Key Points
Question
In patients with long-term response to immunotherapy for advanced non–small cell lung cancer (NSCLC), should treatment be stopped at 2 years or continued indefinitely?
Findings
In this retrospective cohort study of patients on frontline immunotherapy-based treatment, only approximately 1 in 5 patients discontinued immunotherapy at 2 years in the absence of progression. On adjusted analysis, there was no statistically significant difference between patients who received a fixed duration of therapy for 2 years vs those who continued treatment indefinitely.
Meaning
For patients who are progression-free on immunotherapy for NSCLC, it is reasonable to stop therapy at 2 years, rather than continuing indefinitely.
Abstract
Importance
For patients with advanced non–small cell lung cancer (NSCLC) treated with frontline immunotherapy-based treatment, the optimal duration of immune checkpoint inhibitor (ICI) treatment is unknown.
Objective
To assess practice patterns surrounding ICI treatment discontinuation at 2 years and to evaluate the association of duration of therapy with overall survival in patients who received fixed-duration ICI therapy for 2 years vs those who continued therapy beyond 2 years.
Design, Setting, and Participants
This retrospective, population-based cohort study included adult patients in a clinical database diagnosed with advanced NSCLC from 2016 to 2020, who received frontline immunotherapy-based treatment. The data cutoff was August 31, 2022; data analysis was conducted from October 2022 to January 2023.
Exposures
Treatment discontinuation at 2 years (between 700 and 760 days, fixed duration) vs continued treatment beyond 2 years (greater than 760 days, indefinite duration).
Main Outcomes and Measures
Overall survival from 760 days was analyzed using Kaplan-Meier methods. Multivariable Cox regression that adjusted for patient-specific and cancer-specific factors was used to compare survival beyond 760 days between the fixed-duration group and the indefinite-duration group.
Results
Of 1091 patients in the analytic cohort who were still on ICI treatment at 2 years after exclusion criteria for death and progression were applied, 113 patients (median [IQR] age, 69 [62-75] years; 62 [54.9%] female; 86 [76.1%] White) were in the fixed-duration group, and 593 patients (median [IQR] age, 69 [62-76] years; 282 [47.6%] female; 414 [69.8%] White) were in the indefinite-duration group. Patients in the fixed-duration group were more likely to have a history of smoking (99% vs 93%; P = .01) and be treated at an academic center (22% vs 11%; P = .001). Two-year overall survival from 760 days was 79% (95% CI, 66%-87%) in the fixed-duration group and 81% (95% CI, 77%-85%) in the indefinite-duration group. There was no statistically significant difference in overall survival between patients in the fixed-duration and indefinite-duration groups, either on univariate (hazard ratio [HR] 1.26; 95% CI, 0.77-2.08; P = .36) or multivariable (HR 1.33; 95% CI, 0.78-2.25; P = .29) Cox regression. Approximately 1 in 5 patients discontinued immunotherapy at 2 years in the absence of progression.
Conclusions and Relevance
In a retrospective clinical cohort of patients with advanced NSCLC who were treated with immunotherapy and were progression-free at 2 years, approximately only 1 in 5 discontinued treatment. The lack of statistically significant overall survival advantage for the indefinite-duration cohort on adjusted analysis provides reassurance to patients and clinicians who wish to discontinue immunotherapy at 2 years.
This cohort study of advanced non–small cell lung cancer patients assesses practice patterns surrounding immunotherapy discontinuation at 2 years and evaluates the association of duration of therapy with overall survival.
