Hydroxychloroquine and lichen planopilaris: Efficacy and introduction of Lichen Planopilaris Activity Index scoring system

ORIGINAL

ARTICLES

Hydroxychloroquine and lichen planopilaris: Efficacy and introduction of Lichen Planopilaris Activity Index scoring system Charles Chiang, MD,a Deborah Sah, MD,b Bryan K. Cho, MD,c Blanca E. Ochoa,d and Vera H. Price, MDa San Francisco, Fremont, and Mountain View, California; and Houston, Texas See commentary on page 398 Background: Lichen planopilaris (LPP) and its variant frontal fibrosing alopecia (FFA) are primary lymphocytic cicatricial alopecias for which there is no evidence-based therapy. Objective: We assessed the efficacy of hydroxychloroquine in active LPP and FFA using the LPP Activity Index (LPPAI), a numeric score that allows quantification of the symptoms and signs of the condition for statistical comparison. In addition, we determined with the LPPAI if any improvement (reduction) in the numeric score pretreatment and posttreatment reached statistical significance. Methods: This was a retrospective, single-center chart review of 40 adult patients with LPP, FFA, or both who were treated with hydroxychloroquine for up to 12 months from 2004 to 2007 at the University of California, San Francisco Hair Center. Symptoms, signs, activity, and spreading were scored at each visit in the standardized cicatricial alopecia flow chart. A numeric score was assigned to these markers of disease activity and a numeric score was calculated at each visit. Results: There was significant reduction (P \.001) in the LPPAI at both 6 and 12 months. After 6 months, 69% had improved (reduced) symptoms and signs. At 12 months, 83% had improvement (reduction) in symptoms and signs. Limitations: Retrospective analysis and uncontrolled study are limitations. Conclusions: Hydroxychloroquine is effective in decreasing symptoms and signs in LPP and FFA as shown by significant reduction in the LPPAI in 69% and 83% of patients after 6 and 12 months of treatment, respectively. ( J Am Acad Dermatol 2010;62:387-92.) Key words: cicatricial alopecia; frontal fibrosing alopecia; hydroxychloroquine; lichen planopilaris; Lichen Planopilaris Activity Index.

L

ichen planopilaris (LPP) is a primary cicatricial alopecia caused by chronic lymphocytic inflammation around the upper portion of the

From the Department of Dermatology at University of California, San Franciscoa; Palo Alto Medical Foundation, Fremont Center, Fremontb; Department of Dermatology at Palo Alto Medical Foundation, Camino Medical Group, Mountain Viewc; and Department of Dermatology at Baylor College of Medicine, Houston.d Funding sources: None. Conflicts of interest: None declared. Accepted for publication August 14, 2009. Reprint requests: Vera H. Price, MD, Department of Dermatology at University of California, San Francisco, 1701 Divisadero St, Third Floor, San Francisco, CA 94143. E-mail: [email protected]. Published online January 11, 2010. 0190-9622/$36.00 ª 2009 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2009.08.054

Abbreviations used: FFA: LPP: LPPAI: UCSF:

frontal fibrosing alopecia lichen planopilaris Lichen Planopilaris Activity Index University of California, San Francisco

hair follicle. The origin of LPP, and the other primary cicatricial alopecias, remains poorly understood, but they all have in common a targeted folliculocentric attack, which leads to irreversible follicular destruction and permanent hair loss. In 2001, the North American Hair Research Society proposed a working classification of the primary cicatricial alopecias and separated them into lymphocytic, neutrophilic, and mixed groups based on the predominant inflammatory cellular infiltrate.1 LPP has become the prototype 387

