Abstract
Background
The efficacy of osimertinib in previously EGFR-TKI-treated NSCLC without identification of T790M mutational status remains unclear in real-world practice.
Patients and methods
417 patients had stage III–IV NSCLC harboring EGFR mutation and 154 out of 417 patients receiving osimertinib as ≥ second-line EGFR-TKI were identified. The time to treatment failure and risk of death were analyzed.
Results
Higher risk of death was found in EGFR-mutant patients with age ≥ 65 years, non-adenocarcinoma, no surgery or radiation, non-exon 19 deletion/exon 21 L858R, higher ECOG PS (2–4), PD-L1 expression ≥ 50%, and bone/liver/adrenal metastasis (all p < 0.05). Osimertinib as ≥ second-line TKI in patients with/without identification of T790M revealed lower risk of death compared to first-line first/second generation TKI without subsequent osimertinib (p = 0.0002; 0.0232, respectively). However, osimertinib-treated patients with T790M did not have superior survival than those without (p = 0.2803). A higher risk of treatment failure for osimertinib was found in males, patients with first-line TKI duration ≤ 12 months, BMI drop > 10%, and PD-L1 expression ≥ 50% (All p < 0.05). Nonetheless, osimertinib as ≥ second-line TKI in patients without identification of 790 M did not have higher risk of treatment failure than those with T790M (p = 0.1236).
Conclusions
This study demonstrates that osimertinib as second line or subsequent TKI in EGFR-TKI-treated patients without identification of T790M revealed lower risk of death compared to first-line first/second generation TKI without subsequent osimertinib, in real-world practice. Additionally, EGFR-mutant patients with PD-L1 expression ≥ 50% had a higher risk of treatment failure for osimertinib and worse overall survival than those with PD-L1 expression < 50%. These results suggest that osimertinib as second line or subsequent TKI may be a potential alternative option for the treatment of patients without identification of T790M and PD-L1 expression ≥ 50% is associated with a significantly poor outcome in patients receiving osimertinib.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00432-021-03766-5.
Keywords: Lung cancer, EGFR mutation, T790M, PD-L1, Osimertinib
Introduction
Epidermal growth factor receptor (EGFR) mutations are the most common targetable oncogenes in non-small cell lung cancer (NSCLC), being observed in about 10–15% of patients in Caucasian population (Novello et al. 2016; Barlesi et al. 2016) and up to approximately 50% in Asian population (Barlesi et al. 2016; Mok et al. 2009). EGFR tyrosine kinase inhibitors (TKIs) for EGFR-mutant NSCLC have led to improved survival outcome. Three generations of EGFR-TKIs which have been recommended for first-line treatment of advanced EGFR-mutant NSCLC include the first-generation TKIs, gefitinib/erlotinib; the second-generation TKIs, afatinib/dacomitinib; and the third-generation TKI, osimertinib (Hirsh 2018; Wu et al. 2017). Acquired resistance to EGFR-TKI inevitably occurs regardless of which first-line TKI is used. EGFR T790M mutation in exon 20 which accounts for approximately 50–60% of cases, is the most frequent molecular resistance mechanism to first/ second-generation of TKIs (Ohashi et al. 2013; Cortot and Janne 2014).
Osimertinib is a mutant-selective EGFR inhibitor that targets T790M with low affinity for wild-type EGFR, while remaining active against sensitizing EGFR mutations and, therefore, has less side effects than first/second-generation TKIs (Cross et al. 2014). On the basis of impressive efficacy data and favorable safety profile in advanced EGFR T790M-mutant NSCLC that progressed after first/second-generation TKIs from AURA trials, osimertinib is approved by many countries worldwide in this setting (Yang et al. 2017; Mok et al. 2017). According to recent trials (Soria et al. 2018; Ramalingam et al. 2019; Oxnard et al. 2018), although osimertinib has received US FDA approval as first-line EGFR-TKI therapy, many physicians still use 1st/2nd generation EGFR-TKIs as first-line treatment worldwide (Murtuza et al. 2019). The availability of subsequent therapeutic options after disease progression is one of the major considerations in determining treatment strategy. The efficacy of osimertinib in previously EGFR-TKI treated NSCLC patients without identification of T790M mutational status remains unclear in real-world practice. Therefore, we investigated whether osimertinib can provide a survival benefit in EGFR-mutant patients without identification of T790M mutation after 1st/2nd generation TKI treatment.
Patients and methods
A total of 6568 patients with newly diagnosed lung cancer at Mayo Clinic were identified from January 1, 2009 to December 31, 2016 in a prospectively observed cohort (n = 6424); and January 1, 2017 to March 31, 2019 in a consecutive case series of EGFR-mutant lung cancer patients (n = 144) (Luo et al. 2019) (Fig. 1). The prospectively observed cohort and the consecutive case series were comprised of patients in the Rochester Epidemiology Project database with medical records of all people residing in Olmsted County and patients diagnosed at the Mayo Clinic in Minnesota. Medical records of each new patient were reviewed thoroughly for the eligibility of enrollment and relevant clinical data, including demographic characteristics, cigarette smoking history, and clinical characteristics, was collected. In the prospectively observed cohort, each patient was invited to participate in a long-term follow-up program. Treatment data and vital status was collected through detailed medical records data, registration database, death certificates, next-of-kin reports and obituary documents filed in the patients’ medical records as well as through the Mayo Clinic Tumor Registry and Social Security Death Index website (Xie et al. 2018). The above-mentioned information was supplemented with additional data from follow-up questionnaires, interviews, or both. In our previous reports, detailed procedures of patient recruitment, data acquisition, and regular follow-up have been described previously (Yang et al. 2016; Wang et al. 2016).
Fig. 1.
Flow chart for the study population with patient inclusion and exclusion criteria
Patients were enrolled in this study if the following inclusion criteria were met: pathologically confirmed stage III (ineligible for curative surgery or local radiotherapy) or IV NSCLC harboring EGFR mutation. EGFR mutations include those with exon 18 to exon 21 mutations. Patients had EGFR mutation testing performed in their standard-of-care evaluation through polymerase chain reaction (PCR)-based targeted gene assay, sanger sequencing, or next-generation sequencing panels (Leventakos et al. 2016). Patients were excluded if EGFR wild-typed NSCLC was identified; EGFR mutation testing is not recommended by American Society of Clinical Oncology, College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology (Leighl et al. 2014; Kalemkerian et al. 2018); or patients had first-line osimertinib. Upon disease progression, T790M mutation detection based on tissue biopsy, liquid biopsy for identifying circulating tumor DNA in the blood, or both was performed according to the judgment of clinical physicians. The prescription of osimertinib and the testing of PD-L1 immunohistochemistry staining were based on the decision of the clinicians. The PD-L1 expression was recorded for the percentage of positive tumor cells.
Eventually, 417 patients with stage III or IV EGFR-mutant NSCLC, including 154 receiving osimertinib as ≥ 2nd-line EGFR-TKI treatment (95 with T790M mutation and 59 without identification of T790M), and 263 patients without osimertinib treatment, were included for further analyses. Among these, 35 (59.3%) out of 59 osimertinib treated patients without identification of T790M did not receive rebiopsy, including tissue and liquid biopsies, for EGFR mutation testing. This medical record-based study was approved by the Institutional Review Boards (IRB) of Mayo Foundation and Olmsted County Medical Center.
Demographic and clinical characteristics were analyzed by chi-square test and Wilcoxon rank-sum test for categorical and continuous variables, respectively. The primary outcome was to evaluate OS, defined as the date of diagnosis of lung cancer to death from any reason; those alive or lost to follow-up were defined as censored. The secondary outcome was the time-to-treatment failure (TTF) of osimertinib, defined as time from the start date of osimertinib to the date of treatment discontinuation for any cause, such as disease progression, treatment toxicity, or death.
OS and TTF were evaluated using the Kaplan–Meier (KM) method and log-rank test. Univariate Cox proportional hazard (Cox) models were utilized for assessing the association of prognostic factors with estimated five-year survival. Multivariate Cox models were performed using the significant variables (p < 0.10) in the univariate analysis, and hazard ratios (HR) with 95% confidence intervals (CI) were calculated as reported in the literature (Luo et al. 2019). All statistical tests were done as 2-tailed and P levels of < 0.05 were considered statistically significant. Analyses were performed using SAS, v.9.4 (SAS Institute Inc., Cary, NC, USA).
