Abstract
Background
The current therapy for advanced non-small-cell lung cancer (NSCLC) often incorporates frontline use of programmed death (ligand) 1 [PD-(L)1] inhibitors concomitantly or sequentially with platinum-based chemotherapy. For patients whose cancer then progresses, the impact of prior immunotherapy on the efficacy of subsequent chemotherapy remains unclear. This retrospective study aimed to evaluate survival with taxane-containing regimens after prior chemotherapy and immunotherapy.
Methods
Using a US nationwide, longitudinal deidentified database, we selected patients ≥18 years old with unresectable stage IIIB-IV NSCLC, without actionable mutations, who had ECOG performance status of 0-1 when initiating a taxane-containing regimen from 4 March 2015 to 31 May 2021 after prior PD-(L)1 inhibitor and platinum-based chemotherapy, concomitantly or sequentially. Overall survival (OS) was estimated using the Kaplan-Meier method. The data cut-off was 31 May 2022.
Results
Of 587 eligible patients [median age, 68 years, 316 (54%) male, 560 (95%) with nonsquamous NSCLC], 528 (90%) had received one to two prior lines of therapy before taxane monotherapy (195 patients; 33%) or combination therapy (392; 67%). The median patient follow-up to death or data cut-off was 9.2 months (range <0.1-71.8 months). The median OS from taxane initiation was 9.0 months [95% confidence interval (CI) 8.1-9.6 months] and OS rates were 39% and 16% at 12 and 24 months, respectively. The median OS was similar with taxane monotherapy (9.0 months; 95% CI 8.1-11.2 months) and taxane combination therapy (8.8 months; 95% CI 7.5-9.6 months).
Conclusions
Survival with taxane-containing regimens after prior chemotherapy and immunotherapy is consistent with historical outcomes that predate immunotherapy. These findings highlight the need for more effective subsequent treatments after prior PD-(L)1 inhibitor and platinum-based chemotherapy for patients with advanced NSCLC without actionable mutations.
Key words: docetaxel, (nab-)paclitaxel, PD-(L)1 inhibitor, platinum-based chemotherapy, overall survival
Highlights
-
•
It is unclear whether anti-PD-(L)1 therapy for NSCLCs improves response to subsequent therapy.
-
•
Taxane therapy was most common in the second/third line after frontline anti-PD-(L)1 + platinum.
-
•
The median OS was <1 year after the taxane-containing regimen, with 39% survival at 1 year.
-
•
Need for more effective subsequent therapy after progression on anti-PD-(L)1 + platinum.
Introduction
Standard frontline systemic therapy for advanced and metastatic non-small-cell lung cancer (NSCLC) has evolved greatly within the past decade with the development and integration of immunotherapies and targeted therapies.1,2 Platinum-based chemotherapy was historically the standard first-line therapy for advanced and metastatic NSCLC. The current therapy for advanced NSCLC without actionable biomarker mutations now incorporates the early use of immunotherapy with an inhibitor of programmed cell death protein 1 (PD-1) or programmed death-ligand 1 (PD-L1), often concomitantly or sequentially with platinum-based chemotherapy.
For patients whose cancer progresses after a PD-(L)1 inhibitor and platinum-based chemotherapy, treatment options are generally limited to single-agent chemotherapies; and recommendations for subsequent therapies have changed little in recent years.3 For patients with stage IV NSCLC and Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1 who have already received a PD-(L)1 inhibitor and platinum-based chemotherapy, the recommended subsequent therapies (if not previously administered) in the American Society of Clinical Oncology (ASCO) guidelines are as follows: docetaxel, pemetrexed (nonsquamous only), or weekly paclitaxel and bevacizumab.3 The National Comprehensive Cancer Network (NCCN) recommendations are similar and include docetaxel, pemetrexed (nonsquamous only), gemcitabine, ramucirumab and docetaxel, or nanoparticle albumin-bound paclitaxel (nab-paclitaxel) as treatment options (if not previously administered) for patients with advanced or metastatic NSCLC that has progressed after previous immunotherapy.2
The median overall survival (OS) after initiation of these subsequent therapies tends to be poor after platinum-based therapy with no prior or concomitant immunotherapy. For example, among patients in the control arms of recent clinical trials of immunotherapy for previously treated, platinum-refractory advanced NSCLC, the median OS after docetaxel and placebo initiation ranged from 6.0 to 9.6 months.4, 5, 6
Recently, the question has arisen as to whether prior receipt of immunotherapy could result in improved response to subsequent therapy, the proposed theories being that the antitumor effects of immunotherapy may remain active for a time or provide a synergistic benefit with subsequent chemotherapy.7 The findings of two small studies from 2017, each including fewer than 100 patients, suggested that prior exposure to a PD-(L)1 inhibitor resulted in better-than-expected response to subsequent platinum-based or nonplatinum therapy, most commonly docetaxel (also pemetrexed, gemcitabine, and paclitaxel).7,8 The phase III CANOPY-2 trial included patients who received previous therapy with a PD-(L)1 inhibitor and platinum-based chemotherapy; the 117 patients in this study treated with docetaxel and placebo had a median OS of 11.3 months.9 By contrast, two larger, more recent retrospective studies of advanced or metastatic NSCLC without EGFR or ALK genomic alterations did not support this supposition, as the median OS was 6.4 and 7.3 months after subsequent therapy for patients who had received prior immunotherapy and platinum-based chemotherapy, respectively.10,11
Previous observational studies evaluating the impact of prior immunotherapy on the efficacy of subsequent chemotherapy did not follow patients beyond 2019.7,8,10,11 We aimed to update these findings and understand the current use of taxane-containing regimens and survival in clinical oncology practice after prior administration of a PD-(L)1 inhibitor and platinum-based chemotherapy for patients with good performance status who had advanced or metastatic NSCLC without actionable biomarker mutations.
