Abstract
Purpose
Diffuse large B-cell lymphomas (DLBCLs) are rare tumors of the head and neck that often have non-specific presentations and significant morbidity and mortality. In this analysis we use a large cohort to compare the demographic and disease-specific parameters affecting survival and incidence of DLBCLs.
Methods
The United States National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) registry was utilized to extract data regarding sinonasal diffuse large B-cell lymphoma between 1973 and 2009. A total of 852 cases of sinonasal DLBCLs were found. Presenting symptoms, demographics, disease specific survival, relative survival and survival by treatment modality were described for this rare tumor.
Results
Overall disease specific survival (DSS) at 1-year was 84.7% and at 5 years was 68.0%. DSS was significantly lower for those not treated with radiation therapy, with 1- and 5-year survival rates of 77.3% and 62.5%, versus those treated with radiation therapy, with 1- and 5-year rates of 89.2% and 71.5% (p < 0.05). Prognosis was significantly better for patients treated with radiation therapy (HR 0.6, p < 0.05) while it was poorer for patients with involvement of multiple sinuses (HR 1.5, 1.8, p < 0.05).
Conclusions
DLBCLs of the sinonasal tract are rare tumors of the head and neck. Survival is significantly improved for those treated with radiation therapy while the involvement of multiple sinuses is a negative prognostic indicator.
1
Introduction
Sinonasal lymphomas are relatively rare tumors of the head and neck that can be challenging to diagnose for otolaryngologists . Diffuse large B-cell lymphomas (DLBCLs) are among the most common of sinonasal lymphomas. They present most frequently in the paranasal sinuses and rarely as primary nasal tumors. DLBCLs tend to be less aggressive than other lymphoma subtypes, but often have non-specific symptoms and features that vary considerably between cases and even geographic locations . This is partly due to the complex anatomy within the sinonasal region with numerous critical structures, such as the orbit and cranial nerves, which can be affected by local spread.
Presentations may include local effects like nasal obstruction, rhinorrhea, bloody discharge, facial swelling, or visual symptoms. In some cases, presentation can include B symptoms, such as fever and weight loss . Symptoms can be confused with benign inflammatory conditions, leading to delay in diagnosis and treatment . This can be particularly true in the pediatric population where the typical presenting symptoms for DLBCL, such as shortness of breath or high-pitched breathing are among the most common presentations for hospital visits, and may mimic an upper respiratory tract infection . As early stage sinonasal lymphomas have notably better outcomes, this delay in diagnosis, which in some cases may be nearly a year from initial presentation, may significantly increase morbidity and mortality . Mortality is associated with extranodal dissemination, with primary nasal tumors in particular having poor prognosis .
Most of the literature on DLBCL of the sinonasal tract has been limited to single institutional studies. In this report, using data from US National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) database, we analyzed trends in presentation, incidence, and survival of sinonasal DLBCL to further our understanding of this rare tumor.
2
Materials and methods
The SEER Program database using the SEER 18 registry was searched for frequency, incidence and survival data for DLBCL. We extracted data for DLBCL using the International Classification of Diseases for Oncology, 3rd edition (ICD-O-3) codes 9680/3 and 9680/4. Results were restricted to the sinonasal tract using ICD-O-3 anatomic site codes C30.0 Nasal cavity, C31.0 maxillary sinus, C31.1 ethmoid sinus, C31.2 frontal sinus, C31.3 sphenoid sinus, C31.8 overlapping lesions of accessory sinuses (indicating multiple sinus involvement), and C31.9 accessory sinus, not otherwise specified. Preset variables, site-specific factors, and collaborate staging codes were used to collect demographics and disease specific parameters from SEER. All data were extracted using SEER*Stat 8.0.4 (Surveillance Research Program, National Cancer Institute, 2012). Relative survival (RS) for each overall cohort was also calculated using SEER*Stat. Disease specific survival (DSS) was calculated via Kaplan–Meier analysis using JMP Statistical Discovery 10 (SAS Institute, Cary, NC); Hazard ratios were derived using RSs and fitting to multivariate Cox proportional hazards models with all available data (including unknowns) using CanSurv 1.1 (National Cancer Institute, Bethesda, MD). Data were also extracted regarding patients with the diagnosis of DLBCL of all other head and neck sites for the purpose of comparison to sinonasal disease.
Statistical comparison of frequency data was carried out with either a chi-square test (categorical data) or t test (continuous data) using MedCalc 12 (MedCalc Software, Ostend, Belgium) and/or using the two proportion z-test (Vassar Stats). Joinpoint Regression Program 4.0.1 (National Cancer Institute, Bethesda, MD) was used to calculate annual percentage change (APC) and trends in incidence. Microsoft Office Excel 2007 (Microsoft Corporation, Redmond, WA) was used for generation of data tables and simple algebraic operations. Institutional review board approval was not required for this study.