Results
Among 417 study patients with stage III-IV NSCLC harboring EGFR mutation, there was a median follow-up of 4.1 (IQR 2.4–5) years; 154 out of 417 patients received osimertinib as ≥ 2nd-line TKI. The characteristics and comparisons of the 417 EGFR-mutant patients, stratified by osimertinib treatment are shown in Table 1. A higher proportion of patients receiving osimertinib had PD-L1 expression < 50%, exon 19 deletion, exon 21 L858R point mutation, EGFR T790M mutation, 2nd/3rd biopsy, chemotherapy, received 2 to 4 types of EGFR-TKIs, and afatinib as 1st-line EGFR-TKI, compared with those without osimertinib treatment. The characteristics of 182 patients undergoing 2nd or 3rd biopsies are shown in the Supplementary Table 1. The mutation status of second biopsy in patients without EGFR T790M mutation receiving osimertinib is shown in the Supplementary Table 2. A higher proportion of patients with acquired T790M had tumor in left lung, tumor biopsy, exon 19 deletion, no surgery, and 1st-line TKI duration ≥ 6 months. After adjusting for significant variables (defined as p < 0.05) in univariable analysis, the multivariable analysis revealed that the odds ratio for T790M from blood specimens by liquid biopsy compared to tumor biopsies was 0.301 (95% CI 0.144–0.63, p = 0.001; Table 2). The odds ratio for T790M in patients with tumor of left lung was 2.429 (95% CI 1.251–4.715, p = 0.009) compared to those without.
Table 1 .
Patient summary by osimertinib treatment
| 1st/2nd generation TKI with subsequent osimertinib (N = 154) | 1st/2nd generation TKI without subsequent osimertinib (N = 203) | No EGFR-TKI treatment (N = 60) | Total (N = 417) | p value | |
|---|---|---|---|---|---|
| Age | 0.0635 | ||||
| N | 154 | 203 | 60 | 417 | |
| Mean (SD) | 63.6 (11.5) | 64.6 (12.7) | 67.7 (11.1) | 64.7 (12.1) | |
| Median | 63.5 | 65.1 | 69.7 | 65.1 | |
| Q1, Q3 | 56.4, 72.7 | 57.4, 74.5 | 59.5, 75.3 | 57.4, 73.9 | |
| Range | (31.6–88.5) | (31.3–88.5) | (37.1–85.5) | (31.3–88.5) | |
| Gender | 0.2643 | ||||
| Female | 109 (70.8%) | 127 (62.6%) | 39 (65.0%) | 275 (65.9%) | |
| Male | 45 (29.2%) | 76 (37.4%) | 21 (35.0%) | 142 (34.1%) | |
| Race | 0.1922 | ||||
| White | 126 (81.8%) | 173 (85.2%) | 55 (91.7%) | 354 (84.9%) | |
| Non-White | 28 (18.2%) | 30 (14.8%) | 5 (8.3%) | 63 (15.1%) | |
| Cigarette smoking status | 0.5267 | ||||
| Never | 93 (60.4%) | 120 (59.1%) | 33 (55.0%) | 246 (59.0%) | |
| Former | 54 (35.1%) | 65 (32.0%) | 23 (38.3%) | 142 (34.1%) | |
| Current | 7 (4.5%) | 18 (8.9%) | 4 (6.7%) | 29 (7.0%) | |
| ECOG performance status | 0.3222 | ||||
| 0 | 68 (44.2%) | 71 (35.0%) | 24 (40.0%) | 163 (39.1%) | |
| 1 | 70 (45.5%) | 103 (50.7%) | 24 (40.0%) | 197 (47.2%) | |
| 2 | 11 (7.1%) | 20 (9.9%) | 6 (10.0%) | 37 (8.9%) | |
| 3 | 5 (3.2%) | 8 (3.9%) | 5 (8.3%) | 18 (4.3%) | |
| 4 | 0 (0.0%) | 1 (0.5%) | 1 (1.7%) | 2 (0.5%) | |
| Stage | 0.0950 | ||||
| III | 9 (5.8%) | 17 (8.4%) | 9 (15.0%) | 35 (8.4%) | |
| IV | 145 (94.2%) | 186 (91.6%) | 51 (85.0%) | 382 (91.6%) | |
| Cell type | 0.0764 | ||||
| Adenocarcinoma | 150 (97.4%) | 188 (92.6%) | 60 (100.0%) | 398 (95.4%) | |
| Squamous | 1 (0.6%) | 7 (3.4%) | 0 (0.0%) | 8 (1.9%) | |
| Other NSCLC | 3 (1.9%) | 8 (3.9%) | 0 (0.0%) | 11 (2.6%) | |
| Tumor in left lung | 0.6992 | ||||
| No | 72 (46.8%) | 102 (50.2%) | 27 (45.0%) | 201 (48.2%) | |
| Yes | 82 (53.2%) | 101 (49.8%) | 33 (55.0%) | 216 (51.8%) | |
| Tumor in right lung | 0.6217 | ||||
| No | 66 (42.9%) | 77 (37.9%) | 23 (38.3%) | 166 (39.8%) | |
| Yes | 88 (57.1%) | 126 (62.1%) | 37 (61.7%) | 251 (60.2%) | |
| Other cancer | 0.3060 | ||||
| No | 119 (77.3%) | 166 (81.8%) | 44 (73.3%) | 329 (78.9%) | |
| Yes | 35 (22.7%) | 37 (18.2%) | 16 (26.7%) | 88 (21.1%) | |
| Other pulmonary disease | 0.5988 | ||||
| No | 134 (87.0%) | 172 (84.7%) | 49 (81.7%) | 355 (85.1%) | |
| Yes | 20 (13.0%) | 31 (15.3%) | 11 (18.3%) | 62 (14.9%) | |
| Any other disease | 0.9890 | ||||
| No | 29 (18.8%) | 37 (18.2%) | 11 (18.3%) | 77 (18.5%) | |
| Yes | 125 (81.2%) | 166 (81.8%) | 49 (81.7%) | 340 (81.5%) | |
| Family history of lung cancer | 0.1159 | ||||
| No | 154 (100.0%) | 200 (98.5%) | 58 (96.7%) | 412 (98.8%) | |
| Yes | 0 (0.0%) | 3 (1.5%) | 2 (3.3%) | 5 (1.2%) | |
| PD-L1 expression | 0.0001 | ||||
| < 50% | 61 (39.6%) | 50 (24.6%) | 8 (13.3%) | 119 (28.5%) | |
| ≥ 50% | 10 (6.5%) | 5 (2.5%) | 1 (1.7%) | 16 (3.8%) | |
| Not tested | 83 (53.9%) | 148 (72.9%) | 51 (85.0%) | 282 (67.6%) | |
| 1st EGFR mutation subtype | 0.0007 | ||||
| Exon 19 deletion | 83 (53.9%) | 96 (47.3%) | 26 (43.3%) | 205 (49.2%) | |
| Exon 21 L858R | 60 (39.0%) | 68 (33.5%) | 16 (26.7%) | 144 (34.5%) | |
| Others | 11 (7.1%) | 39 (19.2%) | 18 (30.0%) | 68 (16.3%) | |
| T790M | < 0.0001 | ||||
| No | 59 (38.3%) | 184 (90.6%) | 59 (98.3%) | 302 (72.4%) | |
| Yes | 95 (61.7%) | 19 (9.4%) | 1 (1.7%) | 115 (27.6%) | |
| Number of biopsies | < 0.0001 | ||||
| 1 | 39 (25.3%) | 136 (67.0%) | 59 (98.3%) | 234 (56.1%) | |
| 2 | 91 (59.1%) | 55 (27.1%) | 1 (1.7%) | 147 (35.3%) | |
| 3 | 24 (15.6%) | 12 (5.9%) | 0 (0.0%) | 36 (8.6%) | |
| Surgery | 0.4665 | ||||
| No | 140 (90.9%) | 189 (93.1%) | 53 (88.3%) | 382 (91.6%) | |
| Yes | 14 (9.1%) | 14 (6.9%) | 7 (11.7%) | 35 (8.4%) | |
| Radiation | 0.1422 | ||||
| No | 144 (93.5%) | 178 (87.7%) | 52 (86.7%) | 374 (89.7%) | |
| Yes | 10 (6.5%) | 25 (12.3%) | 8 (13.3%) | 43 (10.3%) | |
| Chemotherapy | < 0.0001 | ||||
| No | 6 (3.9%) | 16 (7.9%) | 27 (45.0%) | 49 (11.8%) | |
| Yes | 148 (96.1%) | 187 (92.1%) | 33 (55.0%) | 368 (88.2%) | |
| Numbers of used EGFR-TKI | < 0.0001* | ||||
| 0 | 0 (0.0%) | 0 (0.0%) | 60 (100.0%) | 60 (14.4%) | |
| 1 | 0 (0.0%) | 159 (78.3%) | 0 (0.0%) | 159 (38.1%) | |
| 2 | 122 (79.2%) | 43 (21.2%) | 0 (0.0%) | 165 (39.6%) | |
| 3 | 27 (17.5%) | 1 (0.5%) | 0 (0.0%) | 28 (6.7%) | |
| 4 | 5 (3.2%) | 0 (0.0%) | 0 (0.0%) | 5 (1.2%) | |
