Primary central nervous system (CNS) lymphoma is defined as lymphoma limited to the cranial-spinal axis, including the brain, spinal cord, cerebrospinal fluid, and vitreoretinal space, without systemic disease. This disease has increasingly been seen among patients with HIV and other immunocompromised people. Immunosuppression-related primary CNS lymphomas are almost always associated with Epstein-Barr virus. Unlike patients with other primary CNS lymphomas, patients with immunosuppression-related disease almost never have an activated B-cell cell-of-origin phenotype. Computed tomography (CT) scans may show ring enhancement in 50% of patients with AIDS, while patients without AIDS almost always show only homogeneous enhancement. Median overall survival in published trials generally ranges from 2 to 5 years. Older age (>65 years) and HIV positivity are the most clinically relevant poor prognostic factors, but the prognosis for HIV-associated primary CNS lymphoma has improved with the use of combination antiretroviral therapy. Such patients are then treated according to a protocol for patients who do not have HIV or immunosuppression.

Prognostic Factors

Poor prognostic factors for primary CNS lymphoma include the following:

• Age older than 60 years.
• Elevated serum level of lactate dehydrogenase.
• Elevated cerebrospinal fluid protein concentration.
• Involvement of nonhemispheric areas of the brain (periventricular, basal ganglia, brainstem, and cerebellum).
• Intraocular disease and concomitant brain involvement.

Diagnostics

When tumor progression occurs, it is usually confined to the CNS and/or the eye. Occult systemic disease can be excluded by positron emission tomography-CT scans of the chest, abdomen, and pelvis, and sometimes with bone marrow biopsy.

In one prospective case series of 282 patients, 17% were found to have meningeal dissemination by cytomorphology, polymerase chain reaction of rearranged immunoglobulin heavy-chain genes, or meningeal enhancement on magnetic resonance imaging.

Pathogenesis

Although more than 95% of patients with primary CNS lymphoma have lymphoma of B-cell origin, 45 patients with CNS lymphoma of T-cell origin showed no difference in presentation or outcome in a retrospective series with data collected from 12 cancer centers. Almost all primary CNS lymphomas are diffuse large B-cell lymphomas of the activated B-cell nongerminal center subtype.

Primary CNS lymphoma closely resembles the activated B-cell subtype of large B-cell lymphoma, with additional mutations in the B-cell receptor signaling pathway. A retrospective case series of 40 patients with low-grade primary CNS lymphoma derived from 18 cancer centers in five countries reported a better long-term outcome (median survival, 7 years) than is associated with the usually aggressive CNS lymphoma. Anecdotal cases of primary CNS Hodgkin lymphoma have also been reported.

Radiation Therapy

Because of the diffuse nature of central nervous system (CNS) lymphomas, aggressive surgical decompression with partial or gross total removal of the tumor is of no benefit to the patient. Median survival with surgery alone ranges from only 1 to 5 months. Until the mid-1990s, radiation therapy was the standard treatment, with doses of up to 45 Gy using standard fractionation. A prospective trial by the Radiation Therapy Oncology Group (RTOG-8315) used 40 Gy whole-brain radiation therapy (WBRT) and a 20-Gy boost to the tumor. The results were no better than those previously reported, with a median survival of 1 year and 28% of the patients surviving 2 years. Disease recurs in the brain in 92% of patients despite high doses of radiation. The addition of spinal-axis radiation does not affect survival because it does not prevent cerebral relapse.

Combined Chemotherapy and Radiation Therapy

Two multicenter prospective trials (including RTOG-8806) used preirradiation cyclophosphamide, doxorubicin, vincristine, and dexamethasone followed by WBRT. Median survival times were no better than for radiation therapy alone. The failure of these and other combined-modality trials has been attributed to poor penetration of standard drugs through the blood-brain barrier and to increased neurological toxic effects. Retrospective reviews suggested improved results with the use of high-dose methotrexate or cytarabine with radiation therapy rather than with other combination regimens. While combinations of high-dose methotrexate with WBRT improved progression-free survival (PFS) and overall survival (OS) anecdotally in patients participating in phase II trials, there was unacceptable neurological toxicity.

