Mice were injected subcutaneously with 1 × 105 breast cancer cells in 0.1 ml of PBS. Mice of the control

group (n = 6) were injected with 1 × 106 autologous PBMC, and verum group mice (n = 6) were injected with 1 × 106 autologous CAPRI cells every second day until day 15. PBMC and CAPRI cells were introduced surrounding the injected tumour locations. Mice were observed for 45 days after cancer cell injection. Tumour size was measured for the first time after 21 days. Mice were killed if the maximum tumour diameter was >15 mm unless the tumour killed the mouse before that point. After 45 days, the experiment was completed, and all mice were killed. Pictures were taken with a Konica Minolta Dimage Z3 camera (Konica Minolta Business MAPK Inhibitor Library research buy Solutions Deutschland GmbH, Langenhagen, Deutschland), and figures were prepared with corel PHOTO-PAINT, version 12.0.0.536.,

and Adobe Illustrator CS5, version 3.0.0.400. check details Patient panel, CAPRI cell dose and treatment schedule.  All steps of the production of autologous activated immune cells including the final therapy (treatment attempts) were controlled by the medical doctor (RW) himself. In Germany, medical doctors are allowed to perform such treatment attempts on their own authority. The preparation of CAPRI cells as well as the treatment was performed at the Institute of Immunology of the Ludwig-Maximilians-Universitaet (LMU), München. The patients’ survival data from the Munich Tumor Center were collected from several hospitals, from gynaecologists and from surgeons, independently from the type of treatment, the type of chemotherapy

or radiation therapy. In essence, the data from the Munich Tumor Center are a summary of individual case reports like those from patients treated with CAPRI cells. Each breast cancer patient (T1-4N0-2, G2-3) with diagnosed metastasis (M1, N = 42) who had received at least 500 × 106 CAPRI cells (although higher cell amounts were recommended and often received) was included in the analysis and compared to breast cancer patients with the same tumour staging (T1-4N0-2M1, G2-3) of the Munich Tumor Center (N = 428). Inclusion for treatment was independent of the type of chemotherapy, radiation and/or other therapies. The recommended Amine dehydrogenase treatment schedule included three injections of 60–80 × 106 CAPRI cells per week for 6 months, which was followed by two injections per week for another 6 months. ACT with CAPRI cells has continued for most of the patients once a week for several years. One-third of CAPRI cells were injected i.v., and two-thirds were given i.m. into the forearm in a 1 ml volume of PBS. Statistical analysis.  The slope and y intercept of the regression lines obtained from CML titrations were evaluated using the general linear model (GLM) procedure. The statistical package spss 10.1 (SPSS Inc., Chicago, IL, USA) was used.

The cytoplasmic expression strongly correlated with IL-1α expression (ρ = 0.9583). The cytoplasmic colocalization of HMGB1 and IL-1α was histologically confirmed in cells with collapsing nuclei by the double-staining method. The IgG4/IgG

indexes varied case by case. IL-6 and TLR4 expressions may influence IgG4/IgG index. The nuclei of cells with both IL-1α and HMGB1 expressions in the cytoplasm collapse in the cell death stage. The cooperative high expression of TLR4, IL-6, IL-18, MyD88 and HMGB1 suggest their BEZ235 critical roles in the inflammation circuit. “
“R. D. Jolly, N. R. Marshall, M. R. Perrott, K. E. Dittmer, K. M. Hemsley and H. Beard (2011) Neuropathology and Applied Neurobiology37, 414–422 Intracisternal enzyme replacement therapy in lysosomal storage diseases: routes of absorption into brain Aims: The research concerns enzyme replacement therapy in lysosomal storage diseases with central nervous system involvement. The principle aim was to understand the routes of entry of enzyme into the brain when delivered directly into the cerebrospinal fluid (CSF) via the cerebellomedullary cistern. Methods: Pathways for absorption of replacement enzyme were investigated in dogs with mucopolysaccharidosis IIIA (MPSIIIA) following intracisternal Alvelestat injections of human recombinant N-sulphoglucosamine

sulphohydrolase (rhSGSH, EC3.10.1.1) by light and confocal microscopy using chromogenic and fluorescent immune probes. Results: Enzyme entered the brain superficially by penetration of the pia/glia limitans interface, but the main route was perivascular along large veins, arteries and arterioles extending onto capillaries. It further dispersed into surrounding neuropil to be taken up by neurones, macrophages, astrocytes and oligodendroglia. Enzyme also entered the lateral ventricles adjacent to the choroid plexus, probably also by the tela choroidea and medullary velum, with further spread throughout Rho the ventricular system

