Multiple transplant-related complications (MTRC) represent a severe condition after bone marrow transplantation (BMT) and are supposed to reflect systemic endothelial damage. Soluble thrombomodulin (sTM) and plasminogen activator inhibitor type-1 (PAI-1) were investigated as markers of endothelial dysfunction in 35 patients after autologous or allogeneic BMT and compared with the occurrence of the typical complications sepsis, veno-occlusive disease of the liver (VOD), graft-versus-host disease (GVHD), and capillary leakage syndrome (CLS). PAI-1 was assessed by an assay of functional activity and sTM by antigenic determination. In patients who had undergone allogeneic BMT and had no transplant-related complications (TRC), PAI-1 peaked on day +14 (20 +/- 5 units/ml), and sTM doubled in comparison to the starting range, to 60-80 ng/ml between days +14 and +49. In contrast, PAI-1 and sTM were unchanged following autologous BMT. PAI-1 was increased in sepsis, CLS, and VOD to 39-49 units/ml (p < 0.05, compared with patients without TRC), and in GVHD to 16-47 units/ml (not significant). Soluble TM increased to 63-309 ng/ml in patients with sepsis, VOD, or CLS (p < 0.05, compared with patients without TRC) and to 79-224 ng/ml in GVHD (not significant). The increase of sTM and PAI-1 was also positively correlated to the number of complications, so that in patients with three complications PAI-1 was increased 2.8-fold and sTM 3.5-fold over patients with no complications at all. We conclude that endothelial dysfunction is a feature of VOD, sepsis, CLS, and to lesser extent of GVHD and is worse in patients with multiple complications.

Vascular-leak syndrome (VLS) is a common complication in the first 3 weeks after bone marrow transplantation (BMT). The patients present with weight gain, generalized edema, ascites, pericardial or pleural effusions, tachycardia, arterial hypotonia, and/or pre-renal failure. The aim of our study was to investigate the role of the complement system in VLS. The protein concentrations of C3 and C4 were studied by immunodiffusion, and total hemolytic complement activity was studied by assessment of CH50. C1 esterase inhibitor (C1 Inh), the major inhibitor of the classical pathway of complement, was assessed by a functional test. Activation of complement was assessed by C4d (a C4 activation product). Twelve patients were followed prospectively from start of conditioning therapy to day +21 after bone marrow transplantation. Eight of 12 patients did not develop VLS. These patients had an increase of C3 between day +9 and day +13 (range: 1.3- to 1.5-fold, median: 1.4-fold), C4 (range: 1.3- to 1.9-fold, median: 1.4-fold), CH50 (range: 1.3- to 1.6-fold, median: 1.4-fold), and C1 Inh (range: 1.2- to 1.5-fold, median: 1.3-fold). Four of 12 patients developed VLS. C1 Inh activity was decreased to 0.60- to 0.80-fold. This decrease began 2-6 days prior to clinical diagnosis of VLS (n = 3), or at onset of VLS (n = 1). Patients with VLS showed elevated C4d concentrations (up to 2.4 mg/dl, upper normal threshold value: 0.9 mg/dl). Patients with VLS reveal an activated state of the complement system which is accompanied by a reduced activity of C1 Inh. Insufficient control of complement activation may contribute to VLS in patients after BMT.

Infusions of recombinant human tumor necrosis-alpha plus recombinant human interferon-gamma (rhTNF alpha/rhIFN gamma) were assessed in two patients with Ewing's sarcoma. We analyzed platelet count, coagulation and the terminal complement complex (TCC). During cycles with continuous rhTNF alpha-infusions we found a rapid, marked decrease of platelet count (minus 90% of initial values) and a simultaneous increase of TCC (plus 84% of initial values) at day 4. At days 5-7 a spontaneous increase of platelet count and decrease of TCC were visible. Short-term infusion led to a milder, continuous decrease of platelet counts and to a moderate, progressive increase of TCC at days 5-7. Bleeding occurred only as petechia and mild hemoglobinuria. There was no effect related to the dosage of rhTNF alpha or rhIFN gamma. Relevant differences were seen only in the variable time courses of rhTNF alpha application. Ex vivo analysis of one patient's platelets showed no cytokine-related effect on induced aggregation according to Born. Additionally, we analyzed in vitro effects of the cytokines on platelet count, platelet aggregation, and the assembly of TCC in platelet membranes. No effects were found after incubation of platelet-rich plasma (PRP) with 1000 pg/ml rhTNF alpha and/or 50 pg/ml rhIFN gamma. Fluid-phase and membrane-bound TCC did not change after incubation of PRP with cytokines. A slightly time-dependent increase of TCC without alteration of platelet count and platelet function did not agree with the assumption of a direct injury to platelets. We assume a cytokine-mediated role of the endothelium in platelet loss.

