ABSTRACT
Background
Immune checkpoint inhibitors (ICIs) benefit patients across various tumor types. ICIs block cancer and T-cell interactions whereas cannabinoids may inhibit T-cell activation, reducing lysis of tumor cells. Interactions between cannabinoid use and dual ICI treatment remain unknown.
Methods
Individual patient data from 4 Canadian Cancer Trials Group (CCTG) trials of patients treated with dual ICI ± chemotherapy (n = 684) were pooled. Cochran – Mantel – Haenszel and log-rank tests (stratified by trial/treatment arms) correlated cannabinoid use with clinicopathologic characteristics, Best Overall Response (BOR)/iBOR per RECIST 1.1/iRECIST, Progression-Free Survival (PFS)/iPFS, Overall Survival (OS) and immune-related adverse events (irAEs).
Results
Sixty-five (9.5%) patients took cannabinoids at any time on trial, 32 (4.7%) of which were using cannabinoids at baseline. By multivariate analysis, cannabinoid use at baseline was significantly associated with improved iPFS (0.05), but not iBOR (p = 0.15), PFS (p = 0.12), OS (p = 0.35) or incidence of grade 1/2 or 3/4 irAEs (p = 0.96 and 0.65 respectively). Results were not significantly different with cannabinoid use at any time on trial.
Conclusion
Improved iPFS with cannabinoid use in patients treated with durvalumab plus tremelimumab ± chemotherapy did not translate into OS benefits. This study supports the safe use of cannabinoids in the context of combination ICI therapy.
KEYWORDS: Immunotherapy, Anti-PD-1, Anti-PD-L1, Anti-CTLA-4, immune checkpoint inhibitors, cannabinoids, durvalumab, tremelimumab
Plain Language Summary
Immune checkpoint inhibitors (ICI) have become an important treatment option for cancer patients and has been associated with improved survival outcomes across various tumor types. Cannabinoids are active components of cannabis and include tetrahydrocannabinol (THC) and cannabidiol (CBD). Cannabis use has increased in Canadian cancer patients and is often used for symptom management. Some studies have suggested that cannabinoids can alter the function of the immune system, which could impact the effectiveness of immune checkpoint inhibitors. Therefore, we aimed to evaluate whether cannabinoid use might impact the effectiveness of combination ICI treatment with durvalumab plus tremelimumab (with/without chemotherapy) using data from four clinical trials completed through the Canadian Cancer Trials Group (CCTG). We found no significant difference in response, survival outcomes or adverse events in patients who used cannabinoids with combination ICI treatment compared to cannabinoid non-users. This study supports the safe use of cannabinoids in the context of combination ICI therapy.
1. Introduction
The use of cannabis-derived products for cancer-related symptoms is becoming more common [1–4]. In one province in Canada, prior to the legalization of cannabis, its use was reported in 18% of cancer patients in the 6 months prior to their most recent visit to a cancer clinic, 36% of which were new users [5]. Another study reported that the prevalence of current cannabis use by cancer patients in Canada may have increased by 26% post legalization (although this report was limited by a low number of survey responses) [6,7].
Cannabinoids are biologically active components of cannabis, the two most common of which are tetrahydrocannabinol (THC) and cannabidiol (CBD) [3,8,9]. Humans have an endogenous cannabinoid system regulated by CB1 and CB2 receptors, which are bound by anandamide (AEA) and 2-arachidonylglycerol (2-AG) ligands. CB1 receptors are mainly located in the central nervous system but have roles in other tissues including the immune system [8,10]. CB2 receptors are most densely found in immunological tissues and regulate cell fate [8,10]. THC is a high-affinity agonist at CB1, while CBD acts as a low-affinity inverse agonist at both CB1 and CB2 with interactions even at low concentrations [10]. Preclinical studies have shown inhibition of lymphocyte activation by cannabinoids, and immunosuppressive effects including reduced immune cell-mediated lysis of tumor cells, have been reported in animal models [8,11–15]. AEA has an anti-proliferative effect on human T-cells, induces T-cell apoptosis and inhibits inflammatory cytokine release via CB2 and CD4+ T-cell subsets [16–18]. Furthermore, cannabinoids promote production of anti-inflammatory cytokines including IL-4, IL-5, and IL-10 [19].
Immune checkpoint inhibitors (ICIs) including anti-programmed death 1 (PD-1), anti-programmed death ligand 1 (PD-L1) and anti-cytotoxic T lymphocyte-associated protein 4 (CTLA-4) monoclonal antibodies have been incorporated into the treatment of various solid tumors including melanoma, non-small cell lung cancer (NSCLC) and renal cell carcinoma (RCC), among others [20–23]. However, the depth and duration of treatment response varies across patients [24–27].
Considering the evidence linking cannabinoids to suppression of the adaptive immune system, it is possible that they may affect ICI efficacy [19]. Some retrospective and one prospective observational study in tumors that classically respond to ICI have shown that cannabis use affects treatment response, survival outcomes and incidence of immune-related adverse events (irAEs) in patients treated with ICIs [28–30]. Conversely, a recent retrospective study by Waissengrin et al. suggested no significant impact of cannabis on immune checkpoint inhibitor outcomes [31]. Due to small numbers and limitations associated with retrospective and non-randomized studies, further data is required to describe the effect of cannabis use on outcomes in cancer patients receiving ICIs.
The Canadian Cancer Trials Group (CCTG) is an academic cooperative oncology group that designs and conducts clinical trials testing cancer therapy, supportive care and prevention interventions. Since its inception in 1980, CCTG has grown and evolved into the leading international cooperative clinical cancer research network, with over 85 Canadian member institutions and 2,100 investigators, who have facilitated over 500 trials in collaboration with 700 international centers in 40 countries. Multiple CCTG trials have changed practice in Canada and worldwide. This pooled analysis evaluated the impact of cannabinoid use on clinical outcomes in patients with solid tumors being treated with combination durvalumab (anti-PD-L1) and tremelimumab (anti-CTLA-4), with/without chemotherapy, utilizing data from four CCTG clinical trials. It was hypothesized that cannabinoid use would negatively impact best overall response (BOR), progression-free survival (PFS), and overall survival (OS), and reduce the incidence of irAEs.
2. Methods
2.1. Clinical trials
Study objectives were evaluated by combining clinical data from the BR.34, CO.26, PA.7 and IND.226 CCTG clinical trials [32–35]. Study description and treatment arm details for each trial are described in Supplementary Tables S1 and S2.
2.2. Cannabinoid use
All drugs included in the CCTG concomitant medication dictionary were manually annotated by two independent investigators (AF and POG). Based on their activity toward the cannabinoid receptors CB1 and/or CB2, the following substances were defined as cannabinoids and included in this analysis: cannabis oil, marijuana, marijuana oil, propacetamol, dronabinol and nabilone [36]. Patients were considered users if their concomitant medications included one or more of the above drugs (Supplementary Table S3). Patients were classified into three groups of cannabinoid use: 1) baseline (use starting before and including the date of randomization); 2) any time (use at baseline or use starting during the trial up to the study closure date); 3) non-user. The concomitant use of cannabinoids with ICIs when used at baseline, or at any time, was evaluated.
Concomitant medications taken by the patient at study entry were recorded at baseline across all four clinical trials, as well as 4 weeks after completion of the protocol therapy. While on study, concomitant medications were recorded on Day 1 of each cycle and as clinically indicated in the BR.34 and PA.7 trials. In CO.26, medications were reviewed every 4 weeks until disease progression. The timepoints for review of concomitant medications was not specified in the IND.226 protocol.
2.3. Outcome measures
2.3.1. Response to ICI therapy
Responders as measured by RECIST 1.1 included those with a BOR of complete response (CR), partial response (PR) or stable disease (SD), while non-responders had a BOR of progressive disease (PD) [37]. Responders as measured by iRECIST were those with an iBOR of iCR/iPR/iSD, while non-responders had an iBOR of iPD (iUPD- unconfirmed PD or iCPD- confirmed PD) [38]. Drugs known to affect ICI response were controlled for in this analysis (Supplementary Table S2). Cannabinoid use at any time during trial participation was not assessed for its impact on response as the date of BOR/iBOR may have been determined for some patients prior to starting a cannabinoid drug. Of note, BOR was not required to have been centrally confirmed for BR.34, PA.7 and CO.26, whereas central confirmation of BOR was required for IND.226.
