We designed a SLR to assess the efficacy and safety profile of both topical and IL therapies to treat patients with HS. Following the PICO(T) approach, we defined population (P), intervention (I), comparator (C), outcomes (O), and type of studies (T) as part of this SLR conducted in accordance to the clinical practice guidelines provided by the preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement.13 PROSPERO (ID no. CRD42022361114).14

We included studies on individuals with HS of any sex, age, and ethnicity [Population (P)]. The diagnosis of HS could be defined by the treating physician, or meeting the available diagnostic criteria.15

Any type of topical or IL treatment for HS [Intervention (I)] were included. Therefore, both pharmacological and physical interventions could be included. Destructive therapies, such as laser or surgery were considered out of the scope of this review. Studies evaluating a combination of topical and systemic interventions were also excluded. To facilitate interpretation, a total of five intervention groups were pre-defined: PDT, IL corticosteroids, topical resorcinol, antibiotics, and other interventions. No restrictions were imposed on the comparators (C); placebo, absence of intervention or any other medical interventions (topical or systemic) were included. Studies with no comparator were also included, as were studies comparing between different topical, or IL therapies.

Outcome selection was based on the International Dermatology Outcome Measures (IDEOM) consensus document for developing hidradenitis suppurativa results in health measurements.16 Two co-primary outcomes (O) were defined, one regarding efficacy – self-perceived QoL – and the other one regarding safety – number of patients with, at least, one adverse event (AE). To assess QoL, generic (SF-36 or EQ-5D), dermatology-specific (Dermatology Life Quality Index (DLQI), Skindex 29), and HS-specific scales (Hidradenitis Suppurativa Quality of Life 24, HSQoL-24) were included. Pre-defined secondary efficacy outcomes included patient global assessment, pain, pruritus, disease activity, physician global assessment and treatment satisfaction. Outcomes could be assessed on any scale.

All type of studies (T) were included: randomized, case–control, cohort, and case series. To be included, case series needed to report on more than five patients. Narrative reviews were excluded; previous systematic literature reviews were only used for secondary reference identification. Studies were also excluded if the full paper was unavailable, or written in a language other than English, Spanish, or French.

The search was performed across three main scientific databases (Cochrane Library, MEDLINE and EMBASE) on September 29th, 2022 with no time constraints. Due to the difficulty defining the multiple interventions accurately, the search focused on retrieving papers with the study population (patients with HS). To do so, a combination of standardized (MeSH), and free terms were used. Details of the search are shown in the supporting information (SI, Table 1). A ClinicalTrials.gov search was performed using the terms ‘hidradenitis suppurativa’ and the filter ‘complete’.

Results obtained from the search were added to a citation manager (EndNote® X7). Titles and abstracts of all retrieved abstracts were independently screened by two authors (JCP and RHQ). After title and abstract screening, a full text review of the selected studies was performed. Disagreements between the authors were resolved by a referee (FS). Inclusion criteria followed the PICOT structure described above. The reference list of the studies included in the SLR was, then, checked to identify missed papers. No additional search of the international meetings or key journals was performed.

Data curation was independently performed by JCP and RHQ through an ad hoc designed data extraction sheet, based on the ‘Checklist of items to consider in data collection or data extraction’ available in ‘Cochrane Handbook for Systematic Reviews of Interventions’. Discrepancies were settled by consensus.

The risk of bias and the methodologic quality of the studies included were also evaluated. Due to the heterogeneity of these studies (from case series to RCTs), different tools for proper assessment were established: the Cochrane's ‘risk of bias’ tool (RoB2)17 for clinical trials, and the ‘Newcastle-Ottawa Quality Assessment Scale’ (NOS)17 for observational studies.18 All case series were considered of at a high risk of bias.

If heterogeneity was low, a meta-analysis was planned. Outcomes for dichotomous variables were expressed as relative frequencies, while for continuous variables, outcomes were expressed as means and standard deviations. Publication bias analyses were planned and expressed through a funnel plot. Nevertheless, due to insufficient material, these analyses could not be conducted.

