Primary central nervous system lymphoma in immunocompetent patients

Primary Cerebral Lymphoma: A Retrospective Study in 22 Immunocompetent Patients

Aims and background

Primary CNS lymphomas are uncommon tumors in immunocompetent patients. We describe the radiological features that should orient towards performing a biopsy and analyze the results in our series of patients.

Methods

We reviewed 22 immunocompetent patients with primary central nervous system lymphoma admitted in our Institute between 1977 and 1997. The follow-up period ranged from 2 months (patient deceased) to 69 months. Fourteen patients underwent surgical removal of the tumor and the remainder a biopsy. All patients received radiotherapy and 8 patients radiotherapy plus chemotherapy.

Results

Two of the 14 patients treated by surgical removal of the tumor died. There was no mortality related to biopsy procedures. Patients treated with radiotherapy had 1-year, 2-year and 5-year survival rates of 66%, 41.6% and 16.6%, compared to 87.5%, 62.5% and 50%, respectively, for patients who received radiotherapy and chemotherapy.

Conclusions

At present, there is no definite treatment for these highly malignant brain tumors. The most favorable results seem related to biopsy followed by radiotherapy plus chemotherapy versus surgical removal, which is related to a high risk of severe postoperative deficit for both the deep location and infiltrating nature of these lesions.

Neuroimaging of primary central nervous system lymphoma in immunocompetent patients: comparison of recent and previous findings

The typical neuroimaging features of primary central nervous system lymphoma (PCNSL) have been described as single or multiple intra-axial, homogenous, contrast-enhancing lesions with marked perilesional edema and restricted diffusion, usually contacting the cerebrospinal fluid surface. Necrosis, peripheral enhancement, hemorrhages, and calcifications are unusual. Recently, some of our patients with PCNSL have had atypical neuroimaging features even before treatment. In this article, we review the neuroimaging characteristics of PCNSL in immunocompetent patients and analyzed how imaging findings over the last 10 years differ from those from more than 10 years ago. Neuroimaging findings suggest that PCNSL is a disease that affects the entire brain. Although some imaging findings are characteristic of PCNSL, the frequency of atypical findings on conventional neuroimaging is increasing. Atypical neuroimaging findings do not rule out PCNSL, even in immunocompetent patients.

A Case of Primary T-Cell Central Nervous System Lymphoma: MR Imaging and MR Spectroscopy Assessment

Primary central nervous system lymphomas (PCNSLs) are mainly B-cells lymphomas. A risk factor for the development of PCNSL is immunodeficiency, which includes congenital disorders, iatrogenic immunosuppression, and HIV. The clinical course is rapidly fatal; these patients usually present signs of increased intracranial pressure, nausea, papilledema, vomiting, and neurological and neuropsychiatric symptoms. PCNSL may have a characteristic appearance on CT and MR imaging. DWI sequences and MR spectroscopy may help to differentiate CNS lymphomas from other brain lesions. In this paper, we report a case of a 23-year-old man with T-primary central nervous system lymphoma presenting with a mass in the right frontotemporal lobe. We describe clinical, CT, and MRI findings. Diagnosis was confirmed by stereotactic biopsy of the lesion.

Primary central nervous system lymphoma

Primary central nervous system lymphomas (PCNSL) are aggressive malignancies that arise in distinct anatomical sites, which display unique structural, biological and immunological conditions. So far, despite recent therapeutic advances, these malignancies exhibit one of the worst prognoses among all non-Hodgkin lymphomas (NHL). For a long time, radiotherapy (RT) has been the standard treatment, producing a response rate of 60-65% and a notable neurological improvement in most cases. However, relapse usually occurred within a few months after RT, with a median survival of 14 months and a 5-year survival of approximately 15-24%. Although the introduction of systemic chemotherapy has consistently improved survival, the prognosis of PCNSL is still dismal, with high rates of local relapse and consequent death. Defining the optimum therapeutic management is difficult because of potential selection biases in large retrospective reviews and the limited number of prospective studies. Although studies published on PCNSL are increasing, several therapeutic questions still remain unanswered after a decade of research.

Lymphomatosis Cerebri Presenting as Rapidly Progressive Dementia

BACKGROUND

As the population ages, the incidence of dementing illness is increasing. Accurate and timely diagnosis provides the best hope for instituting appropriate treatment and educating the patient and family members as to prognosis based upon likely etiology in a given patient.