388 Chiang et al

J AM ACAD DERMATOL MARCH 2010

of the lymphocytic group, although it is rare. Its 4 mmedeep punch biopsy specimen down to subreported annual incidence rates in hair referral cencutaneous fat is submitted for horizontal sectioning ters in the United States ranges from 1.15% to 7.59% and hematoxylin-eosin staining. Compared with among all new patients with hair loss.2 Other primary vertical sectioning where 3 to 4 follicles may be lymphocytic cicatricial alopecias include frontal acquired, horizontal sectioning allows up to 30 fibrosing alopecia (FFA), central centrifugal follicles to be examined. The biopsy site is an active, cicatricial alopecia, pseudopelade (Brocq), and symptomatic hair-bearing area with perifollicular chronic cutaneous lupus eryerythema and perifollicular thematous.1 scale, located at the margin CAPSULE SUMMARY Until now, there have of a bare patch, with a posibeen no evidence-based tive anagen pull test result Hydroxychloroquine is significantly studies validating current when possible. effective, after 6 and 12 months of treatments for the primary Histologic features were treatment, in decreasing symptoms and cicatricial alopecias3 because characteristic of LPP and signs in lichen planopilaris and frontal of the lack of meaningful end showed an interface lichenfibrosing alopecia. 4 points. Past studies analyzoid dermatitis affecting the More patients respond to ing the efficacy of hydroxyfollicular epithelium. A varihydroxychloroquine after 12 months chloroquine in LPP5-7 used ably dense lymphocytic infilthan 6 months. subjective end points such trate surrounded the as ‘‘positive’’ and ‘‘fair’’ that infundibulum, isthmus, and A standardized flow chart is essential to are difficult to interpret and bulge region of active aftrack symptoms and signs and facilitate validate. In this retrospective fected hair follicles.8 Loss of treatment decisions. sebaceous glands and perisurvey we report our experiThe Lichen Planopilaris Activity Index is follicular lamellar fibrosis ence with hydroxychlorointroduced as a numeric summary of were invariably present.9 quine in 40 patients with symptoms and signs, and allows The inflammatory reaction LPP and FFA seen at the statistical comparison of disease activity eventually results in permaUniversity of California, San pretreatment and posttreatment. nent destruction of the folliFrancisco (UCSF) between cle and replacement with October 2004 and October fibrotic scar tissue, manifested clinically as loss of 2007. We also introduce the LPP Activity Index follicular ostia. (LPPAI), which allows numeric scoring of the symptoms, signs, activity, and spreading of the condition for statistical comparison. Using the LPPAI, we show Frontal fibrosing alopecia that treatment with hydroxychloroquine resulted in FFA is currently considered a variant of LPP.1 It is statistically significant improvement. characterized by frontal hairline recession, perifollicular erythema and scaling around solitary residual terminal hairs along the receded hairline and frontal Lichen planopilaris scalp, and loss of follicular ostia (Fig 2). Partial or Our 40 patients developed LPP at a mean age of 53 complete loss of eyebrows, present in 9 of 11 of our years; 80% were female and 20% were male. LPP patients, is common and may precede, accompany, usually presents as irregular patchy hair loss, with or follow the hairline recession. Occasionally paloss of follicular ostia, a hallmark of all cicatricial tients have both FFA and LPP, including 4 patients in alopecias. Less commonly the hair loss is diffuse this study (Table I). Histologic features are similar to rather than patchy. Perifollicular erythema and perthose of LPP. Although FFA was first described in ifollicular scale are typically present at the periphery postmenopausal women,10 at the UCSF Hair Center of active lesions (Fig 1). Cutaneous lichen planus between October 2003 and May 2009 35 women papules may develop before, during, or after the presented with FFA and, of these, 6 were premenoonset of LPP. Oral and genital lesions may also occur. pausal (Price et al, unpublished data). This is conMany patients have a history of scalp scaling, often sistent with a recent FFA series.11 considered seborrheic dermatitis, for many years before the diagnosis of LPP. Active, untreated LPP is Hydroxychloroquine often intensely symptomatic with severe pruritus, Hydroxychloroquine is an antilymphocytic, antipain, tenderness, and burning. malarial drug that has been widely used since the Diagnosis of LPP always requires clinicopatho1950s for a range of dermatologic conditions includlogic correlation and cannot be made by the above ing lupus erythematosus, porphyria cutanea tarda, clinical signs and symptoms alone. At least one d

d

d

d

Chiang et al 389

J AM ACAD DERMATOL VOLUME 62, NUMBER 3

Table I. Demographics of 40 study patients

Fig 1. Patchy hair loss in lichen planopilaris. Note loss of follicular ostia in central bare areas and perifollicular scale and perifollicular erythema at periphery of patches.

Sex Male Female Race Caucasian Asian Hispanic Unclassified Diagnosis LPP Age of onset, mean [range] FFA Age of onset, mean [range] Both LPP and FFA Age of onset, mean [range]

20% (n = 8) 80% (n = 32) 85% (n = 34) 8% (n = 3) 2.5% (n = 1) 5% (n = 2) 72.5% (n = 29) 54 y [19-74] 17.5% (n = 7) 62 y [40-77] 10% (n = 4) 56 y [52-64]

FFA, Frontal fibrosing alopecia; LPP, lichen planopilaris.

