Histopathologic Study of Hidradenitis Suppurativa Following Long-Pulsed 1064-nm Nd:YAG Laser Treatment

Hidradenitis suppurativa (HS) is a chronic, recurrent inflammatory disease primarily affecting the skin flexures, such as the axilla, inframammary, and inguinal areas. It can lead to clinically significant morbidity and has been shown to negatively affect quality of life substantially more than skin conditions such as psoriasis and atopic dermatitis.^1,2 Although the etiology of this condition has been debated, recent histopathologic studies demonstrate follicular occlusion as the inciting event, with resulting perifollicular inflammation, rupture of the follicular infundibulum, formation of painful nodules and abscesses, and, finally, sinus tract formation and scarring. The apocrine glands seem to be secondarily involved rather than the origin of the disease.³ Hidradenitis suppurativa is also termed acne inversa and is part of the follicular occlusion tetrad, in conjunction with acne conglobata, dissecting cellulitis, and pilonidal sinuses.⁴

Various medical and surgical regimens have been tried and are dependent on the stage and severity of the disease according to the Hurley scaling system.^5,6 Stage I disease consists of 1 or more abscesses with no sinus tracts or scarring; stage II disease is characterized by 1 or more widely separated recurrent abscesses with tract formation and scarring; and, finally, stage III disease consists of multiple interconnected tracts and abscesses throughout the area.⁷ Mild disease can generally be managed with lifestyle changes, such as weight loss, smoking cessation, and wearing loose-fitting clothes. Pharmacologic treatments, such as topical and oral antibiotics, as well as intralesional corticosteroids, can be added.^8,9 Hormone therapy, retinoids, and immunosuppressive and biologic therapies constitute other medical treatments that have been tried in small series with variable results.^10-13 Haslund et al¹⁴ performed a systematic review of articles studying the treatment of HS with tumor necrosis factor inhibitors. They found that in the majority of cases the treatment was effective; however, in most publications, follow-up was insufficient to allow a systematic exploration, and therefore the authors recommended more standardized reporting of the outcomes as well as randomized controlled trials in HS disease. Surgical treatments range from incision and drainage procedures to debridement and wide surgical excision of interconnected tracts and abscesses in stage II and stage III disease.¹⁵ Other treatment modalities have been reported, including radiotherapy and carbon dioxide and diode laser treatments, as well as photodynamic therapy.^16,17 However, many of these treatment options arise from small retrospective or observational series, and well-designed, randomized controlled trials both with and without histologic characteristics are lacking.¹⁸ Effective treatments are limited, and recurrences often occur, even with wide surgical removal of involved tissue.⁵ A retrospective follow-up study¹⁹ described 34 patients with Hurley stage II HS who were treated with scanner-assisted carbon dioxide laser surgery. The mean duration of follow-up time after laser surgery was 34.5 months. The mean healing time was about 4 weeks, with 4 patients having recurrences at the surgical site. Twenty-five patients had flares of HS lesions in an area other than the treated site.¹⁹ Another study, by Hazen and Hazen,²⁰ describing the effect of treatment of HS using carbon dioxide laser excision and marsupialization, showed effective therapy for management of persistent or late-stage HS and minimal risk of recurrence within the treated areas.

Recently, our group published a report²¹ of a prospective, randomized controlled trial showing the effectiveness of the long-pulsed 1064-nm Nd:YAG laser in treating HS. We postulated that if the hair follicle is the nidus for recurrent chronic inflammation, then using the Nd:YAG hair removal laser would lead to disease improvement. The study²¹ was conducted in 22 patients with stage II to stage III disease and showed that after 3 monthly sessions with the laser, there was a statistically significant improvement in HS severity of 65.3% averaged over all body sites treated. In a follow-up study,²² our group showed that after 4 monthly laser sessions, patients were in remission for at least 2 months thereafter. We also provided preliminary pathologic evidence in 4 patients, suggesting that the laser works through selective photothermolysis of the follicular unit and subsequently decreases inflammatory lesions.²²

The primary goal of this study is to further examine the mechanism by which the 1064-nm Nd:YAG laser works to treat HS lesions by investigating histologic changes in the skin after laser treatment at different time points. A second goal is to determine whether clinical improvement in disease measured by the modified HS Lesion Area and Severity Index (LASI) scoring system described by Sartorius et al²³ (Table 1) correlated with improvement in inflammation seen on pathologic specimens.

