|Ahead of print publication
Checkpoint inhibitors in advanced nonsmall-cell lung cancer; a Bayesian network meta-analysis
Hakan Bozcuk1, Mustafa Yıldırım2, Özlem Sever3, Hasan Mutlu1, Mehmet Artaç4
1 Department of Medical Oncology, Medical Park Hospital, Antalya, Turkey
2 Department of Medical Oncology, Medical Park Hospital, Gaziantep, Turkey
3 Department of Internal Medicine, Sanko University School of Medicine, Gaziantep, Turkey
4 Department of Medical Oncology, Necmettin Erbakan University, Konya, Turkey
|Date of Submission||26-Jun-2019|
|Date of Decision||21-Nov-2019|
|Date of Acceptance||19-Dec-2019|
|Date of Web Publication||07-Jul-2020|
Department of Medical Oncology, Medical Park Hospital, Lara, Antalya
Source of Support: None, Conflict of Interest: None
Background: Checkpoint inhibitors (CPIs) have improved survival compared to chemotherapy alone in advanced nonsmall-cell lung cancer (NSCLC). This article aims to compare indirect evidence and rank the effect of different CPIs in this setting.
Materials and Methods: In this network meta-analysis, we searched for trials comparing CPIs in advanced NSCLC. Figures for survival endpoints were extracted. In addition, a network meta-regression analysis was carried out.
Results: A total of 9220 patients from 16 trials were included in the analysis. In the first-line setting, for the overall survival endpoint, the chemotherapy + Pembrolizumab combination had the highest effectivity rank probability as compared to chemotherapy (hazard ratio = 0.788, 95% credential interval = 0.728–0.855). For the second-line setting, and also for the efficacy in terms of progression-free survival, various CPIs and their combinations were ranked.
Conclusion: Some degree of differences in terms of efficacy exists between different types, dosages, settings, and combinations of CPI. We quantify these differences to guide clinical practice.
Keywords: Atezolizumab, avelumab, checkpoint inhibitor, nivolumab, nonsmall-cell lung cancer, pembrolizumab
|How to cite this URL:|
Bozcuk H, Yıldırım M, Sever &, Mutlu H, Artaç M. Checkpoint inhibitors in advanced nonsmall-cell lung cancer; a Bayesian network meta-analysis. J Can Res Ther [Epub ahead of print] [cited 2020 Aug 12]. Available from: http://www.cancerjournal.net/preprintarticle.asp?id=289133
| > Introduction|| |
Chemotherapy has been the mainstay of management in metastatic nonsmall-cell lung cancer (NSCLC) for decades. Only a couple of years ago, we understood that monoclonal antibodies inhibit checkpoints of the immune system and could improve survival as compared to chemotherapy in NSCLC. First, in 2012, ipilimumab, an anti-cytotoxic T-lymphocyte-associated molecule-4 agent, then nivolumab, an anti-programmed cell death receptor-1 class drug, showed more efficacy than chemotherapy., Of note, the risk of death was 41% lower with nivolumab than with docetaxel in chemotherapy-refractory patients with squamous-cell NSCLC in the CheckMate 017 trial. Agents like atezolizumab target programmed cell death ligand-1 (PD-L1) and have also shown activity in NSCLC. The number of CPIs available is increasing with documented treatment benefits in NSCLC and other cancers.
A number of CPIs work in this setting, but we do not have evidence of their relative efficacies from direct testing of these agents with each other. Moreover, testing available checkpoint inhibitors (CPIs) with each other in NSCLC would be unrealistic, because this would require recruiting a large number of patients in such studies, enormous financial support, and a long-time period for completion. Therefore, this study aims, in the frame of a network meta-analysis, to compare CPIs and their combinations with other CPIs or chemotherapy to gauge their efficacy in terms of prolongation of survival in metastatic NSCLC patients.
