Due to the rarity of the disease, prospective studies regarding treatment of AR-PCNSL are very scarce and literature consists mainly of retrospective studies and case series. In immunocompetent hosts, treatment of PCNSL has evolved from radiotherapy alone to high-dose methotrexate (HD-MTX)-based chemotherapy with or without radiotherapy(15). Indeed, WBRT has been associated with irreversible cognitive dysfunction(16, 17). Patients commonly present at diagnosis with a poor functional status, which impairs chemotherapy planning, as was the case for our patient.
Most studies regarding chemotherapy in AR-PCNSL involved the pre-HAART era and were retrospectively analyzed. Several studies showed that HD-MTX alone or combined with other agents was active and relatively well tolerated with a response rate between 30 to 57% but a low overall survival of about 3 months(18, 19).
More recently, Gupta et al described in the HAART era a multi-center retrospective cohort of 20 patients with AR-PCNSL treated with HD-MTX (alone or in combination with other agents) and HAART without WBRT, with an overall survival and progression free survival that exceeded 60 months(20). This result compared favorably with the pre-HAART era. Chemotherapy was relatively well tolerated except for 2 deaths secondary to sepsis during induction in patients with poor performance status. Other immunochemotherapeutic approaches that combined HAART, HD-MTX and rituximab also lead to long-term remission in more than 70% of patients and this without WBRT(21).
WBRT alone has been for a long time the gold standard treatment of AR-PCNSL. Several studies in the pre-HAART era showed that WBRT was effective, especially with doses ≥ 30 Grays, but responses were generally short-lived. Indeed, median survival after radiotherapy was reported between 2 and 5 months. (10, 22, 23) In the HAART era, several studies showed improved overall survival in AR-PCNSL for patient treated with WBRT and HAART(12, 13, 24). A recent retrospective cohort study of 23 patients with AR-PCNSL between 2002 and 2008 treated with HAART and WBRT showed a 3-year overall survival of 64% and that WBRT had an independent positive impact on survival. Performance status was also a major independent factor of survival as patients with good performance status presented a 3-year overall surviving rate of 100% against only 38% in patients with poor performance status. Importantly, 21% of patients that survived more than 12 months after radiation developed side effects such as leukoencephalopathy (grade ≥2)(24). This is a real concern, as patients with AR-PCNSL might be particularly susceptible for radiation-induced brain injury(13) especially when neurocognitive disorder is already a well-known CNS complication of HIV disease(25).
Finally, HAART is now an important part of treatment of AR-PCSNL. The majority of AR-PCSNL are EBV related and it is accompanied by impaired specific T-cell responses against EBV antigens(26). Treatment of HIV infection leading to immune restoration allows a better immunoregulation of EBV infection and thus, potentially, reversal of immune impairment may contribute to the treatment of brain lymphoma. Similarly, HAART with good CNS penetration may better protect the brain from HIV-related injury and cognitive impairment by reducing cerebro-spinal fluid (CSF) viral load.(25) This supports the hypothesis of a potential effect of HAART on EBV-infected cells in AR-PCNSL.
McGowan and Shah published the first case report that showed CR under HAART alone(27). This report was followed with several other reports showing the same results(3, 28, 29). Recently, Alvarez-Pinzon et al also showed CR with combined HAART and Gamma knife radiosurgery(30). Several multicenter retrospective series also showed improvement in survival in patients treated with HAART. Hoffman et al studied 29 HIV-infected patients with histologically confirmed PCNSL. In this cohort, 12 of the 29 patients were treated with WBRT alone, 6 with HAART alone and 11 did not receive any treatment. Survival of patients treated with HAART differed significantly from those receiving no therapy (1093 and 33 days, respectively)(13). These findings were confirmed in two other retrospectives series(12, 31).