and spinal canal. There was secondary spread back across the ependyma into nervous tissue of brain and spinal cord. Conclusions: Enzyme mainly enters the brain by a perivascular route involving both arteries and veins with subsequent spread within the neuropil from where it is taken up by a proportion of neurones and other cells. Penetration of enzyme through the pia/glia limitans is minor and superficial. “
“I. El Ayachi, N. Baeza, C. Fernandez, C. Colin, D. Scavarda, P. Pesheva and D. Figarella-Branger (2010) Neuropathology and Applied Neurobiology36, 399–410 KIAA0510, the 3′-untranslated region of the tenascin-R gene, and tenascin-R are overexpressed in pilocytic astrocytomas Aims: Studying the molecules and signalling pathways regulating glioma invasiveness is a major challenge because these processes determine malignancy, progression, relapse and prognosis.

Covariates were included in the multivariate models based upon clinical importance. The power of the statistics for the RDW differences between the quartiles of prostate volume was 1.0. Statistical analysis was performed using the PASW Statistics 18.0 for Windows (SPSS Inc., Chicago, click here IL, USA). The statistical significance was set at P < 0.05. The demographic characteristics of the 942 patients were analyzed in four groups that were stratified according to the quartiles of prostate volume. These characteristics are summarized in Table 1. Age, IPSS, storage and voiding subscores,

quality of life (QOL) score, PSA, voided volume, peak flow and PVR were significantly different between patients in prostate volume quartiles. GW 572016 The mean prostate volume was 66.6 ± 34.2 mL. For this registry cohort, the mean RDW, WBC, CRP, and ESR were 14.8 ± 1.7%, 7.7 ± 2.1 × 103, 0.8 ± 2.0 mg/dL, and 13.4 ± 12.9 mm/h respectively. The

RDW was significantly related to the WBC and CRP (P = 0.001 and P = 0.014, respectively). Red cell distribution width was significantly correlated with IPSS (P = 0.012), voiding (P = 0.002) and storage subscores (P = 0.020). The relationships between the prostate volume and RDW, WBC, CRP, and ESR are shown in Table 2. The RDW and WBC were significantly associated with the prostate volume in the multivariate linear regression model that was adjusted for age and hemoglobin. The RDW was significantly different between patients in prostate volume quartiles (Table 2). The relationship between RDW and prostate volume can be seen in Figure 1. The IPSS was significantly correlated with the RDW, CRP, and ESR. The RDW had a significant relationship to the IPSS after only adjusting for age. However, in the model adjusted for both age and prostate volume the RDW was not significantly related to the IPSS (P = 0.081) (Table 3). The RDW was significantly elevated in patients choosing to go to surgery rather

than medical therapy (RDW = 15.3% vs. 14.6%, P = 0.001). The relationship between the RDW and the treatment type selleck kinase inhibitor (surgical or medical) is shown in Table 4. The RDW and PSA were significantly associated with the surgical treatment in the multivariate linear regression model that was adjusted for age and prostate volume. This study has disclosed a new scenario for the clinical usefulness of the RDW. The new data from this study suggest a correlation between an increased RDW and prostate volume was. The association remained after adjusting for age and hemoglobin. A graded and independent association of the baseline RDW with the prostate volume was also identified. Finally, the RDW was found to be increased in patients going to surgery for the treatment of BPH. To our knowledge, this is the first study to report a relationship between prostate volume and an elevated RDW.

837. On behalf of the British Neuropathological Society, the editorial team and our publishers, Wiley-Blackwell, I would like to thank Dr Wharton for all of his hard work leading to these achievements. We both appreciate the vital role that the editorial team and reviewers have played in this success and extend our gratitude to all those who have contributed to these activities. The constant professional support of our publishers, see more Wiley-Blackwell; in particular, Ms Elizabeth Whelan and her team, has been invaluable. Neuropathology and Applied Neurobiology, the Journal of the British Neuropathological Society, was established in 1975, 25 years after

the founding of the Society, under the editorship of Professor John Cavanagh who served in this position until 1989. The Journal was subsequently under the energetic leadership of Professors Roy Weller and James Lowe who have

