P5954Standardizing laboratory monitoring of patients undergoing isotretinointreatment using a cost-effective and best evidence approach

Timothy Hansen, MD, Penn State Dermatology, Hershey, PA, United States;Andrea Zaenglein, MD, Penn State Dermatology, Hershey, PA, United States;Jeffrey Miller, MD, MBA, Penn State Dermatology, Hershey, PA, United States

Background: Prescribers of isotretinoin use varied protocols for laboratory moni-toring and dose escalation for their patients. A survey of prescribers in ourdepartment revealed 6 different monitoring patterns. The goal of this study is todevelop a standardized approach to laboratory monitoring of acne patients beingtreated with isotretinoin, thereby eliminating unnecessary testing and reducingoverall cost while continuing to safely administer the medication.

Results: 515 acne patients treated with 574 courses of isotretinoin were identifiedusing the iPLEDGE database and chart review from May 2003 through July 2011 atPennsylvania State University Hershey Medical Center Department of Dermatology.The following laboratory abnormalities during treatment were identified: hypertri-glyceridemia (24.1%), hypercholesterolemia (26.8%), elevated liver transaminases(5.4%), leukopenia (2.7%), and thrombocytopenia (1.6%). Mean duration of treat-ment before abnormalities were detected was 56.33 days for hypertriglyceridemia,50.1 days for hypercholesterolemia, and 61.9 days for alanine transaminitis.Leukopenia and thrombocytopenia remained stable or resolved over time despitecontinued treatment. Clinically significant laboratory abnormalities (grade 2 asdefined by the National Cancer Institute Common Terminology Criteria for AdverseEvents v3.0) requiring alterations in therapy were noted in 8 (1.6%) patients withhypertriglyceridemia, 3 (0.6%) patients with transaminitis, and 1 (0.2%) patient withhypercholesterolemia. 67.7% (95% confidence interval, 60.7%%-74.1%) of labora-tory abnormalities are detected within 60 days of starting isotretinoin, and only 9.5%(95% confidence interval, 5.8%%-14.4%) are detected within 30 days.

Conclusion: A standardized approach to isotretinoin laboratory monitoring willenable monitoring of outcomes and quality more effectively and minimize cost andpatient discomfort. Based on our data and review of the literature, the followinglaboratory monitoring is our initial recommendation: baseline measurement ofserum lipids and transaminase levels, followed by repeat lipid levels at 1month and 2months, with repeat transaminase levels at least 2 months into therapy. A completeblood count may be unnecessary, unless symptoms warrant. Additional testingshould be individualized based on patient symptoms, history, and comorbidities.

AB6

cial support: None identified.

Commer

PD04—SKIN CANCER

P65929p21 loss detected by FISH is associated with BRAF mutation in advancedmelanoma

Beatrice Nardone, MD, PhD, Department of Dermatology, NorthwesternUniversity, Chicago, IL, United States; Alfred Rademaker, PhD, Department ofPreventive Medicine, Northwestern University, Chicago, IL, United States;Pedram Gerami, MD, Department of Dermatology, Northwestern University,Chicago, IL, United States; Pedram Yazdan, MD, Department of Dermatology,Northwestern University, Chicago, IL, United States; Victoria Godinez-Puig, MD,Department of Dermatology, Northwestern University, Chicago, IL, United States

Background: Understanding the molecular pathways and mutations in cutaneousmelanoma (CM) may help stratify cases into subgroups with similar molecularfeatures, treatment responses and prognosis. A number of specific gene mutationsand chromosomal copy number aberrations have been described in melanoma. Inthis study, our goal was to look at links between specific gene mutations andsubsequent chromosomal copy number aberrations and evaluate their impact onprognosis.

Methods: A case control study using archived melanoma samples fromNorthwestern University was performed. The cases consisted of 48 melanomasfrom 48 patients with documented metastasis as well as 36 melanomas from 36patients with a minimum follow-up of 5 years and no disease recurrence/metastasis.All patients had a Breslow depth of at least 0.55 mm (except for 1 case frommetastatic group). FISH for chromosomal aberrations frequently seen in melanomaand mutation analysis for BRAF V600E mutations was performed for each specimen.Medical records and slides were reviewed to obtain clinical course, Breslow depth,age, sex, anatomic site of melanoma, presence or absence of ulceration, mitoticcount, and Clark’s level were recorded.

Results: Among the total 84 melanoma cases, BRAF mutation was present in 49%(41/84). No statistically significant differencewas foundwhen comparing frequencyof BRAF mutations in metastasizing versus nonmetastasizing CM groups. Among the11markers studied by FISH in all cases, only 9p21 frequency was significantly higherin the BRAF mutation samples compared to wild type samples (P ¼ .013).

Conclusion: Our data suggest that the presence or absence of a BRAF mutation doesnot correlate with the likelihood for metastasis or disease progression in advancedmelanoma. Moreover, the high frequency of 9p21 loss in the BRAF mutation groupsuggests that these 2 molecular aberrations may often be found together and becomplementary aberrations that may lead to the development of melanoma. Ascombined therapy is frequently required during melanoma treatment, understand-ing which pathways are likely to be working cooperatively may help orientcombination therapy towards the most relevant molecules.

cial support: None identified.