| 1st-line EGFR-TKI | < 0.0001* | ||||
| No EGFR-TKI treatment | 0 (0.0%) | 0 (0.0%) | 60 (100.0%) | 60 (14.4%) | |
| Erlotinib | 124 (80.5%) | 177 (87.2%) | 0 (0.0%) | 301 (72.2%) | |
| Gefitinib | 2 (1.3%) | 5 (2.5%) | 0 (0.0%) | 7 (1.7%) | |
| Afatinib | 28 (18.2%) | 21 (10.3%) | 0 (0.0%) | 49 (11.8%) | |
| Lung metastasis | 0.0970 | ||||
| No | 90 (58.4%) | 137 (67.5%) | 33 (55.0%) | 260 (62.4%) | |
| Yes | 64 (41.6%) | 66 (32.5%) | 27 (45.0%) | 157 (37.6%) | |
| Pleural metastasis | 0.3299 | ||||
| No | 123 (79.9%) | 155 (76.4%) | 51 (85.0%) | 329 (78.9%) | |
| Yes | 31 (20.1%) | 48 (23.6%) | 9 (15.0%) | 88 (21.1%) | |
| CNS metastasis | 0.3990 | ||||
| No | 130 (84.4%) | 160 (78.8%) | 48 (80.0%) | 338 (81.1%) | |
| Yes | 24 (15.6%) | 43 (21.2%) | 12 (20.0%) | 79 (18.9%) | |
| Bone metastasis | 0.2351 | ||||
| No | 78 (50.6%) | 113 (55.7%) | 38 (63.3%) | 229 (54.9%) | |
| Yes | 76 (49.4%) | 90 (44.3%) | 22 (36.7%) | 188 (45.1%) | |
| Liver metastasis | 0.8540 | ||||
| No | 138 (89.6%) | 181 (89.2%) | 55 (91.7%) | 374 (89.7%) | |
| Yes | 16 (10.4%) | 22 (10.8%) | 5 (8.3%) | 43 (10.3%) | |
| Peritoneum/abdomen LN metastasis | 0.2464 | ||||
| No | 141 (91.6%) | 193 (95.1%) | 58 (96.7%) | 392 (94.0%) | |
| Yes | 13 (8.4%) | 10 (4.9%) | 2 (3.3%) | 25 (6.0%) | |
| Adrenal metastasis | 0.8521 | ||||
| No | 142 (92.2%) | 187 (92.1%) | 54 (90.0%) | 383 (91.8%) | |
| Yes | 12 (7.8%) | 16 (7.9%) | 6 (10.0%) | 34 (8.2%) | |
| Distant metastasis number | 0.4712 | ||||
| Distant metastatic site = 0–1 | 88 (57.1%) | 108 (53.2%) | 37 (61.7%) | 233 (55.9%) | |
| Distant metastatic site > 1 | 66 (42.9%) | 95 (46.8%) | 23 (38.3%) | 184 (44.1%) | |
CNS central nervous system, ECOG Eastern Cooperative Oncology Group, EGFR epidermal growth factor receptor, LN lymph node, NSCLC non-small cell lung cancer, PD-L1 programmed cell death ligand 1, SD standard deviation, TKI tyrosine kinase inhibitor
*Fisher’s Exact Test
Table 2.
Multivariable analysis for prediction of T790M
| Variables (No. of patients) | Multivariable logistic regression | P value | |
|---|---|---|---|
| Odds ratio | 95% CI | ||
| Tumor in left lung (Yes/no = 94/88) | 2.429 | 1.251–4.715 | 0.009 |
| 1st EGFR mutation subtype | |||
| Others (14) | 1 (reference) | ||
| Exon 19 deletion (97) | 3.126 | 0.886–11.026 | 0.076 |
| L858R (71) | 2.006 | 0.560–7.191 | 0.285 |
| Liquid biopsy (blood specimens/tumor = 46/136) | 0.301 | 0.144–0.630 | 0.001 |
| Surgery (Yes/no = 15/167) | 0.256 | 0.077–0.851 | 0.026 |
| 1st TKI treatment > 6 month (Yes/no = 156/26) | 1.558 | 0.615–3.950 | 0.350 |
CI confidence interval, EGFR epidermal growth factor receptor, TKI tyrosine kinase inhibitor
Significant variables in univariate Cox models of TTF for osimertinib were analyzed in multivariate Cox models (Table 3), demonstrating that higher risk of treatment failure for osimertinib was found in male patients (HR 1.72; 95% CI 1.06–2.8; p = 0.0327), with 1st-line TKI duration ≤ 12 months (HR 2.16; 95% CI 1.33–3.5; p = 0.0019), BMI drop > 10% (HR 1.85; 95% CI 1.11–3.08; p = 0.0207), PD-L1 expression ≥ 50% (HR 4.28; 95% CI 1.83–10.02; p = 0.0008, compared to PD-L1 expression < 50%), and 1st-line TKI with afatinib (HR 2.19; 95% CI 1.2–3.98; p = 0.0136). However, osimertinib as ≥ 2nd-line TKI in patients without identification of 790 M did not have higher risk of treatment failure than those with T790M (p = 0.1236).
Table 3.
Time to treatment failure (TTF) for osimertinib (n = 154)
| Variable | N | Events | Median Months |
2-year survival % (95% CI) |
Cox univariate hazard ratio (95% CI) |
Cox univariate score p value |
Cox multivariate hazard Ratio (95% CI) |
Cox multivariate likelihood ratio p value (n = 154) |
|---|---|---|---|---|---|---|---|---|
| Age under 64 (median age) | 0.4976 | |||||||
| No | 77 | 41 (53%) | 13.8 | 36.2% (23.2%, 49.1%) | – | |||
| Yes | 77 | 43 (56%) | 18.2 | 41.9% (29.0%, 54.8%) | 0.86 (0.56, 1.32) | |||
| Gender | 0.0176 | 0.0327 | ||||||
| Female | 109 | 52 (48%) | 16.9 | 42.6% (31.1%, 54.2%) | – | – | ||
| Male | 45 | 32 (71%) | 10.2 | 30.4% (15.6%, 45.1%) | 1.70 (1.09, 2.64) | 1.72 (1.06, 2.80) | ||
| Race | 0.8212 | |||||||
| White | 126 | 70 (56%) | 14.1 | 39.5% (29.4%, 49.6%) | – | |||
| Non-white | 28 | 14 (50%) | 16.9 | 36.7% (14.2%, 59.1%) | 0.94 (0.53, 1.66) | |||
| Cigarette smoking status | 0.2781 | |||||||
| Never | 93 | 49 (53%) | 16.0 | 43.5% (31.6%, 55.4%) | – | |||
| Ever | 61 | 35 (57%) | 12.4 | 31.5% (17.1%, 45.8%) | 1.27 (0.82, 1.96) | |||
| Stage | 0.1484 | |||||||
| III | 9 | 2 (22%) | NR | 77.8% (50.6%, 100.0%) | 0.37 (0.09, 1.51) | |||
| IV | 145 | 82 (57%) | 14.1 | 36.9% (27.6%, 46.3%) | – | |||
| Cell type | 0.4601 | |||||||
| Adenocarcinoma | 150 | 82 (55%) | 14.6 | 39.1% (29.8%, 48.4%) | – | |||
| Non-adenocarcinoma | 4 | 2 (50%) | 19.7 | 0.0% (0.0%, 0.0%) | 0.59 (0.15, 2.42) | |||
| Surgery | 0.4342 | |||||||
| No | 140 | 79 (56%) | 14.1 | 37.8% (28.2%, 47.3%) | – | |||
| Yes | 14 | 5 (36%) | 52.2% (20.3%, 84.1%) | 0.70 (0.28, 1.73) | ||||
| Radiation | 0.8283 | |||||||
| No | 144 | 81 (56%) | 14.6 | 39.3% (30.0%, 48.7%) | – | |||
| Yes | 10 | 3 (30%) | 19.7 | 0.0% (0.0%, 0.0%) | 0.88 (0.28, 2.80) | |||
| T790M | 0.1236 | |||||||
| No | 59 | 29 (49%) | 11.8 | 21.2% (0.0%, 42.5%) | 1.44 (0.90, 2.30) | |||
| Yes | 95 | 55 (58%) | 16.6 | 43.7% (32.8%, 54.7%) | – | |||
| 1st EGFR mutation subtype | 0.5372 | |||||||
| Exon 19 deletion | 83 | 43 (52%) | 18.8 | 45.3% (32.6%, 58.1%) | – | |||
| Exon 21 L858R | 60 | 34 (57%) | 11.8 | 28.5% (13.9%, 43.0%) | 1.28 (0.82, 2.02) | |||
| Others | 11 | 7 (64%) | 11.7 | 40.0% (9.6%, 70.4%) | 1.23 (0.55, 2.74) | |||
| 1st TKI > 6 months | 0.5960 | |||||||
| No | 21 | 12 (57%) | 7.6 | 41.9% (20.4%, 63.4%) | 1.18 (0.64, 2.18) | |||
| Yes | 133 | 72 (54%) | 16.0 | 38.9% (28.9%, 49.0%) | – | |||