REVIEW SUMMARY

We present a case of an elderly patient referred to our tertiary-care center for further evaluation of a rapidly progressive dementia, whose definitive diagnosis was delayed by nonspecific MRI findings, presence of 14-3-3 protein in the CSF, and nonspecific cutaneous lesions. At brain biopsy, he was thought to have a diffusely infiltrating lymphoma, with distinctive immunohistochemical features.

CONCLUSION

This case is notable in that it presents a patient with progressive dementia whose diagnosis of primary central nervous system lymphoma (PCNSL) was delayed because of the lymphoma's atypical diffusely infiltrating nature. Awareness of this unique presentation may hasten the time between clinical presentation, diagnosis, and subsequent treatment.

Primary central nervous system lymphoma in Korea: comparison of B- and T-cell lymphomas

The great majority of primary central nervous system lymphoma (PCNSL) is known to be of B-lineage, with T-cell PCNSL (T-PCNSL) accounting for <5%. We report an unusually high incidence of T-cell lymphoma among the PCNSLs originated in a large general-care hospital in the metropolitan Seoul area. PCNSLs (n = 42) accrued from April 1995 through June 2001 were reviewed for histologic and clinical features, and immunohistochemical staining was done for CD3, CD20, CD4, CD8, Bcl-6, and CD10. Clonal rearrangements of the TCR-gamma and IgH genes were studied with semi-nested PCR in all seven cases of T-PCNSL and seven of 35 B-cell PCNSL (B-PCNSL). Formalin-fixed, paraffin-embedded specimens were used in all these studies. By immunohistochemical staining and molecular studies, seven cases (16.7%) were diagnosed as T-PCNSL, each displaying clonal rearrangement of the TCR-gamma gene, and 35 (83.3%) as B-PCNSL. Radiologically, T-PCNSL was significantly correlated with the superficial and subcortical lobar location (p <0.001), solitary mass formation (p = 0.001), presence of rim enhancement (p <0.001), and peritumoral edema (p = 0.029). Involvement of cerebrospinal fluid was observed only in B-PCNSL (n = 17) but not in T-PCNSL (p = 0.010). Histologically, T-PCNSL was characterized by a population of mixed predominantly small- and occasionally medium-sized cells (p <0.001), which were loosely scattered without forming a solid mass (p = 0.024), and perivascular infiltration was frequent (p = 0.007), in contrast to predominantly large cells of B-PCNSL, i.e., diffuse large B-cell lymphoma (DLBCL), in which the cells tended to aggregate to form monomorphous sheets (p = 0.024). In T-PCNSL, staining for CD8 was positive in five, including one with coexpression of CD4, and two were negative for CD4 and CD8. Of 24 DLBCLs tested, the pattern of Bcl-6+ tumor cells was diffusely dense, similar to that of the germinal center in nine cases (37.5%), with coexpression of CD10 in three of the nine cases. T-PCNSL accounted for 16.7% of the PCNSLs; thus, in Korea it may not be as rare as previously known. The T-PCNSL presented with certain clinical and pathologic features that were distinct from B-PCNSL and displayed preponderance of CD8 expression. DLBCL of the germinal center B-cell derivation defined by bcl-6 expression comprised 37.5% of DLBCL of the brain.

Summary statement on primary central nervous system lymphomas from the Eighth International Conference on Malignant Lymphoma, Lugano, Switzerland, June 12 to 15, 2002

Under the sponsorship of the International Extranodal Lymphoma Study Group, a Multidisciplinary Workshop on primary CNS lymphoma (PCNSL) with over 50 participants from Europe, North America, Israel, and Australia was held as part of the Eighth International Conference on Malignant Lymphoma in Lugano, Switzerland (June 12 to 15, 2002). The main purposes of the Workshop were to exchange the latest scientific information, to analyze methodologic issues in the design of clinical trials, to reach a consensus on treatment recommendations and prognostic factors, to discuss clinical and molecular targets for future studies, and to establish an international collaborative group to conduct laboratory and clinical investigations in PCNSL. This article summarizes the contents of the Workshop, analyzes the current knowledge on the most relevant biologic and clinical issues in PCNSL, and focuses on fundamental challenges to be addressed in future studies.

