A, Kaplan-Meier estimates of overall survival (OS) for all cases. B, Disease-specific survival (DSS) for all cases. C, Kaplan-Meier analysis of OS stratified by histological subtype. D, Kaplan-Meir analysis of DSS stratified by histological subtype. Adenocarcinoma (ADC), mucoepidermoid carcinoma (MEC), and adenoid cystic carcinoma (ACC) have poorer outcomes compared with squamous cell carcinoma (SCC) in terms of OS (P = .01) and DSS (P = .005).
Mallen-St. Clair J, Arshi A, Tajudeen B, Abemayor E, St John M. Epidemiology and Treatment of Lacrimal Gland TumorsA Population-Based Cohort Analysis. JAMA Otolaryngol Head Neck Surg. 2014;140(12):1110-1116. doi:10.1001/jamaoto.2014.2846
Primary tumors of the lacrimal gland are rare and are associated with substantial morbidity and mortality. The literature regarding these tumors is limited to case series and case reports.
To examine the incidence, treatment, and overall survival (OS) and disease-specific survival (DSS) of patients with cancer of the lacrimal gland.
Design, Setting, and Participants
Population-based cohort analysis using the Surveillance, Epidemiology, and End Results (SEER) database to identify patients with primary tumors of the lacrimal gland from 1973 to 2010.
Main Outcomes and Measures
Overall survival and DSS.
A total of 321 patients with nonlymphoid tumors of the lacrimal gland were identified. The most common histological subtypes were adenoid cystic carcinoma (ACC) (32.1%) and squamous cell carcinoma (SCC) (29.9%). Survival analysis revealed a 5-year OS and DSS for all lacrimal gland tumors of 60% and 75%, respectively. On univariate analysis, low tumor grade (P = .04) and surgical treatment (P < .001) were associated with significantly better OS. For ACC tumors, surgery (P = .009), but not radiotherapy (P = .44), was found to significantly improve OS. For SCC tumors, surgical treatment significantly improved both OS (P < .001) and DSS (P = .004); radiation therapy also significantly improved OS (P = .03). Using a multivariable analysis model, age (hazard ratio [HR], 1.03 [95% CI, 1.01-1.04]; P < .001), surgery (HR, 0.43 [95% CI, 0.25-0.75]; P = .003), and T stage at presentation (HR, 1.18 [95% CI, 1.01-1.37]; P = .03) were found to be independent predictors of OS. For ACC alone, age (HR, 1.04 [95% CI, 1.02-1.06]; P < .001) and surgery (HR, 0.35 [95% CI, 0.13-0.91]; P = .03) were independent predictors of OS. For SCC, age (HR, 1.05 [95% CI, 1.02-1.09]; P = .005), surgical resection (HR, 0.31 [95% CI, 0.12-0.83]; P = .02), and radiation therapy (HR, 0.33 [95% CI, 0.14-0.80]; P = .01) were independent predictors of OS.
Conclusions and Relevance
Our study demonstrates that ACC is the most common malignant epithelial neoplasm of the lacrimal gland. Determinants of survival for tumors of the lacrimal gland include age at diagnosis and surgical therapy. Radiation therapy is associated with improved DSS in SCC but not in ACC.
Tumors of the lacrimal gland represent between 5% and 25% of orbital tumors and can be subdivided into lymphoproliferative, benign epithelial, malignant epithelial, and malignant nonepithelial lesions. Epithelial lesions represent between 23% and 70% of biopsied cases, and 55% of these lesions are malignant.1 Whereas pleomorphic adenoma is the most common benign lesion of the lacrimal apparatus, adenoid cystic carcinoma (ACC) is the most common malignant epithelial tumor.2,3 Other malignant epithelial tumors of the lacrimal gland include mucoepidermoid carcinoma, adenocarcinoma, acinic cell carcinoma, and squamous cell carcinoma (SCC).4 Nonepithelial malignant tumors include lymphomas and rarely sarcomas, malignant melanoma, and other mixed tumor types.
Primary malignant epithelial neoplasms of the lacrimal gland are rare and are associated with substantial morbidity and mortality due to early dissemination via bone. The optimal management of these tumors is controversial: from orbital-sparing conservative surgical approaches followed by postoperative radiation to orbital exenteration and craniofacial approaches.5- 7 Intra-arterial cytoreductive chemotherapy has also been proposed as an adjuvant treatment modality.8,9 Regardless of treatment modality, these tumors continue to portend a serious prognosis, with a reported 5-year overall survival (OS) ranging from 40% to 80%.7,10,11
Whereas there are a myriad of small case series and reports, fewer than 250 cases have been reported in the international literature and population data are scarce.12- 18 The goal of this study was to determine the incidence and survival of patients with primary malignant epithelial tumors of the lacrimal gland in a larger population. We used data from the population-based US National Cancer Institute’s Surveillance Epidemiology and End Results (SEER) cancer registry to analyze a number of patient and disease characteristics to determine factors affecting both OS and disease-specific survival (DSS) after diagnosis.