| 1st TKI > 12 months | < 0.0001 | 0.0019 | ||||||
| No | 65 | 45 (69%) | 6.8 | 19.0% (6.9%, 31.0%) | 2.60 (1.68, 4.02) | 2.16 (1.33, 3.50) | ||
| Yes | 89 | 39 (44%) | 24.4 | 53.7% (41.2%, 66.3%) | – | – | ||
| BMI drop > 10% | 0.0101 | 0.0207 | ||||||
| No | 118 | 57 (48%) | 17.9 | 43.5% (32.5%, 54.5%) | – | – | ||
| Yes | 36 | 27 (75%) | 8.2 | 24.5% (8.9%, 40.2%) | 1.81 (1.14, 2.86) | 1.85 (1.11, 3.08) | ||
| BMI increase > 10% | 0.3199 | |||||||
| No | 141 | 77 (55%) | 16.0 | 40.0% (30.5%, 49.6%) | – | |||
| Yes | 13 | 7 (54%) | 9.9 | 0.0% (0.0%, 0.0%) | 1.48 (0.68, 3.25) | |||
| Tumor in left lung | 0.9783 | |||||||
| No | 72 | 39 (54%) | 15.4 | 41.0% (27.3%, 54.7%) | 1.01 (0.65, 1.55) | |||
| Yes | 82 | 45 (55%) | 14.0 | 37.0% (24.5%, 49.5%) | – | |||
| Tumor in right lung | 0.6841 | |||||||
| No | 66 | 33 (50%) | 16.0 | 35.8% (20.8%, 50.7%) | 0.91 (0.59, 1.42) | |||
| Yes | 88 | 51 (58%) | 14.6 | 40.7% (28.9%, 52.5%) | – | |||
| Other cancer | 0.0755 | 0.4506 | ||||||
| No | 119 | 62 (52%) | 18.2 | 43.9% (33.3%, 54.5%) | – | – | ||
| Yes | 35 | 22 (63%) | 11.0 | 20.8% (3.4%, 38.3%) | 1.56 (0.95, 2.54) | 1.25 (0.70, 2.24) | ||
| Other pulmonary disease | 0.8054 | |||||||
| No | 134 | 72 (54%) | 14.6 | 38.2% (28.3%, 48.1%) | – | |||
| Yes | 20 | 12 (60%) | 17.9 | 41.9% (16.3%, 67.5%) | 0.93 (0.50, 1.71) | |||
| Any other disease | 0.3836 | |||||||
| No | 29 | 16 (55%) | 16.9 | 36.5% (15.9%, 57.0%) | 0.79 (0.46, 1.35) | |||
| Yes | 125 | 68 (54%) | 14.0 | 39.7% (29.4%, 49.9%) | – | |||
| ECOG performance status | 0.3443 | |||||||
| 0–1 | 138 | 75 (54%) | 16.0 | 40.2% (30.6%, 49.9%) | – | |||
| 2–4 | 16 | 9 (56%) | 9.9 | 0.0% (0.0%, 0.0%) | 1.40 (0.70, 2.82) | |||
| Lung metastasis | 0.1997 | |||||||
| No | 90 | 44 (49%) | 18.2 | 42.7% (30.2%, 55.3%) | – | |||
| Yes | 64 | 40 (63%) | 14.1 | 34.7% (21.4%, 48.0%) | 1.32 (0.86, 2.03) | |||
| Pleural metastasis | 0.8133 | |||||||
| No | 123 | 66 (54%) | 15.4 | 37.1% (26.7%, 47.6%) | – | |||
| Yes | 31 | 18 (58%) | 19.2 | 45.0% (25.0%, 65.0%) | 0.94 (0.56, 1.58) | |||
| CNS metastasis | 0.3727 | |||||||
| No | 130 | 67 (52%) | 16.1 | 40.2% (29.8%, 50.6%) | – | |||
| Yes | 24 | 17 (71%) | 12.5 | 32.7% (13.0%, 52.4%) | 1.27 (0.75, 2.17) | |||
| Bone metastasis | 0.0811 | 0.7694 | ||||||
| No | 78 | 38 (49%) | 19.2 | 48.7% (35.5%, 61.8%) | – | – | ||
| Yes | 76 | 46 (61%) | 10.9 | 29.1% (16.7%, 41.5%) | 1.46 (0.95, 2.25) | 1.08 (0.63, 1.85) | ||
| Liver metastasis | 0.0124 | 0.4444 | ||||||
| No | 138 | 72 (52%) | 16.6 | 41.7% (31.8%, 51.7%) | – | – | ||
| Yes | 16 | 12 (75%) | 6.6 | 16.0% (0.0%, 36.0%) | 2.16 (1.16, 4.00) | 1.34 (0.64, 2.84) | ||
| Peritoneum/abdomen LN metastasis | 0.7897 | |||||||
| No | 141 | 77 (55%) | 15.4 | 38.4% (28.7%, 48.1%) | – | |||
| Yes | 13 | 7 (54%) | 5.9 | 41.7% (13.8%, 69.6%) | 1.11 (0.51, 2.41) | |||
| Adrenal metastasis | 0.2315 | |||||||
| No | 142 | 77 (54%) | 16.1 | 40.1% (30.5%, 49.8%) | – | |||
| Yes | 12 | 7 (58%) | 9.8 | 22.2% (0.0%, 49.4%) | 1.60 (0.74, 3.48) | |||
| Distant metastasis number | 0.0578 | 0.4356 | ||||||
| Distant metastatic site = 0–1 | 88 | 43 (49%) | 18.2 | 41.6% (28.8%, 54.5%) | – | – | ||
| Distant metastatic site > 1 | 66 | 41 (62%) | 11.8 | 34.6% (21.5%, 47.7%) | 1.51 (0.98, 2.32) | 1.25 (0.71, 2.19) | ||
| PD-L1 expression | < 0.0001 | 0.0019 | ||||||
| < 50% | 61 | 34 (56%) | 11.8 | 29.3% (13.7%, 44.8%) | – | – | ||
| ≥ 50% | 10 | 8 (80%) | 3.3 | 0.0% (0.0%, 0.0%) | 3.12 (1.43, 6.81) | 4.28 (1.83, 10.02) | ||
| Not tested | 83 | 42 (51%) | 19.2 | 47.2% (34.9%, 59.5%) | 0.64 (0.41, 1.02) | 0.77 (0.48, 1.25) | ||
| Numbers of used EGFR-TKI before osimertinib | 0.5775 | |||||||
| 0–1 | 125 | 64 (51%) | 15.4 | 40.2% (29.8%, 50.7%) | – | |||
| 2–3 | 29 | 20 (69%) | 14.0 | 35.0% (16.2%, 53.7%) | 1.15 (0.70, 1.91) | |||
| 1st-line EGFR-TKI | < 0.0001 | 0.0136 | ||||||
| Erlotinib/Gefitinib | 126 | 64 (51%) | 18.2 | 44.2% (33.9%, 54.5%) | – | – | ||
| Afatinib | 28 | 20 (71%) | 6.3 | 0.0% (0.0%, 0.0%) | 3.16 (1.86, 5.40) | 2.19 (1.20, 3.98) | ||
| 1st-line EGFR-TKI | 0.1236 | |||||||
| Osimertinib as ≥ 2nd-line TKI without T790M | 59 | 29 (49%) | 11.8 | 21.2% (0.0%, 42.5%) | – | |||
| Osimertinib as ≥ 2nd-line TKI with T790M | 95 | 55 (58%) | 16.6 | 43.7% (32.8%, 54.7%) | 0.69 (0.43, 1.11) | |||
BMI body mass index, CI confidence interval, CNS central nervous system, ECOG Eastern Cooperative Oncology Group, EGFR epidermal growth factor receptor, LN lymph node, NR not reached, PD-L1 programmed cell death ligand 1, TKI tyrosine kinase inhibitor
Significant variables in univariate Cox models of OS for EGFR-mutant patients were analyzed in multivariate models (Table 4), demonstrating that higher risk of death was found in patients with age ≥ 65 years (HR 1.44; 95% CI 1.09–1.90; p = 0.0095), non-adenocarcinoma (HR 1.87; 95% CI 1.05–3.33; p = 0.048), no surgery (HR 2.1; 95% CI 1.14–3.84; p = 0.0086), no radiation (HR 2.36; 95% CI 1.36–4.12; p = 0.0008), non-exon 19 deletion/exon 21 L858R mutation (HR 1.46; 95% CI 1.03–2.06; p = 0.0375), higher ECOG PS (2–4) (HR 2.07; 95% CI 1.45–2.95; p = 0.0002), PD-L1 expression ≥ 50% compared to PD-L1 expression < 50%) (HR 2.69; 95% CI 1.32–5.52; p = 0.0109), bone metastasis (HR 1.53; 95% CI 1.12–2.1; p = 0.0076), liver metastasis (HR 1.65; 95% CI 1.07–2.55; p = 0.0292), and adrenal metastasis (HR 1.72; 95% CI 1.1–2.68; p = 0.0222). Furthermore, osimertinib as ≥ 2nd-line TKI treatment in patients with or without identification of T790M revealed lower risk of death compared to 1st/2nd generation TKI treatment without subsequent osimertinib in multivariate analysis (HR 0.33 [95% CI 0.16–0.69]; 0.46 [0.22–0.96], and p = 0.0002; 0.0232, respectively). Nonetheless, among patients receiving osimertinib as ≥ 2nd-line TKI treatment, patients with T790M did not have superior survival than those without (p = 0.2803) (Table 4).