continued to play an active part in recent years. The influence and work of Professor Cavanagh has been honoured by the Society with the foundation of the Cavanagh Prize, awarded every two years to a young neuroscientist who has made a significant contribution to the field of neuropathology. In his opening editorial Professor Cavanagh commented that the discussions of the Society ‘are in the forum of the world’. I believe that this message remains as important today as it was 38 years ago; that the goal of Neuropathology and Applied Neurobiology is to further our understanding of neuropathology Dabrafenib order and underlying disease mechanisms by publishing high quality scientific research GNA12 and to be in the forefront of scientific discussion in this field. Neuropathology and Applied Neurobiology plays an important role in the British Neuropathological Society, of which I have been an active member for many years. I look forward to working with the President, Professor Seth Love, and his successors to maintain the mutual

support between the Society and the Journal. Together we aim to continue the approach of sponsoring lectures at meetings including the International Society of Neuropathology and the European Confederation of Neuropathological Societies, to promote neuropathology on the international stage. Looking forward I will continue to develop the international profile of Neuropathology and Applied Neurobiology. The readily available measure of the impact factor is clearly important for all authors and journals but I believe that there are other markers of quality. Service to our authors and adherence to ethical standards in publishing should be paramount. For authors it is important to have an efficient and fair review process with rapid indexing and availability on-line after acceptance. I will work with the editorial team and publishers to facilitate this.

The amount of phosphorylated p38, ERK 1/2 or STAT5 was calculated as stimulation index equal to the median fluorescence intensity (MFI)stimulated cells/MFIunstimulated cells.[16] Acquisition was performed using an LSR II flow cytometer (BD Bioscience); 5 × 103 events were

collected for analysis. To enumerate CD34+ cells, we used an established multiparameter gating strategy as previously described.[12] Methylcellulose colony assays were completed as previously described[12] using enriched CB CD34+ cells https://www.selleckchem.com/products/obeticholic-acid.html at a plating concentration of 2 × 104 cells/35 mm × 10 mm culture dish (Falcon Plastics) in duplicate. Duplicate cultures were also grown in the presence of supernatant (1/10 final dilution in culture) for 14 days (5% CO2, 37°C). The role of GM-CSF and IL-5 in

supernatant stimulated Eo/B CFU formation was confirmed by adding 5 μg/ml anti-GM-CSF or anti–IL-5 (Peprotech, Rocky Hill, NJ) monoclonal antibodies to the supernatant-stimulated methylcellulose cultures. Eo/B colonies were defined as tight, round refractile cell aggregates of 40 cells or more, staining pink with eosin using Wright–Giemsa (Diff-Quik; Seimens, Newark, DE) and visualized by inverted light microscopy (Olympus CK 40, Olympus Co. Ltd, Tokyo, Japan).[17] Freshly isolated CD34+ progenitor cells were cultured in RPMI complete medium Caspase-dependent apoptosis in the absence or presence of LPS overnight. After overnight incubation

(37°C, most 5% CO2), the cell-free supernatant was harvested and stored at − 80°C for subsequent analysis. Multi-analyte profiling was performed and acquired using a Perkin Elmer CS 1000 Autoplex Analyzer (Luminex XMAP Technology; Austin, TX). A bioplex cytokine assay was used that simultaneously measured the concentrations of GM-CSF and IL-5 in culture supernatant using a human cytokine/chemokine MILLIPLEX MAP kit (Millipore, Mississauga, ON, Canada). The assay sensitivities of these cytokines were 2·3 and 0·1 pg/ml respectively. All analyses were performed according to the manufacturer’s instructions. To determine the mechanism of GM-CSF secretion, CD34+ cells were stimulated with 50 μm STAT5 inhibitor[18] or 50 μm PD98059[19] (ERK 1/2 inhibitor), or 20 μm SB203580[5] (p38 MAPK inhibitor) (Calbiochem, Cambridge, MA) or DMSO vehicle control for 45 min before LPS was added for overnight stimulation to induce GM-CSF secretion. These concentrations were found to be non-toxic to cells and of optimal dosage as determined by preliminary experiments. Data were analysed using IBM SPSS Statistics version 20·0 (Chicago, IL) and presented in figures as mean ± SEM.

