Towards effective CAIX-targeted radionuclide and checkpoint inhibition combination therapy for advanced clear cell renal cell carcinoma

Background: Immune checkpoint inhibitors (ICI) are routinely used in advanced clear cell renal cell carcinoma (ccRCC). However, a substantial group of patients does not respond to ICI therapy. Radiation is a promising approach to increase ICI response rates since it can generate anti-tumor immunity. Targeted radionuclide therapy (TRT) is a systemic radiation treatment, ideally suited for precision irradiation of metastasized cancer. Therefore, the aim of this study is to explore the potential of combined TRT, targeting carbonic anhydrase IX (CAIX) which is overexpressed in ccRCC, using [177Lu]Lu-DOTA-hG250, and ICI for the treatment of ccRCC. Methods: In this study, we evaluated the therapeutic and immunological action of [177Lu]Lu-DOTA-hG250 combined with aPD-1/a-CTLA-4 ICI. First, the biodistribution of [177Lu]Lu-DOTA-hG250 was investigated in BALB/cAnNRj mice bearing Renca-CAIX or CT26-CAIX tumors. Renca-CAIX and CT26-CAIX tumors are characterized by poor versus extensive T-cell infiltration and homogeneous versus heterogeneous PD-L1 expression, respectively. Tumor-absorbed radiation doses were estimated through dosimetry. Subsequently, [177Lu]Lu-DOTA-hG250 TRT efficacy with and without ICI was evaluated by monitoring tumor growth and survival. Therapy-induced changes in the tumor microenvironment were studied by collection of tumor tissue before and 5 or 8 days after treatment and analyzed by immunohistochemistry, flow cytometry, and RNA profiling. Results: Biodistribution studies showed high tumor uptake of [177Lu]Lu-DOTA-hG250 in both tumor models. Dose escalation therapy studies in Renca-CAIX tumor-bearing mice demonstrated dose-dependent anti-tumor efficacy of [177Lu]Lu-DOTA-hG250 and remarkable therapeutic synergy including complete remissions when a presumed subtherapeutic TRT dose (4 MBq, which had no significant efficacy as monotherapy) was combined with aPD-1+aCTLA-4. Similar results were obtained in the CT26-CAIX model for 4 MBq [177Lu]Lu-DOTA-hG250 + a-PD1. Ex vivo analyses of treated tumors revealed DNA damage, T-cell infiltration, and modulated immune signaling pathways in the TME after combination treatment. Conclusions: Subtherapeutic [177Lu]Lu-DOTA-hG250 combined with ICI showed superior therapeutic outcome and significantly altered the TME. Our results underline the importance of investigating this combination treatment for patients with advanced ccRCC in a clinical setting. Further investigations should focus on how the combination therapy should be optimally applied in the future.


Cell viability assay
Radiosensitivity to TRT was assessed using cell viability assays using the CellTiter-Glo assay (Promega, G7570) according to manufacturer's instructions.Cells were seeded in 96-well plates to attach overnight, treated for 24h with 0-6 MBq/mL [ 177 Lu]Lu-DOTA-hG250 in 200 µL culture medium, and cultured for another 8 days in fresh medium.Luminescence values were normalized to untreated cells and the data was described using the log(inhibitor)-response variable slope model with constrained bottom (0.0) and top (1.0) values.Differences in IC 50 between cell lines was statistically tested using one-way ANOVA with Tukey's multiple comparison correction.

Animal experiments
Protocols for each experiment were registered at preclinicaltrials.eu (PCTE0000440-PCTE0000445). For each animal experiment, the number of animals injected with tumor cells was determined accounting for an expected tumor take rate of 83% for Renca-CAIX and 71% for CT26-CAIX.For ex vivo biodistribution studies, sample size (n=5/group, 30 total) was selected based on previous biodistribution studies [3], but there was a dropout of 8 animals due to low tumor take resulting in the group sizes presented in Table S3.A priori sample size calculations for therapy studies were based on an effect size of 75% decline in normalized area under the curve (nAUC) compared to control, significance level of 5% and power of 80%.This resulted in a sample size of 10 mice per group for all therapy experiments.In the experiment with CT26-CAIX tumor-bearing mice, failed tumor cell injection causing intraperitoneal tumor growth resulted in 3 dropouts, resulting in the group sizes presented in Table S4.Experiments for ex vivo analyses were determined on 5 (endpoint day 0 or 5) or 6 (endpoint day 8, to account for possible humane endpoints).Initially, nine extra mice were included, because tumor take was higher and humane endpoints were lower than expected.However, two mice were excluded from analyses because of failed tumor injection and complete tumor remission, resulting in the group sizes presented in the heatmaps of Figures 6-7.Selection of mice for each groups was determined by block randomization based on tumor volume, using a random number generator.Mice from different groups were housed together to minimize confounding effects and cages were stored randomly.The order in which mice were treated and measured was random, as well as the measurements of tumor microenvironment characteristics ((immunohistochemistry, flow cytometry, RNA analysis).Biotechnicians, who performed all injections and measurements and determined whether an animal must be sacrificed due to a humane endpoint, were blinded for group allocation.The investigator was blinded for group allocation during assessment of the tumor microenvironment (immunohistochemistry, flow cytometry, RNA analysis), by concealing of animal number.

Dosimetry estimations
Time-activity curves (TAC) for tumors were fitted to a power law function followed by a monoexponential decay in Python.Tumor growth was modelled with a mono exponential function and used to correct both bound activity (%IA/g) over time and define a time-dependent S-value (e.g.absorbed dose rate per unit activity), evaluated by Geant4 11.02 as previously reported [4].The tumor-absorbed dose was calculated by numerical integration of the dose-rate curve (Figure S1).The absorbed dose rate error was determined by propagating uncertainties from activity measurements and tumor growth curve fitting, affecting both bound activities and S-value.The difference between the areas under the dose rate error bounds was used to calculate the uncertainty in tumor-absorbed doses.The time-integrated activity coefficient for livers was calculated according to the trapezoid integration method [5].Normal organ S-values for 177 Lu were obtained from OLINDA/EXM 2.0 and multiplied with obtained TIACs for each organ to calculate absorbed doses according to Medical Internal Radiation Dose (MIRD) Committee methodology [6,7].

Ex vivo flow cytometry
For panel 3, cells were first stained with AH1 dextramer according to the manufacturer's instructions.For all panels, cells were incubated with live/death marker in PBS for 20 min on ice, and subsequently incubated for 10 min on ice with CD16/CD32 (1:800, BD, 553142) for Fc blocking.Cells were incubated with the given extracellular antibodies (Table S2) for 30 min on ice.For panel 1, cells were incubated with BV510-streptavidin secondary antibody (1:300, BD, 563261) for 15 min on ice.For all panels, cells were fixed for 30 min on ice (foxp3/transcription factor staining buffer set, Invitrogen, 00-5523).For panel 2, cells were incubated with FoxP3 antibody for 30 min at RT.Samples were analyzed on a FACSCanto II Flow Cytometry System (BD Biosciences) and results were analyzed with FlowJo using the specified gating strategies (Figure S2).

Figure S2 .
Figure S2.Gating strategies for all flow cytometry panels.Gating for all panels started with general gating (upper panels) followed by panel-specific gating as indicated.Gate setting for each marker was based on Fluorescence Minus One (FMO) controls

Figure S4 .
Figure S4.In vitro radiosensitivity to TRT.Viability of cells after 24 h of treatment with 0-6 MBq/mL [ 177 Lu]Lu-DOTA-hG250 and 8 days additional culturing.Data represent mean ± SD of 2 independent experiments.Nonlinear regression using the log(inhibitor)-response variable slope model was used to fit the data and 95% confidence bands are shown (dashed lines).

Figure S6 .Figure S7 .
Figure S6.Renca-CAIX tumor weights after dissection, before processing for flow cytometry (left panel) or immunohistochemistry and RNA expression profiling (right panel).Data represents mean + SD of all mice (dots) per group.*Missing value.Day 0 Day 5 Day 8

Figure S10 .
Figure S10.Flow cytometry analyses of AH antigen-specific CD8+ T cells in lymph nodes (upper panels) and tumors (lower panels).(A) Dot plots for AH dextramer-positive CD8+ T cells.(B) Graphs show mean (lines) percentage of AH1 dextramer+ CD8+ T cells of individual mice (dots) per group.

Figure S11 .
Figure S11.Undirected global significance scores for nanostring-annotated gene sets for pair-wise comparisons of treatment groups with control group and combination with ICI treatment groups, as determined by RNA expression nanostring analysis and obtained from the Rosalind Platform.

Table S2 .
Antibody panels for flow cytometry

Table S3 .
Statistical parameters for radiosensitivity assays.α/β ratios derived from a linear quadratic model non-linear regression fit to clonogenic survival data and surviving fraction at 2 Gy (SF2) after EBRT and IC50 values after TRT are shown.Data represents mean ± SEM of 2 independent experiments and p-values are given for comparison between the cell lines using one-way ANOVA.