Dual Immunostaining With SOX10 and AE1/AE3 to Confirm Perineural Invasion on Mohs Sections

March 2019 | Volume 18 | Issue 3 | Original Article | 262 | Copyright © March 2019


Matthew R. Donaldson MDa,b and L. Arthur Weber MDa

aMountain West Dermatology, Grand Junction, CO bDepartment of Dermatology, University of Utah, Salt Lake City, UT

Abstract
Perineural invasion (PNI) is associated with high risk keratinocyte carcinomas. Identification of PNI during Mohs surgery is important for staging and post-adjuvant treatment decisions but can be challenging. To confirm or exclude PNI suspected on hematoxylin and eosin sections, we performed immunohistochemical double staining on Mohs frozen sections. Neural marker SOX10 and pan-cytokeratin marker AE1/AE3 were combined in a simultaneous assay using species-specific (mouse and rabbit) antibodies and horseradish peroxidase and alkaline phosphatase detection systems. Of 23 Mohs cases with suspected PNI, 18 were confirmed to have definitive nerve involvement by tumor using double staining. Double staining frozen tissue is feasible and can be beneficial for real time confirmation of PNI during Mohs.

J Drugs Dermatol. 2019;18(3):262-264.

INTRODUCTION

Perineural invasion (PNI) is associated with high risk keratinocyte carcinomas.1 This histologic finding correlates with increased risk of recurrence for basal cell carcinoma (BCC) and increased recurrence, metastasis, and disease-specific death for squamous cell carcinoma (SCC).2 Estimates of PNI vary by study and tumor type, but are likely between 2-6% for BCC and SCC.3 Both the AJCC 8th edition and the Brigham and Women’s Hospital systems for staging cutaneous SCC consider PNI involving nerves > 0.1 mm in diameter a high risk feature for upstaging tumors.4Identifying PNI on frozen sections during Mohs micrographic surgery (MMS) can be challenging, especially in the context of perineural inflammation. Pancytokeratin marker AE1/AE3 can be used to identify tumor in regions obscured by inflammation during MMS.5 A rapid 19-minute frozen section protocol showed similar reliability in highlighting suspected PNI to formalin sections.6 Rapid cytokeratin immunohistochemistry (IHC) to investigate regions of suspected keratinocyte carcinoma confirms the keratinocyte lineage of suspected tumor cells. However, verification that nerve (rather than other spindled cell structure or inflammation) is involved by tumor can be difficult in some cases.7-9Definitive identification of PNI at the time of MMS can be helpful for timely triage and management of high-risk tumors. To more precisely characterize PNI suspected on hematoxylin and eosin (H/E) sections during MMS, we performed a rapid double staining protocol combining AE1/AE3 with the neural crest marker SOX10. We used antibodies raised in different species with different detection systems to allow simultaneous staining.

MATERIALS AND METHODS

Cases of suspected perineural invasion were identified during routine MMS for keratinocyte carcinomas on H/E sections. Deeper levels of fresh frozen sections were cut at 5 um and dual stained by combining rabbit alkaline phosphatase and mouse horseradish peroxidase assays. The basic protocol we used is outlined in Table 1. Buffer rinses were performed after each step using 1XTBS/Tween until chromogen incubation, after which tap water was used to rinse. Primary incubation was performed in a humidity chamber under mild agitation with an orbital shaker with 1:500 rabbit pancytokeratin AE1/AE3 (Thermoscientific) and 1:50 mouse SOX10 (Biocare). The sections were then incubated with AP reagent (Vector ImmPRESS) and polymer amplifier or polymer HRP (Vector ImmPRESS Excel). The initial chromogen applied was Vector Red (Vector Labs) for rabbit AP system followed by 3,3’,5,5’-tetramethylbenzidine (TMB, Vector Labs) for the mouse HRP system. No counter stain was used.

RESULTS

From August 2016 to July 2018, 23 MMS cases with suspected PNI were tested using simultaneous double staining protocol on fresh frozen tissue. PNI was confirmed by establishing tumor adjacent to nerve in 18 cases with findings similar to Figure 1. Five cases showed no relationship of tumor to nerve using dual staining. Of these cases, four showed perineural inflammation with adjacent tumor on H/E and one had spindled structures suspicious for nerve surrounded by tumor on H/E shown to be squamous pearls on dual staining.

DISCUSSION

In this case series, we found that rapid double immunostaining can be performed on MMS frozen sections. We combined neural and keratinocyte markers to improve confidence in the diagno