Orbital lymphomas are mostly indolent, low-grade lesions. Radiotherapy (RT) is the most effective treatment for local disease either as the sole treatment for low-grade lymphoma or in combination with chemotherapy for intermediate- and high-grade lymphoma. The orbital contents are at risk of radiation toxicity ranging from painful lesions such as dry eye and glaucoma to vision impairment secondary to either cataract formation or, more rarely, other untreatable causes. Since 1997, all patients with orbital lymphoma referred to the RT unit at Sheba Medical Center have undergone conformal CT-based RT planning. Patients with lymphoma limited to one part of the orbit were treated to partial orbital volumes to lessen the radiation dose to the unaffected areas of the orbit, and thereby reduce the possible treatment toxicity. Patients with more extensive intraorbital lymphomas received whole orbit RT with lens shielding when appropriate. To test the hypothesis that partial orbit RT is effective and less toxic than whole orbit RT, we reviewed the charts of patients with orbital lymphoma treated during this period and compared the outcome and toxicity in the group of patients receiving partial orbit RT with those who received whole orbit RT. Conformal radiotherapy (RT) has been used for all patients with orbital lymphoma treated at our institution since 1997. We retrospectively reviewed the charts of 23 consecutive patients to test the hypothesis that partial orbit RT is effective and less toxic than whole orbit RT.

Methods and materials

Patients underwent contrast-enhanced, thin-slice, diagnostic CT of the orbit before treatment. For the RT planning, patients were immobilized with an Aquaplast mask, and CT was obtained in the treatment position on a dedicated CT scanner in the RT suite. Before 2001, a Multidata DSS 2.5 planning system was used. Since 2001, treatment planning has been done on a Varian Eclipse treatment planning system. RT was delivered in fractions of 170–200 cGy five times weekly. The median dose was 25.2 Gy (range, 20–30 Gy) for low-grade lesions in both the whole orbit and the partial orbit RT groups and 24–40 Gy (median, 39.6 Gy) for intermediate- and high-grade lesions

Twelve patients had lesions limited to part of the orbit (usually unilateral), allowing them to receive RT that spared uninvolved areas of the orbit (Fig. 1). Eleven patients (12 eyes) with more widespread lesions received RT to the entire orbit. Eight were treated via an anterior photon or electron field, four with a pencil beam block, and three with conformal fields planned to reduce the dose to the lens and brain. The prechemotherapy volume was used to plan RT for the 2 patients who had received chemotherapy; one received whole orbit RT and one, with high-grade lymphoma in the lacrimal gland, received partial orbit RT.

Twelve patients with limited lesions were treated to partial orbital volumes and 11 patients (1 with bilateral disease) with more extensive lesions received whole orbit RT. The dose was 20–30 Gy (median, 25.2 Gy) for 19 patients with low-grade lymphoma and 24–40 Gy (median, 39.6 Gy) for 5 patients with intermediate- to high-grade lymphoma. The follow-up was 12–68 months (median, 34 months).

Results

All patients had a complete response to RT. Intraorbital recurrence developed in previously uninvolved areas not included in the initial target volume in 4 patients (33%) treated with partial orbit RT. All were salvaged by repeat RT or surgery. No patient treated with whole orbit RT developed intraorbital recurrence. The acute and long-term toxicity was similar in both groups. All but 1 patient retained good vision.

Toxicity was graded from the patients' charts according to the Common Toxicity Criteria, version 2.0, 1998. All patients had conjunctivitis and tearing that was mostly Grade 1. Four patients (33%) treated to the entire orbit and 3 (25%) who received partial orbit RT had Grade 2 toxicity. Acute toxicity resolved within a few weeks of completing therapy. Late toxicity included dry eye requiring artificial tears (Grade 2) in 2 patients in each group and Grade 2 corneal erosion followed by asymptomatic scarring in 1 patient who had received 25.2-Gy whole orbit RT. All patients retained good vision, except for 1 patient with intermediate-grade lymphoma who had received systemic (cyclophosphamide, hydroxydaunomycin, vincristine, and prednisone) and intrathecal (methotrexate) chemotherapy and had deteriorating vision at the time of referral for RT. She became blind in the affected eye during the first week of RT and went on to receive a dose of 40 Gy to the entire eye. A Grade 2 cataract developed in a 70-year-old patient who had received 25.2 Gy RT to the entire orbit and in a 54-year-old patient with high-grade lymphoma who had received 40 Gy to the area of the lacrimal gland and lateral orbit via a three-field conformal technique. The dose to the lens ranged from 16 to 19 Gy.