Introduction
Treatment with immune checkpoint inhibitors (ICIs) has transformed the management of patients with advanced non–small cell lung cancer (NSCLC). The optimal duration of ICI use, however, is not known. Patients in key pivotal studies informing the use of first-line immunotherapy use were treated with ICI therapy for up to 2 years (maximum of 35 cycles every 3 weeks); in clinical practice, however, many patients continue therapy beyond 2 years.1
Long-term follow-up of large randomized clinical trials has shown that durable responses can be maintained after fixed-duration ICI therapy. In 5-year follow-up reports of the Keynote-010 and Keynote-024 trials, respectively, 83% and 82% of patients were still alive 3 years after completing 2 years of pembrolizumab, including 48% and 46% without disease progression or subsequent therapy.2,3 Furthermore, for patients whose disease progressed and were rechallenged with pembrolizumab, the majority of both studies (81% and 83%, respectively) derived clinical benefit from rechallenge with either stable disease or objective response.2,3
Despite these encouraging results, there has been substantial hesitation around ICI discontinuation, some of which stems from the results of the Checkmate-153 study in the second-line setting.4 In this study, patients who exhibited no evidence of disease progression and who were randomized to 1-year fixed-duration therapy with nivolumab had significantly shorter progression-free survival and overall survival compared with those treated with indefinite-duration therapy.5 Furthermore, of patients in the 1-year group whose disease progressed and then were rechallenged with ICIs, only 36% (14 of 39) were alive after 13.5 months of follow-up. Although these results were obtained in a second-line treatment setting and may have been related to a 1-year ICI duration rather than a 2-year ICI duration, they have led many clinicians to prefer indefinite-duration therapy over fixed-duration therapy.1,4
Compared with fixed-duration therapy, indefinite-duration ICI therapy has several inherent disadvantages including the risk of additional medical and financial toxic effects.6,7,8 Compared with those treated with fixed-duration ICI, patients treated with indefinite-duration ICI in the Checkmate-153 study experienced higher rates of treatment-related adverse events (TRAEs) (48% vs 26.4%) including grade 3 or higher TRAEs (9.4% vs 3.2%) and TRAEs leading to treatment discontinuation (9.4 vs 1.6%).5 Financial toxicity is also a major consideration with indefinite therapy; many countries such as the United Kingdom have limited ICI therapy to a maximum of 2 years regardless of disease status.4
We leveraged a nationwide US oncology database to investigate practice patterns and survival outcomes associated with 2-year ICI treatment discontinuation. We compared survival between long-term ICI responders whose therapy was discontinued at 2 years in the absence of death or progression (fixed-duration group) and those who continued ICI beyond 2 years (indefinite-duration group). We hypothesized that among patients in a clinical cohort who were progression-free at 2 years, there would be no difference in overall survival between fixed-duration and indefinite-duration ICI treatment.
Methods
Study Population
This retrospective cohort study used the longitudinal Flatiron Health database derived from the electronic health record (EHR), which includes deidentified patient-level structured and unstructured data curated via technology-enabled abstraction, originating from approximately 280 cancer clinics (approximately 800 sites of care) throughout the US.9 The present study cohort included adult patients (aged at least 18 years) with a new diagnosis of advanced or metastatic NSCLC between 2016 and 2021 who received frontline ICI therapy (alone or in combination with chemotherapy). Patients with driver alterations in EGFR, ALK, or ROS1 were excluded. The University of Pennsylvania institutional review board approval of the study protocol was obtained, and informed consent was waived because the data were deidentified. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. The data cutoff was August 31, 2022; data analysis was conducted from October 2022 to January 2023.
Definition of Variables
We described patient demographics (age, sex, race, smoking status), ECOG (Eastern Cooperative Oncology Group) performance status (PS; range, 0-5; describes a patient’s level of functioning in terms of their ability to care for themself, daily activity, and physical ability), tumor histologic type and PD-L1 expression, practice setting, insurance type, and treatment type (immunotherapy or chemoimmunotherapy). Race, as recorded in the EHR as part of clinical care, was included given previously demonstrated associations with outcomes in NSCLC. Clinical progression events included both RECIST (response evaluation criteria in solid tumors—a standard set of tests and scans to measure how cancer responds to different treatments, including immunotherapy) and non-RECIST progression events from medical record abstraction.10 The primary end point was clinical overall survival, defined as the time from initiation of frontline therapy to death, censoring at the last follow-up, or data cutoff (August 31, 2022). Death was defined from a composite mortality variable developed by Flatiron Health using structured and unstructured EHR-derived data linked to a commercial death data source and US Social Security Death Index.11 Each patient’s time on treatment was defined as the time from the first to the last documented administration of frontline treatment (including maintenance therapy). Structured activity or follow-up was defined as a record of patient vitals, medication administration, and reported laboratory tests and results.
