A 2010 Cochrane review «Rada G, Capurro D, Pantoja T, ym. Non-hormonal int...»1 included 16 randomized controlled trials (RCTs) on the efficacy of non-hormonal therapies for vasomotor symptoms (VMS) in breast cancer (BC) survivors. Three of the studies included also women at high risk for BC, or concerns of BC. The use of tamoxifen varied from 51% to 80% and in three studies all participants used tamoxifen. The use of aromatase inhibitors (2 studies) ranged from 6% to 23%. Pharmacological treatments included six RCTs on venlafaxine, paroxetine, fluoxetine or sertraline, two on clonidine and one on gabapentin. Inconsistent reporting of outcomes prevented a pooled analysis of the data and, thus, no sub-group or sensitivity analyses could be made.
Statistically significant efficacy compared to placebo in reducing the both the frequency and severity of VMS were reported for fluoxetine 20mg (19% and 24%, respectively), paroxetine 10mg (41% and 46%, respectively), paroxetine 20mg (52% and 56%, respectively), oral clonidine (38% and 45%, respectively), transdermal clonidine (44% and 56%, respectively), and gabapentin 900mg (mean change in frequency –2.1; 95% CI –2.95 to –1.23) and severity (–4.88; 95% CI –7.23 to –2.53). Venlafaxine showed a dose-dependent effect in reducing both the the frequency and severity of VMS (Table 1).
| Venlafaxine mg per day |
Reduction in VMS frequency (95% CI) |
Reduction in VMS severity score (95% CI) |
|---|---|---|
| 37.5 | 30% (22-53%) | 37% (26-54%) |
| 75 | 46% (36-63%) | 61% (50-68%) |
| 150 | 58% (4-75%) | 61% (48-75%) |
| Placebo | 19% (14-28%) | 27% (11-34%) |
Comment: The quality of evidence is downgraded by inconsistency (heterogeneity in interventions and outcomes) and imprecision (limited study sizes) and upgraded by dose-response relationship. Of note, follow-up periods in these studies were short.
A narrative review «Biglia N, Bounous VE, De Seta F, ym. Non-hormonal ...»2 summarized data on efficacies and side effects of non-hormonal drugs used for the treatment of VMS. A decrease in VMS may be experienced as early as after two weeks of treatment. Frequently reported side effects include nausea, asthenia, dizziness, xerostomia, constipation and sexual dysfunction, and venlafaxine and duloxetine can also increase blood pressure.
| Drug and dose | Effectiveness Reduction in hot flash frequency (HFF) or severity (HFS) |
Side effects | Inhibitory effect on CYP2D6 enzyme (e.g. interaction with tamoxifen) |
|---|---|---|---|
| Venlafaxine 37.5–150 mg/day 5 studies |
HFF: 30%–58% HFS: 37%–61% |
Nausea, constipation, dry mouth, headache, sleeplessness and decreased appetite | No effect |
| Desvenlafaxine* 100–150 mg/day 4 studies |
HFF: 60%–66% HFS: 24%– 29% | Nausea, dizziness and headache during 1st week of treatment | No effect |
| Citalopram 10–20 mg/day 2 studies |
HFS: 49%-55% | 20% withdrawal rate |
Weak |
| Escitalopram 10–20 mg/day 2 studies |
HFF: 47% | 4% withdrawal rate Nausea, weakness and drowsiness |
Weak |
| Sertraline 25–100 mg/day 3 studies |
HFS: Modest effect | 10% dropout rate Nausea and decreased sexual function |
Weak |
| Duloxetine 30–120 mg/day 1 study |
HFF: 56% HFS: 62% |
Nausea, weakness, drowsiness, insomnia, mouth dryness and constipation | Moderate |
| Paroxetine 10–25 mg/day 5 studies |
HFS: 64% Also sleep improvement |
Nausea at the 20 mg dose Low withdrawal rate with low doses |
Strong |
| Fluoxetine 10–30 mg/day 4 studies |
HFF: 19% HFS: 24% |
18% withdrawal rate due to ineffectiveness rather than to side effects | Strong |
| Gabapentin 300-2700 mg/day 2 studies |
HFF: 44-57% HFS: 46-67% at a dose of 900mg/day |
Drowsiness, unsteadiness, dizziness in up to 50% Withdrawal rate 12-17% |
No effect |
Comment: Being a narrative review this paper serves as an updated overview on the topic.
Clinical comment: It should be noted that duloxetine, paroxetine and fluoxetine inhibit the drug-metabolizing enzyme CYP2D6 and can possibly lower the level of the tamoxifen, which is used as adjuvant endocrine therapy for breast cancer. Furthermore, genetic variance in CYP2C19 and CYP2D6 genes significantly affects serum concentrations (and thus, response and tolerability) of escitalopram, paroxetine, citalopram, and venlafaxine.
A systematic review and meta-analysis «Shan D, Zou L, Liu X, ym. Efficacy and safety of g...»3 included 19 RCTs and 2 cross-over studies on gabapentin as treatment for treating VMS. Data are given as mean difference (MD) [95% confidence interval]. At 4 weeks hot flash frequency was reduced in both naturally menopausal women (-1.82 [-2.56 to -1.08]; 2 studies, 243 participants, I2 0%) and in BC survivors (-1.62 [-2.56 to -0.68]; 2 studies, 432 participants, I2 77%). The dose of gabapentin ranged from 300 to 2400mg and a daily dose of 900mg was used in 7 studies. This dose effectively reduced both hot flash frequency (MD -2.01, -2.60 to -1.43, I2 %) and the hot flash composite score (standardized MD 0.48, -0.67 to -0.29, I218%) at 4 weeks. Adverse events caused stopping the gabapentin treatment compared with placebo, most common being dizziness (RR 4.45, [2.50-7.94]; 7 studies, participants n=2155, I2 38%) and somnolence (RR 3.29, [1.97-5.48]; 6 studies, participants n=2066, I2 8%). Treatment with estrogen was more effective in reducing hot flash frequency than gabapentin (MD 1.11, [0.60 to 1.52[; 2 studies) and the hot flash severity score (SMD 0.50, 0.14 to 0.85; 2 studies).
Comment: The quality of evidence is downgraded by heterogeneity of the data, selection bias and unclear forms of bias, and short follow-up times.
General comment: The level of evidence was downgraded due to heterogeneity of the data.