We wanted to find out how well treatments for infections caused by the Streptococcus bacteria worked, and how safe they were, when compared with no treatment, placebo (an identical but inactive treatment), or each other, in people with acute guttate or chronic plaque psoriasis.
Chronic plaque psoriasis is a long-term condition that causes patches of red, flaky skin, covered with scales (called plaques); it is the most frequent form of psoriasis, and is more common in adults.
Guttate psoriasis is characterised by smaller lesions, and is more common in children and young people. Some studies have suggested that guttate psoriasis occurs in less than 30% of people with psoriasis.
The cause of psoriasis is unknown, but Streptococcal infection may trigger guttate psoriasis or flare-ups of chronic plaque psoriasis.
Tonsillectomy may prevent or reduce the severity of throat infections, and limit the Streptococcus reservoir. Antibiotics work by destroying the bacteria that appear to trigger psoriasis.
The evidence is current to January 2019.
We included five studies (162 participants); three were conducted in hospital dermatology departments. Participants were 12 to 77 years old (100 males; 62 females). One study was funded by a pharmaceutical company. The severity of the condition ranged from mild to severe. Streptococcus bacteria were found in the throats of 14% of people.
We classed outcomes measured within eight weeks of the start of treatment as short-term, and those measured at least one year after the start of treatment as long-term. The antibiotic trials in guttate psoriasis patients were all short-term in duration; the antibiotic trial in chronic plaque psoriasis was 48 weeks long.
Three studies included participants with guttate psoriasis, and assessed the short-term effects of antibiotics: penicillin (20 participants), or erythromycin compared to no treatment (43 participants), and rifampicin compared to placebo (20 participants).
Two studies included participants with chronic plaque psoriasis. One study assessed azithromycin (antibiotic) versus vitamin C at 48 weeks (50 participants); one assessed tonsillectomy versus no intervention at eight weeks and 24 months (29 participants).
These results are backed by very low-quality evidence, so we are not certain of their accuracy. Each result is based on only one study.
No studies measured our main outcome of interest, the time taken for the skin to be clear or almost clear of lesions, or the risk of relapsing at least once during long-term follow-up.
No side effects were seen when penicillin was compared with no treatment in people with guttate psoriasis. Side effects were not measured for the comparisons of rifampicin versus placebo, or erythromycin versus no treatment.
In participants with chronic plaque psoriasis, one trial assessed azithromycin versus vitamin C, and 10 participants in the azithromycin group complained of nausea or mild stomach upset. A trial of tonsillectomy versus no treatment reported one case of minor bleeding in the tonsillectomy group.
Two studies in participants with chronic plaque psoriasis measured the number of participants achieving a 75% reduction on the Psoriasis Area and Severity Index (PASI 75). In one, 18/30 participants in the azithromycin group reached PASI 75 versus none in the vitamin C group. In the other, 3/15 in the tonsillectomy group reached PASI 75 versus none in the no treatment group. The guttate psoriasis trials did not assess this outcome.
We are uncertain whether the number of participants with guttate psoriasis achieving clear or almost clear skin differs between those given penicillin and those receiving no treatment. Only one participant with chronic plaque psoriasis achieved almost clear skin in the tonsillectomy group compared to none in the no treatment group. The other three trials did not measure this outcome.
Quality of the evidence
Many of our main outcomes were not assessed. Those that were assessed were based on very low-quality evidence, meaning we are not sure of their accuracy. The studies were very small, and had a high risk of bias because participants and trial assessors were aware of treatment allocation. More studies are needed to see if antibiotic treatment of Streptococcal infection shortens the duration of acute guttate psoriasis, stopping it from turning into a long-term condition (chronic plaque psoriasis).
We found only five trials (N = 162), which assessed the effects of five comparisons (systemic antibiotic treatment (penicillin, azithromycin) or tonsillectomy). Two comparisons (erythromycin compared to no treatment, and rifampicin compared to placebo) did not measure any of the outcomes of interest. There was very low-quality evidence for the outcomes that were measured, Therefore, we are uncertain of both the efficacy and safety of antistreptococcal interventions for guttate and chronic plaque psoriasis.
The included trials were at unclear or high risk of bias and involved only a small number of unrepresentative participants, with limited measurement of our outcomes of interest. The studies did not allow investigation into the influence of Streptococcal infection, and a key intervention (amoxicillin) was not assessed.
Further trials assessing the efficacy and tolerance of penicillin V or amoxicillin are needed in children and young adults with guttate psoriasis.