Patients and methods
Data source
This study drew on the United States (US) nationwide, longitudinal Flatiron Health database, an electronic health record (EHR)-derived database containing deidentified (pseudonymous), patient-level structured and unstructured data, curated via technology-enabled abstraction. As previously described,12, 13, 14 this database provides real-world patient characteristics, treatment patterns, and outcomes of cancer therapy, refreshed every month, and has been used frequently for oncological research.15, 16, 17 At the time of the study, the deidentified data originated from ∼280 cancer clinics (∼800 sites of care) throughout the US.
Institutional Review Board (IRB) approval of the study protocol was obtained from the WCG IRB,18 a centralized, commercial board of IRB experts in the US that provides independent ethical review, with a waiver of informed consent granted for working with deidentified data. Safeguards were maintained to prevent patient reidentification during the analyses, as identifiers were removed, masked/hashed, and recoded by Flatiron Health before delivery of the dataset for analysis.
Patients
Patients 18 years and older with advanced or metastatic NSCLC, with no documented EGFR-sensitizing mutation or ALK or ROS1 rearrangement, who initiated a taxane-containing regimen by 31 May 2021 were eligible for this study if they had received prior therapy for advanced NSCLC with a PD-(L)1 inhibitor (≥2 doses) and platinum-based chemotherapy, concomitantly or sequentially. In addition, eligible patients had ECOG PS of 0 or 1 at the time of taxane initiation, as per usual clinical trial inclusion criteria. To ensure that prior PD-(L)1 inhibitor utilization was aligned with regulatory-approved therapy, the study period began on 4 March 2015, when the first PD-(L)1 inhibitor (nivolumab) was approved in the US for patients with previously treated NSCLC. Data cut-off was 31 May 2022, to allow for a minimum of 12 months of potential follow-up after taxane initiation.
Advanced NSCLC was histologically confirmed as unresectable stage IIIB, IIIC, or IV at initial diagnosis or upon recurrence. For nonsquamous NSCLC, eligibility required confirmation of wild-type EGFR and ALK, plus no known ROS1 rearrangement, whereas for squamous and not otherwise specified NSCLC, no known EGFR mutation or ALK or ROS1 rearrangement was sufficient (molecular testing was not required). A taxane-containing regimen was defined as docetaxel, paclitaxel, or nab-paclitaxel administered as monotherapy or as part of a combination regimen, excluding any immunotherapy or targeted therapy. The latter category was defined as agents targeting genomic alterations, excluding anti-vascular endothelial growth factor (anti-VEGF) therapies such as bevacizumab and ramucirumab.
Patients enrolled in a clinical trial or with no clinic visit within 90 days of the advanced NSCLC diagnosis were excluded from the analyses. We also excluded patients who had received more than one type of PD-(L)1 inhibitor and those with evidence of inadequate renal, hepatic, or hematologic function based on laboratory values chosen in line with values specified for the LEAP-008 clinical trial (NCT03976375). We defined adequate renal function as a creatinine level of ≤1.5 times the upper limit of normal (ULN) for the patient’s laboratory or a creatinine clearance of ≥30 ml/min, calculated using the Cockcroft–Gault method. Adequate hepatic function was defined as aspartate aminotransferase and alanine aminotransferase levels ≤2.5 times the ULN and total bilirubin level ≤1.5 times the ULN for the patient’s laboratory (or direct bilirubin ≤ULN for patients with total bilirubin >1.5 times the ULN). Adequate hematologic function was defined per KEYNOTE-010 inclusion criteria as a hemoglobin level of ≥9.0 g/dl (≥90 g/l), platelet count ≥100 × 103/μl (≥100 × 109/l), and absolute neutrophil count ≥1500/μl (≥1.5 × 109/l).19
Study assessments and analyses
Patient characteristics at initiation of taxane therapy, together with details of prior systemic therapy and taxane-containing regimens, were summarized using descriptive statistics. The Charlson comorbidity index was derived from listed comorbidities,20,21 and lines of therapy were determined using oncologist-defined rules, as described in prior publications.15,17
The Kaplan-Meier method was used to estimate OS from initiation of the taxane-containing regimen to date of death, and patients with no recorded death were censored at the last recorded clinical activity before data cut-off. The date of death was captured using the validated Flatiron Health mortality 2.0 endpoint, a composite of structured and unstructured data from EHRs, external commercial sources, and the US Social Security Death Index.22,23
Sample size and power calculations were not performed for this descriptive study, and all eligible patients were included in the analyses, which were conducted using SAS software, version 9.4 (SAS Institute, Cary, NC, USA).