LPP Activity Index

Fig 2. Frontal fibrosing alopecia showing perifollicular erythema and recession along frontal hairline. Note presence of freckles on lower chronically sun-exposed forehead.

polymorphous light eruption, dermatomyositis, cutaneous sarcoidosis, granuloma annulare, and lichen planus.12 It is generally well tolerated. Side effects include gastrointestinal symptoms (nausea, vomiting, diarrhea), neuromuscular symptoms (headaches, myalgia, fatigue), and blue-gray to black skin pigmentation that resolves after cessation of therapy. Patients with glucose-6-phosphate deficiency and porphyria cutanea tarda are rare but can experience hemolysis and acute hepatitis, respectively. Reversible leukopenia can also occur. The American Academy of Ophthalmology hydroxychloroquine screening recommendations published in 2002 included a baseline retinal examination.13,14 A large retrospective study of 1207 patients observed no retinal toxicity from hydroxychloroquine.15 Our standard protocol for patients who start treatment with hydroxychloroquine is to obtain a complete blood cell count, liver function tests, and baseline ophthalmologic examination including a retinal exanimation. These blood tests are not usually repeated, and the follow-up retinal examinations are based on the recommendation of the ophthalmologist and are invariably repeated after 6 to 12 months.

The UCSF Hair Center has been using a cicatricial alopecia standardized flow chart to document the course and response to treatment at every visit in patients with cicatricial alopecia16 (Fig 3). The standardized flow chart records symptoms (pruritus, pain, burning) and signs (erythema, perifollicular erythema, perifollicular scale), a measure of activity (the anagen pull test), and spreading of the condition. We assigned numeric values to these subjective and objective surrogate markers of LPP to establish a numeric summary of disease activity, the LPPAI. This provided a quantitation of pretreatment and posttreatment responses that could be computed and statistically compared. The weights given to the symptoms (30%), signs (30%), anagen pull test (25%), and presence of spreading (15%) led to the equation: LPPAI (0-10) = (pruritus 1 pain 1 burning)/3 1 (scalp erythema 1 perifollicular erythema 1 perifollicular scale)/3 1 2.5 (pull test) 1 1.5 (spreading/2). Symptoms and signs are recorded on a 4-point scale: 0 = absent, 1 = mild, 2 = moderate, and 3 = severe. The anagen pull test, when present, is a reliable measure of local disease activity. It involves grasping a small group of 10 to 20 hairs between the thumb, second finger, and third finger at the scalp end of the hair shafts, and pulling away from the scalp with a slow, firm perpendicular force to slide the fingers to the ends of the hair. The result is recorded both as a binary value (0 for no anagen hairs and 1 for the presence of anagen hairs) and as anagen hairs/total hairs pulled. Last is the assessment of spreading, recorded as 0 (no spreading) versus 1 (indeterminate) versus 2 (spreading). When the hair loss is difficult to judge, the issue of spreading is recorded as indeterminate.

390 Chiang et al

J AM ACAD DERMATOL MARCH 2010

Fig 3. Cicatricial alopecia standarized flow chart. Alk Phos, Alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BUN, blood urea nitrogen; CBC, complete blood count; CR, creatinine; F/u, follow up; G6PD, glucose-6-phosphate dehydrogenase; LPPAI, Lichen Planopilaris Activity Index; pt, patient.

METHODS Patient selection Medical records of patients seen between October 2004 and October 2007 with a diagnosis of lichen planus (International Classification of Diseases, Ninth Revision 697.0) were carefully reviewed after receiving institutional review board approval. The

database search obtained 214 records with a diagnosis of LPP or FFA. We excluded patients whose clinical and histologic findings changed over time, whose biopsy specimen showed neutrophils or plasma cells at any time, or who did not have follow-up or documentation during the 6- and 12month windows. The inclusion criteria for the study

Chiang et al 391

J AM ACAD DERMATOL VOLUME 62, NUMBER 3

Table II. Results of hydroxychloroquine treatment after 6 and 12 months 6 mo

LPP FFA LPP and FFA Total

N

Responders

27 7 4 38*

0 14% (1/7) 0 3% (1/38)