Twenty patients (17 women and 3 men) ages 23 to 54 years clinically diagnosed as having HS were enrolled in the study conducted in the Department of Dermatology, Henry Ford Medical Center, Detroit, Michigan, from June 2008 to May 2009. This study was approved by our institutional review board. A sample size calculation was performed with power analysis (α = .05).

Patients with Hurley stage II disease (≥1 widely separated recurrent abscesses with tract formation and scars) were included in the study. Exclusion criteria were (1) concomitant use of systemic treatments for HS, which must be discontinued for 2 weeks prior to start of the study, and (2) exacerbation of the patient's original condition expressed clinically by a shift from Hurley stage II to stage III, which may necessitate systemic treatment. Topical treatments were allowed.

Two monthly laser sessions were performed using a long-pulsed 1064-nm Nd:YAG laser (Cutera Inc, Brisbane, California), treating the entire area with a single pulse and then double stacking on inflamed lesions at the first treatment and triple stacking lesions at the second treatment. A topical cooling gel was used, as well as ice packs before and after laser treatment. No anesthetic was used. A contact metal-tipped cooling device was immediately placed onto the area before and after irradiation (Cutera Inc). Settings were chosen based on Fitzpatrick skin type as follows: skin type I to III: spot size, 10 mm; pulse duration, 20 ms; fluence, 40 to 50 J/cm²; and skin type IV to VI: spot size, 10 mm; pulse duration, 35 ms; fluence, 25 to 35 J/cm².

Treatment response was scored using the modified HS LASI described by Sartorius et al²³ (Table 1). In addition, patient symptoms were assessed by adding scores for the extent of erythema, edema, pain, and purulent drainage at each anatomic site. Erythema and edema were graded on a visual analog scale of 0 to 3 (0, absent; 1, mild; 2, moderate; 3, severe). Purulent drainage was assessed through measuring the surface area that showed purulent discharge compared with the total surface area affected using the following scale: 0, no drainage; 1, 25% of involved area draining; 2, 50% of involved area draining; 3, 100% of involved area draining. Tenderness was graded by patient assessment of duration of daily pain as follows: 0, no pain; 1, 12 or less hours of pain daily; 2, more than 12 hours of pain daily; 3, pain wakes patient at night.²⁴

Patients were randomized according to anatomic site of disease: bilateral axilla, inframammary region, or groin. If only 1 anatomic site was involved, then 1 side was treated and the contralateral side served as the untreated control. If 2 anatomic sites were involved, then both sides were treated for 1 anatomic site and the other anatomic site served as control.

Four-millimeter skin punch biopsy specimens were obtained from active lesions on the treated site before laser treatment, after 24 hours (for the first 7 specimens) and at day 7, day 30, and day 60 (for all patients). On day 60, an active lesion from an untreated site was biopsied as control. All hematoxylin-eosin–stained slides were reviewed by a board-certified dermatopathologist (D.A.M.). The percentage of improvement was calculated for the laser-treated and control sides by comparing the modified HS-LASI score at baseline with the score at each laser treatment (on days 30 and 60).

The mean (SD) percentage of change was calculated for each patient over all anatomic sites combined and for each anatomic site individually. The P value was calculated for the mean percentage of change values using Wilcoxon signed rank. The Pearson correlation coefficient with accompanying P value was calculated for clinicopathologic correlation. The mean (SD) values for these correlation graphs were calculated with a paired t test to determine the resultant P value.

Twenty patients were enrolled in this study. One female patient did not complete the study because her active lesions resolved before the end of it. Nineteen patients completed the 2 months of study treatments. Five skin punch biopsy specimens were obtained from 18 patients, and 4 punch biopsy specimens were obtained from the remaining patient as she did not return for her 2-month visit because she moved to another state. There were a total of 94 pathology specimens. The mean age of the patients was 37 years, and there was a greater proportion of women (n = 16 [84.2%]) than men (n = 3 [15.8%]) in the study. Patients had Fitzpatrick skin types ranging from II to VI (Table 2). The distribution of treated anatomic sites included 11 patients (57.9%) with groin disease and 8 patients (42.1%) with axillary disease. For the 11 patients with inguinal disease, the entire groin area was treated, and 1 side (the left or right side of the groin) was chosen for serial biopsies. The more inflamed side was chosen as the biopsy site. The specimen for the control biopsy at month 2 was taken from a different involved site, including the axilla, suprapubic region, or gluteal folds. Of the 8 patients with axillary disease, 7 received laser treatments to both axilla, and 1 side (left or right axilla) was chosen for serial biopsies. The specimen for the control biopsy at month 2 was taken from an involved suprapubic or groin region. One patient with axillary disease as the only involved anatomic site received treatment to her left axilla, and her right axilla served as the control biopsy site at month 2.