Results
Patients and prior systemic therapy
Of 4979 patients with advanced NSCLC with no documented EGFR, ALK, or ROS1 genomic alteration who initiated a taxane-containing regimen in second-line or later by the end of May 2021, 1066 patients (21%) had received prior concomitant or sequential therapy with a PD-(L)1 inhibitor and platinum-based chemotherapy. Of these 1066 patients, 587 (55%) had ECOG PS of 0 or 1 and adequate renal, hepatic, and hematologic function at the start of taxane therapy and thus were eligible for the study (Figure 1).
Figure 1.
Flow diagram depicting patient selection from the Flatiron Health database.
aA record of wild-type EGFR and ALK status was required for patients with nonsquamous tumors.
1L, first-line; 2L+, second-line or later; ECOG PS, Eastern Cooperative Oncology Group performance status; NSCLC, non-small-cell lung cancer; PD-1, programmed cell death protein 1; PD-L1, programmed death-ligand 1.
Slightly over one-half of the 587 eligible patients were male (54%), the median age overall was 68 years, and 80% of patients were White; 95% of patients were treated at community oncology clinics (rather than academic centers; Table 1). Most patients (95%) had nonsquamous NSCLC, and 84% had advanced NSCLC at initial diagnosis, including 462 (80%) at stage IV; 81 patients (14%) had brain metastases. The characteristics of the 195 patients (33%) who were prescribed taxane monotherapy were similar to those of the 392 patients (67%) who were prescribed taxane combination therapy, although 28% of patients who received taxane monotherapy were ≥75 years (versus 19% in the taxane combination cohort; Table 1).
Table 1.
Characteristics of patients with advanced NSCLC at initiation of taxane-containing therapy in the second line or later
| Characteristic | All patients (N = 587) | Taxane monotherapy (n = 195) | Taxane combination regimen (n = 392) |
|---|---|---|---|
| Sex, male | 316 (53.8) | 113 (57.9) | 203 (51.8) |
| Age, median (range), years | 68 (35-84) | 69 (47-82) | 66 (35-84) |
| Age group (years) | |||
| <75 | 456 (77.7) | 140 (71.8) | 316 (80.6) |
| ≥75 | 131 (22.3) | 55 (28.2) | 76 (19.4) |
| Racea | |||
| White | 431 (79.8) | 149 (82.3) | 282 (78.6) |
| Black or African American | 50 (9.3) | 15 (8.3) | 35 (9.7) |
| Other | 59 (10.9) | 17 (9.4) | 42 (11.7) |
| Unknown, n | 47 | 14 | 33 |
| Smoking status | |||
| Positive history of smoking | 537 (91.5) | 181 (92.8) | 356 (90.8) |
| No history of smoking | 50 (8.5) | 14 (7.2) | 36 (9.2) |
| Practice type | |||
| Community | 560 (95.4) | 189 (96.9) | 371 (94.6) |
| Academic or both | 27 (4.6) | 6 (3.1) | 21 (5.4) |
| ECOG performance status | |||
| 0 | 170 (29.0) | 43 (22.1) | 127 (32.4) |
| 1 | 417 (71.0) | 152 (77.9) | 265 (67.6) |
| Charlson comorbidity index | |||
| Mean (SD) | 2.1 (3.2) | 1.6 (2.8) | 2.3 (3.3) |
| Median (range) | 0 (0-11) | 0 (0-11) | 0 (0-11) |
| NSCLC histologic type | |||
| Nonsquamous | 560 (95.4) | 183 (93.8) | 377 (96.2) |
| Squamous | 8 (1.4) | 3 (1.5) | 5 (1.3) |
| Not otherwise specified | 19 (3.2) | 9 (4.6) | 10 (2.6) |
| Record of brain metastasis | 81 (13.8) | 26 (13.3) | 55 (14.0) |
| Advanced stage at initial diagnosisa,b | 485 (83.9) | 156 (81.7) | 329 (85.0) |
| PD-L1 expressiona | |||
| ≥50% | 111 (26.6) | 24 (19.8) | 87 (29.3) |
| 1%-49% | 136 (32.5) | 43 (35.5) | 93 (31.3) |
| <1% | 171 (40.9) | 54 (44.6) | 117 (39.4) |
| Unknown, n | 169 | 74 | 95 |
| BRAF status (nonsquamous only), na | 560 | 183 | 377 |
| BRAF mutant | 31 (6.7) | 9 (6.0) | 22 (7.0) |
| BRAF wild type | 433 (93.3) | 140 (94.0) | 293 (93.0) |
| Unknown, n | 96 | 34 | 62 |
| KRAS status (nonsquamous only), na | 560 | 183 | 377 |
| KRAS mutant | 185 (41.1) | 59 (41.0) | 126 (41.2) |
| KRAS wild-type | 265 (58.9) | 85 (59.0) | 180 (58.8) |
| Unknown, n | 110 | 39 | 71 |
Note: Data are n (%) unless otherwise noted. Percentages may not add up to 100 because of rounding.