Long-term remission of primary central nervous system lymphoma by intensified methotrexate chemotherapy

High-dose (1-3.5 g/m2) methotrexate (MTX) followed by whole-brain radiation therapy (WBRT) has consistently improved length of survival in primary central nervous system lymphoma (PCNSL), but the prognosis remains dismal. To optimize and enhance the dose intensity of MTX, we applied MTX at 8 g/m2 to 20 patients with PCNSL. In an effort to lower the risk of neurotoxic treatment sequelae, the WBRT dose was reduced to 30 Gy in cases of complete remission after MTX therapy. Further, omission of WBRT and administration of stereotactic radiotherapy (SRT) were undertaken in 3 older patients. The overall response rate to the MTX therapy was 83%. The median progression free survival (PFS) was 54 months with a median overall survival (OS) of 57 months. Achieving a complete response after MTX therapy was significantly associated with a longer PFS. Late neurotoxicity was encountered in 4 (50%) of 8 patients who were aged 60 years or older and received WBRT, but in none of 12 patients who were aged less than 60 years or avoided WBRT. All older patients who underwent SRT sustained complete remission without a dementing disease. Intensifying the MTX dosage to 8 g/m2 appears more promising in comparison to results reported with MTX doses of 1-3.5 g/m2. In younger patients, the establishment of complete remission by intensified MTX therapy and subsequent WBRT with a relatively lower dose could promise durable tumor remission with an acceptable neurotoxicity. In older patients, WBRT should be avoided to sustain a meaningful survival, and SRT may provide a valid strategy in terms of enhancing local disease control without undue risk.

Prognostic scoring system for primary CNS lymphomas: the International Extranodal Lymphoma Study Group experience

PURPOSE

To identify survival predictors and to design a prognostic score useful for distinguishing risk groups in immunocompetent patients with primary CNS lymphomas (PCNSL).

PATIENTS AND METHODS

The prognostic role of patient-, lymphoma-, and treatment-related variables was analyzed in a multicenter series of 378 PCNSL patients treated at 23 cancer centers from five different countries.

RESULTS

Age more than 60 years, performance status (PS) more than 1, elevated lactate dehydrogenase (LDH) serum level, high CSF protein concentration, and involvement of deep regions of the brain (periventricular regions, basal ganglia, brainstem, and/or cerebellum) were significantly and independently associated with a worse survival. These five variables were used to design a prognostic score. Each variable was assigned a value of either 0, if favorable, or 1, if unfavorable. The values were then added together to arrive at a final score, which was tested in 105 assessable patients for which complete data of all five variables were available. The 2-year overall survival (OS) +/- SD was 80% +/- 8%, 48% +/- 7%, and 15% +/- 7% (P =.00001) for patients with zero to one, two to three, and four to five unfavorable features, respectively. The prognostic role of this score was confirmed by limiting analysis to assessable patients treated with high-dose methotrexate-based chemotherapy (2-year OS +/- SD: 85% +/- 8%, 57% +/- 8%, and 24% +/- 11%; P =.0004).

CONCLUSION

Age, PS, LDH serum level, CSF protein concentration, and involvement of deep structures of the brain were independent predictors of survival. A prognostic score including these five parameters seems advisable in distinguishing different risk groups in PCNSL patients. The proposed score and its relevance in therapeutic decision deserve to be validated in further studies.

Relevance of intraocular involvement in the management of primary central nervous system lymphomas

BACKGROUND

Reported data regarding intraocular lymphoma (IOL) management are anecdotal. Cases of IOL included in an international multicentre series of 378 immunocompetent patients with primary central nervous system lymphomas (PCNSLs) were reviewed.

PATIENTS AND METHODS

Staging included slit-lamp examination in 170 patients: IOL was diagnosed in 22 cases (13%). A concomitant brain lesion was detected in 21 cases. Planned treatment was chemotherapy followed by radiotherapy in 13 cases, chemotherapy alone in three and radiotherapy, followed by or not by chemotherapy in five; one patient was not treated. Chemotherapy included high-dose methotrexate in 12 cases. Ten patients received intrathecal chemotherapy. Radiotherapy consisted of whole brain irradiation, followed by or not by a tumour bed boost; ocular irradiation was planned in 15 cases. Irradiation in one patient without brain lesions was limited to the orbits only (50 Gy).