2
Materials and methods
The SEER Program database using the SEER 18 registry was searched for frequency, incidence and survival data for DLBCL. We extracted data for DLBCL using the International Classification of Diseases for Oncology, 3rd edition (ICD-O-3) codes 9680/3 and 9680/4. Results were restricted to the sinonasal tract using ICD-O-3 anatomic site codes C30.0 Nasal cavity, C31.0 maxillary sinus, C31.1 ethmoid sinus, C31.2 frontal sinus, C31.3 sphenoid sinus, C31.8 overlapping lesions of accessory sinuses (indicating multiple sinus involvement), and C31.9 accessory sinus, not otherwise specified. Preset variables, site-specific factors, and collaborate staging codes were used to collect demographics and disease specific parameters from SEER. All data were extracted using SEER*Stat 8.0.4 (Surveillance Research Program, National Cancer Institute, 2012). Relative survival (RS) for each overall cohort was also calculated using SEER*Stat. Disease specific survival (DSS) was calculated via Kaplan–Meier analysis using JMP Statistical Discovery 10 (SAS Institute, Cary, NC); Hazard ratios were derived using RSs and fitting to multivariate Cox proportional hazards models with all available data (including unknowns) using CanSurv 1.1 (National Cancer Institute, Bethesda, MD). Data were also extracted regarding patients with the diagnosis of DLBCL of all other head and neck sites for the purpose of comparison to sinonasal disease.
Statistical comparison of frequency data was carried out with either a chi-square test (categorical data) or t test (continuous data) using MedCalc 12 (MedCalc Software, Ostend, Belgium) and/or using the two proportion z-test (Vassar Stats). Joinpoint Regression Program 4.0.1 (National Cancer Institute, Bethesda, MD) was used to calculate annual percentage change (APC) and trends in incidence. Microsoft Office Excel 2007 (Microsoft Corporation, Redmond, WA) was used for generation of data tables and simple algebraic operations. Institutional review board approval was not required for this study.
3
Results
3.1
Study population
A total of 852 patients with DLBCL of the sinonasal tract were identified in the SEER database from 1973 to 2009. The average age of the patients identified was 65.8 years old. Males accounted for 54.5% of the patients, giving a male to female ratio of 1.2:1. Whites accounted for the vast majority of cases at 80.9%. “Others” (American Indian/AK native, Asian/Pacific Islander) accounted for 12.3% and blacks accounted for 6.2% of cases. Case numbers and percentages are shown in Table 1 .
| Characteristic | Number | Percent |
|---|---|---|
| Gender | ||
| Male | 464 | 54.5% |
| Female | 388 | 45.5% |
| Age at Diagnosis, mean | 65.8 | |
| Race | ||
| White | 689 | 80.9% |
| Black | 53 | 6.2% |
| Other | 105 | 12.3% |
| Anatomic Site | ||
| Nasal cavity | 290 | 34.0% |
| Maxillary sinus | 314 | 36.9% |
| Ethmoid Sinus | 74 | 8.7% |
| Frontal Sinus | 20 | 2.3% |
| Sphenoid sinus | 35 | 4.1% |
| Overlapping lesion of accessory sinuses | 37 | 4.3% |
| Accessory sinus, NOS | 82 | 9.6% |
| Treatment | ||
| No Radiation Therapy | 329 | 38.6% |
| Radiation Therapy | 493 | 57.9% |
| Unknown | 30 | 3.5% |
Incidence of DLBCL from 2000 to 2009 ranged from as high as 0.1 case per 100,000 people in 2006 to as low as 0.06 case per 100,000 people in 2007 ( Fig. 1 ). However, the annual percentage change (APC) indicated that there has been no significant change in the incidence of DLBCL over the course of this period (p > 0.05). This period was selected to study incidence due to availability/quality of data in the various SEER registries and due to evolving diagnostic guidelines for DLBCL.
3.2
Analysis of survival rates
Overall DSS for DLBCL of the sinonasal tract at 1-year was 84.7% and at 5 years was 68.0% ( Fig. 2 ). One-year and 5-year DSS for males was 83.7% and 67.4%, respectively, while for females 1- and 5-year DSS was 85.8% and 68.7%. By race DSS at 1-year was 85.3%, 71.6%, and 86.8% for patients of white, black and “other” race, respectively, with 5-year DSS rates of 68.1%, 61.8%, 69.7%. Differences in survival by sex or age were not statistically significant (p > 0.05). Relative survival (RS), which takes into account baseline survival in non-cancer individuals in the population, was also calculated for all patients with DLBCL of the sinonasal tract. Overall 1- and 5-year RS rates were 83.5% and 65.5%. For male patients, 1- and 5-year RS was 82.4% and 62.2%, respectively, while for females 1- and 5-year RS was 84.9% and 69.4%. One-year relative survival for white, black, and “other” race was 84.0%, 71.3%, and 86.0%, respectively, while 5-year relative survival rates were 65.8%, 57.6%, 65.5%.