2.3.2. Survival outcomes: PFS and OS
Patients on ICI with suspected progression from a radiological scan (iUPD) were to subsequently confirm progression radiologically a minimum of 4 weeks later (iCPD). Only patients with iCPD were counted as iPD except in the following scenarios: i) protocol treatment was stopped because a patient was not considered to be clinically stable; ii) no further response assessments were done (patient refusal or protocol noncompliance or patient death); iii) the next time point responses were all iUPD and iCPD never occurred; iv) the patient died of cancer. PFS was defined as the time from date of randomization to date of PD or death, whichever came first. iPFS was defined as the time from the date of randomization to the date of iPD or death, whichever came first. Patients who had not developed PD/iPD by study closure were censored at the date of the last disease assessment in each trial’s final analysis database. OS was defined as the time from date of randomization to date of death. Patients alive at the time of study closure were censored at the date last known to be alive.
2.3.3. Immune-related adverse events
The Common Terminology Criteria for Adverse Events (CTCAE) version 4 was used for grading of irAEs [39]. Adverse events considered immune-related are listed in Supplementary Table S2.
2.3.4. Quality of life (QOL)
QOL was evaluated with the EORTC QLQ-C30 instrument in the BR.34, PA.7 and CO.26 studies (no QOL data was available for IND.226). The impact of baseline cannabinoid use on QOL was assessed. For each functional domain in the EORTC QLQ-C30 questionnaire, a change in score of 10 points from baseline was defined as clinically relevant. Patients were considered improved if a reported score was 10-points or better than baseline at any time point in the QOL assessment. Conversely, patients were considered worsened if they reported a score minus 10-points or worse than baseline at any time point in the QOL assessment without the above-defined improvement being observed. Patients whose scores were between 10-point changes from baseline at every QOL assessment were considered as stable [40].
2.4. Statistical design and analysis
Tables present patient baseline factors by cannabinoid use status (no vs. yes) both at baseline and at any time during trial participation. Patient clinical characteristics were summarized using descriptive statistics. The p-value was determined using the Cochran – Mantel – Haenszel test and stratified by trial for categorical factors and was determined using ANOVA for continuous factors.
The distribution of PFS/iPFS and OS by cannabinoid use groups was assessed using the Kaplan-Meier method, with an estimate of median survival and its 95% confidence interval (CI). The Cochran – Mantel – Haenszel test stratified by trial and treatment arm was used to assess the association of cannabinoid use with baseline factors, as well as BOR/iBOR and irAEs with an estimate of odds ratio (OR) and its 95% CI. Multivariable analysis was completed using the Cox proportional hazards model/logistic regression model stratified by trials and treatment arms, and was adjusted for the following factors: age, sex, smoking status, tumor type, sites of metastases (e.g., brain and liver), ECOG performance status, comorbidity index, body mass index (BMI), presence/absence of combined modality treatment (i.e., ICI + chemotherapy), number of prior systemic therapies, presence of other medications known to affect ICI response, and the number of concomitant medications. The minimum Akaike information criterion (AIC) was used for model selection.
The statistically significant p-value target was p ≤ 0.05. All statistical analyses were performed using SAS v9.4 software.
3. Results
3.1. Baseline patient characteristics by trial
The analysis included 684 patients in total. Baseline patient characteristics by trial are presented in Supplementary Table S3. A total of 65 patients took cannabinoids at any time during trial participation, 32 of whom were taking them at baseline. The majority of patients taking cannabinoids at baseline were from the BR.34 (43.8%; n = 14 NSCLC patients) and PA.7 (43.8%; n = 14 PDAC patients) trials (Supplementary Table S3). No other trial-specific differences were observed.
3.2. Relationship between patient characteristics and disease factors and cannabinoid use
Table 1 describes the relationship between baseline and any time cannabinoid use status with patient characteristics and disease factors. There was a statistically significant difference in the use/nonuse of cannabinoids for the following baseline factors by univariate analysis: participation by clinical trial (p = 0.0002), age (p = 0.03), smoking status (p = 0.03) and number of prior therapies (p = 0.05). Patients were more likely to be taking cannabinoids if they were <65 years of age, had ever smoked cigarettes or had only one prior line of therapy. When data was controlled for baseline confounders, multivariate analysis revealed that only age <65 years (p = 0.01) was significantly associated with cannabinoid use (Table 2).
Table 1.
Factors associated with baseline and any time use of cannabinoids by univariate analysis.
| Cannabinoid use at baseline |
Cannabinoid use at any time during participation on the trial |
|||||
|---|---|---|---|---|---|---|
| Factor | No N (%) |
Yes N (%) |
p-values | No N (%) |
Yes N (%) |
p-values |
| Trials | 0.0002a | <0.0001a | ||||
| BR.34 (297) | 283 (41.4) | 14 (2.0) | 270 (39.5) | 27 (3.9) | ||
| CO.26 (118) | 115 (16.8) | 3 (0.4) | 110 (16.1) | 8 (1.2) | ||
| IND.226 (150) | 149 (21.8) | 1 (0.1) | 146 (21.3) | 4 (0.6) | ||
| PA.7 (119) | 105 (15.3) | 14 (2.0) | 93 (13.6) | 26 (3.8) | ||
| Total: 684 | 652 (95.3) | 32 (4.7) | 619 (90.5) | 65 (9.5) | ||
| Age | 0.03a | 0.12 | ||||
| <65 | 335 (49.0) | 23 (3.4) | 318 (46.5) | 40 (5.8) | ||
| ≥65 | 317 (46.3) | 9 (1.3) | 301 (44.0) | 25 (3.7) | ||
| Sex | 0.35 | 0.22 | ||||
| Female | 292 (42.7) | 17 (2.5) | 275 (40.2) | 34 (5.0) | ||
| Male | 360 (52.6) | 15 (2.2) | 344 (50.3) | 31 (4.5) | ||
| Smoking Status | 0.03a | 0.01a | ||||
| Never | 59 (8.6) | 0 | 58 (8.5) | 1 (0.1) | ||
| Ever | 373 (54.5) | 15 (2.2) | 358 (52.3) | 30 (4.4) | ||
| Unknown | 220 (32.2) | 17 (2.5) | 203 (29.7) | 34 (5.0) | ||
| ECOG PS | 0.10 | 0.64 | ||||
| 0 | 191 (27.9) | 5 (0.7) | 179 (26.2) | 17 (2.5) | ||
| 1 | 461 (67.4) | 27 (3.9) | 440 (64.3) | 48 (7.0) | ||
| BMI (kg/m2) | 0.06 | 0.57 | ||||