ECOG, Eastern Cooperative Oncology Group; PD-L1, programmed death-ligand 1; SD, standard deviation.
Percentages for race, advanced stage at initial diagnosis, PD-L1 expression, BRAF mutation status, and KRAS mutation status represent the percentages of patients with available data.
Advanced stage at initial diagnosis included stages IIIB, IIIC, and IV. (Nine patients had no recorded stage at diagnosis.)
Approximately 90% of patients had received either one or two prior lines of therapy; 8% had received three prior lines, and 2% had received four or five prior lines (Table 2). A total of 524 patients (89%) had received a PD-(L)1 inhibitor in the line of therapy immediately preceding taxane-containing therapy. Pembrolizumab was the most common prior PD-(L)1 inhibitor (68%), followed by nivolumab (31%; Table 2).
Table 2.
Prior therapy for patients with advanced NSCLC treated with taxane-containing therapy in the second line or later
| Prior therapy | All patients (N = 587) | Taxane monotherapy (n = 195) | Taxane combination regimen (n = 392) |
|---|---|---|---|
| Prior lines of therapy, n (%) | |||
| 1 | 274 (46.7) | 74 (37.9) | 200 (51.0) |
| 2 | 254 (43.3) | 96 (49.2) | 158 (40.3) |
| 3 | 46 (7.8) | 17 (8.7) | 29 (7.4) |
| 4-5 | 13 (2.2) | 8 (4.1) | 5 (1.3) |
| PD-(L)1 inhibitor in prior linea, n (%) | 524 (89.3) | 171 (87.7) | 353 (90.1) |
| In combination with platinum | 313 (59.7) | 82 (48.0) | 231 (65.4) |
| Not with platinum | 211 (40.3) | 89 (52.0) | 122 (34.6) |
| Prior PD-(L)1 inhibitor therapy, n (%) | |||
| Pembrolizumab | 399 (68.0) | 114 (58.5) | 285 (72.7) |
| Nivolumab | 179 (30.5) | 77 (39.5) | 102 (26.0) |
| Atezolizumab | 8 (1.4) | 4 (2.1) | 4 (1.0) |
| Durvalumab | 1 (0.2) | 0 | 1 (0.3) |
PD-1, programmed cell death protein 1; PD-L1, programmed death-ligand 1.
PD-(L)1 inhibitor in the prior line denotes administration of PD-(L)1 inhibitor in the line of therapy immediately preceding taxane-containing therapy.
Taxane therapy and survival outcomes
A total of 195 patients (33%) received taxane monotherapy and 392 (67%) received combination therapy (Table 3). Docetaxel was the most common taxane monotherapy, administered to 69% of patients who received monotherapy; and docetaxel–ramucirumab was the most common taxane combination regimen (57%). The majority of patients initiated the taxane-containing regimen in 2018 or later (77%; Table 3).
Table 3.
Taxane-containing regimens administered in the second line or later
| Variable | Values (N = 587) |
|---|---|
| Year of taxane initiation | |
| 2015-2016 | 48 (8.2) |
| 2017 | 85 (14.5) |
| 2018 | 121 (20.6) |
| 2019 | 138 (23.5) |
| 2020 | 141 (24.0) |
| 2021 | 54 (9.2) |
| Taxane monotherapy, n | 195 |
| Docetaxel | 134 (68.7) |
| Paclitaxel/nab-paclitaxel | 61 (31.3) |
| Taxane combination regimen, n | 392 |
| Docetaxel + ramucirumab | 223 (56.9) |
| Carboplatin + paclitaxel/nab-paclitaxel | 80 (20.4) |
| Bevacizumab + carboplatin + paclitaxel/nab-paclitaxel | 29 (7.4) |
| Docetaxel + gemcitabine | 13 (3.3) |
| Carboplatin + docetaxel | 6 (1.5) |
| Bevacizumab + nab-paclitaxel | 5 (1.3) |
| Paclitaxel + ramucirumab | 4 (1.0) |
| Bevacizumab-awwb + carboplatin + paclitaxel | 2 (0.5) |
| Gemcitabine + nab-paclitaxel | 2 (0.5) |
| Other taxane combinationa | 28 (7.1) |
Note: Data are n (%) unless otherwise noted.
