Furuzawa‐Carballeda 2012.

Methods	This was a prospective, double‐blind, randomised controlled trial with 3‐arm comparisons, conducted in Mexico. The aim of this study was to evaluate the clinical effect of polymerised collagen (PC) vs. methylprednisolone (MP) in the treatment of morphea, and to determine the influence of PC on Th1, Th2, Th17 and Treg peripheral cells, and on the expression in skin of TGF‐b1, IL‐17A, IL‐22 and Foxp3+ cells.
Participants	The participants were enrolled over a period of 2 years (2008 to 2010) at the Dermatologic Centre Ladislado de la Pascua. Histological, immunohistochemical and flow cytometric evaluation was conducted at the Department of Immunology and Rheumatology of the National Institute of Medical Sciences and Nutrition Salvador Zubirán. Inclusion criteria: individuals aged > 18 years with a diagnosis of morphea based on clinical findings and histological evaluation; at least 1 skin lesion; and a negative reaction to a standard forearm skin test for PC administration. This study also included 10 healthy, untreated individuals (control group) to compare subsets of CD4+ peripheral T cells. Exclusion criteria: pregnancy or breastfeeding; presence of lesions of lichen sclerosis et atrophicus; fibrosis induced by L‐tryptophan, bleomycin, vinyl to PC or its components; history of photosensitivity; use of topical steroids within the 2 months prior to the study, use of systemic corticosteroids or immunosuppressant drugs within 3 months prior to the study; concomitant chronic or malignant disease (melanoma or non‐melanoma skin cancer), any relevant abnormalities in baseline laboratory assessment at baseline, or serological evidence of Borrelia burgdorferi infection. Women of childbearing potential were required to use an acceptable means of contraception. Number of participants randomised: 31 (15 group 1 + 16 group 2) Number of participants analysed: 27 (13 group 1 + 14 group 2) Women: 25 (12 group 1 + 13 group 2) Men: 6 (3 group 1 + 3 group 2) Age: mean age was 38.6 ± 15.0 years, range 18 to 73 years (35.5 ± 14.2 in group 1, 18 to 62; 41.4 ± 15.6 in group 2, 18 to 73). Ethnicity: authors did not report this information. Morphea Type: 30 participants presented with circumscribed morphea and 1 participant in group 1 presented with linear scleroderma. Lenght of illness: 40.5 ± 37.8 months (median 24)
Interventions	Group 1: a. Therapy and dosage: weekly injections of PC ranging from 0.2 mL (equivalent to 1.66 mg collagen) for a lesion of 50 mm in size, up to a maximum of 1.0 mL (8.3 mg collagen) for a lesion > 100 mm in size. b. Administration: subcutaneous (intralesional) c. Duration of treatment: 3 months d. Follow‐up after treatment: 6 months Group 2: a. Therapy and dosage: monthly subcutaneous injections of 0.1 mL methylprednisolone (MP, maximum dose of 20 mg or 5 mL ⁄ month) and weekly subcutaneous injections of 0.1 mL placebo (PVP citric ⁄ citrate buffer). b. Administration: subcutaneous (intralesional) c. Duration of treatment: 3 months d. Follow‐up after treatment: 6 months Groups 1 and 2: additional treatment was restricted to the use of emollients, but the composition was not specified. Group 3: no intervention in healthy participants (not included in the review), only a blood sample.
Outcomes	Complete blood cell count, serum analyses including glucose and electrolyte measurement, and liver function tests and urinalysis. Skin appearance was assessed using a score adapted from the modified Rodnan Scoring System (mRSS), which is an aggregate skin score (0 = normal; 1 = thickened skin (able to pinch skin fold); 2 = thickened skin (unable to pinch skin fold); or 3 = unable to move skin) assessed for each of 17 areas (face, anterior chest, abdomen plus each limb (comprising upper arm, forearm, dorsum of hand, fingers, thighs, lower legs, dorsum of foot)). The maximum potential score is 51. All individual lesions of morphea were graded from 0 to 4 (0 = normal thickness; 1 = mild thickening; 2 = moderate thickening; 3 = severe thickening; 4 = extreme thickening). The total score for all the lesions was then calculated. A decrease in the adapted mRSS exceeding 35% from baseline was considered to reflect a clinically significant improvement. A biopsy site was chosen according to clinical criteria, and was required to be an indurate and inflammatory area, excluding areas with close proximity to vasculature or tendons. The biopsy was observed and also study through Immunohistochemistry to determine expression of IL‐17A, IL‐22, TGF‐b1 and Foxp3. A sample (10 mL) of venous blood was obtained from each subject and a single blood sample was also collected from each of the 10 healthy controls. Peripheral blood mononuclear cells (PBMCs) were obtained by gradient centrifugation. Treatment safety was determined by the occurrence of systemic and local adverse events (AEs) related to administration of the study drug. Safety monitoring included records of vital signs and clinical laboratory tests (blood chemistry, urinalysis and liver function tests). Assessments: laboratory exams were performed at baseline, at the end of treatment and at the last follow‐up visit. At each subsequent visit, efficacy evaluations were conducted and AEs recorded before the scheduled administration of the study medication. Participants attended for follow‐up every 4 weeks for 6 months after the end of the treatment period.
Notes	Trial registration: authors did not register a protocol for this study. Ethics committee approval: from both institutions, the National Institute of Medical Sciences and Nutrition Salvador Zubirán and the Dermatologic Centre Ladislado de la Pascua. Funding source: authors did not report this information. However, authors responded to our contact and informed there was no funding or grant. Declarations of interest: authors declared no conflict of interest.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients were allocated using random number generation and block randomisation to two parallel groups, to receive either PC or steroid". Comment: the method used to generate the random sequence was adequate to produce comparable groups.
Allocation concealment (selection bias)	Low risk	Authors did not provide enough information to judge if the intervention allocation could be foreseen before or during the recruitment of participants. However, authors responded to our contact and reported using sealed envelopes to conceal the assignment.
Blinding of participants and personnel (performance bias) Subjective outcomes	Low risk	Quote: "Both researchers and patients were blinded to the study preparations and had no access to them". Quote: "The placebo experimental preparation was visually identical, and its viscosity was very similar to PC". Comment: participants and personnel were effectively blinded from knowledge of which intervention a participant received. It is unlikely that the blinding could have been broken.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "All patients were assessed before and after treatment by two dermatologists blinded to the treatment groups". Comment: outcome assessors were effectively blinded from knowledge of which intervention a participant received. It is unlikely that the blinding could have been broken.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "For the primary analysis, the means of the scores were compared between the two treatment groups on an intention‐to‐treat (ITT) basis (all patients who received a dose of study medication and had at least one efficacy observation recorded after treatment)". Quote: "In total, 13 patients (87%) in the PC group and 14 (88%) in the MP group completed the study, and were valid for ITT analysis" Comment: authors performed ITT analysis. In addition, authors reported no numerical data for adverse events, but responded to our contact and provided them.
Selective reporting (reporting bias)	Unclear risk	The study protocol is not available.
Other bias	Unclear risk	Authors reported the baseline mean clinical score for all participants, but responded to our contact and provided this data for each group (no statistical difference). However, the clinical tool used (adapted mRSS) assessed individual lesions instead of all the lesions of the individual, which could affect the result.