| <18.5 | 26 (3.8) | 4 (0.6) | 26 (3.8) | 4 (0.6) | ||
| 18.5 to 24.9 | 272 (39.8) | 13 (1.9) | 260 (38.0) | 25 (3.7) | ||
| 25 to 29.9 | 212 (31.0) | 12 (1.8) | 199 (29.1) | 25 (3.7) | ||
| ≥30 | 142 (20.8) | 3 (0.4) | 134 (19.6) | 11 (1.6) | ||
| No. of disease sitesb | 0.90 | 0.57 | ||||
| 1 | 95 (13.9) | 4 (0.6) | 91 (13.3) | 8 (1.2) | ||
| 2 | 202 (29.5) | 11 (1.6) | 189 (27.6) | 24 (3.5) | ||
| 3 | 163 (23.8) | 8 (1.2) | 156 (22.8) | 15 (2.2) | ||
| 4 | 102 (14.9) | 3 (0.4) | 99 (14.5) | 6 (0.9) | ||
| ≥5 | 75 (11.0) | 5 (0.7) | 70 (10.2) | 10 (1.5) | ||
| No. concurrent medications | 0.50 | 0.03a | ||||
| ≤1st tertile | 314 (45.9) | 12 (1.8) | 305 (44.6) | 21 (3.1) | ||
| 1st to 2nd tertile | 118 (17.3) | 7 (1.0) | 109 (15.9) | 16 (2.3) | ||
| ≥2nd tertile | 220 (32.2) | 13 (1.9) | 205 (30.0) | 28 (4.1) | ||
| No. comorbidities | 0.40 | 0.07 | ||||
| ≤3 | 254 (37.1) | 16 (2.3) | 238 (34.8) | 32 (4.7) | ||
| 4, 5 | 220 (32.2) | 10 (1.5) | 207 (30.3) | 23 (3.4) | ||
| ≥6 | 178 (26.0) | 6 (0.9) | 174 (25.4) | 10 (1.5) | ||
| Combined modality tx | 0.67 | 0.65 | ||||
| No | 269 (39.3) | 12 (1.8) | 256 (37.4) | 25 (3.7) | ||
| Yes | 383 (56.0) | 20 (2.9) | 363 (53.1) | 40 (5.8) | ||
| Histology | 0.07 | 0.04a | ||||
| ADE | 186 (27.2) | 3 (0.4) | 178 (26.0) | 11 (1.6) | ||
| DUC | 2 (0.3) | 0 | 2 (0.3) | 0 | ||
| EPI | 3 (0.4) | 0 | 3 (0.4) | 0 | ||
| MXB | 2 (0.3) | 0 | 2 (0.3) | 0 | ||
| NOS | 3 (0.4) | 0 | 3 (0.4) | 0 | ||
| NSQ | 234 (34.2) | 10 (1.5) | 224 (32.7) | 20 (2.9) | ||
| OTH | 18 (2.6) | 0 | 18 (2.6) | 0 | ||
| SCD | 2 (0.3) | 0 | 2 (0.3) | 0 | ||
| SMA | 19 (2.8) | 0 | 19 (2.8) | 0 8 (1.2) |
||
| SQU | 71 (10.4) | 5 (0.7) | 68 (9.9) | |||
| TRA | 1 (0.1) | 0 | 1 (0.1) | 0 | ||
| UN | 105 (15.4) | 14 (2.0) | 93 (13.6) | 26 (3.8) | ||
| UND | 3 (0.4) | 0 | 3 (0.4) | 0 | ||
| UUU | 3. (0.4) | 0 | 3. (0.4) | 0 | ||
| No. prior therapies | 0.05a | 0.07 | ||||
| 0c | 450 (65.8) | 24 (3.5) | 424 (62.0) | 50 (7.3) | ||
| 1 | 15 (2.2) | 3 (0.4) | 14 (2.0) | 4 (0.6) | ||
| 2 | 36 (5.3) | 1 (0.1) | 35 (5.1) | 2 (0.3) | ||
| 3+ | 151 (22.1) | 4 (0.6) | 146 (21.3) | 9 (1.3) | ||
| TNM Stage | 0.79 | 0.28 | ||||
| IVA | 96 (14.0) | 6 (0.9) | 89 (13.0) | 13 (1.9) | ||
| IVB | 187 (27.3) | 8 (1.2) | 181 (26.5) | 14 (2.0) | ||
| Unknownd | 369 (53.9) | 18 (2.6) | 349 (51.0) | 38 (5.6) | ||
| Corticosteroid use | 0.26 | 0.42 | ||||
| No | 608 (88.9) | 32 (4.7) | 577 (84.4) | 63 (9.2) | ||
| Yes | 44 (6.4) | 0 | 42 (6.1) | 2 (0.3) | ||
| PPI use | 0.24 | 0.53 | ||||
| No | 581 (84.9) | 31 (4.5) | 552 (80.7) | 60 (8.8) | ||
| Yes | 71 (10.4) | 1 (0.1) | 67 (9.8) | 5 (0.7) | ||
| Antibiotic use | 1.00 | 0.60 | ||||
| No | 643 (94.0) | 32 (4.7) | 611 (89.3) | 64 (9.4) | ||
| Yes | 9 (1.3) | 0 | 8 (1.2) | 1 (0.1) | ||
| ß-blocker use | 0.39 | 0.35 | ||||
| No | 621 (90.8) | 32 (4.7) | 589 (86.1) | 64 (9.4) | ||
| Yes | 31 (4.5) | 0 | 30 (4.4) | 1 (0.1) | ||
| Aspirin use | 0.72 | 0.72 | ||||
| No | 606 (88.6) | 31 (4.5) | 573 (83.8) | 64 (9.4) | ||
| Yes | 46 (6.7) | 1 (0.1) | 46 (6.7) | 1 (0.1) | ||
| Statin use | 1.00 | 1.00 | ||||
| No | 595 (87.0) | 30 (4.4) | 564 (82.5) | 61 (8.9) | ||
| Yes | 57 (8.3) | 2 (0.3) | 55 (8.0) | 4 (0.6) | ||
| Serotonergic drug use | 0.71 | 0.79 | ||||
| No | 614 (89.8) | 30 (4.4) | 583 (85.2) | 61 (8.9) | ||
| Yes | 38 (5.6) | 2 (0.3) | 36 (5.3) | 4 (0.6) | ||
There was a statistically significant difference in the baseline use/nonuse of cannabinoids for the following factors by univariate analysis: participation by clinical trial (p = 0.0002), age (p = 0.03), smoking status (p = 0.03) and number of prior therapies (p = 0.05). For use of cannabinoids at any time, there was a statistically significant difference in the following baseline factors by univariate analysis: participation by clinical trial (p < 0.0001), smoking status (p = 0.01), number of concomitant medications (continuous variable) (p = 0.03) and histology (p = 0.04).
aDenotes statistical significance.
bSixteen patients did not have data on the number of disease sites.
cPatients from the BR.34 and PA.7 trials did not receive previous chemotherapy in the metastatic setting, but could have had (neo)adjuvant chemotherapy or prior RT therapy in the non-metastatic setting.
dOnly patients from BR.34 collected data regarding TNM stage. Patients from IND.226 had to have histologically confirmed advanced, metastatic, recurrent or unresectable cancer for which there were no curative therapy options. Patients from CO.26 had locally advanced or metastatic histologically confirmed adenocarcinoma of the colon or rectum that was unresectable and had received all available standard systemic therapies. Patients from PA.7 had to have histologically or cytologically confirmed metastatic pancreatic ductal adenocarcinoma.
ADE- adenocarcinoma, DUC- infiltrating ductal (breast), EPI- epithelial sarcoma (mesothelioma), MXB- mixed/biphasic (mesothelioma), NOS- not otherwise specific (NSCLC NOS x2, mesothelioma NOS x1), NSQ- non-squamous (NSCLC), OTH- other, SCD- sarcomatoid (mesothelioma), SMA- SCLC, SQU- squamous, TRA- transitional (bladder), UN- unknown (pancreatic ductal adenocarcinoma- all of PA.7), UND- undifferentiated (bladder x1, uterine sarcoma x1, NSCLC x1), UUU – unknown (thymic unknown/unspecified histology x1, SCLC unknown/unspecified histology x1, solid pseudopapillary tumor of the pancreas unknown/unspecified histology x1). BMI- body mass index. ECOG PS- Eastern Cooperative Oncology Group Performance Status. No.- number. PPI- proton pump inhibitor. TNM- tumor, node, metastasis. Tx- treatment.
Table 2.
Statistically significant factors associated with baseline and any time use of cannabinoids by multivariate analysis.
| Baseline Cannabinoid Use |
Anytime Cannabinoid Use |
||||
|---|---|---|---|---|---|
| Factor | Effect (OR, 95% CI) | p-value | Factor | Effect (OR, 95% CI) | p-value |
| Age | No. comorbidities | 0.67 (0.47, 0.96) | 0.03a | ||
| <65 | 0.35 (0.16, 0.78) | 0.01a | |||
| ≥65 | 1 | ||||
| No. Concurrent medications | 1.39 (1.01, 1.90) | 0.04a | |||
After controlling for baseline confounders, multivariate analysis revealed that only age <65 years (p = 0.01) was significantly associated with baseline cannabinoid use, while having more comorbidities (p = 0.03) and taking more concomitant medications (p = 0.04) were significantly associated with cannabinoid use at any time.
aDenotes statistical significance.
CI- confidence interval. No.- number. OR- odds ratio.
