We wanted to investigate the effects of treatment for chronic (lasting longer than six weeks) pruritus (itching) of unknown cause in children and adults. We assessed all treatments, as long as they were compared against each other, placebo (an identical but inactive treatment), a sham procedure, or no treatment (or equivalent, e.g. waiting list). We were particularly interested in assessing safety and itch intensity as reported by the patient or the parent.
Pruritus, or itching, is an unpleasant sensation that provokes a desire to scratch. It can be caused by diseases of the skin or other parts of the body. We searched the medical literature up to July 2019 to determine the effects of drug and non-drug therapies (e.g. phototherapy) used for treatment of itching of unknown cause.
We included one study (257 participants) that investigated the safety and efficacy of three different doses of a drug called serlopitant (5 mg, 1 mg, and 0.25 mg, taken by mouth once daily for six weeks) versus placebo for severe chronic pruritus (participants had a score of 7 cm or higher on the visual analogue scale (VAS)). The age of included participants ranged from 18 to 65; 60.6% were women; 55% suffered from itching of unknown origin; and approximately 45% presented a dermatological diagnosis (atopic dermatitis/eczema 37.3%, psoriasis 6.7%, acne 3.6%, among other diagnoses). A pharmaceutical company funded this study, which was undertaken across 25 centres in the United States (clinical research centres and universities). The study lasted 10 weeks in total (six weeks of treatment plus four weeks of post-treatment follow-up).
We found no eligible studies for the main treatments we sought to assess, which included emollient creams, cooling lotions, topical corticosteroids (a class of steroid hormones) or antidepressants, systemic antihistamines (medicines used to relieve symptoms of allergies) or antidepressants, anticonvulsants (antiseizure drugs), and phototherapy.
Participants who received serlopitant at doses of 0.25 mg, 5 mg, and 1 mg may be more likely to experience reduced itch intensity, as reported by the patient, when compared with participants given placebo (low-certainty evidence). However, for serlopitant 1 mg and 0.25 mg, the range of possible results indicates there may be little to no difference between groups.
We are uncertain of the effects of serlopitant (in the three doses) on side effects, health-related quality of life, and sleep disturbances due to very low-certainty evidence.
The most commonly reported side effects were sleepiness, diarrhoea, headache, and upper respiratory tract infection, among others.
All outcomes were measured at the end of treatment (six weeks from baseline) with the exception of adverse events, which were monitored throughout the study.
The included study did not report the effects of this drug on depression and patient satisfaction.
Certainty of the evidence
Certainty of the evidence was low for patient-reported itch intensity because 45% of participants had an identifiable skin disease and 55% had itch of poorly defined cause. Additionally, the number of study participants was small and there were few occurrences of the outcomes, or results were imprecise or were not meaningful; therefore, the study was at risk of random errors.
Certainty of the evidence was very low for three outcomes (adverse events, quality of life, and sleep disturbances) due to additional concerns that measurement of these outcomes was not pre-planned. Also, no information was available to assess bias from missing data for the outcome of sleep disturbances.
We found lack of evidence to address our review question: for most of our interventions of interest, we found no eligible studies. The neurokinin 1 receptor (NK1R) antagonist serlopitant was the only intervention that we could assess. One study provided low-certainty evidence suggesting that serlopitant may reduce pruritus intensity when compared with placebo. We are uncertain of the effects of serlopitant on other outcomes, as certainty of the evidence is very low.
More studies with larger sample sizes, focused on patients with CPUO, are needed. Healthcare professionals, patients, and other stakeholders may have to rely on indirect evidence related to other forms of chronic pruritus when deciding between the main interventions currently used for this condition.
Pruritus is a sensation that leads to the desire to scratch; its origin is unknown in 8% to 15% of affected patients. The prevalence of chronic pruritus of unknown origin (CPUO) in individuals with generalised pruritus ranges from 3.6% to 44.5%, with highest prevalence among the elderly. When the origin of pruritus is known, its management may be straightforward if an effective treatment for the causal disease is available. Treatment of CPUO is particularly difficult due to its unknown pathophysiology.
To assess the effects of interventions for CPUO in adults and children.
We searched the following up to July 2019: Cochrane Skin Group Specialised Register, CENTRAL, MEDLINE, Embase, and trials registries. We checked the reference lists of included studies for additional references to relevant trials.