Table 4.
Overall survival of all EGFR-mutant patients (n = 417)
| Variable | N | Events | Median Months |
5-year survival % (95% CI) |
Cox univariate hazard ratio (95% CI) |
Cox univariate score p value |
Cox multivariate hazard ratio (95% CI) |
Cox multivariate likelihood ratio p value (n = 417) |
|---|---|---|---|---|---|---|---|---|
| Age under 65 (median age) | 0.0208 | 0.0095 | ||||||
| No | 209 | 119 (57%) | 29.3 | 29.0% (21.1%, 37.0%) | 1.36 (1.05, 1.76) | 1.44 (1.09, 1.90) | ||
| Yes | 208 | 109 (52%) | 39.2 | 32.5% (24.5%, 40.6%) | – | – | ||
| Gender | 0.0145 | 0.4426 | ||||||
| Female | 275 | 139 (51%) | 36.5 | 35.8% (28.6%, 42.9%) | – | – | ||
| Male | 142 | 89 (63%) | 30.0 | 21.1% (12.4%, 29.8%) | 1.39 (1.07, 1.82) | 1.12 (0.84, 1.47) | ||
| Race | 0.9011 | |||||||
| White | 354 | 193 (55%) | 34.3 | 31.9% (25.7%, 38.1%) | – | |||
| Non-white | 63 | 35 (56%) | 35.0 | 23.2% (9.0%, 37.3%) | 1.02 (0.71, 1.47) | |||
| Cigarette smoking status | 0.5398 | |||||||
| Never | 246 | 129 (52%) | 34.3 | 32.6% (25.0%, 40.1%) | – | |||
| Ever | 171 | 99 (58%) | 33.1 | 28.4% (20.0%, 36.9%) | 1.09 (0.84, 1.41) | |||
| Stage | 0.0593 | 0.7187 | ||||||
| III | 35 | 14 (40%) | NR | 52.3% (33.6%, 71.1%) | 0.60 (0.35, 1.03) | – | ||
| IV | 382 | 214 (56%) | 32.9 | 28.4% (22.5%, 34.3%) | – | 1.11 (0.61, 2.03) | ||
| Cell type | 0.0545 | 0.0480 | ||||||
| Adenocarcinoma | 398 | 214 (54%) | 34.4 | 31.3% (25.4%, 37.1%) | – | – | ||
| Non-adenocarcinoma | 19 | 14 (74%) | 25.1 | 18.4% (0.0%, 38.8%) | 1.69 (0.98, 2.90) | 1.87 (1.05, 3.33) | ||
| Surgery | 0.0092 | 0.0086 | ||||||
| No | 382 | 216 (57%) | 32.5 | 28.1% (22.3%, 33.9%) | 2.13 (1.19, 3.81) | 2.10 (1.14, 3.84) | ||
| Yes | 35 | 12 (34%) | NR | 57.6% (37.4%, 77.9%) | – | – | ||
| Radiation | 0.0040 | 0.0008 | ||||||
| No | 374 | 213 (57%) | 32.4 | 27.9% (22.1%, 33.7%) | 2.12 (1.25, 3.58) | 2.36 (1.36, 4.12) | ||
| Yes | 43 | 15 (35%) | NR | 56.3% (38.7%, 73.9%) | – | – | ||
| Chemotherapy | 0.6574 | |||||||
| No | 49 | 22 (45%) | 29.1 | 45.7% (29.4%, 61.9%) | – | |||
| Yes | 368 | 206 (56%) | 34.3 | 29.4% (23.5%, 35.4%) | 0.91 (0.58, 1.41) | |||
| T790M | 0.0015 | 0.5189 | ||||||
| No | 302 | 168 (56%) | 30.8 | 27.8% (21.0%, 34.6%) | – | – | ||
| Yes | 115 | 60 (52%) | 44.1 | 37.5% (27.0%, 48.0%) | 0.62 (0.46, 0.84) | 1.23 (0.66, 2.30) | ||
| 1st EGFR mutation subtype | 0.0014 | 0.0375 | ||||||
| Exon 19 deletion/exon 21 L858R | 349 | 183 (52%) | 36.7 | 32.1% (25.7%, 38.5%) | – | – | ||
| Others | 68 | 45 (66%) | 24.4 | 22.2% (10.2%, 34.2%) | 1.69 (1.22, 2.34) | 1.46 (1.03, 2.06) | ||
| Tumor in left lung | 0.2524 | |||||||
| No | 201 | 99 (49%) | 35.2 | 35.8% (27.1%, 44.6%) | 0.86 (0.66, 1.12) | |||
| Yes | 216 | 129 (60%) | 32.9 | 27.3% (20.1%, 34.5%) | – | |||
| Tumor in right lung | 0.4122 | |||||||
| No | 166 | 94 (57%) | 32.4 | 29.2% (20.6%, 37.9%) | 1.12 (0.86, 1.45) | |||
| Yes | 251 | 134 (53%) | 34.7 | 31.7% (24.2%, 39.1%) | – | |||
| Other cancer | 0.2188 | |||||||
| No | 329 | 178 (54%) | 35.9 | 31.7% (25.3%, 38.1%) | – | |||
| Yes | 88 | 50 (57%) | 29.8 | 26.1% (13.5%, 38.7%) | 1.22 (0.89, 1.67) | |||
| Other pulmonary disease | 0.3884 | |||||||
| No | 355 | 196 (55%) | 33.6 | 29.7% (23.6%, 35.8%) | – | |||
| Yes | 62 | 32 (52%) | 35.9 | 36.5% (21.7%, 51.2%) | 0.85 (0.58, 1.23) | |||
| Any other disease | 0.3512 | |||||||
| No | 77 | 44 (57%) | 31.8 | 23.7% (11.4%, 36.1%) | 1.17 (0.84, 1.62) | |||
| Yes | 340 | 184 (54%) | 34.7 | 32.0% (25.7%, 38.3%) | – | |||
| Family history of lung cancer | 0.0481 | 0.9600 | ||||||
| No | 412 | 224 (54%) | 34.4 | 30.9% (25.2%, 36.7%) | – | – | ||
| Yes | 5 | 4 (80%) | 21.1 | 0.0% (0.0%, 0.0%) | 2.62 (0.97, 7.06) | 0.97 (0.35, 2.73) | ||
| ECOG performance status | < 0.0001 | 0.0002 | ||||||
| 0–1 | 360 | 185 (51%) | 37.7 | 33.3% (27.0%, 39.6%) | – | – | ||
| 2–4 | 57 | 43 (75%) | 19.7 | 14.1% (3.3%, 25.0%) | 2.25 (1.61, 3.14) | 2.07 (1.45, 2.95) | ||
| PD-L1 expression | 0.0488 | 0.0469 | ||||||
| < 50% | 119 | 46 (39%) | 43.0 | 31.6% (17.8%, 45.5%) | – | – | ||
| ≥ 50% | 16 | 10 (63%) | 22.8 | 11.3% (0.0%, 31.8%) | 1.983 (1.00, 3.93) | 2.69 (1.32, 5.52) | ||
| Not tested | 282 | 172 (61%) | 32.0 | 30.2% (23.9%, 36.5%) | 1.41 (1.02, 1.96) | 1.08 (0.77, 1.53) | ||
| Lung metastasis | 0.3188 | |||||||
| No | 260 | 142 (55%) | 31.7 | 32.1% (25.0%, 39.3%) | – | |||
| Yes | 157 | 86 (55%) | 38.5 | 28.1% (18.9%, 37.2%) | 0.87 (0.67, 1.14) | |||
| Pleural metastasis | 0.4407 | |||||||
| No | 329 | 176 (53%) | 35.0 | 31.3% (24.8%, 37.8%) | – | |||
| Yes | 88 | 52 (59%) | 29.4 | 28.2% (16.9%, 39.5%) | 1.13 (0.83, 1.54) | |||
| CNS metastasis | 0.0021 | 0.1258 | ||||||
| No | 338 | 167 (49%) | 37.8 | 34.4% (27.7%, 41.1%) | – | – | ||
| Yes | 79 | 61 (77%) | 26.3 | 18.5% (9.0%, 28.0%) | 1.58 (1.18, 2.12) | 1.34 (0.92, 1.93) | ||
| Bone metastasis | 0.0010 | 0.0076 | ||||||
| No | 229 | 114 (50%) | 39.9 | 35.7% (28.0%, 43.5%) | – | – | ||
| Yes | 188 | 114 (61%) | 28.0 | 23.8% (15.3%, 32.3%) | 1.55 (1.19, 2.01) | 1.53 (1.12, 2.10) | ||
| Liver metastasis | 0.0001 | 0.0292 | ||||||
| No | 374 | 195 (52%) | 36.5 | 33.3% (27.1%, 39.4%) | – | – | ||
| Yes | 43 | 33 (77%) | 20.5 | 9.2% (0.0%, 19.1%) | 2.05 (1.41, 2.96) | 1.65 (1.07, 2.55) | ||
| Peritoneum/abdomen LN metastasis | 0.3727 | |||||||
| No | 392 | 211 (54%) | 34.3 | 31.5% (25.7%, 37.4%) | – | |||
| Yes | 25 | 17 (68%) | 27.5 | 18.9% (0.5%, 37.4%) | 1.25 (0.76, 2.05) | |||
| Adrenal metastasis | 0.0032 | 0.0222 | ||||||
| No | 383 | 201 (52%) | 35.2 | 32.8% (26.8%, 38.8%) | – | – | ||
| Yes | 34 | 27 (79%) | 21.6 | 7.1% (0.0%, 19.5%) | 1.82 (1.21, 2.72) | 1.72 (1.10, 2.68) | ||
| Distant metastasis number | 0.0002 | 0.5086 | ||||||
| Distant metastatic site = 0–1 | 233 | 108 (46%) | 41.1 | 38.2% (30.2%, 46.2%) | – | – | ||
| Distant metastatic site > 1 | 184 | 120 (65%) | 27.8 | 21.1% (13.2%, 28.9%) | 1.64 (1.27, 2.13) | 0.88 (0.60, 1.29) | ||
| Numbers of used EGFR-TKI | < 0.0001 | 0.6466 | ||||||
| 0–1 | 219 | 128 (58%) | 26.2 | 25.7% (18.2%, 33.2%) | – | – | ||
