A supportive care in cancer unit reduces costs and hospitalizations for transfusions in a comprehensive cancer center.

PURPOSE: Among patients with solid or hematologic malignancies undergoing oncologic therapies, blood product transfusions (BPT) are a relevant reason for planned/unplanned hospitalizations, as well as a possible cause of delay in administration of the oncologic therapies. Furthermore, they create additional costs for the healthcare system (HCS). The aim of this study was to compare the costs of performing BPT (erythrocytes and platelets) in medical units/wards to the costs derived from the administration of BPT in a dedicated outpatient supportive care in cancer unit (SCCU). METHODS: Costs were analyzed from June 3, 2009 (when the SCCU started), until December 2013. Four inpatient oncologic units (bone marrow transplantation, radiotherapy, medical oncology I and II) were compared to the SCCU. Data regarding the transfusions performed by the SCCU of the patients who were previously hospitalized for transfusions were extracted, checked, and analyzed through a cross-check on the tax codes. Therefore, patients were considered suitable for the analysis if they had received BPT in the SCCU after a previous hospitalization for transfusion in one of the 4 units/wards. The average daily cost deriving from blood product units and from the hospitalization in each ward (irrespective of pharmaceutical expenses) was compared with the average daily cost deriving from blood product units and from the management of patients in the SCCU. RESULTS: We analyzed 227 patients (112 female) with a mean age of 60 years (range 20-90) with hematologic malignancies in 79% of cases. The number of transfusions performed by the SCCU has grown constantly and consistently over the years, reaching 1,402 transfusions in 2013, thus exceeding the other considered units. The total savings for the HCS was €282.204.71, €151.182.85 in 2013 only. We saved €124.319,26 for each patient transfused at the SCCU. CONCLUSIONS: A dedicated outpatient SCCU, aimed at monitoring and treating cancer therapy-related toxicities and comorbidities and in which it is also possible to perform BPT promptly and effectively, reduces the number of hospitalizations and provides an economical benefit for HCS.

Application of Controlled Shear Stresses on the Erythrocyte Membrane as a New Approach to Promote Molecule Encapsulation.

Human red blood cells (RBCs) have a remarkable capacity to undergo reversible membrane swelling. Resealed erythrocytes have been proposed as carriers and bioreactors to be used in the treatment of various diseases. This work is aimed at developing a setup allowing the encapsulation of test molecules into erythrocytes by inducing reversible pore formation on the RBC membrane through the application of controlled mechanical shear stresses. The designed setup consists of two reservoirs connected by a glass capillary. Each reservoir is connected to a compressor; during the tests, the reservoirs were in turn pressurized to promote erythrocyte flow through the capillary. The setup was filled with a suspension of erythrocytes, phosphate buffer, and FITC-dextran. Dextran was chosen as the diffusive molecule to check membrane pore dimensions. Samples of the suspension were withdrawn at scheduled times while the setup was operating. Flow cytometry and stereo-optical microscopy analyses were used to evaluate the erythrocyte dextran uptake. The setup was shown to be safe, well controlled, and adjustable. The outcomes of the experimental tests showed significant dextran uptake by RBCs up to 8%. Microscopy observations highlighted the formation of echinocytes in the analyzed samples. Erythrocytes from different donors showed different reactions to mechanical stresses. The experimental outcomes proved the possibility to encapsulate test molecules into erythrocytes by applying controlled mechanical shear stresses on the RBC membrane, encouraging further studies.

Biological quality control for extracorporeal photochemotherapy: Assessing mononuclear cell apoptosis levels in ECP bags of chronic GvHD patients.

Extracorporeal photochemotherapy (ECP) is a treatment approved by the FDA for cutaneous T-cell lymphoma, and it is currently used off-label for graft-versus-host disease (GvHD) and other conditions. In agreement with good practices for the therapeutic use of human cells, quality control has to be performed to validate the ECP procedure with the off-line technique. Since no gold-standard biological test is available, we assessed the apoptosis generated in the ECP bag using a flow cytometric analysis. Thirty-one ECP procedures performed on 13 patients with chronic GvHD were studied by monitoring the induction of mononuclear cell (MNC) apoptosis using annexin V/propidium iodide double staining; residual lymphocyte proliferation to standard mitogens was also measured in 17 of the procedures. The kinetics of apoptosis was analyzed at different times in MNCs untreated or treated with 8-methoxy-psoralen plus ultraviolet A; the variation (ΔAPOPTOSIS ) after 24 h revealed the efficacy of the treatment. In 88.6% of the 31 ECP procedures, ΔAPOPTOSIS was >15% (the "alerting" threshold for ΔAPOPTOSIS was set at 15% on the basis of our data); in the remainder (19.4%), the increment in apoptosis was lower. In four procedures, the proliferation assay was useful for assessing the effect of ECP on the apheretic bag. In conclusion, both flow cytometric assays enabled a biologically significant result to be obtained. In our opinion, the apoptosis test-being faster and easier than the proliferation test-could be a reliable way to validate ECP procedures.

