Mutant KRAS status in non-small cell lung cancer (NSCLC) correlated with significantly worse progression-free survival (PFS) and resistance to platinum-based chemotherapy, data from a randomized trial showed.

KRAS mutations were associated with almost a 40% decrement in PFS compared with patients who had wild-type KRAS.

The outcome was consistent across all platinum-containing combinations the patients had received as first-line therapy, although patients treated with gemcitabine did worse than others, Marina C. Garassino, MD, reported here at the European Lung Cancer Conference.

“KRAS is a strong prognostic factor in first-line treatment of non-small cell lung cancer patients,” said Garassino, of the Fatebenefratelli and Ophthalmic Hospital of Milan. “The numbers are too small to determine whether there is a mutation with worse prognosis than others,” she added. “It is of major importance to develop new strategies for patients with KRAS mutations.” KRAS is the most frequently mutated gene in humans, and the mutation rate is especially high in pancreatic, colon, and lung cancers. Mutant KRAS is a necessary step for tumor initiation and progression in several preclinical models, and is widely considered to be “undruggable,” said Garassino. An Italian trial of second-line chemotherapy in NSCLC included prespecified substudy of EGFR and KRAS mutations and their impact on treatment (Clin Lung Cancer 2011; 12: 138-141). Garassino presented results of the KRAS analyses. Eligible patients who had received first-line platinum-based chemotherapy were randomized to second-line treatment with either erlotinib (Tarceva) or docetaxel. The primary endpoint of the parent trial was overall survival, and secondary endpoints included PFS and response rate. PFS was the primary endpoint of the substudy, and the secondary endpoint was the impact of the three most common KRAS mutations (G12C, G12V, and G12D) on prognosis. The 306 evaluable patients received first-line therapy consisting of cisplatin or carboplatin paired with gemcitabine, vinorelbine, or pemetrexed (Alimta). The most commonly used doublets were cisplatin-gemcitabine (N=135), cisplatin-pemetrexed (N=70), and carboplatin-gemcitabine (N=52). Analysis of KRAS mutations showed that G12C occurred in 36.56% of specimens, G12V in 22.58%, and G12D in 15.05%. The next most frequent mutation was G12A (9.68%). A majority of patients (253) had adenocarcinoma histology, followed by squamous in 80 cases. Overall, 93 tumors harbored KRAS mutations, most of which (71 tumors, or 76.34%) occurred in adenocarcinomas. Mutations were uncommon in squamous-cell tumors (5 of 80, 5.38%). Data on tumor grade were available for 249 patients, including 60 patients with KRAS mutations. Garassino reported that 40 of the 60 mutations were associated with grade 3 disease. The primary analysis showed that patients with KRAS mutations had a significantly increased risk of progression compared with patients who had wild-type KRAS (HR 1.38, P=0.041). “This means that patients with a KRAS mutation had a 38% increase in the probability of progression during first-line treatment with platinum-based chemotherapy,” said Garassino. Analysis of PFS by first-line regimen showed that KRAS mutations increased the progression hazard for all regimens, but the effect appeared more pronounced with gemcitabine, an observation that did not have an obvious explanation, she said. Comparison of PFS by the type of KRAS mutations was not informative because of small numbers. However, Garassino pointed out a possible trend toward worse outcomes in patients who had the G12V mutation. In a multivariate analysis that included performance status and histology, KRAS mutations remained a significant predictor of worse PFS (P=0.03). The impact of mutations was unaffected by age, sex, performance status, or tumor grade. Consistent with previously reported observations, KRAS mutations were associated with current smoking (P=0.0036). However, Garassino noted that 11% of nonsmokers also had KRAS mutations. Mutations also were associated with lower body mass index at diagnosis (P=0.013) and adenocarcinoma histology (P<0.0001). Invited discussant Tetsuya Mitsudomi, MD, of Aichi Cancer Center in Nagoya, Japan, took issue with Garassino's conclusion that KRAS mutations have a strong association with worse PFS. Citing the absence of a nonchemotherapy control group and lack of uniformity in chemotherapy regimens, Mitsudomi said that distinguishing between predictive and prognostic implications was impossible. "KRAS mutations may be weakly prognostic," he said. "KRAS mutations may be predictive for platinum-doublet therapy, but the data are conflicting." "KRAS gene testing can be used to exclude other driver gene mutations, but making treatment decisions based on KRAS is not recommended," Mitsudomi added.