A population-based cohort analysis for patients with a diagnosis of a primary malignant tumor of the lacrimal gland was performed using the case-listing session of the SEER 18 database (www.seer.cancer.gov); the database is a widely used cancer registry that covers an estimated 28% of the US population, including 23% of African Americans and 40% of Hispanics. Geographic regions covered include San Francisco–Oakland, Connecticut, metropolitan Detroit, Hawaii, Iowa, New Mexico, Seattle (Puget Sound), Utah, metropolitan Atlanta, San Jose–Monterey, Los Angeles, Alaska, rural Georgia, California, Kentucky, Louisiana, New Jersey, and greater Georgia. No internal review board approval was required because the database uses publicly available information with no personal identifiers.
Patients with a primary tumor of the lacrimal gland from 1973 through 2010, the widest date range available in the latest version of the SEER software, were identified using the primary site label C69.5. The following primary data were extracted from the database for analysis: age at diagnosis, sex, race, histologic subtype (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] code), tumor extent and tumor size from both extent of disease and collaborative stage coding methods, tumor grade, tumor T stage, treatment with surgery and/or radiation therapy, and OS and DSS in months. Well-differentiated and moderately differentiated histological subtypes were grouped as low grade, while poorly differentiated and undifferentiated histological subtypes were grouped as high grade. T stage was retroactively determined where possible using extent of disease and collaborative stage staging codes for tumor size and location following the classification protocol determined by the American Joint Committee on Cancer; tumors were defined as T1 for size less than 2 cm with no extraglandular extension, T2 for size between 2 and 4 cm with no extraglandular extension, T3 for size greater than 4 cm, T4a for invasion of the periosteum, T4b for invasion of the orbital bone, and T4c for invasion of adjacent structures. Cases with a histological subtype corresponding to lymphoma (ICD-9-CM codes 9500 and greater) were excluded from all analysis.
Primary outcome was defined as time in months from diagnosis to death from any cause for OS and as time from diagnosis to death specific to the cancer-related diagnosis for DSS. Descriptive statistics were calculated for all variables. The OS and DSS curves were calculated using the Kaplan-Meier method. The differences were formally tested for using the log-rank test. Covariates were assessed for predictive performance with univariate and multivariate Cox proportional hazards regression models with regard to OS and DSS. Comparisons between groups were deemed statistically significant at the P < .05 threshold. Covariates were chosen for multivariate analysis on the basis of factors identified as significant or with log rank P < .25 on univariate analysis. This method was chosen to minimize the total number of covariates, thus improving the generalizability of the findings and minimizing instability in the model. As a default, age and sex were included in all multivariate models. Using this methodology, there were no less than 10 events per covariate for each model. Statistical analyses were performed in SPSS, version 21 (IBM Corp).
The SEER database search revealed 754 patients with a primary malignant tumor of the lacrimal gland from 1973 to 2010. Of these, 321 patients had primary carcinomas or sarcomas of the lacrimal gland and composed the cohort used in the analysis; the remaining 433 patients had lymphomas and were excluded from the remainder of the data analysis.19
Of our selected cohort, 55.5% were males and 74.8% were white (Table 1). The mean age at diagnosis was 57.7 years. Patients presented with various histological subtypes of lacrimal neoplasms. The most common histological subtypes were ACC (32.1% of cases) and SCC (29.9% of cases). Other less common but notable histological categories included mucoepidermoid carcinoma (8.7%), adenocarcinoma (8.4%), and transitional cell carcinoma (4.4%); all other documented histological subtypes had fewer than 10 cases total. At presentation, 29.3% of all tumors were poorly differentiated or undifferentiated tumors of high grade, and 35.8% demonstrated periosteal, bony, or extraorbital invasion corresponding to stage 4 (T4). Surgical treatment was used for 87.5% of patients, and 54.8% were treated with radiation therapy.