RESULTS

IOL was positively correlated to systemic symptoms and meningeal disease. Fifteen patients (71%) achieved an objective response; 16 patients experienced a failure (2-year failure-free survival 34+/-10%). Failures involved the eyes in eight cases, with a 2-year time to ocular relapse of 59+/-11%. Ocular failure was less common in patients treated with chemotherapy plus ocular irradiation and was associated with a significantly shorter survival. Seven patients are alive [median follow-up 53 months, 2-year overall survival (OS): 39+/-11%] , five of whom were treated with ocular irradiation. The patient with isolated IOL is alive and disease-free at 14 months. OS of the entire series was similar to that of PCNSL patients with negative slit-lamp examination.

CONCLUSIONS

IOL is usually associated with concomitant brain disease and shows a survival similar to that of the rest of PCNSLs. Chemotherapy combined with ocular irradiation resulted in better control of ocular disease, which seems to be associated with survival. In view of the potential role of ocular irradiation, the use of chemotherapy alone in phase II trials should be critically reconsidered in PCNSL patients with ocular disease.

Therapeutic management of primary central nervous system lymphoma: lessons from prospective trials

Primary central nervous system lymphomas (PCNSL) are aggressive malignancies, exhibiting one of the worst prognoses among lymphomas. The best treatment modality for PCNSL has not yet been identified. Several therapeutic questions still remain unanswered, and some methodological pitfalls in clinical trials prevent definitive conclusions from being drawn. In this review, certain aspects of trial design as well as emerging therapeutic guidelines are analyzed, and future perspectives are discussed. In the vast majority of prospective trials, general criteria for treatment of aggressive lymphomas were adopted, choosing primary chemotherapy (CHT) followed by radiotherapy (RT) as therapeutic modality. This strategy produced a five-year survival of 22%- 40% in comparison to the 3%-26% reported with RT alone. Systemic high-dose methotrexate (HD-MTX) seems to be the most effective drug, producing a response rate of 80%-90% and a two-year survival of 60%-65%. To date, the addition of other drugs at conventional doses have not consistently improved outcome. With a few exceptions, any regimen without HD-MTX comprehensively performed no better than RT alone. In combined treatment. RT doses should be decided on the bases of response to primary CHT and the number of lesions, and, until definitive conclusions from well-designed trials are available, RT parameters should follow the widely accepted principles used for other aggressive lymphomas. CHT as exclusive treatment, keeping RT for relapses or persistent disease, appears to be an attractive strategy. However, the worldwide experience with this modality is still limited, and corroborating data are needed. Intrathecal CHT still has not found a defined role in PCNSL management. Preliminary data seem to indicate that adequate meningeal treatment with HD-MTX, but without intrathecal CHT, could also be suitable in positive-cerebrospinal fluid patients. Future efforts should be addressed to identify new active drugs and more efficient CHT combinations, to evaluate the efficacy of high-dose CHT supported by autologous peripheral blood stem cells transplantation, and to clarify the impact of RT delay in complete responders, the usefulness of intrathecal CHT, and the best management for elderly patients. The assessment of impact of treatment on neuropsychological functions and quality of life is a mandatory endpoint in clinical trials.

Prevention of recurrence and prolonged survival in primary central nervous system lymphoma (PCNSL) patients treated with adjuvant high-dose methylprednisolone

Five patients at risk for primary central nervous system lymphoma (PCNSL) recurrence were treated with high-dose methylprednisolone (HDMP) to prevent 'trafficking' of malignant lymphocytes into the central nervous system (CNS). HDMP was chosen because of its ability to stabilize the 'blood brain barrier (BBB)'. Three men with newly diagnosed PCNSL, ages 62, 76 and 78y, whose survival was projected to be 6.6 months, began treatment after achieving complete response (CR) to initial radiation therapy alone and survived 27, 37 and 59 months after treatment. In none was death from recurrent disease in CNS but one patient did die of systemic non-Hodgkin's lymphoma (NHL) five years after PCNSL diagnosis. A 20 y old man was treated with HDMP after successful combined modality therapy and is alive 75+ months after initial diagnosis without evidence of disease recurrence. A 34 y old man relapsed after combined modality initial treatment and failed to respond to HDMP when treatment was begun after unsuccessful salvage therapy; he died of disease 12 months after initial diagnosis. There were no treatment complications. The promising results in this pilot study from the basis for a North Central Cancer Treatment Group (NCCTG) 96-73-51, a Phase 2 clinical trial of brain radiotherapy and HDMP for PCNSL patients 70y of age and older, a group of patients at high risk for toxicity from intensive combined modality therapy.