| > Materials and Methods|| |
Search strategy and selection criteria
This is a network meta-analysis that used summary data. Four reviewers used multiple search criteria to ensure accurate selection of clinical trials for the network analysis. All articles in the English language were systemically searched from database inception to April 1, 2019, in PubMed, Science of Web, and “Clinicaltrials.gov.” In addition, congress proceeding books from the past 3 years as well as online presentations of the American Society of Clinical Oncology and European Society for Medical Oncology congresses were searched. Finally, we also looked for each CPI and relevant results of clinical trials on the web. As this was the first search strategy, all randomized clinical trials (RCTs) that evaluated the use of any monoclonal antibody in the treatment of lung cancer were taken into account. Second, relevant RCTs of individual drugs – Nivolumab, Pembrolizumab, Ipilimumab, Atezolizumab, Avelumab, Durvalumab, and Tremelimumab – were also searched. In addition, references of the selected studies as well as clinical studies in the relevant reviews were evaluated. Finally, we conducted a Google search for any possible early press release from drug companies about the efficacy of the CPIs studied. Studies satisfying the selection criteria above were considered only if they reported oral allergy syndrome (OAS) and/or Pollen Fruit Syndrome (PFS) results. Conflicts over the inclusion of trials were resolved by discussion among the authors. We followed the PRISMA NMA guidelines for this analysis.
We used the GeMTC software which is freely available on “https://gemtc.drugis.” This software is an online tool for Bayesian network meta-analysis., With it, a researcher can upload their own dataset and perform the various analyses supported by the GeMTC R package. We performed consistency models for PFS and OAS analyses both in the first and subsequent line usage of CPIs. OAS was the main and PFS the secondary endpoint in this analysis. The principal summary measure was the hazard ratio (HR). The logarithmic transformation and the corresponding standard errors were calculated and loaded on the database. Multi-arm trials were also allowed in the analysis, and the standard error of the (absolute) effect in the base was specified. Standard errors were calculated using previously published methods.,
We collected data on the main and secondary endpoints as well as the type of treatment received, PDL-1 status, histology of cancer, size of trial, and year of publication. PDL-1 status was calculated for the CPI-treated arm in each trial by multiplying the fraction of the highest PDL-1-expressing subgroup by the strength of PDL-1 expression. In addition, the fraction with the nonsquamous histology for each CPI-treated arm in each trial was recorded. We also conducted a network meta-regression analysis to explain any potential sources of heterogeneity in the outcome. For the pairwise comparison of CPIs and their various combinations, we used fixed analysis assuming a Poisson/log type of likelihood/link. For run-length parameters for the analyses, 5000 was set as the number of burn-in iterations and 20,000 as the number of inference iterations.
Funnel plots, to check for heterogeneity, and rank probability plots, to rank CPIs and their combinations, were constructed. To assess model fit, model-fit statistics (residual deviance, leverage, deviance information criterion, and number of data points) were separately computed for each model.
| > Results|| |
A total of 16 clinical trials recruiting a total of 9220 patients (sum of chemotherapy and CPI-treated cases) were selected.,,,,,,,,,,,,,,, [Figure 1] shows the PRISMA flowchart for the selection of the studies. Notably, we also used the web release information regarding OAS results of the Javelin Lung 200 trial, which compared avelumab with chemotherapy. One of these 16 clinical trials had unpublished data.
Among the 16 clinical trials studied, 10 evaluated the effects of CPIs compared to chemotherapy first line, and six, after the first line, in the chemotherapy refractory setting. Five different CPIs (nivolumab, pembrolizumab, ipilimumab, atezolizumab, and avelumab) and their various combinations with another CPI or chemotherapy, or different dosages of the same CPI, had been used in these trials. [Table 1] shows the details of these 16 clinical trials.
|Table 1: Details and outcomes of randomised controlled trials on the effect of Monoclonal Antibody Checkpoint Inhibitors in advanced Non-Small Cell Lung Cancer|
Click here to view
First line efficacy
For the first-line efficacy of CPIs (chemotherapy sensitive setting) on OAS, six different treatment types incorporating CPIs (chemotherapy + Atezolizumab, chemotherapy + Ipilimumab (concurrent), chemotherapy + Ipilimumab (phased), chemotherapy + Pembrolizumab, Nivolumab and Pembrolizumab) were evaluated. In addition, for the PFS endpoint, the Nivolumab + Ipilimumab treatment type was also taken into account in addition to the previous six treatment types, totaling seven treatment types for PFS. [Figure 2] shows the corresponding network graphs.