Nab-paclitaxel, nanoparticle albumin-bound paclitaxel.
A total of 28 patients each received a different taxane combination regimen, 17 of which (61%) included docetaxel.
Taxane-containing therapy was administered a median of 1 year after the advanced NSCLC diagnosis (Table 4).
Table 4.
Overall survival for patients with advanced NSCLC after initiating a taxane-containing regimen as a second-line or later therapy
| Variables | All patients (N = 587) | Taxane monotherapy (n = 195) | Taxane combination regimen (n = 392) |
|---|---|---|---|
| Advanced diagnosis to taxane initiation, median (range), months | 12.4 (1.7-96.7) | 14.1 (2.5-96.7) | 11.1 (1.7-72.3) |
| Study follow-up, median (range), monthsa | 37.9 (12.2-81.4) | 41.4 (12.6-80.3) | 36.3 (12.2-81.4) |
| Patient follow-up, median (range), monthsa | 9.2 (<0.1-71.8) | 9.3 (<0.1-60.0) | 9.1 (<0.1-71.8) |
| Overall survival (OS) | |||
| Events, n (%) | 481 (81.9) | 161 (82.6) | 320 (81.6) |
| Median OS (95% CI), months | 9.0 (8.1-9.6) | 9.0 (8.1-11.2) | 8.8 (7.5-9.6) |
| OS rate (95% CI), % | |||
| At 12 months | 39.0 (35.0% to 43.1%) | 40.1 (33.1% to 47.1%) | 38.5 (33.6% to 43.4%) |
| At 24 months | 16.2 (13.0% to 19.6%) | 17.4 (12.0% to 23.7%) | 15.5 (11.8% to 19.7%) |
| At 36 months | 8.2 (5.6% to 11.4%) | 6.5 (2.9% to 12.0%) | 9.2 (6.0% to 13.3%) |
Study follow-up was defined as the duration of follow-up from taxane regimen initiation to database cut-off (31 May 2022). Patient follow-up was defined as the time from taxane initiation to the date of death or data cut-off, whichever occurred first.
The median patient follow-up from taxane initiation to death or data cut-off, whichever occurred first, was 9.2 months (range <0.1-71.8 months) and was similar for patients who received taxane monotherapy or combination therapy (Table 4). The overall median study follow-up time from initiation of taxane-containing therapy to data cut-off was just over 3 years (37.9 months).
The median OS was 9.0 months [95% confidence interval (CI) 8.1-9.6 months], with OS rates of 39% and 16% at 12 and 24 months, respectively, after the initiation of taxane-containing therapy (Table 4). Survival rates and patterns appeared similar in the taxane monotherapy and combination therapy cohorts: median OS was 9.0 months (95% CI 8.1-11.2 months) with taxane monotherapy and 8.8 months (95% CI 7.5-9.6 months) with taxane combination therapy (Table 4; Figure 2).
Figure 2.
Overall survival by taxane monotherapy or taxane combination regimen.
CI, confidence interval; OS, overall survival.
For the 134 patients who received docetaxel monotherapy, the median OS was 10.5 months (95% CI 8.2-13.1 months), and OS rates were 45% at 12 months and 19% at 24 months, as depicted in Supplementary Methods, Table S1. For the 223 patients who received docetaxel–ramucirumab combination therapy, the median OS was 7.6 months (95% CI 6.6-8.9 months), and OS rates were 33% and 15% at 12 and 24 months, respectively (Supplementary Methods, Table S1).
Discussion
This retrospective study included almost 600 patients with advanced or metastatic NSCLC and no actionable biomarker mutations who were treated in US clinical practice with a taxane-containing regimen initiated from March 2015 to May 2021 after exposure to both a PD-(L)1 inhibitor and platinum-based chemotherapy. We were able to describe current prescribing patterns for taxane-containing therapies, which were administered in the second or third line to 90% of studied patients and most commonly included docetaxel as monotherapy or in combination with ramucirumab. The median study follow-up was over 3 years (37.8 months) from taxane initiation to data cut-off. The median patient follow-up after taxane therapy initiation was 9.2 months to death or data cut-off, and the median OS of 9.0 months for patients in this study, most of whom had nonsquamous NSCLC, was similar to historical controls in the immunotherapy-naïve setting, such as that reported for the immunotherapy-naïve docetaxel arm (9.4 months) of the CheckMate 057 trial, investigating nivolumab therapy versus docetaxel for previously treated nonsquamous advanced NSCLC.24
Our findings support the results of two earlier retrospective studies that also used the Flatiron Health database to examine treatment patterns and survival after PD-(L)1 inhibitor and platinum-based chemotherapy for advanced NSCLC without documented EGFR or ALK genomic alterations.10,11 Bains et al.10 reported a median OS of 6.4 months (95% CI 5.3-7.6 months) after initiation of second-line therapy for 145 patients in their study (cohort 2) previously treated with first-line combination PD-(L)1 inhibitor and platinum-based chemotherapy; the second-line therapy was initiated from 4 March 2015 to 31 October 2018, with data cut-off on 30 April 2019. In the second study, presented at the 2021 World Conference on Lung Cancer, Smith et al.11 reported findings for 2294 patients with advanced NSCLC diagnosed from 2015 to February 2020. The median OS was 7.3 months (95% CI 6.9-7.7 months) overall, including 29% of patients who were treated with a regimen containing docetaxel. For those who received docetaxel monotherapy and docetaxel plus ramucirumab, the median OS was 6.7 and 6.5 months, respectively, while in the present study, the median OS was 10.5 and 7.6 months with docetaxel monotherapy and docetaxel plus ramucirumab, respectively. The study by Smith et al.11 differed in several ways from our study, namely, (i) ECOG PS was not limited to 0 or 1; (ii) patients were not excluded for inadequate renal, hepatic, or hematological function; (iii) there was no requirement for two or more doses of PD-(L)1 inhibitor; and (iv) subsequent therapy was not limited to taxane-containing regimens.
