Purpose/Objectives: Lymph node yield(LNY) has been reported as an independent prognostic factor for patients with oral cavity cancer. This study evaluates the prognostic significance of LNY in patients with HPV(-) and HPV(+) oropharynx cancers.
Materials/Methods: A retrospective cohort of adult oropharynx cancer patients from the National Cancer Database(NCDB) diagnosed from 2004-2014 was obtained. Patients with squamous cell carcinoma, who underwent primary surgery were included. Those who had previous malignancy, underwent pre-operative radiation or those with an indication for bilateral neck dissections were excluded. Following descriptive analysis, a multivariate analysis was performed to evaluate the relationship between LNY and all-cause mortality.
Results: In 93 patients with HPV(-) tumors and a pathologically negative neck(pN0) dissection after a mean follow-up of 32.3 months, the hazard ratio for death among those with a LNY of <5 was 32 when compared to those with a LNY of ≥5(95% CI: 2-601; p=0.019). This benefit decreased in a step-wise fashion with increased lymph nodes but remained significant until LNYs of ≥15. There was no statistically significant LNY found for HPV(-) cancers with a clinically negative neck(cN0). For HPV(+) tumors and pN0 dissections, no obvious trend was observed relating LNY and survival. However, in 210 patients with HPV(+) tumors and cN0 neck with a mean follow-up of 30 months, the hazard ratio for death among those with a LNY of ≥31 was 3.2 when compared with those with a LNY of <31 (95% CI: 1-9; p=0.027) This benefit decreased, again in a step-wise fashion, as LNY increased LNY but remained significant until LNYs of 34. Among HPV(-) patients, 81% switched from cN0 to pN+ compared to 84% of HPV(+) patients (p=0.056)
Conclusion: Our results suggest that among individuals with oropharynx cancer LNY is an independent predictor of survival in pN0 HPV(-) cases. A threshold of 15 lymph nodes provides a statistically significant benefit to survival and may be used as a minimum to ensure an adequate dissection. This is not seen in the cN0 HPV(-) cohort, suggesting that the risk of pathological under-staging may be a determinant of decreased survival. In HPV(+) oropharynx cancer, there was no threshold found amongst pN0 neck dissections. This may be due to an increased rate of adjuvant therapy for negative tumor characteristics as opposed to lymph node characteristics, or due to the decreased impact of pathologically overlooked lymph node metastasis on survival in HPV(+) as compared to HPV(-) oropharynx cancer. In cN0 necks, a threshold of ≥31 lymph nodes negatively impacted survival. This may be a result of surgeons recognizing intra-operatively the presence of suspicious lymph nodes and subsequently extending their dissection, or when a pathologically positive lymph node was encountered in a cN0 specimen, the pathologist could identify more lymph nodes for examination.