Partial responders

70% 57% 50% 66%

(19/27) (4/7) (2/4) (25/38)

12 mo Nonresponders

30% 29% 50% 32%

(8/27) (2/7) (2/4) (12/38)

N

Responders

29 7 4 40

14% 57% 25% 23%

(4/29) (4/7) (1/4) (9/40)

Partial responders

72% 29% 25% 60%

(21/29) (2/7) (1/4) (24/40)

Nonresponders

14% 14% 50% 18%

(4/29) (1/7) (2/4) (7/40)

FFA, Frontal fibrosing alopecia; LPP, lichen planopilaris; nonresponders, LPPAI decreased \25%; partial responders, LPPAI decreased 25%-85%; responders, LPPAI decreased [85%. *Two patients did not follow-up at 6 mo.

were patients with a scalp biopsy specimen that confirmed a primary lymphocytic cicatricial alopecia, who had good clinicopathologic correlation, and had treatment with hydroxychloroquine (200 mg twice daily) initiated by the UCSF Hair Center. Forty patients with LPP, FFA, or both fulfilled these requirements and were included in the study (Table II). Calculation of the LPPAI The LPPAI was calculated pretreatment and at 6 and 12 months posttreatment using the standardized flow chart (Fig 3). Because of the far distances that patients often travel to the UCSF Hair Center, they often cannot be followed up at precise intervals as a result of scheduling and travel constraints. Thus, patients seen between 3 and 9 months after beginning hydroxychloroquine were grouped into the 6month window for statistical comparison. Patients seen between 9 and 15 months were included in the 12-month window. Response definitions Patients with an LPPAI reduction greater than 85% over pretreatment values were considered responders, with 25% to 85% reduction were considered partial responders, and with less than 25% reduction were considered nonresponders. The physician assessment was made by a single observer (V. H. P.) at our referral center. It was recorded separately from the LPPAI and was a clinical evaluation of whether the patient was well controlled, moderately controlled, or uncontrolled. Patients were considered well controlled if symptoms had resolved and hydroxychloroquine was discontinued. Patients were considered moderately controlled if symptoms and signs were improved but required continued hydroxychloroquine. Patients were classified as uncontrolled if they were not improved after 3 months and required changing to an alternative systemic medication.

Data analysis Pretreatment LPPAI scores were compared with 6- and 12-month posttreatment LPPAI scores for each patient. The UCSF Biostatistics, Research Design, Ethics, and Data Management used a paired t test to determine if there was a significant change between pretreatment and posttreatment LPPAI scores. Cohen’s Kappa is a statistic used to assess agreement between two categorical measures. In this study, UCSF Biostatistics, Research Design, Ethics, and Data Management calculated it to assess agreement between patient response category determined by LPPAI score (responder, partial responder, or nonresponder) and category by physician assessment (well controlled, moderately controlled, or uncontrolled). The presence or absence of significant agreement was used to assess the validity of the LPPAI score as a measure of clinical outcome.

RESULTS The LPPAI scores improved significantly (P \ .001) from pretreatment to 6- and 12-month posttreatment. After 6 months, 69% of patients (26 of 38) were full or partial responders. Two patients did not follow up (Table II). After 12 months, 83% of patients (33 of 40) were full or partial responders. Of the 9 responders at 12 months (Table II), 6 were able to discontinue hydroxychloroquine and had a sustained remission with no systemic medication for at least 1 year. UCSF Biostatistics, Research Design, Ethics, and Data Management calculated a Cohen’s Kappa statistic to assess agreement between the two categorical measures of patient response category determined by LPPAI score (responder, partial responder, or nonresponder) versus category by physician assessment (well controlled, moderately controlled, or uncontrolled). The Cohen’s Kappa statistic of 0.483 suggested moderate agreement (95% confidence interval: 0.22-0.74).17 This statistically significant (P \ .05) agreement between the

392 Chiang et al

J AM ACAD DERMATOL MARCH 2010

two categories validated using the LPPAI score as a measure of clinical outcome. Adverse events One patient had headaches that resulted in the patient withdrawing from treatment at 4 months and the headaches resolved. No retinopathy was encountered in any patient.

