Radiation Therapy. The role of RT for patients with primary CNS lymphoma continues to evolve. Early studies demonstrated that these tumors were radiosensitive and that complete and partial responses could be obtained using doses ranging from 3000 to 5000 cGy. However, the responses were brief, and patients often developed recurrent disease within a matter of months. These findings prompted the RTOG to design a dose-intensification study in which 41 patients with non-AIDS-related primary CNS lymphoma received 4000 cGy of whole-brain irradiation plus a 2000-cGy boost to involved regions. The median survival for the cohort was only 12.2 months, and tumors recurred frequently in the boosted field. Similar limitations of efficacy for high-dose RT have been noted by DeAngelis as well as by other investigators. Therefore, the currently recommended dose of RT for cerebral primary CNS lymphoma is between 4000 and 5000 cGy (whole brain), without a boost. However, RT should be avoided in patients older than 60 years. For patients with ocular lymphoma, irradiation is the treatment of choice; 3600 cGy should be administered to both eyes. However, high-dose methotrexate may also be effective.

Chemotherapy. In general, methotrexate-based chemotherapy regimens are more effective against primary CNS lymphoma than non-methotrexate regimens. Most of these non-methotrexate protocols are based on the CHOP model and feature cyclophosphamide in combination with other drugs, usually doxorubicin, vincristine, and prednisone. Compared with methotrexate-based regimens, the progression-free and overall survival rates are lower in these non-methotrexate regimens, and they are often associated with an increased incidence of neurologic toxicity. The major reason cited for the decreased efficacy of CHOP and similar protocols is the poor penetration of the intact blood-brain barrier by cyclophosphamide and doxorub

Timing of Radiotherapy and Chemotherapy The major controversy regarding RT for primary CNS lymphoma first-line therapy, in combination with chemotherapy, or should it be withheld in selected patients and not used until the time of recurrence? Although this issue has not been resolved, most authors recommend irradiation after some form of initial chemotherapy.  A multicenter RTOG study demonstrated improved survival with the combination of chemotherapy, using high-dose methotrexate plus RT when compared with previous reports of RT alone; delayed neurotoxicity remains a risk of this approach. Neuwelt and colleagues suggest withholding RT until recurrence or progression in all patients to decrease the risk of radiation-related neuropsychological sequelae.  Others have similar recommendations but only for primary CNS lymphoma patients older than 50 years. DeAngelis and associates found that it was very uncommon for patients younger than 50 years to develop radiationinduced neurotoxicity. The addition of chemotherapy has significantly improved disease-free and overall survival in patients with primary CNS lymphoma. With RT alone, median survival is approximately 12 months. When some form of chemotherapy has been added to the treatment regimen, median survival is extended; it ranges from 30 to 41 months. In many of these investigations, chemotherapy was administered before RT and often resulted in complete or partial responses.

Methotrexate appears to be the most effective drug and can be administered via the intravenous or intra-arterial route. Gabbai and associates  reported a series of 13 patients given high-dose intravenous methotrexate before radiation. They noted complete responses in 9 patients and partial responses in 4 patients, with an overall median survival of greater than 9 months. DeAngelis and colleagues also used preradiation methotrexate (intravenous and intrathecal) plus cytarabine in 31 patients with primary CNS lymphoma.  The overall median survival for the cohort was 42.5 months, with partial responses in 17 patients and stable disease in 5 patients. Neuwelt  administered methotrexate via the intra-arterial route, in combination with osmotic blood-brain barrier disruption, cyclophosphamide, and procarbazine, to a series of 16 patients. In all of these patients, irradiation was withheld until the time of disease progression (and was eventually administered to 9 of 16 patients). Chemotherapy induced complete responses in 13 patients and partial responses in 3 patients, with an over-all median survival of 44.5 months. Neuropsychological follow-up of the responding patients who did not undergo irradiation demonstrated stable cognitive function.

Preradiation versus Postradiation Chemotherapy.  Once the diagnosis of primary CNS lymphoma has been established and the extent of disease determined, treatment should be initiated as soon as possible. Preradiation chemotherapy, as opposed to postradiation chemotherapy, has been emphasized for several theoretical reasons. At least for agents such as methotrexate and cisplatin, some data (albeit in the pediatric literature) indicate that pre-radiation chemotherapy is less neurotoxic than postradiation chemotherapy. Additionally, drug delivery to a primary CNS lymphoma may be increased before radiation, when the blood-brain barrier is maximally disrupted by the tumor, than after RT, which results in tumor regression as well as partial repair and closure of the blood-brain barrier behind the regressing tumor. Finally, preradiation chemotherapy allows one to assess the efficacy of chemotherapy without the confounding variable of irradiation.

For patients with extremely poor KPS (< 40) or creatinine clearance less than 50, it is recommended that treatment consist of whole-brain irradiation (45 Gy) in order to rapidly induce a response, diminish neurologic morbidity, and optimize quality of life. Chemotherapy is also an option. If the lumbar puncture or spinal MRI is positive, consider intrathecal chemotherapy plus focal spinal RT for the few patients ithis group who have an excellent response to whole-brain irradiation and who achieve a reasonable quality of life with improved PS; systemic chemotherapy could be considered when their disease recurs. For patients in this group who do not achieve a significant benefit from RT and whose disease progresses, palliative care is suggested.