Our goal in selecting a study population with ECOG PS of 0 or 1 and adequate renal, hepatic, and hematologic function was to study patients with at least fair underlying health and functional status so that the survival outcomes would reflect the impact of the primary treatment (i.e. taxane-based regimen). We took special consideration in this regard because one of the objectives was to understand if taxane therapy, when used following a PD-(L)1 inhibitor, would result in better outcomes (versus when used after chemotherapy). Therefore, we wanted to ensure that the real-world cohort would reflect the control arms of the relevant trials (e.g. OAK, CheckMate 057, CheckMate 017, and KEYNOTE-0104, 5, 6) as much as was feasible for retrospective analyses using a deidentified data source.
The OS results of our analyses are similar to those recorded previously for patients with platinum-refractory advanced NSCLC.4, 5, 6 These findings suggest that patients still have poor survival outcomes if their disease has progressed on prior PD-(L)1 inhibitor and platinum therapy, highlighting the need for more effective systemic therapy after prior immunotherapy and platinum-based chemotherapy. Ongoing phase III clinical trials are currently evaluating the efficacy and safety of PD-(L)1 inhibitor-containing combination regimens for this patient population, including LEAP-008 (NCT03976375), SAPPHIRE (NCT03906071), and CONTACT-01 (NCT04471428). Of note, a recent phase II study found that subsequent therapy with the combination of ramucirumab and pembrolizumab significantly improved OS when compared with standard-of-care therapy (hazard ratio 0.69; 80% CI 0.51-0.92), as chosen by investigators, including docetaxel, ramucirumab/docetaxel, gemcitabine, or, for nonsquamous NSCLC only, pemetrexed.25
The strengths of our study are the description of contemporary prescribing patterns for second-line and later taxane-containing regimens at US community oncology practices over a period when PD-(L)1 inhibitor therapy became a well-established frontline therapy for advanced NSCLC. With taxane initiation captured until the end of May 2021, and follow-up until the end of May 2022, we were able to analyze outcomes with a minimum of 1 year of potential patient follow-up. The Flatiron Health database is well-regarded and frequently used for oncologic research, and the patient population was well-characterized, including ECOG PS, smoking history, and NSCLC histologic type.
A key study limitation, however, is that information was not available in the database regarding the disease progression status of patients during their previous therapy. Moreover, most patients were treated at community oncology centers; therefore study results may not be generalizable to academic centers or represent outcomes for patients treated outside of the Flatiron Health EHR system. In addition, selection bias cannot be ruled out, as patients with unknown ECOG PS were excluded; thus the care trajectory and outcomes for the selected patients with recorded PS may differ (from those without recorded PS), a possibility suggested by a study in which patients with advanced NSCLC tested for biomarkers had longer OS than those not tested.26 As for all EHR-based studies, errors in EHR data recording at oncology clinics are possible. Finally, all analyses were descriptive, using data from a nonrandomized patient population with NSCLC, and no inferential statistical analysis was conducted; therefore no conclusions regarding causality can be drawn directly from this work.
Conclusions
Taxane-containing regimens are currently used primarily in the second- and third-line settings after exposure to immunotherapy and platinum-based chemotherapy. The OS estimates in this real-world cohort of patients with prior exposure to a PD-(L)1 inhibitor and platinum-based chemotherapy who received a taxane-containing regimen, as either monotherapy or combination therapy, are consistent with historical data of patients with prior exposure to platinum before the availability of immunotherapy. These findings suggest a continuing area of unmet need for effective subsequent therapy for patients with immunotherapy- and platinum-refractory advanced NSCLC without actionable molecular biomarkers.
Acknowledgements
We gratefully acknowledge Yu (Julia) Feng and Mary Anne Rutkowski for programming support (both of Merck & Co., Inc., Rahway, NJ, USA). Medical writing and editorial assistance were provided by Elizabeth V. Hillyer, DVM (freelance). This assistance was funded by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.
