Pre- and post-operative anti-PD-L1 plus anti-angiogenic therapies in mouse breast or renal cancer models of micro- or macro-metastatic disease

Abstract

Background: There are phase 3 clinical trials underway evaluating anti-PD-L1 antibodies as adjuvant (postoperative) monotherapies for resectable renal cell carcinoma (RCC) and triple-negative breast cancer (TNBC); in combination with antiangiogenic VEGF/VEGFR2 inhibitors (e.g., bevacizumab and sunitinib) for metastatic RCC; and in combination with chemotherapeutics as neoadjuvant (preoperative) therapies for resectable TNBC.

Methods: This study investigated these and similar clinically relevant drug combinations in highly translational preclinical models of micro- and macro-metastatic disease that spontaneously develop after surgical resection of primary kidney or breast tumours derived from orthotopic implantation of murine cancer cell lines (RENCA^luc or EMT-6/CDDP, respectively).

Results: In the RENCA^luc model, adjuvant sunitinib plus anti-PD-L1 improved overall survival compared to either drug alone, while the same combination was ineffective as early therapy for unresected primary tumours or late-stage therapy for advanced metastatic disease. In the EMT-6/CDDP model, anti-PD-L1 was highly effective as an adjuvant monotherapy, while its combination with paclitaxel chemotherapy (with or without anti-VEGF) was most effective as a neoadjuvant therapy.

Conclusions: Our preclinical data suggest that anti-PD-L1 plus sunitinib may warrant further investigation as an adjuvant therapy for RCC, while anti-PD-L1 may be improved by combining with chemotherapy in the neoadjuvant but not the adjuvant setting of treating breast cancer.

Fig. 1

Combining an anti-PD-L1 antibody with sunitinib in the adjuvant (postoperative) therapy setting for renal cancer. At 7 days after intra-renal implantation of 10⁵ RENCA^luc cells, primary orthotopic (intra-renal (i.r.)) kidney tumours were surgically resected. Adjuvant therapies began on day 10 for sunitinib and its vehicle control, and on day 11 for the anti-PD-L1 antibody (clone 10F.9G2) and its isotype control. a Kaplan–Meier survival analyses and log-rank tests show PD-L1 inhibition prolonging overall survival when added to adjuvant sunitinib therapy. b Bioluminescent imaging of treatment effects on lung metastatic burden. The P values displayed are derived from unpaired t tests of log₁₀-transformed thoracic bioluminescent fluxes measured from images taken between 19 and 21 days post implantation (this experiment was replicated three times, with n = 17–19 mice per treatment group in total; see Suppl. Fig. S1 for further details). Means ± SEM are shown

Fig. 2

Combining an anti-PD-L1 antibody with an anti-VEGF-A antibody in the adjuvant (postoperative) therapy setting for renal cancer. At 7 days after intra-renal implantation of 10⁵ RENCA^luc cells, primary orthotopic (intra-renal (i.r.)) kidney tumours were surgically resected. Adjuvant therapies began on day 10 for the anti-VEGF-A antibody (clone B20-4.1.1) and on day 11 for the anti-PD-L1 antibody (clone 6E11), along with their respective isotype controls. a Kaplan–Meier survival analyses and log-rank tests show no difference in overall survival with the addition of PD-L1 inhibition to anti-VEGF-A therapy. b Bioluminescent imaging of treatment effects on lung metastatic burden. The P values displayed are derived from unpaired t tests of log₁₀-transformed thoracic bioluminescent fluxes measured from images taken at day 16 post implantation (n = 7–8 mice per treatment group). Means ± SEM are shown

Fig. 3

Sunitinib plus anti-PD-L1 combination therapy in the unresected primary tumour setting and postsurgical advanced metastatic disease setting of renal cancer. a, b 10⁵ RENCA^luc cells were orthotopically implanted on day 0. Kaplan–Meier survival analyses show that adding PD-L1 inhibition to sunitinib therapy does not improve survival either in the primary tumour therapy setting or in the late-stage therapy setting. a In the presence of unresected orthotopic primary kidney tumours, sunitinib and anti-PD-L1 (clone 10 F.9G2) therapies were started at 5 and 6 days post implantation (DPI) respectively. Three mice in the anti-PD-L1 monotherapy group experienced severe acute toxicity (laboured breathing and lethargy) immediately after administration of their 4th dose (17DPI), requiring immediate killing. b With the primary kidney tumours resected at 7 DPI, sunitinib (60 mg/kg/d) and anti-PD-L1 (clone 10F.9G2) treatments were initiated at 14 and 15 DPI respectively, i.e., after the establishment of advanced metastatic disease. c Tissues from the control group of the primary tumour therapy experiment (a) were collected at endpoint, sectioned, and subjected to IHC and IF staining for PD-L1; normal kidney tissue and lung metastases showed negligible positivity while the necrotic areas within RENCA^luc primary tumours were often positively stained; see Suppl. S3A for monochrome IF images

Fig. 4

Combining anti-PD-L1 with antiangiogenic agents and/or chemotherapy in the adjuvant (postoperative) setting for breast cancer. Orthotopic (intra-mammary fat pad (imfp)) implantation of 2 × 10⁵ EMT-6/CDDP cells occurred on day 0. The resulting primary breast tumours were resected on day 11 (a) or day 12 (b, c, showing data from one experiment displayed across 2 panels). Adjuvant treatments began 1 day later, involving either 5 weeks (10 doses; a) or 2 weeks (4 doses; b) of the anti-PD-L1 antibody (clone 6E11); 4 weeks of sunitinib (b); 4 weeks of paclitaxel (c); 4 weeks of anti-VEGF antibody (clone B20-4.1.1; a, c). Kaplan–Meier survival curves are shown, along with P values derived from log-rank tests and hazard ratios for relevant comparisons

Fig. 5

Combining anti-PD-L1 with chemotherapy, with or without anti-VEGF-A, in the neoadjuvant (preoperative) setting of breast cancer. At 6 days after the orthotopic implantation of 2 × 10⁵ EMT-6/CDDP cells, neoadjuvant therapies were administered according to the dosing schedules (black arrows) depicted above tumour growth curves in (a). Primary breast tumours were then resected on day 11 post implantation, after which only one group received adjuvant anti-VEGF-A (clone B20-4.1.1) therapy which resumed on day 13, as depicted over the Kaplan–Meier survival curves shown in (b). The resected primary breast tumours were dissociated and subjected to flow cytometry to quantify the % of VEGFR2⁺CD31⁺CD45⁻ endothelial cells (c) and the ratio of CD45⁻ non-immune cells (mostly tumour cells) vs. CD45⁺CD3⁺CD8⁺ T cells (d); *P < 0.05 and **P < 0.01 as calculated by the Kruskal–Wallis test and Dunn’s post-test; means are also depicted. See Suppl. Fig. S5C for other measures of postsurgical outcomes