Survival analysis from Kaplan-Meier curves (Figure) revealed that the 5-year OS and DSS for all lacrimal gland tumors was 60% and 75%, respectively (Table 2); the median OS and DSS was 8.9 and 10.0 years, respectively (Table 2). Both OS and DSS showed a statistically significant difference in survival based on tumor type (P = .01 and .005, respectively) (Table 3). We focused the remainder of our analysis on ACC and SCC, which had the largest cohort size for analysis. The median OS was 7.6 years for ACC and 16.5 years for SCC (Table 2). Univariate survival analysis revealed that low tumor grade (P = .04) and surgical resection (P < .001) were associated with significantly better OS. White race (P = .047) showed a significant effect on DSS. Univariate analysis was performed for each variable on cases of ACC (n = 103) and SCC (n = 96), the 2 most common histological subtypes of lacrimal gland malignant neoplasms (Table 3). For ACC, surgical resection (P = .009) was found to have a significant effect on OS, and race (P = .03), on DSS. For SCC, surgical resection significantly improved both OS (P < .001) and DSS (P = .004); radiation therapy also significantly improved OS (P = .03).
In our multivariate analysis model (Table 4), age (hazard ratio [HR], 1.03 [95% CI, 1.01-1.04]; P < .001), surgical resection (HR, 0.43 [95% CI, 0.25-0.75]; P = .003), and T stage at presentation (HR, 1.18 [95% CI, 1.01-1.37]; P = .03) were found to be independent predictors of OS. Tumor grade was not statistically significant (P = .10), which may be due to fewer cases having grade data available from the SEER database. T stage was the only independent predictor of DSS (HR, 1.32 [95% CI, 1.04-1.68]; P = .02).
The multivariate analysis model was then used to ascertain the independent effects of case variables on patients with ACC and SCC. For ACC (n = 103), age (HR, 1.04 [95% CI, 1.02-1.06]; P < .001) and surgical resection (HR, 0.35 [95% CI, 0.13-0.91]; P = .03) were found to be independent predictors of OS. Age alone was an independent predictor of DSS (HR, 1.03 [95% CI, 1.01-1.05]; P = .01). Radiation therapy was not associated with improved survival outcomes in ACC. For SCC (n = 96), age (HR, 1.05 [95% CI, 1.02-1.09]; P = .005), surgical resection (HR, 0.31 [95% CI, 0.12-0.83]; P = .02), and radiation therapy (HR, 0.33 [95% CI, 0.14-0.80]; P = .01) were independent predictors of OS, whereas only surgical resection (HR, 0.13 [95% CI, 0.02-0.79]; P = .03) was an independent predictor of DSS.
Tumors of the lacrimal gland are a rare oncological entity with varied histopathological characteristics. Optimal treatment of these tumors is controversial and is likely to vary according to the histopathological subtype of the tumor. Currently, our knowledge of the behavior of these tumors is based on case reports and single-institution studies.5- 7,20- 22 Use of the SEER database allows for the analysis of treatment and outcomes of rare malignant neoplasms with a greater statistical power. The SEER database has been used to find determinants of survival in many types of head and neck cancer and has been validated in previous studies.23- 28 However, there are inherent limitations in such a study design. The database lacks information on patient comorbidities, margin status, and administration of chemotherapy, and there are concerns regarding misclassification because there is no centralized review by a head and neck pathologist. This report represents, to our knowledge, the first effort analyzing tumors of the lacrimal gland from an epidemiological perspective at the population level.
Our data reinforce the finding that ACC represents the most common epithelial malignant neoplasm of the lacrimal apparatus.3,29 In this series, ACC represented 31.2% of epithelial malignant neoplasms. Our analysis revealed a surprisingly high rate of SCC of the lacrimal gland, which represented 29.9% of the epithelial tumors. Previously, SCC of the lacrimal apparatus has only been noted in case reports.30,31 Primary SCC of the lacrimal gland, similar to that of the salivary glands, is thought to be a rare entity in comparison with metastatic SCC from primary SCC of the skin. According to SEER methodology, these tumors were coded as primary tumors; however, it is possible that the high rate of SCC found in this cohort could be the result of misclassification of metastatic disease. Interestingly, an institutional review of malignant neoplasms of the lacrimal gland performed at University of California, Los Angeles, revealed a similar high rate of primary SCC (33.3%, J.M.-S.C., A.A., M.S.J., unpublished data, 2014). Other notable histopathological subtypes encountered in this analysis included mucoepidermoid carcinoma (8.7%), adenocarcinoma (8.4%), and transitional cell carcinoma (4.4%). Given the rarity of these other histopathological subtypes and insufficient numbers for statistical power, the remainder of our analysis focused on ACC and SCC.