Funding
This work was supported by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA. Employees of the funder participated in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Disclosure
SVL reports serving in an advisory and/or consultant role for AbbVie, Amgen, AstraZeneca, Boehringer-Ingelheim, Bristol-Myers Squibb, Catalyst, Daiichi Sankyo, Eisai, Elevation Oncology, Genentech, Gilead, Guardant Health, Janssen, Jazz Pharmaceuticals, Merus, MSD, Novartis, Regeneron, Sanofi, Takeda, and Turning Point Therapeutics; and receiving research funding for his institution from Alkermes, Blueprint, Bristol-Myers Squibb, Elevation Oncology, Genentech, Gilead, MSD, Merus, Nuvalent, Pfizer, Rain Therapeutics, RAPT, and Turning Point Therapeutics. XH, DC, and AS report full-time employment with Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA; and stock ownership of Merck & Co., Inc., Rahway, NJ, USA. At the time of this study, WM was a full-time employee of Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA; and reports stock ownership of Merck & Co., Inc., Rahway, NJ, USA.
Data sharing
The data that support the findings of this study have been originated by Flatiron Health, Inc. Requests for data sharing by license or by permission for the specific purpose of replicating results in this manuscript can be submitted to publicationsdataaccess@flatiron.com.
Supplementary Methods
Table S1.
Overall survival for patients with advanced NSCLC after initiating second-line/later taxane-containing therapy, by common regimen types
| Overall survival | Taxane monotherapy (n = 195) |
Taxane combination (n = 392) |
||
|---|---|---|---|---|
| Docetaxel (n = 134) | Paclitaxel or nab-paclitaxel (n = 61) | Docetaxel + ramucirumab (n = 223) | Other combinationa(n = 169) | |
| Events, n (%) | 107 (79.9) | 54 (88.5) | 185 (83.0) | 135 (79.9) |
| Median OS (95% CI), mo | 10.5 (8.2-13.1) | 8.1 (7.1-9.4) | 7.6 (6.6-8.9) | 10.2 (8.4-13.1) |
| OS rate, % (95% CI%) | ||||
| At 12 months | 45.4 (36.6-53.8) | 29.0 (18.2-40.7) | 32.8 (26.6-39.2) | 45.9 (38.1-53.3) |
| At 24 months | 18.5 (11.9-26.2) | 15.4 (7.0-26.8) | 14.7 (10.0-20.3) | 16.6 (11.0-23.3) |
| At 36 months | 9.4 (4.4-16.7) | n/ab | 7.5 (3.7-13.1) | 11.3 (6.4-17.8) |
mo, months; n/a, not available; OS, overall survival.
Taxane combination regimen other than docetaxel + ramucirumab.
Insufficient numbers of patients for analysis.
Supplementary data
References
- 1.Howlader N., Forjaz G., Mooradian M.J., et al. The effect of advances in lung-cancer treatment on population mortality. N Engl J Med. 2020;383(7):640–649. doi: 10.1056/NEJMoa1916623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Referenced with permission from the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines®) for Non-Small Cell Lung Cancer V.1.2023. © National Comprehensive Cancer Network, Inc. 2022. All rights reserved. Accessed February 3, 2023. To view the most recent and complete version of the guideline, go online to NCCN.org. NCCN makes no warranties of any kind whatsoever regarding their content, use or application and disclaims any responsibility for their application or use in any way.
- 3.Singh N., Jaiyesimi I.A., Ismaila N., et al. Therapy for stage IV non-small-cell lung cancer without driver alterations: ASCO Living Guideline, version 2023.1. J Clin Oncol. 2023;41(15):e51–e62. doi: 10.1200/JCO.23.00282. [DOI] [PubMed] [Google Scholar]
- 4.Herbst R.S., Garon E.B., Kim D.W., et al. Five year survival update from KEYNOTE-010: pembrolizumab versus docetaxel for previously treated, programmed death-ligand 1-positive advanced NSCLC. J Thorac Oncol. 2021;16(10):1718–1732. doi: 10.1016/j.jtho.2021.05.001. [DOI] [PubMed] [Google Scholar]
- 5.Horn L., Spigel D.R., Vokes E.E., et al. Nivolumab versus docetaxel in previously treated patients with advanced non-small-cell lung cancer: two-year outcomes from two randomized, open-label, phase III trials (CheckMate 017 and CheckMate 057) J Clin Oncol. 2017;35(35):3924–3933. doi: 10.1200/JCO.2017.74.3062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Rittmeyer A., Barlesi F., Waterkamp D., 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(10066):255–265. doi: 10.1016/S0140-6736(16)32517-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Park S.E., Lee S.H., Ahn J.S., Ahn M.J., Park K., Sun J.M. Increased response rates to salvage chemotherapy administered after PD-1/PD-L1 inhibitors in patients with non-small cell lung cancer. J Thorac Oncol. 2018;13(1):106–111. doi: 10.1016/j.jtho.2017.10.011. [DOI] [PubMed] [Google Scholar]
