or P-NENs

Key alterations associated with sex in patients with GI- or P-NENs are illustrated in Figure 2C and D. Mutations in TP53, FBXW7, TERT (promoter), and genomic loss of heterozygosity (gLOH) were more prevalent among males than females with GI-NENs (TP53 42.4% vs 32.4%, FBXW7 4.64% vs 2.58%, TERT 3.99% vs 0.93%, LOH 10.37% vs 5.57%, all with P < .05, q > 0.05 in males and females, respectively). PIK3CA mutations and mismatch repair deficient/microsatellite instability high (dMMR/MSI-H) were more prevalent in females than males with P-NENs (PIK3CA 1.92% vs 0%, dMMR/MSI-H 2.94% vs 0.34% all with P < .05, q > 0.05 in males and females, respectively).

Immune microenvironment associated with age and sex

Immune checkpoint gene expression (ICG) analysis can aid in developing immunotherapies for cancer by allowing us to identify potential immune checkpoint targets. Comparing ICG expression profile for GI-NENs by age of diagnosis, we demonstrate that LAG3, CD80 and HAVCR2 increased with increasing age at diagnosis [LAG3 0.887 vs 1.001 vs 1.094, CD80 1.527 vs 1.672 vs 1.811, and HAVCR2 2.984 vs 3.071 vs 3.169 all with P < .05, q > 0.05 in Groups A, B and C respectively, with medians reported in log₂(TPM + 1); Figure 3A]. Furthermore, ICG expressions of LAG3 was higher in males than females with GI-NENs (1.09 vs 0.97 with P < .05, q > 0.05, in males and females, respectively; Figure 3C). Immune cell estimates of anti-tumorigenic M1 macrophages in GI-NENs increased with increasing age (0.006 vs 0.008 vs 0.01, P < .05, q > 0.05 in Groups A, B and C, respectively) while M2 macrophages (pro-tumorigenic) decreased with increasing age (0.042 vs 0.036 vs 0.033, P < .05, q > 0.05 in Groups A, B, and C, respectively; Figure 4A). No associations of immune cell estimates with sex were found (Figure 4C).

Figure 3.

Boxen plots of the log of the median value of immune related gene transcript in patients with GI-NENs by age (A) and sex (C) and in patients with P-NENs by age (B) and sex (D). Significance is represented as follows: Dashed lines: statistically trending with asterisks reflective of the corresponding P-value; *: P < .05.

Figure 4.

Boxen plots of % of immune cell infiltrates in patients with GI-NENs by age (A) and sex (C) and in patients with P-NENs by age (B) and sex (D). Significance is represented as follows: Dashed lines: statistically trending with asterisks reflective of the corresponding P-value; *: P < .05; Abbreviation: TME: tumor microenvironment.

For P-NENs, no differences in ICG expression were found with age or sex (Figure 3B and D). With regards to immune cell estimates, M2 macrophages were found to decrease with increasing age (0.049 vs 0.041 vs 0.037, P < .05, q > 0.05 in Groups A, B, and C, respectively) while CD4+ T cells were found to increase with increasing age (0.063 vs 0.14 vs 0.141 P < .05, q > 0.05 in Groups A, B, and C, respectively; Figure 4B). There were no associations of immune cell estimates with sex (Figure 4D). A summary of the statistically significant or statistically trending results is illustrated in Figure 5.

Figure 5.

Summary of the mutational, ICG expression, and immune cell infiltrate differences with age and sex in patients with GI- and P- NENs. Of note, bold text represents statistically significant (P < .05, q < 0.05), while regular text represents statistically trending (P < .05, q > 0.05).

Discussion

Our study demonstrates distinct age- and sex- associated molecular profiles in both GI- and P-NENs as well as distinct age- associated immune profiles in GI-NENs. Briefly, in GI-NENs, mutations such as TP53 and RB1and ICG expressions of LAG3, CD80, and HAVCR2 increased with increasing age while APC mutations decreased with increasing age. In P-NENs, mutations such as KRAS and SMAD4 increased with increasing age while MUTYH and NTHL1 mutations were less prevalent with increasing age. In GI-NENs, TP53 and FBXW7 mutations were increased among males while in P-NENs, PIK3CA mutations were increased in females.

NENs are a diverse group of tumors that exhibit variability in their appearance, cellular differentiation, and proliferation capacity. Busico et al.’s retrospective series examined the molecular heterogeneity of poorly differentiated NENs and found that markers of TP53, KRAS, and RB1 were significantly enriched.¹² Specifically, TP53 mutations concurrent with RB1 mutations were more prevalent in high-grade, poorly differentiated NENs. Our study also found an association between TP53/RB1 mutations and NENs of increasing age, as observed in both the GI and P cohorts. Given that RB1/TP53 mutations are associated with more aggressive NENs, it is possible to infer that late-onset NENs may have worse outcomes and require a different set of treatment targets.

Our study found that KRAS mutations tend to be more prevalent with increasing age in P-NENs. Hijioka et al. showed that the KRAS mutation was detected in about half of their cohort of high-grade P-NEN patients, with a poorer response to platinum-based regimen seen in patients lacking these mutations.¹³ In addition, previous studies have shown that young-onset (before age 50) P-NENs have better survival outcomes,¹⁴ and our findings suggest that KRAS expression related to age may have a part to play in this. Taken together, our study provides important insights into the molecular heterogeneity of P-NENs, and points to age being a risk factor for aggressive disease.

