Targetable landscape in pancreatic cancer varies greatly by KRAS status
Click Here to Manage Email Alerts
CHICAGO — Potentially targetable oncogenic rearrangements are more prevalent among patients with KRAS wild-type pancreatic cancer than those with KRAS-mutated disease, according to study results presented at ASCO Annual Meeting.
Homologous recombination deficiency, high tumor mutation burden and microsatellite instability-high disease also appeared more common in the KRAS wild-type population.
In addition, early-onset pancreatic cancers — those diagnosed before age 50 years — appeared more likely to be KRAS wild-type, exhibit homologous recombination deficiency phenotype and have germline BRCA1/BRCA2 alterations.
“When we initiated this project, one of the clinically meaningful outcomes we aimed to achieve was to identify and quantify the rates of potentially targetable alterations in hopes of informing development of new therapies and clinical trials,” researcher Todd C. Knepper, PharmD, associate member in the department of individualized cancer management at Moffitt Cancer Center and a HemOnc Today Editorial Board member, told Healio.
Background and methods
Pancreatic adenocarcinoma is the fourth leading cause of cancer mortality in the United States.
Incidence is increasing among individuals aged younger than 50 years, and early-onset pancreatic cancers may have a unique biology.
Results of prior small cohort studies suggest these tumors may be more likely than later-onset pancreatic cancer to be KRAS wild-type and harbor enrichment of targetable mutations or fusions.
In addition, therapies that target oncogenic fusions in KRAS wild-type tumors have induced meaningful responses, according to study background.
“There is increasing recognition of molecular differences between KRAS-mutated and KRAS wild-type pancreatic cancers,” Knepper said. “We previously reported our single-center experience with patients with pancreatic cancer who had next-generation sequencing, and we identified an exclusivity of potentially targetable gene fusions among patients with KRAS wild-type disease, including one patient who was successfully treated and achieved complete response to therapy on the basis of a MET fusion. We wanted to explore this further in a much larger cohort.”
Knepper and colleagues assessed the prevalence of fusions, mutations and homologous recombination deficiency in KRAS wild-type pancreatic cancer, with the goal of identifying potential treatment targets. They also aimed to assess differences between early-onset and non-early-onset pancreatic cancers.
The retrospective analysis included de-identified records from 4,956 patients (median age, 67 years; range, 59-73; 53% men; 83% white) with pancreatic cancer. Among the 3,952 patients with known disease stage, 78% had stage IV disease and 22% had stage I/II/III.
All patients’ tumor biopsies underwent sequencing with the Tempus xT (Tempus) solid tumor assay.
Researchers detected fusions from RNA sequencing data via the Tempus bioinformatics pipeline. They identified germline and somatic single-nucleotide mutation variants and insertions/deletions.
Investigators used RNA sequencing data from the Tempus HRD test (Tempus) to determine homologous recombination deficiency status.
Results by KRAS status
Researchers determined 79% of patients had KRAS-mutant tumors and 21% had KRAS wild-type tumors.
Homologous recombination deficiency status appeared more than twice as prevalent in KRAS wild-type than KRAS-mutated tumors (7.9% vs. 3.1%; P < .001).
Results showed several statistically significant differences in the somatic mutational landscape between KRAS wild-type and KRAS-mutated cohorts. These included pathogenic/likely pathogenic mutations in TP53 (21.4% vs. 79.3%), CDKN2A (5.4% vs. 25.4%), SMAD4 (4.5% vs. 20.3%), ARID1A (3.6% vs. 8.6%), RNF43 (1.7% vs. 4.9%), KMT2C (1.4% vs. 4.5%), KDM6A (0.5% vs. 4.1%), KMT2D (1.8% vs. 3.7%), MEN1 (2.2% vs. 0.2%) and BRAF (5% vs. 0.2%).
Knepper and colleagues determined actionable rearrangements were enriched among patients with KRAS wild-type tumors (10% vs. 2.1%; P < .001). These included NRG1, (3% for KRAS wild-type vs. 1.5% for KRAS-mutated), BRAF (2% vs. 0%), FGFR2 (1% vs. 0.1%), NTRK3 (0.6% vs. 0.1%), RET (0.6% v. 0%), NTRK1, (0.5% vs. 0%), RAF1 (0.4% vs. < 0.1%), FGFR1 (0.4% vs. < 0.1%), and MET (0.2% vs. 0%).
The most common actionable fusion gene pairs included LDAH-NRG1 (2.5% for KRAS wild-type vs. 1.4% for KRAS-mutated), SND1-BRAF (0.6% vs. 0%), BRAF-CCNY (0.3% vs. 0%), ETV6-NTRK3 (0.3% vs. 0%) and NTRK3-EML4 (0.2% vs. < 0.1%).
Researchers also reported a higher prevalence of tumor mutational burden-high status (2.9% vs. 1.7%) — defined as at least 10 mutations/megabase — and microsatellite instability-high status (1% vs. 0.4%) in the KRAS wild-type cohort (P = .015).
“These results validate what we and others have previously shown in terms of increased rates of fusions in KRAS wild-type pancreatic cancer,” Knepper said. “One of the key takeaways for me is the diversity of gene fusions that occur in this population. It’s not just one recurrent fusion; there are several and they occur at very different rates.”
Results by age at onset
Researchers characterized 382 patients as having early-onset pancreatic cancer. KRAS wild-type tumors occurred more frequently among those with early-onset disease (30%) than those with average onset (22%) — defined as 50 to 70 years — or those with late onset (17%), defined as older than 70 years.
Patients with early-onset pancreatic cancers also were more likely than those with average or late onset to have germline alterations in BRCA1 (2% vs. 0.4% vs. 0.4%; P = .015) or BRCA2 (4.5% vs. 1.7% vs. 0.8%; P < .001).
Researchers reported no statistically significant differences in actionable genes between the early-onset cohort and the average-onset and late-onset cohorts.
Homologous recombination deficiency occurred more frequently among patients with early-onset pancreatic cancer (5.4%) than average onset (4.7%) or late onset (2.5%; P = .047).
Rearrangements also occurred more frequently in the KRAS wild-type early-onset cohort than the KRAS-mutated early-onset cohort (10% vs. 1.1%; P < .001).
Next steps
For many cancer types, established molecular alterations confer potential eligibility for targeted therapies.
“That is not currently the case for pancreatic cancer, so it’s not quite standard of care in all cases to do next-generation sequencing,” Knepper said. “Now that we are gaining a deeper understanding of the alterations that occur in pancreatic cancer, it does change the conversation.”
Because gene fusions are one of the more common potentially targetable alterations that can be detected, it is important for physicians and patients who opt for tumor sequencing to use an assay that has RNA sequencing capability to ensure improved detection of gene fusions, Knepper said.
Perspective
Back to Top
Igor Astsaturov, MD, PhD
The significance of this work is that genetic testing of pancreatic tumors allows the identification of rare variants for which existing drug therapies already are available. Further, it shows a dramatic heterogeneity — and yet commonalities — for pancreatic tumors. Addressing the pathogenic mutations in KRAS and TP53 will remain the biggest challenge for us in fight against this cancer.
Collapse
Knepper TC, et al. Abstract 4155. Presented at: ASCO Annual Meeting; June 3-7, 2022; Chicago.
Click Here to Manage Email Alerts