Whole body PD-L1 PET in patients with NSCLC and melanoma

• Published in: Journal of clinical oncology Vol. 36; no. 5_suppl; p. 139
• Main Authors: Leung, David K., De Langen, Joop, Raunig, David, Niemeijer, Anna-Larissa N., Smit, Egbert F., Boellaard, Ronald, Vallez-Garcia, David, Dongen, G.a.M.S. van, Windhorst, A.D., Huisman, Marc C., Glaudemans, Andor W. J. M., Hendrikse, N. Harry, Smith, Ralph Adam, Poot, Alex J., Lipovsek, Dasa, Donnelly, David J., Bonacorsi, Samuel J., Velasquez, Linda M., Du, Shuyan, Hayes, Wendy
• Format: Journal Article
• Language: English
• Published: 10.02.2018
• ISSN: 0732-183X; 1527-7755
• DOI: 10.1200/JCO.2018.36.5_suppl.139

Abstract only 139 Background: PD-(L)1 immunotherapy is effective in multiple tumors, including NSCLC and melanoma, but tumor PD-L1 IHC correlates only moderately with treatment outcome. This study aims to assess 1) safety of 18 F-BMS-986192 ( 18 F-PD-L1) in human, 2) PD-L1 quantification in tumors using 18 F-PD-L1 PET, 3) PD-L1 PET correlation with IHC and treatment outcome, and 4) intra and inter subject tracer uptake variability. Methods: Pts with NSCLC (N = 10) and melanoma (N = 3) were included. At baseline, pts received a static or multiphase dynamic whole body PET scan after injecting 200 MBq 18 F-BMS-986192. For NSCLC pts, (1) SUV(max, peak and mean) were measured for each delineable tumor (N = 32, 1-7 tumors/pt), (2) PD-L1 IHC (28.8 assay) was performed on the biopsy, and (3) response to Nivolumab therapy assessed by RECIST 1.1. Intra and inter subject variability and intraclass correlation were calculated using SUVs of all assessed tumors. Equal variance for PD-L1 status was evaluated by a Levene’s test. Four (3 female) pts underwent dosimetry study (ICRP 60). Results: No AEs related to radiotracer was observed. Dosimetry study demonstrated whole body exposure of 30 mGy at dose > 1400 MBq. Biodistribution among pts is comparable. PD-L1 IHC from 13 biopsied lesions were evaluated, 5 < 1%, 4 ≥1%, and 4 ≥50%. Tumor tracer uptake was measured in NSCLC pts and categorized by PDL-1 IHC as ≥50% or < 50%. Clinical trial information: 2015-004760-11. Tumor SUVs did not correlate with RECIST 1.1 assessment. Lesion heterogeneity was reflected in both inter and intra pt variability (CVinter = 41%, CVintra = 53%, ICC = 0.41 for SUVpeak). Levene’s test showed no significance in variability between the two PD-L1 categories. Conclusions: PET-imaging with 18 F-BMS-986192 is safe and feasible in pts with NSCLC and melanoma. Pts with higher PD-L1 PET SUV have higher PD-L1 by IHC. Intra pt variability is similar to inter pt variability. With limited number of pts, no clear correlation of PET PD-L1 and tumor response is observed. A prospective study with this tracer is underway to further investigate 18 F-BMS-986192 in understanding of PD-L1 expression.[Table: see text]