Chemotherapy Alone

Trials using chemotherapy alone were justified because of the unsatisfactory results of using WBRT alone, and the neurological toxic effects seen using chemotherapy combined with WBRT. Numerous phase I and phase II studies over two decades established the following active drugs for induction therapy or for treatment of relapsing disease. The following drugs have been used as single agents and in combinations:

• High-dose methotrexate.
• High-dose cytarabine.
• Rituximab.
• Thiotepa.
• Ibrutinib.
• Lenalidomide.
• Lenalidomide + rituximab.
• Pomalidomide.
• Nivolumab.

Severe delayed neurological toxic effects were rarely seen in chemotherapy-only trials in the absence of subsequent radiation therapy. However, salvage radiation can be given for relapsed or refractory disease, sometimes at reduced dosage.

The International Extranodal Lymphoma Study Group investigated three different induction combinations in 227 patients with newly-diagnosed HIV-negative primary CNS lymphoma who were randomly assigned to one of three groups:

• High-dose methotrexate + high-dose cytarabine.
• High-dose methotrexate + high-dose cytarabine + rituximab.
• High-dose methotrexate + high-dose cytarabine + rituximab + thiotepa (the MATRix regimen).

With a median follow-up of 30 months, patients who received the four-drug combination had a complete remission rate of 49% (95% confidence interval , 38%‒60%) compared with 23% (interquartile range , 14%‒31%) for patients who received the two-drug combination (hazard ratio , 0.46; 95% CI, 0.28‒0.74) and 30% (IQR, 21%‒42%) for patients who received the three-drug combination (HR, 0.61; 95% CI, 0.40‒0.94).

In a randomized, nonblinded, multicenter trial, 79 patients were randomly assigned to receive either high-dose methotrexate or high-dose methotrexate plus cytarabine. While the 3-year PFS was better for patients who received the two-drug regimen (HR, 0.54; 95% CI, 0.31–0.92; P = .01), there was no difference in the 3-year OS rate (46% for the two-drug regimen vs. 32% for the one-drug regimen; HR, 0.65; 95% CI, 0.38–1.13; P = .07).

In a randomized, prospective, multicenter trial, 200 patients were randomly assigned to receive intravenous high-dose methotrexate, carmustine, teniposide, and oral prednisone with or without rituximab. With a median follow-up of 32.9 months, there was no difference in the 1-year event-free survival rate: 52% (95% CI, 42%−61%) with rituximab and 49% (95% CI, 39%−58%) without rituximab (HR, 1.00; 95% CI, 0.70−1.43; P = .99).

The DA-TEDDI-R regimen incorporates temozolomide, etoposide, liposomal doxorubicin, dexamethasone, ibrutinib, and rituximab. Among 18 patients who received this regimen (five previously untreated), the complete remission rate was 86%, but high rates (39%) of invasive aspergillosis were reported. Further studies of this regimen are under way (NCT03964090 and NCT02203526). This approach requires access to intravenous antifungal agents not available outside of a clinical trial.

In a phase II study of patients with relapsed or refractory primary CNS lymphoma, treatment with rituximab plus lenalidomide resulted in a 36% overall response rate.

Additional randomized trials are needed to establish the optimal chemotherapy combination for induction therapy. The optimal length of induction therapy, the use of maintenance therapy, and the use of consolidation therapy are all areas of controversy that await further trial results.

Consolidation After Induction Chemotherapy

Several phase II studies have investigated consolidation with intensive chemotherapy supported by autologous stem cell transplantation (SCT). This approach is most applicable for younger patients with few comorbidities and good performance status, who also respond well to induction therapy.

Several prospective randomized trials are comparing or have compared the value of WBRT and the value of autologous SCT as consolidation after high-dose methotrexate induction therapy: International Extranodal Lymphoma Study Group 32 (IELSC32 ), Pragmatic–Explanatory Continuum Indicator Summary (PRECIS ) (a phase II randomized trial of 97 patients), Cancer and Lymphoma Group B/Alliance (CALGB 51101 ), and International Extranodal Lymphoma Study Group 43 (IELSG43 ).

