Saving the best to the last: Can we wait for second-line?

Abstract

284 Background: A common perception of some oncologists is that the vast majority of their patients with metastatic disease will receive second-line treatment upon progression. Therefore, “saving” good treatment options for the future may be acceptable. We aimed to examine whether this perception correlates with real-life. Methods: Using an oncology electronic database, consisting of >27,000 patients treated at our institution, we selected consecutive patients with metastatic or locally advanced lung, colon, pancreatic, bile duct and gastric cancers who started standard first-line. We then assessed the correlation between proceeding to second-line therapy and demographic and clinical variables, including age, gender, initial performance status, BMI, hemoglobin, WBC, creatinine, glucose, calcium, as well as duration on first line therapy and survival. Results: A total of 492 patients met the inclusion criteria. Their median age was 67 and 285 were men. Their diagnoses were colon (169), lung (102), pancreas (101), bile duct (65) and gastric (55) cancers. Only 52% (255) received second-line treatment for at least 30 days (36% colon, 26% lung, 18% pancreas, 9% bile duct and 11% gastric). Receipt of second-line therapy was associated with disease site (P=0.001) as well as with age, with patients who received second-line being 5 years younger compared to those who did not (64 vs. 69 years, P=0.004). Patients who reached second-line had better performance status and higher hemoglobin level at presentation, additionally their median duration on first-line chemotherapy was substantially longer (P<0.007 for all comparisons). Survival of patients not starting second-line was significantly shorter across all tumor types (19.8 vs 6.5 months P=0.001). General deterioration and toxicity were the major reasons for avoiding second-line therapy at progression, 43% and 30% respectively. Conclusions: These real-life data indicate that only half of the patients starting standard doublet or triplet treatment for advanced cancers will commence second-line therapy; and this can be hardly predicted in advance using standard clinical and laboratory characteristics. Our data challenge the practice of saving good treatment options for subsequent lines, and call for the development of tools enabling prediction of response and tolerance to treatment, pursuing for better patient selection and patient-tailored therapy.