Of the 65 patients reporting cannabinoid use at any time on trial, 33 patients started them after randomization (Supplementary Table S4). The majority of patients taking cannabinoids at any time participated in the BR.34 (n = 27, 41.5%) and PA.7 trials (n = 26, 40.0%) (Table 1, Supplementary Table S4). There was a statistically significant difference in the use of cannabinoids at any time for the following baseline factors by univariate analysis: participation by clinical trial (p < 0.0001), smoking status (p = 0.01), number of concomitant medications (continuous variable) (p = 0.03) and histology (p = 0.04) (Table 1). Patients were more likely to be taking cannabinoids if they had ever smoked cigarettes, were taking more concomitant medications (continuous variable) and had the following histologies: adenocarcinoma, PDAC, and non-squamous NSCLC. Multivariate analysis revealed that having more comorbidities (p = 0.03) and taking more concomitant medications (p = 0.04) were significantly associated with cannabinoid use (Table 2).
3.3. Baseline cannabinoid use and best overall response
There was no statistically significant difference between responders versus those with PD based on cannabinoid use at baseline by univariate (p = 0.09) (Supplementary Table S5) or multivariate analysis (p = 0.15) (Table 3). However, multivariate analysis showed that patients using proton pump inhibitors (PPIs) or having fewer comorbidities were less likely to respond (p = 0.03 and 0.05 respectively) (Table 3).
Table 3.
Factors associated with a BOR or iBOR of CR/PR/SD or iCR/iPR/iSD respectively, adjusted for cannabinoid use at baseline.
| BOR |
iBOR |
||||
|---|---|---|---|---|---|
| Factor | Effect (OR, 95% CI) | p-value | Factor | Effect (OR, 95% CI) | p-value |
| Baseline cannabinoid use | Baseline cannabinoid use | ||||
| Yes | 1.90 (0.79, 4.56) | 0.15 | Yes | 1.99 (0.56, 7.10) | 0.29 |
| No | 1 | No | 1 | ||
| PPI use | Corticosteroid use | ||||
| Yes | 0.54 (0.31, 0.93) | 0.03a | Yes | 0.14 (0.03, 0.68) | 0.02a |
| No | 1 | No | |||
| No. comorbidities | 1.24 (1.00, 1.54) | 0.05a | |||
There was no statistically significant difference between responders versus those with progressive disease based on baseline cannabinoid use by multivariate analysis. However, multivariate analysis showed that patients using proton pump inhibitors or having fewer comorbidities were less likely to respond based on BOR measured by RECIST 1.1; whereas, patients using corticosteroids were less likely to respond to treatment by iBOR.
aDenotes statistical significance.
BOR/CR/PR/SD- best overall response/complete response/partial response/stable disease as measured by RECIST 1.1. iBOR/iCR/iPR/iSD- the same outcomes as measured by iRECIST. CI- confidence interval. No.- number. OR- odds ratio. PPI- proton pump inhibitor.
There was no statistically significant difference between responders versus those with iPD in patients using or not using cannabinoids at baseline by univariate (p = 0.15) (Supplementary Table S4) or multivariate analysis (p = 0.29) (Table 3). Multivariate analysis demonstrated that patients using corticosteroids (p = 0.02) were less likely to respond to treatment (Table 3).
3.4. Baseline cannabinoid use and effect on PFS, iPFS and OS
Univariate analysis showed a statistically significant benefit in PFS for patients taking cannabinoids at baseline versus non-users (HR 0.66, 90% CI 0.47–0.93, p = 0.046) (Figure 1). Median PFS was 6.97 months versus 3.91 months, respectively (90% CI 3.58–9.20 months and 3.48–4.44 months). However, Cox proportional hazards analysis did not confirm significance when controlling for baseline confounders (p = 0.12). Instead, female sex (p = 0.03), fewer disease sites (p < 0.001) and not taking antibiotics (p = 0.02) were significantly associated with a longer PFS (Table 4).
Figure 1.
Baseline cannabinoid use significantly improves progression free survival (PFS). PFS using RECIST 1.1 criteria was evaluated based on baseline cannabinoid use and was stratified by clinical trial. Data were plotted on a Kaplan–Meier curve.
Table 4.
PFS/iPFS and OS adjusted for cannabinoid use at baseline and at any time during participation in the clinical trial (Cox’s proportional hazards analysis).
| PFS Baseline use |
PFS Any time use |
|||
|---|---|---|---|---|
| Factor | Effect (OR, 95% CI) |
p-value | Effect (OR, 95% CI) |
P-value |
| Cannabinoid use | ||||
| Yes | 0.72 (0.48, 1.08) | 0.12 | 0.74 (0.52, 1.04) | 0.08 |
| No | 1 | 1 | ||
| Sex | ||||
| Male | 1.20 (1.01, 1.42) | 0.03a | 1.19 (1.01, 1.41) | 0.04a |
| Female | 1 | 1 | ||
| No. of disease sites | 1.19 (1.12, 1.26) | <0.001a | 1.19 (1.12, 1.26) | <0.0001a |
| Antibiotic use | ||||
| Yes | 2.34 (1.14, 4.79) | 0.02a | 2.33 (1.14, 4.79) | 0.02a |
| No |
1 |
|
1 |
|
| |
iPFS Baseline use |
iPFS Any time use |
||
| Cannabinoid use | ||||
| Yes | 0.60 (0.36, 1.00) | 0.05a | 0.70 (0.47, 1.03) | 0.07 |
| No | 1 | 1 | ||
| Sex | ||||
| Male | 1.28 (1.06, 1.55) | 0.01a | 1.28 (1.06 – 1.55) | 0.01a |
| Female | 1 | 1 | ||
| No. of disease sites | 1.23 (1.15, 1.32) | <0.0001a | 1.24 (1.15, 1.32) | <0.0001a |
| Antibiotic use | ||||
| Yes | 2.22 (1.04, 4.76) | 0.04a | 2.23 (1.04, 4.78) | 0.04a |
| No | 1 | |||
| ECOG PS | ||||
| 1 | 1.33 (1.08, 1.65) | 0.008a | 1.31 (1.06, 1.63) | 0.01a |
| 0 |
1 |
|
1 |
|
| |
OS Baseline use |
OS Any time use |
||
| Cannabinoid use | 0.81 (0.51, 1.27) | 0.90 (0.63, 1.28) | ||
| Yes | 1 | 0.35 | 1 | 0.56 |
| No | ||||
| Sex | ||||
| Male | 1.26 (1.04, 1.52) | 0.02a | 1.26 (1.05, 1.52) | 0.01a |
| Female | 1 | 1 | ||
| No. of disease sites | 1.26 (1.16, 1.35) | <0.0001a | 1.27 (1.19, 1.35) | <0.0001a |
| ECOG | ||||
| 1 | 1.59 (1.28, 1.97) | <0.0001a | 1.58 (1.27, 1.96) | <0.0001a |
| 0 | 1 | 1 | ||
There was no significant difference in PFS or OS based on baseline or any time cannabinoid use when controlling for baseline confounders. Instead, female sex, fewer disease sites and not taking antibiotics were significantly associated with a longer PFS, while female sex, ECOG 0 and fewer disease sites were significantly associated with a longer OS. There was a statistically significant difference in iPFS for patients taking versus not taking cannabinoids at baseline (p = 0.05), and female sex, ECOG performance status 0, fewer disease sites and not taking antibiotics were also significantly associated with a longer iPFS.
aDenotes statistical significance.
CI- confidence interval. ECOG PS- Eastern Cooperative Oncology Group Performance Status. No.- number. OR- odds ratio. OS- overall survival. PFS- progression free survival where progression was determined using RECIST 1.1. iPFS- PFS where progression was determined using iRECIST.
There was a statistically significant difference in iPFS for patients taking versus not taking cannabinoids at baseline (HR 0.55, 90% CI 0.36–0.83, p = 0.017) (Figure 2). Median PFS was 10.91 months versus 8.31 months, respectively (90% CI 6.97–25.10 months and 7.23–9.26 months). Significance remained with Cox proportional hazards analysis controlling for baseline confounders (p = 0.05). Female sex (p = 0.01), ECOG performance status 0 (p = 0.008), fewer disease sites (p < 0.0001) and not taking antibiotics (p = 0.04) were also significantly associated with a longer iPFS (Table 4).
Figure 2.
Baseline cannabinoid use significantly improves progression free survival (iPFS) where immune responses were assessed using iRECIST. iPFS was evaluated based on cannabinoid use and was stratified by clinical trial. Data were plotted on a Kaplan–Meier curve.
