Alberto Agazzi

Angiogenic growth factors and endostatin in non-Hodgkin's lymphoma

A number of clinical studies have demonstrated the prognostic significance of angiogenesis and angiogenic growth factors in solid tumours; however, very little is known about the relevance of these parameters in haematological malignancies. We evaluated circulating levels of angiogenic growth factors and endostatin in 36 non‐Hodgkins lymphoma (NHL) patients. Baseline vascular endothelial growth factor (VEGF) levels of patients in complete remission (CR) after a median follow‐up of 21 months were significantly lower than those of patients with progressive disease (P = 0.016). Event‐free survival (EFS) rate was significantly higher in patients who had baseline VEGF and basic‐fibroblast growth factor (b.FGF) levels below the median values of 147 and 19.5 pg/ml (P = 0.018 and 0.039 by log‐rank test, respectively). Conversely, the levels of endostatin, angiogenin and leptin were not different in CR patients compared to relapsed patients and did not correlate with EFS. Our data suggest that b‐FGF and, particularly, VEGF might be considered prognostic factors in NHL staging and management.

The absence of swirling in platelet concentrates is highly predictive of poor posttransfusion platelet count increments and increased risk of a transfusion reaction.

Quality control of platelet concentrates (PCs) may involve the determination of pH level and the evaluation of platelet count, size, and morphology and/or the response to aggregating agents and hypotonic shock.1 These assays are invasive, require dedicated laboratory devices and personnel, and cannot be performed on all PCs before they are released for transfusion. Platelet discoid shape correlates well with in vivo platelet viability, and discoid platelets, viewed through a light source, manifest the “swirling” phenomenon. Swirling is easily evaluated without any instrumentation by examining a PC against a light source while gently rotating the platelet bag for a few seconds. Two multicenter studies2,3 recently suggested that evaluation of swirling might replace invasive assays for routine PC quality control. In one study,3 involving 5366 PCs in 13 blood transfusion centers in Europe, the United States, Canada, and Japan, 0 to 18 percent of the PCs were found to be negative for swirling. The presence of swirling correlated with pH values that have been associated with adequate platelet viability in vivo. In the absence of data correlating swirling, posttransfusion platelet count increments, and transfusion reactions, we routinely performed quality control of PCs supplied to our hospital by the community blood center. Swirling and other measures of PC quality were evaluated in 47 single-donor PCs collected by apheresis (Excel Cell Separator, Dideco, Mirandola, Italy), which were transfused from October 1998 to April 1999 to 34 cancer patients in our institution. PCs were stored in polyvinylchloride-tri (2ethylhexyl) trimellitate (TOTM) containers (Dideco) in 235 ± 50 mL of plasma. PCs were white cell reduced though a filter (PXL3, Pall, Glen Cove, NY) at the bedside. Five to 15 minutes before transfusion, swirling was determined to be present or absent by visual inspection,3 and the presence or absence of swirling was correlated with 30minute or 1-hour posttransfusion corrected count increments (CCIs) and PC pH values measured at 22 ± 2°C at the end of the transfusion. Swirling was absent in 28 percent of the PCs (Table 1). The percentage of swirling-negative PCs was higher in our experience than that in other studies.3 On the other hand, Schooneman and Claise4 reported that PCs produced by the Excel may have a lower pH during storage than do PCs prepared from other cell separators. Although some PCs were negative for swirling from the first day of storage, statistical analysis (Mann-Whitney U test) demonstrated that swirling-negative PCs were stored significantly longer before transfusion (median, 3 days vs. 1 day) and had a significantly lower pH (6.1 vs. 7.0). There was a trend indicating a higher number of platelets in swirling-negative PCs (4.3 × 1011 vs. 3.5 × 1011). Enumeration of the platelets in PCs with decreased pH may lead to inappropriately low results, because of platelet agglutination. Taken together, these data suggest that the absence of swirling was consistent with previous observations of accumulation of lactic acid, decrease in pH, and loss of platelet discoid shape when there is a lack of oxygen during platelet storage.5 CCIs were significantly lower after the transfusion of swirling-negative PCs (1.2 × 109 vs. 10.1 × 109/L, p<0.001). A CCI of <7.5 × 109 per L is considered to represent a failed PC transfusion.6 CCIs were always <7.0 × 109 per L after the transfusion of swirling-negative PCs. Moreover, febrile (n = 2) or anaphylactic (serous anginoedema of the glottis, n = 2; cutaneous rash, n = 2; hypotension, n = 1) reactions occurred after 2 (6%) of 34 transfusions of swirling-positive PCs and after 5 (38%) of 13 transfusions of swirling-negative PCs (p = 0.019 by chi-square test). To our knowledge, increased rates of transfusion reactions have not been reported after the transfusion of PCs with low pH. It should be noted that swirling-negative PCs were older and might have had increased cytokine levels secondary to the presence of contaminating white cells. Our data indicate that the absence of swirling immediately before transfusion was highly predictive of low pH, poor posttransfusion CCI, and increased risk of transfusion reaction. Evaluation of swirling appears to be a useful firstline quality control of PCs. Swirling-negative PCs should be investigated further and discarded if they fail to meet other standards of quality. Francesco Bertolini, MD, PhD f.bertolini@agora.stm.it Alberto Agazzi, MD Fedro Peccatori, MD Giovanni Martinelli, MD Hematology-Oncology Unit Maria Teresa Sandri, MD Laboratory Medicine Unit IRCCS European Institute of Oncology via Ripamonti 435 20141 Milan, Italy