Graft-versus-lymphoma effect after allogeneic peripheral blood stem cell transplantation for primary central nervous system lymphoma

Allogeneic peripheral blood stem cell transplantation (allo PBSCT) is a recognized treatment modality for hematological malignancies resistant to conventional chemoradiotherapy. The post-transplant immune-mediated graft-versus-leukemia effect has major curative potential. In this case presentation, the allogeneic approach to resistant recurrent primary central nervous system (CNS) lymphoma using peripheral blood stem cells from an HLA identical sibling after immuno-suppressive non-myeloablative conditioning, was examined clinically. The patient in question had relapsing refractory primary CNS lymphoma and is the first to be treated with this modality. She developed early skin and liver-localized grade II graft-versus-host disease after allo PBSCT, which then responded to short-term treatment. Chimeric studies at the time showed 100% donor cells and repeated magnetic resonance imaging of the brain revealed gradual shrinkage of the tumor. Three months after transplant the cerebral mass was no longer evident and currently, 30 months after transplantation, the patient continues to be disease free. The absence of any signs of malignancy suggests the development of a durable graft-versus-lymphoma effect in this brain tumor and indicates that this effect may be achieved even after non-myeloablative conditioning.

Primary central nervous system lymphoma in Japan: a nationwide survey

PURPOSE

To analyze clinical features, treatment results, and prognostic factors of primary central nervous system lymphoma (PCNSL) in Japan, we conducted a nationwide survey.

METHODS AND MATERIALS

We analyzed 466 patients with histologically proven PCNSL treated between 1985 and 1994 at 62 institutions, including 56 medical schools.

RESULTS

Patient and tumor characteristics of the 466 patients were not greatly different from those reported previously, except for the relatively high proportion of T-cell lymphoma (8.5%). The median survival time of the 466 patients was 18 months, and the 5- and 10-year survival rates were 15.2% and 8.2%, respectively. Complete response was observed in 63% of evaluable patients, but 64% of the complete responders developed recurrence (77% within the irradiated volume). Among patient- or tumor-related factors, higher age, worse performance status, presence of B symptom, multiple lesions, presence of meningeal dissemination, and elevated lactate dehydrogenase (LDH) level were associated with poorer survival, whereas no significant difference was observed in prognosis with respect to sex, T/B phenotype, or histological subclassification. In 410 patients receiving at least 40 Gy to the tumor, there was no difference in survival with respect to total radiation dose or field. Patients receiving 2 or more cycles of systemic chemotherapy had a slightly longer median survival time (22 months) and higher 5-year survival rate (20%) than those receiving radiotherapy alone (18 months and 17%, respectively), but the difference was not significant (p = 0.13). No chemotherapy protocol appeared to be better than any other.

CONCLUSIONS

In addition to age and performance status, which are well-known prognostic factors, B symptom, tumor number, presence of meningeal dissemination, and serum LDH level also seemed to influence survival. Higher radiation dose was not associated with better prognosis. The role of chemotherapy could not be clarified because of the use of various protocols, but it seemed that the influence of various prognostic factors is greater than the effect of chemotherapy.

Systemic chemotherapy with vincristine, cyclophosphamide, doxorubicin and prednisolone following radiotherapy for primary central nervous system lymphoma: a phase II study

We treated 23 patients with primary central nervous system lymphoma with a protocol of conventional radiation up to 55 +/- 5 Gy followed by 4 to 6 cycles of intravenous doxorubicin (30 mg/m2), vincristine (1 mg/m2) and cyclophosphamide (350 mg/m2), and oral prednisolone (8-30 mg/m2) (VEPA chemotherapy) repeated at 2-week intervals. The median age of the 23 patients was 59 years, and the median World Health Organization performance status score was 2. Seventeen patients received 4 or more courses of the chemotherapy, but 6 received only 1 or 2 courses for various reasons. The median survival time for all 23 patients was 25.5 months and their 5-year survival rate was 23%. These values were 34 months and 32%, respectively, for the 17 patients who received 4-6 courses of chemotherapy. After treatment, decline in performance status unaccompanied with tumor recurrence was observed in 44% of the patients; the incidence was apparently higher in older than in younger patients. The survival results obtained with this combined radiochemotherapy regimen appear to be better than those reported in most previous studies of patients treated with radiation alone. Post-irradiation VEPA chemotherapy appears to be worthy of further evaluation.

