A, Clinical examination revealed proptosis and downward displacement of the right eye and swelling of the upper eyelid. B, Computed tomography revealed a tumor mass of the right lacrimal gland that extended into the posterior orbit (arrow). C, A uniform population of small cells (centrocytes) with irregular nuclei (arrows) consistent with a grade 1 follicular lymphoma (hematoxylin-eosin, original magnification ×400). D, The tumor cells of the follicles express CD10, which is characteristic for follicular lymphoma (original magnification, ×100).
Patients treated with EBRT (n = 29) had a significantly better outcome than patients receiving chemotherapy combined with EBRT (n =22 ) (P = .02, log-rank test). The × symbols indicate censored patients.
eTable 1. Staging of 69 Primary Ocular Adnexal Follicular Lymphomas According to the TNM-Based Staging System
eTable 2. Current Guidelines for External Beam Radiation Therapy of Early-Stage Follicular Ocular Adnexal Lymphomas According to the Participating Eye Cancer Centers
eFigure 1. Disease-specific survival of all 98 ocular adnexal follicular lymphoma patients classified as primary- (n = 69), disseminated (n = 19), and relapse patients (n = 10).
eFigure 2. Disease-specific survival of all 98 ocular adnexal follicular lymphoma patients depicted according to the Ann Arbor stage of lymphoma as stage IE/IIE (n = 78) and stage IIIE/IV (n = 20).
Rasmussen PK, Coupland SE, Finger PT, Graue GF, Grossniklaus HE, Honavar SG, McKelvie P, Mulay K, Prause JU, Ralfkiaer E, Sjö LD, Heegaard S. Ocular Adnexal Follicular LymphomaA Multicenter International Study. JAMA Ophthalmol. 2014;132(7):851-858. doi:10.1001/jamaophthalmol.2014.376
The clinical features of the follicular subtype of ocular adnexal lymphoma (OAL) have not been previously evaluated in a large cohort.
To characterize the clinical features of follicular OAL.
Design, Setting, and Participants
We performed a retrospective multicenter study that involved 6 eye cancer centers from January 1, 1980, through December 31, 2010. A total of 105 patients with follicular OAL were identified, of which 7 patients were excluded because of missing clinical data. The median follow-up time was 52 months (range, 13-118 months).
Main Outcomes and Measures
Overall survival, disease-specific, and progression-free survivals were the primary end points.
Ninety-eight eligible patients with follicular OAL were included; 60 (61%) were women. The median patient age was 63 years (range, 32–96 years). Sixty-nine patients (70%) had primary OAL, 19 (19%) had OAL in conjunction with a concurrent systemic lymphoma, and 10 (10%) presented with an ocular adnexal relapse. The lacrimal gland (28%), conjunctiva (28%), and orbit (28%) were the most frequently involved sites. Of the 69 patients with primary follicular lymphoma, 38 (55%) presented with Ann Arbor stage IE lymphoma, and 31 (45%) had stage IIE lymphoma. Patients with disseminated lymphoma had stage IIIE (9 of 19 [47%]) and stage IV (10 of 19 [53%]) disease, whereas patients with a relapse of systemic lymphoma presented with stage IE (8 of 10 [80%]), stage IIE (1 of 10 [10%]), and stage IIIE (1 of 10 [10%]) disease. Patients with primary follicular lymphoma (n = 69) and those with isolated ocular relapse (n = 9) were treated with external beam radiation therapy (EBRT) (35 of 78 [45%]) or EBRT plus chemotherapy (22 of 78 [28%]). Patients presenting with stage IIIE-IV follicular lymphoma (n = 20) most frequently received chemotherapy (9 of 20 [45%]) or EBRT plus chemotherapy (4 of 20 [20%]). The 10-year overall survival for the entire study cohort was 60%. Primary patients treated with EBRT had a better disease-specific survival compared with patients receiving ERBT plus chemotherapy (10-year disease-specific survival, 94% vs 40%; P = .02 by log-rank test).
