New analyses validate TAP and CPS scores for PD-L1 expression

Exploratory post-hoc analyses of the RATIONALE-305 and the RATIONALE-306 studies showed no significant difference in the results of PD-L1 subgroups when they were defined by either PD-L1 tumour area positivity (TAP) score or combined positive score (CPS). Because generating a TAP score can be automated, this could substantially improve workflows and be easy to implement.

PD-L1 scoring based on the combined positive score (CPS) has shown predictive value for checkpoint inhibitors [1,2]. Although CPS is well-established, it remains a labour-intensive process with heterogeneous results between centres.

The TAP scoring system [3], which evaluates both immune and tumour cells to generate a score for PD-L1 tumour area, was validated for advanced gastric or gastro-oesophageal junction (G/GEJ) cancer in the RATIONALE-305 study (NCT03777657) and for advanced or metastatic oesophageal squamous cell carcinoma (ESCC) in RATIONALE-306 (NCT03783442) [4,5].

In the phase 3 RATIONALE-305 trial (n=1,657), tislelizumab plus chemotherapy demonstrated a significant overall survival (OS) benefit versus placebo plus chemotherapy as first-line therapy, in all randomised participants (HR 0.80; 95% CI 0.70–0.92; P=0.001) and participants with PD-L1 TAP scores ≥5% (HR 0.71; 95% CI 0.58–0.86) [6].

The phase 3 RATIONALE-306 trial recently demonstrated a superior OS with first-line tislelizumab plus chemotherapy compared with placebo plus chemotherapy, in all participants (HR 0.66; 95% CI 0.54–0.80) and in participants with PD-L1 TAP scores ≥10% (HR 0.62; 95% CI 0.44–0.86) [7].

In both trial cohorts, researchers directly compared the prognostic value of CPS versus TAP score [4,5]. The results showed that PD-L1 status was comparable across arms by TAP score or CPS under different thresholds. Both TAP and CPS scores similarly predicted OS and PFS in patients with PD-L1 1%, 5%, and 10% cut-off thresholds (see Figure). In addition, a good correlation was observed between TAP score and CPS based on the interclass correlation coefficient (ICC 0.81 and 0.85 in respective trials). TAP and CPS scores also showed substantial concordance in terms of Cohen’s Kappa and overall percent agreement (OPA) at matched thresholds for each score.

Figure: OS improvement for tislelizumab plus chemotherapy versus placebo plus chemotherapy across PD-L1 subgroups by TAP score and CPS in RATIONALE-306 [5]



CPS, combined positive score; CT, chemotherapy; HR, hazard ratio; OS, overall survival; PBO, placebo; TAP, tumour area positivity; TIS, tislelizumab.

The conclusions from these analyses are that both TAP and CPS scores are viable for PD-L1 expression measurement in patients with G/GEJ cancer and ESCC. TAP score and CPS at matched thresholds exhibited substantial concordance among patients. Tislelizumab plus chemotherapy improved OS and PFS of patients within prespecified PD-L1 subgroups by TAP score, and demonstrated comparable OS and PFS results in PD-L1 subgroups by matched CPS.

1. Shitara K, et al. Nature. 2022;603(7903):942-948.
2. Rha SY, et al. Lancet Oncol. 2023;24(11):1181-1195.
3. Liu C, et al. Diagn Pathol. 2023;18:48.
4. Moehler M, et al. Tislelizumab (TIS) plus chemotherapy (CT) vs placebo (PBO) plus CT in HER2-negative advanced or metastatic gastric or gastro-esophageal junction adenocarcinoma (GC/GEJC): PD-L1 biomarker analysis from RATIONALE-305. Abstract 397MO, ESMO Gastrointestinal Cancers Congress 2024, 26–29 June, Munich, Germany.
5. Raymond E, et al. Tislelizumab (TIS) + chemotherapy (CT) vs placebo (PBO) + CT in advanced or metastatic esophageal squamous cell carcinoma (ESCC): PD-L1 biomarker analysis from RATIONALE-306. Abstract 395MO, ESMO Gastrointestinal Cancers Congress 2024, 26–29 June, Munich, Germany.
6. Qiu MZ, et al. BMJ. 2024;385:e078876.
7. Xu J, et al. Lancet Oncol. 2023;24(5):483-495.

Copyright ©2024 Medicom Medical Publishers



Posted on 21 August 202421 August 2024