Factors affecting Fontan length of stay: Results from the Single Ventricle Reconstruction trial.

BACKGROUND: In the Single Ventricle Reconstruction trial, infants with hypoplastic left heart syndrome (HLHS) who received a right-ventricle-to-pulmonary-artery shunt (RVPAS) versus a modified Blalock-Taussig shunt (MBTS) had lower early postoperative mortality, but more complications at 14 months. We explored the effect of shunt type and other patient, medical, and surgical factors on postoperative length of stay (LOS) after the Fontan operation. METHODS: Fontan postoperative course was ascertained from medical record review. Cox proportional hazards modeling was used to identify factors associated with LOS. RESULTS: Of 327 subjects who underwent Fontan, 323 were analyzed (1 death, 1 biventricular repair, 2 with missing data). Median age and weight at Fontan were 2.8 years (interquartile range [IQR]: 2.3, 3.4) and 12.7 kg (IQR: 11.4, 14.1), respectively. Fontan type was extracardiac in 55% and lateral tunnel in 45%; 87% were fenestrated. The RVPAS and MBTS subjects had similar LOS (median 11 days [IQR: 9, 18] vs 10 days [IQR: 9, 13]; P = .23). Independent risk factors for longer LOS were treatment center (P < .01), LOS at stage II (hazard ratio [HR] 1.02 for each additional day; P < .01), and pre-Fontan complications (HR 1.03 for each additional complication; P = .04). Use of deep hypothermic circulatory arrest at Fontan (HR 0.64; P = .02) was independently associated with shorter LOS. When center was excluded from the model, pre-Fontan complications and use of circulatory arrest were no longer significant; instead, older age at stage II (HR 1.08 for each additional month; P = .01) predicted longer LOS. In 254 subjects who had a pre-Fontan echocardiogram, at least moderate tricuspid regurgitation was independently associated with longer LOS, both with center (HR 1.72; P < .01) and without center in the model (HR 1.49; P = .02). CONCLUSIONS: In this multicenter prospective cohort of subjects with HLHS, Norwood shunt type was not associated with Fontan LOS. Rather, global measures of earlier medical complexity indicate greater likelihood of longer LOS after the Fontan operation.

Relationship among chelator adherence, change in chelators, and quality of life in thalassemia.

PURPOSE: Thalassemia, a chronic blood disease, necessitates life-long adherence to blood transfusions and chelation therapy to reduce iron overload. We examine stability of health-related quality of life (HRQOL) in thalassemia and adherence to chelation therapy over time, especially after changes in chelator choice. METHODS: Thalassemia Longitudinal Cohort participants in the USA, UK, and Canada completed the SF-36v2 (ages 14+) and the PF-28 CHQ (parents of children <14 years). Chelation adherence was defined as self-reported percent of doses administered in the last 4 weeks. RESULTS: Two hundred and fifty-eight adults/adolescents (mean 29.7 years) and 133 children (mean 8.5 years) completed a mean of 2.8-years follow-up. Children made few chelator changes, whereas a mean of 2.2 changes was observed among the 37% of adults/adolescents who made chelator changes, mainly due to patient preference or medical necessity. Physical HRQOL improved among those with lower iron burden (better health status) at baseline who made a single change in chelator, but declined among participants with multiple changes and/or high iron burden (worse health status). Mental health improved among participants with lower iron burden, but iron overload was negatively associated with social functioning. Adherence did not significantly change over follow-up except for an increase after a change from deferoxamine (DFO) infusion to oral deferasirox (p = 0.03). Predictors of lower adherence for adults/adolescents at follow-up included side effects, smoking, younger age, problems preparing DFO, increased number of days per week DFO prescribed, and lower physical quality of life . CONCLUSIONS: Strategies to balance medical needs with family, work, and personal life may assist in adherence.

Doppler flow patterns in the right ventricle-to-pulmonary artery shunt and neo-aorta in infants with single right ventricle anomalies: impact on outcome after initial staged palliations.