Psoriasis is a chronic skin disease that affects approximately two per cent of the general population. Plaque psoriasis is the most common form: it usually appears as raised, red patches of inflamed skin, covered with silvery white scales. The patches often occur in a symmetrical pattern. Guttate psoriasis is a particular form of psoriasis with widespread, small erythematosquamous lesions. Streptococcal infection is suspected to be a triggering factor for the onset of guttate psoriasis, and flare-up of chronic plaque psoriasis. The previous Cochrane Review on this topic was published in 2000; it required an update because antistreptococcal treatment continues to be used to treat psoriasis, especially for the acute form of guttate psoriasis.
To assess the effects of antistreptococcal interventions for guttate and chronic plaque psoriasis.
We searched Cochrane Skin Specialised Register, Cochrane Register of Studies Online, CENTRAL, MEDLINE, Embase, LILACS, and five trials registers (January 2019). We checked the reference lists of included and excluded studies and searched conference proceedings from the American Academy of Dermatology, Society for Investigative Dermatology, and European Academy of Dermatology and Venereology.
We considered randomised controlled trials (RCTs) assessing antistreptococcal interventions (tonsillectomy or systemic antibiotic treatment) in people with clinically diagnosed acute guttate and chronic plaque psoriasis compared with placebo, no intervention, or each other.
We used standard methodological procedures expected by Cochrane. Primary outcome measures were: 1) time-to-resolution; achieving clear or almost clear skin (Physician Global Assessment (PGA) 0 or 1 or Psoriasis Area and Severity Index (PASI) 90 or 100); 2) proportion of participants with adverse effects and severe adverse effects. Secondary outcomes were: 1) proportion of participants achieving clear or almost clear skin; 2) proportion of participants achieving PASI 75 or PGA 1 to 2; 3) risk of having at least one relapse at long-term follow-up. Short-term assessment was defined as within eight weeks of the start of treatment; long-term was at least one year after the start of treatment.
We included five trials (162 randomised participants); three were conducted in a hospital dermatology department. One study declared funding by a pharmaceutical company. Participants' ages ranged from 12 to 77 years; only two participants were younger than 15 years. Mean PASI score at baseline varied from 5.7 (i.e. mild) to 23 (i.e. severe) in four studies. Twenty-three of 162 participants had streptococcus-positive throat swab culture. We did not perform a meta-analysis due to heterogeneity of participants' characteristics and interventions.
None of the trials measured our efficacy primary outcome, time-to-resolution, or the secondary outcome, risk of having at least one relapse at long-term follow-up.
We rated the quality of the results as very low-quality evidence, due to high risk of bias (absence of blinding of participants and caregivers, and high risk of outcome reporting bias) and imprecision (single study data with a low number of events). Hence, we are very uncertain about the results presented.
One three-armed trial (N = 43) assessed penicillin (50,000 international units (IU)/kg/day in three doses) versus erythromycin (250 mg four times per day) versus no treatment (treatment for 14 days, with six-week follow-up from start of treatment). Adverse events and the proportion of participants achieving clear or almost clear skin were not measured.
One trial (N = 20) assessed penicillin (1.6 MU (million units) intramuscularly once a day) versus no treatment (six weeks of treatment, with eight-week follow-up from start of treatment). At six-week (short-term) follow-up, no adverse events were observed in either group, and there was no statistically significant difference between the two groups in the proportion of participants with clear or almost clear skin (risk ratio (RR) 2.00, 95% confidence interval (CI) 0.68 to 5.85).
One trial (N = 20) assessed rifampicin (300 mg twice daily) versus placebo (14-day treatment duration; six-week follow-up from start of treatment); none of the review outcomes were measured.
These trials did not measure the proportion of participants achieving PASI 75 or PGA 1 to 2.
Chronic plaque psoriasis
One trial (N = 50) assessed long-term azithromycin treatment (500 mg daily dose) versus vitamin C. Adverse events were reported in the azithromycin group (10 out of 30 had nausea and mild abdominal upset), but not in the vitamin C group. The proportion of participants who achieved clear or almost clear skin was not measured. In the azithromycin group, 18/30 versus 0/20 participants in the vitamin C group reached PASI 75 at the end of 48 weeks of treatment (RR 25.06, 95% CI 1.60 to 393.59).
One trial (N = 29) assessed tonsillectomy versus no treatment, with 24-month follow-up after surgery. One participant in the tonsillectomy group had minor bleeding. At eight-week follow-up, 1/15 in the tonsillectomy group, and 0/14 in the no treatment group achieved PASI 90; and 3/15 participants in the tonsillectomy group, and 0/14 in the no treatment group achieved PASI 75 (RR 6.56, 95% CI 0.37 to 116.7).