PALB2 sequencing in Italian familial breast cancer cases reveals a high-risk mutation recurrent in the province of Bergamo.

PURPOSE: Monoallelic germ-line deleterious mutations of PALB2 (partner and localizer of BRCA2) are associated with breast cancer risk and have been found in several populations, with carrier frequencies of ~1-2%. Initially, these mutations were considered to have moderate penetrance, but accumulating evidence now indicates that they are associated with much higher risk. METHODS: In this study, we sequenced the PALB2 coding regions unlinked to BRCA (breast cancer) genes in 575 probands from Italian breast cancer families recruited in Milan. RESULTS: We found 12 carriers (2.1%) of deleterious mutations, and none of the mutations was found in 784 controls collected in Milan. One of these mutations, the c.1027C>T (p.Gln343X), was found to be recurrent in the province of Bergamo in northern Italy, being detected in 6/113 (5.3%) familial breast cancer cases and 2/477 (0.4%) controls recruited in this area (Fisher's exact test: P < 0.01). CONCLUSIONS: Our data provide confirmatory findings that, in the Italian population also, deleterious mutations of PALB2 are relatively frequent predisposing factors for breast cancer and may be associated with high risk of the disease.

Association between CASP8 -652 6N del polymorphism (rs3834129) and colorectal cancer risk: results from a multi-centric study.

The common -652 6N del variant in the CASP8 promoter (rs3834129) has been described as a putative low-penetrance risk factor for different cancer types. In particular, some studies suggested that the deleted allele (del) was inversely associated with CRC risk while other analyses failed to confirm this. Hence, to better understand the role of this variant in the risk of developing CRC, we performed a multi-centric case-control study. In the study, the variant -652 6N del was genotyped in a total of 6,733 CRC cases and 7,576 controls recruited by six different centers located in Spain, Italy, USA, England, Czech Republic and the Netherlands collaborating to the international consortium COGENT (COlorectal cancer GENeTics). Our analysis indicated that rs3834129 was not associated with CRC risk in the full data set. However, the del allele was under-represented in one set of cases with a family history of CRC (per allele model OR = 0.79, 95% CI = 0.69-0.90) suggesting this allele might be a protective factor versus familial CRC. Since this multi-centric case-control study was performed on a very large sample size, it provided robust clarification of the effect of rs3834129 on the risk of developing CRC in Caucasians.

Peripheral blood CD34+ cell monitoring after cyclophosphamide and granulocyte-colony-stimulating factor: an algorithm for the pre-emptive use of plerixafor.

Plerixafor "on demand" after chemotherapy plus granulocyte-colony-stimulating factor (G-CSF) is efficient in peripheral stem cell mobilization, but the timing of administration and criteria for patient selection are under investigation. To devise an algorithm for the "on demand" use of plerixafor at the first mobilization attempt, we analyzed the kinetics of hematopoietic recovery and peripheral blood CD34+ cells in 107 patients treated with high-dose cyclophosphamide plus G-CSF. Fifty-one patients with myeloma were treated with cyclophosphamide 3-4 g/m(2) on day 0 followed by G-CSF 10 µg/kg from day + 6, and 56 patients with lymphoma received cyclophosphamide 6-7 g/m(2) followed by G-CSF 5 µg/kg from day + 1. Peripheral blood CD34+ cell monitoring was started on day + 8 in patients with myeloma and day + 10 in patients with lymphoma. The outcome of interest was a collection of ≤ 2 × 10(6) CD34+/kg. By a multivariate logistic regression model, CD34+ cell count < 10/µL at leukocyte recovery (> 1000/µL) or leukocyte count < 1000/µL after day + 12 in myeloma and day + 14 in lymphoma predicted the failure of mobilization by 2.7 and 2.8 times (p = 0.001 and p = 0.02) with a sensitivity of 89% and specificity of 88%, respectively. Plerixafor "on demand" may be considered in patients with myeloma and lymphoma with delayed hematopoietic recovery and < 10/µL CD34+ cells, as a first-line mobilization strategy.

