Patients
Of 1257 patients who were assessed for eligibility, 1003 treatment-naive patients from 9 sites (Supplemental Table 1) who initiated 1L therapy for aNSCLC between June 1, 2016, and March 31, 2018, were included in the study population (Table 1). Patient data were not obtained from 1 of 10 planned sites because of capacity issues; 2 sites (Nottingham University Hospitals NHS Trust and the Clatterbridge Cancer Centre NHS Foundation Trust) contributed 56% of patients. In the study population, the median age was 68 years (range, 28–93 years), 53.9% were male, Eastern Cooperative Oncology Group performance status score was 0–1 in 75.7% and ≥ 2 in 24.3%, and tumor histology was non-squamous in 63.9%, squamous in 24.2%, and unknown in 11.9% (Table 2). All patients had metastatic disease at diagnosis.
Table 1
Summary of Patient Numbers in the Study Population and Reasons for Exclusion (Study Attrition)
| Patients |
n | % |
Adults who received 1L treatment for NSCLC | 1257 | – |
Excluded patients | 254 | 100 |
NSCLC not advanced or metastatic at diagnosis | 151 | 59.4 |
Treatment started outside of study period | 47 | 18.5 |
ECOG PS score missing | 15 | 5.9 |
Randomized trial participant | 9 | 3.5 |
Response data missing | 8 | 3.1 |
Never received treatment | 7 | 2.8 |
Duplicate patient | 5 | 2.0 |
Age missing | 4 | 1.6 |
Sex missing | 3 | 1.2 |
Not NSCLC | 2 | 0.8 |
Date of death or last hospital follow-up missing | 1 | 0.4 |
Histological diagnosis missing | 1 | 0.4 |
Diagnosis date unknown | 1 | 0.4 |
Met inclusion criteria | 1003 | – |
1L, first line; ECOG PS, Eastern Cooperative Oncology Group performance status; NSCLC, non-small cell lung cancer. |
Table 2
Patient Demographics in the Overall Population and in Subgroups Defined by 1L Drug Class Received
| All patients (n = 1003) | 1L chemotherapy (n = 698) | 1L IO monotherapy (n = 179) | 1L targeted therapy (n = 126) |
Proportion of study population, % | 100 | 69.6 | 17.8 | 12.6 |
Median follow-up (range), months | 9.2 (0.0-42.7) | 7.9 (0.0-42.7) | 12.7 (0.1–37.3) | 16.3 (0.1–37.1) |
Median age at diagnosis (range), years | 68 (28–93) | 68 (28–88) | 67 (48–90) | 70 (32–93) |
Sex, n (%) | | | | |
Male | 541 (53.9) | 395 (56.6) | 94 (52.5) | 52 (41.3) |
Female | 462 (46.1) | 303 (43.4) | 85 (47.5) | 74 (58.7) |
Tumor histology, n (%) |
Adenocarcinoma | 635 (63.3) | 387 (55.4) | 131 (73.2) | 117 (92.9) |
Squamous cell carcinoma | 243 (24.2) | 202 (28.9) | 38 (21.2) | 3 (2.4) |
Large cell carcinoma | 6 (0.6) | 4 (0.6) | 2 (1.1) | 0 |
Not specified | 119 (11.9) | 105 (15.0) | 8 (4.5) | 6 (4.8) |
TNM stage at diagnosis, n (%) |
T |
T X-4 | 938 (93.5) | 647 (92.7) | 170 (95.0) | 121 (96.0) |
N/A | 65 (6.5) | 51 (7.3) | 9 (5.0) | 5 (4.0) |
N |
N X-3 | 939 (93.6) | 648 (92.8) | 170 (95.0) | 121 (96.0) |
N/A | 64 (6.4) | 50 (7.2) | 9 (5.0) | 5 (4.0) |
M |
M1a | 524 (52.2) | 351 (50.3) | 114 (63.7) | 59 (46.8) |
M1a | 166 (16.6) | 120 (17.2) | 22 (12.3) | 24 (19.0) |
M1b | 310 (30.9) | 224 (32.1) | 43 (24.0) | 43 (34.1) |
M1c | 3 (0.3) | 3 (0.4) | 0 | 0 |
ECOG PS score at diagnosis, n (%) |
0–1 | 759 (75.7) | 513 (73.5) | 157 (87.7) | 89 (70.6) |
2+ | 244 (24.3) | 185 (26.5) | 22 (12.3) | 37 (29.4) |
EGFR + status, n (%)b |
Documented | 19 (1.9) | 1 (0.1) | 0 | 18 (14.3) |
Assumed | 89 (8.9) | 0 | 0 | 89 (70.6) |
ALK + status, n (%)b |
Documented | 2 (0.2) | 0 | 0 | 2 (1.6) |
Assumed | 17 (1.7) | 0 | 0 | 17 (13.5) |
PD-L1 + status, n (%)b |
Documented | 10 (1.0) | 3 (0.4) | 7 (3.9) | 0 |
Assumed | 172 (17.1) | 0 | 172 (96.1) | 0 |
