Luspatercept for the treatment of anaemia in patients with lower-risk myelodysplastic syndromes (PACE-MDS): a multicentre, open-label phase 2 dose-finding study with long-term extension study.

BACKGROUND: Myelodysplastic syndromes are characterised by ineffective erythropoiesis. Luspatercept (ACE-536) is a novel fusion protein that blocks transforming growth factor beta (TGF ß) superfamily inhibitors of erythropoiesis, giving rise to a promising new investigative therapy. We aimed to assess the safety and efficacy of luspatercept in patients with anaemia due to lower-risk myelodysplastic syndromes. METHODS: In this phase 2, multicentre, open-label, dose-finding study (PACE-MDS), with long-term extension, eligible patients were aged 18 years or older, had International Prognostic Scoring System-defined low or intermediate 1 risk myelodysplastic syndromes or non-proliferative chronic myelomonocytic leukaemia (white blood cell count <13 000/µL), and had anaemia with or without red blood cell transfusion support. Enrolled patients were classified as having low transfusion burden, defined as requiring less than 4 red blood cell units in the 8 weeks before treatment (and baseline haemoglobin <10 g/dL), or high transfusion burden, defined as requiring 4 or more red blood cell units in the 8 weeks before treatment. Patients received luspatercept subcutaneously once every 21 days at dose concentrations ranging from 0·125 mg/kg to 1·75 mg/kg bodyweight for five doses (over a maximum of 12 weeks). Patients in the expansion cohort were treated with 1·0 mg/kg luspatercept; dose titration up to 1·75 mg/kg was allowed, and patients could be treated with luspatercept for a maximum of 5 years. Patients in the base study were assessed for response and safety after 12 weeks in order to be considered for enrolment into the extension study. The primary endpoint was the proportion of patients achieving modified International Working Group-defined haematological improvement-erythroid (HI-E), defined as a haemoglobin concentration increase of 1·5 g/dL or higher from baseline for 14 days or longer in low transfusion burden patients, and a reduction in red blood cell transfusion of 4 or more red blood cell units or a 50% or higher reduction in red blood cell units over 8 weeks versus pre-treatment transfusion burden in high transfusion burden patients. Patient data were subcategorised by: luspatercept dose concentrations (0·125-0·5 mg/kg vs 0·75-1·75 mg/kg); pre-study transfusion burden (high transfusion burden vs low transfusion burden, defined as ≥4 vs <4 red blood cell units per 8 weeks); pre-study serum erythropoietin concentration (<200 IU/L, 200-500 IU/L, and >500 IU/L); presence of 15% or more ring sideroblasts; and presence of SF3B1 mutations. Efficacy analyses were carried out on the efficacy evaluable and intention-to-treat populations. This trial is currently ongoing. This study is registered with ClinicalTrials.gov, numbers NCT01749514 and NCT02268383. FINDINGS: Between Jan 21, 2013, and Feb 12, 2015, 58 patients with myelodysplastic syndromes were enrolled in the 12 week base study at nine treatment centres in Germany; 27 patients were enrolled in the dose-escalation cohorts (0·125-1·75 mg/kg) and 31 patients in the expansion cohort (1·0-1·75 mg/kg). 32 (63% [95% CI 48-76]) of 51 patients receiving higher dose luspatercept concentrations (0·75-1·75 mg/kg) achieved HI-E versus two (22% [95% CI 3-60]) of nine receiving lower dose concentrations (0·125-0·5 mg/kg). Three treatment-related grade 3 adverse events occurred in one patient each: myalgia (one [2%]), increased blast cell count (one [2%]), and general physical health deterioration (one [2%]). Two of these treatment-related grade 3 adverse events were reversible serious grade 3 adverse events: one patient (2%) had myalgia and one patient (2%) had general physical health deterioration. INTERPRETATION: Luspatercept was well tolerated and effective for the treatment of anaemia in lower-risk myelodysplastic syndromes and so could therefore provide a novel therapeutic approach for the treatment of anaemia associated with lower-risk myelodysplastic syndromes; further studies are ongoing. FUNDING: Acceleron Pharma.

Mammalian-target of rapamycin inhibition with temsirolimus in myelodysplastic syndromes (MDS) patients is associated with considerable toxicity: results of the temsirolimus pilot trial by the German MDS Study Group (D-MDS).