The search retrieved a total of 11,363 references, 4184 of which were obtained from MEDLINE, 6807 from EMBASE, 312 from Cochrane Library, and 60 from Clinicaltrials.gov. After duplicate removal and screening by title and abstract, a total of 43 studies were selected for full text review. Of these, 31 fulfilled all inclusion criteria (Tables 1–5),19–49 and 12 were excluded (SI, Table 2)50–61 (Fig. 1). The 31 studies selected reported on a total of 1143 patients with HS, 62% of whom were women. The mean age of participants was 34.1 years old. The mean number of participants per study was 36.2, ranging from 520,21 to 131,37 and the mean follow-up was 113.2 days. Most patients had chronic HS with a mean course of the disease of 11.7 years, and moderate severity (21.4% were Hurley I and 62.9% Hurley II) (SI, Tables 3 and 4).

Figure 1.

PRISMA flowchart.

(0,12MB).

Most papers reported on case series (n=25), with only five RCTs and a single cohort study. RCTs explored a variety of interventions: topical PDT-comparing a niosomal formulation of 1% methylene blue (NMB) to free methylene blue (FMB),24 IL corticosteroids-comparing triamcinolone acetonide (TCA) to normal saline,29 topical clindamycin-compared to either oral tetracycline44 or placebo,43 and botulinum toxin (BTX) vs placebo.45 The cohort study compared topical clindamycin to topical resorcinol.42 The remaining 25 case series evaluated the interventions, and others such as the combination of IL corticosteroids and lincomycin,46 topical ichthammol,47 topical triethyl citrate,48 or IL hypertonic saline.49

Considering the primary outcomes, 14 studies provided data on QoL, 12 of which used the DLQI.19,22,25–27,33,34,39,40,42,45–47,49 Meanwhile 23 reported the frequency of patients with, at least, one AEs.19–21,23,25–28,30–32,35,37–46,49 The remaining outcomes were assessed in a lower number of studies: patient global assessment of disease status (n=7)24,30,33,43–45,49, pain (n=20)22,23,25,26,29–34,36,38–42,44,45,49, pruritus31 (n=1), disease activity (n=17),20,22–27,30,34,36,39,40,42,44,45,48,49 physician global assessment (n=16),19,21–23,28–32,34,35,41,44,46,47,49 and treatment satisfaction (n=4)29,33,37,46(SI, Table 5).

The RoB for the clinical trials included was considered low for three and high for two44,45 studies: the first with a high risk of attrition bias due to unbalanced withdrawals and a per-protocol analysis,44 the second with a high risk of performance and detection bias due to its open design45 (Fig. 2). The RoB of the only cohort study was also considered low.42 The remaining 25 case series were considered high risk by default.

Figure 2.

Risk of Bias 2 (RoB2).

(0,27MB).

As far as we know, this is the first SLR ever conducted on topical and IL treatments for HS. Several systematic reviews in the past five years have evaluated the efficacy or safety profile of treatments for HS: Robert,62 Gracia Cazaña,63 and Cuenca-Barrales.64 The first focuses on systemic therapies and laser (including PDT), the second on light treatments, and therefore includes data on PDT, while the last one focus exclusively on IL therapies. Since this last SLR, three new studies on IL treatments have been published and collected in our SLR.34,35,49 A prior Cochrane review published by Ingram et al., 8 years ago evaluated all interventions available for HS10; following Cochrane recommendations, the authors exclusively included RCTs, the best design to provide data on therapeutical options. However, the results show scarce evidence on understudied therapies, such as topical and IL therapies, with only three publications43,44,65 being included. Data from the remaining 29 studies included in our SLR were not considered, including two recently published RCTs. Finally, we should mention that none of these reviews included studies on topical resorcinol nor BTX.