Commer

J AM ACAD DERMATOL

P6483Characterizing peritumoral clefts in basal cell carcinoma with histologicstaining and reflectance confocal microscopy

Anna Levin, MD, Mayo Clinic Arizona, Scottsdale, AZ, United States; DavidDiCaudo, MD, Mayo Clinic Arizona, Scottsdale, AZ, United States; DavidSwanson, MD, Mayo Clinic Arizona, Scottsdale, AZ, United States

Background: Basal cell carcinoma (BCC) is histologically characterized by basaloidtumor noduleswith peripheral palisading of cells and nuclei. Frequently, nodules areseparated from surrounding connective tissue by empty appearing clefts. Typically,these are assumed to be an artifact of tissue preparation. The recently developedreflectance confocal microscopy (RCM) has been proposed as a new tool in theevaluation of dermatologic lesions in vivo, including BCC. Previous research hassuggested that small areas of low refractility separating tumor islands fromsurrounding tissue in vivo correlated with peritumoral clefts seen on routinehistology and may correspond to deposited, amorphous material such as mucin.Mucin itself is a mixture of acidic glycosaminoglycans normally produced in smallquantities by the skin to maintain fluid and salt balance, but is also a component ofpathological cutaneous conditions. It may be present as dermal mucin, produced byfibroblasts, or epithelial mucin, usually found in the eccrine and apocrine secretorycells of the skin. Identification of the substance in peritumoral clefts may provideclues to its role in BCC. Our objective is to correlate histological clefting to RCMsamples, and furthermore to study the staining patterns of this material, allowingfurther classification as mucin of dermal or epithelial origin.

Methods: A database of RCM images was evaluated and narrowed down to 23 BCCswhich showed optimal peritumoral clefting. Staining was performed on corre-sponding histological slides with toluidine blue, Alcian blue, and periodicacideSchiff (PAS) for the detection and characterization of clefting and peritumoralmucin.

Results: Fifteen of 23 samples revealed clefting on routine histology. An additional 2samples were indeterminate. Eight of the 15 stained positively for mucin with Alcianblue, none of the samples were positive for PAS or toluidine blue.

Conclusion: Our results show that peritumoral clefts in BCC are present in vivo. Theabsence of clefting on routine histology in samples with postive RCM images may becaused by sampling error in the preparation of histologic slides. However, theinconsistent presence of mucin suggests that mucin is not only substance present inthese clefts. In addition, the staining patterns were not consistent with knownpatterns of dermal or epithelial mucin, and more information regarding the materialwithin the clefts may be elucidated in further study.

cial support: None identified.

Commer

P5959Correlation of hOGG1 protein expression and 8-OhDG adduct levels inactinic keratosis and cutaneous squamous cell carcinoma

Yu-Ping Hsiao, MD, Institute of Medicine, Chung Shan Medical University, andDepartment of Dermatology, Chung Shan Medical University Hospital, Taichung,Taiwan; Chung-Hung Tsai, PhD, Department of Pathology, Chung Shan MedicalUniversity Hospital, Taichung, Taiwan, Taichung, Taiwan; Jen-Hung Yang, PhD,School of Medicine, Tzu Chi University, and Department of Dermatology,Buddhist Tzu Chi General Hospital, Hualien, Taiwan, Hualien, Taiwan; Ya-WenCheng, PhD, Institute of Medicine, Chung Shan Medical University, Taichung,Taiwan, Taichung, Taiwan

Background: 8-hydroxy-29-deoxyguanosine (8-OHdG) is a photoproduct of oxida-tively damaged DNA formed by UV radiation. Human genome 8-Oxoguanineglycosylase (hOGG1) is a base excision repair system and a key componentresponsible for removal of 8-OHdG. Although the role of 8-OHdG and hOGG1 in thelung and kidney malignances had been defined, their role in the neoplastic cascadein epithelial cancer is not entirely clear.

Objective: To investigate the oxidative DNA damage (8-OHdG adducts) and repairgene (hOGG1) expressions in carcinogenesis of epithelial skin cancer.

Methods: From 2002 to 2011, formalin-fixed paraffin-embedded tissues of cutaneoussquamous cell carcinoma (cSCC; n ¼ 78), actinic keratosis (AK; n ¼ 34), and normalskin samples (NS; n¼ 85)were collected. The studywas approved by the Chung ShanMedical University Hospital (IRB No CS11077) institutional review board. Theparaffin-embedded normal skin samples were obtained from the facial benignsubcutaneous tumors and excluded malignancy and autoimmune diseases by chartreview. Immunohistochemistry was performed using antibodies against 8-OHdG andhOGG1 in cSCC, AK, and NS. Nuclear and cytoplasmic expressions were assessed byintensity (SI) and percentage positivity (PC). Themean of the total score, calculated byadding intensity and percentage positivity, was used for statistical analyses.

Results: Themean age of cSCC, AK, and NSwere 72.3, 75.6, and 42.6. There were nogender differences between groups. hOGG1-expression in the nucleus of cSCC(mean, 0.65 6 1.19) was significantly lower than in AK (mean, 2.00 6 1.30; P\.001) and higher than in normal skin (mean, 3.33 6 1.04; P\.001). A statisticallysignificant inverse linear trend of cytoplasmic hOGG1 expression in cSCC (mean,0.53 6 1.08), AK (mean, 2.00 6 1.65), and NS (mean, 2.59 6 1.12) was alsoobserved (P\.001). In cSCC, AK, and NS, mean nuclear 8-OHdG expression was4.05, 5.56, and 3.72, respectively (P ¼ no significance for cSCC versus AK and AKversus NS) and cytoplasmic 8-OHdG expression was 3.49, 3.03, and 0.84, respec-tively (P\.01 for cSCC versus AK, and AK versus NS).

Conclusion: From the clinicopathologic point of view, our results provide evidencethat the expression of cytoplasmic 8-OHdG in AK and cSCC could be an alternativepathological marker in mitochondrial damage and correlate with skin carcinogen-esis. Reduced cytoplasmic and nuclear hOGG1expressions in cSCC and AK suggeststhat hOGG1 inactive may be of relevance to disease progression.

cial support: None identified.

Commer

APRIL 2013