The percentage of change in modified HS-LASI score after 2 months of treatment was –31.6% (P < .001) averaged over all anatomic sites, –24.4% (P = .008) for the axillary region, and –36.8% for the inguinal region (P = .001) (Figure 1). Overall, the mean HS-LASI scores trended downward from baseline to 1 month and 2 months after treatment for both axilla and inguinal sites (Table 3).

For our first 7 patients, serial biopsy specimens were taken at baseline, 24 hours, 1 week, and 2 months after the first laser treatment to determine the temporal changes that occurred at the histologic level after treatment (Table 4). At 24 hours, there was dense perifollicular inflammation with a mixed infiltrate of neutrophils, eosinophils, plasma cells, and giant cells. Granulation tissue was prominent, and only 1 specimen showed periapocrine gland inflammation. All specimens at this time point showed acute inflammation and did not differ substantially from control specimens. Therefore, for the next set of 12 patients, the 24-hour time point was removed from our protocol, and instead the serial biopsy specimens were taken at baseline, 1 week, 1 month, and 2 months after treatment. The 1-month time point was added for these remaining 12 patients.

At 1 week after treatment, there was more pronounced perifollicular inflammation and granulation tissue compared with baseline. A superficial and deep mixed inflammatory infiltrate was again present. At 1 month after treatment, the inflammatory reaction decreased with markedly less perifollicular inflammation, occasional mixed inflammatory infiltrate with eosinophils, and remaining dilated hair follicles, some of which contained broken hair shafts. At 2 months, the dermis was replaced by scarring and fibrosis, and inflammation was minimal. If present, the inflammation was in the deeper dermis and mostly spared the apocrine glands. Of note, most of the specimens at various time points showed follicular plugging.

For the patients who responded most to laser treatment, their clinical improvement in HS-LASI scores correlated with decreased inflammation and fibrosis seen on final pathologic findings at 2 months after laser treatment (Figures 1, 2, 3, and 4 show clinical and pathologic photographs of 2 representative patients). Overall, the response of 15 of the 18 patients transitioned from a marked superficial and deep inflammatory response seen on pathologic findings at baseline to a minimal inflammatory response (mainly deep if present) and/or scarring and fibrosis seen at 2 months after laser treatment. This change at the microscopic level corresponded with a decrease in the HS-LASI scores of these patients. This correlation is evidenced by 15 patients having higher scores at baseline (month 0) when corresponding pathologic findings showed superficial and deep inflammation, and lower scores at 2 months after treatment when pathologic findings showed minimal inflammatory reaction and/or scarring. The Pearson correlation coefficient was –0.481 (P = .007). The mean (SD) LASI score at baseline was 35.5 (13.6) and at month 2 was 21.7 (11.3). This change in mean LASI scores was significant with P < .001.

However, 3 patients continued to show inflammation at the microscopic level, and their HS-LASI scores did not show clinically significant improvement, although the statistical analysis for this group is limited by the small sample size. The Pearson correlation coefficient was –0.359 (P = .49). The mean (SD) LASI score at baseline was 33.7 (8.1) and at month 2 was 27.7 (10.8). This change in mean LASI scores was nonsignificant (P = .10), likely owing to the limited statistical power from the very small sample size of 3 patients.

Hidradenitis suppurativa is primarily a follicular disease with secondary apocrine gland involvement. In a groundbreaking study by Sellheyer and Krahl,³ 176 specimens of HS from the axillae or inguinal regions were reviewed. They noted that early lesions are similar to the open comedones in acne and show follicular hyperkeratosis and infundibula dilatation. Apocrine glands are not involved. At a later stage, the dilated follicular infundibulum ruptures, mainly at the inferior portion, spilling keratin debris into the surrounding dermis. This evokes an acute inflammatory response around the follicles. The acute neutrophilic infiltrate is gradually replaced by granulomatous infiltration with multinucleated foreign body giant cells. If the follicular rupture is more florid, an abscess forms, which may extend deep into the subcutaneous tissue. Inflammation around the apocrine gland occurs as an extension of the inflammatory process, leading to apocrine gland destruction. After extensive tissue destruction has occurred, naked hair shafts may remain. Residual follicular epithelium may proliferate in long-standing lesions and form sinus tracts, surrounded by fibrosis, which connect to the epidermis. These sinus tracts may also rupture, later forming additional coalescent sinus tracts, fistulas, and scarring.^3,5 A histologic study by Jemec et al²⁵ showed a homogeneity of HS lesions, in which no primary apocrine involvement was seen, and eccrine involvement was seen more often than apocrine involvement. Those findings again supported the reclassification of HS as a primary follicular disease. Boer and Weltevreden²⁶ suggested that the use of the term acne inversa is more appropriate than HS because the results of their study revealed that the primary event in HS is an infundibulofolliculitis with secondary involvement of apocrine glands. Another retrospective pathologic study of 118 skin resection specimens from 101 patients with HS showed that follicular occlusion by keratinous material, with subsequent active folliculitis and secondary destruction of the skin adnexae and subcutis, occurs as an integral step in the pathogenesis of HS.²⁷