BACKGROUND: A Pediatric Heart Network trial compared outcomes in infants with single right ventricle anomalies undergoing Norwood procedures randomized to modified Blalock-Taussig shunt (MBTS) or right ventricle-to-pulmonary artery shunt (RVPAS). Doppler patterns in the neo-aorta and RVPAS may characterize physiologic changes after staged palliations that affect outcomes and right ventricular (RV) function. METHODS: Neo-aortic cardiac index (CI), retrograde fraction (RF) in the descending aorta and RVPAS conduit, RVPAS/neo-aortic systolic ejection time ratio, and systolic/diastolic (S/D) ratio were measured early after Norwood, before stage II palliation, and at 14 months. These parameters were compared with transplantation-free survival, length of hospital stay, and RV functional indices. RESULTS: In 529 subjects (mean follow-up period, 3.0 ± 2.1 years), neo-aortic CI and descending aortic RF were significantly higher in the MBTS cohort after Norwood. The RVPAS RF averaged <25% at both interstage intervals. Higher pre-stage II descending aortic RF was correlated with lower RV ejection fraction (R = -0.24; P = .032) at 14 months for the MBTS cohort. Higher post-Norwood CI (5.6 vs 4.4 L/min/m(2), P = .04) and lower S/D ratio (1.40 vs 1.68, P = .01) were correlated with better interstage transplantation-free survival for the RVPAS cohort. No other Doppler flow patterns were correlated with outcomes. CONCLUSIONS: After the Norwood procedure, infants tolerated significant descending aortic RF (MBTS) and conduit RF (RVPAS), with little correlation with clinical outcomes or RV function. Neo-aortic CI, ejection time, and S/D ratios also had limited correlations with outcomes or RV function, but higher post-Norwood neo-aortic CI and lower S/D ratio were correlated with better interstage survival in those with RVPAS.

DTPA Fe(III) decreases cytokines and hypotension but worsens survival with Escherichia coli sepsis in rats.

OBJECTIVE: Nonselective inhibition of nitric oxide (NO) with NO synthase antagonists decreases hypotension but worsens outcome clinically. We investigated whether iron (III) complex of diethylenetriaminepentaacetic acid [DTPA Fe(III)], a scavenger of NO as well as other oxidant mediators, has similar divergent effects in E. coli challenged rats. METHODS: Animals with venous and arterial catheters and challenged with intrabronchial or intravenous E. coli were randomized to treatment with DTPA Fe(III) in doses from 3 to 800 mg/kg or placebo. Mean blood pressure (MBP) was measured in all animals and plasma NO, cytokines, and blood and lung leukocyte and bacteria counts in animals administered intrabronchial E. coli and DTPA Fe(III) 50 mg/kg or placebo. Animals received antibiotics and were observed 168 h. RESULTS: Independent of drug regimen or infection site, compared to placebo, DTPA Fe(III) increased MBP although this was greater with high vs. lower doses. Despite increased MBP, DTPA Fe(III) worsened the hazards ratio of survival . At 6 and 24 h DTPA Fe(III) decreased NO but not significantly and decreased four cytokines (tumor necrosis factor-alpha, interleukins 1 and 10, and macrophage inflammatory protein 3alpha) and lung lavage neutrophils. From 6 to 24 h DTPA Fe(III) increased blood bacteria. CONCLUSIONS: DTPA Fe(III) while increasing blood pressure has the potential to worsen outcome in sepsis. Further preclinical testing is required before this agent is applied clinically.

Effective dosing of lipid A analogue E5564 in rats depends on the timing of treatment and the route of Escherichia coli infection.

