Proteomic Signature for EGFR Inhibitor Therapy Predicts Survival Benefit of Second-Line Chemotherapy vs Erlotinib in NSCLC
There are conflicting data on whether epidermal growth factor receptor (EGFR) inhibitor therapy is beneficial in second-line treatment of lung cancer patients with unknown or wild-type EGFR status. In a phase III trial (PROSE) reported in The Lancet Oncology, Gregorc et al assessed the predictive value of a proteomic signature serum protein test for likely outcome of EGFR inhibitor therapy in non–small cell lung cancer (NSCLC) patients receiving second-line therapy with the EGFR inhibitor erlotinib (Tarceva) vs chemotherapy. They found that the test was predictive of differential survival benefit for erlotinib vs chemotherapy, with patients classified by the test as likely to have poor outcome on EGFR inhibitor therapy having better outcome on chemotherapy.
The serum protein test (available as VeriStrat, Biodesix) uses mass spectrometry to identify patients likely to have good or poor survival on EGFR tyrosine kinase inhibitors. Test status is categorized as good or poor by comparison of the intensity of eight regions in mass spectra obtained from pretreatment serum samples with intensity of those of a reference set.
Study Details
In the trial, 285 patents with histologically or cytologically confirmed second-line stage IIIB or IV NSCLC at 14 centers in Italy were randomly assigned between February 2008 and April 2012 to receive erlotinib at 150 mg/d or chemotherapy (IV pemetrexed [Alimta] at 500 mg/m2 or docetaxel at 75 mg/m2 every 21 days).
All patients underwent proteomic testing prior to treatment. Patients were stratified by Eastern Cooperative Oncology Group (ECOG) performance status, smoking history, center, and masked pretreatment serum protein test classification. The proteomic test classification was masked for patients and treating investigators, and treatment allocation was masked for investigators who determined the proteomic classification.
The primary endpoint was overall survival, and the primary hypothesis was that there would be a significant interaction between the serum protein test classification and treatment. Analyses were performed in the per-protocol population.
A total of 129 chemotherapy patients and 134 erlotinib patients were included in the per-protocol population; of these, 68% and 72% had a proteomic test classification of good. The groups were otherwise generally balanced for age (median, 64 and 66 years), sex (71% and 74% male) ECOG performance status (0 in 50% and 54%, 1 in 43% and 40%), smoking status (13% and 16% never, 58% and 57% former, 29% and 27% current), histology (adenocarcinoma in 71% and 57%, squamous in 12% and 23%), stage (IIIB in 13% and 9%, IV in 85% and 90%), and EGFR status (mutated in 5% and 6%, wild-type in 65% and 59%, not available in 30% and 35%).
Compared with the 79 patients with a test classification of poor, the 184 patients with test classification of good were more likely to have better ECOG performance status (P = .002), to be female (P = .007), and to be never-smokers (P = .052).
Significant Interaction of Treatment and Test Classification for Overall Survival
Median overall survival was 9.0 months in the chemotherapy group and 7.7 months in the erlotinib group, with no significant difference between groups on analysis adjusted for stratification factors (adjusted hazard ratio [HR] = 1.22, P = .148).
There was a significant interaction between treatment and proteomic classification in both unadjusted (P = .031 for interaction) and adjusted analysis (P = .017 for interaction). Patients with a test classification of poor had significantly worse overall survival on erlotinib vs chemotherapy (median, 3.0 vs 6.4 months, adjusted HR = 1.72, P = .022), whereas there was no difference between the two groups among patients with a test classification of good (median 11.0 vs 10.9 months, adjusted HR = 1.06, P = .714).
Independent Predictors
The proteomic test classification by treatment interaction remained a significant independent predictor of overall survival (HR = 1.98, P = .022) after adjustment for other clinical characteristics and EGFR mutation status, as did proteomic test classification (HR = 1.88, P = .003, for poor vs good) and ECOG performance status (HR = 2.67, P = .0003, for 0 or 1 vs 2).
Third-line treatment included erlotinib in 26% of patients in the chemotherapy group and chemotherapy in 49% of patients in the erlotinib group; 59% and 49% received no third-line treatment.
The most common grade 3 or 4 adverse events were neutropenia in the chemotherapy group (15% vs < 1%) and skin toxicity in the erlotinib group (16% vs < 1).
The investigators concluded, “Our findings indicate that serum protein test status is predictive of differential benefit in overall survival for erlotinib versus chemotherapy in the second-line setting. Patients classified as likely to have a poor outcome have better outcomes on chemotherapy than on erlotinib.”
Vanesa Gregorc, MD, of Ospedale San Raffaele, Milan, is the corresponding author for the Lancet Oncology article.
The study was supported by the Italian Ministry of Health, Italian Association of Cancer Research, and Biodesix. Julia Grigorieva, PhD, Maxim Tsypin, PhD, Joanna Roder, PhD, and Heinrich Roder, PhD, are employees of Biodesix.
The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.