Based on these pathologic studies, our group chose to use the long-pulsed Nd:YAG laser as a noninvasive treatment option for moderate HS.²² The wavelength of the long-pulsed 1064-nm Nd:YAG laser in the near-infrared electromagnetic spectrum selectively targets hair shafts and follicles via absorption by the melanin chromophore, while relatively sparing melanin in the superficial epidermis owing to its deeper dermal penetration of approximately 5 to 7 mm and lower absorption of melanin in the epidermis owing to its lower concentration at this wavelength.^28,29 This longer wavelength can penetrate deeper to reach the hair bulb in most anatomical areas than other lasers used for selective photothermolysis of hair follicles, such as the ruby (694-nm), alexandrite (755-nm), and diode (810-nm) lasers.^28-31 The long pulse duration is closer to the hair follicle's thermal relaxation time (10-100 ms).³⁰ Our previous study²² showed that this laser led to clinical improvement after 3 to 4 monthly treatments as well as remission of disease for up to 2 months after treatment cessation.

In this study, we sought to determine the mechanism of action of the Nd:YAG laser by examining histologic changes over time. On initial histologic findings before laser treatment, active lesions were noted to have a brisk mixed inflammatory infiltrate centered around the follicle with surrounding granulation tissue and some fibrosis. No major histologic difference was noted 24 hours after treatment from baseline; therefore, this data point was removed from our study for the remaining 12 patients. At 1 week after treatment, the superficial and deep inflammatory reaction increased, with more perifollicular inflammation and granulation tissue noted, suggesting acute effects of photothermolysis. Some patients reported an increase in tenderness and drainage during this first week, which was then followed by rapid disease improvement. This was noted to be clinically similar to a surgical incision and drainage procedure. One and 2 months after treatment, there was markedly decreased perifollicular inflammation, and the dermis was replaced by scarring and fibrosis in most patients. If present, inflammation involved the deeper dermis only and mostly spared the apocrine glands. Based on these findings, we postulate that the wavelength of the laser could not penetrate deeply enough to affect lesions in the lower dermis and dermal subcutaneous junction. However, most of the active lesions biopsied showed inflammation predominantly in the mid-reticular dermis, which may explain the effectiveness of the laser. The gross effects of the Nd:YAG laser on granulation tissue, when double pulsing, was to cause coagulation, Also, histologically there was increased coagulation of the granulation tissue. Our histologic analysis also showed that as early as 1 month after treatment, there was a significant decrease in inflammation, followed by fibrosis at 2 months after treatment. This suggests that Nd:YAG laser treatment may have resulted in the inflammatory lesions progressing more quickly through their cycle of acute, then granulomatous, inflammation to fibrosis, without the formation of abscesses and sinus tracts. Moreover, after 2 months, no recurrence of the superficial inflammation was present, signifying stabilization of disease. The Nd:YAG laser may also be decreasing inflammation by targeting water as a secondary minor chromophore and generating heat in the dermis via photothermolysis, and thus disrupting the inflammatory infiltrate. This may explain why even patients with Fitzpatrick skin types II and III and lighter hairs responded well to the laser treatments. The decrease in inflammation explains the relative decrease in the postinflammatory scarring and fibrosis that occurs as a final step in the disease progression. In addition, the photothermolysis theory also explains the improvement of the already fibrosed and scarred lesions by maximizing the tissue effects of normalization of both neocollagenesis and collagenolysis leading to optimum scar remodeling. A similar mechanism was demonstrated in a previous study³² of laser treatment of surgical scars. However, because of the small sample size of this group of patients, we could not perform a statistical analysis by Fitzpatrick type with enough power. It may be interesting to characterize the treatment response in larger subsets of patients with lighter skin types in future studies.