The mammalian-target of rapamycin (also termed mechanistic target of rapamycin, mTOR) pathway integrates various pro-proliferative and anti-apoptotic stimuli and is involved in regulatory T-cell (TREG) development. As these processes contribute to the pathogenesis of myelodysplastic syndromes (MDS), we hypothesized that mTOR modulation with temsirolimus (TEM) might show activity in MDS. This prospective multicentre trial enrolled lower and higher risk MDS patients, provided that they were transfusion-dependent/neutropenic or relapsed/refractory to 5-azacitidine, respectively. All patients received TEM at a weekly dose of 25 mg. Of the 9 lower- and 11 higher-risk patients included, only 4 (20%) reached the response assessment after 4 months of treatment and showed stable disease without haematological improvement. The remaining patients discontinued TEM prematurely due to adverse events. Median overall survival (OS) was not reached in the lower-risk group and 296 days in the higher-risk group. We observed a significant decline of bone marrow (BM) vascularisation (P = 0·006) but were unable to demonstrate a significant impact of TEM on the balance between TREG and pro-inflammatory T-helper-cell subsets within the peripheral blood or BM. We conclude that mTOR-modulation with TEM at a dose of 25 mg per week is accompanied by considerable toxicity and has no beneficial effects in elderly MDS patients.

Altered treatment of chronic lymphocytic leukemia in Germany during the last decade.

Chronic lymphocytic leukemia (CLL) is the most common subtype of adult leukemia in the western world. We here report a nationwide survey monitoring the treatment decisions concerning CLL patients in Germany in 2011 and compare treatment trends to sequential surveys performed previously during the last decade. The rate of patients diagnosed in early stages (Binet A/B) notably increased (2006: 66 %, 2009: 71 %, 2011: 77 %) over the years. From 2006 to 2009, the most frequent applied regime switched from chlorambucil to fludarabine containing regimes (2006 chlorambucil: 32 %, 2009: 14 %, fludarabine 2006: 23 %, 2009: 37 %). In 2011, the combination of rituximab with bendamustine (31 %) was most frequent used followed by the rituximab-fludarabine-cyclophosphamide (22 %) regime. Further, immune-chemotherapies were administered significantly more often over the observation period (2006: 15 %, 2011: 73 %). Taken together, this data reflects the change of treatment strategies over the last decade in clinical reality.

Brain and whole-body FDG-PET in diagnosis, treatment monitoring and long-term follow-up of primary CNS lymphoma.

BACKGROUND: Positron emission tomography (PET) with F-18-labeled fluorodeoxyglucose (FDG) provides remarkable accuracy in detection, treatment monitoring and follow-up of systemic malignant lymphoma. Its value in the management of patients with primary central nervous system lymphoma (PCNSL) is less clear. PATIENTS AND METHODS: In a prospective trial, 42 FDG-PET examinations were performed in ten immunocompetent patients with newly diagnosed or recurrent PCNSL before and repeatedly during and after the treatment. Brain and whole body FDG-PET were compared to brain MRI and extra-cerebral CT, respectively. RESULTS: Before the treatment, 6 of 10 patients had congruent findings on FDG-PET and MRI of the brain. Three patients had lesions on brain MRI, not detected by FDG-PET. One patient had additional FDG-PET positive lesions inconspicuous in MRI. The follow-up suggested FDG-PET to be false positive in these lesions. After the treatment, brain PET was in agreement with MRI in 6 of 8 patients. In the remaining 2 patients there were persistent lesions in brain MRI whereas FDG-uptake was reduced to normal values. In the long-term follow-up of 5 patients (63-169 weeks), 3 patients retained normal in both PET and MRI. In 2 patients a new focal pathologic FDG-uptake was detected 69 and 52 weeks after the end of the treatment. In one of these patients, recurrence was confirmed by MRI not until 9 weeks after PET. CONCLUSIONS: Brain FDG-PET may contribute valuable information for the management of PCNSL, particularly in the assessment of the treatment response. Integration of FDG-PET into prospective interventional trials is warranted to investigate prognostic and therapeutic implications.

Long-Term Efficacy and Safety of Luspatercept for Anemia Treatment in Patients With Lower-Risk Myelodysplastic Syndromes: The Phase II PACE-MDS Study.