Coagulation Disorders in Patients with Cancer: Nontunneled Central Venous Catheter Placement with US Guidance—A Single-Institution Retrospective Analysis

PURPOSE To assess the feasibility and safety of ultrasonographic (US) guidance in the placement of nontunneled central venous catheters (CVCs) in patients with cancer who had altered coagulation profiles. MATERIALS AND METHODS The study was approved by the institutional review board; informed consent was obtained. Medical charts of all patients with cancer who underwent nontunneled CVC placement at the European Institute of Oncology, Milan, from September 2001 to August 2008 were retrospectively reviewed. Patients were considered to have coagulation disorders or risk of bleeding when they had the following: prothrombin time more than 1.2 times normal or activated partial thromboplastin time more than 1.2 times normal and/or platelet count less than 150 x 10(9)/L. Patients with a prothrombin time and partial thromboplastin time more than 2.2 times normal and/or a platelet count less than 50,000/mm(3) were considered to be at high risk for bleeding. Two hundred thirty-nine nontunneled CVCs were placed with US guidance in 157 patients. RESULTS One hundred twenty-two (51%) of 239 nontunneled CVCs were inserted in patients with cancer who had hemostasic disorders. Forty-five (37%) of 122 nontunneled CVCs were implanted in patients considered to be at high risk for bleeding. All catheters were successfully placed at the first needle pass with no major complications such as bleeding or pneumothorax. Two hundred thirty-three (97%) nontunneled CVCs were placed in the subclavian vein, and six (3%) were placed in the internal jugular vein. No patient underwent any correction for an abnormal coagulation profile. CONCLUSION In patients with cancer who had coagulation disorders, nontunneled CVC placement with US guidance was feasible and safe and did not require correction of coagulation parameters.

Lymphocytic Toxicity in Patients After Peptide-Receptor Radionuclide Therapy (PRRT) with 177Lu-DOTATATE and 90Y-DOTATOC

PURPOSE Peptide-receptor radionuclide therapy (PRRT) with somatostatin analogs is an efficient new tool in patients with neuroendocrine tumors, with low risk of toxicity. Since lymphocytes express somatostatin receptors, the aim of this study was to evaluate lymphocytic toxicity after PRRT. METHODS From May 2005 to May 2007, 16 patients affected by neuroendocrine tumors received PRRT with (90)Y-DOTATOC (9), (177)Lu-DOTATATE (5), or both (2). Absolute count, percentage of leukocytes and lymphocytes, and lymphoid subsets (B, T, and NK) were tested at baseline and until 90 days after treatment. RESULTS A significant lymphoid toxicity (G2-3), mainly affecting B-cells, was observed. It was particularly evident after (90)Y-DOTATOC. Toxicity resulted in being transient and resolved completely at the end of the follow-up (90 days). CONCLUSION Lymphocyte toxicity in PRRT is mainly due to the selective targeting on B-cells. The relative sparing of T-lymphocytes could explain the absence of clinical side-effects in these patients, such as increased risk of infections. These findings open interesting perspectives in the treatment of B-cell lymphoproliferative disorders.

