Background: Decreased lymph node count (LNC) from neck dissection (ND) specimens for head and neck squamous cell carcinoma (HNSCC), is correlated with decreased survival. The cause of this correlation is unknown. Undernutrition is the most common cause of secondary immunodeficiency worldwide but relatively rare in developed countries. HNSCC is unique as it is one of the few cancers that directly interferes with feeding. Therefore, in contrast to other cancers, weight loss due to mechanical obstruction or odynophagia and relatively low body mass index (BMI) is common in HNSCC. Suppression of immune function due to nutritional deprivation could be causative with respect to both decreased LNC and survival in HNSCC, independent of LNC as an indicator of surgical completeness. We hypothesized that lower BMI (<22) due to larger cancers in the upper aerodigestive tract in HNSCC patients undergoing ND would be correlated with decreased LNC reflecting generalized immunosuppression.
Methods: We conducted our study in a single tertiary care institution to reduce variations in surgical and pathological technique. Retrospective review of UTSW database identified 428 Head and Neck SCC patients who underwent ND between 2006-2017. Clinical records and pathology reports were reviewed to quantify number of levels dissected and number of lymph nodes obtained. Variables such as Age, BMI, Tobacco history, Diabetes, T-stage and sex were collected. Total LNC per surgery was normalized to the number of neck levels dissected to further reduce the impact of surgical technique. Stepwise linear regression analysis was performed to identify independent significant factors associated with the number of lymph nodes obtained. Stepwise logistic regression analysis was conducted to identify the independent significant factors associated with higher T-stage (3 or 4).
Results: There was a significant difference in average LNC per level dissected at 6.161 for patients with a BMI <22 and 7.736 for patients with a BMI ≥22 (p-value 0.0005). Multivariate analysis showed that among the variables tested only BMI (<22 vs. ≥22) was significantly associated with decreased LNC per level. (p=0.006) and that low BMI (≤22 vs. ≥22) was significantly associated with higher T-stage (p=0.0002) after controlling for the effects of tobacco, smoking, sex, Charlson-Age comorbidity index, and the number of lymph nodes. Altering the BMI threshold to 21 (p<0.0001), or 23 (p=0.0005) did not change the significance of this relationship.
Conclusions: Low BMI in HNSCC is associated with higher t-stage and lower LNC per level of neck dissected possibly reflecting a generalized acquired immunodeficiency due to undernutrition.