Clinical trials frequently include multiple end points that mature at different times. The initial report, typically based on the primary end point, may be published when key planned co-primary or secondary analyses are not yet available. Clinical Trial Updates provide an opportunity to disseminate additional results from studies, published in JCO or elsewhere, for which the primary end point has already been reported.Luspatercept has high clinical activity in patients with transfusion-dependent lower-risk myelodysplastic syndromes (LR-MDS) and ring sideroblasts (RS) relapsed or refractory to erythropoietin. We report long-term luspatercept safety and efficacy in 108 patients with LR-MDS in the PACE-MDS study, including 44 non-RS and 34 non-transfusion-dependent or previously untreated patients. The primary end point was safety. Secondary end points included rates of hematologic improvement (HI) erythroid (HI-E), HI neutrophil, and HI platelet. Exploratory end points included erythropoiesis biomarker quantitation and mutation data. Median duration of luspatercept exposure was 315 days (range, 21-1,934 days). No new safety signals emerged. HI-E was observed in 53.7% of patients, including 36.4% of non-RS and 70.6% of non-transfusion-dependent patients. HI neutrophil and HI platelet were observed in 33.3% and 9.5% of patients, respectively. An almost three-fold increase in bone marrow late to early progenitor cell ratio accompanied HI-E response, irrespective of RS status. Lower baseline erythropoietin levels in non-RS patients (69.6 v 623.3 IU/L; P = .0077) and higher late to early erythroid progenitor cell ratio (10.44 v 4.48; P = .0106) in RS patients were associated with HI-E. This study highlights luspatercept's effects across LR-MDS subtypes, including untreated MDS-RS, serving as a platform for future trials.

Topotecan and ifosfamide systemic chemotherapy for CNS involvement of solid tumors.

The prognosis of patients with CNS involvement of solid tumors is poor. In these patients, systemic chemotherapy has a theoretical advantage of concurrent treatment of systemic disease and reduced risk of neurotoxicity. Here, we report on the efficacy and toxicity of topotecan/ifosfamide (TOPO/IFO) combination chemotherapy in patients treated for CNS involvement of different solid malignancies. Fourteen patients with CNS manifestations (seven with brain metastases, two meningeal carcinomatosis, and five both) of solid tumors (seven with breast cancer, six lung cancer, and one unknown primary cancer) received TOPO/IFO treatment. Eleven patients each were pretreated with 1-6 systemic therapy regimens and whole-brain irradiation. Patients received a total of 34 (median 2) TOPO/IFO cycles. TOPO dosage was 3.6 mg/m(2) (1.2 mg/m(2), days 1-3) and IFO dosage 3,000 mg/m(2) (1,500 mg/m(2), days 1-2) per cycle. Of 12 patients with brain metastases, one patient had partial remission, three stable disease, two progressed, and six had no radiologic CNS response evaluation. Response of meningeal carcinomatosis was found in two and progressive disease in two (three patients not evaluated). Neurologic improvement or stabilization was observed in six of twelve evaluable patients. No systemic tumor response was seen in seven evaluated patients. Grade 3/4 toxicities in eleven evaluable patients were leukopenia (n = 9), infection (n = 6), and thrombopenia (n = 5). Median time to treatment failure was 43 days and median overall survival 107 days. Symptom control was frequently achieved with TOPO/IFO systemic chemotherapy despite a low objective response rate. The feasibility of this treatment is impaired by severe hematotoxicity.

First report on a prospective trial with yttrium-90-labeled ibritumomab tiuxetan (Zevalin) in primary CNS lymphoma.

Most patients with primary CNS lymphoma (PCNSL) relapse after primary therapy. Standard salvage treatment has not yet been established in PCNSL. Anti-CD20 immunotherapy has expanded treatment options in systemic B-cell lymphoma; however, its use is limited by reconstitution of the blood-brain barrier after tumor shrinkage. The aim of this phase II trial was to evaluate the therapeutic efficacy, toxicity, and biodistribution of yttrium-90 ((90)Y) ibritumomab tiuxetan in PCNSL. Ten patients with relapsed PCNSL were included in a phase II trial and treated with the (90)Y-labeled anti-CD20 antibody ibritumomab tiuxetan. Nine patients actually received the planned radioimmunotherapy. In six patients, biodistribution of the antibody was measured by indium-111 ((111)In) ibritumomab tiuxetan whole-body scans and single-photon-emission CT (SPECT) of the brain. All patients were evaluated for toxicity and response at least 4 weeks after therapy. Four patients responded: one patient had a complete response lasting 30+ months, and three patients had short-lived responses of

Systemic high-dose methotrexate plus ifosfamide is highly effective for central nervous system (CNS) involvement of lymphoma.