Cohort Definition
Among 14 406 patients with advanced NSCLC who received frontline ICI-based therapy, we excluded 13 315 patients who discontinued ICI therapy before 700 days or approximately 2 years. To avoid immortal time bias,12 we also excluded patients who died or were lost to follow-up before 760 days, the starting point of the survival analysis (n = 52). Of the remaining 1039 patients who were still on treatment at 700 days, alive, and in follow-up at 760 days, we categorized patients into 2 groups: fixed duration and indefinite duration. Patients in the fixed-duration group had a time on frontline treatment of 700-759 days; ie, stopped treatment at 2 years (±30 days) after initiation of ICI-based therapy. Patients in the indefinite-duration group had at least 760 days on frontline treatment; ie, continued treatment beyond 2 years. To restrict the analysis to patients with ongoing response to immunotherapy, patients in either cohort were excluded if they had documented progression event(s) (including both RECIST and non-RECIST progression events)13 anytime from the start of ICI as frontline treatment to 760 days. Finally, patients in the fixed-duration group were excluded if they died in the same month as, or the month after, their end of treatment, as these patients’ treatment cessation was likely related to death rather than a decision to pursue fixed-duration therapy.
Statistical Analysis
Baseline Patient Characteristics
Patient, cancer, and treatment characteristics were summarized and compared between fixed-duration and indefinite-duration groups by unpaired t-test (for continuous variables) or Pearson χ2 test (for categorical variables).
Analysis of Clinical Overall Survival
Kaplan-Meier methods were used to estimate overall survival in the fixed-duration and indefinite-duration groups starting from a landmark time of 760 days after initiation of frontline treatment therapy. Survival between patients in the fixed-duration and indefinite-duration groups was compared using univariate and multivariable Cox regression models.
Multivariable Cox regression models were used to estimate adjusted hazard ratios (HRs) for overall survival in the fixed relative to the indefinite treatment duration group. Models were adjusted for age (continuous), sex, race (Asian, Black/African American, White, and other), ECOG PS (≤1 or >1), smoking history, squamous or nonsquamous histologic type, PD-L1 status (<1%, 1%-49%, or ≥50%), ICI monotherapy vs chemoimmunotherapy, insurance status (commercial, Medicare/Medicaid, or other/unknown), and academic vs community practice site. Proportional hazards assumption was assessed using Schoenfeld residuals. Multiple imputation with chained equations was used to account for variables with missing data (ECOG PS, race, PD-L1). Multiple imputation with chained equations has been shown to produce unbiased estimates when data are missing at random14 and provides advantages over alternate imputation techniques including flexibility in handling different data types.15,16 All covariates, including variables with and without missing data, were used for the multiple imputation process. Analyses were conducted within each of the 50 multiply imputed data sets, and results were combined across imputations using Rubin rules.17
Secondary Analyses
Probability of Treatment Discontinuation in the Absence of Progression
Within the entire cohort of patients who started frontline immunotherapy-based treatment, we estimated the probability of discontinuation of therapy not related to disease progression or death using a cumulative incidence function and life table approach. For this analysis, time 0 was the start of frontline therapy; and the event of interest was defined as discontinuation of frontline treatment (1) without documented progression within 60 days of discontinuation, (2) without documented death within 6 months of discontinuation, and (3) with at least 90 days of structured follow-up after discontinuation. Patients who had progression or death within these intervals around treatment discontinuation were considered to have stopped therapy due to progression or death, and classified as having a competing event that precluded the occurrence of the primary event. Patients without progression or death, but with less than 90 days of structured follow-up after discontinuation, were censored.
Outcomes With ICI Rechallenge
Within the fixed-duration cohort, we identified patients who were rechallenged with ICI therapy after at least 30 days without ICI treatment. The length of time from cessation of frontline ICI therapy to progression, ICI rechallenge, and subsequent progression or death were visually displayed on a swimmer plot. Time from ICI rechallenge to subsequent progression or death (PFS2) was estimated using Kaplan-Meier methods.
Statistical significance was defined as a 2-sided P < .05. Statistical analyses were performed using Stata statistical software, version 17 (StataCorp).
Results
Study Population
Of 14 406 patients who started frontline immunotherapy-based treatment for NSCLC, 13 315 discontinued therapy before 2 years; the majority of these died (n = 8522; 64%) or had disease progression and/or initiation of second-line therapy (n = 2663; 20%) within the first 2 years of treatment. Among a total of 1091 patients who were still on treatment at 2 years after exclusion criteria were applied, 113 patients (median [IQR] age, 69 [62-75] years; 62 [54.9%] female; 86 [76.1%] White) were included in the fixed-duration group, and 593 patients (median [IQR] age, 69 [62-76] years; 282 [47.6%] female; 414 [69.8%] White) were in the indefinite-duration group (Figure 1).
