Analysis of Time-to-treatment Discontinuation of Targeted Therapy, Immunotherapy, and Chemotherapy in Clinical Trials of Patients With Non-Small-Cell Lung Cancer

G. M. Blumenthal; Y. Gong; K. Kehl; P. Mishra-Kalyani; K. B. Goldberg; S. Khozin; P. G. Kluetz; G. R. Oxnard; R. Pazdur


Ann Oncol. 2019;30(5):830-838. 

In This Article


Selection Criteria

An internal FDA database was used to identify new or supplemental NDA or BLA submissions for the treatment of patients with mNSCLC submitted between 2007 and 2014. We found 18 randomized controlled trials, consisting of 37 treatment arms (see Table 1 for study characteristics)

End Point Definition

OS was defined as the time from random assignment to death. For patients who were alive at the cut-off date, OS was censored at the last follow-up date. PFS was defined as the time from random assignment to progression or death. Patients who were alive and who had not experienced progression as of the analysis cut-off date were censored at the last disease assessment. PFS was determined by RECIST version 1.0 or 1.1. TTD was defined as the time from random assignment until the last dose recorded or death. Patients alive who had not experienced treatment discontinuation as of the analysis cut-off date were censored at the analysis cut-off date.

Analysis Methods

First, individual trial-level Kaplan–Meier estimates of TTD and PFS were performed. For the pooled analysis, we divided trial arms based on therapeutic class assignment and/or molecular subtype (see Figure 1). Notably, five treatment arms were excluded because the regimen included a limited number of cycles of therapy with no scheduled maintenance therapy, thus confounding the interpretation of TTD. The remaining trial arms were grouped as follows:

Figure 1.

Diagram of patients included and excluded from the pooled analysis.

  1. EGFRm TKI—EGFR mutation (EGFRm)-positive patients receiving tyrosine kinase inhibitor (TKI) (n = 5 arms)

  2. ALK TKI—ALK+ patients receiving ALK TKI (n = 6 arms)

  3. EGFRwt TKI—EGFR wild-type patients receiving EGFR TKI (n = 2 arms)

  4. ICI—patients receiving ICI (n = 7 arms)

  5. Chemo doublet—patients receiving platinum doublet chemotherapy with maintenance chemo (with or without anti-EGFR mAB) (n = 4 arms)

  6. Chemo mono—patients receiving docetaxel monotherapy (with or without anti-VEGF mAB) (n = 8 arms).

To understand the relationship between TTD, PFS, and OS, we performed patient-level correlation analysis in all patients, as well as within the six therapeutic subgroups. We also conducted several sensitivity analyses, including an exposure analysis adding additional treatment duration based on the schedule of the therapeutic (adding 2 weeks for an every 2 week intravenously administered drug and adding 3 weeks for an every 3 week intravenously administered drug). In an additional sensitivity analysis, we moved up the TTD censoring window 2 weeks and 3 weeks from the data cut-off date to explore impact on TTD results.

Finally, an outlier analysis of cases where TTD exceeded PFS or where PFS exceeded TTD was performed. 'Late TTD' was defined as TTD that was at least 3 months longer than PFS. 'Early TTD' was defined as TTD date that was at least 3 months shorter than PFS date.