Patients with malignant central nervous system (CNS) involvement of lymphoma have a poor prognosis with intrathecal chemotherapy and radiation. In this paper, we report the results we obtained in such patients by intravenous chemotherapy with high-dose methotrexate and ifosfamide (HDMTX/IFO). The study involved a review of all patients who received HDMTX/IFO for CNS involvement of malignant lymphoma at our hospital. Therapy consisted of 4 g/m(2) of MTX (4 h infusion on day 1) and 1.5-2 g/m(2)/day of IFO (3 h infusion on days 3-5). The study included 20 patients with a median age of 65 years (range, 30-83) and CNS relapse of a malignant lymphoma. Seventeen patients had been pretreated with up to two chemotherapy regimens. The objective response rate was 90% with 12 complete or unconfirmed complete (CR and CRu) and six partial remissions. All patients had at least stabilization of their neurological symptoms. Myelosuppression was the most common toxicity. The median follow-up time was 14.9 months. The median time to neurological progression was 8.9 months. Twelve patients received subsequent therapy, including high-dose chemotherapy with autologous stem cell transplantation in five cases. The median overall survival was not reached. Systemic chemotherapy with HDMTX/IFO is a feasible and promising treatment modality for CNS relapse of a malignant lymphoma.

Prediction of qualitative outcome of oligonucleotide microarray hybridization by measurement of RNA integrity using the 2100 Bioanalyzer capillary electrophoresis system.

RNA quality is critical to achieve valid results in microarray experiments and to save resources. The RNA integrity number (RIN) can be measured with minimal sample consumption by microfluidics-based capillary electrophoresis. To determine whether RIN can predict the qualitative outcome of microarray hybridization, we measured RIN in total RNA samples from 484 different experiments by the 2100 Bioanalyzer system and correlated with the percentage of present calls (%pc) of downstream oligonucleotide microarrays. The correlation coefficient for RNA and %pc in all 408 samples for which the bioanalyzer algorithm was able to produce an RIN was 0.475 (p < 0.05), ranging from 0.039 to 0.673 for different tissue- and assay-type subgroups. Multivariate analysis found RIN to be the best predictor of microarray quality as assessed by %pc, outperforming the 28S to 18S ratio. For a %pc threshold of 25% and 35%, we determined optimal cut points for RIN at 7.15 and 8.05, respectively. Using the suggested cut points, RIN can support the final decision whether a certain RNA sample is appropriate for successful microarray hybridization.

Intrasplenic Pancreatic Pseudocyst after Chemoradiation of a Pancreatic Adenocarcinoma Mimicking Progressive Disease: A Case Report and Review of the Literature.

Chemoradiation is one of the therapeutic options in palliative treatment of locally advanced pancreatic adenocarcinoma, with a well-known safety profile. In this case report, we describe the treatment-related occurrence of an intrasplenic pancreatic pseudocyst which was successfully removed by gastrocystic drainage. This rare complication should be considered in the follow-up and clinical management of patients, particularly if left-sided complaints occur.

High-dose methotrexate is beneficial in parenchymal brain masses of uncertain origin suspicious for primary CNS lymphoma.

In patients with parenchymal brain masses of uncertain origin responsive to corticosteroids, primary CNS lymphoma (PCNSL) should be considered. PCNSL is a rare but aggressive brain tumor that is highly sensitive to high-dose methotrexate (HDMTX)-based chemotherapy. We report a series of six patients with brain masses without histologic confirmation suspicious for PCNSL based on clinical and radiomorphologic criteria after exclusion of some infectious conditions. All patients were treated with HDMTX. We observed two complete responses, two partial responses, and one stable disease. One patient had progressive disease and received rescue whole-brain irradiation. All patients were alive without disease progression 12-48 months after HDMTX start. No symptoms of late neurotoxicity have occurred so far. The response and survival data in this small series of patients are encouraging and suggest a benefit for patients with suspected PCNSL after initial treatment with HDMTX.

Isolated B-cell lymphoproliferative disorder at the dura mater with B-cell chronic lymphocytic leukemia immunophenotype.