There was no statistically significant difference in OS for patients taking or not taking cannabinoids at baseline (HR 0.75, 90% CI 0.51–1.10, p = 0.21) (Figure 3), consistent with Cox proportional hazards analysis (p = 0.35). Median OS was 14.23 months versus 11.30 months, respectively (90% CI 6.41–19.38 months and 10.18–12.65 months). Cox proportional hazards analysis demonstrated that female sex (p = 0.02), ECOG 0 (p < 0.0001) and fewer disease sites (p < 0.0001) were significantly associated with a longer OS (Table 4).
Figure 3.
Baseline cannabinoid use does not affect overall survival (OS). OS was evaluated based on cannabinoid use and was stratified by clinical trial. Data were plotted on a Kaplan–Meier curve.
3.5. Cannabinoid use at any time and effect on PFS, iPFS and OS
The any time use of cannabinoids was not significantly associated with PFS, iPFS or OS (p = 0.08, p = 0.07, and p = 0.56, respectively) (Table 4). Cox proportional hazards analysis demonstrated that female sex (p = 0.04), having fewer disease sites (p < 0.0001) and not using antibiotics (p = 0.02) were significantly associated with a longer PFS (Table 4) whereas, female sex (p = 0.01), ECOG 0 (p = 0.01), fewer disease sites (p < 0.0001) and not taking antibiotics (p = 0.04) were significantly associated with longer iPFS (Table 4).
Cox proportional hazards analysis demonstrated that female sex (p = 0.01), fewer disease sites (p < 0.0001), and ECOG 0 (p < 0.0001) were significantly associated with longer OS (Table 4).
3.6. Baseline cannabinoid use and incidence of irAEs
The observed patterns of irAEs in patients receiving cannabinoids at any time during treatment with combination immunotherapy were consistent with the known pattern for these agents, with no new irAE signals. There was no impact of cannabinoid use at baseline on the incidence of irAEs by univariate analysis (p = 0.71) (Supplementary Table S6) or multivariate analysis (grade 1/2 irAEs, p = 0.96; grade 3/4, p = 0.65) (Table 5). However, antibiotic use was significantly associated with increased rates of grade 3/4 irAEs (p = 0.01) on multivariate analysis (Table 5).
Table 5.
Incidence of irAE adjusted for cannabinoid drug use at baseline.
| Factor | irAE grade (vs none) | Odds Ratio | 95% CI | p-value |
|---|---|---|---|---|
| Cannabinoid use | 3 or 4 | |||
| Yes | 0.71 | 0.16 – 3.10 | 0.65 | |
| No | 1 | |||
| 1 or 2 | ||||
| Yes | <0.001 | <0.001 – >999 | 0.96 | |
| No | 1 | |||
| Antibiotic use | 3 or 4 | |||
| Yes | 6.75 | 1.50 – 30.36 | 0.01a | |
| No | 1 | |||
| 1 or 2 | ||||
| Yes | <0.001 | <0.001 – >999 | 0.99 | |
| No | 1 |
There was no impact of cannabinoid use at baseline on the incidence of irAes; however, antibiotic use was significantly associated with increased rates of grade 3/4 irAEs (p = 0.01) on multivariate analysis.
aDenotes statistical significance.
CI- confidence interval.
3.7. Baseline cannabinoid use and QOL
Cannabinoid use at baseline appeared to be associated with improved EORTC QLQ-C30 scores in the following domains: Emotional, Cognitive, Sleep and Appetite Loss (Supplementary Table S8). Importantly, there was no significant decline in any of the functional domains or symptom items in patients who had baseline cannabinoid use.
3.8. Cannabinoid use in BR.34 patients
Any effects of cannabinoid use on outcomes in the overall study population may be obfuscated by data from patients with tumor histologies that are classically poorly responsive to ICI including PDAC (17% of patients). BR.34 patients exclusively had NSCLC, which classically responds to ICI. BR.34 patients comprised 43.8% of patients taking cannabinoids at baseline (vs 43.8% PDAC) and 41.5% at any time (vs 40.0% PDAC). A separate analysis of the BR.34 trial was done with similar results to the overall study population, and the use of cannabinoids was not associated with response, PFS/iPFS or OS in this subset of patients (Supplementary Table S7).
4. Discussion
This pooled analysis of individual patient data from four CCTG clinical trials (BR.34, CO.26, PA.7 and IND.226) evaluated the effect of cannabinoid use on clinical outcomes in patients treated with durvalumab and tremelimumab (with/without chemotherapy). On multivariate analysis, baseline use of cannabinoids was not associated with differences in treatment response (BOR/iBOR). Although cannabinoid use at baseline was associated with a trend toward longer PFS and iPFS, there was no significant impact on OS. Furthermore, any time cannabinoid use was not associated with any differences in PFS, iPFS or OS. Finally, cannabinoid use was not associated with any differences in the incidence of any grade irAEs, and baseline cannabinoid use was associated with improved EORTC QLQ-C30 scores in the Emotional, Cognitive, Sleep and Appetite Loss domains.
In contrast to the study hypothesis, data showed a significant association of cannabinoid use with better iPFS using Cox proportional hazards analysis, contrary to previously published preclinical data and observational studies [28–30]. This may be due to multiple factors. One possibility is that median PFS may underestimate the activity of ICIs in patients with prolonged stable disease as tumor control may develop more slowly over time [41]. Alternatively, while the majority of patients in previously described observational studies received ICI monotherapy, this pooled analysis included patients exposed to dual ICI exclusively [28–30]. Therefore, the anti-inflammatory effects of cannabinoids may be less clinically relevant in patients taking dual ICI therapy. As well, differences in pharmacology among cannabinoids may explain variations in the reported impacts of cannabinoids on ICI efficacy. Nabilone, a synthetic THC analog, is most likely to bind with high affinity to CB1 receptors [42]. Cannabis oil formulations, which contain CBD and THC in various ratios, will bind CB1 as well as CB2 receptors [42]. CB2 receptors are present at higher levels in immune tissues and several previous in vitro studies have shown that CB2 receptors cause mainly anti-inflammatory effects when bound by AEA (similar to THC), and pro-inflammatory effects when bound by 2-AG (similar to CBD) [17,43]. Despite different mechanisms of action, one study in which a majority of patients taking synthetic THC alone and two in which patients were taking cannabis alone (THC + CBD), have published similar results where cannabinoid use was associated with reduced response rates or OS [28–30]. This highlights the need for further studies to better understand the mechanism by which cannabinoids might impact ICI signaling on T-cells (or vice versa). Finally, it is possible that the route of ingestion (e.g., inhaled, ingested) could affect absorption, distribution and therefore effects on the immune system; however, data is currently limited and further studies are needed.
There has been recent interest in nociceptor neurons, which can affect cancer immunosurveillance by producing neuropeptide signaling molecules that contribute to exhaustion of cytotoxic CD8+ T-cells [44]. As cannabinoids are often used for cancer-related pain control, they may indirectly potentiate the ability of the immune system to eliminate cancer cells by decreasing the release of neuropeptides that contribute to CD8+ T-cell exhaustion. Cannabinoids also affect other endogenous signaling pathways that interact with the immune system (e.g., serotonergic, opioid). It is possible that effects of cannabinoids on these pathways, or medications commonly taken by cancer patients targeting these pathways (e.g., SSRIs, opioids) are at play [45–47]. To address this, concomitant medications taken by patients that affect the immune system were controlled for in the analyses (Supplementary Table S2). Of these, only PPIs, antibiotics and corticosteroids were independently associated with outcomes, though none with differences in OS. Given the complex interactions between cannabinoids and several immunomodulatory pathways, further research is needed to better understand the impact of various types of cannabinoids (e.g., THC vs CBD) on the immune system.
With respect to ICI toxicities, cannabinoid use was not significantly associated with lower rates of irAEs, as hypothesized. However, antibiotic use at baseline was associated with a significantly higher incidence of grade 3/4 irAEs, counter to previously published studies citing no effect [48,49]. One explanation may be that patients presenting with symptoms related to their cancer (e.g., SOB/cough/hypoxia) are often treated for multiple differential diagnoses including infection, until an irAE, progression or other diagnosis is confirmed and the antibiotic is discontinued. Interpretation of this data is limited by the small number of patients using antibiotics at baseline (1.3%).