We found differences in immune cell estimates in GI-NENs, with younger patients showing higher estimates of M2 macrophages and older patients showing higher estimates of M1 macrophages. M1 macrophages tend to be proinflammatory and anti-tumorigenic, with a role in inhibiting tumor growth through processes such as directly phagocytosing cancerous cells. M2 macrophages, by contrast, are anti-inflammatory and pro-tumorigenic, often promoting tumor growth and metastasis via remodeling, angiogenesis, and tissue repair.¹⁵ These differences may fall in line with our prior results of potentially higher-grade NENs being more common in older patients (based on mutation profile).

In our sex-based analysis, we observed a higher prevalence of dMMR/MSI-H in females than males with P-NENs. This has not been previously reported and requires further validation given the potential important therapeutic implications. However, the minimal differences noted, while statistically significant, may not be clinically significant, implying that sex may not be a major determinant of molecular variation in NENs.

The strength of this study is its design, which evaluates sex- and age- based molecular disparities in a large, multi-center cohort of NENs and correlates these findings with survival. Key limitations include its retrospective nature and absence of tumor grade data, such as Ki-67 score and differentiation details. In this setting, we are unable to determine whether age-associated molecular patterns, such as the elevations in TP53, RB1, and KRAS, reflect intrinsic molecular biology associated with demographics, or if they represent higher prevalence of high-grade tumors in older patients. Therefore, these results should be interpreted with caution. We also have limited information on timing of tumor profiling as compared to time of diagnosis and treatment characteristics of this patient cohort which may mean certain molecular signatures have changed over time since initial diagnosis.

Finally, because we did not validate our immune findings with histologic or single-cell approaches, the macrophage differences we report should be considered exploratory, and future studies incorporating tissue-level or single-cell validation will be essential to confirm these observations.

Growing evidence suggests that demographic variables influence outcomes in NENs. Tan et.al. recently demonstrated sex-based differences in the demographics (race and primary payer) and diagnoses present during hospital stay (gastrointestinal, headache, stomatitis and mucositis) among the 2019 National Inpatient Sample of patients with NENs.¹⁶ Their work highlights that sex may not only shape presentation and care patterns, but also the underlying tumor biology. Building on this, this study set out to investigate whether molecular differences parallel this observed clinical difference and has identified age- and sex- associated differences in mutational profile and immune checkpoint gene expression across GI- and P-NENs.

This data has potential therapeutic implications down the road. Currently, the presence of a TP53 mutation in neuroendocrine tumors does not change management.¹⁷ Meanwhile, the presence of TP53 mutation changes management for other cancers such as chronic lymphocytic leukemia, in which the presence of del(17p)/TP53 mutation changes the treatment regimen from chemotherapy to Bruton’s tyrosine kinase inhibitors such as acalabrutinib or zanubrutinib.¹⁸ With additional validation studies, the presence of TP53 mutation in NENs may eventually be hypothesis-generating for future clinical trials.

Conclusion

Our study reveals significant insights into the molecular and immune landscapes of neuroendocrine neoplasms (NENs) influenced by age and sex. We identified distinct molecular profiles, such as increased TP53 and RB1 mutations in older patients and higher KRAS mutations in P-NENs with age, indicating age as a key factor in tumor aggressiveness. We also noted differences in immune cell infiltration and molecular pathways driving early- versus late-onset NENs, highlighting the need for further exploration of treatments and biomarkers specific to these age groups. Finally, we noted molecular alterations such as TP53 and FBXW7 were more prevalent among males than females with GI-NENs and other alterations such as PIK3CA mutations and dMMR/MSI-H were more prevalent in females than males with P-NENs, identifying potential sex-specific molecular targets. Overall, our research underscores the importance of examining molecular profiling through the lens of key demographic factors like age and sex. Leveraging genomic profiling can improve therapeutic strategies and patient outcomes. Future studies should continue exploring these demographic differences to refine treatment approaches for patients with NENs.

Author contributions

Shuning Li (Data curation, Writing—review & editing), Lisa Liu (Data curation, Formal analysis, Writing—original draft, Writing—review & editing), Nishant Gandhi (Data curation, Formal analysis, Writing—original draft, Writing—review & editing), Alex Farrell (Data curation, Formal analysis, Writing—review & editing), Emil Lou (Data curation, Formal analysis, Writing—review & editing), Heloisa Soares (Data curation, Formal analysis, Writing—review & editing), Bassel Nazha (Data curation, Formal analysis, Writing—review & editing), Jeffrey Swensen (Data curation, Formal analysis, Writing—review & editing), Matthew Oberley (Data curation, Formal analysis, Writing—review & editing), Mark G. Evans (Data curation, Formal analysis, Writing—review & editing), Pamela Kunz (Conceptualization, Formal analysis, Methodology, Writing—original draft, Writing—review & editing), and Namrata Vijayvergia (Conceptualization, Formal analysis, Methodology, Project administration, Writing—original draft, Writing—review & editing)

Funding

Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under Award Number P30CA006927.

Conflicts of interest

N.V: Honoraria for Ad board - Exelixis, Novartis, Boerhinger Ingram, Pfizer, Jazz Pharma; Reseach Support- ITM, Exelixis, Arcus BioSciences, Puma Therapeutics. N.G.: Employed by and own stock in Caris Life Sciences. H.S.: Honoraria for ad board related to Ipsen, tersera, Novartis, ITM, Exelixis. M.G.E.: Caris Life Sciences (full-time employment, stock/stock options, travel/speaking expenses).

Data Availability

Summary level data may be obtained by contacting the corresponding author of the study.