BACKGROUND: E5564, a competitive lipid A antagonist, inhibits endotoxin-stimulated inflammation and is under study in patients with sepsis. METHODS: We tested whether clinically relevant variables, including the timing of treatment and the route of infection, influenced the effective dosing of E5564 in Escherichia coli-challenged rats. RESULTS: All E5564 doses (0.3, 1.0, 2.0, and 3.0 mg/kg intravascular bolus followed by 10% of the bolus dose infused hourly for 24 h) administered 1 h before intravascular E. coli challenge similarly reduced the risk of death. Delaying the start of E5564 to 1 or 3 h after intravascular E. coli challenge significantly reduced the beneficial effect of the doses tested. However, increasing the dose of E5564 reversed some loss of efficacy for delayed treatment (P=.004, for increasing benefit with increasing dose at 1 h). During intrabronchial or intraperitoneal (extravascular) E. coli challenge, the pattern of effective E5564 dosing was the inverse of that for intravascular E. coli challenge (P=.001, for the interaction)--lower doses of E5564 were beneficial and higher doses were not (0.03, 0.3, 1.0, 2.0, and 3.0 mg/kg bolus followed by infusion) (P=.05, for decreasing benefit with increasing dose at 1 h). CONCLUSION: These findings suggest that, for maximal clinical benefit, E5564 should be given early and that dosing should be adjusted upward for intravascular infection and downward for extravascular infection.

Neutrophil stimulation with granulocyte colony-stimulating factor worsens ventilator-induced lung injury and mortality in rats.

BACKGROUND: Based on the association between the neutrophil and ventilator-induced lung injury, the authors hypothesized that neutrophil inhibition with fucoidin would be beneficial and stimulation with granulocyte colony-stimulating factor (G-CSF) would be harmful in a rat model of lethal ventilator-induced lung injury. METHODS: Animals (n = 111) were randomly assigned to be pretreated with fucoidin, G-CSF, or placebo (control) before 4 h of low-tidal-volume (10 ml/kg) or high-tidal-volume (40 ml/kg) mechanical ventilation. RESULTS: All low-volume animals survived. With high volumes, compared with controls, fucoidin did not improve survival (3 of 20 control animals and 5 of 20 fucoidin animals died; P = 0.51) but G-CSF significantly worsened it (18 of 22 animals died; P < 0.001). Circulating neutrophils were increased early with G-CSF and late with fucoidin with low and high tidal volumes (P < 0.05 for each treatment and tidal volume). Fucoidin decreased lung neutrophils, but these were only significant with high tidal volumes, whereas G-CSF increased lung neutrophils but only significantly with low tidal volumes (P < or = 0.01 for each). Fucoidin did not alter any cardiopulmonary measure significantly. Compared with control, G-CSF increased airway pressures with high tidal volumes and worsened lung edema and arterial oxygen with both tidal volumes (P < 0.05 for each). CONCLUSIONS: In this model, neutrophil stimulation by G-CSF increased lung dysfunction and with high tidal volumes worsened survival rates. Extrapolated clinically, neutrophil stimulation either by agents such as G-CSF or conditions such as sepsis may aggravate ventilator-induced lung injury.

Prophylactic high-dose N(omega)-monomethyl-L-arginine prevents the late cardiac dysfunction associated with lethal tumor necrosis factor-alpha challenge in dogs.

We investigated nitric oxide (NO) as a possible cause of the cardiac dysfunction associated with high, lethal doses of tumor necrosis factor-alpha (TNF-alpha) in dogs. Eighty-seven awake, 2-year-old (10-12 kg), purpose-bred beagles were randomized to receive an infusion of saline or N-monomethyl-L-arginine (L-NMMA), a nonselective NO synthase (NOS) inhibitor, as a 40 mg kg bolus followed by a 40 mg kg(-1) h(-1) infusion for 3 to 6 h 3 h before (prophylactic) or 3 h after (therapeutic) challenge with TNF-alpha (60 microg kg(-1)) or vehicle. Serial radionuclide-heart scans and thermodilution pulmonary artery catheter hemodynamic measurements were performed. The effects of prophylactic L-NMMA on TNF-alpha-induced cardiac dysfunction as measured by decreases in mean left ventricular (LV) ejection fraction and downward and rightward shifts of LV function plots (peak systolic pressure versus end systolic volume index and LV stroke work index versus end diastolic volume index) were significantly different comparing early (3-6 h) and delayed (24 h) time points (P = 0.02). Prophylactic L-NMMA therapy did not appear to fully prevent early (3-6 h) TNF-alpha-induced cardiac dysfunction, but at 24 h, complete protection was seen. Therapeutic L-NMMA did not appear to fully protect the heart from TNF-alpha-induced early or delayed cardiac dysfunction (P = NS). Similarly, L-NMMA given prophylactically, but not therapeutically, blocked TNF-alpha-induced increases in exhaled NO flow rates and plasma nitrite and nitrate concentrations (both P = 0.02). These data suggest that TNF-alpha initiates two phases of cardiac injury: an early (3-6 h) phase that may be partially NO independent and a delayed (24 h) phase that is NO dependent. The delayed, more persistent dysfunction can be completely blocked by high doses of a nonselective NOS inhibitor administered before TNF-alpha.