|Figure 2: Network graphs of check point inhibitors and their effect on survival at first line in Nonsmall cell lung cancer. (a) Check point inhibitors and their effect on overall survival at first line, (b) Check point inhibitors and their effect on progression-free survival at first line (Che=Chemotherapy, Pem: Pembrolizumab, Niv=Nivolumab, Ip=Ipilimumab phased, Ic=Ipilimumab concurrent, I=Ipilimumab, Ate=Atezolizumab)|
Click here to view
In terms of OAS, chemotherapy + Atezolizumab, chemotherapy + Pembrolizumab and Pembrolizumab were more effective than chemotherapy alone. In addition, Pembrolizumab was marginally more effective than Nivolumab (HR = 0.890 95% credential interval [CrI] = 0.796–0.997), but less effective than chemotherapy + Pembrolizumab (HR = 1.141 [95%] CrI = 1.034–1.258). Rank probability plot identified chemotherapy + Pembrolizumab as the most effective treatment; with regard to chemotherapy, chemotherapy + Pembrolizumab had an HR of 0.788 (95% CrI = 0.728–0.855). [Table 2] and [Figure 3] show the details of OAS-related effects of CPIs.
|Table 2: 1st line comparative activity in terms of overall survival of Monoclonal Antibody Check Point Inhibitors in Nonsmall cell lung cancer|
Click here to view
|Figure 3: Rank probabilities plot of effect of monoclonal antibody check point inhibitors on survival given first line in Nonsmall cell lung cancer. (a) Check point inhibitors and first line overall survival, (b) Check point inhibitors and first line progression-free survival (Che=Chemotherapy, Ate=Atezolizumab, Ic=Ipilimumab concurrent, Ip=Ipilimumab phased, Pem=Pembrolizumab, Niv=Nivolumab, Niv + I=Nivolumab + Ipilimumab. Higher rank numbers imply higher efficacy)|
Click here to view
Chemotherapy + Atezolizumab, chemotherapy + Ipilimumab (phased), chemotherapy + Pembrolizumab and Nivolumab + Ipilimumab were associated with better PFS as compared to chemotherapy alone. Nivolumab and Pembrolizumab were less effective than chemotherapy + Pembrolizumab with (HR = 1.387 [95%] CrI = 1.234–1.561) and (HR = 1.277 [95%] CrI = 1.171–1.395), respectively. Chemotherapy + Pembrolizumab again appeared to be the most influential treatment type in the rank probability plot as compared to chemotherapy; chemotherapy + Pembrolizumab had an HR of 0.766 (95% CrI = 0.716–0.819). The details of the related efficacy figures and ranking details are shown in [Table 3] and [Figure 3], respectively.
|Table 3: First line comparative activity in terms of progression-free survival of monoclonal antibody check point inhibitors in nonsmall-cell lung cancer|
Click here to view
To assess the second-line efficacy of CPI (chemotherapy refractory setting) both for the OAS and PFS analysis, five treatment types (nivolumab, pembrolizumab, pembrolizumab (high dose), atezolizumab, and avelumab) had been tested. [Figure 4] shows the associated network graphs.
|Figure 4: Network graphs of Check Point Inhibitors and their effect on survival at 2nd line in Nonsmall cell lung cancer. (a) Check point inhibitors and their effect on overall survival at 2nd line, (b) Check point inhibitors and their effect on progression-free survival at 2nd line (Che=Chemotherapy, Niv=Nivolumab, Pem=Pembrolizumab, Pem (h)=Pembrolizumab high dose; 10 mg/kg, Ate=Atezolizumab, Ave=Avelumab)|
Click here to view
Nivolumab, pembrolizumab, atezolizumab, and pembrolizumab (high dose) were comparably potent and associated with better OAS in contrast to chemotherapy alone. Avelumab was less effective than pembrolizumab (high dose) (HR = 1.150 (95% CrI = 1.175–1.230). [Table 4] gives the OAS endpoint-related details. For the PFS point, nivolumab and pembrolizumab (high dose) were more efficient with reference to chemotherapy. [Table 5] shows the specifics of the effect of CPIs on the PFS outcome.