Primary central nervous system non-Hodgkin's lymphoma (PCNSL): survival advantages with combined initial therapy? A final report of the North Central Cancer Treatment Group (NCCTG) Study 86-72-52

PURPOSE

We herein report updated survival and toxicity data on the entire cohort of 53 eligible patients treated on North Central Cancer Treatment Group (NCCTG) protocol 86-72-52, which is now closed.

METHODS AND MATERIALS

An initial report was published in this journal in 1995. No substantive changes in the conclusions of that report were identified in this analysis. Median survival was 9.6 months for the entire cohort; median survival for the 20 patients who completed the prescribed protocol treatment was 20.7 months. The hematologic and non-hematologic toxicity distributions are virtually the same as those reported in the original paper.

RESULTS

Results are given for the entire group and for subsets defined by age < or = 60 versus > 60 years, and < 70 versus > or = 70 years of age.

CONCLUSIONS

No significant differences were observed in any of the outcome variables by age group. There was, however, a nonsignificant suggestion of poorer outcome in those who were > 60 years of age.

Second-line treatment for primary central nervous system lymphoma

Failure after first-line treatment was reported in 35-60% of immunocompetent patients with primary central nervous system lymphoma (PCNSL). There are currently no reports focusing on salvage therapy. This review analyses prognostic factors and the efficacy of salvage therapy by focusing on data from papers reporting results of first-line treatment in 355 cases. The study group consisted of 173 patients presenting treatment failure. The interval between failure and death (TTD) was compared for age at relapse (< or =60 vs. >60 years), type of failure (relapse vs. progression), time to relapse (< or =12 vs. >12 months) and salvage treatment (yes vs no). Median TTD was similar in younger and older patients (P = 0.09). Relapsed patients had a longer TTD than patients with progressive disease (P = 0.002). Early relapse led to a shorter TTD than late relapse (P = 0.005). Median TTD was 14 months for patients who underwent salvage therapy and 2 months for untreated cases (P<0.00001). A multivariate analysis showed an independent prognostic role for salvage therapy and time to relapse. Age and type of failure had no predictive value. Salvage therapy significantly improves outcome and, possibly, quality of life. As many different treatments were used conclusions cannot be made regarding an optimal treatment schedule.

Primary central nervous system lymphomas in immunocompetent individuals: histology, Epstein-Barr virus genome, Ki-67 proliferation index, p53 and bcl-2 gene expression

Epstein-Barr virus (EBV) has been detected in the large majority of HIV-related primary central nervous system lymphomas (PCNSL) suggesting a pathogenetic role of the virus. Unlike HIV-related PCNSL, conflicting data exist with regard to the presence of EBV in non immunodeficiency-related (sporadic) PCNSL. For this reason, a population based material of 41 sporadic PCNSL was analysed for the presence of EBV genome (EBER, BHLF) using RNA in situ hybridisation (RISH). Furthermore, the expression of the gene products of the bcl-2 oncogene and the p53 tumor suppressor gene and the tumor growth fraction reactive with the monoclonal antibody Ki-67 have been evaluated. All cases but two were EBV genome negative. In the two positive cases less than 5% of tumor cells showed EBER positivity. In contrast, more than 75% of cells morphologically belonging to the tumor-cell population stained positively for EBER in two cases of HIV related PCNSL. Immunostaining for the bcl-2 oncoprotein was positive in 28 (72%) of 39 cases examined. In most cases more than 75% of tumor cells showed cytoplasmic expression. Of 37 cases investigated for p53 expression, 21 (57%) stained positively. However, in the large majority of positive cases less than 10% of the neoplastic cells stained. The percentage of Ki-67 positive cells ranged between 10% and 80% with a mean of 50%. The expression of the p53 and bcl-2 oncoproteins and the growth fraction did not have any prognostic impact. We conclude that the EBV genome is rarely detected in sporadic PCNSL, indicating that a pathogenetic role of EBV is unlikely. Like extracerebral B-cell lymphomas a large fraction of PCNSL expresses the p53 and bcl-2 oncoproteins, a feature, however, which does not seem to have prognostic implications.