Thirty RCTs were included (1,760 patients were randomised but 394 patients were subsequently excluded by the original authors, leaving 1,366 evaluable patients).
The average sample size was 46 patients (range: 8 to 139).
Anti-emetic efficacy (10 RCTs).
There was wide variability in event rates with both cannabinoids and controls. The scatter in the event rates suggested increased efficacy with cannabinoids and relative homogeneity of the data.
Cannabinoids versus active control. Cannabinoids were more effective than active comparators for complete control of nausea (7 RCTs) and for complete control of vomiting (6 RCTs); the RRs were 1.38 (95% CI: 1.18, 1.62; NNT=6.4) and 1.28 (95% CI: 1.08, 1.51; NNT=8.0), respectively.
Cannabinoids versus placebo control. Cannabinoids were more effective than placebo for complete control of nausea (4 RCTs) and for complete control of vomiting (4 RCTs); the RRs were 1.21 (95% CI: 1.03, 1.42; NNT=8.0) and 1.84 (95% CI: 1.42, 2.38; NNT=3.3), respectively.
After the exclusion of studies in which the control groups showed extreme event rates (less than 25% or greater than 75%), cannabinoids were superior to conventional anti-emetics. The RRs for complete control of nausea and complete control of vomiting were 1.70 (95% CI: 1.32, 2.18; NNT=3.4) and 1.26 (95% CI: 1.07, 1.48; NNT=6.6), respectively.
After limiting the analysis to those trials with extreme event rates in the control group, there was no significant difference between cannabinoids and active comparators. These studies included one RCT that used low emetogenic drugs (vincristine and fluorouracil) and had low rates (less than 25%) of nausea in the control group, and 6 RCTs that had high rates (greater than 75%) of nausea in the placebo control groups. The latter 6 RCTs used regimes involving the following: high-dose methotrexate; doxorubicin and cytoxan; cisplatin; or a combination of cyclophosphamide, methotrexate and fluorouracil.
Patients' preference (18 crossover RCTs).
Significantly more patients expressed a preference to be treated with cannabinoids for further chemotherapy cycles rather than either placebo or active comparator. The RR versus placebo (4 RCTs) was 5.67 (95% CI: 3.95, 8.15; NNT=1.6) and the RR versus active control (14 RCTs) was 2.39 (95% CI: 2.05, 2.78; NNT=2.8).
Side-effects.
Side-effects were significantly more common with cannabinoids. The potentially beneficial side-effects found were 'high' (RR 10.6, 95% CI: 6.86, 16.5), sedation or drowsiness (RR 1.66, 95% CI: 1.46, 1.89), and euphoria (RR 12.5, 95% CI: 3.00, 52.1). The definitely harmful side-effects found were as follows: dysphoria or depression (RR 8.06, 95% CI: 3.38, 19.2), dizziness (RR 2.97, 95% CI: 2.31, 3.83), hallucinations (RR 6.10, 95% CI: 2.41, 15.4), paranoia (RR 8.58, 95% CI: 6.38, 11.5), and arterial hypotension (RR 2.23, 95% CI: 1.75, 2.83).
Withdrawals due to side-effects were significantly more common with cannabinoids than with the control; the RR was 4.67 (95% CI: 3.07, 7.09; NNT=11).