Prostaglandin E2, Tumor Necrosis Factor α, and Pro-opiomelanocortin Genes as Potential Mediators of Cancer Pain in Cutaneous Squamous Cell Carcinoma of Organ Transplant Recipients

Pain is a frequently reported symptom in keratinocyte cancer.^1,2 In particular, squamous cell carcinoma (SCC) is associated with spontaneous pain or pain evoked by touching the lesion. The association of pain with SCC has been reported in the general population as well as in organ transplant recipients (OTRs).^1,3,4 The recent SCOPE ITSCC PAIN study³ reported that pain as a symptom predicts a histological diagnosis of SCC in the high-risk group of OTRs in 75% of biopsy specimens. In addition, the presence of pain separated SCC well from other skin lesions, like basal cell carcinoma, seborrheic keratosis, SCC in situ (Bowen disease), and actinic keratosis.³ Thus, the aim of our study was to identify possible mediators of pain in SCC to elucidate potential mechanism of SCC-associated pain in OTRs.

Candidate mediators were selected for their known involvement in inflammation or pain sensation of the skin. Based on literature and expert advice, 6 target genes were identified as candidate mediators of pain. In 2008 and 2009 the skin samples were collected from patients surveyed as part of the SCOPE ITSCC PAIN study.³ The use of excision specimens stored in pathology laboratories was approved by the institutional review board (Kantonale Ethikkommission Zürich). For this study, SCC tissue samples were retrieved from the pathology laboratories of the University Hospital of Zurich and the University Medical Center of Leiden. All specimens’ diagnoses were confirmed by a certified dermatohistopathologist. Messenger RNA (mRNA) was isolated from these formalin-fixed and paraffin-embedded SCC samples and analyzed by quantitative real-time polymerase chain reaction. We applied analysis of variance, corrected for multiple testing to compare gene expression levels quantified by cycling time (CT) values (housekeeping gene 36B4 minus target gene) in tissue samples derived from painful vs nonpainful tumors and Pearson χ²-test for comparison of categorical variables. We also conducted a univariate analysis to identify factors associated with pain. We included covariates in the multivariable logistic regression model to control for possible confounding. Two-sided P values are reported for statistical tests.

Thirty-four SCCs from OTRs (18 lesions associated with pain; 16 lesions not associated with pain) were analyzed. The Figure shows the expression of mRNA for the 6 targets: prostaglandin E 2 (PGE2), pro-opiomelanocortin (POMC), tumor necrosis factor α (TNFA), calcitonin gene-related peptide (CGRP), interleukin 1β (IL1β) and nerve growth factor (NGF) in SCC, according to the presence of pain. PGE2, POMC, and TNFA present more abundant mRNA expression in painful SCC. Multivariate analysis showed that SCC with pain was associated with increased levels of PGE2 compared with SCC without pain, with an odds ratio (OR) of 1.9 (95% CI, 1.1-3.4; P = .002), adjusted for age and sex. A similar association was found for POMC (adjusted OR, 1.5; 95% CI, 1.1-2.3; P = .003) and TNFA (adjusted OR, 1.4; 95% CI, 0.99-2.0; P = .05), as shown in the Table.

Successful mRNA isolation from formalin-fixed, paraffin-embedded tissue allowed proper quantitative polymerase chain reaction amplification showing more abundant mRNA expression for PGE2, TNFA, and POMC in painful SCC lesions. Our other candidates, CGRP, IL1β, and NGF, did not differ in mRNA expression levels. The technical constraints of RNA extraction, along with full clinical information on the samples, limited the number of biopsies available for analysis. Our differential results, however, confirm that a sufficient sample size was achieved to observe meaningful differences. While previous studies have identified these and other mediators systemically, our data based on the local tumor microenvironment newly suggest a role for these mediators in mediating pain in SCC.

Corresponding Author: Sara Regina Frauenfelder, MD, Dermatologische Klinik, UniversitätsSpital Zürich, Spitalstrasse 66, Wetzikon, Zurich 8620, Switzerland (sara.frauenfelder@bluewin.ch).

Accepted for Publication: October 11, 2016.

Published Online: December 7, 2016. doi:10.1001/jamadermatol.2016.4775

Author Contributions: Drs Frauenfelder and Hofbauer 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.

Concept and design: Frauenfelder, Bouwes Bavinck, Serra, Hofbauer.

Acquisition, analysis, or interpretation of data: Frauenfelder, Freiberger, Bouwes Bavinck, Quint, Genders, Hofbauer.

Drafting of the manuscript: Frauenfelder, Genders, Serra, Hofbauer.

Critical revision of the manuscript for important intellectual content: Freiberger, Bouwes Bavinck, Quint, Serra, Hofbauer.

Statistical analysis: Serra.

Obtained funding: Hofbauer.

Administrative, technical, or material support: Frauenfelder, Freiberger, Quint, Hofbauer.

Supervision: Freiberger, Bouwes Bavinck, Genders, Serra, Hofbauer.

Conflict of Interest Disclosures: None reported.

Additional Contributions: We thank Dr Michaela Kress, Department of Physiology, Medical University of Innsbruck; Dr Claudia Sommer, Department of Neurology, University of Würzburg; Dr Peter W. Reeh, Department of Physiology and Pathophysiology, University of Erlangen-Nürnberg; Dr Dominik Straumann, Department of Neurology, University Hospital Zürich; and Dr Hanns Ulrich Zeilhofer, Department of Pharmacology and Toxicology, University of Zürich, for providing discussion and feedback on the selection of pain mediators in skin cancer pain. We thank Ines Kleiber, laboratory assistant, Department of Dermatology, University Hospital Zürich, for contributing experience and precision in performing the histologic analysis. We thank Guergana Iotzova-Weiss, Department of Dermatology, University Hospital Zürich, for providing polymerase chain reaction reagents and primers. No additional compensation was provided.