Difficult-to-treat scabies: oral ivermectin

Classical or uncomplicated scabies

A Cochrane systematic review has evaluated topical and systemic drugs for treating scabies. It included 22 randomised controlled trials (RCTs) that compared drug treatments, herbal or traditional treatments, or any combination of these with placebo, no intervention or a different intervention. Participants included children or adults with a clinical or parasitological diagnosis of scabies and their contacts. Most trials were conducted in countries with healthcare systems that are very different to the UK, such as India. The primary outcome was treatment failure, defined as the persistence of original lesions, the appearance of new lesions or confirmation of a live mite. The secondary outcome was persistence of patient-reported itch.

The review included 9 RCTs of oral ivermectin. One trial compared oral ivermectin with placebo, 5 trials compared oral ivermectin with benzyl benzoate, and one of these plus another trial compared oral ivermectin with permethrin. Two trials compared oral ivermectin with lindane but because this was withdrawn from the UK market in 1996, these trials are not discussed further.

Macotela-Ruíz and Peña-González (1993) compared a single dose of 200 micrograms/kg body weight oral ivermectin with placebo in 55 participants aged over 5 years. There were fewer treatment failures in clinically diagnosed cases in the ivermectin group at 7 days. Treatment failure occurred in 21% of people treated with ivermectin (6/29) compared with 85% of people treated with placebo (relative risk [RR] 0.24, 95% confidence interval [CI] 0.12 to 0.51).

Five trials, with data for a total of 462 participants, compared a single dose of oral ivermectin with various strengths and applications of topical benzyl benzoate. Three trials found no statistically significant difference between the treatments, 1 trial found benzyl benzoate to be more effective, and 1 trial found ivermectin to be more effective.

Brooks and Grace (2002) compared a single dose of 200 micrograms/kg body weight ivermectin with a single application of 10% benzyl benzoate in 110 children aged 6 months to 14 years. No statistically significant difference in treatment failure in clinically diagnosed cases between the groups was found at 3 weeks in the 80 children who completed follow-up. Treatment failed in 44% of children treated with ivermectin (19/43) compared with 49% of people treated with benzyl benzoate (18/37) (RR 0.91, 95% CI 0.57 to 1.46). This trial also reported on itch persistence, and no statistically significant difference in the number of participants who reported night-time itch at 3 weeks was found (30% with ivermectin [10/33] and 56% with benzyl benzoate [14/25] [RR 0.54, 95% CI 0.29 to 1.01]).

Glaziou et al. (1993) compared a single dose of 100 micrograms/kg body weight ivermectin with 2 applications of 10% benzyl benzoate in 44 participants aged 5–56 years. At 30 days, no statistically significant difference in treatment failure in clinically diagnosed cases was found between the groups. Treatment failure occurred in 30% of people treated with ivermectin (7/23) and 52% of people treated with benzyl benzoate (11/21) (RR 0.58, 95% CI 0.28 to 1.22).

Two trials compared 200 micrograms/kg bodyweight ivermectin with 25% benzyl benzoate. Bachewar et al. (2009) initially compared a single dose of ivermectin with 2 applications of benzyl benzoate 25% in 69 adults (a further 34 adults were randomised to 1 application of permethrin cream). Treatment failure was assessed after 1 week, and no statistically significant difference between groups was found in the 52 adults who were followed up. Treatment failure in clinically diagnosed cases occurred in 52% of people treated with ivermectin (14/27) compared with 24% of people treated with benzyl benzoate (6/25) (RR 2.16, 95% CI 0.98 to 4.74). Nnoruka and Agu (2001) compared a single dose of ivermectin with a 72-hour application of 25% benzyl benzoate in 58 participants aged 5–63 years. They assessed treatment failure after 30 days. This trial found a statistically significant difference in favour of ivermectin, with treatment failure occurring in 7% of people treated with ivermectin (2/29) compared with 52% of people treated with benzyl benzoate (15/29) (RR 0.13, 95% CI 0.03 to 0.53).

Ly et al. (2009) compared 150–200 micrograms/kg body weight ivermectin with 1 and 2 applications of 12.5% benzyl benzoate in 181 participants aged 5–65 years. After 14 days, a statistically significant difference in favour of benzyl benzoate was found in the 162 participants who completed follow up. Treatment failure in clinically diagnosed cases occurred in 70% of people treated with ivermectin (38/54) compared with 35% of people treated with benzyl benzoate (38/108) (RR 2.00, 95% CI 1.47 to 2.72).

There was significant heterogeneity between trials, which could be explained by the different drug regimens and follow-up periods.

