Abstract
Background: The use of medical cannabis has rapidly increased among cancer patients worldwide. Cannabis is often administered concomitantly with cancer medications, including immune checkpoint inhibitors (ICIs). As the cannabinoid receptors are abundantly expressed and modulate immune cells, it has been hypothesised that cannabis may attenuate the activity of ICIs. We aimed to assess the effect of cannabis on ICIs’ efficiency in patients having non-small cell lung cancer (NSCLC).
Method: The murine model of CT26 tumour-bearing mice treated with an anti-PD-1 antibody and Δ9-tetrahydrocannabinol (THC) was used to evaluate the interaction between THC and ICIs in vivo. Correlation between use of medical cannabis and clinical outcome was evaluated in a cohort of 201 consecutive metastatic NSCLC patients treated with monotherapy pembrolizumab as a first-line treatment.
Results: Median overall survival (OS) of the mice receiving a control vehicle, THC, anti-PD-1 antibody or their combination was 21, 24, 31 and 54 days, respectively (p < 0.05 for the combination treatment compared to a control vehicle), indicating that THC did not reduce the efficacy of anti-PD-1 therapy. Of 201 NSCLC patients treated with first-line monotherapy pembrolizumab for metastatic disease, 102 (50.7%) patients received licence for cannabis within the first month of treatment. Cannabis-treated patients were younger compared to the cannabis naïve patients (median age 68 versus 74, p = 0.003), with female predominance (62, 60.8% versus 34, 34.3%, p = 0.002) and with more prevailing brain metastasis (15.7% versus 5%, p = 0.013). Similar distribution of histology, smoking status, ECOG (Eastern Cooperative Oncology Group) and programmed death-ligand 1 expression was noted between the groups. Liver metastases were marginally significant (19.6% versus 10.1%, p = 0.058). The most common indication for cannabis was pain (71%) followed by loss of appetite (34.3%). Time to tumour progression was similar for cannabis-naive and cannabis-treated patients (6.1 versus 5.6 months, respectively, 95% confidence interval, 0.82 to 1.38, p = 0.386), while OS was numerically higher in the cannabis-naive group (54.9 versus 23.6 months) but did not reach statistical significance (95% confidence interval 0.99 to 2.51, p = 0.08). In multivariate analyses, we did not identify cannabis use as an independent predictor factor for mortality. Conclusions: Preclinical and clinical data suggest no deleterious effect of cannabis on the activity of pembrolizumab as first-line monotherapy for advanced NSCLC. The differences in OS can most likely be attributed to higher disease burden and more symptomatic disease in the cannabis-treated group. These data provide reassurance regarding the absence of a deleterious effect of cannabis in this clinical setting.