Evidence strengths and limitations
Only 3 small randomised controlled trials (RCTs) were identified that assessed dimethyl sulfoxide (DMSO) bladder instillation for interstitial cystitis. None of the trials included long-term follow-up. All 3 trials were likely to be subject to significant bias.
Perez-Marrero et al. (1988) compared instillation with DMSO with instillation with saline (placebo) in a small randomised crossover trial in 33 people with biopsies suggestive of interstitial cystitis. The method of randomisation and whether randomisation was concealed was not reported. However, the groups were comparable at baseline.
Participants were blinded to treatment allocation but the garlic halitus (breath smelling of garlic) caused by DMSO bladder instillations makes blinding difficult to maintain. The group randomised to DMSO in the first phase maintained its numbers throughout the study. One patient in group B was found to be pregnant shortly after first instillation of placebo and was removed from the study. Two participants dropped out after placebo before crossing over to DMSO.
Outcomes were assessed using both subjective and objective criteria, and results were reported for both phases of the study. Assessors for the objective response were blinded to treatment allocation.
Peeker et al. (2000) and Sairanen et al. (2009) compared instillation with DMSO with instillation with bacillus Calmette-Guerin (BCG) (off-label use). It should be noted that the European Association of Urology 2012 guidelines on chronic pelvic pain concluded that bladder instillation with BCG is not effective for bladder pain syndrome and that its use for this indication is not recommended.
Peeker et al. (2000) compared instillation with DMSO with instillation with BCG in a small double-blind crossover RCT. The number of participants randomised to each group and the method of randomisation were not reported. It is unclear if randomisation was concealed.
A Cochrane review on intravesical treatments for interstitial cystitis (Dawson and Jamison 2007) identified the Peeker et al. (2000) study but the results from this study did not contribute to the review. This was because of the poor quality of data in the study. The study did not specify the number of participants in either group during the initial instillation and treatment crossover only occurred if the first treatment was ineffective.
Participants were blinded to treatment allocation but as in Perez-Marrero et al. (1988) garlic halitus (breath smelling of garlic) caused by DMSO bladder instillation makes blinding difficult to maintain. The length of time DMSO instillations were held in the bladder was not reported, but the BCG was held for 2 hours. In Perez-Marrero et al. (1988) DMSO was held for 15 minutes. If this was the case, the difference in the length of time the instillations were held for could have led to unblinding.
Four participants who were randomised to BCG for the initial treatment dropped out before crossover to DMSO. No participants who were randomised to DMSO dropped out, but the condition of 2 patients responded to treatment and they did not cross over. The primary outcome was subjective improvement. This was not defined.
The effect of treatment on maximal functional capacity, urinary frequency (both assessed according to patient voiding diaries) and pain (VAS scores) was also reported, according to whether participants had ulcerative or non-ulcerative interstitial cystitis. Data after the first treatment only were not presented. The number of participants in each subgroup, and whether they received BCG or DMSO as the initial treatment, was not reported. It is not clear whether the washout period (at least 7 weeks) used in this study was long enough because it has been reported that a 6-month follow-up after treatment with BCG is needed to define response.
Sairanen et al. (2009) compared instillation with DMSO with instillation with BCG in a randomised open-label trial in 75 people meeting National Institute of Diabetes and Digestive and Kidney Diseases criteria for painful bladder syndrome/interstitial cystitis. The main aim of this study was to evaluate a health-related quality of life questionnaire in painful bladder syndrome/interstitial cystitis.
The dose of BCG administered was not reported, nor was the length of time the instillations were held for. The method of randomisation and whether randomisation was concealed was not reported. The authors reported that the groups were comparable at baseline, however patients in the BCG group had not been symptomatic on average for as long as those in the DMSO group.
This was an open-label trial, so participants were aware of whether they were receiving instillations with DMSO or BCG. Six patients in the DMSO group and 7 patients in the BCG dropped out of the study; the reasons for doing so were not reported.
The primary outcome of the study was alleviation of pain. However, data for this outcome were not reported. Data were reported on subjective global assessment of treatment response after 3 months. Health-related quality of life data were also reported, but outcomes from the end of the 3-month trial could not be extracted. At the end of the 3-month trial, participants were allowed to switch treatments if they wished to, without a specified washout period. Health-related quality of life data were combined for first and second treatments.