|Table 4: Second-line comparative activity in terms of overall survival of Monoclonal Antibody Check Point Inhibitors in nonsmall-cell lung cancer|
Click here to view
|Table 5: 2nd line comparative activity in terms of progression-free survival of monoclonal antibody check point Inhibitors in nonsmall cell lung cancer|
Click here to view
Rank probability plots identified pembrolizumab (high dose) as the treatment type with a higher chance of being the most effective strategy in this setting. Nivolumab was the second higher-ranking treatment type. [Figure 5] specifies the details of efficacy-ranking probabilities of CPIs in terms of OAS and PFS.
|Figure 5: Rank Probabilities Plot of effect of Monoclonal Antibody Check Point Inhibitors on survival given 2nd line in Nonsmall cell lung cancer. (a) Check point inhibitors and 2nd line overall survival, (b) Check point inhibitors and 2nd line progression-free survival (Che=Chemotherapy, Pem=Pembrolizumab, Pem (h)=Pembrolizumab high dose, Niv=Nivolumab, Ave=Avelumab, Ate=Atezolizumab. Higher rank numbers imply higher efficacy)|
Click here to view
Meta-regression analysis and heterogeneity
Meta-regression analysis indicated that line of therapy ( first line versus beyond first line) was the only significant predictor of OAS variation (β = −0.1, 95% confidence interval [CI] = −0.185–−0.018). PDL-1 status was not significant for PFS (β = −0.136, 95% CI = −0.345–0.076). In addition, again, line of chemotherapy (β = −0.080, 95% CI = −0.153–−0.008), size of trial (β = 0.183, 95% CI = 0.098–0.272), and year of publication (β = 0.151, 95% CI = 0.061–0.183) were significantly associated with PFS variation. [Table 6] shows the details and interpretation.
|Table 6: Metaregression analysis of the effect of Check Point Inhibitors on survival in advanced Non.Small Cell Lung Cancer|
Click here to view
There was some evidence of inconsistency in the first-line and second-line PFS analysis, as the residual deviance and number of data points criteria appeared to be unequal. Conversely, there was no evidence of inconsistency in the first- and second-lineOAS analysis. Funnel plots did not show evidence of heterogeneity.
| > Discussion|| |
Our analysis shows that combining chemotherapy and CPI frontline may be a better approach in advanced NSCLC than administering CPIs alone, chemotherapy alone or combining CPIs. Thus, pembrolizumab with chemotherapy is one of the most effective treatment strategies in metastatic NSCLC patients without driver mutations. We find that the use of pembrolizumab alone is associated with a 14% increase in the risk of death and 28% increase in the risk of progression and/or death as compared to Pembrolizumab + chemotherapy, raising the possibility of omission of chemotherapy may be hazardous frontline. Of note, the superiority of Pembrolizumab + chemotherapy has been demonstrated in the first-line setting, both for nonsquamous and squamous NSCLC, as is the case in Keynote 407 and Keynote 189 trials., Studies suggest that the benefit from Pembrolizumab + chemotherapy is irrespective of PD-L1 status, whereas Pembrolizumab alone appears to be effective if PD-L1 tumor proportion score (TPS) is at least ≥1%., In addition, the use of tumour mutational burden (TMB) as a biomarker in Checkmate 227, where nivolumab in combination with ipilimumab is tested against chemotherapy, is promising since patients with TMB ≥10 mut/Mb derive substantial benefit from nivolumab in combination with ipilimumab over chemotherapy, regardless of PD-L1 status or tumor histologic type. Thus, in addition to PD-L1 status, new biomarkers like the TMB status may further shed light on the biology-driven frontline selection of the best CPI-based treatment approach in metastatic NSCLC. However, the most recent TMB data are disappointing and question the utility of TMB in this setting. Thus, we think we need new and better predictive biomarkers, such as immune gene signatures and their prospective testing in this setting, since this may theoretically allow selection of the best CPI-based treatment frontline; whether it is chemotherapy + Pembrolizumab as our analysis shows, or a CPI alone as seen in Keynote studies 024 and 042, or a combination of CPIs as in Checkmate 227.,,,
In chemotherapy refractory patients, our results favor Pembrolizumab (high dose: 10 mg/kg) or nivolumab, pembrolizumab having a marginally higher probability of being the better treatment in terms of OAS endpoint. However, pairwise comparison of these two agents, and also of pembrolizumab fixed dose, and atezolizumab with pembrolizumab (high dose) do not show any significant differences. As the Food and Drug Administration approved dosage is fixed at 200 mg every 3 weeks and not 10 mg/kg, all three agents, nivolumab, pembrolizumab fixed-dose or atezolizumab, need to be recognized as equally good treatment strategies in chemotherapy refractory patients. There is a clinical need in metastatic NSCLC, at the second line, to see the results of chemotherapy + nivolumab or another CPI compared to nivolumab or another CPI alone since we do not have such data yet. As we demonstrated in our study, the frontline contribution to survival by chemotherapy or by the addition of a second CPI may also be valid at the second-line setting. We believe this hypothesis needs prospective testing.