Figure 1. CONSORT Diagram.
Abbreviations: NSCLC, non–small cell lung cancer; ICI, immune checkpoint inhibitor.
aExclusion criterion applied to enable landmark Kaplan-Meier analysis starting at 760 days.
bPatients in the fixed-duration group were excluded if death occurred in the same month as, or the month after, their end of treatment, as these patients’ treatment cessation was more likely related to death than the decision to pursue fixed-duration therapy.
Patients in the fixed-duration group vs the indefinite-duration group were more likely to have a history of smoking, respectively (112 of 113 [99%] vs 550 of 593 [93%]; P = .01) and be treated in an academic center (25 of 113 [22%] vs 65 of 593 [11%]; P = .001) (Table 1). Approximately half of patients in both groups were treated initially with immunotherapy alone vs chemoimmunotherapy; ICI monotherapy was slightly numerically more common in the fixed-duration group vs the indefinite-duration group, respectively (59 of 113 [52%] vs 279 of 593 [46%]; P = .39). The PD-L1 status distribution was similar in both groups (PD-L1 ≥ 50%: 51 of 113 [45%] vs 293 of 593 [49%] ; P = .19), as was ECOG PS (ECOG PS 0 or 1: 90 of 113 [77%] vs 458 of 593 [80%] ; P = .68).
Table 1. Baseline Cohort Characteristics.
| Characteristic | Patients, No. (%)a | P valueb | |
|---|---|---|---|
| Fixed duration (n = 113) |
Indefinite duration (n = 593) |
||
| Median (IQR) age, y | 69 (62-75) | 69 (62-76) | .66 |
| Sex | |||
| Female | 62 (54.9) | 282 (47.6) | .15 |
| Male | 51 (45.1) | 311 (52.4) | |
| Racec | |||
| Asian | 0 | 5 (0.8) | .38 |
| Black/African American | 16 (14.2) | 68 (11.5) | |
| White | 86 (76.1) | 414 (69.8) | |
| Other | 7 (6.2) | 58 (9.8) | |
| Missing | 4 (3.5) | 48 (8.1) | |
| ECOG performance status | |||
| 0 | 34 (30.1) | 204 (34.4) | .68 |
| 1 | 56 (49.6) | 254 (42.8) | |
| ≥2 | 14 (12.4) | 78 (13.3) | |
| Missing | 9 (8.0) | 57 (9.6) | |
| PD-L1, % | |||
| 0 | 10 (8.8) | 90 (15.2) | .19 |
| 1-49 | 25 (22.1) | 110 (18.5) | |
| ≥50% | 51 (45.1) | 293 (49.4) | |
| Missing | 27 (23.9) | 100 (16.9) | |
| Smoking status | |||
| Former/current | 112 (99.1) | 550 (92.7) | .01 |
| Never | 1 (0.9) | 43 (7.3) | |
| Histologic type | |||
| Nonsquamous | 79 (69.9) | 463 (78.1) | .15 |
| Squamous | 29 (25.7) | 107 (18.0) | |
| NSCLC histology NOS | 5 (4.4) | 23 (3.9) | |
| Practice setting | |||
| Community | 88 (77.9) | 528 (89.0) | .001 |
| Academic | 25 (22.1) | 65 (11.0) | |
| Insurance | |||
| Commercial | 56 (49.6) | 305 (51.4) | .53 |
| Medicare/Medicaid | 43 (38.1) | 197 (33.2) | |
| Other/unknown | 14 (12.4) | 91 (15.3) | |
| Treatment | |||
| Immunotherapy | 59 (52.2) | 279 (46.0) | .39 |
| Chemoimmunotherapy | 54 (47.8) | 314 (53.0) | |
Abbreviations: NSCLC, non–small cell lung cancer; NOS, not otherwise specified; ECOG, Eastern Cooperative Oncology Group; EHR, electronic health record; PD-L1, programmed death-ligand 1.
Number (%) for categorical variables; median (IQR) for continuous variables.
Chi square test for categorical variables; t-test for continuous variables.
Race was collected when recorded in the EHR as part of routine clinical care. “Other” includes a myriad of unspecified terms describing race in the EHR and cannot be further categorized.