Granulocyte colony-stimulating factor has differing effects comparing intravascular versus extravascular models of sepsis.

BACKGROUND: Previously, neutrophil stimulation with granulocyte colony-stimulating factor (G-CSF) pretreatment increased survival rates in canines challenged with intraperitoneal or intrabronchial Escherichia coli and in rats challenged with intrabronchial Staphylococcus aureus. We investigated whether G-CSF pretreatment would be beneficial with intravascular challenge in these models. METHODS: Animals were randomized to G-CSF or placebo pretreatment followed by intravenous E. coli challenge in canines (n = 24) or intravenous or intrabronchial S. aureus challenge in rats (n = 273). All animals were treated with antibiotics. RESULTS: In canines, G-CSF before intravenous E. coli did not decrease mortality rates (7 of 12 [58%] G-CSF vs. 5 of 12 [42%] controls), which contrasted with prior reductions during extravascular infection (10 of 35 [29%] G-CSF vs. 37 of 65 [57%] controls). Consistent with the present and previously published studies in canines, in rats, G-CSF decreased mortality rates with intrabronchial S. aureus (22 of 90 [24%] G-CSF vs. 26 of 51 [51%] controls, p = 0.009) but did not decrease them with intravenous infection (34 of 67 [50%] G-CSF vs. 27 of 65 [42%] controls, p = 0.2) in patterns that were very different (p = 0.005 for the effects of G-CSF with intravascular vs. intrabronchial S. aureus). CONCLUSION: In contrast to extravascular infection, sepsis with intravascular E. coli in canines and S. aureus in rats may not provide a compartmentalized nidus of bacteria on which G-CSF-stimulated neutrophils can exert a beneficial antimicrobial effect. Extrapolated clinically, a proinflammatory agent like G-CSF may be most beneficial with sepsis related primarily to a compartmentalized extravascular site of infection.

Changes in the use of postmenopausal hormone therapy after the publication of clinical trial results.

BACKGROUND: The recent publication of clinical trial results has led to a dramatic shift in the evidence about postmenopausal hormone therapy. OBJECTIVE: To examine whether the publication of clinical trial results, specifically the Heart and Estrogen/progestin Replacement Study (HERS) in 1998 and the Women's Health Initiative (WHI) in 2002, has influenced the use of hormone therapy among postmenopausal women. DESIGN: Observational cohort (1997 to 2003). SETTING: San Francisco Mammography Registry, San Francisco, California. PARTICIPANTS: Postmenopausal women between the ages of 50 and 74 years without a personal history of breast cancer who underwent mammography (151862 mammograms). MEASUREMENTS: Self-reported current use of hormone therapy. RESULTS: Among menopausal women who had mammography, it was estimated that 41% were currently using hormone therapy in 1997. Before the publication of HERS, the use of hormone therapy was increasing at a rate of 1% (95% CI, 0% to 2%) per quarter. After the publication of HERS, use decreased by 1% (CI, -3% to 0%) per quarter. In contrast, the publication of the WHI in 2002 was associated with a more substantial decline in the use of hormone therapy of 18% (CI, -21% to -16%) per quarter. Similar associations were observed for most subgroups of women, including women older than 65 years of age; women with a previous hysterectomy; and women who described their race or ethnicity as white, African American, Latina, Chinese, or Filipina. CONCLUSIONS: The release of the HERS data was temporally associated with a modest decline in the use of hormone therapy. In contrast, the release of the principal findings from the WHI was associated with a more substantial decline in use by postmenopausal women. The reason for the differences in decline may relate to the fact that the WHI results were widely publicized or were more applicable to most postmenopausal women because the WHI study was performed in healthy women.