Lineage markers were anti-CD3 (clone 145-2C11) and anti-CD19 (clone 1D3) (BD Pharmingen), anti-CD4 (clone RM4-5), anti-CD8 (clone 53-6.7), anti-Gr1 (clone Rb6-8C5) and anti-TER119 (clone

TER119) (kindly provided by Dr. B. Fazekas de St. Groth, Sydney, Australia). Second step reagentia used were streptavidin-allophycocyanin (APC) and streptavidin-APC-Cyanine-7 (BD Pharmingen). For flow cytometric analysis, cells were incubated Cilomilast chemical structure with mAb combinations. The FcγR was blocked by preincubation of cells with saturating amounts of anti-CD16/CD32 mAb to avoid aspecific binding. Cells were analyzed using a FACSCalibur or a LSRII flow cytometer (Becton Dickinson Immunocytometry Systems, CA, USA) with the CellQuest or FACSDiva software program (Becton Dickinson Immunocytometry Systems), respectively. To determine the absolute NK cell numbers, cell suspensions harvested from the different organs were first counted in a counting chamber. Viable cells were discriminated from dead cells using trypan blue and the total viable cell number was calculated. PI was added prior to flow cytometric analysis. Cells were gated on PI-negative cells and then on the lymphocyte gate based on forward and side scatter. Pexidartinib mw In the viable lymphocyte gate, the NK cell percentage was determined by gating on CD3−NK1.1+CD122+ cells. Multiplication of the total viable cell number by the percentage of viable lymphocytes and by the percentage of

CD3−NK1.1+CD122+ cells gives the absolute NK cell number. For detection of granzyme B expression, cells were first cell membrane labelled, permeabilized in Cytofix/Cytoperm reagent (BD Biosciences, Fludarabine supplier CA, USA) and stained with anti-granzyme B mAb. For detection of cytokine-induced IFN-γ production, hepatic leukocytes or DX5-enriched splenocytes were plated in a U-bottomed, 96-well microtitre plate

at 50 000 (liver leukocytes) or 300 000 (splenocytes) cells per well in 200 μL complete medium supplemented with 5 ng/mL IL-12 (R&D Systems) and 2.5 ng/mL IL-18 (Medical & Biological Laboratories, Nagoya, Japan). Plates were incubated at 37°C and 5% CO2. After 3 h, 1/4000 brefeldin A (Golgiplug™, BD Biosciences) was added to each well. After a total culture period of 6 h, cells were collected and stained with anti-NK1.1 and anti-CD3. Cells were permeabilized in Cytofix/Cytoperm reagent (BD Biosciences) and stained with anti-IFN-γ mAb. For NK1.1-stimulated IFN-γ production, 96-well flat-bottomed, non-tissue culture microtitre plates were coated with 0, 6 or 25 μg/mL purified anti-NK1.1 antibody (clone PK136, BD Pharmingen) overnight at 4°C. Afterwards, plates were washed three times and blocked with 2% bovine serum albumin for 30 min. Plates were washed once with medium. A total of 250 000 (liver leukocytes) or 300 000 (splenocytes) cells were added per well in 200 μL complete medium supplemented with 1000 U/mL IL-2 (R&D Systems). Plates were incubated at 37°C and 5% CO2.

Antigen specificity and memory are two essential features of adaptive immunity. A lack of presentation of tumour antigens by DC in vivo in patients with cancer has long been suggested based on findings from early studies in animal models 11, 40, 41. In support of this, abnormalities in DC functional phenotype, with a downregulated expression of MHC class I and class II molecules, have been further demonstrated

in cancer-bearing individuals 42. These findings could thus explain at least in part the insufficient induction of T-cell-mediated anti-tumour immunity observed in patients with cancer 40, 43. Indeed, the very objective learn more initially proposed for DC-based tumour therapy was selleck inhibitor to improve the in vivo presentation of tumour antigens, in an attempt to expand those rare tumour-specific T cells in these patients

11. To maximise the efficiency and stability of antigen presentation by DC, several strategies have been developed. These include the use of various forms of tumour antigens for DC loading, means by which DC were loaded with tumour antigens, and ways through which the antigen-loaded DC were delivered into the patients 11, 44. Moreover, DC transduced with tumour-derived RNA 45, DNA 46 or fused directly with tumour cells 47 have also been tested and shown to be more effective in delivering the tumour-specific signals, and for the induction of anti-tumour responses in vitro and in vivo. One important issue which was not