Survival
With a median (range) follow-up of 14.0 (0.1-50.9) months starting 760 days after treatment initiation, there was no statistically significant difference in overall survival between patients treated with fixed-duration and indefinite-duration ICI therapy on either unadjusted or adjusted analysis (Figure 2). Survival probabilities at time points after 760 days from the start of therapy in the fixed vs indefinite-duration groups are shown in Table 2. Two-year overall survival from 760 days (ie, 4 years total from frontline treatment initiation) was 79% (95% CI, 66%-87%) in the fixed-duration group and 81% (95% CI, 77%-85%) in the indefinite-duration group. The HR for death associated with fixed-duration therapy was 1.26 (95% CI, 0.77, 2.08) on unadjusted analysis, and 1.33 (95% CI, 0.78, 2.25) on adjusted multivariable analysis compared with the indefinite-duration group (Table 2). Proportional hazards assumption was not violated in these analyses. In multivariable analyses, both Black/African American race and having unknown/other insurance status were significantly associated with worse overall survival (eTable 1 in the Supplement).
Figure 2. Overall Survival.
Kaplan-Meier curve of overall survival from 2 years (760 days) from immune checkpoint inhibitor (ICI) treatment initiation in the fixed-duration cohort (stopped treatment at 2 years; 700-759 days of treatment) and indefinite-duration cohort (at least 760 days of treatment).
Table 2. Overall Survival in Fixed-Duration Treatment Cohort and Indefinite-Duration Treatment Cohort.
| Survival Characteristic | Fixed duration (n = 113) | Indefinite duration (n = 593) |
|---|---|---|
| Overall survival probability | ||
| 3 y (12 mos from 760 d) | 0.89 (0.81-0.94) | 0.91 (0.88-0.94) |
| 4 y (24 mos from 760 d) | 0.79 (0.66-0.87) | 0.81 (0.77-0.85) |
| Hazard ratio for death | ||
| Unadjusted | 1.26 (0.77-2.08) | 1 [Reference] |
| P value | .36 | |
| Adjusteda | 1.33 (0.78-2.25) | 1 [Reference] |
| P value | .29 | |
Multivariable Cox regression adjusted for age, sex, race, ECOG (Eastern Cooperative Oncology Group) performance status (PS), PD-L1 (programmed death-ligand 1), history of smoking, histologic type, immunotherapy vs chemoimmunotherapy, insurance, and academic vs community site. Multiple imputation by chained equations was used for variables with missing data (race, ECOG PS, PD-L1).
Secondary Analyses
Treatment Discontinuation
The cumulative incidence of treatment discontinuation in the absence of progression or death after initiation of ICI therapy is presented in Figure 3. Among patients still on ICI treatment at 2 years, only approximately 1 in 5 discontinued treatment in the absence of associated progression or death by the next 2-month time point. The discontinuation rates of ICI were not substantially higher at 2 years than at other time points on treatment (Figure 3; eTable 2 in the Supplement).
Figure 3. Treatment Discontinuation Over Time and ICI Rechallenge.
A, The cumulative incidence of treatment discontinuation in the absence of progression or death over time (in months) from treatment initiation. Patients with progression within 60 days of discontinuation or death within 6 months of discontinuation were classified as having a competing event for this analysis. B, Swimmer plot with fixed-duration treatment group patients rechallenged with ICI-based therapy. Abbreviations: ICI, immune checkpoint inhibitor; PD, progressive disease.
ICI Rechallenge
A total of 11 patients in the fixed-duration cohort subsequently had progression and were rechallenged with ICI after at least 30 days without treatment: 8 with ICI monotherapy and 3 with ICI in combination with chemotherapy. All but 1 patient was rechallenged with the same ICI used in the frontline setting. The median (range) time from cessation of frontline treatment to initiation of second-line therapy was 7.4 (1.8-26.9) months. After ICI rechallenge, median PFS2 was 8.1 months.
Discussion
This retrospective cohort study evaluated the association between fixed 2-year vs indefinite-duration ICI therapy and overall survival among patients with advanced NSCLC on long-term ICI. In this clinical cohort study of patients treated with frontline ICI-based therapy, we found no evidence of a difference in overall survival among patients who received fixed-duration ICI-based therapy for 2 years compared with indefinite therapy. This finding remained robust in multivariable analyses adjusting for covariates such as smoking, PD-L1 status, and histologic type. Overall, approximately 1 in 5 patients discontinued ICI treatment at 2 years.