The studies included in our SLR had a high degree of heterogeneity regarding patient characteristics, type of lesions included, assessed outcomes, and the follow-up period. This stresses the need to reach consensus on the core outcome domains that must be assessed in studies of patients with HS.66 This has proven to be a successful path in other diseases such as rheumatoid arthritis establishing a set of core outcome domains through OMERACT, which have been universally applied in the studies conducted ever since.67 Additionally, US was used in 12 out of 31 studies included in this review, both as an aid to treatment delivery or to assess the efficacy profile.19,20,23,30–32,34,39–41,49 This finding reinforces the utility of US in the management of patients with HS. On the other hand, the natural history of HS complicates the assessment of the efficacy profile of topical treatments for nodules and abscesses. Von der Werth described that nodules and abscesses often resolve within 7–10 days even without treatment,68 highlighting the need for a comparator to establish efficacy. For example, all lesions in which resorcinol was used were nodules and abscesses. Nonetheless, certain outcomes such as patient satisfaction can indirectly support efficacy.37

The last decade has seen an increase in the treatment options for moderate and severe HS, with the approval of adalimumab by the EMA (European Medicines Agency), and several new drugs seeking approval. This increased effort from the pharmaceutical industry has focused on moderate or severe forms of HS, resulting in a greater number of high-quality scientific studies (RCTs) in these subtypes of patients.69–72 However, the evidence supporting the treatment of mild forms of the disease remains poor. Mild HS is often the steppingstone towards moderate or severe forms. Additionally, even mild forms of the disease have an impact on QoL.73 The use of topical and IL drugs, commonly used in clinical practice in mild forms is not in general supported by RCTs. This fact is highlighted by this SLR and calls for the development of RCTs to determine the efficacy and safety profile of such treatments, regardless of pharmaceutical interests. Even though AE seem to be associated with the use of topical and IL therapies, these are, overall, mild and do not require treatment discontinuation.

The results regarding PDT suggest that it might be an effective and well-tolerated procedure, being the main AE, pain. Regarding IL corticosteroids, although the study conducted by Fajgenbaum is methodologically well-designed, two limitations hamper its external validity, the small volume of corticosteroids used (0.1mL) and the small sample size. Publications on resorcinol might suggest good efficacy and safety profile, but they lack a control group. A single cohort study confirmed the superiority of resorcinol over topical clindamycin, highlighting a therapeutic alternative that could avoid the promotion of antibiotic resistance. Finally, a single RCT with BTX suggests efficacy; a possible mechanism of action might be its anhidrotic effect that could change the skin microbiota.

The guidelines from the British Association of Dermatologists on the management of patients with HS only includes two topical or IL treatments in its recommendations: topical clindamycin and IL corticosteroids, the latter, exclusively for acute lesions. Both treatments are classified as strong recommendations, the benefits outweighing the risks. The European guidelines on the management of HS suggests74 the use of topical resorcinol, adapalene, and azelaic acid. It also mentions the use of IL corticosteroids, and, as experimental treatments, BTX, and PDT. However, only topical clindamycin for mild forms is included in its final algorithm.8 The North American guidelines on the management of HS, recommends skin cleaners and keratolytics, topical clindamycin, and IL corticosteroids.11 As we can see, guidelines are heterogeneous in their recommendation of topical and IL therapies, many times based on expert preferences and not on evidence support.

This SLR is hampered by some limitations. Every precaution was taken to gather all information available, with a wide search across three databases. The studies included show a wide degree of heterogeneity with respect to intervention, population, study design, outcomes and scales, and data analysis. In fact, the heterogeneity of interventions, together with the lack of a comparator in most of the studies, has prevented the development of the pre-planned meta-analysis. Additionally, follow-up periods in many of these studies are short, without further insight into the prolonged administration of these therapies.

In conclusion, although there is an increasing interest and studies on HS, the quality of evidence supporting the use of individual topical, or IL in HS is very low, with conditional support towards topical clindamycin, PDT with NMB and BTX. Further adequately designed RCTs, with standardized outcomes and homogeneous patient populations and lesions are welcomed to support clinical practice decisions.

JCP declared to be a consultant for Abbvie, Novartis and UCB in the field of HS. RHQ, VSG, AVM, IB, and FS declared no conflicts of interest whatsoever.