Predictors of CD34+ cell mobilization and collection in adult men with germ cell tumors: implications for the salvage treatment strategy.

BACKGROUND: High-dose chemotherapy with tandem or triple carboplatin and etoposide course is currently the first curative choice for relapsing GCT. The collection of an adequate amount of hematopoietic (CD34(+)) stem cells is a priority. PATIENTS AND METHODS: We analyzed data of patients who underwent HDCT at 2 referral institutions. Chemotherapy followed by myeloid growth factors was applied in all cases. Uni- and multivariable models were used to evaluate the association between 2 prespecified variables and mobilization parameters. Analyses included only the first mobilizing course of chemotherapy and mobilization failures. RESULTS: A total of 116 consecutive patients underwent a mobilization attempt from December 1995 to November 2012. Mobilizing regimens included cyclophosphamide (CTX) 7 gr/m(2) (n = 39), cisplatin, etoposide, and ifosfamide (PEI) (n = 42), paclitaxel, cisplatin, and gemcitabine (TPG) (n = 11), and mixed regimens (n = 24). Thirty-seven percent were treated in first-line, 50% (n = 58) in second-line, 9.5% (n = 11) and 3.4% (n = 4) in third- and fourth-line settings, respectively. Six patients did not undergo HDCT because they were poor mobilizers, 2 in first- and second-line (1.9%), and 4 beyond the second-line (26.7%). In the multivariable model, third-line or later setting was associated with a lower CD34(+) cell peak/µL (P = .028) and a lower total CD34(+)/kg collected (P = .008). The latter was also influenced by the type of mobilizing regimen (P < .001). CONCLUSION: A decline in significant mobilization parameters was found, primarily depending on the pretreatment load. Results lend support to the role of CD34(+) cell mobilization in the therapeutic algorithm of relapsing GCT, for whom multiple HDCT courses are still an option, and potentially a cure.

Screening of related donors and peripheral blood stem cell collection practices at different Italian apheresis centres.

BACKGROUND: Recommendations on eligibility criteria for donation of haematopoietic stem cells, management of collection of the cells and follow-up mainly concern unrelated donors. The aim of this study was to analyse the screening of related donors and collection practices at different Italian apheresis centres. MATERIALS AND METHODS: A questionnaire regarding eligibility criteria for related haematopoietic stem cell donors, their peripheral blood collections and early follow-up was sent to several apheresis units. Data from the full charts of 500 candidates, screened between May 2005 and December 2009, were retrospectively evaluated. RESULTS: The donors' records, eligibility criteria, collections and follow-up are managed differently in each centre. Of the 500 evaluable candidates (51.2% male, 49.8% female; median age 47 years, range 13-77), 26.4% underwent thorough screening according to Italian Bone Marrow Donor Registry standards, while local protocols were applied to 73.6%; 91 candidates (18.2%) proved ineligible for donation. In the end, 352 donors (53.4% male, 46.6% female; median age 45 years, range 16-76) underwent 508 leukaphereses. Central venous catheters were used in 8.0% of donors, mainly in one centre. Unsuitable pre-apheresis peripheral blood parameters were reported in 38.7% of the aphereses. Leukapheresis-related adverse events were recorded in 23.0% of the procedures, with a drop-out rate of 0.2% for severe events. No donation-related fatalities occurred. The CD34+ cell yield was <2×10(6)/kg of recipient's body weight from 1.1% of donors ≥70 years old. DISCUSSION: More uniformity in donor screening procedures, management of peripheral blood collection and follow-up should be planned at a national level to maximise the safety of related donors.

Detection of minimal residual disease in hematopoietic progenitor cell harvests: lack of predictive value of peripheral blood and bone marrow analysis in mantle cell and indolent lymphoma.