Univariate analysis revealed a statistically significant difference in OS with tumor type, surgical resection, and low tumor grade. Surprisingly, SCC demonstrated improved median survival relative to ACC (16.5 vs 7.6 years). Importantly, the apparent survival benefit for SCC is not explained by younger age at diagnosis in these patients. In our cohort, patients with ACC were statistically insignificantly younger at diagnosis and there was no significant difference in median age at diagnosis among patients with the major subtypes of tumors of the lacrimal gland, with a median age at diagnosis of 52, 60, and 61 years for ACC, SCC, and mucoepidermoid carcinoma, respectively. Previous reports have demonstrated improved survival of ACC relative to SCC in epithelial tumors, highlighting the unique and poorly understood clinical behavior of these head and neck tumors.32,33 Surgical resection was associated with improved OS, which is consistent with the established role of primary surgical resection in the management of these tumors.7 It should be noted that the exclusion of patients who are not surgical candidates because of advanced disease or substantial comorbidities at presentation may bias the survival advantage seen in this cohort. To our knowledge, there is no report of grade as a prognostic factor in tumors of the lacrimal gland; however, our finding that well-differentiated and moderately differentiated tumors have better survival outcomes is consistent with findings in other head and neck malignant neoplasms.28,32,34,35
Univariate analysis performed on ACC and SCC again revealed that surgical management improved OS and DSS in SCC and OS alone in ACC. Surgical resection with wide local excision is the mainstay of treatment of tumors of the lacrimal gland.5,14 These results again justify surgical management of tumors of the lacrimal gland with the aforementioned caveat that patients who are not candidates for surgery may have poor prognosis. The SEER database does not provide sufficient detail to allow complete investigation of outcomes dependent on surgical technique. Given the controversy surrounding the surgical approach (eg, orbit preservation vs exenteration vs orbitofacial resection), additional studies investigating the optimal surgical approach for each histopathological subtype will be important to optimize treatment.7,36- 40
Multivariate analysis of all epithelial tumors of the lacrimal gland demonstrated that age, surgical resection, and T stage at presentation are important determinants of DSS. These data are in accordance with previous studies noting poor outcomes among elderly patients and may reflect increased stage at presentation, decreased access, and limited treatment options in this population.41 T stage, considered to be a strong prognostic indicator in most cancers, was an independent predictor of survival and should be used to guide management of patients with lacrimal neoplasms.
The multivariate analysis model was used to analyze determinants of survival in patients with ACC and SCC. In ACC, age and surgical resection were found to be independent predictors of OS. Age alone was an independent predictor of DSS. In SCC, age, surgical resection, and radiation therapy were independent predictors of OS, whereas only surgical resection was an independent predictor of DSS. As in univariate analysis, radiation therapy improved survival in SCC but not in ACC. These results corroborate the apparent inefficiency of radiation and nonsurgical therapy in patients with ACC.42 However, analysis of the SEER database does not allow determination of whether radiation was used with curative attempt or as adjuvant therapy, which might bias the results. Nevertheless, our data support primary surgical management of epithelial lacrimal malignant neoplasms and adjuvant radiation therapy in patients with SCC.
Our study independently reinforces the finding that ACC is the most common malignant epithelial neoplasm of the lacrimal gland. Herein we report, for the first time to our knowledge, a surprisingly high rate of SCC within the lacrimal apparatus. Determinants of survival include age at diagnosis and surgical therapy. Radiation therapy is associated with improved DSS in SCC but not in ACC. These data underline the import of multimodality therapy for SCC of the lacrimal gland.
Submitted for Publication: August 10, 2014; final revision received September 17, 2014; accepted September 28, 2014.
Corresponding Author: Jon Mallen-St. Clair, MD, PhD, Department of Head and Neck Surgery, University of California, Los Angeles, 62-132 CHS, 10833 Le Conte Ave, Los Angeles, CA 90095-1624 (firstname.lastname@example.org).
Published Online: November 13, 2014. doi:10.1001/jamaoto.2014.2846.
Author Contributions: Dr Mallen-St. Clair and Mr Arshi had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Tajudeen, Abemayor, St John.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Mallen-St. Clair, Arshi, Abemayor.
Critical revision of the manuscript for important intellectual content: Mallen-St. Clair, Tajudeen, Abemayor, St John.
Statistical analysis: Arshi, Tajudeen.
Obtained funding: St John.
Administrative, technical, or material support: St John.
Study supervision: Mallen-St. Clair, Tajudeen, Abemayor, St John.
Conflict of Interest Disclosures: None reported.
Funding/Support:National Institutes of Health/National Center for Advancing Translational ScienceUniversity of California, Los Angeles, Clinical and Translational Science Institute grant ULITR000124.
Role of the Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Previous Presentation: This study was presented at the Fifth World Congress of the International Federation of Head and Neck Oncologic Societies and the Annual Meeting of the American Head & Neck Society; July 26, 2014; New York, New York.