Young non-smokers with oral cancer: What are we missing and Why?

Presentation: S005
Topic: Oral Cavity
Type: Oral
Date: Friday, July 23, 2021
Session: 1:00 PM - 1:50 PM Skullbase Plus
Authors: Cristina Valero1; Avery Yuan1; Daniella K Zanoni1; Snjezana Dogan2; Jatin P Shah1; Luc G Morris1; Richard J Wong1; Aviram Mizrachi3; Snehal G Patel1; Ian Ganly1
Institution(s): 1Head and Neck Service, Memorial Sloan Kettering Cancer Center, New York, NY, USA; 2Department of Pathology, Memorial Sloan Kettering Cancer Center, New York, NY, USA; 3Department of Otorhinolaryngology - Head and Neck Surgery, Rabin Medical Center, Petah Tikva, Israel


Outcomes are perceived to be worse in young non-smokers (YNS) with oral cancer (OSCC). However, there is meager data to support this. We hypothesize that the difficulty in proving this is related to inappropriate weighting of prognostic factors and ignoring host immune status. Moreover, the etiology of OSCC is also likely distinct in YNS, since they develop OSCC even when lacking the main risk factors (age and smoking). This may be related to an impaired host immune status. The aim of this study was to analyze outcomes in YNS OSCC patients using adequate weighting of important cofactors and to explore if peripheral blood leukocytes (as a surrogate for immune status) could explain why these patients develop OSCC.


From a database of 2073 OSCC patients treated with primary surgery between 1985-2015, four groups were categorized by age and smoking: ≤40 years-old/non-smokers (n=100), ≤40 years-old/smokers (n=80), >40 years-old/non-smokers (n=595) and >40 years-old/smokers (n=1298). Baseline covariates including sex, alcohol use, comorbidities, tumor site, pathological overall stage, and adjuvant treatment were used to calculate propensity scores. Following this, stabilized inverse probability of treatment weights were calculated and cumulative incidence functions for the outcomes of interest were plotted. A weighted multivariable Cox proportional hazard model was used to quantify the impact of tumor and host covariates. Lastly, to explore if having a worse peripheral blood leukocytes profile (higher neutrophils, monocytes and neutrophil-to-lymphocyte ratio [NLR]) may partially explain why YNS develop OSCC and have poor outcomes, a control cohort of thyroid cancer patients (a more indolent cancer) was matched using covariates that influence leukocyte counts (age, sex, and smoking status). Depending on data distribution, two-sample t-test or Kruskal-Wallis test was used to compare the mean/median of peripheral blood leukocytes between these 2 groups.


Before adjusting, YNS had a lower probability of death compared to young smokers
(Figure 1A)

. After adjusting, YNS had a higher probability of death
(Figure 1B)

. This poorer outcome was driven by higher incidence of regional and distant recurrences. In a multivariable analysis controlling for tumor and host factors, YNS maintained the same trend, having worse outcomes than young smokers, for OS (HR=2.02; 95% CI 0.99-4.12; P=.054) and disease-specific survival (DSS) (HR=1.60; 95% CI 0.63-4.02; P=.320). Host factors such as peripheral blood leukocytes showed a strong association with both OS and DSS suggesting YNS may have an impaired immune system that led to OSCC development and poor outcomes. To explore this, we compared peripheral blood data between YNS with OSCC to a control group (YNS with thyroid cancer) and report that OSCC patients had worse peripheral blood leukocytes profile, with higher neutrophil (p=.018) and monocyte counts (p=.009), and higher NLR (p=.004).


 When adjusted by relevant covariates, YNS with OSCC have a poorer survival due to higher incidence of regional and distant recurrences.  Using peripheral blood leukocytes as surrogate of the balance between the tumor and the host, our results suggest that an impaired immune system may be partly responsible for OSCC development and poorer outcomes in YNS.