Two trials compared 200 micrograms/kg body weight oral ivermectin with 5% topical permethrin cream. Both trials reported more treatment failures in clinically diagnosed cases in the ivermectin group. Usha and Gopalakrishnan Nair (2000) compared a single dose of ivermectin with a single application of permethrin cream in 88 participants aged over 5 years. Treatment failure was reported in 30% of people treated with ivermectin (12/40) compared with 2% of people treated with permethrin (1/45) at 2 weeks (RR 13.50, 95% CI 1.84 to 99.26). Bachewar et al. (2009) compared a single dose of ivermectin with 1 application of permethrin cream in 68 adults (a further 35 adults were randomised to 1 application of benzyl benzoate [see above]). They reported treatment failure in 52% of people treated with ivermectin (14/27) compared with 18% of people treated with permethrin (5/28) at 1 week follow-up (RR 2.90, 95% CI 1.21 to 6.96). The combined relative risk of treatment failure for the 2 trials comparing ivermectin with permethrin (n=140) was 4.61 (95% CI 2.07 to 10.26, fixed effect model), favouring permethrin.

Since the publication of the Cochrane systematic review 4 RCTs comparing the efficacy of oral ivermectin with permethrin in classical or uncomplicated scabies have been published.

Chhaiya et al. (2012) was an open-label RCT that initially compared a single application of topical 5% permethrin cream, a single application of topical 1% ivermectin lotion and a single dose of oral ivermectin (200 micrograms/kg body weight) in 315 people aged 5 to 80 years in India. People were followed up at 1, 2, 3 and 4 weeks. If scabies was not cured, the same intervention was repeated at each follow-up. The primary end point of the trial was clinical cure of scabietic lesions, the secondary end point was complete relief of pruritus.

At the end of the first week, after 1 dose or application of treatment, the clinical cure rate was statistically significantly lower with oral ivermectin (30/100 [30%]) compared with permethrin (75/99 [75%], p<0.05) or topical ivermectin (69/101 [69%], p<0.05). In other words, treatment failure was 70% with oral ivermectin and 25% with permethrin.

At the end of the second week, when people who had not previously been cured had received a second dose or application of treatment, the cure rate was still statistically significantly lower with oral ivermectin (63%) compared with permethrin (99%, p<0.05) or topical ivermectin (100%, p<0.05). At the end of the third week, when people who had not previously been cured had received a third dose or application, there was no statistically significant difference between the clinical cure rates (p=0.367): 99% with oral ivermectin, 100% with permethrin and 100% with topical ivermectin (100%). These rates remained the same at the end of the fourth week.

The cure rate for itching was statistically significantly better with permethrin and topical ivermectin compared with oral ivermectin at the end of weeks 2 and 3 (p<0.05), but itching was cured in a similar proportion of people in all groups at the end of week 4.

Goldust et al. (2012) was a single-blind RCT that compared 2 applications of 5% permethrin cream (1 week apart) with a single dose of 200 micrograms/kg body weight ivermectin in 272 people aged 2–84 years in Iran. Cure was defined as the absence of new lesions and all old lesions healed. At 2 weeks, ivermectin was as effective as permethrin. Of the 242 participants followed-up, cure was seen in 112/121 people (93%) treated with 5% permethrin cream and 104/121 people (86%) treated with ivermectin (p=0.42). In other words, treatment failure occurred in 9/121 (7%) of people treated with 5% permethrin cream and 17/121 (14%) of people treated with ivermectin.

Saqib et al. (2012) was an open-label RCT that compared a single dose of 200 micrograms/kg body weight ivermectin with a single application of topical 5% permethrin in 120 adults in Pakistan. Cure was defined as the absence of itching, lesions and microscopic evidence of mites. After 1 and 2 weeks, the number of participants considered to be cured was similar with both treatments. After 1 week, the cure rate was 73% (44/60) with permethrin and 68% (41/60) with ivermectin (p=0.54), and after 2 weeks it was 67% (40/60) with both permethrin and ivermectin (p=1.00). In other words, treatment failure was 27% with permethrin and 32% with ivermectin after 1 week, and 33% with both treatments after 2 weeks.

Sharma and Singal (2011) was a double-blind RCT that compared a single application of topical 5% permethrin cream with a single dose of 200 micrograms/kg body weight ivermectin and 2 doses of 200 micrograms/kg body weight ivermectin (2 weeks apart) in 120 people aged over 5 years in India. Placebo tablets and cream were given to ensure blinding. Complete clinical cure was defined as reduction in both the number of lesions and the grade of pruritus by at least 50%, and negative microscopy. Treatment was considered to have failed if, at the end of the 4 weeks, there was no improvement in pruritus and skin lesions, there were new lesions or there was microscopic evidence of mites.

