Survival curves for all patients. A, Disease-free survival (DFS) curve for all patients. Kaplan-Meier estimates at 48 and 72 months were 74.8% and 69.1%, respectively. B, Overall survival (OS) curve for all patients. Kaplan-Meier estimates at 16 and 36 months were 93.4% and 89.0%, respectively. Orange hatch marks indicate censored data during follow-up.
Survival curves according to the histopathologic classification. A, Disease-free survival (DFS) according to the histopathologic classification (hazard ratio, 5.128; P < .001). B, Overall survival (OS) according to the histopathologic classification (hazard ratio, 18.868; P < .001). The green line indicates patients with mucosa-associated lymphoid tissue (MALT) lymphoma; blue line, patients with non-MALT lymphoma. Orange hatch marks indicate censored data during follow-up.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Oh D, Kim Y. Lymphoproliferative Diseases of the Ocular Adnexa in Korea. Arch Ophthalmol. 2007;125(12):1668–1673. doi:10.1001/archopht.125.12.1668
Copyright 2007 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2007
To evaluate the clinical features, treatments, and outcomes of patients with ocular lymphoproliferative disease classified according to the World Health Organization classification and to determine prognostic factors of this disease in South Korea.
Between March 1, 1995, and December 31, 2005, 128 cases of patients with ocular lymphoid tumors treated at the Samsung Medical Center were retrospectively reviewed.
The mean patient age was 46.3 years (range, 1-87 years). The major histopathologic subtypes were mucosa-associated lymphoid tissue (MALT) lymphoma in 96 patients (75.0%), lymphoid hyperplasia in 11 (8.6%), diffuse large B-cell lymphoma in 6 (4.7%), and mantle cell lymphoma in 4 (3.1%). Ocular lymphoproliferative lesions were located in the conjunctiva (53 patients), the eyelid (33 patients), and the orbit (42 patients). Twenty patients had tumor relapses (15.6%), and 9 died of lymphoma during follow-up (7.0%). Regarding the analysis of prognostic factors, most patients with MALT lymphoma evidenced local disease, required local treatment, and exhibited a superior prognosis.
Lymphomas of the MALT type constitute most ocular adnexal lymphoproliferative diseases and occur more frequently in South Korea than in Western countries. Patients with MALT lymphoma have favorable outcomes compared with patients with other types of lymphoma.
Lymphoproliferative disease of the ocular adnexa is relatively uncommon. Ocular lymphoma constitutes less than 1% of all non-Hodgkin lymphomas (NHLs) and 5% to 15% of all extranodal cases.1 Most of these are primary tumors and are usually NHLs of B-cell type. It has been recognized that most ocular adnexal lymphomas are low-grade, extranodal, marginal zone B-cell lymphomas of the mucosa-associated lymphoid tissue (MALT) type.2,3
Previous studies have reported that the relative incidence of MALT lymphomas may vary in different populations. In a National Institutes of Health–sponsored epidemiologic study4 of 1378 NHLs using the REAL (Revised European American Lymphoma) classification, the relative incidence of MALT lymphomas varied considerably from 3% in London to 10% in Hong Kong, although these differences were not reported to be statistically significant. Similarly, some geographic variation in the distribution of ocular lymphoma may exist. The incidence of MALT lymphoma varies from 35% to 70% in Western populations and from 79.5% to 86% in Asians.2,3,5- 8 The primary objectives of this study are to summarize the clinical features of ocular lymphoproliferative diseases in Korea and to determine prognostic factors that affect survival.
We conducted a retrospective review of patients with ocular lymphoproliferative diseases who were treated at the Samsung Medical Center between March 1, 1995, and December 31, 2005. Diagnoses of ocular lymphoma were based on the results of histopathologic examinations using immunohistochemical analysis via surgical biopsy. Patients with follow-up of less than 6 months were excluded from this study unless they had died.