| 2–4 | 198 | 100 (51%) | 44.1 | 35.7% (27.2%, 44.2%) | 0.55 (0.42, 0.72) | 0.89 (0.56, 1.44) | ||
| Type of 1st-line EGFR-TKI | 0.3573 | |||||||
| No EGFR-TKI treatment | 60 | 31 (52%) | 1.7 | 36.7% (22.4%, 50.9%) | – | – | ||
| Erlotinib/Gefitinib | 308 | 174 (56%) | 3.0 | 30.1% (23.6%, 36.5%) | 0.76 (0.52, 1.11) | – | ||
| Afatinib | 49 | 23 (47%) | 2.9 | 28.9% (7.6%, 50.2%) | 0.77 (0.45, 1.33) | – | ||
| 1st-line EGFR-TKI | < 0.0001 | 0.0005 | ||||||
| No EGFR-TKI treatment | 60 | 31 (52%) | 20.5 | 36.7% (22.4%, 50.9%) | – | – | ||
| 1st/2nd generation TKI without subsequent osimertinib | 203 | 133 (66%) | 26.1 | 17.1% (10.3%, 23.9%) | 1.12 (0.76, 1.66) | 1.05 (0.68, 1.61) | ||
| Osimertinib as ≥ 2nd-line TKI without T790M | 59 | 22 (37%) | 50.8 | 48.7% (30.6%, 66.8%) | 0.47 (0.27, 0.82) | 0.46 (0.22, 0.96) | ||
| Osimertinib as ≥ 2nd-line TKI with T790M | 95 | 42 (44%) | 52.8 | 46.6% (34.8%, 58.3%) | 0.44 (0.28, 0.70) | 0.33 (0.16, 0.69) | ||
BMI body mass index, CI confidence interval, CNS central nervous system, ECOG Eastern Cooperative Oncology Group, EGFR epidermal growth factor receptor, LN lymph node, NR not reached, PD-L1 programmed cell death ligand 1, TKI tyrosine kinase inhibitor
The KM estimates of TTF for patients receiving osimertinib (n = 154) showed that patients with PD-L1 expression ≥ 50% remained on osimertinib for a significantly shorter time than those with PD-L1 expression < 50% (p = 0.0008) (Fig. 2A). Patients with 1st-line TKI treatment duration ≤ 12 months had inferior TTF of osimertinib than those with treatment duration > 12 months (p = 0.0019) (Fig. 2B). The KM estimates of OS for EGFR-mutant lung cancer patients (n = 417) demonstrated that patients with PD-L1 expression < 50% survived longer than those with PD-L1 expression ≥ 50% (p = 0.0109) (Fig. 2C). Osimertinib as ≥ 2nd-line TKI treatment in patients with or without identification of T790M achieved better survival than first-line 1st/2nd generation TKI treatment without subsequent osimertinib (p = 0.0002, 0.0232, respectively). However, no significant different OS was found between osimertinib-treated patients with T790M and those without (p = 0.2803) (Fig. 2D).
Fig. 2.
Kaplan–Meier (KM) estimates of time-to-treatment failure (TTF) for patients receiving osimertinib (n = 154) revealed that A patients with PD-L1 expression of 50% or more remained on osimertinib for a shorter time than those with PD-L1 expression less than 50% (adjusted p = 0.0008). B Patients with 1st-line TKI treatment duration ≤ 12 months had inferior TTF than those with treatment duration > 12 months (p = 0.0019). KM estimates of OS for EGFR-mutant lung cancer patients (n = 417) demonstrated that C patients with PD-L1 expression < 50% survived longer than those with PD-L1 expression ≥ 50% (adjusted p = 0.0109). D Osimertinib as ≥ 2nd-line TKI treatment in patients with or without identification of T790M showed better survival than first-line 1st/2nd generation EGFR-TKI treatment without subsequent osimertinib (adjusted p = 0.0002, 0.0232, respectively). There is no significantly different OS between osimertinib-treated patients with T790M and those without (adjusted p = 0.2803)
Discussion
This study tested the hypothesis that osimertinib can provide survival benefit in EGFR-mutant patients without identification of a T790M after 1st/2nd generation TKI treatment in real-world practice, suggesting that osimertinib as ≥ 2nd line treatment led to better survival in previously EGFR-TKI treated NSCLC patients without identification of T790M mutation when compared to those receiving 1st-line 1st/2nd generation TKI without subsequent osimertinib in terms of OS. Osimertinib has been approved and recommended for the 1st-line treatment of advanced EGFR-mutant NSCLC patients in the U.S. and Europe (Soria et al. 2018; Ramalingam et al. 2019), but, currently, many patients still undergo 1st/2nd generation TKIs as the 1st-line treatment due to drug availability or financial burden in developing regions (Cai et al. 2019; Wu et al. 2018). Osimertinib is also approved for previously TKI treated NSCLC harboring T790M. Nevertheless, the challenges of T790M detection include inadequate medical condition for rebiopsy, inaccessible progressive sites, insufficient tumor tissue, and false negative results due to tumor heterogeneity or limited sensitivity (Tokaca et al. 2018). Consequently, liquid biopsy for detecting genomic alteration from circulating tumor DNA has emerged as the initial approach to discover T790M (Popat 2018). T790M required for subsequent osimertinib was identified in up to 50% of patients receiving 1st-line 1st/2nd generation TKIs (Takeda and Nakagawa 2019). Therefore, owing to the lack of T790M, approximately 50% of patients whose diseases had progressed on 1st-line 1st/2nd generation TKIs would not be eligible for osimertinib. However, our findings reveal that osimertinib could provide a survival benefit in this setting. The potential reasons for the favorable efficacy of osimertinib in EGFR-TKI treated NSCLC without identification of T790M comprise false-negative results of T790M, inability to perform rebiopsy, and greater potency of osimertinib for tumors harboring exon 19 deletion/L858R than erlotinib in preclinical studies (Masuzawa et al. 2017).
Advanced NSCLC leads to a tremendous disease and economic burden for patients and communities due to the cost of EGFR-TKIs (Mortality 2015). The treatment strategy of first-line gefitinib/erlotinib followed by osimertinib has been reported to be more cost-effective than first-line osimertinib for advanced EGFR-mutant NSCLC in developing regions (Cai et al. 2019). The unfavorable cost-effectiveness was mainly derived from the high cost of osimertinib (Wu et al. 2018). Therefore, later-line osimertinib treatment could be an alternative economical therapy for previously EGFR-TKI treated NSCLC with T790M and may be a potential alternative option for TKI treated NSCLC without identification of T790M.