Conclusions and Relevance
Follicular OAL was more commonly found in elderly female patients. These tumors were equally noted to involve the conjunctiva, lacrimal gland, and orbit. Patients with ocular adnexal follicular lymphoma primarily treated with EBRT had a more favorable long-term disease-specific survival.
Follicular lymphoma (FL) accounts for approximately 20% of all lymphomas, with the incidence being highest in Western Europe and the United States.1 Most patients with FL have generalized lymphadenopathy and bone marrow involvement.2 The disease course is usually indolent but exhibits frequent relapses. Before the approval of the anti-CD20 antibody rituximab for the treatment of FL, the 10-year survival rates for patients with FL were approximately 71%, 51%, and 36% according to the low-, intermediate-, and high-risk groups, respectively, constituting the Follicular Lymphoma International Prognostic Index (FLIPI).3 The FLIPI score is based on 5 prognostic factors: age (≥60 years), Ann Arbor stage of disease (stage III-IV), number of nodal sites (>4), hemoglobin level (<12 g/dL [to convert to grams per liter, multiply by 10]), and elevated serum lactate dehydrogenase level. With novel therapies using combination chemotherapy and rituximab, the 8-year survival rates of FL in the 3 risk groups have improved to 94%, 84%, and 59%, respectively.4
Follicular lymphoma may also arise as primary tumors in extranodal sites, such as the gastrointestinal tract, skin, breast, and testis.5 Primary extranodal FL has been suggested to constitute a clinicopathologic distinct subgroup. For example, extranodal FL has a lower frequency of BCL2/IGH rearrangement6 and may be associated with a more favorable outcome.7- 10 In addition, most FLs are successfully managed with external beam radiation therapy (EBRT).11,12 However, whether there is a prognostic difference between extranodal FL at different site-specific locations remains uncertain.7,8
The ocular adnexal region (ie, orbit, eyelids, conjunctiva, lacrimal gland, and lacrimal sac) is regarded as the extranodal region. The frequency of FL in this region varies among different geographic areas. In the United States, FL accounts for 23% of all ocular adnexal lymphomas (OALs),13 whereas in Western Europe, frequencies between 5% and 10% have been reported.14- 16 Although ocular adnexal FL has been described to exhibit morphologic and immunophenotypical features similar to nodal FL,13 the clinical outcome has not been evaluated extensively. Therefore, the aim of the present international multicenter study is to evaluate the clinical features of ocular adnexal FL in a large cohort from 6 eye cancer centers.
Eligible patients with a histologic diagnosis of FL that involved the ocular adnexal region were identified from the databases of 6 eye cancer centers in a search period ranging from January 1, 1980, through December 31, 2010. For histopathologic examination, all specimens were stained with hematoxylin-eosin and analyzed immunohistochemically using the following panel of antibodies: BCL2, BCL6, CD3, CD5, CD10, CD20, CD23, CD79α, cyclin D1, and MUM1. In addition, the Ki-67 labeling index (cellular marker for proliferation) was assessed in all specimens. The local eye cancer centers reviewed the samples and classified the specimens according to World Health Organization Classification of Tumours of Haematopoietic and Lymphoid Tissues, Fourth Edition.17 All Danish specimens were analyzed by fluorescence in situ hybridization for the t(14;18) chromosomal abnormality.
The study followed the tenets of the Declaration of Helsinki, and institutional review board and health information privacy agency approvals were obtained from the Danish Data Protection Agency for inclusion of anonymized data in the present study. Informed consent was not required by the Danish Data Protection Agency because most of the patients were dead at the time of the study.
Evidence of systemic involvement according to the Ann Arbor staging classification18 was registered for each patient. In addition, the following data were recorded: year of diagnosis, sex, age, symptoms and clinical findings, treatment modalities and response to therapy, recurrence-free periods, time and localization of relapse, survival duration, disease stage at last follow-up, and time and cause of death.