All the patients were categorized according to the classifications established by the World Health Organization.9 Data collected included patient age, sex, initial symptoms and signs, and sites of involvement. Staging workups were performed after the consultation with the hemato-oncology department. These patients were staged according to the Ann Arbor classification.10 The performance status of the patients was determined according to the Eastern Cooperative Oncology Group classification,11 and a search for B symptoms (fever, night sweats, and weight loss) was included in the staging workup. When data were available, the International Prognostic Index (IPI) score12 was defined for each patient. Lymphoma grade was classified according to the working formulation of the Non-Hodgkin's Lymphoma Pathologic Classification Project.13
Treatment modalities and efficacies are also analyzed herein. In cases in which radiotherapy was applied, ophthalmic complications resulting from local radiotherapy were examined. In cases of recurrence, the site and time of recurrence was determined. The mortality rate and cause of death were recorded.
In this study, all the patients were reclassified into either group A (low grade) or group B (high grade) for analysis of the prognosis. Low-grade lymphoma was evaluated according to the working formulation, and lymphoid hyperplasia was reclassified into group A. Intermediate- and high-grade lymphomas, as determined by the working formulation, were reclassified into group B. Correlations were analyzed between reclassified grades and a variety of variables, including age, sex, involvement site, metastasis, stage, B symptoms, histopathologic type, IPI scores, and treatment modalities.
Disease-free survival (DFS) was defined as the interval between the date of diagnosis of lymphoproliferative disorder and the date of the first relapse or death. Overall survival (OS) was defined as the time elapsing from the date of diagnosis of lymphoproliferative disorder to the date of death. Survival curves were estimated using the Kaplan-Meier method, and the level of significance among different outcomes was evaluated via log-rank test. Univariate analyses were performed to identify prognostic factors for DFS and OS. Multivariate analyses were conducted whenever possible.
A total of 149 cases were reviewed, and 128 were included in this study. Twenty-one patients were excluded from the study owing to follow-up loss after diagnosis in 11 patients (MALT type in 9 patients, lymphoid hyperplasia in 1, and diffuse large B-cell lymphoma in 1) and transfer to another hospital for regional reasons in 10 patients with stage IE MALT lymphoma for radiotherapy.
There were 61 women and 67 men (Table 1). The mean age of the patients was 46.3 years (range, 1-87 years; median, 44.5 years). Mean follow-up was 35.1 months (range, 6-120 months; median, 22.5 months). The symptoms and signs exhibited initially included a palpable mass in 45 patients, a salmon patch in 45, proptosis in 27, and eyelid swelling in 24. The tumor was located in the conjunctiva in 53 patients (41.4%), the orbit in 42 (32.8%), and the eyelid in 33 (25.8%). Bilateral involvements were observed in 26 patients (20.3%).
The histopathologic review according to the World Health Organization classification showed 96 patients (75.0%) with MALT lymphoma, 11 (8.6%) with lymphoid hyperplasia, 6 (4.7%) with diffuse large B-cell lymphoma, 4 (3.1%) with mantle cell lymphoma, and others (Table 2). Excluding 11 patients with lymphoid hyperplasia, 117 patients with malignant lymphomas were graded according to the working formulation. Ninety-seven of these patients (82.9%) were classified as low grade. Ninety-six of these 97 patients were found to have MALT lymphoma, and the remaining patient had lymphoplasmacytic lymphoma. Ten of 117 patients (8.5%) each (20 patients total) had intermediate- and high-grade lymphomas.
The staging workup was conducted on 115 patients (Table 1). Ninety-eight patients (85.2%) had stage IE disease. Only 17 patients (14.8%) showed metastasis at the time of diagnosis. The metastasis sites included bone marrow in 11 patients, adjacent lymph nodes in 7, distant lymph nodes in 5, the chest in 4, and the liver in 3. Six of 115 patients (5.2%) exhibited B symptoms.