Tailored therapy of adult acute leukaemia in Jehovah's Witnesses: unjustified reluctance to treat

Treatment of acute leukaemia in adult Jehovahs Witnesses (JW) is challenging because of ‘a priori’ refusal of most physicians to apply diagnostic and therapeutic procedures to haematological abnormalities resembling acute leukaemia. Rejection of blood transfusions by individuals of this faith is usually blamed to justify this attitude, thus leading to severe personal, medical and psychological distress related to the lack of care. We therefore intended to verify whether a standard (tailored) chemotherapy, without the use of prophylactic blood product transfusions, could be applied during treatment of acute leukaemia under such circumstances. Eleven consecutive JW adult patients with acute leukaemia, all of whom had been denied care in other institutions, were treated at the European Institute of Oncology (EIO) in Milan, Italy. Five had acute lymphoblastic leukaemia (ALL) (one bcr/abl positive), six had acute myeloid leukaemia (AML) with immunophenotype and/or cytogenetic intermediate‐high risk features, except one patient with acute promyelocytic leukaemia (APML). Standard induction chemotherapy [cytosine arabinoside (ARA‐C) and daunorubicin (DNR) for AML, vincristine (VCR), DNR and prednisone (PDN) for ALL, all‐trans retinoic acid (ATRA) and DNR for APML] with the antracycline dose of at least 30 mg/sqm were used. All patients experienced severe anaemia after induction chemotherapy despite erythropoietin. Median haemoglobin nadir for patients with ALL and AML was 4.5 g/dL (range 1.3–6.9) and 5.1 g/dL (range 2.6–6.8), respectively. Median platelet nadir counts for all patients was 14.5 × 109)/L (range 1–24). One patient died during induction probably due to haemorrhage. Four of five patients with ALL achieved a complete remission (CR) (including the bcr/abl case) while among patients with AML only the one with APML achieved CR. Three patients (APML = 1 and ALL = 2) are still alive and disease‐free. This small series of adult patients with leukaemia illustrates difficulties in treating patients who are practising JW, yet nevertheless provides a significant argument against the prejudicial decision leading to evasion of treatment in these patients.

Mammaglobin Expression in Leukapheresis Products Is a Predictive Marker of Poor Prognosis in Women with High-Risk Breast Cancer

Purpose: The purpose of this study was to investigate the incidence and prognostic relevance of tumor cell detection in granulocyte colony-stimulating factor–mobilized peripheral blood progenitor cell collections (PBPCCs) using cytokeratin (CK), maspin (MAS), and mammaglobin (MAM) genes as epithelial cell markers. The population on which the study was conducted was drawn from stage III breast cancer patients undergoing high-dose chemotherapy and autologous transplantation with PBPCCs. Experimental Design: One hundred and ninety-four patients were enrolled in the study and analyzed for tumor cell detection on the basis of 481 PBPCCs gathered before administration of chemotherapy. CK, MAS, and MAM gene expressions were investigated by means of the reverse transcription nested polymerase chain reaction, and those samples expressing CK were further hybridized with a radiolabeled internal probe to reduce false-positive results. Sensitivity and specificity were assessed on 37 controls (12 cell lines, 12 healthy donors, and 13 nonepithelial malignancies). Each of the known prognostic variables (age, stage, lymph node status, receptor status, c-ErbB2 status, and Ki67 status) was then analyzed (both individually and together with CK, MAS, and MAM expression on PBPCCs) in relation to patient overall survival (OS) and relapse-free survival (RFS). Results: After a 3-year follow-up, an estimated 83% (95% confidence interval, 77.1–88.8%) of the patients were alive and an estimated 67% (95% confidence interval, 60.1–74.6%) were free of relapse. One hundred and seventy-six of the 194 patients (91%) had contaminated PBPCCs evidenced by at least one positive sample for any of the markers evaluated. The PBPCC frequency of CK, MAS, and MAM positivity (+) was 71%, 36%, and 16%, respectively. MAM expression on PBPCC was associated with an increased risk of relapse (P = 0.003), whereas CK and MAS expressions were not associated with changes in either RFS or OS. Conclusions: MAM gene expression on leukapheresis products of high-risk breast cancer patients is an indicator of poor prognosis. The method of evaluation is simple and reproducible and provides new tools for evaluating the role played by tumor cells in apheresis products and their potential in causing metastasis.

The prevalence and clinical implications of c-kit expression in plasma cell myeloma

Aim:  To evaluate the clinical implications of c‐kit (CD117) expression in plasma cell myeloma (PCM).

Peripheral T-lymphocyte subsets in patients treated with Rituximab-Chlorambucil combination therapy for indolent NHL.