- 8.Leger P.D., Rothschild S., Castellanos E., Pillai R.N., York S.J., Horn L. Response to salvage chemotherapy following exposure to immune checkpoint inhibitors in patients with non-small cell lung cancer (abstract) J Clin Oncol. 2017;35(suppl 15):9084. [Google Scholar]
- 9.Paz-Ares L., Goto Y., Lim W.D.T., et al. 1194MO Canakinumab (CAN) + docetaxel (DTX) for the second- or third-line (2/3L) treatment of advanced non-small cell lung cancer (NSCLC): CANOPY-2 phase III results (abstract) Ann Oncol. 2021;32(suppl 5):S953–S954. [Google Scholar]
- 10.Bains S., Kalsekar A., Amiri K.I., Weiss J. Real-world treatment patterns and outcomes among patients with metastatic NSCLC previously treated with programmed cell death protein-1/programmed death-ligand 1 inhibitors. JTO Clin Res Rep. 2022;3(2) doi: 10.1016/j.jtocrr.2021.100275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Smith T., Knoll S., Martinalbo J., Ye F., Kolaei F. P10.07 real-world US treatment patterns and clinical outcomes in advanced NSCLC after prior platinum chemotherapy and immunotherapy (abstract) J Thorac Oncol. 2021;16(10 suppl):S1001. [Google Scholar]
- 12.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. Preprint (not peer-reviewed) Available at. https://www.medrxiv.org/content/10.1101/2020.03.16.20037143v3. 2020. Accessed February 7, 2024.
- 13.Khozin S., Abernethy A.P., Nussbaum N.C., et al. Characteristics of real-world metastatic non-small cell lung cancer patients treated with nivolumab and pembrolizumab during the year following approval. Oncologist. 2018;23(3):328–336. doi: 10.1634/theoncologist.2017-0353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hess L.M., Krein P.M., Haldane D., Han Y., Sireci A.N. Biomarker testing for patients with advanced/metastatic nonsquamous NSCLC in the United States of America, 2015 to 2021. JTO Clin Res Rep. 2022;3(6) doi: 10.1016/j.jtocrr.2022.100336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Abernethy A.P., Arunachalam A., Burke T., et al. Real-world first-line treatment and overall survival in non-small cell lung cancer without known EGFR mutations or ALK rearrangements in US community oncology setting. PLoS One. 2017;12(6) doi: 10.1371/journal.pone.0178420. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Benedum C.M., Sondhi A., Fidyk E., et al. Replication of real-world evidence in oncology using electronic health record data extracted by machine learning. Cancers (Basel) 2023;15(6):1853. doi: 10.3390/cancers15061853. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Canavan M., Wang X., Ascha M., et al. End-of-life systemic oncologic treatment in the immunotherapy era: the role of race, insurance, and practice setting. J Clin Oncol. 2023;41(30):4729–4738. doi: 10.1200/JCO.22.02180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.WCG IRB. Available at. https://www.wcgirb.com/. Accessed February 7, 2024.
- 19.Herbst R.S., Baas P., Kim D.W., 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(10027):1540–1550. doi: 10.1016/S0140-6736(15)01281-7. [DOI] [PubMed] [Google Scholar]
- 20.Quan H., Sundararajan V., Halfon P., et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130–1139. doi: 10.1097/01.mlr.0000182534.19832.83. [DOI] [PubMed] [Google Scholar]
- 21.Deyo R.A., Cherkin D.C., Ciol M.A. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45(6):613–619. doi: 10.1016/0895-4356(92)90133-8. [DOI] [PubMed] [Google Scholar]
- 22.Carrigan G., Whipple S., Taylor M.D., et al. An evaluation of the impact of missing deaths on overall survival analyses of advanced non-small cell lung cancer patients conducted in an electronic health records database. Pharmacoepidemiol Drug Saf. 2019;28(5):572–581. doi: 10.1002/pds.4758. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zhang Q., Gossai A., Monroe S., Nussbaum N.C., Parrinello C.M. 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]
- 24.Borghaei H., Paz-Ares L., Horn L., et al. Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer. N Engl J Med. 2015;373(17):1627–1639. doi: 10.1056/NEJMoa1507643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Reckamp K.L., Redman M.W., Dragnev K.H., et al. Phase II randomized study of ramucirumab and pembrolizumab versus standard of care in advanced non-small-cell lung cancer previously treated with immunotherapy-lung-MAP S1800A. J Clin Oncol. 2022;40(21):2295–2306. doi: 10.1200/JCO.22.00912. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wu N., Ge W., Quek R.G., et al. Trends in real-world biomarker testing and overall survival in US patients with advanced non-small-cell lung cancer. Future Oncol. 2022;18(39):4385–4397. doi: 10.2217/fon-2022-0540. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.