A total of 106 patients were treated with local radiotherapy (Table 3). Radiotherapy constituted the sole treatment for 92 of the 128 patients (71.9%) and was combined with chemotherapy in 14 patients. Radiotherapy doses ranged from 2000 to 4500 cGy. Twelve patients (9.4%) received chemotherapy. One hundred eleven patients (86.7%) exhibited complete regression after initial treatment. Radiotherapy was generally well tolerated. However, ophthalmic complications developed in 41 patients. Severe dry eye was noted in 25 patients. Nine patients developed cataracts, and 4 underwent cataract surgery. Radiation retinopathy was noted in 7 patients.
Twenty of the 128 patients (15.6%) who had initially shown complete responses relapsed. Eight patients had local tumor relapses, and 12 (9.4%) developed systemic lymphoma. Only 10 patients (10.4%) of the 96 patients with MALT lymphoma experienced recurrence: 4 (4.2%) had distant recurrences and the other 6 had local relapses. In contrast, 10 of 21 patients (47.6%) with non-MALT lymphoma had recurrences, and 8 (38.1%) had distant recurrences. During mean follow-up of 35.1 months, DFS for all patients was 87.8%, 74.8%, and 69.1% at 16, 48, and 72 months, respectively (Figure 1A). Overall survival at 16 and 36 months for all patients was 93.4% and 89.0%, respectively (Figure 1B). Of the 10 patients (7.8%) who died during follow-up, 9 died of lymphoma and 1 of another cause. Of those, 8 patients died of their high-grade lymphoma during chemotherapy. The remaining patient with MALT lymphoma died of recurrence of the disease.
In this study, 108 patients were reclassified into group A (low grade), and the other 20 into group B (high grade) for prognostic analysis (Table 1). Group A lymphomas were significantly associated with conjunctival involvement, stage IE disease, B-symptom negativity, lower IPI scores, and lack of metastasis. The DFS and OS rates were significantly better for low-grade lymphomas than for high-grade lymphomas.
We analyzed data regarding variable factors and the hazard ratio affecting survival according to the results of Kaplan-Meier estimates, log-rank tests, and multiple Cox proportional hazards regression models (Table 4). The univariate analysis for DFS uncovered several significant factors. Non-MALT type (Figure 2), higher stage, metastasis, the presence of B symptoms, and higher IPI scores were found to adversely affect DFS. The univariate analysis for OS yielded results similar to those for DFS, but patient age older than 60 years was included as a significant factor for poor prognosis. Male sex and eyelid involvement were factors suggestive of poor outcomes. However, these 2 factors did not have statistical significance. In the multivariate models for DFS, the only variable with statistical significance was advanced patient age at the time of diagnosis (risk ratio, 5.56). Regarding OS results, age older than 60 years (risk ratio, 21.74) and higher IPI scores (risk ratio, 166.67) were correlated with unfavorable outcomes.
This retrospective study of patients treated at a large medical center in South Korea reviewed 128 cases of patients with ocular lymphoid tumors. Most tumors were MALT lymphomas, and the incidence of these lymphomas was higher than in Western countries. The latter finding may be due to genetic factors associated with the South Korean population, since there is general agreement that racial variations may affect the characteristics and prognosis of ocular adnexal lymphomas.