1L, first line; ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; IO, immuno-oncology; NSCLC, non-small cell lung cancer; PD-L1, programmed death-ligand 1. |
a Includes 77 patients with clinician-defined stage IV NSCLC. |
b Biomarker status was based either on hospital test results (documented) or treatment regimen (assumed, ie, patients who received an EGFR or ALK inhibitor were assumed to have a tumor harboring an EGFR or ALK mutation, and patients receiving IO therapy were assumed to have a PD-L1 + tumor). |
Treatment Patterns and Sequencing
First-line treatment comprised chemotherapy in 698 patients (69.6%), IO monotherapy in 179 patients (17.8%), and targeted therapy in 126 patients (12.6%). Among chemotherapy-treated patients, 674 (96.6%) received platinum-based chemotherapy, and carboplatin-based doublet or triplet chemotherapy was the most commonly administered regimen (n = 499 [71.5% of chemotherapy-treated patients]; Table 3). Among patients who received 1L IO monotherapy or targeted therapy, pembrolizumab (n = 174 [97.2% of the 1L IO subgroup]) and afatinib (n = 67 [53.2% of the 1L targeted therapy subgroup]) were the most commonly administered agents, respectively. During the time period analyzed (June 2016 to March 2018), the proportions of patients receiving 1L IO monotherapy or targeted therapy increased (from 0–25.9% for IO therapy, and from 11.8–15.9% for targeted therapy), whereas the proportion of 1L chemotherapy-treated patients decreased (from 88.2–58.2%; Fig. 1).
Table 3
First-Line and Second-Line Treatment Regimens
Regimen | Patients, n (%) |
1L therapy (n = 1003) | 2L therapy (n = 287) |
Chemotherapy | | |
Carboplatin-based doublet or triplet therapya | 499 (49.8) | 57 (19.9) |
Carboplatin | 3 (0.3) | 0 |
Cisplatin-based doublet or triplet therapya | 172 (17.1) | 7 (2.4) |
Docetaxel | 5 (0.5) | 17 (5.9) |
Docetaxel + nintedanib | 4 (0.4) | 16 (5.6) |
Gemcitabine | 3 (0.3) | 1 (0.3) |
Nintedanib | 0 | 1 (0.3) |
Paclitaxel | 0 | 4 (1.4) |
Pemetrexed | 9 (0.9) | 0 |
Vinorelbine | 3 (0.3) | 1 (0.3) |
Immuno-oncology therapy | | |
Atezolizumab | 0 | 32 (11.1) |
Nivolumab | 5 (0.5) | 20 (7.0) |
Pembrolizumab | 174 (17.3) | 96 (33.4) |
Targeted therapyb | | |
Afatinib | 67 (6.7) | 5 (1.7) |
Alectinib | 2 (0.2) | 2 (0.7) |
Ceritinib | 6 (0.6) | 3 (1.0) |
Crizotinib | 11 (1.1) | 4 (1.4) |
Erlotinib | 12 (1.2) | 5 (1.7) |
Gefitinib | 24 (2.4) | 3 (1.0) |
Osimertinib | 4 (0.4) | 13 (4.5) |
1L, first line; 2L, second line; ALK, anaplastic lymphoma kinase; EGFR, epidermal growth factor receptor; ROS, ROS proto-oncogene 1. |
a Triplet therapy indicates carboplatin (IV) + vinorelbine (IV) + vinorelbine (oral), cisplatin (IV) + vinorelbine (IV) + vinorelbine (oral), or cisplatin (IV) + etoposide (IV) + etoposide (oral). |
b EGFR, ALK, or ROS inhibitor. |
In the overall population, 287 patients (28.6%) received 2L therapy, which was chemotherapy in 104 (36.2%), IO monotherapy in 148 (51.6%), and targeted therapy in 35 (12.2%; Table 3). The most common 2L therapy was pembrolizumab (n = 96 [33.4%]). The most common treatment sequence was 1L chemotherapy, followed by 2L IO monotherapy (n = 146 [20.9% of those who received 1L chemotherapy]) or 2L chemotherapy (n = 74 [10.6% of those who received 1L chemotherapy]; Supplemental Fig. 2). Of patients who had received 1L IO monotherapy, chemotherapy was the most common 2L treatment (n = 26 [14.5% of those who received 1L IO monotherapy]). Of patients who received 1L targeted therapy, targeted therapy was also the most commonly used 2L treatment class (n = 26 [20.6% of those who received 1L targeted therapy]).