Our network meta-regression analysis clearly shows the importance of a line of CPI usage in metastatic NSCLC, both in terms of OAS and PFS. In chemotherapy refractory patients, the benefit of a CPI compared to chemotherapy is marginally better than, or at least comparable to, that in chemotherapy-sensitive patients. This illustrates the importance of the use of CPIs in NSCLC regardless of the line of use, and that CPIs should not be denied to all suitable patients. However, CPIs as monoclonal antibodies are costly treatments, and due to the lack of coverage by health systems in many developing countries, strategies await to be developed to enable wider access to CPIs. We also find that earlier and smaller trials demonstrate better PFS advantage with CPIs. Some earlier trials are randomized phase 2 studies and larger effect sizes observed may be due to a play of chance resulting from a smaller patient size or due to unknown factors.
We did not show the effect of PDL-1 status on OAS and PFS results by CPIs, but its association with OAS is at the boundary of significance. In the literature, there are studies showing CPI benefit is both correlated and contrarily, unrelated to PD-L1 expression. The nonsignificance of PDL-1 status in our study may stem from our definition and calculation of PDL-1 status or from different interactions of PDL-1 status with the efficacy of different CPIs. In this manner, in our analysis, any significant association regarding PDL-1 and efficacy may have been lost. In the literature, different methods to denote PD-L1 expression have been used, like the PD-L1 Immunohistochemistry 73–10 pharmDx assay, PD-L1 IHC 22C3 pharmDx assay (Agilent Technologies), and VENTANA SP142 PD-L1 immunohistochemistry assay (Ventana Medical Systems, Inc., Tucson, AZ, USA). In addition, different PD-L1 assessment methods take different proportions of tumor cells or tumor-infiltrating immune cells into account. We believe this fact contributes to the difficulty of questioning the predictive value of PD-L1 in this setting.
Optimal patient selection for CPIs remains to be an important clinical problem. To enable better and easier patient selection for CPI-based therapies, liquid biopsy has recently emerged as a powerful tool for tumor genotyping, and also enable tracking of lung tumor evolutions over time, but, the best way to implement this technology into the current practice is still unknown.
Another interesting point is the optimal duration of treatment with CPIs in metastatic NSCLC patients. There is the suggestion that in patients with advanced melanoma responding to CPIs, after a variable predefined period, treatment may be discontinued. It would be interesting to see if such an algorithm would be feasible in metastatic NSCLC. We think a similar approach may decrease the duration of treatment on CPIs and theoretically lessen financial and toxicity burden in selected patients, raising the possibility of further research in NSCLC.
The main strength of our study is the network meta-analysis framework, and utilization of the meta-regression analysis for some of the possible predictors of treatment efficacy with CPIs, like the PDL-1 status, line of therapy, and tumor histology. However, some of the limitations to consider are short follow-up times for some of the trials included in this article, and the availability and usage of different CPIs in the first- and second-line settings. Obviously, an individual patient data meta-analysis will be more informative if conducted for the same CPI (pembrolizumab, for example), and at the same setting; like first-line treatment.