Most ocular lymphomas are non-Hodgkin B-cell type, and less than 1% to 3% are T-cell type.2,14 Burkitt and Hodgkin lymphomas in the ocular adnexa are rare except in endemic regions.15 In this study, almost all the patients (96.1%) had B-cell NHLs, and only 5 (3.9%) had T-cell types. The major histopathologic subtype was MALT lymphoma (75.0%), followed by lymphoid hyperplasia (8.6%), diffuse large B-cell lymphoma (4.7%), and mantle cell lymphoma (3.1%). In this study of malignant lymphomas, 82.1% were MALT type, much higher than in Western countries, where the incidence is reported to be 35% to 70%.2,3,5- 8 However, according to previous studies in Asian populations, MALT lymphomas have a 79.5% to 91% incidence,2,3,5- 8 similar to what we found. In the West, follicular and lymphoplasmacytic lymphomas are relatively common, with an incidence of 11%to 53%2,3,16 and 4% to 24%,2,3,16 respectively. Only 1 patient was found to have lymphoplasmacytic lymphoma in the present study. In the West, diffuse large B-cell lymphomas represent approximately 7% to 15% of all ocular lymphomas, and they are the second most common lymphoma in the ocular adnexa.5,16- 19 In the present study of South Korean patients, diffuse large B-cell lymphomas represented 4.7% of all lymphoproliferative diseases and were the second most common lymphoma. These findings support the idea that racial differences affect lymphoma incidence, although the underlying cause of such differences is still unknown. Long-term antigen stimulation has been implicated as a causative agent in the development of some mature B-cell lymphoproliferative processes, as with Helicobacter pylori infection and MALT lymphoma of the stomach.20,21 Similarly, Borrelia burgdorferi and Campylobacter jejuni infections are associated with cutaneous marginal zone B-cell lymphoma and immunoproliferative disease, respectively.22,23 A possible association of Chlamydia psittaci with ocular MALT lymphoma has been reported. Ferreri et al24 detected C psittaci DNA in 80% of patients with ocular adnexal MALT lymphoma but in only 12% with reactive lymphoid hyperplasia. However, other studies25,26 of larger cohorts did not confirm this association. Recently, Chanudet et al27 suggested that there may be geographic variability in C psittaci distribution and that in different subtypes of ocular adnexal lymphoma, the bacterium seemed to be preferentially associated with MALT lymphoma. Infection with H pylori has not been widely associated with ocular adnexal lymphoma, although H pylori DNA was detected in tumor cells in 4 of 5 patients with conjunctival MALT lymphoma.28 Other studies29 have suggested that the presence of gastric H pylori did not affect clinical manifestation, even if they were common in patients with ocular adnexal lymphoma. Helicobacter pylori seropositivity has been reported in 45% of B-cell NHLs diagnosed in southern Switzerland, and it has been suggested that H pylori contributes to the pathogenesis of MALT lymphomas arising at different extranodal organs.30,31 As many as 90% of South Koreans are H pylori positive,32 much higher than in Western populations, where 43% of healthy individuals are H pylori positive.24,29,30 Thus, H pylori infection could be related to the higher prevalence of ocular adnexal lymphoma of MALT type in South Koreans. A study should be performed to elucidate the association of chronic H pylori infection with the differences in the incidence of ocular adnexal lymphomas in different populations.
The mean age of the South Korean patients with lymphoma in this study was 46.3 years, younger than the mean age of patients in Western studies (60-70 years)5,6,16,17 but similar to the mean age of Asian patients in other studies (40-50 years).7,8,33 These findings indicate that ocular adnexal lymphomas develop in Asians at a younger age than in Westerners and that the higher proportion of MALT lymphomas may be related to a younger age at onset. In addition, the mean age at onset of follicular or lymphoplasmacytic lymphoma is 50 to 70 years,17,19 and these lymphomas were rare in this study.
Although some studies have reported a higher rate of orbital lymphomas than those at other sites, the present study showed a higher rate of conjunctival involvement. The frequency of involvement has been reported as follows: orbit, 46% to 74%; conjunctiva, 20% to 33%; and eyelid, 5% to 20%.2,17,34 Differences may be attributed to the fact that in South Korean patients, most MALT lymphomas show conjunctival involvement. Prognosis for ocular adnexal lymphomas is related to the primary location. Eyelid lymphomas have a more aggressive course than orbital or conjunctival disease, and conjunctival lymphomas have a better prognosis than orbital or eyelid disease.6,18,35 In the present study, conjunctival MALT lymphomas also showed a better prognosis than those at other sites, although the univariate and multivariate analyses did not show that the site was a statistically significant predictor of survival.