DISCUSSION Before this retrospective review, the efficacy of hydroxychloroquine in treating LPP was uncertain largely because of the lack of a systematic way of monitoring patients while on treatment. Clinically, notes alone are not adequate in following up the symptoms, signs, and progression of LPP. The standardized cicatricial alopecia flow chart that we have used for several years17 has been key in monitoring patient progress. By assigning numeric values to the subjective and objective markers in the flow chart, a single numeric summary of disease activity, the LPPAI, can be calculated. This provides both clinician and patient an outcome measure with which to monitor the efficacy of LPP treatment. Hydroxychloroquine improved the LPPAI scores significantly (P \ .001) in 69% and 83% of patients after 6 and 12 months, respectively. It is noteworthy that the nonresponders had high LPPAI pretreatment scores of 6.7 to 8.7 compared with the average pretreatment LPPAI score of 4.5. This is the first evidence-based study reporting the significant (P\.001) efficacy of hydroxychloroquine in the treatment of LPP and FFA after 6 and 12 months of treatment. We have introduced a numeric score, the LPPAI, that allows statistical comparison of pretreatment and posttreatment response. It is our hope that by using a standardized flow chart and the LPPAI scoring system, dermatologists will have a useful numeric summary of LPP activity to document pretreatment and posttreatment response and to guide treatment decisions.

REFERENCES 1. Olsen E, Bergfeld W, Cotsarelis G, Price V, Shaprio J, Sinclair R, et al. Summary of North American Hair Research Society (NAHRS)-sponsored workshop on cicatricial alopecia, Duke University Medical Center, February 10 and 11, 2001. J Am Acad Dermatol 2003;48:103-10. 2. Ochoa B, King L, Price V. Lichen planopilaris: annual incidence in four hair referral centers in the United States. J Am Acad Dermatol 2008;58:352-3. 3. Shapiro J. Cicatricial alopecias. Dermatol Ther 2008;21:211. 4. Harries M, Sinclair R, Macdonald-Hull S, Whiting D, Griffiths C, Paus R. Management of primary cicatricial alopecias: options for treatment. Br J Dermatol 2008;159:1-22. 5. Chieregato C, Zini A, Barba A, Magnanini M, Rosina P. Lichen planopilaris: report of 30 cases and review of the literature. Int J Dermatol 2003;42:342-5. 6. Mehregan D, Van Hale H, Muller S. Lichen planopilaris: clinical and pathologic study of forty-five patients. J Am Acad Dermatol 1992;27:935-42. 7. Cervasco N, Bergfeld W, Remzi B, de Knott H. A case-series of 29 patients with lichen planopilaris: the Cleveland Clinic Foundation experience on evaluation, diagnosis, and treatment. J Am Acad Dermatol 2007;57:47-53. 8. Templeton S, Solomon A. Scarring alopecia: a classification based on microscopic criteria. J Cutan Pathol 1994;21:97-109. 9. Tan E, Martinka M, Ball N, Shapiro J. Primary cicatricial alopecias: clinicopathology of 112 cases. J Am Acad Dermatol 2004;50:25-32. 10. Kossard S. Postmenopausal frontal fibrosing alopecia: scarring alopecia in a pattern distribution. Arch Dermatol 1994;130:770-4. 11. Tan K, Messenger A. Frontal fibrosing alopecia: clinical presentation and prognosis. Br J Dermatol 2009;160:75-9. 12. Kalia S, Dutz J. New concepts in antimalarial use and mode of action in dermatology. Dermatol Ther 2007;20:160-74. 13. Marmor MF, Carr RE, Easterbrook M, Farjo AA, Mieler WF. American Academy of Ophthalmology. Recommendations on screening for chloroquine and hydroxychloroquine retinopathy: a report by the American Academy of Ophthalmology. Ophthalmology 2002;109:1377-82. 14. Marmor MF. New American Academy of Ophthalmology recommendations on screening for hydroxychloroquine retinopathy. Arthritis Rheum 2003;48:1764. 15. Levy G, Munz S, Paschal J, Cohen H, Pince K, Peterson T. Incidence of hydroxychloroquine retinopathy in 1,207 patients in a large multicenter outpatient practice. Arthritis Rheum 1997;40:1482-6. 16. Mirmirani P, Willey A, Price VH. Short course of oral cyclosporine in lichen planopilaris. J Am Acad Dermatol 2003;49:667-71. 17. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-74.