Dr. DeAngelis has written a succinct and accurate assessment of management of primary central nervous system lymphoma. Non-AIDS-related primary central nervous system lymphoma is a rare, highly malignant primary brain tumor. However, its incidence is increasing, as are AIDS-related tumors of this type. Patients with primary central nervous system lymphoma usually have widespread infiltration throughout the brain, rendering aggressive resection of no benefit. There is a significant incidence of leptomeningeal and ocular involvement (the latter a "sanctuary site," thus posing a particular therapeutic dilemma).
Aggressive whole head radiotherapy in primary central nervous system lymphoma results in gratifying initial responses, but unfortunately these are universally transient . This contrasts with the long-term efficacy of radiotherapy in localized systemic lymphoma. The addition of postradiation chemotherapy in primary central nervous system lymphoma, particularly methotrexate(Drug information on methotrexate), but also other standard lymphoma regimens, also results in gratifying but transient efficacy, and significant cognitive loss .
Current Clinical Trials in Non-AIDS-Related Central Nervous System Lymphoma
Dr. DeAngelis emphasizes that the majority of current clinical trials recognize that the blood-brain barrier is an impediment to drug delivery. Even though primary central nervous system lymphoma is a highly infiltrative malignancy, dense tumor cell aggregates often perturb blood-brain barrier integrity, as evidenced by the characteristic homogeneous, periventricular enhancement seen on imaging studies . Therefore, DeAngelis et al treat with high-dose systemic and intraventricular chemotherapy initially, while the blood-brain barrier is still leaky, then follow with whole head radiotherapy, and often further chemotherapy. Changing the sequence of standard-dose systemic chemotherapy and whole head radiotherapy also can extend survival, presumably by improving drug delivery .
DeAngelis et al at Memorial Sloan-Kettering have reported the largest series to date of central nervous system lymphoma patients undergoing conventional chemotherapy followed by radiation. Similar to the recent report by Miller recently reported on a series of 104 patients , almost all patients at Sloan-Kettering had B-cell lymphoma, consistent with previous reports of primary central nervous system lymphoma pathologic classification.
Patients initially received systemic and intraventricular methotrexate, then whole brain radiation, and subsequently high-dose cytosine arabinoside. The survival of 31 patients treated in this manner had a predicted median time to recurrence of 41 months, which was statistically better than the survival of patients treated at Sloan-Kettering during the same period with radiotherapy alone. No plateau in the survival curve was present. Age, functional status, and prior radiation were significant prognostic factors. Cognitive impairment was a large problem, although not quantitated with formal testing.
Osmotic Blood-Brain Barrier Disruption to Enhance Drug Delivery
Over the past 13 years at the Oregon Health Sciences University, we have employed enhanced chemotherapy delivery to obviate the need for, or at least delay, whole head radiotherapy [6-8]. Osmotic blood-brain barrier disruption increases drug delivery not only to tumor, but also to brain around tumor, brain distant to tumor, and to the cerebral spinal fluid (CSF). In dense tumor where the blood-brain barrier is already leaky, drug delivery with blood-brain barrier disruption may increase only two- to threefold; however, in brain around tumor and brain distant to tumor, which often contain infiltrating tumor cells, delivery is increased as much as fifty- to one hundred fold.
This magnitude of enhanced delivery is far greater than can be attained with high-dose chemotherapy, even with autologous bone marrow rescue. Furthermore, osmotic blood-brain barrier dis- ruption can be used repeatedly. Since barrier integrity is restored in dense tumor after the initial courses of chemotherapy, repeated drug cycling may be crucial.
In our current series of 58 non-AIDS-related lymphoma patients, prognostic factors and toxicity have been carefully assessed (manuscript submitted). In contrast to other series that have included patients with a presumptive diagnosis, all patients in our series had a tissue diagnosis from brain, CSF, or vitreous. Systemic involvement, albeit rare, was present in five additional patients.
The Oregon Health Sciences University series was similar to DeAngelis' in several factors: incidence of leptomeningeal disease, ocular involvement, and predominance of large B-cell lymphoma. Similar to DeAngelis' series, age, functional status, and prior radiation were statistically significant prognostic factors. The distribution of these factors is also comparable to those previously described in other reported series .
Chemotherapy given with osmotic blood-brain barrier disruption is more intensive than other preirradiation chemotherapy regimens. However, in our series of barrier disruptions in non-AIDS-related primary central nervous system lymphoma, there were no procedure-related deaths. The neurologic (primarily focal seizures) and non-neurologic (primarily granulocytopenic fever) complications usually have been reversible and without sequelae, in contrast to the delayed but irreversible cognitive sequelae associated with whole head radiotherapy.
The Key Issue
The key issue, then, is a comparison of the duration and quality of survival in patients undergoing osmotic blood-brain barrier disruption with regimens that include radiotherapy. Both preradiation chemotherapy, as outlined by DeAngelis, and enhanced chemotherapy delivery with blood-brain barrier disruption, improve survival when compared to radiotherapy with or without subsequent chemotherapy.
The Oregon Health Sciences University series has follow-up to 12 years and, in those 39 patients treated with initial chemotherapy (19 alive, including 18 tumor free), a plateau in survival is apparent. However, no statistical survival difference is yet present when these patients are compared with the DeAngelis' series. The Oregon Health Sciences University series is unique in the absence of cognitive loss, even in patients more than 60 years old (a group particularly susceptible to cognitive impairment with preradiation chemotherapy, as reported in other series).
In DeAngelis' series, 27% of complete responders who are more than 50 years old have clinical evidence of dementia, even without formal cognitive testing. As a result, direct comparison of preradiation chemotherapy (with blood-brain barrier disruption- enhanced drug delivery) with radiation is problematic: after radiation, patients are rarely rescued with blood-brain barrier disruption, and they are predisposed to dementia. We have initiated a multi-institutional clinical trial to replicate the Oregon Health Sciences University results achieved with an osmotic blood-brain barrier disruption chemotherapy regimen in primary central nervous system lymphoma.
As Dr. DeAngelis correctly concludes, primary central nervous system lymphoma is an infiltrative, malignant brain neoplasm that cannot be controlled with radiotherapy alone, and in which drug delivery and cognitive outcome are key issues. Complete and durable responses for up to 12 years without cognitive loss and without radiotherapy have now been attained by enhancing drug delivery with osmotic blood-brain barrier disruption.