Primary OAL was defined as follows: a biopsy-verified lymphoma limited to the ocular adnexal region (stage IE) with or without involvement of unilateral preauricular and/or submandibular lymph nodes and/or adjacent structures, such as the parotid gland (stage IIE), and no history of lymphoma disease. The extent of ocular adnexal involvement of primary lymphomas was also evaluated according to the American Joint Committee on Cancer and Union for International Cancer Control TNM-based clinical staging system of OALs.19 Treatment response criteria were assessed as complete remission, partial remission, and stable disease.20
Overall survival, disease-specific, and progression-free survivals were considered primary end points. Disease-specific survival was defined as the time from the date of diagnosis to the date of death from lymphoma or to the date of last contact, the latter being a censored event. Overall survival was defined as the time from the date of diagnosis to the date of death from any cause or the date of last contact, the latter being a censored event. Progression-free survival was calculated as the time from the date of diagnosis to either the date of first relapse or progression after initial treatment to the date of death from any cause or to the date of last contact, with the last 2 being censored events. Kaplan-Meier plots were used to visualize the disease-specific survival and progression-free survival, and different risk groups were compared using the log-rank test. Individual risk factors were compared using the χ2 test. Statistical calculations were performed using the IBM SPSS Package, version 20 (IBM Corporation). P < .05 was considered significant.
A total of 105 patients with FL of the ocular adnexal region were identified from the databases of 6 eye cancer centers: Liverpool, United Kingdom (n = 30); Copenhagen, Denmark (n = 24); Melbourne, Australia (n = 19); Hyderabad, India (n = 15); New York, New York (n = 14); and Atlanta, Georgia (n = 3). The 14 patients from New York represent a subset of patients from the 2013 study by Graue et al.21 Seven patients from Melbourne were excluded because of missing clinical data. Consequently, 98 patients were included in the analysis.
Thirty-eight of the patients (39%) were men, and 60 (61%) were women (Table 1). The median age was 63 years (range, 32-96 years). The median ages according to the eye cancer centers were as follows: 65.5 years in Liverpool, 62.9 years in Copenhagen, 56.9 years in Hyderabad, 65 years in Melbourne, 66 years in New York, and 71 years in Atlanta. The male-female ratio according to the centers were as follows: 0.2 in Liverpool, 0.7 in Copenhagen, 2.8 in Hyderabad, 0.7 in Melbourne, 0.8 in New York, and 0.5 in Atlanta.
Of the 98 patients, 69 (70%) were diagnosed as having primary FL (median age, 63 years), 19 had disseminated lymphoma (stage II and above) (median age, 63 years), and 10 presented with a relapse of systemic lymphoma in the ocular adnexal region. Of the 69 patients with primary FL, 44 (64%) were women, whereas 10 of the 19 patients (53%) with disseminated lymphoma were men. Primary lesions located only in the ocular adnexal region (stage IE) were found in 38 patients, whereas bilateral ocular adnexal region involvement and/or ipsilateral regional lymph node involvement (stage IIE) were found in 31 (Table 1).
In primary FL, the conjunctiva (20 of 69 [29%]), the orbit (18 of 69 [26%]), and the lacrimal gland (18 of 69 [26%]) were the most frequent involved sites (Table 2). The conjunctiva (7 of 19 [37%]) was the most common site in patients with disseminated lymphoma, whereas the lacrimal gland (6 of 10 [60%]) and orbit (4 of 10 [40%]) were the sites most frequently affected in patients with a systemic relapse in the ocular adnexal region. Swelling (62%) was the most common symptom (Figure 1). Symptom duration ranged from 1 to 108 months (median, 5.5 months).
The TNM staging was performed in the 69 patients with primary lymphoma (eTable 1 in the Supplement). Of these patients, 24 had stage T1 disease (conjunctival involvement only), 36 had stage T2 disease (orbital with or without the lacrimal gland), 7 had stage T3 disease (preseptal eyelids), and 2 had stage T4 disease (extra-orbital). The palpebral conjunctiva (T1bN0M0) was the most frequently involved site in stage T1 disease, whereas the anterior orbit and lacrimal gland (T2bN0M0) were the most frequently involved sites in stage T2 disease. The TNM stages were not associated with disease progression or survival.
