Tan 2014.

Methods	Parallel design, randomised controlled trial Duration of the trial: from November 2011 to July 2013
Participants	266 subjects from 29 centres in Canada Inclusion criteria: male or female subject of any race, aged 12 to 35 years inclusive; subject weighing between 50 and 110 kg; subject with severe acne (IGA at least 4), which in the opinion of the investigator is appropriate for treatment with oral isotretinoin (severe nodular acne, severe inflammatory acne, recalcitrant acne; all unresponsive to conventional first‐line therapies); subject with at least 5 nodules on the face Exclusion criteria: subject with clinically abnormal results to blood tests performed at screening; subject with acne conglobata, acne fulminans, secondary acne (chloracne, drug‐induced acne, etc.), pyoderma faciale, sinus tracks; female subject who is pregnant, nursing or planning a pregnancy during the study; subject with known history of hepatic and/or renal insufficiency, to be confirmed by blood tests; subject with known metabolic or structural bone disease (for 12‐17 years old population); subject with bowel disease and/or with hypervitaminosis A; subject who presents with treated or untreated depression or has a history of depression including a family history of major depression; subject with a wash‐out period (from baseline for topical treatment on the face) of less than two weeks (corticosteroids, antibiotics, antibacterials, antiseptics, retinoids, other anti‐inflammatory drugs or other acne treatments), one week (cosmetic procedures), three months (photodynamic therapy and laser therapy for acne); subject with a wash‐out period from baseline for systemic treatment of less than: corticosteroids, antibiotics (4 weeks), progesterone for contraception (3 months); spironolactone (3 months), other acne treatments (6 months), cyproterone acetate (6 months). Age: for all participants, ranged from 12 to 35 years Age: by group, mean ± standard deviation (years) D+A/BPO Group (n = 133): 19.5 ± 5.0 ISO Group (n = 133): 19.3 ± 4.5 Gender: D+A/BPO Group (n = 133): 115 (male)/18 (female) ISO Group (n = 133): 112 (male)/21 (female) ‐ Duration of acne: not provided ‐ Acne severity: severe
Interventions	D+A/BPO Group (n = 133): adapalene 0.1%/benzoyl peroxide 2.5% gel plus doxycycline 200mg once daily for 20 weeks ISO Group (n = 133): oral isotretinoin once daily (0.5 mg/kg/day for the first 4 weeks with scalation to 1 mg/kg/day for subsequent 16 weeks ‐ mean cumulative dose of 136 mg/kg) plus vehicle gel
Outcomes	Efficacy outcomes: Percent change from baseline in facial nodules, papules/pustules, and total lesion counts (sum of comedones, papules/pustules, and nodules) at weeks 2, 4, 8, 12,16, and 20* Investigator global assessment success rate defined as improvement of at least 2 grades from baseline, measured at weeks 2, 4, 8, 12, 16, and 20* Safety outcomes: ­ Incidence of clinical adverse effects observed by the investigator or reported by the subject at weeks 2, 4, 8, 12, 16, and 20* ­ Incidence of laboratory adverse effects screened by standard laboratory tests at weeks 4, 16, and 20* ­ Incidence of depression assessed by Major Depression Inventory (MDI/ICD‐10) (Bech 2001)* Composite efficacy/safety outcome: Overall success, a composite endpoint including efficacy and safety measurements. The composite endpoint was developed and predefined based on precedent of nodule counts as a standard efficacy measure for oral isotretinoin clinical trials, and the safety criteria based on consultation with study investigators * Indicates outcomes which matched those prespecified for this review
Funding body	Galderma
Notes	Authors declared being advisor, consultant, investigator, speaker, or employee of a company with commercial interest in the results of the study
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Prior to the start of the study, the randomisation list was generated by a statistician. The RANUNI routine of the Statistical Analysis System (SAS; Institute Inc., Cary, NC) was used for the kit number generation. Subjects were randomised in a 1:1 ratio for each group." Comment: The study adopted an adequate method to generate the random sequence
Allocation concealment (selection bias)	Low risk	Quote: "The randomisation list was secured in a locked cabinet and in an electronic file with restricted access to only the designated personnel directly responsible for labelling and handling the study treatments..." Comment: There was an adequate concealment of the allocation sequence
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: "This phase IIIb, non‐inferiority, multicentre, randomised, investigator‐blinded, controlled, parallel group study recruited subjects of any race". Comment: The authors reported clearly that participants were not blinded
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Investigators did not have access to the randomisation list and study treatments were dispensed by the designated study drug dispenser – someone other than the investigator/rater. Both study drug dispenser and subject were instructed not to discuss the study treatments with the investigator/rater." Comment: There was an explicit report of blinding of outcome assessment, despite not having reported an evaluation of the success of blinding
Incomplete outcome data (attrition bias) All outcomes	High risk	Quote: "Three study populations were analysed: safety, intent‐to‐treat (ITT), and per‐protocol (PP) populations. The last observation carried forward method was used to impute missing efficacy values. For the composite endpoint, missing values were considered as unsuccessful." Comment: Since missing data were not equally distributed across groups (regarding both number of missing subjects and reasons for incomplete outcome data), it was unlikely that data analysis was unbiased, even with the use of well recognised methods of data imputation
Selective reporting (reporting bias)	Low risk	Comment: There was an available protocol on a clinical trials registry and outcomes listed in the methods section of the protocol were adequately reported in the final publication
Other bias	Unclear risk	Comment: The study might be at risk of inappropriate influence of funders, as it was promoted by a pharmaceutical company