Lymphoma manifestations of the dura mater are extremely rare and have mostly been attributed to extranodal marginal zone lymphoma of mucosa-associated lymphoid tissue (MALT) type. We report a patient with an isolated meningeal tumor, identified as a B-cell lymphoproliferative disorder with typical B-cell chronic lymphocytic leukemia immunophenotype. Because of the subclinical detection of trisomy 3 in the bone marrow by cytogenetic analysis and interphase fluorescence in situ hybridization, CD5(+) MALT is an important differential diagnosis; however, to our knowledge, this entity has never been reported in the context of dural lymphoma.

Phase I trial of the trifunctional anti-HER2 x anti-CD3 antibody ertumaxomab in metastatic breast cancer.

PURPOSE: Ertumaxomab is an intact bispecific antibody targeting HER2/neu and CD3 with selective binding to activatory Fcgamma type I/III receptors, resulting in the formation of a tri-cell complex between tumor cells, T cells, and accessory cells. Patients with metastatic breast cancer were enrolled into a multicenter phase I dose-escalating trial. EXPERIMENTAL DESIGN: Three ascending doses of ertumaxomab (10-200 microg) were administered i.v. on day 1, 7 +/- 1, and 13 +/- 1. Safety and tolerability were the primary objectives. Secondary objectives were antitumor activity and different immunologic variables. RESULTS: Fifteen out of 17 enrolled patients completed the study. One hundred micrograms was identified as the maximal tolerable single dose. Most drug-related adverse events were mild and transient including fever (94%), rigors (47%), headache (35%), nausea (29%), vomiting (29%). Grades 3 and 4 (Common Toxicity Criteria) were lymphocytopenia (76%) and elevation of liver enzymes (47%). One patient (200 mug dose) developed severe hypotension and respiratory distress syndrome, another patient (150 mug dose) developed a systemic inflammatory response syndrome and acute renal failure. Aggravation of congestive heart failure was seen in one patient with preexisting ventricular dysfunction after administration of the third dose (200 microg). All adverse events were fully reversible. Antitumor response was seen in 5 out of 15 evaluable patients (one with a complete response, two with partial responses, two with stable disease) at dose levels of > or = 100 microg. Measurements of cytokines (interleukin-6, interleukin-2, tumor necrosis factor-alpha, and IFN-gamma) suggest a strong T helper cell type 1-associated immune response. The induction of human anti-mouse/anti-rat antibodies was detected in 5 out of 16 (31%) patients. DISCUSSION: Treatment with triple infusions of ertumaxomab yields a strong immunologic response. Doses up to 100 microg can be safely infused with close monitoring of patients. The observed clinical responses are encouraging and indicate antitumor efficacy.

Phase II Trial of Temsirolimus for Relapsed/Refractory Primary CNS Lymphoma.

PURPOSE: In this phase II study (NCT00942747), temsirolimus was tested in patients with relapsed or refractory primary CNS lymphoma (PCNSL). PATIENTS AND METHODS: Immunocompetent adults with histologically confirmed PCNSL after experiencing high-dose methotrexate-based chemotherapy failure who were not eligible for or had experienced high-dose chemotherapy with autologous stem-cell transplant failure were included. The first cohort (n = 6) received 25 mg temsirolimus intravenously once per week. All consecutive patients received 75 mg intravenously once per week. RESULTS: Thirty-seven eligible patients (median age, 70 years) were included whose median time since their last treatment was 3.9 months (range, 0.1 to 14.6 months). Complete response was seen in five patients (13.5%), complete response unconfirmed in three (8%), and partial response in 12 (32.4%) for an overall response rate of 54%. Median progression-free survival was 2.1 months (95% CI, 1.1 to 3.0 months). The most frequent Common Toxicity Criteria ≥ 3° adverse event was hyperglycemia in 11 (29.7%) patients, thrombocytopenia in eight (21.6%), infection in seven (19%), anemia in four (10.8%), and rash in three (8.1%). Fourteen blood/CSF pairs were collected in nine patients (10 pairs in five patients in the 25-mg cohort and four pairs in four patients in the 75-mg cohort). The mean maximum blood concentration was 292 ng/mL for temsirolimus and 37.2 ng/mL for its metabolite sirolimus in the 25-mg cohort and 484 ng/mL and 91.1 ng/mL, respectively, in the 75-mg cohort. Temsirolimus CSF concentration was 2 ng/mL in one patient in the 75-mg cohort; in all others, no drug was found in their CSF. CONCLUSION: Single-agent temsirolimus at a weekly dose of 75 mg was found to be active in relapsed/refractory patients with PCNSL; however, responses were usually short lived.