Radiotherapy has been advocated for orbital lymphoma for many years. As in other series of orbital lymphoma, the great majority of our patients had low-grade, small-cell lymphoma Low-grade orbital lymphomas are highly radiosensitive lesions with overall local control rates of >90% in reported series (reviewed by Bhatia  and Martinet). In older series, many patients were treated with radiation doses of around 40 Gy, but recently, the trend has been toward the use of lower radiation doses without any reduction in the local control rate. Little data are available regarding the lower limit of effective doses for low-grade orbital lymphoma, although in one report, 3 of 11 patients treated with doses <20 Gy developed recurrence compared with none of those who received >30 Gy, usually via anterior fields (5). In another series, 1 of 6 patients who received <24 Gy developed recurrence compared with none of 14 who received >24 Gy

Our series, albeit nonrandomized and with a small number of patients, is the first published series that analyzed the relapse pattern in patients treated in the modern era with conformal partial orbit RT. Before the widespread availability of three-dimensional conformal RT, most patients with orbital lymphoma were treated with RT to the entire orbit, although lens-sparing techniques such as an anterior field with a central lens block for unilateral disease or split beam lateral fields posterior to the lens for patients with bilateral lesions were commonly used. In recent years, several groups have published their results of RT for orbital lymphoma using modern techniques, but none of these series investigated the role of partial orbit RT in these patients.

Several individual cases of intraorbital recurrence outside the target volume or in areas of the orbit that were underdosed have been noted, but it is sometimes difficult to determine the denominator of total cases treated with partial orbit RT in these series. Bolek reported on 1 patient with low-grade lymphoma in the lacrimal gland that recurred in the untreated part of the orbit. Uno  reported on 1 patient who developed conjunctival recurrence both under and adjacent to the shielded area of the eye. The Rare Cancer Network multicenter study of 90 orbital lymphomas reported that 20 patients were treated with focal RT . One of these, an intermediate-grade lymphoma, recurred in the orbit and was considered a geographic miss. Four (33%) of 12 patients in our series who were treated with partial orbit RT developed relapse at sites in the ipsilateral orbit that were not included in the original target volume.

Two of 14 patients with low-grade lymphoma in the Rare Cancer Network study had recurrence in the contralateral orbit. Contralateral disease is seen in >10% of patients with orbital lymphoma . It is, therefore, reasonable to suggest that some patients with seemingly focal intraorbital lymphoma as seen on CT may harbor occult microscopic disease at other intraorbital sites.

The use of conformal RT did not result in a lower incidence of either acute or chronic side effects compared with the group receiving whole orbit RT. The incidence of long-term toxicity at doses of 25.2 Gy has been minimal. Grade 2 conjunctivitis was seen in about one-third of patients in both groups in our study. The Rare Cancer Network study reported 56 of 90 patients with conjunctivitis, but these patients had received a median dose of 34 Gy. Stafford  recently updated the Mayo Clinic experience and noted a 52% rate of acute complications even in patients who received a dose of <30 Gy. However, no late complications (except for asymptomatic cataract development and mild dry eye) were seen at doses of <35 Gy. At doses of ?25 Gy, by far the most common long-term side effect of orbit RT is cataract formation. Henk   studied 40 patients with orbital lymphoma and found that the median toxic dose for cataract formation was 15 Gy and that lens opacities usually appeared 3–9 years after RT. In the Rare Cancer Network study, the incidence of cataract formation (irrespective of the use of lens shielding) was 30%. The follow-up in our series was relatively short (median follow-up 34 months); therefore, it is highly possible that more patients will develop symptomatic cataracts with additional observation.

The incidence of severe dry eye syndrome at doses of <30 Gy to the lacrimal apparatus is almost nonexistent. In our series, moderate symptoms requiring the use of artificial tears were seen in 2 patients treated with partial orbit RT whose target volume included the lacrimal gland compared with 3 patients treated with an anterior field to the entire orbit. Intermediate- and high-grade orbital lymphoma are usually treated with higher radiation doses of around 30–40 Gy, but even with doses of 40 Gy delivered in daily fractions of <2 Gy the incidence of severe dry eye syndrome or retinopathy is low.

Conclusion

Patients with orbital lymphoma should be treated to the entire orbit. An effective dose of RT for low-grade lesions is 25 Gy, which results in minimal morbidity even when delivered to the entire orbit.