Elimination of neoplastic cells from peripheral blood progenitor cells (PBPCs) is an important issue in transplantation-based high-dose chemotherapy in non Hodgkin's lymphoma (NHL). The capacity to reliably assess the presence of residual lymphoma cells in PBPCs is mandatory in designing this type of protocols. Polymerase chain reaction (PCR) amplification of molecular rearrangements is widely used to detect minimal residual disease (MRD) in NHL patients. Although concordant data can be obtained in most of the cases from peripheral blood (PB) and bone marrow (BM) at diagnosis, the relationship between these two compartments and the role of their analysis in predicting the molecular status of PBPCs is still an open issue. Here we report data about MRD analysis in BM, PB and PBPCs in a series of mantle cell and indolent NHL patients who underwent high-dose chemotherapy: discordant results were obtained comparing PB, BM and PBPC molecular data. In addition, differences were noted among these results if molecular analysis was performed using well-known rearrangements (i.e., bcl-1/IgH and bcl-2/IgH) or patient specific oligonucleotides. We conclude that neither BM nor PB are reliable in predicting the molecular status of PBPCs and that caution must be adopted in interpreting molecular data obtained using patient specific oligonucleotides.

Tumor-reactive CD8+ early effector T cells identified at tumor site in primary and metastatic melanoma.

CD8(+) T cells at the earliest stage of effector generation have not been identified at tumor site of melanoma patients. Such early effectors, if present, should be characterized by a specific phenotype, distinct from that expressed at later stages of the antigen-induced differentiation program, by short-lived effector cells, memory precursors, and terminal effectors. Here, we show that neoplastic tissues from primary and metastatic lesions of melanoma patients contain a subset of CD8(+) T cells expressing FOXP3. CD8(+) FOXP3(+) CD25(+) T lymphocytes were found in tumor-invaded lymph nodes (TILN), s.c. metastases, and advanced primary lesions. Their frequency was significantly higher in TILN compared with tumor-free lymph nodes or with peripheral blood and in primary tumors compared with TILN. CD8(+) FOXP3(+) T cells did not express markers of regulatory [CTLA-4, CCL4, interleukin-10 (IL-10), transforming growth factor-ß1], exhausted (PD-1), or senescent (CD57) CD8(+) T lymphocytes. Instead, this subset showed an antigen-experienced "EM1" phenotype (CCR7(-) CD45RA(-) CD28(+) CD27(+)) and exhibited a CD127(-), KLRG1(-), HLA-DR(+), CD38(+), T-bet(+), perforin(+) "early effector" profile predicted by current models. CD8(+) FOXP3(+) T cells produced IFN-γ on short in vitro activation, recognized autologous tumor by CD107a mobilization, and expressed Ki-67 on ex vivo analysis. In response to autologous tumor plus IL-2/IL-15, the CD8(+) FOXP3(+) T cells proliferated promptly and showed competence for differentiation (downregulation of CD27 and upregulation of T-bet). These results suggest development of early phases of antitumor immunity even in advanced melanoma. Moreover, the CD8(+) FOXP3(+) "early effector" subset may be an invaluable tool for monitoring immunity at tumor site.

The lipid moiety of haemozoin (Malaria Pigment) and P. falciparum parasitised red blood cells bind synthetic and native endothelin-1.

Endothelin1 (ET-1) is a 21-amino acid peptide produced by the vascular endothelium under hypoxia, that acts locally as regulator of vascular tone and inflammation. The role of ET-1 in Plasmodium falciparum malaria is unknown, although tissue hypoxia is frequent as a result of the cytoadherence of parasitized red blood cell (pRBC) to the microvasculature. Here, we show that both synthetic and endothelial-derived ET-1 are removed by parasitized RBC (D10 and W2 strains, chloroquine sensitive, and resistant, resp.) and native haemozoin (HZ, malaria pigment), but not by normal RBC, delipidized HZ, or synthetic beta-haematin (BH). The effect is dose dependent, selective for ET-1, but not for its precursor, big ET-1, and not due to the proteolysis of ET-1. The results indicate that ET-1 binds to the lipids moiety of HZ and membranes of infected RBCs. These findings may help understanding the consequences of parasite sequestration in severe malaria.

Single-nucleotide polymorphisms inside microRNA target sites influence tumor susceptibility.