Information regarding the histology grade was available in 93 of the 98 patients. Of these, 75 (81%) had grade 1 to 2 FL, 17 had grade 3A FL (Figure 1), and 1 had grade 3B FL. More relapses occurred in patients with primary FL with grade 3 tumors (n = 16) compared with those with grade 1 to 2 FL (n = 49) (P = .07, log-rank test).
Information on growth pattern was available in 78 of the 98 patients. Of these, 53 (68%) had a follicular growth pattern (follicular or follicular and diffuse), whereas 25 (32%) had a predominantly diffuse pattern (focally follicular or diffuse) (Figure 1). Neither disease-specific survival (P = .82, log-rank test) nor disease progression (P = .98, log-rank test) was associated with the growth pattern.
BCL2/IGH gene arrangement status was available in 21 of the 24 Danish patients with ocular adnexal FL. Sixteen patients (76%) tested t(14;18) positive, and 5 patients (24%) tested t(14;18) negative. The t(14;18) status was not associated with progression-free survival or overall survival.
Of the 69 primary FLs, 42% (29 of 69) were treated with EBRT only or with EBRT combined with chemotherapy (cyclophosphamide, hydroxydaunorubicin, vincristine [Oncovin], and prednisone [CHOP], alkylating agents, or unspecified chemotherapy) (26 of 69 [38%]). The EBRT plus chemotherapy regimen was more frequently used for patients with primary FL who were from Liverpool (18 of 28 [64%]) and India (3 of 9 [33%]) compared with Denmark (1 of 10 [10%]) and New York (0 of 4). Furthermore, EBRT and chemotherapy were more frequently used for patients with stage IIE disease compared with patients with stage IE disease (P = .01) (Table 3). The 19 patients with disseminated lymphoma primarily received systemic chemotherapy (n = 13), whereas patients with a localized relapse (n = 9) were most frequently treated with EBRT (n = 6). Data on initial therapy are summarized in Table 1.
Of the 98 patients with follicular OAL, information on EBRT doses was available in 16 Danish patients. The median dose was 26 Gy (range, 26-45 Gy). In the guidelines of all the participating eye cancer centers, the radiation doses varied from 20 to 45 Gy (eTable 2 in the Supplement).
Relapse or disease progression was observed in 32 of the 69 patients (46%) with primary lymphoma (Table 1). The time to relapse was accessible in 16 of these patients, with a median of 52 months during a follow-up period that ranged from 13 to 118 months. The median progression-free survival from the date of diagnosis was 5.3 years. Of the 32 patients with a relapse, 10 had recurrence within the ocular adnexal region, 9 had ocular adnexal relapse in conjunction with nodal and/or extranodal involvement, and 13 had recurrence in nodal and/or extranodal sites outside ocular adnexal region. Two of the 69 primary FLs (3%) transformed to diffuse large B-cell lymphoma. However, the survival of these 2 patients was not different from patients whose FL did not transform.
The overall survival rates at 10 and 20 years were 60% and 15% (median overall survival, 11 years), respectively, for the entire patient group, whereas the 10- and 20-year disease-specific survival rates were 69% and 29% (median disease-specific survival, 20 years), respectively. The disease-specific survival was not different in patients with primary, disseminated, and relapsed lymphoma (P = .47, log-rank test) (eFigure 1 in the Supplement). Furthermore, the disease-specific survival was not associated with the Ann Arbor stage of lymphoma (P = .60, log-rank test) (eFigure 2 in the Supplement).