It is important to note that a recent report by Piper et al. highlighted concerns regarding inaccuracy of data reporting and statistical analyses in two published studies that reported a lower response rate [28] and shorter time to progression and overall survival [30] when cannabis was used concomitantly with immune checkpoint inhibitors [50]. It was suggested that important patient characteristics (e.g., burden of metastatic disease or smoking status) might not have been accurately addressed and could have played a role in the reported differences in outcomes [50]. Waissengrin et al. reported a numerically lower OS (although not a statistically significant difference) with concomitant ICI and cannabis use; however, it was felt this could have been related to higher rates of brain and liver metastases in this group [31]. In the current study, multiple known and potential confounders were controlled for by way of multivariate analyses. The prognostic role of previously well-described factors on OS including ECOG performance status and the number of disease sites was affirmed [51–53]. The only other patient-related factor significantly associated with outcomes (poorer PFS/iPFS/OS) in this pooled analysis was patient sex, which may reflect sex-based differences in patients’ innate and adaptive immune responses in the context of ICI therapy [54]. Therefore, well-known prognostic factors may be more clinically relevant than the presence or absence of concomitant medications in informing patient prognosis.
This study has limitations. Firstly, due to the retrospective nature of the analysis, details on the CBD/THC ratio, frequency, duration, amount and route of administration of the cannabinoid agents used were not available across all trials. Data on the route and dose was collected for the IND.226 trial but only four patients from this trial had documented cannabinoid use (Supplementary Table S4); therefore, a meaningful analysis on the impact of route of administration or dosage could not be completed. This is a common limitation reported in other published retrospective analyses evaluating the impact of cannabis on ICI outcomes as well [31]. Secondly, results are further limited by the small number of cannabinoid users, which was lower than expected as compared to national surveys evaluating cannabis use among Canadian adults with cancer published at the same time of accrual to trials included in this pooled analysis (i.e., 34% in 2015–2016 and 18% in 2017, with accrual to trials included in this pooled analysis beginning in 2015–2017) [5,55]. Possible explanations include: i) data reported here is exclusively from patients participating in clinical trials who tend to have a better ECOG performance status due to eligibility requirements, and therefore may have fewer symptoms for which cannabinoid use is sought; ii) patients may have been less likely to report cannabinoid use if it was not purchased commercially; iii) patients were not likely to have been systematically asked if they were taking cannabinoids, therefore report of their use relied on patient disclosure. Furthermore, underreporting of cannabinoid use may have occurred overall as trial enrollment occurred mainly before legalization. Thirdly, by separating patients into baseline and any time users, we assume i) those with baseline cannabinoid use continued to use them after ICI start, and ii) any time cannabinoid users started them while on immunotherapy and far enough prior to disease progression/treatment stoppage to impact the measured outcomes. Making these assumptions could infer a relationship that is not real or mask a relationship that is. Finally, as the bulk of patients taking cannabinoids at any time had either NSCLC (41.5%) or PDAC (40.0%), this potentially limits the generalizability of these findings to all cancer types.
5. Conclusions
In this pooled analysis of individual patient data, improved iPFS with cannabinoid use in patients treated with durvalumab plus tremelimumab with/without chemotherapy did not translate into OS benefits. Despite the role of cannabinoids in modifying the adaptive immune response via inhibition of lymphocyte activation, proliferation and cytokine release, the use of cannabinoids in the context of combination ICI appears safe, however given the limitations discussed above, further study is warranted. Future studies could focus on patients with more classically immunogenic solid tumors, patients treated with single agent ICIs, or investigate cannabinoids which act only on CB2 receptors, so as to separate direct versus indirect effects of cannabinoids on the immune system.
Supplementary Material
Acknowledgments
The authors gratefully acknowledge the patients and their family members, as well as all investigators and site staff who participated in and contributed to the four included clinical trials. The authors would also like to acknowledge the contribution of administrative, technical and material support by the Central Offices of the Canadian Cancer Trials Group and Xiaoqun Sun, SAS programmer.
Funding Statement
This study, and the trials this study was based on, were supported by a grant to the Canadian Cancer Trials Group from the Canadian Cancer Society Research Institute (CCSRI; Grant #707213). AstraZeneca provided drug and funding support for the four clinical trials included in this pooled analysis. CCSRI and AstraZeneca did not contribute to the design and conduct of this study; the collection, management, analysis or interpretation of the data; preparation, review or approval of the manuscript; nor the decision to submit the manuscript for publication.
Article highlights
Immune checkpoint inhibitors (ICIs) are used in the treatment of various cancer types. ICIs block cancer and T-cell interactions. However, cannabinoids might inhibit lymphocyte activation, reduce immune cell-mediated lysis of tumor cells, have an anti-proliferative effect on T-cells, induce T-cell apoptosis, inhibit inflammatory cytokines and promote production of anti-inflammatory cytokines, all of which are thought to reduce the efficacy of ICIs. Interactions between clinical cannabinoid use and dual ICI treatment remain unknown.
Individual patient data from 4 Canadian Cancer Trials Group (CCTG) trials of patients treated with dual ICI (durvalumab plus tremelimumab) ± chemotherapy were pooled to assess Best Overall Response (BOR)/iBOR per RECIST 1.1/iRECIST, Progression-Free Survival (PFS)/iPFS, Overall Survival (OS) and immune-related adverse events (irAEs) in cannabinoid users vs. non-users.
Improved iPFS with cannabinoid use in patients treated with durvalumab plus tremelimumab with/without chemotherapy did not translate into OS benefits. There was no significant difference in BOR, PFS, OS or incidence of grade 1/2 or 3/4 irAEs in patients with cannabinoid use at baseline. Other well-known prognostic factors such as performance status, sex and disease burden were the only factors significantly associated with OS.
Although overall numbers of cannabinoid users were small and future larger studies are warranted, this pooled analysis supports the safe use of cannabinoids in the context of combination ICI therapy.
Author contributions
Dr. Gaudreau, Dr. Fung, Dr. Coschi and Dr. Ding had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Dr. Gaudreau and Dr. Fung.
Acquisition and analysis of data: Dr. Gaudreau, Dr. Fung, Dr. Coschi and Dr. Ding.
Statistical analysis: Dr. Ding.
Interpretation of data: All authors.
Drafting of the manuscript: Dr. Coschi, Dr. Tong, Dr. Gaudreau and Dr. Fung.
Critical revision of the manuscript for important intellectual content: All authors.
Obtained funding: Dr. Gaudreau, Dr. Ding, Dr. Tu, Dr. O’Collaghan, Dr. Seymour.
Administrative, technical, or material support: Central Offices of the Canadian Cancer Trials Group and Xiaoqun Sun, SAS programmer.
Supervision: Dr. Gaudreau and Dr. Fung.
Disclosures
RAJ is on the advisory board to Amgen, AstraZeneca, Bayer, Bristol Myers Squibb, EMD Serono, Jazz Pharmaceuticals, Johnson and Johnson, Lilly, Merck Sharp and Dohme, Novartis, Pfizer, Roche, Sanofi and Takeda. RAJ has received honoraria from Amgen, AstraZeneca, Bristol Myers Squibb, Jazz Pharmaceuticals, Merck Sharp and Dohme, Novartis and Takeda. RAJ has completed contracted research for Amgen, Astellas, AstraZeneca, Bold Therapeutics, Bristol Myers Squibb, Debio Pharmaceuticals, Fusion Pharmaceuticals, Johnson and Johnson, Macrogenics and Merck Sharp and Dohme. RAJ is a consultant for Merck Sharpe and Dohme. EC has received honoraria from Bayer, Eisai, GSK, Merck, AZ, Taiho, Roche, Ipsen and Amgen. EC has participated in clinical trials sponsored by Roche, Merck, AZ, Novartis, Zymeworks, BMS, Mirati, 1Global, Jansen, Repare, Pfizer and GSK. DJR reports grants and personal fees from Roche and Bayer outside the submitted work. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.
Data sharing statement
Clinical trial data can be requested through the Canadian Cancer Trials Group in accordance with its data sharing policy at https://www.ctg.queensu.ca/public/policies. This original research is the secondary analysis of clinical trial data that have been shared with the authors.
Supplementary material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/1750743X.2025.2485012.