Single-nucleotide polymorphisms (SNP) associated with polygenetic disorders, such as breast cancer (BC), can create, destroy, or modify microRNA (miRNA) binding sites; however, the extent to which SNPs interfere with miRNA gene regulation and affect cancer susceptibility remains largely unknown. We hypothesize that disruption of miRNA target binding by SNPs is a widespread mechanism relevant to cancer susceptibility. To test this, we analyzed SNPs known to be associated with BC risk, in silico and in vitro, for their ability to modify miRNA binding sites and miRNA gene regulation and referred to these as target SNPs. We identified rs1982073-TGFB1 and rs1799782-XRCC1 as target SNPs, whose alleles could modulate gene expression by differential interaction with miR-187 and miR-138, respectively. Genome-wide bioinformatics analysis predicted approximately 64% of transcribed SNPs as target SNPs that can modify (increase/decrease) the binding energy of putative miRNA::mRNA duplexes by >90%. To assess whether target SNPs are implicated in BC susceptibility, we conducted a case-control population study and observed that germline occurrence of rs799917-BRCA1 and rs334348-TGFR1 significantly varies among populations with different risks of developing BC. Luciferase activity of target SNPs, allelic variants, and protein levels in cancer cell lines with different genotypes showed differential regulation of target genes following overexpression of the two interacting miRNAs (miR-638 and miR-628-5p). Therefore, we propose that transcribed target SNPs alter miRNA gene regulation and, consequently, protein expression, contributing to the likelihood of cancer susceptibility, by a novel mechanism of subtle gene regulation.

Peripheral blood stem cell collection in multiple myeloma: a retrospective analysis of 6 years leukapheresis activity in 109 patients treated at the Istituto Nazionale dei Tumori of Milan.

Double autologous stem cell transplantation is the standard treatment in newly diagnosed multiple myeloma (MM) patients younger than 65 years; therefore, optimization of leukapheresis is crucial. We performed a retrospective analysis of 297 leukaphereses comparing semiautomated (V4.7 in 20% of collections) versus automated (V6.0 in 80%) Caridian (COBE) Spectra versions and analyzing the influence of M-protein on the outcome. Both methods gave comparable collection efficiencies (CE%) (53.4% vs. 55.7% in V6.0 and V4.7, respectively) with similar leukapheresis time and processed volume. Harvest volume was higher in V4.7 (P < 0.0001) with similar contamination of red blood cells (RBCs) (P = 0.77) and platelets (P = 0.09) when compared with V6.0. In patients with higher peripheral white blood cells (WBCs), V6.0 with adjusted harvest volume (<700 mL), achieved similar CD34(+) CE% (P = 0.39) and better enrichment of nucleated cells (P < 0.0,002) but higher RBCs (P < 0.0,001) and platelets contamination (P = 0.001), when compared with a larger cycle volume in patients with lower WBCs. In hard to mobilize patients, CD34(+) CE% was significantly more efficient with V4.7 than V6.0 (P < 0.0,001). CD34(+) CE% was unaffected by serologic M-protein, but platelet CE% was higher in the absence of M-protein (P = 0.0,003), without any reduction in peripheral patients platelets. We, therefore, conclude that in the setting of MM patients with a high WBCs count and/or low percentage of peripheral CD34(+) cells, collections with V4.7 or adjusted cycle volume V6.0 gave comparable result in CD34(+) CE%. RBCs and platelets contamination is higher if low cycle volume is chosen. In hard to mobilize patients, V4.7 is advisable.

Impaired STAT phosphorylation in T cells from melanoma patients in response to IL-2: association with clinical stage.

PURPOSE: To assess the extent of signal transducer and activator of transcription (STAT) activation in response to interleukin 2 (IL-2) in melanoma patients' T cells, along with clinical stage of tumor progression. EXPERIMENTAL DESIGN: T lymphocytes from peripheral blood of healthy donors and of American Joint Committee on Cancer stage I to IV melanoma patients, as well as from metastatic lymph nodes of patients, were evaluated for responsiveness to IL-2. CFSE assays and single-cell phospho-STAT-specific flow cytometry screening were used. Results. T cells from advanced melanoma patients, in comparison with healthy donors, showed reduced proliferation to IL-2 and IL-15, but not to anti-CD3 monoclonal antibody. Impaired response occurred in CCR7(+) and CCR7(-) T-cell subsets, but not in CD3(-) CD8(+) natural killer (NK) cells, and was not explained by induction of apoptosis, increased cytokine consumption, or altered IL-2R subunit expression in patients' T lymphocytes. By phospho-specific flow cytometry, defective STAT1 and STAT5 activation in response to IL-2 was found mainly in T lymphocytes from peripheral blood and/or tumor site of American Joint Committee on Cancer stage III and IV patients, compared with stage I and II patients and to donors, and in melanoma antigen-specific T cells isolated from metastatic lymph nodes. At tumor site, impaired STAT activation in T cells did not correlate with frequency of CD4(+) CD25(+) Foxp3(+) T cells. Serum from advanced melanoma patients inhibited IL-2-dependent STAT activation in donors' T cells and a neutralizing monoclonal antibody to transforming growth factor beta1 counteracted such inhibition. CONCLUSIONS: These results provide evidence for development of impaired STAT signaling in response to IL-2, along with clinical evolution of the disease, in melanoma patients' T cells.

