The Histopathology of Oral Cancer Pain in a Mouse Model and a Human Cohort

Abstract

Oral cancer patients often have severe, chronic, and mechanically induced pain at the site of the primary cancer. Oral cancer pain is initiated and maintained in the cancer microenvironment and attributed to release of mediators that sensitize primary sensory nerves. This study was designed to investigate the histopathology associated with painful oral cancers in a preclinical model. The relationship of pain scores with pathologic variables was also investigated in a cohort of 72 oral cancer patients. Wild-type mice were exposed to the carcinogen, 4-nitroquinoline 1-oxide (4NQO). Nociceptive (pain) behavior was measured with the dolognawmeter, an operant device and assay for measuring functional and mechanical allodynia. Lesions developed on the tongues and esophagi of the 4NQO-treated animals and included hyperkeratoses, papillomas, dysplasias, and cancers. Papillomas included lesions with benign and dysplastic pathological features. Two histologic subtypes of squamous cell carcinomas (SCCs) were identified-SCCs with exophytic and invasive components associated with papillary lesions (pSCCs) and invasive SCCs without exophytic histology (iSCCs). Only the pSCC subtype of tongue cancer was associated with nociceptive behavior. Increased tumor size was associated with greater nociceptive behavior in the mouse model and more pain experienced by oral cancer patients. In addition, depth of invasion was associated with patient-reported pain. The pSCC histology identifies 4NQO-induced tongue cancers that are expected to be enriched for expression and release of nociceptive mediators.

Keywords: 4-nitroquinoline 1-oxide; nociception; nociceptive pain; oral SCC; oral cancer pathology; oral carcinogenesis.

Figure 1.

Multiple lesions on a tongue from a 4-nitroquinoline 1-oxide (4NQO)–treated animal. (A) Longitudinal sections (anterior to the left) of a bisected tongue. Tongues were sectioned from the center outward. The scanned images are from slide 40, so that the top and bottom portions of the tongue are separated by 390 µm. Lesions include acanthosis, hyperkeratosis (H), field dysplastic changes (D), papillomas (P1, P2), and cancers (C1, C2, C3). (B) High-power image of P1, a dorsal sessile papilloma. (C) High-power image of P2, a dorsal pedunculated papilloma with finger-like epithelial proliferations with fibrovascular cores (arrowheads). (D) High-power image of C1, a dorsal papillary squamous cell carcinoma (pSCC) with exophytic (black arrow) and invasive (white arrow) components. (E) High-power image of C2, a ventral invasive squamous cell carcinoma (iSCC). High-power images of additional lesion types from other tongue sections. (F) Acanthosis and hyperkeratosis. (G) Hyperkeratosis with low-grade dysplasia and (H) high-grade dysplasia. Scale bars: A = 500 µm, B–E = 250 µm, F = 200 µm, G and H shown in G = 100 µm.

Figure 2.

Deeply invasive squamous cell carcinoma (diSCC) with aggressive features. (A) Longitudinal section (anterior to the left) shows a diSCC involving the entire dorsoventral thickness of the tongue with depth of invasion greater than 2 mm. High-power views of inset boxes 1, 2, and 3 shown in panels B–D. (B) Inset box 1: islands of diSCC (white arrows) invading stroma with interspersed inflammatory cells (turquoise arrows). (C) Inset box 2: Perineural invasion (PNI) with cancer (white arrows) surrounding greater than 50% of a nerve at multiple foci. Inflammation (turquoise arrows) is interspersed between nerves and cancer. (D) High-power image of PNI, cancer (white arrows), and inflammation (turquoise arrows). (E) Inset box 3: large blood vessels surrounded by cancer islands (white arrows) and inflammation (turquoise arrows). Scale bars: A = 100 µm; B–E = 50 µm.

Figure 3.

Tongues from 4-nitroquinoline 1-oxide (4NQO)–treated animals harbor multiple lesions. Shown are individual mice in columns (arranged in order of pain scores from left to right) and with numbers of indicated lesions in rows.

Figure 4.

Scatterplots showing correlation of pain scores (x-axis) and pathologic variables (y-axis) in the human cohort of 72 oral cancer patients. (A) Correlation of pain scores with tumor size (greatest dimension, n = 71 patients, r = 0.47, P < 0.001). (B) Correlation of pain scores with depth of invasion (DOI; n = 49 patients, r = 0.37, P = 0.01). r = Pearson’s coefficient of correlation; P is significant at P ≤ 0.05.