Veristrat as Predictor of Benefit of First Line Non Small Cell Lung Cancer (NSCLC) Patients From Standard Chemotherapy

VeriStrat®, a pretreatment blood-based test correlated with clinical outcome after EGFR-TKI therapy in non-small cell lung cancer (NSCLC) patients, was developed and validated in a multi-institutional study of advanced NSCLC patients treated with gefitinib \[Taguchi et al\] . The VeriStrat algorithm was developed using a training set of pre-treatment serum samples from patients who then experienced either long term stable disease or early progression on gefitinib therapy. Mass spectra from these patients' serum samples were used to define eight MS features (i.e. peaks), differentiating these two outcome groups. An algorithm (VeriStrat) utilizing these features and based on a k-nearest neighbors (KNN) classification scheme was created and its parameters were optimized using additional spectra from the training cohort. All aspects of VeriStrat were frozen after development. VeriStrat assigns each spectrum a "Good" or "Poor" label. VeriStrat was validated in a blinded fashion on two independent cohorts of patients who were treated with gefitinib or erlotinib. These studies confirmed that patients classified as "Good" had better outcome than patients classified as "Poor" (HR of death = 0.47 P = 0.0094 in one cohort, HR of death = 0.33 P = 0.0007 in the other). In the original study, VeriStrat was shown to correlate with clinical outcome following EGFR-TKI therapy, but not in the chemotherapy or post-surgery setting: No statistically significant difference was seen in the overall survival of patients classified as "Good" or "Poor" from the serum from patients collected before second-line chemotherapy (HR = 0.74, 95%, P = 0.42 in one cohort (cohort B) and HR = 0.81, P = 0.54 in another (cohort C)). In a third control cohort of patients with resected early-stage NSCLC, the hazard ratio for overall survival was 0.90 (P = 0.79). However, further analysis of the subsets of chemotherapy samples demonstrated that separation between "Good" and "Poor" arms may depend on a particular type of chemotherapy. Thus a retrospective subset analysis of the cohort C showed that while patients treated with docetaxel in second line did not show any sign of separation, patients receiving a combination of platinum-based agents with either vinorelbine or gemcitabine or paclitaxel had a trend to separation between the two arms. The working hypothesis for the mechanism is that the VeriStrat "Poor" label is related to the activation of canonical and non-canonical MAPK pathways downstream from receptor tyrosine kinases, with possible cross-talk activation of the NF-kB pathway. This means that VeriStrat may demonstrate different predictive performance depending on the particular chemotherapy treatment and its associated with cell pathway interactions.