Liquid biopsy has emerged as a promising and useful biopsy method, allowing for genetic profiling, disease monitoring and personalized therapy (Saarenheimo et al. 2019). Liquid biopsy is also clinically useful in advanced NSCLC and confers potential advantages over tumor biopsy, because that it reflects the spatial/temporal heterogeneity of tumor and is minimally invasive so repeat testing is easier and safer for monitoring treatment efficacy, disease progression, and drug resistance (Mathai et al. 2019). High concordance rates of EGFR mutation between tumor and liquid biopsies have been reported (He et al. 2017; Veldore et al. 2018). However, one study demonstrated that only 66% of somatic mutations identified by tumor biopsies were detected by next-generation sequencing-based liquid biopsies (Iwama et al. 2017). Our data showed that a lower percentage of patients with acquired T790M received liquid biopsy compared to tumor biopsy and a significantly lower odds ratios (0.301) for detecting T790M by liquid biopsy than tumor biopsy was found. Collectively, these findings support that liquid biopsy could be a valid tool for detection of T790M and repeat tumor biopsy should be considered if liquid biopsy is uninformative (Rolfo et al. 2018, 2021).
In the previous study with centrally confirmed T790M, osimertinib demonstrated significantly longer PFS in patients with T790M than those without, after disease progression to initial EGFR-TKIs (Janne et al. 2015). Our real-world data showed that later-line osimertinib in patients with 790 M did not have a longer TTF than those without identification of T790M. This inconsistent finding may result from false-negative T790M results or that some patients without rebiopsy (59.3%), including tissue and liquid biopsies, for EGFR mutation test could have T790M. The independent factors related to the development of acquired T790M in the prior report include first-line EGFR-TKI duration > 13 months, male, and liver metastasis; (Lin et al. 2019) our findings show that no surgery treatment is also related factors. Moreover, in our study, patients with 1st-line TKI duration > 12 months had longer TTF than those with 1st-line TKI duration ≤ 12 months, suggesting that shorter 1st-line TKI treatment duration is significantly associated with poor efficacy of osimertinib. Our results also showed that osimertinib-treated patients receiving 1st-line TKI with afatinib had shorter TTF than those receiving Erlotinib/Gefitinib, but OS analysis shows that patients receiving 1st-line TKI with afatinib did not have worse survival than those receiving 1st-line TKI with erlotinib/gefitinib (p = 0.9126). A previous report showed that 1st-line afatinib followed by osimertinib in patient with acquired T790M demonstrated 27.6 months of median duration from afatinib initiation to the last dose of osimertinib or death in real-world practice; so, collectively, 1st-line afatinib with subsequent osimertinib is an effective treatment (Hochmair et al. 2018).
The PFS of EGFR-TKI was reported to be significantly worse in EGFR-mutant lung cancer with PD-L1 expression ≥ 1% (Yoneshima et al. 2018). Our results show that EGFR-mutant NSCLC with PD-L1 expression ≥ 50% had worse TTF on osimertinib and OS than those with PD-L1 expression < 50%. PD-L1 expression ≥ 50% is associated with a significantly poor outcome in patients receiving osimertinib. EGFR activation was shown to promote PD-L1 expression and lead to tumor evasion from immune response, and accumulating evidence suggests that activated EGFR signal pathway is essential to the activation of regulatory T cells and dendritic cells (Yu et al. 2018). Immune suppression and tolerance in EGFR-mutant lung cancer with high PD-L1 expression could explain the worse survival observed in our study. Several trials of later-line treatment for advanced NSCLC have shown that EGFR-mutant lung cancer patients did not benefit from PD-1/L1 inhibitors (Rittmeyer et al. 2017; Herbst et al. 2016). Recently, the atezolizumab plus chemotherapy and bevacizumab significantly improve PFS among patients with EGFR-mutant lung cancer in the exploratory analyses (Socinski et al. 2018). Taken together, the optimal therapeutic agents aside from EGFR-TKI are critical in EGFR-mutant patients with high PD-L1 expression. For EGFR-mutant NSCLC patients with PD-L1 expression ≥ 50%, further studies are warranted to clarify the best treatment strategies following EGFR-TKI treatment.
The limitations of this study were that the investigation was conducted in a single tertiary-care referral center and patient numbers are relatively small. The current study includes the prospective cohort and the consecutive case series. Therefore, selection bias was a concern and EGFR mutation rate in this study might not be completely representative of the general lung cancer population. Furthermore, first-line osimertinib treatment in advanced EGFR-mutant NSCLC patients was recently approved; (Soria et al. 2018; Ramalingam et al. 2019) however, patients receiving first-line osimertinib were not enrolled in this study because the follow-up time is unlikely to be sufficient at present. Although rebiopsy is not possible to be performed among every EGFR-TKI treated NSCLC patients in real-world practice, 35 (59.3%) out of 59 osimertinib treated patients without identification of T790M did not receive rebiopsy, resulting in potentially undiscovered T790M among patients without identification of T790M in this research. Additionally, the survival analysis of the study population was potentially confounded by the evolving treatment paradigm of lung cancer in recent years.
Conclusions
This study demonstrates that osimertinib as second line or subsequent TKI in EGFR-TKI treated patients without identification of T790M revealed lower risk of death compared to first-line 1st/2nd generation TKI without subsequent osimertinib in real-world practice. Additionally, EGFR-mutant patients with PD-L1 expression ≥ 50% appear to have a higher risk of treatment failure for osimertinib and worse OS than those with PD-L1 expression < 50%. These results suggest that osimertinib as second line or subsequent TKI may be a potential alternative option for the treatment of patients without identification of T790M and PD-L1 expression ≥ 50% is associated with a significantly poor outcome in patients receiving osimertinib.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We thank Barbara A Abbott for assisting access Rochester Epidemiology Project resource that was made available to this study. The authors appreciate Chia-Sui Weng for her technical assistance with the manuscript.
Abbreviations
- CI
Confidence intervals
- Cox
Cox proportional hazard
- EGFR
Epidermal growth factor receptor
- HR
Hazard ratios
- IRB
Institutional Review Boards
- KM
Kaplan–Meier
- NSCLC
Non-small cell lung cancer
- TTF
Time-to-treatment failure
- TKIs
Tyrosine kinase inhibitors
Author contributions
Y and L had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design Y, L, Y. Acquisition, analysis, or interpretation of data Y, L, L, W, L, P, L, L, Y. Drafting of the manuscrip Y, L. Critical revision of the manuscript for important intellectual content Y, T, C, Y, C. Statistical analysis Y, W, L, L. Obtained funding Y, L. Administrative, technical, or material support: Y, W, L, Y, L. Supervision Y, C.
Funding
This study was supported by the National Institutes of Health [Grant Numbers R03 CA77118, R01 CA80127 and R01 CA84354]; the National Institute on Aging [Grant Numbers R01 AG034676 and R01 AG052425]; the Mayo Clinic Foundation; and Yin Shu-Tien Foundation Taipei Veterans General Hospital-National Yang-Ming University Excellent Physician Scientists Cultivation Program [Grant Number 108-V-A-012]; Ministry of Health and Welfare, Taiwan [Grant Number MOHW109-TDU-B-211-134019]; Ministry of Science and Technology (MOST) (Grant Numbers 108-2628-B-075-007 and 109-2628-B-075-023), Taipei Veterans General Hospital (Grant Numbers V108E-006-3[109] and V109C-123).
Data availability
All authors confirm that all data and materials are available and support our claims.
Code availability
All authors confirm that custom code is available and support our claims.
Declarations
Conflict of interest
The authors have no conflicts of interest.
Ethical approval
This medical record-based study was approved by the Institutional Review Boards (IRB) of Mayo Foundation and Olmsted County Medical Center.