Neisseria meningitidis infection may present as meningitis or as severe, fulminant sepsis. In order to classify individual patients early according to the expected course of the disease, we developed a score named Neisseria sepsis index [NESI]. The NESI was defined using the parameters heart rate, mean arterial blood pressure, base excess and presence of acute subcutaneous bleeding and/or skin necroses (minimal value [= no evidence for sepsis] NESI 0; maximum value [= most severe sepsis] NESI 8). Seventeen patients with documented, systemic N. meningitidis infection were prospectively assessed for the terminal complement complex (TCC), serum tumour necrosis factor alpha (TNF alpha) levels (as laboratory parameters for severity of sepsis) and NESI score. The evaluation was immediately performed when the patients were admitted to the hospital. The 17 patients showed the following distribution of data: NESI 0 (n = 4), NESI 1 (n = 6), NESI 2 (n = 0), NESI 3 (n = 1), NESI 4 (n = 2), NESI 5 (n = 2), NESI 6 (n = 0), NESI 7 (n = 1), NESI 8 (n = 1). Mortality was 4/17 patients, all had NESI > or = 5. TCC values ranged from 647-6461 ng/ml (normal range: 130-360 ng/ml); and was not correlated to NESI. TNF alpha values ranged from 10-910 pg/ml and were correlated to NESI (r2 = 0.71, n = 17, P < 0.001). In patients with fatal outcome, TNF alpha was 600 +/- 160 pg/ml (mean +/- SEM) and in surviving patients 130 +/- 50 pg/ml (mean +/- SEM). TNF alpha was increased in 15/17 patients when compared to normal controls (< 27 pg/ml). CONCLUSION. The NESI is based on few clinical, objective data, that are available in every hospital. NESI appears to offer an instrument: (1) for making decisions in regard to appropriate monitoring and treatment of vital organ function; and (2) for assessing the quality of care for this life-threatening infection.

Capillary leakage syndrome (CLS) is a severe side effect of intravenous interleukin-2 (IL-2) therapy. Twenty-seven cycles of IL-2 therapy [six (day I), nine (day 2), and 12 x lo6 U/m2 body surface (days 3 to 3, given as continuous infusion] were analyzed in children and adolescents. The anaphylatoxin C5a was assessed as an early predictor for CLS. CLS developed in 11 of 27 cycles of IL-2 infusion. C5a at day 2 of IL-2 infusion (0.8s9.43 pg/L; median, 1.8 p.g/L) was increased in CLS patients when compared with baseline values (0.21-0.74 pg/L; median, 0.40 pg/L; p = 0.01) and when compared with C5a at day 2 in non-CLS patients (0.44-1.2 pg/L; median, 0.62 pg/L;p < 0.01). Ten of 1 1 CLS patients showed C5a levels > 1 .O p.g/L, whereas 14 of 16 patients who did not develop CLS showed C5a < 1 .O p.g/L (predictive value positive 83% for CLS).

The concentrations of the complement components C3 and C4 and their activation products C3dg and C4d were determined in EDTA-stabilized serum of 25 premature and term infants. EDTA plasma and EDTA serum obtained from 30 normal blood donors were used as controls. According to clinical, laboratory and/or microbiological findings, six of the 25 children had infections. The mean scatter range of the C3 and C4 values was from 30% (in the 30th week of pregnancy) to 80% (in term infants) of the normal value for adults. In all the children, irrespective of gestational age, the C3dg concentrations were of the same order of magnitude as in healthy adults. As regards the C3, C4, and C3dg values, there was no difference between the newborns with and without infections. The C4d values of the newborns without infections, on the other hand, (range 0.1-1.4 mg/dl, mean 0.8 mg/dl, n = 19) were significantly lower than those of the newborns with infections (range 1.3-2.4 mg/dl, mean 1.95 mg/dl, n = 6). Observation of the course and comparison with CrP showed that elevated C4d values may occur earlier. In the authors' view, these findings indicate that in bacterial infections of premature and term infants the fourth complement component is activated, while the extent to which the third complement component is involved in the activation process is not measurable. Further studies are needed to establish whether early diagnosis of neonatal sepsis can be improved by determining C4d.

To see whether the clinical manifestation of thrombotic events or hemorrhagic infarctions appears as a relevant problem when treating children with acute lymphoblastic leukemia (ALL) concerning the COALL therapy-protocol, we started an inquiry of the participating hospitals. The mentioned protocol was designed by the German Society for Pediatric Oncology and Hematology to treat ALL in childhood. All participants gave us information about the treatment period from January 1989 to December 1992. In 6 from 286 treated patients a thromboses appeared in clinical terms. None of them was connected with a lethal outcome. There was no observation of a hemorrhagic infarction. The overall thromboses frequency was 2.1%. In 1.4% patients "symptomatic" thrombosis developed close to a continuous venous catheter, which can be considered as a thrombogene risk factor. About 0.6% (2/286) of the patients developed the thrombotic events without another risk factor. They can be regarded as "idiopathic". 1/2 idiopathic thromboses led to a life threatening situation. There are two important factors that can enhance thromboses: 1) the therapy period, especially induction therapy and application of asparaginase and 2) a continuous venous catheter. The fact that asparaginase is not used during the induction therapy is a characteristic of the COALL protocol. It seems to be useful to differentiate between "idiopathic" and "symptomatic" thrombotic events, because "symptomatic" thromboses appear also in non-leukemic diseases quite frequently.

No abstract