Relapses were more frequent in patients with primary FL treated with chemotherapy plus EBRT compared with EBRT only (P = .001) (Table 3). This tendency persisted when patients were subdivided into those with stage IE (P = .003) and stage IIE (P = .06) lymphomas. In addition, the disease-specific survival was significantly better for patients treated with EBRT compared with patients treated with chemotherapy plus EBRT (10-year disease-specific survival, 94% vs 40%; P = .02, log-rank test) (Figure 2).
For the entire study group, 26 patients with FL (27%) died during follow-up. Of these, 16 patients (62%) died of lymphoma-related causes. The median time from initial diagnosis to lymphoma-related death was 5 years (range, 2-20 years).
The 10-year disease-specific survivals according to the eye cancer centers were 60% in Liverpool, 80% in Copenhagen, 100% in Melbourne, 83% in Hyderabad, 100% in New York, and 100% in Atlanta. The disease-specific survival was not significantly different among the centers (P = .30, log-rank test). However, the median follow-up time varied across the centers as follows: 5.5 years in Liverpool, 7.8 years in Copenhagen, 2.1 years in Melbourne, 5.5 years in Hyderabad, 1 year in New York, and 0.7 years in Atlanta.
The present cohort of 98 patients with ocular adnexal FLs was obtained from 6 eye cancer centers. This international, multicenter collaborative study is currently the largest reported collection of clinical data and outcomes of patients with ocular adnexal FLs. Most ocular adnexal FLs occurred in middle-aged and elderly patients (median age, 63 years), with a female predominance (female-male ratio, 1.6).
This finding is consistent with other published studies that contain wider subtypes of OALs13 and with studies of extranodal FLs.7,9 However, the median age was lower in patients from Hyderabad (India) compared with those from the other participating centers (predominantly Western countries). Patients from India also had a marked male predominance (male-female ratio, 2.8). Although the ratio of men alive at 63 years equals that of women,22 the relatively lower median age and male predominance observed in Indian patients are probably related to known variations with age and sex among lymphoma patients in different geographic regions.23
The female overrepresentation has been described to be particularly common in patients with lymphomas of the lacrimal gland.24 The lacrimal gland was more frequently involved in primary FLs compared with the most common OAL subtype, the extranodal marginal zone lymphoma, which primarily involves the orbit and conjunctiva.13,14 In addition, the frequency of FL has been reported to be relatively high in salivary gland lymphomas in conjunction with a higher frequency of female patients.25,26 Consequently, we may suggest that the distribution of lymphoma subtypes and male-female ratios of the lacrimal gland resemble those of the salivary glands.
Interestingly, the conjunctiva was the most common location affected in the group of ocular adnexal FLs being part of concurrent disseminated lymphoma. The conjunctiva is particularly important with regard to recirculation and homing of lymphocytes to the ocular adnexal region because the conjunctiva, along with the lacrimal gland and lacrimal drainage system, constitutes the eye-associated lymphoid tissues.27 Lymphoma dissemination has been found to be closely related to lymphocyte recirculation.28
The present study found the prognosis of FL that involves the ocular adnexal region in general to be favorable. Neither the Ann Arbor staging nor primary, disseminated, or relapsed lymphomas were correlated with recurrence or survival because of the indolent nature of FL.
External beam radiation therapy provided excellent disease control in primary ocular adnexal FL with a 10-year disease-specific survival rate of 94%, which is in accordance with studies of early-stage extranodal FL outside the ocular adnexal region.9,11 Interestingly, a significantly better disease-specific survival was found for patients with primary ocular adnexal FL treated with EBRT only compared with patients receiving EBRT plus chemotherapy (P = .02, log-rank test). This difference could not be explained by the different Ann Arbor stages (stage IE vs IIE), the TNM-based staging system for OAL (TN0M0 vs TN1M0 and/or TN2M0), or age. Consequently, either we are not differentiating the group of patients in an adequate way or the chemotherapy itself results in poorer survival. Similar observations have been noted in a recent OAL cohort study by Graue et al.21 In the present study, these questions could not be answered with the data available. The high frequency of patients with primary ocular adnexal FL receiving EBRT and chemotherapy instead of EBRT only is because one center (Liverpool) treated 18 of the 28 patients with primary ocular adnexal FL with EBRT combined with chemotherapy. We acknowledge that the present study is an analysis of a retrospective cohort series that spans a 30-year period, which limits interpretation when comparing different types of treatments.
