Pretreatment Pain as a Prognostic Predictor in Oral Tongue Squamous Cell Carcinoma: The Mediating Role of Perineural Invasion

Abstract

The potential effects of pretreatment pain on prognosis of tongue squamous cell carcinoma (TSCC) and the role that perineural invasion (PNI) plays in this process remains unclear.

BACKGROUND

The potential effects of pretreatment pain on prognosis of tongue squamous cell carcinoma (TSCC) and the role that perineural invasion (PNI) plays in this process remains unclear.The purposes of this study are to estimate the prognostic value of pretreatment pain and identify whether PNI is a mediating factor in the relationship between pretreatment pain and prognosis in TSCC.

PURPOSE

The purposes of this study are to estimate the prognostic value of pretreatment pain and identify whether PNI is a mediating factor in the relationship between pretreatment pain and prognosis in TSCC.This retrospective cohort study included TSCC patients who underwent first resections of primary lesions at the Beijing Stomatological Hospital of Capital Medical University between January 2009 and December 2019. Patients who had incomplete medical records and pathological data, received neoadjuvant radiotherapy and chemotherapy before surgery, and did not receive a TSCC diagnosis, were excluded.

STUDY DESIGN, SETTING, SAMPLE

This retrospective cohort study included TSCC patients who underwent first resections of primary lesions at the Beijing Stomatological Hospital of Capital Medical University between January 2009 and December 2019. Patients who had incomplete medical records and pathological data, received neoadjuvant radiotherapy and chemotherapy before surgery, and did not receive a TSCC diagnosis, were excluded.The predictor variables are pretreatment pain and PNI. The visual analog scale (VAS) was used to assess pretreatment pain levels, and the PNI status was evaluated by pathological section.

PREDICTOR VARIABLE

The predictor variables are pretreatment pain and PNI. The visual analog scale (VAS) was used to assess pretreatment pain levels, and the PNI status was evaluated by pathological section.The main outcome variables were the 3-year disease-specific survival (DSS) and disease-free survival (DFS).

MAIN OUTCOME VARIABLE(S)

The main outcome variables were the 3-year disease-specific survival (DSS) and disease-free survival (DFS).Covariates included age, sex, smoking history, alcohol history, growth pattern, and T-stage.

COVARIATES

Covariates included age, sex, smoking history, alcohol history, growth pattern, and T-stage.The χ² test was used to describe the baseline data. Kaplan-Meier analysis was used to estimate the 3-year DSS and DFS. The Cox regression model was adapted for univariate and multivariate analysis. The association between VAS score and PNI was analyzed using logistic regression analysis and mediation analysis. P value less than .05 indicated statistical significance.

ANALYSES

The χ2 test was used to describe the baseline data. Kaplan-Meier analysis was used to estimate the 3-year DSS and DFS. The Cox regression model was adapted for univariate and multivariate analysis. The association between VAS score and PNI was analyzed using logistic regression analysis and mediation analysis. P value less than .05 indicated statistical significance.The study included 307 subjects with a mean age of 52 (±12.1) years, and 164 (53.4%) were male. There were 65 (21.2%) with high VAS (>5) and 242 (78.8%) with low VAS (≤5). The DSS and DFS of high VAS were 64.6% (95% CI: 23.6 to 80.9%) and 52.3% (95% CI: 35.3 to 92.8%), of patients with PNI were 62.7% (95% CI: 19.6 to 64.5%) and 46.7% (95% CI: 25.9 to 66.9%), respectively. The group high VAS/with PNI had lower DSS and DFS than group high VAS/without PNI (55.3 and 40.4% vs 88.9 and 83.3%). The VAS and PNI were identified as independent factors associated with prognosis (P < .05). Mediation analysis revealed that the indirect effect of VAS on DFS was 0.071 (95% CI: 0.011 to 0.135, P = .024), while the total effect was 0.187 (95% CI: 0.074 to 0.296, P < .001), VAS score affected the DFS of TSCC through the mediating effect of PNI.

RESULTS

The study included 307 subjects with a mean age of 52 (±12.1) years, and 164 (53.4%) were male. There were 65 (21.2%) with high VAS (>5) and 242 (78.8%) with low VAS (≤5). The DSS and DFS of high VAS were 64.6% (95% CI: 23.6 to 80.9%) and 52.3% (95% CI: 35.3 to 92.8%), of patients with PNI were 62.7% (95% CI: 19.6 to 64.5%) and 46.7% (95% CI: 25.9 to 66.9%), respectively. The group high VAS/with PNI had lower DSS and DFS than group high VAS/without PNI (55.3 and 40.4% vs 88.9 and 83.3%). The VAS and PNI were identified as independent factors associated with prognosis (P < .05). Mediation analysis revealed that the indirect effect of VAS on DFS was 0.071 (95% CI: 0.011 to 0.135, P = .024), while the total effect was 0.187 (95% CI: 0.074 to 0.296, P < .001), VAS score affected the DFS of TSCC through the mediating effect of PNI.Our findings confirmed that pretreatment pain is associated with worse outcomes in TSCC. Patients with TSCC and severe pretreatment pain are more likely to be diagnosed with PNI, which results in a worse prognosis.

CONCLUSION AND RELEVANCE

Our findings confirmed that pretreatment pain is associated with worse outcomes in TSCC. Patients with TSCC and severe pretreatment pain are more likely to be diagnosed with PNI, which results in a worse prognosis.