Adherence to TNF-alpha inhibitors in patients with hidradenitis suppurativa

Hidradenitis suppurativa (HS) is a chronic, inflammatory, painful skin disease characterized by abscesses, nodules, fistulas and scarring in the axillary, groin and infra-mammary areas (1). Severe HS remains a therapeutic challenge and in many patients treatment with topical or systemic antibiotics or retinoids is ineffective or not well tolerated. The discovery of tumor necrosis factor alpha (TNF-α) in the pathophysiology of HS holds promise that treatment with TNF-α inhibitors provides symptomatic relief in these patients (2).

We read with interest the systematic review by van Rappard et al. concerning off-label treatment of HS with TNF-α inhibitors (3). Based on 65 studies, mainly case reports and case series, involving 459 patients, and including four randomized controlled trials (RCTs) it was concluded that infliximab and adalimumab – but not etanercept – can be a useful supplement to the treatment of recurrent severe HS (evidence 1B).

Treatment recommendations based mainly on case reports (or case series) tend to appreciate the intervention more than if the recommendation was built on a higher level of evidence, for example, cohort studies or RCTs. In contrast, RCTs often include selected patient groups and therefore do not reflect the broad spectrum of patients represented in daily clinical practice. Consequently, data on adherence and sustainability of treatment of HS with TNF-α inhibitors are lacking, and importantly, long-term follow-up data from real-life studies are needed to establish further the position of TNF-α inhibitors in the therapeutic hierarchy of HS.

Since June 2007, we have treated 27 HS-patients at a tertiary referral center (Bispebjerg Hospital, Copenhagen, Denmark) with TNF-α inhibitors. These patients did not respond sufficiently to other recommended systemic agents such as antibiotics (tetracycline, clindamycin, rifampicin and dapsone) or retinoids (isotretinoin, acitretin, etc.). All patients had frequent, recurrent and severe HS of the axillary, groin and/or infra-mammary areas. Prior to therapy all patients screened negative for tuberculosis and hepatitis. The mean age of the population was 40.9 years (range 21–64); 17 (63%) were women (mean age 37.9 years) and 10 (37%) were men (mean age 46.0 years). One patient was HIV-positive.

A total of 19 patients (70%) initially received adalimumab 40 mg twice monthly; five patients (19%) received etanercept 50 mg once weekly; two patients received certulizumab twice monthly; and one patient received infliximab 500 mg once every second month. The choice of primary drug was not based on a pre-specified belief of superiority of one drug over another.

The median adherence time summed over all drugs was 269 d for men and 182 d form women (log-rank p value for difference between men and women = 0.311). Age was a positive predictor for adherence, lowering the risk of adherence failure with 3% per year increase in age, however, not statistically significantly, HR (per year) = 0.97 (0.91–1.03), p = 0.269. After 6 months, 60% of the patients were still treated with the initial drug, whereas after 12 months around 30% of the patients were still adherent to treatment. There was no statistically significant difference in adherence rate between the various drugs, however, the sample was too small to explore this fully. Comparing adalimumab with the other drugs combined showed a trend towards a longer adherence in the adalimumab group (median: 269 versus 214 d), p = 0.922.

Due to insufficient response to the primary TNF-α inhibitor, or unacceptable side effects, a total of 10 (37%) patients were subsequently treated with one or more different TNF-α inhibitors or with the IL-12/23 inhibitor ustekinumab, once every third month, before obtaining sustained reduction in disease activity or discontinued treatment due to insufficient response or side effects. The median adherence rate in the second round of treatment was only 91 d; 33 d for ustekinumab; 39 d for etanercept; 71 d for infliximab; and 354 d for adalimumab (p = 0.414 for difference between drugs). After 6 months, 40% of the patients were still treated with the second drug, whereas after 12 months all patients had failed treatment with the second drug due to insufficient effect or unacceptable side effects.

TNF-α inhibitors are a promising group of agents for various off-label indications including HS. However, we show that adherence to, and consequently sustainability and long-term efficacy of these drugs in a real-life setting is less convincing compared to results from previously published reports (1). In particular, our data do not support a second round of treatment with another TNF-α inhibitor, or ustekinumab, when having failed treatment with an initial TNF-α inhibitor.

There are several possible explanations for the relatively low adherence rate observed in our patient sample. First, it has been shown that patients with severe HS (Hurley grade 3), in contrast to patients with milder disease, lack up-regulation of β-defensin-3 in lesional HS skin, providing a reason for suboptimal treatment effect in these patients (4). Second, several recommended therapies were not tried in our sample, which could have influenced adherence. In particular, a large armamentarium of treatment options are available for HS, most of which are reviewed by Scheinfeld (5). For grade 1 and 2 HS, first-line treatment combines rifampicin with either clindamycin or minocycline. Other HS treatments include: broad spectrum antibiotics such as linezolid, fluoroquinolones, rifampicin, ertapenem, ceftriaxone and metronidazole (6); oral dapsone, zinc, acitretin, alitretinoin (7), hormone inhibitors (oral contraceptives, spironolactone, finasteride, and dutasteride), and oral prednisone (5). For severe HS, systemic immunosuppressive treatment, for example cyclosporine A and TNF-α inhibitors are often required, whereas isotretinoin, isoniazid, lymecycline, sulfasalazine, methotrexate, metformin, colchicine, clarithromycin, intravenous immunoglobulins and thalidomide are less favored treatments (5). The role of botulinum toxin is uncertain, whereas invasive treatment options such as de-roofing of fluctuant nodules and injection of intralesional corticosteroids (triamcinolone acetonide) ameliorates the disease and perhaps, if done regularly, improves HS more permanently (5). Surgical excision, possibly in conjunction with hyperbaric oxygen therapy (8) and CO₂ laser ablation are more definitive treatments (5). Specifically, combination of medical treatment and surgical interventions may provide additional benefit. Establishing the long-term effect of TNF-α inhibitors, and the possibly combined efficacy of these with other available agents for HS, awaits evidence from future clinical trials.

Declaration of interest

The authors report no conflicts of interest.