Importance  Oral isotretinoin has been associated with several adverse effects, but evidence-based estimates of laboratory changes during isotretinoin therapy in large patient samples are limited.

Objective  To develop estimates of the laboratory changes that occur during isotretinoin therapy for acne using extant data and meta-analytic methods.

Data Sources  A comprehensive search strategy using Ovid/MEDLINE, EMBASE, and gray literature was conducted (1960-August 1, 2013) to identify all relevant studies of isotretinoin use in acne vulgaris. Terms related to acne treatment, isotretinoin, and diagnostic procedures were searched with all available synonyms.

Study Selection  Inclusion criteria consisted of clinical trials using oral isotretinoin, doses of 40 mg/d or more, duration of at least 4 weeks, patients aged 9 to 35 years with acne vulgaris, and 10 or more participants. Studies from all countries published in any language were included. Exclusion criteria were use of modified isotretinoin products, isotretinoin therapy for conditions other than acne vulgaris, and concomitant acne therapy. The initial search yielded 342 records; 116 of these were screened for full-text examination.

Data Extraction and Synthesis  Two authors independently reviewed the publications to determine eligibility, and disagreements were resolved by a third author. Generated weighted means and 99% CIs were calculated using the reported means (SDs or SEs). A random effects model was created, and statistical heterogeneity was quantified. Data were analyzed from August 25, 2014, to December 4, 2015.

Main Outcomes and Measures  Laboratory values for lipid levels, hepatic function, and complete blood cell count were evaluated.

Results  Data from 61 of the 116 studies were evaluated; 26 studies (1574 patients) were included in the meta-analysis. The mean (99% CI) values during treatment (nonbaseline) for triglycerides was 119.98 mg/dL (98.58-141.39 mg/dL); for total cholesterol, 184.74 mg/dL (178.17-191.31 mg/dL); for low-density lipoprotein cholesterol, 109.23 mg/dL (103.68-114.79 mg/dL); for high-density lipoprotein cholesterol, 42.80 mg/dL (39.84-45.76 mg/dL); for aspartate aminotransferase, 22.67 U/L (19.94-25.41 U/L); for alanine aminotransferase, 21.77 U/L (18.96-24.59 U/L); for alkaline phosphatase, 88.35 U/L (58.94-117.76 U/L); and for white blood cell count, 6890/µL (5700/µL-8030/µL). This meta-analysis showed that (1) isotretinoin is associated with a statistically significant change in the mean value of several laboratory tests (white blood cell count and hepatic and lipid panels), yet (2) the mean changes across a patient group did not meet a priori criteria for high-risk and (3) the proportion of patients with laboratory abnormalities was low.

Conclusions and Relevance  The evidence from this study does not support monthly laboratory testing for use of standard doses of oral isotretinoin for the standard patient with acne.