Management of poor peripheral blood stem cell mobilization: incidence, predictive factors, alternative strategies and outcome. A retrospective analysis on 2177 patients from three major Italian institutions.

CD34+ peripheral blood hematopoietic stem cells (HSC) are usually collected following mobilization therapy accomplished by using growth factors (GF) such as rHuG-CSF or rHuGM-CSF with or without chemotherapy. A target dose of yielded CD34+ is usually prescribed by the attending physician depending on different protocols, which may include single or double transplantation. HSC collection usually is performed when at least 20 CD34+ HSC/microL are detected by means of flow cytometry. A cumulative dose of at least 2 x 10(6)/Kg/bw CD34+ HSC has been considered as the threshold to allow a prompt and persistent hematopoietic recovery. Unfortunately, this goal is not achieved by the totality of patients undergoing mobilization regimen. In fact, 5-46% of patients who underwent mobilization therapy fail HSC collection due to very low peripheral blood HSC CD34+ count. Patients' characteristics, including age, sex, stage of the underlying disease (complete or partial remission), diagnosis, previously administered radio/chemotherapy regimens, time-lapse from last chemotherapy before mobilization and mobilization schedule (including dose of GF) were considered as possibly predictive of poor or failed mobilization. We performed a retrospective analysis in 2177 patients from three large Italian academic institutions to assess the incidence of poor mobilizers within our patients' series. Therefore, a patient who fails a first mobilization (and when an HLA-compatible related on unrelated donor is not available) could undergo a second attempt either with different mobilization schedule or by using different GF, such as stem cell factor, growth hormone (GH), or more recently newly introduced drugs such as AMD3100, alone or in combination with rHuG- or -rHuGM-CSF. Thus, we investigated the fate of those who failed a first mobilization and subsequently underwent a second attempt or alternative therapeutic approaches.

No salvage using high-dose chemotherapy plus/minus reirradiation for relapsing previously irradiated medulloblastoma.

PURPOSE: Myeloablative regimens were frequently used for medulloblastoma relapsing after craniospinal irradiation (CSI): in 1997-2002, we used repeated surgery, standard-dose and myeloablative chemotherapy, and reirradiation. METHODS AND MATERIALS: In 10 patients, reinduction included sequential high-dose etoposide, high-dose cyclophosphamide/vincristine, and high-dose carboplatin/vincristine, then two myeloablative courses with high-dose thiotepa (+/- carboplatin); 6 other patients received two of four courses of cisplatin/etoposide. Hematopoietic precursor mobilization followed high-dose etoposide or high-dose cyclophosphamide or cisplatin/etoposide therapy. After the overall chemotherapy program, reirradiation was prescribed when possible. RESULTS: Seventeen patients were treated: previous treatment included CSI of 19.5-36 Gy with posterior fossa/tumor boost and chemotherapy in 16 patients. Fifteen patients were in their first and 2 in their second and third relapses, respectively. First progression-free survival had lasted a median of 26 months. Relapse sites included leptomeninges in 9 patients, spine in 4 patients, posterior fossa in 3 patients, and brain in 1 patient. Three patients underwent complete resection of recurrence, and 10 underwent reirradiation. Twelve of 14 patients with assessable tumor had an objective response after reinduction; 2 experienced progression and were not given the myeloablative courses. Remission lasted a median of 16 months. Additional relapses appeared in 13 patients continuing the treatment. Fifteen patients died of progression and 1 died of pneumonia 13 months after relapse. The only survivor at 93 months had a single spinal metastasis that was excised and irradiated. Survival for the series as a whole was 11-93 months, with a median of 41 months. CONCLUSIONS: Despite responses being obtained and ample use of surgery and reirradiation, second-line therapy with myeloablative schedules was not curative, barring a few exceptions. A salvage therapy for medulloblastoma after CSI still needs to be sought.