Progressive multifocal leukoencephalopathy with detection of JC virus in a patient with chronic lymphocytic leukemia parallel to onset of fludarabine therapy.

Along with the establishment of more intense chemotherapeutic regimens including fludarabine for the treatment of indolent lymphoproliferative diseases like chronic lymphocytic leukemia (CLL), an increasing amount of cases with progressive multifocal leukoencephalopathy (PML) due to JC virus have been observed. We report a patient with CLL who developed PML parallel to the onset of fludarabine therapy. Spinal fluid was tested positive for JC virus. Despite virostatic treatment with cidofovir, neurologic symptoms were progressive and the disease ultimately fatal. The present case suggests that immunosuppression caused by chronic lymphoproliferative malignancies alone may be a factor in the development of PML. Chemotherapy with fludarabine may act as an additional trigger. The question remains whether serologic screening for JC virus in patients with chronic lymphoproliferative disease undergoing intense chemotherapy might be valuable once sufficient antiviral treatment has been established.

Efficacy and pharmacologic data of second-generation tyrosine kinase inhibitor nilotinib in BCR-ABL-positive leukemia patients with central nervous system relapse after allogeneic stem cell transplantation.

Central nervous system (CNS) involvement is a severe complication of BCR-ABL-positive leukemia after allogenic stem cell transplantation (alloSCT) associated with fatal outcome. Although second-generation tyrosine-kinase inhibitors (TKI) such as nilotinib have shown activity in systemic BCR-ABL(+) disease, little data exists on their penetration and efficacy within the CNS. Four patients (3 male, 1 female; age 15-49) with meningeal relapse after alloSCT and subsequent treatment with nilotinib were identified. A total of 17 cerebrospinal fluid (csf) and serum samples were assessed for nilotinib concentration and patient outcome was recorded. Nilotinib concentrations showed a low median csf/plasma ratio of 0.53% (range 0.23-1.5%), yet pronounced clinical efficacy was observed with long-lasting responses (>1 year) in three patients. Comparison with historical data showed a trend towards superior efficacy of nilotinib versus imatinib. Despite poor csf penetration, nilotinib showed significant clinical activity in CNS relapse of BCR-ABL(+) leukemias. As nilotinib has a high protein-binding affinity, the low-protein concentration in csf could translate into a relatively higher amount of free and therefore active nilotinib in csf as compared to blood, possibly explaining the observed efficacy. Thus, treatment with a 2nd generation TKI warrants further investigation and should be considered in cases of CNS relapse of BCR-ABL-positive leukemia after alloSCT.

Outcome of elderly patients with primary CNS lymphoma in the G-PCNSL-SG-1 trial.

OBJECTIVE: To assess the outcome of elderly patients with primary CNS lymphoma (PCNSL) treated within the G-PCNSL-SG-1 trial. METHODS: We reviewed response, toxicity, and survival of patients with PCNSL aged 70 or more enrolled in the G-PCNSL-SG-1 trial. RESULTS: A total of 126 of the 526 eligible patients (24%) and 66 of 318 patients (21%) in the per protocol population were aged 70 or more. Among all eligible patients, the rate of complete and partial responses (CR+PR) to HD-MTX-based chemotherapy was 44% in the elderly vs 57% in the younger patients (p = 0.016). Toxicity was age-independent except for a higher rate of grade III/IV leukopenia in the elderly (34% vs 21%, p = 0.007). Death on therapy was more frequent (18% vs 11%; p = 0.027), and progression-free survival (PFS) (4.0 vs 7.7 months, p = 0.014) and overall survival (12.5 vs 26.2 months, p < 0.001) inferior, in the elderly. A striking difference between younger and elderly patients was the PFS of CR patients of 35.0 in the younger vs 16.1 in the elderly patients (p = 0.024). Elderly patients were treated less often and less aggressively at salvage. However, age was not associated with survival from salvage whole brain radiotherapy in patients progressing during primary HD-MTX-based chemotherapy (p = 0.633). CONCLUSIONS: Lower response rate and higher mortality on HD-MTX-based chemotherapy as well as lower PFS of CR patients and less salvage therapy contribute to the poor prognosis of elderly patients with PCNSL.