In this study, 93.8% of MALT lymphomas were stage IEa, and all except 4 were treated with local radiotherapy. Patients with non-MALT lymphomas were treated with chemotherapy or combined chemotherapy and radiotherapy. Among MALT lymphomas, 92.7% of patients showed complete remission. We found no difference between treatment modality and response in patients with MALT lymphoma. In contrast to MALT lymphoma, only 52.4% of patients with non-MALT lymphoma achieved complete remission. Studies17,34 have demonstrated that with sufficiently long follow-up, one-third to one-half of patients with ocular adnexal lymphoma develop systemic lymphoma. Although follow-up was relatively short (median, 22.5 months), only 15.6% of patients had recurrences and 9.4% had systemic progression in this study. The recurrence rate was 10.4% for patients with MALT lymphoma, and only 1 patient with MALT lymphoma died of disease. In contrast, 47.6% of patients with non-MALT lymphoma had recurrence, and 38.1% died. Systemic progression was 4.2% for patients with MALT lymphoma. In contrast, 38.1% of patients with non-MALT lymphoma had systemic progression during follow-up. Thus, MALT lymphomas have an indolent natural history, and radiotherapy alone can achieve excellent local control. A local control rate of 98% to 100% and a 75% distant relapse-free survival for stage IE MALT lymphoma was described recently.36,37 The excellent prognosis of MALT lymphomas in this study is in agreement with previous studies.
Various factors are thought to affect the clinical course of ocular adnexal lymphoma: site, stage, lymphoma subtype, tumor cell growth fraction, presence of p53, and presence of serum lactate dehydrogenase. Several studies have statistically analyzed the outcomes of ocular adnexal lymphoma. Coupland et al38 found that stage of disease at initial examination, cytologic atypia, MIB-1 proliferation rate, and p53-positive tumor cells correlated with lymphoma-related death. Jenkins et al3 found a significant correlation between tumor histologic features and frequency of extraorbital disease and lymphoma-related death. Using multivariate analysis, McKelvie et al17 found that histologic type and stage of disease at initial examination were the only significant independent variables. Finally, Jenkins et al,35 in a larger study, showed that clinical features, including previous or concurrent systemic disease, pain, and bilateral disease, had a significant association with the poor prognosis of lymphoma. In this study, cases were reclassified for disease prognosis analysis. All the patients in reclassified group A except those with reactive lymphoid hyperplasia (n = 11) and lymphoplasmacytic lymphoma (n = 1) had MALT lymphomas (n = 96). In group A, the lymphomas had a tendency toward more conjunctival involvement, stage IEa, negative B symptoms, low IPI scores, and less metastasis (Table 1). In fact, most South Korean MALT lymphomas can be characterized as conjunctival, stage IEa, low-risk lymphomas. These lymphomas mostly showed a benign indolent clinical course and did not develop extraorbital dissemination during follow-up. In contrast, lymphomas that are high grade at the time of diagnosis are significantly correlated with an increased risk ratio for persistent disease, recurrence, and lymphoma-related death. In addition, survival analysis showed that unfavorable predictors of survival and outcome were non-MALT lymphoma, higher stage, metastasis, B-symptom positivity, higher IPI scores, and being older than 60 years (Table 4). In this study, patient sex, site of involvement, and treatment modality did not have a statistically significant correlation with survival or outcome.
In conclusion, most ocular adnexal lymphomas in South Korea are MALT type, low-grade lymphomas that frequently involve the conjunctiva. Follicular and lymphoplasmacytic lymphomas, which are relatively common in Western countries, are rarely found in South Korean patients. The South Korean patients in this study with ocular adnexal lymphomas were younger than Western patients. We found that South Korean patients with MALT lymphomas were likely to have a favorable outcome.
Correspondence: Yoon-Duck Kim, MD, Department of Ophthalmology, Samsung Medical Center, 50 Ilwon-dong, Kangnam-gu, Seoul 135-710, South Korea (firstname.lastname@example.org).
Submitted for Publication: May 31, 2007; final revision received July 3, 2007; accepted July 5, 2007.
Financial Disclosure: None reported.
Previous Presentation: This study was presented at the American Society of Ophthalmic Plastic and Reconstructive Surgery Fall Scientific Symposium; November 15, 2006; Las Vegas, Nevada.