At the end of the first week, more people who received permethrin achieved complete clinical cure (27/40 [68%]) compared with those who were randomised to either a single dose (14/40 [35%]) or 2 doses (12/40 [30%]) of ivermectin (no statistical analysis given). After 2 weeks, clinical cure was achieved by 87% (33/38) of people who received permethrin, 78% (31/40) who received 1 dose of ivermectin and 67% (26/39) who received 2 doses of ivermectin (no statistical analysis given). After 4 weeks, 95% (36/38) of people who received permethrin, 90% (36/40) of people who received 1 dose of ivermectin and 90% (35/39) of people who received 2 doses of ivermectin achieved complete clinical cure. There was no statistically significant difference between these cure rates at 4 weeks.

In other words, of the 117 participants that were followed up at 4 weeks, treatment failure occurred in 2 people randomised to 5% permethrin (2/38 [5%]), 4 people randomised to 1 dose of ivermectin (4/40 [10%]) and 4 people randomised to 2 doses of ivermectin (4/39 [10%]). This difference was not statistically significant (p=0.769).

This study also reported on pruritus, self-assessed on a visual analogue scale of 0 to 10. The improvement in itching was better with permethrin after 1 week. However, at the end of the fourth week there was no statistically significant difference between groups: 36/38 people (95%) who received permethrin, 36/40 people (90%) who received 1 dose of ivermectin and 35/39 people (90%) who received 2 doses of ivermectin reported at least a 50% improvement in pruritus.

Crusted scabies

No RCTs of ivermectin for the treatment of crusted scabies were identified.

Uncontrolled trials and case series with 4 or more participants with crusted scabies that reported cure rates or treatment failures are included in this evidence summary.

Huffam and Currie (1998) was an Australian open-label study of oral ivermectin in combination with topical therapy for crusted scabies that had not responded to previous treatment with topical therapies. Twenty aboriginal people with refractory crusted scabies were hospitalised for 1 week and given 3 supervised applications of overnight 5% permethrin over the first week. Keratolytic therapy with 10% urea and lactic acid 5% cream were applied on the days when permethrin was not used. In addition, people were treated with up to 3 doses of 200 micrograms/kg oral ivermectin at 14-day intervals. Complete response, defined as normal skin 4 weeks or longer after the last dose of ivermectin, occurred in 8 people (40%), 9 people had at least a partial response and 3 had minimal improvement. It is unclear whether the results of this study would be applicable to patients in the UK with crusted scabies.

Larralde et al. (1999) described the use of ivermectin to successfully treat 4 people with crusted scabies in Argentina. Two people had Down's syndrome, and their crusted scabies was refractory to repeated treatment with topical 5% permethrin. Two to 3 weeks after 1 dose of 200 micrograms/kg oral ivermectin, plaques were still present in 1 person, and new plaques appeared in the other person. However, no signs and symptoms of scabies were present 2 weeks after a second dose. Two people were HIV positive and were treated with 2 doses of 200 micrograms/kg oral ivermectin 1 week apart. No signs of crusted scabies developed during 6 months of follow-up.

Alberici et al. (2000) was a retrospective analysis of 39 people with scabies who were HIV positive and were admitted to a hospital in Italy during a scabies epidemic. Eight of these people had crusted scabies. People were treated with topical 15% benzyl benzoate solution applied twice daily for 3 days, a single dose of 200 micrograms/kg ivermectin or a combination of both of these treatments. Complete clinical response was defined as both resolution of itching and either dermatological or microbiological cure, and treatment failure as persistent microbiological evidence of infestation within 4 weeks of treatment. Of the 8 people with crusted scabies, treatment failed in all 3 people treated with benzyl benzoate alone and in the 1 person treated with ivermectin alone. All 4 people treated with a combination of ivermectin and benzyl benzoate had a complete treatment response.

Paasch and Haustein (2000) reported on the management of endemic outbreaks of scabies in 3 residences for the elderly in Germany. They reported on 252 patients, staff and family members living in these residencies who showed recurrent infestations over more than 1 year. Twelve people had crusted scabies and received 12 mg ivermectin once (n=5) or twice (n=7) after an interval of 8 days. No treatment failures were reported in people given ivermectin, although the length of follow-up was not reported and it was unclear whether additional topical treatment was given. The other 240 people received topical treatment with either permethrin cream or allethrin spray.

Nofal (2009) was an uncontrolled study of oral ivermectin for crusted scabies in Egypt. Eight people with crusted scabies were given a single oral dose of 200 micrograms/kg ivermectin and re-examined at 2, 4, 6 and 8 weeks. A second dose of ivermectin was given if treatment failed, defined as persistence of pruritus, or clinical signs or microscopic evidence of scabies, at the end of the second week. A third dose of ivermectin, combined with topical permethrin 5% and salicylic acid 5%, was given at the end of the fourth week to people whose scabies did not respond to the second dose. Two people were cured at the end of week 2, after a single dose of ivermectin. Four people were cured at the end of week 4, after 2 doses of ivermectin. The remaining 2 people were cured at the end of week 6, after 3 doses of ivermectin combined with topical therapy.