The anti-CD20 chimeric monoclonal antibody, rituximab, is a well-established treatment for newly diagnosed or pretreated follicular non-Hodgkin’s lymphoma (NHL) patients and is able to induce molecular response in about 30% of the cases when used as single agent [1]. The addition of rituximab to chemotherapy improves the clinical and molecular remission rate and prolongs the time to treatment failure [2, 3] without any increase in infectious complications in randomized controlled trials. Prolonged rituximab treatment is associated with a depletion of circulating B cells; however, data are not available on T-cells peripheral subset after a rituximab chemotherapy combination treatment. We retrospectively investigated the effects of rituximab– chlorambucil combination therapy on peripheral blood lymphocyte subsets in 25 low-grade or follicular B-cell NHL patients (12 newly diagnosed and 13 relapsed/ refractory). The regimen [4] consisted of chlorambucil 6 mg/m daily for six consecutive weeks, in association with a standard four-weekly rituximab as induction phase. After revaluation, the responding patients received four additional cycles with chlorambucil (6 mg/m daily for 2 weeks monthly) plus rituximab (once monthly). All patients completed the planned treatment and were evaluated for response and toxicity. After induction, we observed absolute G2 and G3 lymphopenia in 13 and 6 patients, respectively. After consolidation, ten patients had G2 and five patients had G3 absolute lymphopenia. After a median of 9 months (3–30 months) from the end of treatment, six patients showed G2 lymphopenia. As expected, after the first administration of rituximab, immunophenotypic analysis showed a reduction of CD19/ 20+ positive cells below 0.1% in all the patients; that reduction was still present at revaluation after induction and consolidation therapy. As shown in Fig. 1, all the patients presented a significant CD4+ reduction (p<0.001), independent of the levels at baseline (median absolute CD4+ count, 246 cell/ μl), and this trend was maintained during the therapy (median absolute CD4+ count, 216 cell/μl). At median follow-up of 9 months in 50% of the patients, absolute CD4+ T cells had recovered within normal range. On the other hand, we observed no significant reduction of CD8+ and CD56+ absolute values. Infections were rare and none was fatal: two patients developed cutaneous HZV, one had perianal abscess, and one patient was hospitalized for HBV reactivation. To our knowledge, this is the first report on the T-cell profile in patients receiving a rituximab chemotherapy combined chemotherapy. Our experience suggests that chlorambucil–rituximab combination therapy induces a selective decrease of absolute CD4+ T cells during treatment. However, no major infection was observed; the lack of a correlation between the drop in CD4+ T cell count and an increase in infection rate could be related to the relative short period of absolute CD4+ low count. On the contrary, as recently observed by Kaplan et al. [5] in AIDSrelated lymphomas, this effect might justify the increased infectious death rates observed with rituximab-CHOP treatment. Ann Hematol (2006) 85:813–814 DOI 10.1007/s00277-006-0170-9

Idarubicin containing regimen in multiple myeloma: preliminary results of a pilot study using a modified "TANDEM" transplant program.

Tandem autologous transplant actually represents a challenge in multiple myeloma treatment, but the best conditioning regimen is still under investigation. With the aim of evaluating the feasibility of a modified tandem transplant strategy, we treated 10 multiple myeloma patients after conventional first line chemotherapy with a two step conditioning regimen consisting of high-dose melphalan (200 mg/m 2 ) followed by high-dose melphalan (180 mg/m 2 ) together with indarubicin (15 mg/sqm 2 c.i. × 3 days) both with peripheral stem cell support. At first transplant, the median age was 62 years, performance status was good and disease status was CR in 2 patients and PR in the rest. At the end of the first transplant, 70% of patients achieved CR and only mild toxicity was observed. After the second transplant further improvement of the response rate was obtained with 90% CR. However, we observed three toxic early infection-related deaths from CMV and legionella pneumonia at day +17, +26, +54 after transplantation. Although this schedule seems to be effective in terms of response rate, the 30% TRM imposes an anthracycline dose-reduction with careful patient selection. This approach could reduce the toxic effects and maintain the efficacy of therapy at the same time.

Local thrombolytic therapy in cancer patients with central venous catheter occlusion in urgent need of antiblastic treatment: A single institution experience

Central venous catheterism is a common and widespread procedure in haematological cancer patients for delivery of chemotherapy, blood products, parenteral nutrition, fluids or when lacking a peripheral venous access [1]. Despite many efforts to reduce endothelial damage following catheter insertion and the use of biocompatible devices, axillosubclavian deep vein thrombosis (DVTs) represent a common feature (almost 20% in some reports) with severe potential complications regarding the development of pulmonary embolism and sepsis [2–5]. Historically systemic thrombolysis has been performed even in localized thrombosis but the risk of serious bleeding complications (i.e. intracranial haemorrhage) must be considered. Moreover the systemic anticoagulation is not indicated in severe thrombocytopenic patients (<50.000 plts/mm3) after chemotherapy because the risk of bleeding could be unacceptable. Alternatively, the thrombosed Central Venous Catheters (CVCs) must be removed and a new device insertion is required. The main consequences are the following: