Following the Methods as detailed in the Supplementary file, four patients with a diagnosis of resectable ccRCC who were recruited to the WIRE trial underwent successful HP 13C-MRI before and after treatment. In brief, the patient was imaged in a 3 T MRI (MR750, GE Healthcare, Waukesha, WI, USA) equipped with a 13C-transmit clamshell coil and a 13C-tuned 8-channel array coil (Rapid Biomedical, Rimpar, Germany) positioned over the tumor. A 250 mM hyperpolarized pyruvate solution (0.4 mL/kg) was administered via a power injector (Medrad) at 5 mL/s. 13C images were acquired starting 12 s after the injection, with a temporal resolution of 4 s for 20 time points, and had a true resolution of 17 × 17 mm.

An overview of the patient clinical characteristics is presented in Table 1.

Representative LAC/PYR ratio maps overlaid on anatomical T1w images before and after treatment are presented in Fig. 1a. Quantitative results are presented in Suppl. Table S1. The mean LAC/PYR ratio change for all four patients showed a slight increase of + 6% post-treatment but with a significant interpatient variation in S.D. of 27%. Patients 1 (cedirinib monotherapy) and 2 (cediranib and olaparib) demonstrated a decrease in the tumor LAC/PYR of −21% (baseline: 0.19; post-treatment: 0.15) and − 14% (baseline: 0.21; post-treatment: 0.18) respectively. Patients 3 (cediranib and olaparib) and 4 (olaparib monotherapy) showed an increase in the tumor LAC/PYR ratio post-treatment of + 21% (baseline: 0.07; post-treatment: 0.09) and + 35% (baseline: 0.17; post-treatment: 0.23) respectively.

The perfusion surrogate, fp, showed the greatest and most consistent changes post-treatment of −31 ± 18%, followed by the hypoxia surrogate, R2*, of + 18 ± 21% (Fig. 1b). The treatment effect on the diffusion coefficient (D0) was variable, with Patients 1 and 3 exhibited a decrease (−14% and − 3%, respectively), and Patients 2 and 4 an increase (+ 4% and + 10%, respectively). Only small alterations were observed in tumor diameters, with a mean decrease of −4% from 9.1 cm at baseline to 8.7 cm post-treatment (S.D. 3%), and the RECIST 1.1 indicating stable disease in all patients. Mean volumetric changes were greater: −10% over the treatment time course (S.D. 18%); Patient 4 (olaparib only) showed a 13% increase in tumor volume post-treatment which was associated with the greatest increase in D0 of + 10% indicating reduced restriction, the largest decrease in the fp of −57%, as well as the highest baseline R2*, which remained high post-treatment (0% change). Correlative analysis as shown in Fig. 1c, with detailed results in Suppl. Table S2, confirmed a significant relationship between the %change in tumor volume and D0 (Pearson r = 0.96, P < 0.05), and though non-significant, a strong negative correlation coefficient with fp (Pearson r = −0.84, P = 0.165) and R2* (Pearson r = −0.92, P = 0.052). Importantly, the LAC/PYR changes were independent of the alterations in other imaging metrics.

a Representative LAC/PYR maps of the four patients overlaid on T1w images, with tumor ROIs annotated in blue (baseline) and green (post-treatment). Percentage changes between the baseline and post-treatment LAC/PYR ratios are presented with arrows for each individual patient. ROI = region of interest. b Paired plots representing changes in the imaging parameters within tumor ROIs pre- (blue) and post-treatment (green). The LAC/PYR ratio and D0 showed variable changes, whereas fp and R2* were most consistently altered post-treatment. c Correlation analyses between these imaging parameters are presented in a heatmap, labelled with respective Pearson coefficients. B = baseline, After-Tx = after-treatment