Abstract 1322P
Background
We evaluated a large real-world dataset for use of IO and chemotherapy (chemo) in clinical practice during a time of evolving treatment landscape in metastatic non-small cell lung cancer (NSCLC), based on PD-L1 tumor proportion score (TPS) in patients from the UK, Germany, France, Italy and Spain.
Methods
De-identified patient data collected from Q3/2018 to Q3/2019 were analyzed. Physicians completed the online data collection form in the Medimix LiveTracker™ based on their patients’ medical records. Patients with metastatic NSCLC without EGFR mutation or ALK translocation and known first line (1L) start date were included. For second line (2L) treatment and 1L to 2L sequencing analysis, a 2L regimen start date was required. Patients were stratified by PD-L1 TPS (<1%, 1-49%, ≥50%).
Results
Of 5,415 patients receiving 1L treatment, 84% of NSCLC tumors were tested for PD-L1 expression. 1L chemo to 2L IO was prescribed to 71% and 84% of patients with PD-L1 TPS <1% and 1-49%, respectively (Table). Patients with PD-L1 TPS ≥50% primarily received IO alone followed by 2L chemo (58%), but 17% received 1L chemo followed by 2L IO. Comparing Q3/2018 versus Q3/2019 1L prescribing, use of IO plus chemo increased from 1% to 17% and 4% to 28% in patients with PD-L1 TPS <1% and 1-49%, respectively. In Q3/2019, no patient with PD-L1 TPS ≥50% received chemo alone compared to 6% a year earlier. Pembrolizumab was the most widely prescribed 1L checkpoint inhibitor, used in 95% of patients receiving IO. In 2L, nivolumab (54%), atezolizumab (23%) and pembrolizumab (23%) were used Table: 1322P
1L and 2L selection and sequence, by PD-L1 TPS
| PD-L1 TPS | |||
| < 1% | 1-49% | ≥ 50% | |
| Median Age, years | 68 | 68 | 68 |
| Male, % | 73 | 71 | 71 |
| 1L and 2L selection, Q3/2018 – Q3/2019: | |||
| 1L, N (%) | 1565 (34) | 1766 (39) | 1218 (27) |
| IO alone, % | 1 | 5 | 80 |
| IO+chemo, % | 3 | 8 | 5 |
| Chemo alone, % | 93 | 84 | 12 |
| 2L, N (%) | 887 (35) | 1163 (45) | 513 (20) |
| IO alone, % | 72 | 88 | 28 |
| IO+chemo, % | 0.4 | 2 | 2 |
| Chemo alone, % | 23 | 9 | 64 |
| 1L→2L sequence, N (%) | 887 (35) | 1163 (45) | 513 (20) |
| 1L chemo→2L IO, % | 71 | 84 | 17 |
| 1L IO→2L chemo, % | 0.6 | 2 | 58 |
| 1L IO→2L IO, % | 0.4 | 1 | 8 |
| 1L chemo→2L chemo, % | 21 | 4 | 2 |
| 1L selection, known PD-L1 TPS, Q3/2018 vs Q3/2019: | |||
| 1L in Q3/2018, N (%) | 283 (37) | 264 (34) | 219 (29) |
| IO alone, % | 0.4 | 2 | 91 |
| IO+chemo, % | 1 | 4 | 1 |
| Chemo alone, % | 95 | 92 | 6 |
| 1L in Q3/2019, N (%) | 59 (31) | 66 (35) | 63 (34) |
| IO alone, % | 5 | 6 | 84 |
| IO+chemo, % | 17 | 28 | 13 |
| Chemo alone, % | 76 | 62 | 0 |
Conclusions
1L and 2L prescribing patterns reflect adherence and adaptation to the emerging clinical trial data and practice guidelines. Our data suggests a clinically significant increase in prescribing IO with or without chemo as 1L treatment in Q3/2019, corresponding with the most recent guideline recommendations.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
Medimix International.
Funding
Medimix International and PPD, LLC.
Disclosure
E. Janowicz, D. Anderson, N. Stoyanov, A. Combest: Full/Part-time employment: PPD. B. Nguyen: Shareholder/Stockholder/Stock options, Full/Part-time employment: PPD. All other authors have declared no conflicts of interest.