Spindle Cell Squamous Carcinoma During BRAF Inhibitor Therapy for Advanced Melanoma: An Aggressive Secondary Neoplasm of Undetermined Biologic Potential

Secondary cutaneous squamous cell carcinomas (cSCCs) are adverse effects of BRAF inhibitor targeted therapy for advanced melanoma. The histologic type most commonly reported during vemurafenib and dabrafenib mesylate BRAF inhibitor therapy is well-differentiated keratoacanthoma-like cSCC (cSCC-KA).^1-3 Lesions of cSCC-KA in BRAF inhibitor therapy are considered to have a low biologic potential (ie, infrequent metastasis or recurrence). Whereas the early follow-up (<3 years) in most patients seems to support this claim, it is unproven. Furthermore, optimal intervention for secondary BRAF inhibitor cSCC is not established. We report multiple biphasic spindle cell squamous carcinomas, also known as sarcomatoid squamous cell carcinomas, in a patient receiving BRAF inhibitor therapy.

A woman in her 50s with no history of cSCC presented in 2004 with a left lower extremity Clark level IV, 1.3-mm Breslow depth melanoma with no disease found in a sentinel lymph node biopsy (T2aN0 stage IB). She presented in 2009 with upper extremity weakness and had multiple brain metastases and adenopathy. A resected pelvic node harbored V600E BRAF–mutant metastatic melanoma, and she was treated with whole brain and stereotactic radiation. In 2011, a brain lesion grew and she was enrolled in a BRAF inhibitor (dabrafenib [GSK2118436]; clinicaltrials.gov Identifier: NCT01266967) clinical trial. Within 1 week of starting therapy (150 mg, twice daily), she developed facial and extremity acrochordons and new nevi on her torso and axillae, along with fever, chills, and fatigue. She experienced enlarging, tender and bleeding lesions on the trunk and extremities. On examination at 4 weeks, she had more than 100 new cutaneous squamous proliferations throughout her body (Figure 1). New verrucous red lesions also erupted on the shoulders, as well as rough lesions on the face. Seven large, tender and indurated lesions were removed for histopathologic assessment using “deep scoop” shave biopsy.

Histologically, lesions revealed a biphasic malignant growth pattern. The superficial portion demonstrated conventional cSCC-KA features of hyperkeratosis, epidermal acanthosis, and central core of glassy eosinophilic keratin with pseudopapillomatosis and a base with focal invasive lobules of cytologically atypical keratinocytes (Figure 2A), consistent with previously reported cSCC-KA. In stark contrast to prior reports, the deep aspects of 6 of 7 lesions showed invasive spindled and epithelioid cells with monomorphic elongated nuclei with condensed chromatin and mitoses (Figure 2B and eFigure 1 in Supplement) consistent with spindle cell squamous carcinoma, an aggressive subtype of squamous cell carcinoma.

Superficial and spindled tumor cells were strongly immunoreactive for cytokeratin CK5/6 and CK903 (eFigure 2A and 2B in Supplement). Both squamous and spindled components were vimentin reactive (eFigure 2C in Supplement) and displayed increased proliferative index (Ki-67/MIB-1) (eFigure 2D in Supplement). The tumor cells lacked immunoreactivity with CD68, actin, SMA, or factor XIIIa (not shown), supporting a diagnosis of spindle cell squamous carcinoma in the setting of BRAF inhibitor therapy. Unlike the expected pattern of a spindle cell melanoma, the spindled tumor cells were also MART-1 and S100 negative (not shown). The patient elected to discontinue the BRAF inhibitor therapy and has had no recurrence of, or new, cSCCs. She had a single melanoma recurrence in her thigh that was resected in early 2013. She is presently disease free.

Cutaneous squamous cell carcinomas with features of keratoacanthoma are common during BRAF inhibitor targeted therapy. These lesions are generally considered indolent and are conservatively managed with electrodessication and/or curettage, topical fluorouracil,⁴ or photodynamic therapy.⁵ In BRAF inhibitor–related cSCC-KA, the mitogen-activated protein kinase (MAPK) cascade signaling is paradoxically increased and augmented by mutant RAS.⁶

Because the clinical appearance of cSCC-KA and the spindle cell squamous carcinomas in our patient are indistinguishable, histologic evaluation of the entire lesion (via saucerization biopsy or incisional biopsy) is vital to prevent inadequate treatment of a deeply invasive process with a probable higher malignant potential. To our knowledge, this is the first report of an invasive spindled cSCC to occur during BRAF inhibitor therapy that clinically mimics cSCC-KA. It serves as a cautionary tale for dermatologists, dermatopathologists, and medical oncologists caring for these patients.

Corresponding Author: Daniel N. Cohen, MD, PhD, Department of Pathology, Microbiology, and Immunology, Vanderbilt University Medical Center, 1161 21st Ave S, MCN CC3322, Nashville, TN 37232-2561 (daniel.n.cohen@vanderbilt.edu).

Published Online: February 26, 2014. doi:10.1001/jamadermatol.2013.7784.

Author Contributions: Drs Cohen and Zwerner 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: Cohen, Lumbang, Zwerner.

Acquisition of data: All authors.

Analysis and interpretation of data: Cohen, Lumbang, Zwerner.

Drafting of the manuscript: Cohen, Lumbang, Sosman.

Critical revision of the manuscript for important intellectual content: All authors.

Administrative, technical, and material support: Lumbang, Boyd, Sosman.

Study supervision: Zwerner.

Conflict of Interest Disclosures: Dr Sosman has served as a consultant to Bristol-Myers Squibb and Genentech and has received grant support from Genentech. No other disclosures are reported.