References
The following references are of considerable interest given that each of these trials were included in this pooled analysis:
- 1.Worster B, Hajjar ER, Handley N.. Cannabis use in patients with cancer: a clinical review. JCO Oncol Pract. 2022;18(11):743–749. doi: 10.1200/OP.22.00080 [DOI] [PubMed] [Google Scholar]
- 2.Taneja S, Guo Y, Slaven M, et al. The perceptions and beliefs of cannabis use among Canadian genitourinary cancer patients. Can Urol Assoc J. 2022;16(2):48–54. doi: 10.5489/cuaj.7197 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Bennett M, Paice JA, Wallace M. Pain and opioids in cancer care: benefits, risks, and alternatives. Am Soc Clin Oncol Educ Book. 2017;37(37):705–713. doi: 10.1200/EDBK_180469 [DOI] [PubMed] [Google Scholar]
- 4.MacDonald E, Farrah K. Medical cannabis use in palliative care: review of clinical effectiveness and guidelines – an update [internet]. Ottawa (ON): Canadian Agency for Drugs and Technologies in Health; 2019. Available from: https://www.ncbi.nlm.nih.gov/books/NBK551867/ [PubMed] [Google Scholar]
- 5.Martell K, Fairchild A, LeGerrier B, et al. Rates of cannabis use in patients with cancer. Curr Oncol. 2018;25(3):219–225. doi: 10.3747/co.25.3983 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hawley P, Gobbo M. Cannabis use in cancer: a survey of the current state at BC cancer before recreational legalization in Canada. Curr Oncol. 2019;26(4):425–432. doi: 10.3747/co.26.4743 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hawley P, Gobbo M, Afghari N. The impact of legalization of access to recreational cannabis on Canadian medical users with cancer. BMC Health Serv Res. 2020;20(1):977. doi: 10.1186/s12913-020-05756-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Khuja I, Yekhtin Z, Or R, et al. Cannabinoids reduce inflammation but inhibit lymphocyte recovery in murine models of bone marrow transplantation. Int J Mol Sci. 2019;20(3):668. doi: 10.3390/ijms20030668 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Borgelt LM, Franson KL, Nussbaum AM, et al. The pharmacologic and clinical effects of medical cannabis. Pharmacotherapy. 2013. Feb;33(2):195–209. doi: 10.1002/phar.1187 PMID: 23386598. [DOI] [PubMed] [Google Scholar]
- 10.Nahtigal I, Blake A, Hand A, et al. The pharmacological properties of cannabis. J Pain Manage. 2016;9(4):481–491. [Google Scholar]
- 11.Klein TW, Cabral GA. Cannabinoid-induced immune suppression and modulation of antigen-presenting cells. J Neuroimmune Pharmacol. 2006;1(1):50–64. doi: 10.1007/s11481-005-9007-x PMID: 18040791. [DOI] [PubMed] [Google Scholar]
- 12.Burnette-Curley D, Marciano-Cabral F, Fischer-Stenger K, et al. Delta-9-tetrahydrocannabinol inhibits cell contact-dependent cytotoxicity of Bacillus Calmette-Guerin-activated macrophages. Int J Immunopharmacol. 1993;15(3):371–382. doi: 10.1016/0192-0561(93)90048-4 [DOI] [PubMed] [Google Scholar]
- 13.Burnette-Curley D, Cabral GA. Differential inhibition of RAW264.7 macrophage tumoricidal activity by delta 9tetrahydrocannabinol. Proc Soc Exp Biol Med. 1995;210(1):64–76. doi: 10.3181/00379727-210-43926 [DOI] [PubMed] [Google Scholar]
- 14.McKallip RJ, Lombard C, Martin BR, et al. Delta(9)-tetrahydrocannabinol-induced apoptosis in the thymus and spleen as a mechanism of immunosuppression in vitro and in vivo. J Pharmacol Exp Ther. 2002;302(2):451–465. doi: 10.1124/jpet.102.033506 [DOI] [PubMed] [Google Scholar]
- 15.Xiong X, Chen S, Shen J, et al. Cannabis suppresses antitumor immunity by inhibiting JAK/STAT signaling in T cells through CNR2. Signal Transduct Target Ther. 2022;7(1):99. doi: 10.1038/s41392-022-00918-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Schwarz H, Blanco FJ, Lotz M. Anadamide, an endogenous cannabinoid receptor agonist inhibits lymphocyte proliferation and induces apoptosis. J Neuroimmunol. 1994;55(1):107–115. doi: 10.1016/0165-5728(94)90152-x [DOI] [PubMed] [Google Scholar]
- 17.Cencioni MT, Chiurchiù V, Catanzaro G, et al. Anandamide suppresses proliferation and cytokine release from primary human T-lymphocytes mainly via CB2 receptors. PLoS One. 2010;5(1):e8688. doi: 10.1371/journal.pone.0008688 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Maccarrone M, Lorenzon T, Bari M, et al. Anandamide induces apoptosis in human cells via vanilloid receptors. Evidence for a protective role of cannabinoid receptors. J Biol Chem. 2000;275(41):31938–31945. doi: 10.1074/jbc.M005722200 [DOI] [PubMed] [Google Scholar]
- 19.Chiurchiù V, Battistini L, Maccarrone M. Endocannabinoid signalling in innate and adaptive immunity. Immunology. 2015;144(3):352–364. doi: 10.1111/imm.12441 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Robert C, Schachter J, Long GV, et al. KEYNOTE-006 investigators. Pembrolizumab versus ipilimumab in advanced melanoma. N Engl J Med. 2015;372(26):2521–2532. doi: 10.1056/NEJMoa1503093 [DOI] [PubMed] [Google Scholar]
- 21.Reck M, Rodríguez-Abreu D, Robinson AG, et al. KEYNOTE-024 investigators. Pembrolizumab versus chemotherapy for PD-L1-Positive non-small-cell lung cancer. N Engl J Med. 2016;375(19):1823–1833. doi: 10.1056/NEJMoa1606774 [DOI] [PubMed] [Google Scholar]
- 22.Motzer RJ, Tannir NM, McDermott DF, et al. CheckMate 214 investigators. Nivolumab plus ipilimumab versus sunitinib in advanced renal-cell carcinoma. N Engl J Med. 2018;378(14):1277–1290. doi: 10.1056/NEJMoa1712126 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Choueiri TK, Powles T, Burotto M, et al. CheckMate 9ER investigators. Nivolumab plus cabozantinib versus sunitinib for advanced renal-cell carcinoma. N Engl J Med. 2021;384(9):829–841. doi: 10.1056/NEJMoa2026982 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Schoenfeld AJ, Hellmann MD. Acquired resistance to immune checkpoint inhibitors. Cancer Cell. 2020;37(4):443–455. doi: 10.1016/j.ccell.2020.03.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Jenkins RW, Barbie DA, Flaherty KT. Mechanisms of resistance to immune checkpoint inhibitors. Br J Cancer. 2018;118(1):9–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Pitt JM, Vétizou M, Daillère R, et al. Resistance mechanisms to immune-checkpoint blockade in cancer: tumor-intrinsic and -extrinsic factors. Immunity. 2016;44(6):1255–1269. doi: 10.1016/j.immuni.2016.06.001 [DOI] [PubMed] [Google Scholar]
- 27.Zhou B, Gao Y, Zhang P, et al. Acquired resistance to immune checkpoint blockades: the underlying mechanisms and potential strategies. Front Immunol. 2021;12:693609. doi: 10.3389/fimmu.2021.693609 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Taha T, Meiri D, Talhamy S, et al. Cannabis impacts tumor response rate to Nivolumab in patients with advanced malignancies. Oncologist. 2019;24(4):549–554. doi: 10.1634/theoncologist.2018-0383 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Biedny A, Szpunar S, Abdalla A, et al. The effect of concomitant cannabinoids during immune checkpoint inhibitor treatment of advanced stage malignancy. J Clin Oncol. 2020;38(15_suppl):e15064–e15064 abstract e15064. doi: 10.1200/JCO.2020.38.15_suppl.e15064 [DOI] [Google Scholar]
- 30.Bar-Sela G, Cohen I, Campisi-Pinto S, et al. Cannabis consumption used by cancer patients during immunotherapy correlates with poor clinical outcome. Cancers (Basel). 2020;12(9):2447. doi: 10.3390/cancers12092447 Erratum in: Cancers (Basel). 2022 Apr 13;14(8):1957. doi: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Waissengrin B, Leshem Y, Taya M, et al. The use of medical cannabis concomitantly with immune checkpoint inhibitors in non-small cell lung cancer: a sigh of relief? Eur J Cancer. 2023;180:52–61. [DOI] [PubMed] [Google Scholar]
- 32.Leighl NB, Laurie SA, Goss GD, et al.; Canadian Cancer Trials Group Lung Disease Site and the Australasian Lung Cancer Trials Group . CCTG BR34: a randomized phase 2 trial of Durvalumab and tremelimumab with or without platinum-based chemotherapy in patients with metastatic NSCLC. J Thorac Oncol. 2022;17(3):434–445. doi: 10.1016/j.jtho.2021.10.023 [DOI] [PubMed] [Google Scholar]; • CCTG BR.34 was a multicenter, randomized phase II study evaluating the efficacy and safety of adding platinum-doublet chemotherapy to first-line treatment with durvalumab and tremelimumab in patients with metastatic NSCLC. The trial addressed the question of the relative contribution of chemotherapy added to first-line checkpoint inhibition in patients who might benefit from checkpoint inhibitor therapy alone.