In a prospective, randomized trial of 551 immunocompetent patients with newly diagnosed primary CNS lymphoma, all patients received induction chemotherapy with six cycles of high-dose methotrexate (4 g/m2) with or without ifosfamide. After the completion of chemotherapy, responders were randomly assigned to receive either WBRT (45 Gy) or no treatment for complete-response patients and cytarabine for partial-response patients. There was no statistical difference in median OS, with 32.4 months for patients who received radiation therapy versus 37.1 months for those who did not receive radiation therapy (HR, 1.06; 95% CI, 0.80–1.40; P = .71). Treatment-related neurotoxic effects were significantly worse on the radiation therapy arm. Such toxicity must be weighed against the possibility that the survival from chemotherapy alone may be marginally inferior to the survival when radiation is added.

In a prospective randomized trial, 410 immunocompetent patients with newly diagnosed primary CNS lymphoma were scheduled to receive high-dose methotrexate. Patients were randomly assigned to receive either WBRT or no radiation therapy. In the intent-to-treat population, WBRT was associated with longer PFS, at 15.4 months versus 9.9 months (HR, 0.79; 95% CI, 0.64–0.98; P = .034), but no difference in OS, at 32.4 months versus 36.1 months (HR, 0.98; 95% CI, 0.79–1.26; P = .98). However, the study lacked the power to exclude a benefit or harm from the WBRT. In this study, 19 patients were diagnosed with intraocular involvement at diagnosis; intraocular lymphoma was an independent negative prognostic indicator.

The prognosis for recurrent primary central nervous system (CNS) lymphoma is poor, with a median survival of about 4 to 6 months. The prognosis is worse for patients older than 60 years, who account for more than 50% of cases.

Patients with recurrence after high-dose chemotherapy with methotrexate or cytarabine may try autologous or allogenic stem cell consolidation after reinduction of remission with single-agent or combination therapy from the following options:

1. Rituximab.
2. Ibrutinib.
3. Temozolomide.
4. Liposomal doxorubicin.
5. Etoposide.
6. DA-TEDDI-R (temozolomide, etoposide, liposomal doxorubicin, dexamethasone, ibrutinib, and rituximab).
7. Chimeric antigen receptor (CAR) T cells.

Dexamethasone should be avoided with ibrutinib single agent or combination therapy due to the risk of serious fungal infections. Patients deemed ineligible for transplantation can be palliated with these agents.

The DA-TEDDI-R regimen incorporates temozolomide, etoposide, liposomal doxorubicin, dexamethasone, ibrutinib, and rituximab. Among 18 patients who received this regimen (five previously untreated), the complete remission rate was 86%, but high rates (39%) of invasive aspergillosis were reported. Further studies of this regimen are under way (NCT03964090 and NCT02203526). This approach requires access to intravenous antifungal agents not available outside of a clinical trial.

A phase I/II clinical trial investigated CD19-directed CAR T-cell therapy using tisagenlecleucel in patients with relapsed primary CNS lymphoma. One-half of the patients (6 of 12) had a complete response. Three patients maintained a complete response at 9 months, 12 months, and 23 months at the time of data cutoff. Five patients had low-grade immune cell–assisted neurotoxicity and one patient had grade 3 neurotoxicity. CAR-T cell therapy provides an option in relapsed primary CNS lymphoma.

Retrospective reviews of selected patients with primary intraocular lymphoma and no evidence of disseminated central nervous system (CNS) disease showed that localized therapy with intraocular methotrexate or ocular radiation therapy or systemic therapy with rituximab were effective in clearing lymphoma cells from the eye. However, most patients had CNS relapse. Anecdotal series reported lower relapse rates when high-dose methotrexate was added, but prospective multicenter trials with even retrospective controls do not exist. Relapsing disease in the rest of the CNS is treated with the same options listed for primary CNS lymphoma in the brain. In a phase III randomized study of whole-brain radiation therapy, patients with intraocular disease and concomitant brain involvement had a worse prognosis than those with brain involvement alone (19 patients with both, 391 patients with brain involvement only).