Long-term outcome of esthesioneuroblastoma: hyams grade predicts patient survival

Abstract

Object Esthesioneuroblastoma (ENB) is a rare malignant neuroendocrine tumor originating from the olfactory neuroepithelium in the cribriform plate. Controversy still exists regarding the role of pathologic grading (Hyams grade) in prognostication. This study was undertaken to describe our experience with ENB and assess the role of pathologic grading in patient outcome. Methods This was a retrospective, single-institution experience, including 109 patients with ENB treated at our institution from 1962 to 2009. Multivariate analysis was performed utilizing Cox regression analysis models utilizing age, gender, modified Kadish stage, and Hyams grade. Results Mean age was 49 ± 16 (median 50) years at presentation (range 12 to 90 years). Median follow up was 5.1 years. All-cause mortality was significantly influenced by Hyams grading in univariate (p = 0.04) and multivariate (p = 0.02) analysis, in addition to proven prognostic factors, Kadish staging, lymph node metastasis, and age. Median survival was 9.8 years compared with 6.9 years with low (grade 1 to 2) versus high (grade 3 to 4) Hyams grade. Median overall survival was 7.2 ± 0.7 years. Conclusion ENB has a variable outcome, which is primarily prognosticated by the extent of involvement at presentation (Kadish stage and lymph node metastasis) and higher Hyams pathologic grade.

Keywords: Hyams grade; esthesioneuroblastoma; olfactory neuroblastoma; outcome; surgical treatment.

Figure 1

Distribution of esthesioneuroblastoma cases at date of earliest pathologic confirmation. Breakdown of number of cases (y-axis) by year in decades (x-axis). There appears to be relative stability in our esthesioneuroblastoma population from 1990 to 2010. Eighty-eight percent of cases occurred after 1980.

Figure 2

Age distribution of cases. There appears to be a unimodal distribution of cases among these 109 cases at age of presentation. The mean age is 49 ± 16 years (95% confidence interval 46 to 52 years). The median age was 50, with a range of 12 to 91 years.

Figure 3

Distribution of (A) Kadish staging and (B) Hyams grading for cases in which this information is available.

Figure 4

Impact of staging (Kadish), age (<65), and pathologic grade (Hyams) on survival. Kaplan-Meier survival curves for (A) modified Kadish staging (p < 0.001). (B) Age >65 (p = 0.02). (C) Hyams low and high grade (p = 0.04). (D) Hyams further separated to Hyams 1/2, 3, and 4, note Hyams grade 4 has a particularly poor outcome (p < 0.001).

Figure 5

Survival, surgical margins, and adjuvant treatment. Ten-year survival based on (A) surgical margins and (B) radiation therapy. There is significant influence of surgical margins (p = 0.04) and radiation, limited only to Kadish (C, D) patients (n = 81) (p < 0.001).

Figure 6

Modified Kadish staging and Hyams grading correlate highly. Modified Kadish stage (x-axis) is shown in proportion to each stage's Hyams grading (gray is low grade [1/2], black is high grade [3/4]). There is a strong correlation with higher-grade tumors and higher stage at presentation, gamma correlation of association 95% confidence interval (0.34 to 0.81).