We sought to include randomised controlled trials and quasi-randomised controlled trials that assessed interventions for CPUO, as defined in category VI ('Other pruritus of undetermined origin, or chronic pruritus of unknown origin') of the International Forum for the Study of Itch (IFSI) classification, in children and adults. Eligible interventions were non-pharmacological or topical or systemic pharmacological interventions, and eligible comparators were another active treatment, placebo, sham procedures, or no treatment or equivalent (e.g. waiting list).
We used standard methodological procedures expected by Cochrane. Our primary outcomes were 'Patient- or parent-reported pruritus intensity' and 'Adverse events'. Our secondary outcomes were 'Health-related quality of life', 'Sleep disturbances', 'Depression', and 'Patient satisfaction'. We used GRADE to assess the certainty of evidence.
We found there was an absence of evidence for the main interventions of interest: emollient creams, cooling lotions, topical corticosteroids, topical antidepressants, systemic antihistamines, systemic antidepressants, systemic anticonvulsants, and phototherapy.
We included one study with 257 randomised (253 analysed) participants, aged 18 to 65 years; 60.6% were female. This study investigated the safety and efficacy of three different doses of oral serlopitant (5 mg, 1 mg, and 0.25 mg, once daily for six weeks) compared to placebo for severe chronic pruritus; 25 US centres participated (clinical research centres and universities). All outcomes were measured at the end of treatment (six weeks from baseline), except adverse events, which were monitored throughout. A pharmaceutical company funded this study.
Fifty-five per cent of participants suffered from CPUO, and approximately 45% presented a dermatological diagnosis (atopic dermatitis/eczema 37.3%, psoriasis 6.7%, acne 3.6%, among other diagnoses). We unsuccessfully attempted to retrieve outcome data from study authors for the subgroup of participants with CPUO. Participants had pruritus for six weeks or longer. Total study duration was 10 weeks.
Participants who received serlopitant 5 mg may have a greater rate of relief of patient-reported pruritus intensity as measured by the visual analogue scale (VAS; a reduction in VAS score indicates improvement) compared to placebo (126 participants, risk ratio (RR) 2.06, 95% confidence interval (CI) 1.27 to 3.35; low-certainty evidence). We are uncertain of the effects of serlopitant 5 mg compared to placebo on the following outcomes due to very low-certainty evidence: adverse events (127 participants; RR 1.48, 95% CI 0.87 to 2.50); health-related quality of life (as measured by the Dermatology Life Quality Index (DLQI); a higher score indicates greater impairment; 127 participants; mean difference (MD) -4.20, 95% CI -11.68 to 3.28); and sleep disturbances (people with insomnia measured by the Pittsburgh Sleep Symptom Questionnaire-Insomnia (PSSQ-I), a dichotomous measure; 128 participants; RR 0.49, 95% CI 0.24 to 1.01).
Participants who received serlopitant 1 mg may have a greater rate of relief of patient-reported pruritus intensity as measured by VAS compared to placebo; however, the 95% CI indicates that there may also be little to no difference between groups (126 participants; RR 1.50, 95% CI 0.89 to 2.54; low-certainty evidence). We are uncertain of the effects of serlopitant 1 mg compared to placebo on the following outcomes due to very low-certainty evidence: adverse events (128 participants; RR 1.45, 95% CI 0.86 to 2.47); health-related quality of life (DLQI; 128 participants; MD -6.90, 95% CI -14.38 to 0.58); and sleep disturbances (PSSQ-I; 128 participants; RR 0.38, 95% CI 0.17 to 0.84).
Participants who received serlopitant 0.25 mg may have a greater rate of relief of patient-reported pruritus intensity as measured by VAS compared to placebo; however, the 95% CI indicates that there may also be little to no difference between groups (127 participants; RR 1.66, 95% CI 1.00 to 2.77; low-certainty evidence). We are uncertain of the effects of serlopitant 0.25 mg compared to placebo on the following outcomes due to very low-certainty evidence: adverse events (127 participants; RR 1.29, 95% CI 0.75 to 2.24); health-related quality of life (DLQI; 127 participants; MD -5.70, 95% CI -13.18 to 1.78); and sleep disturbances (PSSQ-I; 127 participants; RR 0.60, 95% CI 0.31 to 1.17).
The most commonly reported adverse events were somnolence, diarrhoea, headache, and nasopharyngitis, among others.
Our included study did not measure depression or patient satisfaction.
We downgraded the certainty of evidence for all outcomes due to indirectness (only 55% of study participants had CPUO) and imprecision. We downgraded outcomes other than patient-reported pruritus intensity a further level due to concerns regarding risk of bias in selection of the reported result and some concerns with risk of bias due to missing outcome data (sleep disturbances only). We deemed risk of bias to be generally low.