Patient characteristics and clinical presentation in association with treatment procedure
Overall, 103 patients were diagnosed with POAML, of which 11 were excluded from analysis because of lost follow-up, 11 due to missing data, and 6 due to short follow-up within 1 year. As a result, 75 patients were analyzed in this study. Patient characteristics are shown in Table 1. The median age was 68 years (range: 26–92 years), and 41 patients (54.7%) were females. The most common involved sites were the conjunctiva (42.7%), followed by the orbit (36.0%), lacrimal gland (12.0%), and eyelid (8.0%). Of the 75 patients, 69 (92.0%) presented with stage IE disease, followed by 3 (4.0%) with stage IVE, 2 (2.7%) with stage IIE, and 1 with stage IIIE. In addition, 61 patients (81.3%) presented with a unilateral ocular region, 14 (18.7%) presented with bilateral disease regions, 5 (6.7%) had complications with an immunoglobulin G4 (IgG4)-related disease, and 4 of them had lymphoma involvement in the lacrimal gland. According to the IPI, the majority of patients (69, 92.0%) were classified as low risk, 3 patients (4.0%) as low-intermediate risk, 2 (2.7%) as high-intermediate risk, and 1 (1.3%) as high risk. The most frequent complaint at the initial presentation was the recognition of a tumor without symptoms in 29 patients, followed by swelling in 19, hyperemia in 9, discomfort in 7, and other symptoms in 11. At tumor biopsy for histopathologic diagnosis, the tumor was completely resected in 17 patients, while a residual tumor existed after biopsy in 58 patients (Table 1).
WW was the most frequent first-line modality, selected for 29 (38.7%) of 75 patients, including 14 patients with compete tumor resection at biopsy. In the remaining 46 patients (61.3%), the primary treatment modalities were RT in 24 (52.0%) (RT group), rituximab monotherapy in 19 (41.3%) (R group), and other modalities in 3 (6.5%), including systemic chemotherapy with rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisolone (CHOP)-like chemotherapy in 2 patients and discontinuation of methotrexate in 1 patient with rheumatoid arthritis.
Of the 29 patients in the WW group, 19 (65.5%) had primary lesions in the conjunctiva, which was more frequent compared with the RT or rituximab (R) group. In addition, the tumor was completely resected in 14 patients (48.3%) at diagnosis, which was also more frequent compared with the RT or R group. All patients except one had stage ⅠE disease in the WW group. Although 6 (20.7%) of the 29 patients complained of eye discomfort with foreign-body sensation caused by lymphoma, they showed no treatment-emergent symptom that could lead to vision impairment, such as ptosis, diplopia, or proptosis.
In the RT and R groups, 43 of the 46 patients had a residual tumor after biopsy. In the RT group, the orbit was the most frequent site of lymphoma, which caused treatment-emergent subjective symptoms that potentially impair visual function. The median doses of total and fractional radiation were 30 Gy (range: 24–30 Gy) and 2 Gy (range: 1.5–2 Gy), respectively. Rituximab was administered intravenously at a dose of 375 mg/m2 once weekly up to eight doses. Patients with bilateral disease or advanced disease tended to undergo rituximab monotherapy.
Treatment response
As shown in Table 2, the RT group successfully obtained local disease control. A CR or complete metabolic response (CMR) was achieved in 19 (79.2%) of the 24 patients. The ORR in the R group was 73.7%, and the CR or CMR was 42.1%. However, five patients (26.3%) in the R group failed in achieving an objective response and two of them underwent additional RT. Two patients who received systemic chemotherapy combined with rituximab achieved CR, and one patient with methotrexate withdrawal showed a partial response; these three experienced no disease progression or recurrence during the observation period.
Survival outcome
With a median follow-up of 48.8 months, new treatment was added for 20 patients, including 16 with disease progression and 4 without disease progression, who underwent new treatment due to their preference. Overall, the median TTNT was not achieved in both WW and RT groups and was 69.0 months in the R group (Figure 1A). The TTNT tended to be longer in the RT group than in the WW and R groups; however, the differences were not statistically significant. At 60 months, the estimated proportion of patients not requiring new treatment was 69.4% (95% CI: 47.4–83.6) in the WW group, 85.2% (95% CI: 60.2–95.1) in the RT group, and 53.8% (95% CI: 26.0–75.2) in the R group. The median TTNT was 20.3 months (range: 5.9–139.3 months) in the 16 patients with disease progression.
The median PFS was not achieved in both WW and RT groups and was 45.0 months in the R group (WW vs. R group, p = 0.375; WW vs. RT group, p = 0.301; RT vs. R group, p = 0.079) (Figure 1B). The 5–year PFS was 69.5% (95% CI: 43.9%–85.2%) in the WW group, 67.4% (95% CI: 34.7%–86.3%) in the RT group, and 41.4% (95% CI: 17.4%–64.1%) in the R group. All patients enrolled in this study were alive at the time of analysis, so the OS was 100% during the observation period.
Clinical course and predictive factor for disease progression in the WW group
Of the 29 patients in the WW group, 9 (31.0%) showed disease progression or recurrence, including 1 patient with systemic progression (Table 3). Of these nine patients, seven received treatment with a median TTNT of 15.6 months (range: 5.9–139.3 months): three received RT, three underwent surgical resection, and one received rituximab monotherapy, and the disease was well-controlled in all seven. The remaining two patients continued WW even after disease progression, because of no subjective symptom or organ impairment. Neither histological transformation nor mortality was observed. In the remaining 20 patients without disease progression, 1 received RT because of the patient’s preference 19 months after diagnosis and 1 received rituximab monotherapy, expecting improvement of lacrimation 16 months after diagnosis. At the time of the last follow-up, 20 (69.0%) of 29 patients did not require treatment.
To identify predictive factors for disease progression in the WW group, we compared the clinical characteristics of 9 patients with disease progression and 20 without disease progression or recurrence. We analyzed age, gender, involved site, laterality, degree of resection, and tumor size evaluated by the detection of post-biopsy tumor involvement via CT, MRI, or PET-CT. As shown in Table 4, all these factors were not significantly associated with disease progression or recurrence in the WW group. In addition, the degree of resection on diagnosis, the involved site, laterality, and the tumor size did not significantly affect the PFS in the WW group (Figure 2).
AEs with RT and rituximab
In the RT group, 16 of 24 patients (66.7%) developed AEs: conjunctivitis (45.8%), radiodermatitis (41.7%), cataract (33.3%), and dry eye (16.7%). Most AEs were generally mild (grade 1 or 2), while grade 3 cataract occurred in five patients. The median time from RT initiation to onset of cataract was 35 months, and 5 patients underwent cataract surgery. In the R group, infusion-related reaction occurred in only 1 of 19 patients (5.3%). No other AEs due to rituximab occurred.