Cross-sectional imaging utilization for EGFR, EML/ALK, and KRAS mutant non-small cell lung cancer.

Thomas Bomberger,Elias Kikano,Kai Laukamp,Daniel Arnold Smith,Amit Gupta,Kianoush Ansari-Gilani,Afshin Dowlati,Nikhil H. Ramaiya
DOI: https://doi.org/10.1200/JCO.2019.37.15_suppl.e20663
IF: 45.3
2019-10-07
Journal of Clinical Oncology
Abstract:e20663 Background: Targeted therapies for specific genotypic subsets of non-small cell lung cancer (NSCLC) patients have had drastic effects on overall survival at all stages. However, the NCCN guidelines have remained relatively static with respect to initial evaluation, staging, response, and imaging surveillance. In this study we evaluated 3 cohorts of patients, two with a specific targetable mutation (EGFR and EML/ALK) and one without a targetable mutation available (KRAS), for cross-sectional imaging utilization in the initial presentation and surveillance of disease. Methods: Retrospective analysis of imaging studies for patients with NSCLC who received treatment and follow-up between 2007 and 2018 was completed. CT, MRI, and PET imaging data were collected from the picture archiving system. Additional clinical data collected included age, race, smoking history, and dates of diagnosis and death. Results: 153 NSCLC patients were identified with positive EGFR (n = 65), EML/ALK (n = 15), or KRAS (n = 73) mutations. Patients averaged an age of 64 at diagnosis, with a distribution of 38% male, 62% female, 29% current smokers, 48% former smokers, and 23% never smokers. Mean duration between date of diagnosis and death for each group was 1099 ± 770 days (EGFR), 1139 ± 924 days (EML/ALK), and 542 ± 602 days (KRAS). Compared to the KRAS group, the EML/ALK group had higher utilization of chest CTs (p = 0.044), pulmonary embolus CTs (p = < 0.001), abdominopelvic CTs (p = 0.004), and PET CTs (p = 0.006). The EGFR group had significantly fewer pulmonary embolus CTs (p = < 0.001) and higher numbers of non-brain MRIs (p = 0.004) compared to the KRAS group, as well as fewer abdominopelvic CTs (p = 0.034) compared to the EML/ALK group. Conclusions: Different genotypes of NSCLC with available targetable mutations show statistically significant differences in cross-sectional imaging utilization compared to those without targetable mutations. Mean Values CT Head CT Chest 2 CT PE 1,3 CT Abd/Pel 2,3 CT PET 2 MRI Head Other MRI 1 EGFR (n = 65) 0.74 4.45 0.28 2.37 3.43 3.34 1.12 EML/ALK (n = 15) 0.6 4.67 1.33 4.93 4.33 5.8 1.13 KRAS (n = 73) 1.03 3.42 0.9 1.42 1.95 2.62 0.39 1, 2, 3–Significant difference between EGFR and KRAS (1), EML/ALK and KRAS (2), or EGFR and EML/ALK (3)
oncology
What problem does this paper attempt to address?