Protocol for a systematic review and meta-analysis on Janus kinase inhibitors in the management of vitiligo

Vitiligo is a chronic auto-immune skin disorder caused by the destruction of melanocytes resulting in white patches on the skin and hair in all races where its worldwide prevalence in adults varies from 0.5 to 2% ADDIN EN.CITE [1, 2]. Predominantly, younger people under the age of 20 years are affected [1, 2]. Of the 0 to 2% prevalence in children with vitiligo worldwide, evidence suggests that 25% or more of pediatric cases occur up to 10 years of age [3]. However, the prevalence may differ depending on whether its occurrence is reported to healthcare providers, which may be more likely in regions with dark-skinned people such as in Africa, Asia, and Latin America (including Mexico and Brazil) [1]. Consequently, the prevalence of vitiligo may indeed depend on the geographic region [3]. Studies from Benin, Togo, Senegal, Nigeria, Tanzania, and other African countries showed a prevalence of vitiligo ranging from 0.13 to 2.8% [4‐9]. Lu and colleagues reported a 0.1% prevalence in the Chinese population [10]. A study from India reported an 8.8% prevalence of vitiligo [11]. Further, the prevalence of vitiligo in adults ranged from 0.21 to 4% in Mexico and 0.04 to 0.54% in Brazil [12, 13].

Research is ongoing related to the cause of vitiligo, but findings suggest family history, autoimmunity, or extrinsic factors alone or in combination may play a role [14‐17]. Patients usually do not report symptoms associated with vitiligo, but Ezzedine et al. found that almost 20% of patients reported itchiness at the site of a new lesion [1, 18]. Most commonly, vitiligo is found near the orifices, the face, and the upper and lower extremities. The two types of vitiligo depend on the regions of the body affected [1, 18]. Non-segmental vitiligo is the most common type, also known as vitiligo vulgaris; is symmetrical; and may spread from affected regions, while segmental vitiligo affects one side of the body and is unlikely to spread to other body regions [1, 18].

Psychological and social consequences including anxiety and stigmatization can affect the quality of life of affected individuals; especially those with darker skin shades [19‐21]. Specifically, previous studies found that individuals with vitiligo felt stigmatized and experienced low quality of life due to perceiving the visibility of their affected skin parts as socially unacceptable [19, 22, 23]. The quality of life assessed with the vitiligo quality of life (VitiQoL) questionnaire was poor in a sample of adult Nigerian patients with vitiligo [23]. In a cross-sectional study of adults and children with vitiligo in Brazil, the participants reported that mostly stigma from the disease affected their quality of life [22].

Various treatment options are available for vitiligo. A 2015 systematic review [1] described therapies for vitiligo including topical treatments: (1) topical corticosteroids, (2) intralesional corticosteroids, (3) topical vitamin D analogues, (4) topical calcineurin inhibitors, (5) khellin, (6) pseudocatalase and catalase/dismutase superoxide, (7) melagenina (human placental extract), (8) tetrahydrocurcuminoid cream, (9) topical anti-oxidant gel; oral therapies: (10) psoralen and ultraviolet A (UVA); light therapies: (11) punch grafts, minigrafts, and skin thickness grafts, (12) melanocyte transplantation, (13) fractional carbon dioxide (CO2) laser, (14) psychological therapy, and (15) complementary therapies. In 2022, the US Food and Drug Administration, followed by the European Medicines Agency in 2023, approved the topical use of the Janus kinase inhibitor (JAK) ruxolitinib for the treatment of non-segmental vitiligo in individuals 12 years of age and older [24]. Other JAK kinases including upadacitinib, ritlecitinib, brepocitinib, ifidancitinib, cerdulatinib, deglocitinib, baricitinib, and tofacitinib have not been approved for the treatment of vitiligo. In vitiligo, the JAK kinase/signal transducer and activator of transcription (STAT) pathway is activated by interferon (IFN)-gamma-chemokine produced from melanocyte-specified cluster of differentiation 8 (CD8 +)T cells [25]. JAK kinases subsequently phosphorylate STATs that translocate to the nucleus to activate IFN-gamma-inducible genes. Interest in inhibitors of the JAK/STAT pathway has recently increased as a target for vitiligo therapy as this pathway modulates immune cell activation after response to cytokines [25]. Due to the novel application of JAK inhibitors to treat vitiligo, systematic reviews are needed to summarize comprehensively adverse events and patient-centered outcomes surrounding the new treatments.

A recent systematic review and meta-analysis of observational studies and clinical trials by Phan and colleagues described that robust evidence is needed to determine the effectiveness of Janus kinase (JAK) inhibitors in treating vitiligo [26]. However, this systematic review lacked an assessment of patient-centered outcomes, which are of utmost importance surrounding a potentially stigmatizing disease such as vitiligo. The authors did not include an internationally agreed-upon core set of outcomes for vitiligo including tolerability of treatments and cosmetic acceptability of repigmentation [27]. Phan and colleagues also lacked an assessment of the quality of the individual studies that comprised the review, which precludes the weighing of the strength of the evidence against the results. Similarly, a 2023 meta-analysis of JAK inhibitors for the treatment of vitiligo also lacked an assessment of the quality and risk of bias of the evidence and patient-centered outcomes [28]. Additionally, the emerging off-label use of JAK inhibitors for vitiligo warrants further investigation. Thus, we plan to perform an updated systematic review and meta-analysis with an assessment of the quality and risk of bias that will describe patient-centered outcomes deemed important by patients and clinicians and the effectiveness of the most recent available JAK1, JAK2, and JAK 3 inhibitors to treat vitiligo including JAK 1/2 inhibitors ruxolitinib and baricitinib, JAK1/3 inhibitors upadacitinib, tofacitinib, ifidancitinib, and JAK1-3 inhibitor deglocitinib, JAK and spleen tyrosine kinase inhibitor cerdulatinib, JAK 1 and tyrosine kinase 2 inhibitor brepocitinib, and JAK3 and the tyrosine kinase expressed in hepatocellular carcinoma (TEC) kinase family inhibitor ritlecitinib for the treatment of non-segmental and segmental vitiligo [29‐32]. Our Patient, Intervention, Comparator, and Outcome (PICO) (refer to Table 1 Table  review question is, “In children and adults with vitiligo, are topical or systemic JAK inhibitors at various dosages more effective than standard therapy for skin repigmentation within 1 year or longer?”.