TGF-beta1 increases microbial clearance but worsens lung injury during Escherichia coli pneumonia in rats.

We investigated the effects of either intravenous (IV) or intrabronchial (IB) treatment with transforming growth factor beta1 (TGF-beta1) during bacterial pneumonia in rats. Immediately following IB Escherichia coli inoculation (T0), animals (n=270) were randomized to receive a single treatment with human recombinant TGF-beta1 either via IV or IB, or via both IV and IB routes, or to receive placebo (human serum albumin, HSA) only. Blood and lung analysis was done at 6 and 168 h after E. coli inoculation. Other animals (n=40) were administered IV TGF-beta1 or HSA at T0 and 6, 12 and 24 h after E. coli inoculation to investigate the effects of multiple treatments also on survival rates alone. All animals received ceftriaxone daily. Route of administration did not influence TGF-beta1 (p=ns for the effect of TGF-beta1 comparing IV vs IB routes) and we averaged over this variable in analysis. The relative risk of death (mean +/- sem) was not altered by either single treatments administered at T0 (-0.18 +/- 0.25, p=0.47) or multiple treatments (0.40 +/- 0.50, p=0.66) of TGF-beta1. Single treatment with TGF-beta1 first decreased and then increased vascular leukocytes at 6 and 168 h, respectively, but increased alveolar leukocytes at both time points (p=0.02 comparing the differing effects of TGF-beta1 on vascular and alveolar leukocytes at 6 and 168 h). Although TGF-beta1 decreased blood and lung bacteria counts at 6 and 168 h, it also increased serum tumor necrosis factor levels and lung injury scores at these time points (p<0.05 for the effects of TGF-beta1 on each parameter at 6 and 168 h together). Thus, while increases in lung leukocyte recruitment with TGF-beta1 were associated with improved microbial clearance in this rat model of pneumonia, worsened lung injury may have negated these beneficial host defense effects, and overall survival was not significantly improved. Despite these harmful effects, additional studies may be warranted to better define the influence of exogenous TGF-beta1 on host defense during acute bacterial infections.

Protection with antibody to tumor necrosis factor differs with similarly lethal Escherichia coli versus Staphylococcus aureus pneumonia in rats.

BACKGROUND: Differing factors may alter the effects of antibody to tumor necrosis factor (TNF) in infection and sepsis. The authors tested whether bacteria type or treatment route alters antibody to TNF in a rat model of bacterial pneumonia. METHODS: Rats (n = 231) received similarly lethal doses of either intratracheal Escherichia coli or Staphylococcus aureus followed by treatment with either intratracheal or intraperitoneal antibody to TNF or control serum. Animals received antibiotics (cefotiam daily dose, 100 mg/kg) starting 4 h after inoculation and were studied for up to 96 h. RESULTS: Compared with S. aureus, E. coli increased serum TNF and interleukin-6 concentrations, lung lavage TNF concentrations, neutrophil counts, and alveolar-to-arterial oxygen gradients and decreased circulating neutrophils and lymphocytes (P > or = 0.05 for all). Compared with controls, with both bacteria, except for lung lavage TNF concentrations (which decreased with intratracheal but not with intraperitoneal antibody to TNF), treatment route did not alter the effects of antibody to TNF on any parameter (P = not significant for all). Antibody to TNF reduced mortality rates (relative risk of death +/- SEM) with both E. coli (-1.6 +/- 0.6; P = 0.006) and S. aureus (-0.5 +/- 0.04; P = 0.185), but these reductions were greater with E. coli than with S. aureus in a trend approaching statistical significance (P = 0.09). Compared with controls, similarly (P = not significant) with both bacteria, antibody to TNF decreased lung lavage and tissue bacteria concentrations (both P < 0.05) and serum TNF concentration (P < 0.09) and increased circulating neutrophils and lymphocytes (both P < or = 0.01). Compared with S. aureus, with E. coli antibody to TNF decreased alveolar-to-arterial oxygen gradients (P = 0.04) and increased serum interleukin-6 concentrations (P = 0.003). CONCLUSION: Antibody to TNF improved host defense and survival rates with both lethal E. coli and S. aureus pneumonia, but protection was greater with E. coli, where TNF concentrations were higher than with S. aureus. The efficacy of antiinflammatory agents in sepsis may be altered by bacteria type.