- 33.Chen EX, Jonker DJ, Loree JM, et al. Effect of combined immune checkpoint inhibition vs best supportive care alone in patients with advanced colorectal cancer: the Canadian cancer trials group CO.26 study. JAMA Oncol. 2020;6(6):831–838. doi: 10.1001/jamaoncol.2020.0910 [DOI] [PMC free article] [PubMed] [Google Scholar]; • CCTG CO.26 was a multicenter, randomized phase II study to evaluate the efficacy of durvalumab and tremelimumab in patients with advanced refractory colorectal cancer. The trial addressed the question of whether combined blockade of PD-L1 and CTLA-4 would provide greater anticancer activity as compared with best supportive care.
- 34.Renouf DJ, Loree JM, Knox JJ, et al. The CCTG PA.7 phase II trial of gemcitabine and nab-paclitaxel with or without durvalumab and tremelimumab as initial therapy in metastatic pancreatic ductal adenocarcinoma. Nat Commun. 2022;13(1):5020. doi: 10.1038/s41467-022-32591-8. [DOI] [PMC free article] [PubMed] [Google Scholar]; • CCTG PA.7 was a multicenter, randomized phase II study that aimed to assess the safety and efficacy of combination chemotherapy/ICI in metastatic pancreatic ductal adenocarcinoma by comparing gemcitabine, nab-paclitaxel plus durvalumab and tremelimumab versus gemcitabine with nab-paclitaxel alone. As part of an exploratory objective, bioinformatics analysis of baseline ctDNA sequencing was performed to assess the utility of incorporating plasma sequencing as a correlative dataset in prospective metastatic pancreatic ductal adenocarcinoma trials.
- 35.Juergens RA, Hao D, Ellis PM, et al. A phase IB study of durvalumab with or without tremelimumab and platinum-doublet chemotherapy in advanced solid tumours: Canadian cancer trials group study IND226. Lung Cancer. 2020;143:1–11. doi: 10.1016/j.lungcan.2020.02.016 [DOI] [PubMed] [Google Scholar]; • CCTG IND.226 was a multicenter, phase Ib study of dual immune blockade with durvalumab with or without tremelimumab in combination with one of four platinum-doublet chemotherapy regimens, in patients with incurable advanced or metastatic cancer, which evaluated safety, tolerability and defined the recommended phase II dose (RP2D) for each treatment combination.
- 36.Drugbank . Available from: https://go.drugbank.com/
- 37.Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45(2):228–247. doi: 10.1016/j.ejca.2008.10.026 [DOI] [PubMed] [Google Scholar]
- 38.Seymour L, Bogaerts J, Perrone A, et al. RECIST working group. iRECIST: guidelines for response criteria for use in trials testing immunotherapeutics. Lancet Oncol. 2017;18(3):e143–e152. doi: 10.1016/S1470-2045(17)30074-8 Erratum in: Lancet Oncol. 2019 May;20(5):e242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.CTCAE . Available from: https://ctep.cancer.gov/protocoldevelopment/electronic_applications/ctc.htm#ctc_40
- 40.Osoba D, Bezjak A, Brundage M, et al. Quality of life committee of the NCIC CTG. Analysis and interpretation of health-related quality-of-life data from clinical trials: basic approach of the National Cancer Institute of Canada Clinical Trials Group. Eur J Cancer. 2005;41(2):280–287. doi: 10.1016/j.ejca.2004.10.017 [DOI] [PubMed] [Google Scholar]
- 41.Ferrara R, Pilotto S, Caccese M, et al. Do immune checkpoint inhibitors need new studies methodology? J Thorac Dis. 2018;10(Suppl S13):S1564–S1580. doi: 10.21037/jtd.2018.01.131 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Vučković S, Srebro D, Vujović KS, et al. Cannabinoids and pain: new insights from old molecules. Front Pharmacol. 2018;9:1259. doi: 10.3389/fphar.2018.01259 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Turcotte C, Blanchet MR, Laviolette M, et al. The CB2 receptor and its role as a regulator of inflammation. Cell Mol Life Sci. 2016;73(23):4449–4470. doi: 10.1007/s00018-016-2300-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Balood M, Ahmadi M, Eichwald T, et al. Nociceptor neurons affect cancer immunosurveillance. Nature. 2022;611(7935):405–412. doi: 10.1038/s41586-022-05374-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Al-Hasani R, Bruchas MR. Molecular mechanisms of opioid receptor-dependent signaling and behavior. Anesthesiology. 2011;115(6):1363–1381. doi: 10.1097/ALN.0b013e318238bba6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Rios C, Gomes I, Devi LA. Mu opioid and CB1 cannabinoid receptor interactions: reciprocal inhibition of receptor signaling and neuritogenesis. Br J Pharmacol. 2006;148(4):387–395. doi: 10.1038/sj.bjp.0706757 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.De Gregorio D, Rj M, Posa L, et al. Cannabidiol modulates serotonergic transmission and reverses both allodynia and anxiety-like behavior in a model of neuropathic pain. Pain. 2019;160(1):136–150. doi: 10.1097/j.pain.0000000000001386 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Giordan Q, Salleron J, Vallance C, et al. Impact of antibiotics and proton pump inhibitors on efficacy and tolerance of anti-PD-1 immune checkpoint inhibitors. Front Immunol. 2021;12:716317. doi: 10.3389/fimmu.2021.716317 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Zhao S, Gao G, Li W, et al. Antibiotics are associated with attenuated efficacy of anti-PD-1/PD-L1 therapies in Chinese patients with advanced non-small cell lung cancer. Lung Cancer. 2019;130:10–17. doi: 10.1016/j.lungcan.2019.01.017 [DOI] [PubMed] [Google Scholar]
- 50.Piper BJ, Tian M, Saini P, et al. Immunotherapy and cannabis: a harmful drug interaction or reefer madness? Cancers (Basel). 2024;16(7):1245. doi: 10.3390/cancers16071245 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Sehgal K, Gill RR, Widick P, et al. Association of performance status with survival in patients with advanced non-small cell lung cancer treated with pembrolizumab monotherapy. JAMA Netw Open. 2021;4(2):e2037120. doi: 10.1001/jamanetworkopen.2020.37120 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Yin Z, Tang H, Li L, et al. Impact of sites versus number of metastases on survival of patients with organ metastasis from newly diagnosed cervical cancer. Cancer Manag Res. 2019;11:7759–7766. doi: 10.2147/CMAR.S203037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Kim SI, Cassella CR, Byrne KT. Tumor burden and immunotherapy: impact on immune infiltration and therapeutic outcomes. Front Immunol. 2021;11:629722. doi: 10.3389/fimmu.2020.629722 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Ye Y, Jing Y, Li L, et al. Sex-associated molecular differences for cancer immunotherapy. Nat Commun. 2020;11(1):1779. doi: 10.1038/s41467-020-15679-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Abdel-Rahman O. Cannabis use among Canadian adults with cancer (2007–2016): results from a national survey. Expert Rev Pharmacoecon Outcomes Res. 2021;21(5):1025–1029. doi: 10.1080/14737167.2021.1854735 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.