Genomic and Proteomic Profiling Selects Pancreatic Cancer Patients for Phase I Trial

In a few cases, this approach yielded a definite treatment benefit, reported Tamara Saurí, MD, of the Molecular Therapeutics Research Unit at Vall d'Hebron Institute of Oncology, at the 2016 Gastrointestinal Cancers Symposium.

"We saw an overall survival difference of 7 months for patients matched with a drug according to their mutation," said Saurí in an interview withTargeted Oncology.

Advanced pancreatic cancer has a "dismal" prognosis, the investigators noted. The hope is that molecular screening can identify genetic characteristics that lend themselves to targeted treatments, Saurí explained.

She and her colleagues performed molecular screening on archived tumor samples, analyzing them for selected gene mutations using mass spectrometry at first, then updating to next-generation sequencing. PTEN and PD-L1 expression levels were measured with immunohistochemistry. The aim of the molecular screening was to identify potentially targetable alterations in patients eligible for clinical trials with matched targeted therapies.

Between 2010 and 2014, they evaluated 86 patients with chemotherapy-refractory pancreatic cancer in this program. Of these, 61 (71%) had adequate tumor tissue, using formalin-fixed paraffin-embedded samples for profiling (ie, detection of specific molecular aberrations, either from the primary site or metastatic site in 16%). All demographic, treatment, and survival data were extracted retrospectively from electronic medical records.

Patients' median age was 60 years (range 29-78); 69% were diagnosed with early-stage disease and 31% with locally advanced or metastatic disease. Median treatment lines before profiling was 2 (range 1-5).

The most common potentially targeted alterations in the 61 patients wereKRASmutation (n = 42), with theG12Dmutation comprising 62% of theseKRASalterations; low PTEN expression was identified in 4 patients and PTEN loss in 9 patients;CDKN2AandBRAFwere mutated in 2 patients each; andRNF43andSTK11were mutated in one patient each.

"Most of these alterations were typical of the biology of pancreatic cancer," Saurí noted.

A total of 30 patients were enrolled in phase I trials following molecular profiling, 10 of them with a direct-matched therapy. These included 5 patients treated with a PI3K pathway inhibitor, 3 with a PI3K inhibitor plus a MEK inhibitor, one with anti-PD-L1 agent (a patient with high PD-L1 expression), and one with an HDM2 inhibitor (a TP53 wild-type patient).

One patient had a partial response to targeted treatment—a patient with PTEN loss who received a PI3K inhibitor plus chemotherapy. Stable disease was achieved in an additional 4 patients, including 3 receiving a PI3K inhibitor (with/without chemotherapy) and one with a PI3K inhibitor plus a MEK inhibitor.

One such patient was a 55-year-old man with stage IV disease, including liver metastasis, refractory to standard therapy, who had mutations inKRAS,Q61H, andTP53 G245fs,and PTEN loss. He was enrolled in the phase I trial and treated with a PI3K inhibitor plus paclitaxel 80 mg/m²weekly. After 4 cycles, he achieved a partial response, with 40% reduction in tumor. His progression-free survival was 6 months.

"We saw a big difference in the target lesions. He had much improved clinical features," Saurí noted.

Median time on treatment with matched therapy was 2.5 months for the trial population. Time to progression was 5.1 months.

Notably, patients treated on the phase I trial had a significantly longer time from recurrence or metastasis to death: 19.2 months, compared with 12.8 months for patients not entered on the matched-treatment trial (P= .02), she reported.

Saurí added that the targeting of pathways—blocking their effects—is probably more important in pancreatic cancer than targeting mutations. The main mutation—KRAS—she noted, actually has no available targeted agent.

References

1. Saurí T, Macarulla T, Cabrera G, et al. Molecular prescreening (MP) to treat patients (pts) with advanced pancreatic cancer (PC) in early clinical trials.J Clin Oncol, 34, 2016 (suppl 4S; abstr 272).