Percentage Change in Plasma Cytokeratin 18 Is Associated with Clinical Outcomes in Patients Receiving Pemetrexed and Carboplatin for the Adenocarcinoma Subtype of NSCLC.

BACKGROUND: The adenocarcinoma subtype of non-small cell lung cancer (adeno-NSCLC) is routinely treated with chemotherapy if patients do not have molecular aberrations such as epidermal growth factor receptor mutations or anaplastic lymphoma kinase rearrangements. There are currently no validated biomarkers that can predict if patients will gain clinical benefit from chemotherapy, leading to a majority of patients receiving many cycles of unnecessary chemotherapy. We hypothesized that the percentage rise in plasma caspase-cleaved cytokeratin 18 (cCK18) and total cytokeratin 18 (tCK18) assessed before and after chemotherapy correlates with the radiological response to chemotherapy. METHODS: Plasma samples from 40 patients with stage IV adeno-NSCLC, treated with first-line chemotherapy with carboplatin (AUC5) plus pemetrexed (500 mg/m(2)), were collected prior to chemotherapy and 48 h after treatment. ELISA was used to quantify cCK18 and tCK18. RESULTS: The male-to-female ratio was 3:1, and the median age of patients was 63 years. Patients who had a clinical benefit (complete response, partial response or stable disease) at the first radiological assessment following chemotherapy had a significantly higher percentage change in plasma tCK18 levels compared to those who had no clinical benefit, i.e. progressive disease (69.5 ± 75.1 vs. 25.3 ± 30.9%, respectively; p = 0.042). The receiver operating characteristic area was 0.712 (p = 0.039). There was an increase in the percentage change in cCK18 in patients with clinical benefit compared to those without clinical benefit but this was not statistically significant (57.6 ± 112.8 vs. 24.38 ± 45.1%, respectively; p = 0.85). CONCLUSIONS: The percentage change in plasma tCK18 levels before and after the first cycle of pemetrexed and carboplatin chemotherapy is associated with clinical benefit. If validated in larger cohorts, this test can be used to identify patients unlikely to respond to treatment who can thus be offered alternative treatments or entry into clinical trials.