The present findings have several key implications for the clinical management of patients with metastatic NSCLC. First, in line with previous studies that have shown durable benefit after stopping ICI at 2 years,2,3 the present study supports treatment discontinuation at 2 years as a valid approach that does not appear to compromise survival outcomes. In this analysis, the 2 groups (fixed-duration vs indefinite-duration ICI therapy) were well balanced, with the exception that patients in the fixed-duration group were more likely to be treated at an academic center, have a history of smoking, and a trend toward a higher proportion of squamous cell carcinoma and ICI monotherapy. These differences may reflect providers’ comfort level in discontinuing ICI therapy for a subgroup of patients known to respond well to immunotherapy (squamous histologic type, smoking history). These findings augment the literature by demonstrating that even after adjusting for these covariates, there was no overall survival benefit for indefinite-duration therapy.
Second, we demonstrate that patients treated with fixed-duration therapy may benefit from rechallenge with ICI even if progression occurs after ICI cessation. In this small cohort of patients who underwent ICI rechallenge, median progression-free survival after rechallenge was 8.1 months, with more than one-third of patients still on treatment at the data cutoff point. Results from this small cohort in a clinical setting align with the promising results seen with ICI rechallenge in long-term follow-up of randomized clinical trials.2,3
Third, overall rates of fixed-duration ICI treatment in this observational analysis are low; among those without disease progression or death, only approximately 1 in 5 patients discontinued therapy at 2 years. While the 2-year fixed duration of ICI instituted in clinical trials is empirical, we expected to see a spike in discontinuation at 2 years aligned with the fixed 2-year ICI duration specified in the large pivotal trials. However, there was barely a discernible uptick in discontinuations at 2 years, and the difference in probability of discontinuation of therapy was not substantially different at 2 years compared with other points in time.
Limitations
Despite adjustment for baseline covariates, residual confounding remains a possibility, as unmeasured factors may influence the decision to continue or discontinue treatment at 2 years. Patients who discontinued treatment at 2 years due to disease progression occurring just after 760 days may have been incorrectly included and labeled as fixed duration because they are indistinguishable in the database from patients who stopped treatment at 2 years without disease progression and subsequently rapidly progressed. Notably, despite this possible bias against the fixed-duration group, there was still no statistically significant survival difference between the cohorts. Finally, this study solely investigated a clinically relevant 2-year cut point for immunotherapy duration; we did not evaluate other time points of treatment cessation to investigate the question of “how much immunotherapy is enough?” in a continuous manner.
Strengths
Given concerns about medical adverse effects and potentially burdensome financial costs associated with the indefinite continuation of ICI therapy, there is a pressing need to answer the question of whether continuing ICI beyond 2 years offers clinical benefit compared with stopping treatment. Randomized clinical trials of 2-year vs indefinite-duration ICI therapy have been proposed, but they will require prolonged follow-up and will not yield results to guide clinical practice in the foreseeable future. With a growing cohort of long-term responders reaching the 2-year treatment mark, clinical guidance is urgently needed to inform practice now. The present results support the practice of fixed-duration ICI therapy in patients who have achieved and are maintaining clinical benefit, an approach that does not appear to compromise long-term survival.
Conclusions
In this retrospective cohort study of patients with advanced/metastatic NSCLC treated with frontline ICI-based therapy, only approximately 1 in 5 patients who reached 2 years of ICI discontinued treatment in the absence of progression or death. There was no statistically significant difference in overall survival by fixed duration (2 years) vs indefinite duration (>2 years) of ICI therapy. These findings provide reassurance that for patients with advanced NSCLC whose disease is still responding to ICI therapy at 2 years, stopping therapy and monitoring rather than continuing immunotherapy indefinitely is a reasonable strategy with sustained clinical benefit.