In short, we show that, in terms of patient care and also from a policy perspective, for metastatic NSCLC patients without driver mutations and regardless of tumor histology, chemotherapy and a CPI (like pembrolizumab) may be a superior frontline treatment strategy, whereas at the 2nd line, a CPI – nivolumab, pembrolizumab or atezolizumab – alone is comparable in efficacy and superior to chemotherapy. To the best of our knowledge, for the first time, we also determine the relative rankings of the efficacy of CPIs in metastatic NSCLC, both in chemotherapy-sensitive and refractory patients.
| > Conclusion|| |
Our work raises the possibility that CPIs may yield different levels of benefit in various clinical circumstances. In this direction, the clinical endpoint assessed, line of treatment, whether a CPI alone or a combination with chemotherapy or another CPI is utilized as treatment, are important factors that influence the comparative therapeutic effect of CPIs.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Chemotherapy in non-small cell lung cancer: A meta-analysis using updated data on individual patients from 52 randomised clinical trials. Non-Small Cell Lung Cancer Collaborative Group. BMJ 1995;311:899-909.
Lynch TJ, Bondarenko I, Luft A, Serwatowski P, Barlesi F, Chacko R, et al
. Ipilimumab in combination with paclitaxel and carboplatin as first-line treatment in stage IIIB/IV non-small-cell lung cancer: Results from a randomized, double-blind, multicenter phase II study. J Clin Oncol 2012;30:2046-54.
Brahmer J, Reckamp KL, Baas P, Crinò L, Eberhardt WE, Poddubskaya E, et al
. Nivolumab versus docetaxel in advanced squamous-cell non-small-cell lung cancer. N
Engl J Med 2015;373:123-35.
Fehrenbacher L, Spira A, Ballinger M, Kowanetz M, Vansteenkiste J, Mazieres J, et al
. Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): A multicentre, open-label, phase 2 randomised controlled trial. Lancet 2016;387:1837-46.
Hargadon KM, Johnson CE, Williams CJ. Immune checkpoint blockade therapy for cancer: An overview of FDA-approved immune checkpoint inhibitors. Int Immunopharmacol 2018;62:29-39.
Hutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C, et al
. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: Checklist and explanations. Ann Intern Med 2015;162:777-84.
Doi SA, Barendregt JJ. A generalized pairwise modelling framework for network meta-analysis. Int J Evid Based Healthc 2018;16:187-94.
Woods BS, Hawkins N, Scott DA. Network meta-analysis on the log-hazard scale, combining count and hazard ratio statistics accounting for multi-arm trials: A tutorial. BMC Med Res Methodol 2010;10:54.
Langer CJ, Gadgeel SM, Borghaei H, Papadimitrakopoulou VA, Patnaik A, Powell SF, et al
. Carboplatin and pemetrexed with or without pembrolizumab for advanced, non-squamous non-small-cell lung cancer: A randomised, phase 2 cohort of the open-label KEYNOTE-021 study. Lancet Oncol 2016;17:1497-508.
Reck M, Rodríguez-Abreu D, Robinson AG, Hui R, Csőszi T, Fülöp A, et al
. Pembrolizumab versus chemotherapy for PD-L1-Positive non-small-cell lung cancer. N
Engl J Med 2016;375:1823-33.
Carbone DP, Reck M, Paz-Ares L, Creelan B, Horn L, Steins M, et al
. First-line nivolumab in stage IV or recurrent non-small-cell lung cancer. N
Engl J Med 2017;376:2415-26.
Socinski MA, Jotte RM, Cappuzzo F, Orlandi F, Stroyakovskiy D, Nogami N, et al
. Atezolizumab for first-line treatment of metastatic nonsquamous NSCLC. N
Engl J Med 2018;378:2288-301.
Gandhi L, Rodríguez-Abreu D, Gadgeel S, Esteban E, Felip E, De Angelis F, et al
. Pembrolizumab plus chemotherapy in metastatic non-small-cell lung cancer. N
Engl J Med 2018;378:2078-92.
Hellmann MD, Ciuleanu TE, Pluzanski A, Lee JS, Otterson GA, Audigier-Valette C, et al
. Nivolumab plus ipilimumab in lung cancer with a high tumor mutational burden. N
Engl J Med 2018;378:2093-104.