Of 716 patients who did not receive 2L therapy, 77.0% died (88.7%, 56.3%, and 52.1% of those who had received 1L chemotherapy, IO monotherapy, or targeted therapy, respectively), 12.8% had ongoing 1L treatment (1.9%, 31.8%, and 36.5% of those who received 1L chemotherapy, IO monotherapy, or targeted therapy, respectively), and 10.2% stopped treatment and were still alive at the end of the study period (9.4%, 11.9%, and 11.5% of those who had received 1L chemotherapy, IO monotherapy, or targeted therapy, respectively; Supplemental Table 2).
Of the 287 patients who received 2L treatment, 51 subsequently received 3L treatment (5.1% of the total population or 17.8% of the 2L population [21.3%, excluding 47 patients who continued to receive 2L therapy at last follow-up]). Eleven patients received fourth-line (4L) therapy (1.1% of the total population or 21.6% of the 3L population [28.2%, excluding 12 patients who continued to receive 3L therapy at last follow-up]), and 1 patient received fifth-line therapy (0.1% of the total population or 9.1% of the 4L population [10.0%, excluding 1 patient who continued to receive 4L therapy at last follow-up]; Supplemental Fig. 2).
Clinical Outcomes
In the overall population, the median follow-up was 9.2 months (95% CI, 0-42.7 months), with a longer median follow-up in the IO monotherapy group (12.7 months, 95% CI, 0.1–37.3 months) and targeted therapy group (16.3 months, 95% CI, 0.1–37.1 months), and a shorter median follow-up in the chemotherapy group (7.9 months, 95% CI, 0-42.7 months) (Table 2). The median OS was 9.5 months (95% CI, 8.8–10.7 months; Fig. 2A) in the entire population. Within 1L subgroups defined by drug class, median OS was longest in patients who had received 1L targeted therapy (median 20.2 months [95% CI, 16.0-30.5 months]), followed by patients who had received 1L IO monotherapy (median 14.0 months [95% CI, 10.7–20.6 months]), and was shortest in patients who had received 1L chemotherapy (median 8.1 months [95% CI, 7.4–8.9 months]; Fig. 2B). In the overall population, median TTD from 1L was 2.1 months (95% CI, 2.1–2.3 months; Fig. 2C). Within subgroups, the median TTD was longest with 1L targeted therapy (median, 7.6 months [95% CI, 5.8–11.5 months]) and was 5.3 months (95% CI, 4.2–7.2 months) with 1L IO monotherapy and 2.1 months (95% CI, 1.8–2.1 months) with 1L chemotherapy (Fig. 2D). Median TtNT from 1L was 6.7 months (95% CI, 6.3–7.3 months) in the overall study population (Fig. 2E); in 1L subgroups, it was 13.6 months (95% CI, 10.7–18.8 months) with 1L targeted therapy, 8.9 months (95% CI, 7.5–15.8 months) with 1L IO monotherapy, and 5.9 months (95% CI, 5.3–6.3 months) with 1L chemotherapy (Fig. 2F).
In the overall population, 291 patients (29.0%) had an rwTR. Within 1L subgroups, rwTRs occurred in 187 (26.8%) of those who received 1L chemotherapy, 61 (34.1%) of those who received 1L IO monotherapy, and 43 (34.1%) of those who received 1L targeted therapy.