The recommended total EBRT doses according to the guidelines of the participating eye cancer centers differed from 20 to 45 Gy. This difference in EBRT doses did not affect the survival in the present study. Therefore, a total radiation dose of 20 to 32 Gy is recommended to minimize radiation-induced ophthalmic complications.29,30
Current guidelines recommend rituximab in combination with either CHOP or cyclophosphamide, vincristine, and prednisone for patients with FL and a high tumor burden according to the FLIPI.31 Rituximab alone or in combination with alkylating chemotherapy was only applied in 1 patient with primary ocular adnexal FL and in 2 patients with disseminated lymphoma. The relatively low number of patients with disseminated lymphoma who received rituximab reflects that the present study spans a 30-year period.
We evaluated the feasibility of the new TNM-based staging system for OALs in a large cohort of patients with primary ocular adnexal FLs. The TNM-based system enabled a precise definition of the anatomical location of lesions and involvement of regional lymph nodes and adjacent tissues, such as the parotid gland. However, no associations were found between the site-specific FL location within the ocular adnexal region (as depicted by the TNM-based staging system), local or systemic relapse frequency, or disease-specific survival.
This study was not able to identify other adverse factors associated with recurrence or survival, probably because the median follow-up time was only 5.5 years, whereas the median overall survival was 11 years. Thus, it will remain unclear whether the TNM-based staging system is associated with prognosis in ocular adnexal FL until future studies with longer follow-up times are conducted. Therefore, the Ann Arbor staging system and type of treatment remain the best and simplest tools for predicting the course of disease rather than tumor size or ocular adnexal region location.
Follicular lymphoma that involves the ocular adnexal region is relatively rare, predominantly affecting middle-aged to elderly patients and females. Although relapses are relatively frequent in this patient group, the long-term survival is favorable for most patients. Primary EBRT appears most likely to control early-stage FL that involves the ocular adnexal region.
Submitted for Publication: August 25, 2013; final revision received November 21, 2013; accepted December 16, 2013.
Corresponding Author: Steffen Heegaard, MD, DMSc, Eye Pathology Institute, Department of Neuroscience and Pharmacology, University of Copenhagen, Frederik V’s Vej 11, 1, DK-2100, Copenhagen, Denmark (firstname.lastname@example.org).
Published Online: April 24, 2014. doi:10.1001/jamaophthalmol.2014.376.
Author Contributions: Dr Rasmussen had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Rasmussen, Sjö, Heegaard.
Acquisition, analysis, and interpretation of data: Rasmussen, Coupland, Finger, Graue, Grossniklaus, Honavar, McKelvie, Mulay, Prause, Ralfkiaer, Sjö.
Drafting of the manuscript: Rasmussen, Coupland, Finger, Graue.
Critical revision of the manuscript for important intellectual content: Grossniklaus, Honavar, McKelvie, Mulay, Prause, Ralfkiaer, Sjö, Heegaard.
Statistical analysis: Rasmussen, Graue.
Obtained funding: Rasmussen, Finger.
Administrative, technical, or material support: Grossniklaus, Mulay, Sjö.
Study supervision: Coupland, Finger, Graue, Prause, Sjö, Heegaard.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by grants from Fight for Sight Denmark, the Danish Cancer Society, the Danish Eye Research Foundation, Synoptik Foundation, the Danish Foundation for Cancer Research, Engineer Lars Andersens Foundation, the A.P. Møller Foundation for the Advancement of Medical Science, and the Merchant Kjaer and Wife Kjaer, born la Cour-Holmens Foundation (Dr Rasmussen) and The Eye Cancer Foundation Inc (Dr Graue).
Role of the Sponsors: The funding sources 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.