After the calibration period, the remaining trials will be exported to Rayyan [39] as separate files using the RIS export option. Two investigators will then independently determine the eligibility of the trials for inclusion. We will copy and paste the data directly into a Google Excel spreadsheet created for data collection. We would describe the number of studies that have data on each of the characteristics as well as the quality and strength of the evidence. Our variables will comprise extracted data about the baseline characteristics, outcomes of the participants including (patient-centered and disease-specific outcomes), and characteristics of the trials and patients (cohort size, blinding type, phase, intervention and comparator names, study duration, outcomes, adverse events, age, sex, race, and other relevant variables). Missing information will be labeled as such. We aim for an overall agreement in the extracted data of Cohen κ ≥ 80% between at least two extractors from a subsample consisting of 10% of the first selected sources. Unclear cases will be discussed until a consensus is reached. The remaining data extraction will be continued by one of the investigators involved in the data extraction from the subsample.

For data that may be dichotomous such as the number of patients with at least 75% improvement in pigmentation, we will calculate the odds ratio (OR) and 95% confidence interval (CI). For categorical data, we will extract information about the category assessed, number of patients with a particular outcome, and a number of patients with particular characteristics. We will describe the categories of repigmentation as 75%, > 75%, ≥ 75%, 75 to 90%/100%, or 76 to 90%/100% [1]. We will extract continuous or numerical data as means and standard deviations (SDs), medians and interquartile ranges (IQRs), follow-up, and change from baseline and used to calculate the mean differences with 95% CIs. Reporting of results (mean difference, precision estimates [i.e., 95% CI], and P values) from statistical analyses comparing the groups will be extracted.

For adverse events, data for all conditions will be combined. A stratified analysis and meta-regression will be performed to determine whether associations will vary according to treatment type, study design (parallel vs. crossover), repigmentation (categories described above), comparator (active vs. placebo), and duration of follow-up as categories based on a previous systematic review (< 6 months, 6 months to 1 year, < 2 years, ≥ 2 years) [1].

We will group studies with similar indications, treatments, and outcomes. If two or more trials will be within a single grouping, we will pool them using random-effects meta-analysis [40]. Continuous outcomes will be analyzed with the mean difference in change from baseline or the mean difference at follow-up (if the mean difference from change in baseline is not reported or calculations from other data are not possible). Based on a previous study, we will use a single data set instead of more than one to avoid double counting of characteristics [1]. We will select an intervention or dose that is comparable to the other studied interventions or doses in trials with more than one intervention.

We plan to express the measures of the effect of the JAK inhibitors, assess the heterogeneity using the I² statistic using meta-analysis techniques (forest plots) if the I² statistic is less than 80%, and assess publication bias (if there will be numerous studies about JAK inhibitors) [41]. For studies with similar interventions, we will use a random-effects model to summarize the treatment effect across the studies as a meta-analysis. We will describe data in narrative or descriptive form if it is heterogeneous or unsuitable for pooling (e.g., data only reported in graphs) in tables. Similarly, study characteristics and descriptive information will be displayed in tables. We will perform sensitivity analyses to assess the effect of the trial design. Parallel-group trial results will be included in the primary analysis, while results from other trial designs (factorial or non-parallel designs) will be included in additional analyses. We will present forest plots to display the summarized measures of effect. Funnel plots will be constructed to determine publication bias. If data will be dichotomous, we will use Harbord’s test, while for continuous outcomes we plan to use Egger’s test. P values < 0.05 will be considered significant. We propose to use Comprehensive Meta-Analysis version 3.3, Microsoft Excel, and IBM SPSS Statistics for Windows, version 25.0 (IBM Corp., Armonk, NY, USA) or similar programs for the analyses. When the pooled 95% CIs do not cross the line of no effect, we will consider this to indicate statistical significance.

We plan to conduct the proposed systematic review guided by evidence-based guidelines to ensure transparency and replicability of our search. Any deviations from our proposed literature search and strategy, identification and selection of reviews, data collection and extraction, and synthesis of the evidence will be updated in the protocol and described in detail in the final draft of the systematic review. We will additionally collect qualitative evidence, which will provide valuable insights into the nature of data in research studies about vitiligo treatment with JAK inhibitors.

To provide the results of the proposed systematic review to as wide an audience as possible, especially to the lay public and researchers, we plan to publish the results of our review in a peer-reviewed, open-access journal. To disseminate further the results of our systematic review, we plan to present them at international conferences and meetings.

The proposed systematic review may be subject to several limitations. First, although our search strategy will be designed using comprehensive sources, the search may not find all of the trials pertaining to our objective, and review authors may be unavailable for further information.