Sympathetic blockade in a canine model of gram-negative bacterial peritonitis.

We investigated, in a well-established canine model of human sepsis, the effects of two different techniques of sympathetic blockade during bacterial peritonitis on pain relief, hemodynamics, and survival rate. Twenty-two purpose-bred beagles (12-28 months old, weighing 10-12 kg) were studied. Fourteen animals received an epidural infusion of bupivicaine and morphine, and the other eight received either a celiac plexus block (n = 4) or a sham block (n = 4). Eighteen of the 22 animals received an intraperitoneal challenge of Escherichia coli (1-10 x 10(9) CFU kg(-1) body weight). At comparable doses of intraperitoneal-implanted E. coli (2.5-5 x 10(9) CFU kg(-1) body weight), the addition of sympathetic blockade produced a synergistic decrease in survival times (P = 0.002) and mean left ventricular ejection fraction (P = 0.008), and increase in creatinine levels (P = 0.02). There was also a significant increase in tumor necrosis factor (TNF) levels (P = 0.004) and decrease in blood endotoxin clearance (P = 0.006) associated with sympathetic blockade during sepsis. The celiac plexus-blocked animals had no improvement in pain scores, and subjectively looked clinically worse than animals with sepsis without a celiac plexus block. In contrast, the epidural block was effective in blocking the pain and discomfort associated with low lethality doses of intraperitoneal bacteria reflected by no increase in pain scores compared with animals not receiving bacterial challenge. This study shows that during severe bacterial peritonitis, maintenance of sympathetic tone irrespective of pain relief provided is necessary for clearance of bacterial toxins, control of proinflammatory mediator release, hemodynamic stability, and survival.

Inflammation-promoting activity of HMGB1 on human microvascular endothelial cells.

Systemic inflammation because of sepsis results in endothelial cell activation and microvascular injury. High-mobility group protein-1 (HMGB1), a novel inflammatory molecule, is a late mediator of endotoxin shock and is present in the blood of septic patients. The receptor for advanced glycation end products (RAGE) is expressed on endothelium and is a receptor for HMGB1. Here we examine the effects of HMGB1 on human endothelial cell function. Recombinant human HMGB1 (rhHMGB1) was cloned and expressed in Escherichia coli and incubated with human microvascular endothelium. rhHMGB1 caused a dose- and time-dependent increase in the expression of intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule 1 (VCAM-1), and RAGE. rhHMGB1 induced the secretion of tumor necrosis factor-alpha (TNFalpha), interleukin 8 (IL-8), monocyte chemotactic protein-1 (MCP-1), plasminogen activator inhibitor 1 (PAI-1), and tissue plasminogen activator (tPA) (P <.01). rhHMGB1 stimulation resulted in transient phosphorylation of mitogen-activated protein (MAP) kinases, extracellular signal-related kinase (ERK), Jun N-terminal kinase (JNK), and p38, and in nuclear translocation of transcription factors NF-kappaB and Sp1. These effects are partially mediated by TNFalpha autocrine stimulation, as anti-TNFalpha antibodies significantly decrease chemokine and adhesion molecule responses (P