Footnotes
Role of the funding source: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- Barlesi F, Mazieres J, Merlio JP et al (2016) Routine molecular profiling of patients with advanced non-small-cell lung cancer: results of a 1-year nationwide programme of the French Cooperative Thoracic Intergroup (IFCT). Lancet 387:1415–1426 [DOI] [PubMed] [Google Scholar]
- Cai H, Zhang L, Li N et al (2019) Cost-effectiveness of osimertinib as first-line treatment and sequential therapy for EGFR mutation-positive non-small cell lung cancer in China. Clin Ther 41:280–290 [DOI] [PubMed] [Google Scholar]
- Cortot AB, Janne PA (2014) Molecular mechanisms of resistance in epidermal growth factor receptor-mutant lung adenocarcinomas. Eur Respir Rev 23:356–366 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cross DA, Ashton SE, Ghiorghiu S et al (2014) AZD9291, an irreversible EGFR TKI, overcomes T790M-mediated resistance to EGFR inhibitors in lung cancer. Cancer Discov 4:1046–1061 [DOI] [PMC free article] [PubMed] [Google Scholar]
- He J, Tan W, Ma J (2017) Circulating tumor cells and DNA for real-time EGFR detection and monitoring of non-small-cell lung cancer. Future Oncol 13:787–797 [DOI] [PubMed] [Google Scholar]
- Herbst RS, Baas P, Kim DW et al (2016) Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial. Lancet 387:1540–1550 [DOI] [PubMed] [Google Scholar]
- Hirsh V (2018) Turning EGFR mutation-positive non-small-cell lung cancer into a chronic disease: optimal sequential therapy with EGFR tyrosine kinase inhibitors. Ther Adv Med Oncol 10:1758834017753338 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hochmair MJ, Morabito A, Hao D et al (2018) Sequential treatment with afatinib and osimertinib in patients with EGFR mutation-positive non-small-cell lung cancer: an observational study. Future Oncol 14:2861–2874 [DOI] [PubMed] [Google Scholar]
- Iwama E, Sakai K, Azuma K et al (2017) Monitoring of somatic mutations in circulating cell-free DNA by digital PCR and next-generation sequencing during afatinib treatment in patients with lung adenocarcinoma positive for EGFR activating mutations. Ann Oncol 28:136–141 [DOI] [PubMed] [Google Scholar]
- Janne PA, Yang JC, Kim DW et al (2015) AZD9291 in EGFR inhibitor-resistant non-small-cell lung cancer. N Engl J Med 372:1689–1699 [DOI] [PubMed] [Google Scholar]
- Kalemkerian GP, Narula N, Kennedy EB (2018) Molecular testing guideline for the selection of lung cancer patients for treatment with targeted tyrosine kinase inhibitors: American Society of Clinical Oncology Endorsement Summary of the College of American Pathologists/International Association for the Study of Lung Cancer/Association for Molecular Pathology Clinical Practice Guideline Update. J Oncol Pract 14:323–327 [DOI] [PubMed] [Google Scholar]
- Leighl NB, Rekhtman N, Biermann WA et al (2014) Molecular testing for selection of patients with lung cancer for epidermal growth factor receptor and anaplastic lymphoma kinase tyrosine kinase inhibitors: American Society of Clinical Oncology endorsement of the College of American Pathologists/International Association for the study of lung cancer/association for molecular pathology guideline. J Clin Oncol 32:3673–3679 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Leventakos K, Kipp BR, Rumilla KM et al (2016) S768I mutation in EGFR in patients with lung cancer. J Thorac Oncol 11:1798–1801 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lin YT, Chen JS, Liao WY et al (2019) Clinical outcomes and secondary epidermal growth factor receptor (EGFR) T790M mutation among first-line gefitinib, erlotinib and afatinib-treated non-small cell lung cancer patients with activating EGFR mutations. Int J Cancer 144:2887–2896 [DOI] [PubMed] [Google Scholar]
- Luo YH, Luo L, Wampfler JA et al (2019) 5-year overall survival in patients with lung cancer eligible or ineligible for screening according to US Preventive Services Task Force criteria: a prospective, observational cohort study. Lancet Oncol 20:1098–1108 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Masuzawa K, Yasuda H, Hamamoto J et al (2017) Characterization of the efficacies of osimertinib and nazartinib against cells expressing clinically relevant epidermal growth factor receptor mutations. Oncotarget 8:105479–105491 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mathai RA, Vidya RVS, Reddy BS et al (2019) Potential utility of liquid biopsy as a diagnostic and prognostic tool for the assessment of solid tumors: implications in the precision oncology. J Clin Med. 10.3390/jcm8030373 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mok TS, Wu YL, Thongprasert S et al (2009) Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 361:947–957 [DOI] [PubMed] [Google Scholar]
- Mok TS, Wu YL, Ahn MJ et al (2017) Osimertinib or platinum-pemetrexed in EGFR T790M-positive lung cancer. N Engl J Med 376:629–640 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mortality GBD, Causes of Death C (2016) Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet 388:1459–1544 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Murtuza A, Bulbul A, Shen JP et al (2019) Novel third-generation EGFR tyrosine kinase inhibitors and strategies to overcome therapeutic resistance in lung cancer. Cancer Res 79:689–698 [DOI] [PubMed] [Google Scholar]
- Novello S, Barlesi F, Califano R et al (2016) Metastatic non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 27:v1–v27 [DOI] [PubMed] [Google Scholar]
- Ohashi K, Maruvka YE, Michor F, Pao W (2013) Epidermal growth factor receptor tyrosine kinase inhibitor-resistant disease. J Clin Oncol 31:1070–1080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Oxnard GR, Hu Y, Mileham KF et al (2018) Assessment of resistance mechanisms and clinical implications in patients with EGFR T790M-positive lung cancer and acquired resistance to osimertinib. JAMA Oncol 4:1527–1534 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Popat S (2018) Osimertinib as first-line treatment in EGFR-mutated non-small-cell lung cancer. N Engl J Med 378:192–193 [DOI] [PubMed] [Google Scholar]
- Ramalingam SS, Gray JE, Ohe Y et al (2019) Osimertinib vs comparator EGFR-TKI as first-line treatment for EGFRm advanced NSCLC (FLAURA): final overall survival analysis. Ann Oncol 30:v851–v934 [Google Scholar]
- Rittmeyer A, Barlesi F, Waterkamp D et al (2017) Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial. Lancet 389:255–265 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rolfo C, Mack PC, Scagliotti GV et al (2018) Liquid biopsy for advanced non-small cell lung cancer (NSCLC): a statement paper from the IASLC. J Thorac Oncol 13:1248–1268 [DOI] [PubMed] [Google Scholar]
- Rolfo C, Mack P, Scagliotti GV et al (2021) Liquid biopsy for advanced non-small cell lung cancer: a consensus statement from The International Association for the Study of Lung Cancer (IASLC). J Thorac Oncol. 10.1016/j.jtho.2021.06.017 [DOI] [PubMed] [Google Scholar]
- Saarenheimo J, Eigeliene N, Andersen H et al (2019) The value of liquid biopsies for guiding therapy decisions in non-small cell lung cancer. Front Oncol 9:129 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Socinski MA, Jotte RM, Cappuzzo F et al (2018) Atezolizumab for first-line treatment of metastatic nonsquamous NSCLC. N Engl J Med 378:2288–2301 [DOI] [PubMed] [Google Scholar]
- Soria JC, Ohe Y, Vansteenkiste J et al (2018) Osimertinib in untreated EGFR-mutated advanced non-small-cell lung cancer. N Engl J Med 378:113–125 [DOI] [PubMed] [Google Scholar]
- Takeda M, Nakagawa K (2019) First- and second-generation EGFR-TKIs are all replaced to osimertinib in chemo-naive EGFR mutation-positive non-small cell lung cancer? Int J Mol Sci. 10.3390/ijms20010146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tokaca N, Barth S, O’Brien M et al (2018) Molecular adequacy of image-guided rebiopsies for molecular retesting in advanced non-small cell lung cancer: a single-center experience. J Thorac Oncol 13:63–72 [DOI] [PubMed] [Google Scholar]
- Veldore VH, Choughule A, Routhu T et al (2018) Validation of liquid biopsy: plasma cell-free DNA testing in clinical management of advanced non-small cell lung cancer. Lung Cancer (auckl) 9:1–11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang L, Ruan X, Yang P, Liu H (2016) Comparison of three information sources for smoking information in electronic health records. Cancer Inform 15:237–242 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wu YL, Cheng Y, Zhou X et al (2017) Dacomitinib versus gefitinib as first-line treatment for patients with EGFR-mutation-positive non-small-cell lung cancer (ARCHER 1050): a randomised, open-label, phase 3 trial. Lancet Oncol 18:1454–1466 [DOI] [PubMed] [Google Scholar]
- Wu B, Gu X, Zhang Q (2018) Cost-effectiveness of osimertinib for EGFR mutation-positive non-small cell lung cancer after progression following first-line EGFR TKI therapy. J Thorac Oncol 13:184–193 [DOI] [PubMed] [Google Scholar]
- Xie D, Allen MS, Marks R et al (2018) Nomogram prediction of overall survival for patients with non-small-cell lung cancer incorporating pretreatment peripheral blood markers. Eur J Cardio-Thorac Surg 53:1214–1222 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yang P, Wang Y, Wampfler JA et al (2016) Trends in subpopulations at high risk for lung cancer. J Thorac Oncol 11:194–202 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yang JC, Ahn MJ, Kim DW et al (2017) Osimertinib in pretreated T790M-positive advanced non-small-cell lung cancer: aura study phase ii extension component. J Clin Oncol 35:1288–1296 [DOI] [PubMed] [Google Scholar]
- Yoneshima Y, Ijichi K, Anai S et al (2018) PD-L1 expression in lung adenocarcinoma harboring EGFR mutations or ALK rearrangements. Lung Cancer 118:36–40 [DOI] [PubMed] [Google Scholar]
- Yu S, Liu D, Shen B et al (2018) Immunotherapy strategy of EGFR mutant lung cancer. Am J Cancer Res 8:2106–2115 [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All authors confirm that all data and materials are available and support our claims.
All authors confirm that custom code is available and support our claims.