eTable 1. Hazard ratio for death
eTable 2. Proportion of patients without progression or death who stopped treatment at 2-month intervals, starting at 1 year after treatment initiation
Data Sharing Statement
References
- 1.Steuer CE, Ramalingam SS. Advances in immunotherapy and implications for current practice in non–small-cell lung cancer. JCO Oncol Pract. 2021;17(11):662-668. doi: 10.1200/OP.21.00305 [DOI] [PubMed] [Google Scholar]
- 2.Reck M, Rodríguez-Abreu D, Robinson AG, et al. Five-year outcomes with pembrolizumab versus chemotherapy for metastatic non–small-cell lung cancer with PD-L1 tumor proportion score ≥50. J Clin Oncol. 2021;39(21):2339-2349. doi: 10.1200/JCO.21.00174 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Herbst RS, Garon EB, Kim D-W, et al. Long-term outcomes and retreatment among patients with previously treated, programmed death-ligand 1–positive, advanced non–small-cell lung cancer in the KEYNOTE-010 study. J Clin Oncol. 2020;38(14):1580-1590. doi: 10.1200/JCO.19.02446 [DOI] [PubMed] [Google Scholar]
- 4.Friedlaender A, Kim C, Addeo A. Rethinking the optimal duration of immune checkpoint inhibitors in non–small cell lung cancer throughout the COVID-19 pandemic. Front Oncol. 2020;10:862. doi: 10.3389/fonc.2020.00862 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Waterhouse DM, Garon EB, Chandler J, et al. Continuous versus 1-year fixed-duration nivolumab in previously treated advanced non–small-cell lung cancer: CheckMate 153. J Clin Oncol. 2020;38(33):3863-3873. doi: 10.1200/JCO.20.00131 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Almutairi AR, McBride A, Slack M, Erstad BL, Abraham I. Potential immune-related adverse events associated with monotherapy and combination therapy of ipilimumab, nivolumab, and pembrolizumab for advanced melanoma: a systematic review and meta-analysis. Front Oncol. 2020;10;91. doi: 10.3389/fonc.2020.00091 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Thompson JA, Schneider BJ, Brahmer J, et al. NCCN guidelines insights: management of immunotherapy-related toxicities, version 1.2020. J Natl Compr Canc Netw. 2020;18(3):230-241. doi: 10.6004/jnccn.2020.0012 [DOI] [PubMed] [Google Scholar]
- 8.Naidoo J, Page DB, Li BT, et al. Toxicities of the anti-PD-1 and anti-PD-L1 immune checkpoint antibodies. Ann Oncol. 2015;26(12):2375-2391. doi: 10.1093/annonc/mdv383 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ma X, Long L, Moon S, Adamson BJS, Baxi SS. Comparison of population characteristics in real-world clinical oncology databases in the US: Flatiron Health, SEER, and NPCR. medRxiv. Preprint posted online May 30, 2020. doi: 10.1101/2020.03.16.20037143 [DOI]
- 10.Griffith SD, Tucker M, Bowser B, et al. Generating real-world tumor burden endpoints from electronic health record data: comparison of RECIST, radiology-anchored, and clinician-anchored approaches for abstracting real-world progression in non–small cell lung cancer. Adv Ther. 2019;36(8):2122-2136. doi: 10.1007/s12325-019-00970-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zhang Q, Gossai A, Monroe S, Nussbaum NC, Parrinello CM. Validation analysis of a composite real-world mortality endpoint for patients with cancer in the United States. Health Serv Res. 2021;56(6):1281-1287. doi: 10.1111/1475-6773.13669 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Yadav K, Lewis RJ. Immortal time bias in observational studies. JAMA. 2021;325(7):686-687. doi: 10.1001/jama.2020.9151 [DOI] [PubMed] [Google Scholar]
- 13.Griffith SD, Miksad RA, Calkins G, et al. Characterizing the feasibility and performance of real-world tumor progression end points and their association with overall survival in a large advanced non–small-cell lung cancer data set. JCO Clin Cancer Inform. 2019;3(3):1-13. doi: 10.1200/CCI.19.00013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Azur MJ, Stuart EA, Frangakis C, Leaf PJ. Multiple imputation by chained equations: what is it and how does it work? Int J Methods Psychiatr Res. 2011;20(1):40-49. doi: 10.1002/mpr.329 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sterne JA, White IR, Carlin JB, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ. 2009;338:b2393. doi: 10.1136/bmj.b2393 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wells BJ, Chagin KM, Nowacki AS, Kattan MW. Strategies for handling missing data in electronic health record derived data. EGEMS (Wash DC). 2013;1(3):1035. doi: 10.13063/2327-9214.1035 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Rubin DB. Multiple imputation after 18+ years. J Am Stat Assoc. 1996;91(434):473-489. doi: 10.1080/01621459.1996.10476908 [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Hazard ratio for death
eTable 2. Proportion of patients without progression or death who stopped treatment at 2-month intervals, starting at 1 year after treatment initiation
Data Sharing Statement