sufficiently addressed in these early studies, however, was about the abilities of DC to deliver the essential co-stimulatory signals, i.e. in addition Dapagliflozin to the antigen-specific triggers, for T-cell activation. Although the main function of DC is to present antigens to T cells, what make DC special are their potent immunological adjuvanticity and diversified regulatory capacities 7, 14. Importantly, DC can provide both activating and inactivating co-stimulatory signals to the T cells they interact with. These include both the cell surface membrane-bound (e.g. B7) and soluble (e.g. cytokines) molecules. Antigen recognition by T cells in the absence of certain essential co-stimulatory signals may result in T-cell deletion or anergy, and the induction of regulatory T cells 48. The expression or level of expression of these co-stimulatory molecules on DC is again found to be directly associated with the maturation or activation status of the cells. Immature DC are characterised by low surface expression of not only MHC (class I, class II) but also B7 (CD80, CD86) and CD40 molecules 48.

There were dose-related increases in a variety of indicators of pulmonary inflammation, such as number of polymorphonuclear leucocytes, amounts of albumin and lactic dehydrogenase (LDH) in the bronchi and nitric oxide production of alveolar macrophages. Contradictory results were reported from an acute inhalation exposure

in guinea-pigs to non-soluble curdlan, schizophyllan and zymosan (300 µg/m3 for 40 min) [24]. There was no effect on the number of neutrophils in the airways, but a tendency to a decreased number of macrophages and lymphocytes. The discrepancy between the studies is related probably Protease Inhibitor Library chemical structure to the differences in dose levels, where P-glucan in low levels does not induce an inflammatory response. Another reason might be interspecies differences in lung macrophage function [25]. In the present in vitro experiments with PBMC, the dose level per cell was very high compared to environmental exposures. P-glucan caused a large increase in the secretion of IL-6, which was higher among subjects with sarcoidosis. This cytokine is a potent inducer of a general inflammatory response, involving several

cytokines such as IL-17 which CHIR-99021 chemical structure has been related to granuloma formation. Secretion of the anti-inflammatory IL-10, as seen after the stimulation with P-glucan and LPS, will inhibit macrophages and the differentiation of Th2 cells into Th1 effector cells [26]. The secretion was higher among subjects with sarcoidosis, which is in agreement with previous studies where the secretion

of IL-10 from alveolar macrophages was higher among subjects with sarcoidosis compared to controls [27,28]. IL-10 has important anti-inflammatory properties and also supresses granuloma formation [29]. S-glucan was a moderate inducer of cytokines from PBMC. In previous experiments an intratracheal instillation of a soluble β-glucan from www.selleck.co.jp/products/erlotinib.html C. albicans (25–100 ug/animal) was found to induce neutrophil and eosinophil inflammation with increased local expression of a variety of inflammatory cytokines [IL-1β, IL-6, macrophage proteins and regulated upon activation normal T cell expressed and secreted (RANTES)][30]. This suggests that S-glucan and P-glucan trigger different mechanisms for cytokine secretion from PBMC. The relation between the P-glucan-induced release of all the cytokines measured and serum levels of IL-2R and IL-12 connects the PBMC reactivity with two major inflammatory markers of sarcoidosis [6]. The ability of PBMC to secrete IL-12 after stimulation with P-glucan also related to the duration of the disease, reflecting the increasing inflammatory changes developing in sarcoidosis and paralleling the relation between domestic exposure to NAHA and spontaneous secretion of IL-12.

Since the 1980s, the main objective

of VL diagnostics development has been to replace the visual identification of parasites in tissue, a technique that is invasive and requires considerable expertise. Moreover, the use of whole parasite extracts in the serologic tests is limited because of the low reproducibility and low specificity values obtained. Several serological tests have been developed, but none of them are specific for VL, although they have proved to be useful in combination with a clinical case definition (20). So, to obtain a specific diagnosis for VL, some purified recombinant Leishmania antigens have been proposed (21). The present study describes the expression, isolation and purification of two recombinant proteins from L. chagasi, rLci2B and rLci1A that were previously selected from a cDNA library, followed by standardization of an ELISA using South America canine sera, to contribute selleck chemicals llc to the diagnostic of dog’s infection provoked by the protozoa parasite. Both proteins were produced in Escherichia coli, and their sequences were registered by the National Institute of Industrial Property – Brazil (INPI) under paragraph PI0900961-2, with the title: ‘Use of antigens of Leishmania in methods for diagnosis, therapy and vaccine for leishmaniasis’. The rLci2B protein has homology with a parasite cytoskeleton protein kinesin, while rLci1A has homology with the heat shock protein 70 (HSP70). The