Our study further confirmed that clinical improvement in disease as measured by the HS-LASI score corresponded to lesion improvement seen on histopathologic characteristics. The overall percentage of change in modified HS-LASI score for all anatomical sites combined after only 2 monthly laser treatments was −31.6%. As with our previous study,²¹ both the inguinal (−36.8%) and axillary regions (−24.4%) responded well, suggesting that areas with dark, coarse terminal hairs respond to the follicular deocclusion mechanism of the laser. In the current study, there were no patients with mammary or perianal involvement, but our previous study²¹ showed that the mammary region improved the least, at 53.1% reduction in clinical severity compared with inguinal and axillary score severity reductions of 73.4% and 62%, respectively, after 3 laser treatments. In this study, 15 patients showed the best pathologic response, with minimal inflammatory reaction and/or scarring seen at 2 months after laser treatment. This response seen on pathologic examination correlated significantly with clinical improvement as measured by LASI scores (P = .007). Conversely, the 3 patients who continued to show inflammation 2 months after treatment did not have statistically significant changes in their LASI scores (P = .10), although the statistical power in this latter group was limited by the small sample size. Nevertheless, to our knowledge, this is the first study correlating the clinical and pathologic response to Nd:YAG laser treatment, thus elucidating the effectiveness of the laser from a mechanistic viewpoint.

One of the limitations of this study is sampling error of the biopsy sites. Biopsy specimens were taken from an area as close as possible (within 5 mm) to the previously biopsied site to follow the same area longitudinally on pathologic examination. However, this longitudinal follow-up was often difficult, because we could not take a biopsy specimen of exactly the same lesion at baseline and various time points. Also, disease severity may wax and wane owing to the natural course of HS, which complicated our interpretation of the pathologic data. Nevertheless, at follow-up visits, most patients were noted to have significant clinical improvement at 1 and 2 months after treatment. Because this clinical improvement as measured by LASI scores correlated with disease remission seen on pathologic findings, this suggests that our pathologic specimens closely depicted treatment response at the microscopic level. Another limitation is the small sample size; however, the sample size was calculated to be adequate based on a power analysis performed at the beginning of the study.

To our knowledge, this is the largest prospective, controlled histologic study examining the mechanism of the long-pulsed 1064-nm Nd:YAG laser in treating moderately severe (Hurley stage II) HS. Our data suggest that the Nd:YAG laser penetrates deeply enough for selective photothermolysis of the follicular unit and destruction of organized inflammatory lesions in the superficial to mid dermis. The laser seems to speed the healing of active lesions and subsequently induces remission. Our pathologic data further confirm that HS is primarily a follicular disorder, which explains the effectiveness of the long-pulsed Nd:YAG hair removal laser in treating this condition. This is also the first study showing clinicopathologic correlation following laser treatment. In conclusion, the 1064-nm Nd:YAG laser is a noninvasive, well-tolerated, and effective treatment option for patients with this chronic, debilitating disease.

Correspondence: Iltefat H. Hamzavi, MD, Department of Dermatology, Henry Ford Hospital, 3031 W Grand Blvd, Ste 800, Detroit, MI 48202 (ihamzav1@hfhs.org).

Accepted for Publication: July 20, 2010.

Published Online: September 20, 2010. doi:10.1001/archdermatol.2010.245

Author Contributions: All authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Xu, Wright, Mahmoud, Ozog, and Hamzavi. Acquisition of data: Xu, Wright, Mahmoud, Mehregan, and Hamzavi. Analysis and interpretation of data: Xu, Wright, Mahmoud, Mehregan, and Hamzavi. Drafting of the manuscript: Xu, Wright, Mahmoud, and Hamzavi. Critical revision of the manuscript for important intellectual content: Xu, Wright, Mahmoud, Ozog, and Hamzavi. Obtained funding: Xu. Administrative, technical, and material support: Wright, Mahmoud, and Ozog. Study supervision: Hamzavi.

Financial Disclosure: None reported.

Funding/Support: This study was supported by the Henry Ford Resident research grant.

Disclaimer: The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; or in the preparation, review, or approval of the manuscript.

Additional Contributions: We gratefully acknowledge the support from the HS-USA Foundation and the Hidradenitis Suppurativa Foundation. Gordon Jacobsen, BS, performed the statistical analysis.