Weight loss affects outcome in KrasG12C non-small cell lung cancer first line treated patients.

Abstract

e21054 Background: Cachexia is a syndrome characterized by weight loss (WL), reduction of muscle mass, systemic inflammation and is an independent criteria of worse clinical outcomes in cancer. Patients with KRASmut non-small cell lung cancer (NSCLC) experience a higher rate of WL than KRASwt patients. However no data about the prognostic value of each Kras mutation subtype for WL and/or BMI status are available. Methods: We reviewed the clinicopathological data of 110 NSCLC patients harboring KRAS G12 mutation who were treated at Clinical Oncology Unit, Careggi University Hospital in Florence, from March 2016 to March 2020. Patients received either chemotherapy or anti-PD1 pembrolizumab or a combination of chemo-immunotherapy as first-line treatment. The prognostic value of basal BMI and WL (< and > 5%) was analyzed according to each mutation subtype. Results: In the study population, 49.1% of patients harbor the KRAS G12C mutation, 20.9% G12V, 18.2% G12D and 11.2% of patients have other subtypes of KRAS mutations. Overall, our data confirm the better prognostic value of G12C mutation. At a median follow-up of 18 months, median OS for G12C patients was 17.5 months, longer compared to any other mutation subtype ( p = 0.113), whereas G12V mutation was associated with significantly worse outcomes than in the whole population (mOS 9.5 months IC 95% 6.0-17.0, p = 0.018). WL within 6 months of first-line therapy was observed in 29.7% and 35.3% of G12C and G12V-mutated patients, respectively. WL statistically decreased survival outcomes both in G12C (mPFS 4.0 months vs NR, p < 0.001; mOS 12.0 months vs NR, p = 0.009) and in G12V patients (mPFS 3.0 months vs NR, p = 0.073; mOS 8.0 vs 12.0 months, p = 0.994), compared to patients who do not lose weight during treatment. Conclusions: WL significantly worsens clinical outcomes in G12C patients, with a similar trend in G12V. Further analyses in a larger population including a representative sample of less frequent mutations are needed.