The data from a wide range of prospective, randomized phase 3 trials are clear: There is a significant survival benefit from adjuvant cisplatin-based chemotherapy for patients with resected higher-risk non–small-cell lung cancer (NSCLC), often defined as node-positive and/or with the primary tumor measuring 4 cm or greater in diameter.1 As in many clinical settings, however, patients in real-world practice are not well represented by those on the relevant clinical trials, as those in the pivotal adjuvant trials are approximately a decade younger and with a near uniformly good performance status. A clinical trial directly comparing preoperative with postoperative chemotherapy for early-stage NSCLC also demonstrated an attrition rate of 34% for postoperative chemotherapy within the protocol-defined window of 6 to 7 weeks owing to a wide range of complicating factors.2 In practice, this means that many patients who would otherwise be appropriate candidates for adjuvant chemotherapy do not present for consideration of treatment until later than best practices would recommend. Should an intervention with an established survival benefit be deferred in this setting? While we do not have prospective evidence to speak to this, absence of proof is not proof of absence.
Every clinic day, oncologists and other physicians face the challenge of how far to extrapolate from the results of clinical trials with often narrowly defined clinic populations to a broader population of patients in our clinics. Here, the study by Salazar and colleagues,3 a retrospective analysis of survival among 12 473 postoperative patients with results captured in the National Cancer Database, and with the aforementioned criteria defining appropriate level of risk of recurrence, compared survival outcomes for patients who initiated adjuvant chemotherapy early (18-36 days), late (57-114 days), or at a time between these intervals. The results demonstrated no hint of inferior survival among the patients who received chemotherapy in the later time period, beyond which we have data from clinical trials. Moreover, a Cox analysis of nearly 4000 propensity-matched pairs of patients demonstrated that adjuvant chemotherapy was associated with a lower mortality risk than surgery alone, including among patients who initiated adjuvant chemotherapy more than 8 weeks after surgery.
While retrospective data cannot define the benefit of delayed adjuvant chemotherapy with the clarity of a prospective randomized trial, we must remember that in the land of the blind, the one-eyed man is king; these limited data inject an evidence-based answer for a very common clinical question for which we have been forced by necessity to rely only on our best judgments. But real-world data from large populations also provide a valuable corroboration of data obtained in randomized clinical trials of potentially nonrepresentative patients. In the absence of prospective data for every patient, such retrospective data are valuable in offering the best evidence available for clinicians charged with the task of offering the best treatment recommendations for practical management of a broad population of patients that includes many who would not qualify for more restrictive clinical trials on which we rely to establish clinical standards.
Corresponding Author: Howard (Jack) West, MD, Swedish Cancer Institute, 1221 Madison St, Ste 200, Seattle, WA 98104 (firstname.lastname@example.org).
Conflict of Interest Disclosures: None reported.
West H(. Bringing Adjuvant Chemotherapy for Resected Non–Small-Cell Lung Cancer Into Real-world Practice—Better Late Than Never?. JAMA Oncol. 2017;3(5):619. doi:10.1001/jamaoncol.2016.5798