HSP70 family belongs to ID-8 a class of proteins highly conserved throughout evolution and has an immunogenic activity

(22). Antibodies to specific recombinant antigens such Palbociclib as heat shock protein 70 (rHSP70) and kinesin K39 (rK39) have also been shown to be good markers to detect infection (23). Thus, this study will be useful to expand the panel of recombinant antigens for the development of more sensitive and specific serodiagnostic tests for this disease. All reagents used were of analytical grade. Distilled water was filtered and deionized using a Millipore water purification system. The sera used in this study were collected from domestic dogs from three Brazilian regions: Northeast, Southeast and Midwest. The serum samples were collected by venipuncture by veterinaries of three research centres: Centro de Pesquisa Aggeu Magalhães, Pernambuco; Instituto de Pesquisa Evandro Chagas, Rio de Janeiro and Centro Gonçalo Muniz, Bahia, between 2006 and 2008. The sera were stored at −70°C and transferred to our laboratory. The panel of 56 negative sera used for determining the cut-off came from the state of Rio de Janeiro. The multicentre panel of 119 negative sera used in the assay came from the Southeast (79), Midwest (26) and Northeast (14) regions of Brazil. The multicentre panel of 138 positive sera used in the challenge originated in the Southeast (38), Midwest (46) and Northeast (54) regions of Brazil. The panel of 86 negative sera for L.

17 Ofsthun et al. reported a similar analysis of 44 550 prevalent haemodialysis patients from the Fresenius Medical BGJ398 cost Care database.18 The relative risk of death for haemoglobin <90 g/L was 2.11 (P < 0.001) compared with a reference haemoglobin level of 110–120 g/L. The relative risk of death decreased to approximately 1.6 and

1.3 as haemoglobin increased to 90–100 g/L and 100–110 g/L, respectively. There was a 16% reduction in mortality for haemoglobin levels between 120 and 130 g/L (RR 0.84, P = 0.007). Fort et al. prospectively studied the effects of time-dependent haemoglobin and ESA dose on mortality in 2310 incident haemodialysis patients from Spain.19 Using a time-dependent multivariate Cox proportional hazard model, the adjusted HR for death was 1.36 (95% CI 1.01–1.86) for a haemoglobin level <100 g/L compared with a level of 111–120 g/L. In contrast, a haemoglobin

level of >130 g/L was associated with a survival benefit (HR 0.69, 95% CI 0.49–0.97). Analysis of the UK Renal Registry data reported similar outcomes with HRs for death for haemoglobin values <100 g/L and >110 g/L being 1.28 (P < 0.001) and 0.64 (P < 0.001), respectively, compared Selleckchem NVP-BEZ235 with a reference haemoglobin level of 100–110 g/L.20 The HRs decreased as achieved haemoglobin increased (Hb 110–120 g/L HR 0.63; Hb 120–130 g/L HR 0.47, and Hb >130 g/L HR 0.44). Zhang et al. conducted a retrospective study of 94 569 prevalent patients who were on haemodialysis in 2000 and 2001.21 The patients were divided into quartiles of ESA dose (1388–7905 U/week, 7905–13 377 U/week, 13 377–22 068 U/week and >22 068 U/week) and five categories of

haematocrit values (<30%, 30–33%, 33–36%, 36–39% and >39%). Mortality rates decreased as haematocrit values increased. Within each haematocrit category, mortality rates were lowest in the lowest quartile of ESA dose and highest in the highest quartile. A US Medicare study reported outcomes of 393 967 prevalent haemodialysis patients from 2002 to 2004.22 In a fully adjusted Cox proportional hazard model, mortality was higher at all haematocrit levels pheromone below 34.5% compared with the reference haematocrit level of 34.5% to 36%. The HR for death increased from 1.17 (95% CI 1.14–1.20) to 3.11 (95% CI 3.01–3.20) when haematocrit decreased from 33–34.5% to <27%. Similarly, mortality increased at all levels of haematocrit >39%. Mortality was comparable for haematocrit levels between 36% and 39%. When patients were grouped into five categories of erythropoietin dose (0 U/week, 0–6000 U/week, 6000–12 000 U/week, 12 000–18 000 U/week and >18 000 U/week), the HR for death progressively increased with increasing dose of erythropoietin for every level of haematocrit.