Paz-Ares L, Luft A, Vicente D, Tafreshi A, Gümüş M, Mazières J, et al
. Pembrolizumab plus chemotherapy for squamous non-small-cell lung cancer. N
Engl J Med 2018;379:2040-51.
Mok TS, Wu YL, Kudaba I, Kowalski DM, Cho BC, Turna HZ, et al
. Pembrolizumab versus chemotherapy for previously untreated, PD-L1-expressing, locally advanced or metastatic non-small-cell lung cancer (KEYNOTE-042): A randomised, open-label, controlled, phase 3 trial. Lancet 2019;393:1819-30.
Jotte R, Cappuzzo F, Vynnychenko I, Stroyakovskiy D, Abreu DR, Hussein M, et al
. IMpower131: Primary PFS and Safety Analysis of a Randomized Phase 3 Study of Atezolizumab + Carboplatin + Paclitaxel or Nab-Paclitaxel versus Carboplatin + Nab-Paclitaxel as 1L therapy in advanced Squamous NSCLC. ASCO Annual Meeting; 2018. Available from: http://www.ascopost.com/News/58899
. [Last accessed on 2018 Jul 31].
Herbst RS, Baas P, Kim DW, Felip E, Pérez-Gracia JL, Han JY, et al
. Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): A randomised controlled trial. Lancet 2016;387:1540-50.
Borghaei H, Paz-Ares L, Horn L, Spigel DR, Steins M, Ready NE, et al
. Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer. N
Engl J Med 2015;373:1627-39.
Rittmeyer A, Barlesi F, Waterkamp D, Park K, Ciardiello F, von Pawel J, et al
. Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): A phase 3, open-label, multicentre randomised controlled trial. Lancet 2017;389:255-65.
Barlesi F, Vansteenkiste J, Spigel D, Ishii H, Garassino M, de Marinis F, et al
. Avelumab versus docetaxel in patients with platinum-treated advanced non-small-cell lung cancer (JAVELIN Lung 200): An open-label, randomised, phase 3 study. Lancet Oncol 2018;19:1468-79.
Gibney GT, Weiner LM, Atkins MB. Predictive biomarkers for checkpoint inhibitor-based immunotherapy. Lancet Oncol 2016;17:e542-51.
Pai-Scherf L, Blumenthal GM, Li H, Subramaniam S, Mishra-Kalyani PS, He K, et al
. FDA Approval summary: Pembrolizumab for treatment of metastatic non-small cell lung cancer:First-line therapy and beyond. Oncologist 2017;22:1392-9.
Shaughnessy AF. Monoclonal antibodies: Magic bullets with a hefty price tag. BMJ 2012;345:e8346.
Wang SJ, Hung HM, O'Neill RT. Adapting the sample size planning of a phase III trial based on phase II data. Pharm Stat 2006;5:85-97.
Manson G, Norwood J, Marabelle A, Kohrt H, Houot R. Biomarkers associated with checkpoint inhibitors. Ann Oncol 2016;27:1199-206.
Tsao MS, Kerr KM, Kockx M, Beasley MB, Borczuk AC, Botling J, et al
. PD-L1 Immunohistochemistry Comparability Study in Real-Life Clinical Samples: Results of Blueprint Phase 2 Project. J Thorac Oncol 2018;13:1302-11.
Roach C, Zhang N, Corigliano E, Jansson M, Toland G, Ponto G, et al
. Development of a Companion diagnostic PD-L1 immunohistochemistry assay for pembrolizumab therapy in non-small-cell lung cancer. Appl Immunohistochem Mol Morphol 2016;24:392-7.
Russo A, De Miguel Perez D, Gunasekaran M, Scilla K, Lapidus R, Cooper B, et al
. Liquid biopsy tracking of lung tumor evolutions over time. Expert Rev Mol Diagn 2019;19:1099-108.
Jansen YJ, Rozeman EA, Mason R, Goldinger SM, Geukes Foppen MH, Hoejberg L, et al
. Discontinuation of anti-PD-1 antibody therapy in the absence of disease progression or treatment limiting toxicity: Clinical outcomes in advanced melanoma. Ann Oncol 2019;30:1154-61.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]