Experimental validation of a new ultrasound method for the simultaneous assessment of radial and longitudinal myocardial deformation independent of insonation angle.

BACKGROUND: Strain and strain rate have been proposed as tools to quantify regional myocardial function. One of the major pitfalls of the current methodology is its angle dependency. To overcome this problem, we have developed a new method for the estimation of strain, independent of angle. The aim of this study was to validate this new methodology in an experimental setting using sonomicrometry. METHODS AND RESULTS: In 5 open-chest sheep, ultrasound data were acquired. The new methodology was used to perform simultaneous measurements of radial and longitudinal strain in the inferolateral wall. Segment-length sonomicrometry crystals were used as the reference. After baseline acquisitions, deformation was modulated by pharmacologically changing the inotropic state of the myocardium and by inducing ischemia. Ultrasonically estimated radial and longitudinal strain were validated against sonomicrometry by means of Bland-Altman analysis and the intraclass correlation coefficient. For both strain components, good agreements were found between the ultrasound and the sonomicrometry measurements as shown by Bland-Altman statistics. The intraclass correlation coefficients were found to be 0.72 and 0.80 for the radial and longitudinal components, respectively. CONCLUSIONS: A new technique for the estimation of myocardial deformation was validated. It was shown that the current problem of angle dependency was solved and that 2 deformation components could be estimated simultaneously and accurately. Furthermore, the technique was less time-consuming, because anatomic tracking was performed automatically. This approach could potentially accelerate the clinical acceptance of ultrasound deformation imaging in cardiology.

Experimental assessment of a new research tool for the estimation of two-dimensional myocardial strain.

One-dimensional strain imaging has been shown to be angle dependent. To address this problem, a new methodology, 2D-strain, has become available. The aim of this study was to validate this methodology in an in vivo set-up against sonomicrometry. In five open chest sheep, ultrasound gray-scale images were acquired of the inferolateral wall from two different angles. The longitudinal and radial strain components were simultaneously extracted using the novel 2D-strain methodology. The extracted values were compared with sonomicrometry using Bland-Altman statistics and correlation coefficients. A good agreement was found for the longitudinal strain component, while, for the radial strain estimates, the accuracy was less. 2D-strain is a fast and accurate tool to assess longitudinal strain from apical views. Further improvements are needed for the method to be sufficiently accurate in estimating the deformation perpendicular or close to perpendicular to the ultrasound beam.

Association of mannose-binding lectin levels and invasive fungal disease in hematologic malignancy patients receiving myelosuppressive chemotherapy or allogeneic hematopoietic stem cell transplantation.

Several studies have suggested an association of mannose-binding lectin (MBL) deficiency with infections. In this study, we investigated the association between MBL deficiency and invasive fungal disease (IFD) in hematologic malignancy patients receiving myelosuppressive chemotherapy or hematopoietic stem cell transplant. MBL levels were quantified at the start of treatment in 152 patients who were followed for 6 months and scored as developing IFD or not. Forty-five patients (29.6%) developed IFD, of which 21 (46.7% of IFD cases and 13.8% of patients) were proven or probable IFD. Fifty-nine (38.8%) had MBL levels <1000 ng/mL. The rates of all IFD in patients with MBL levels below and above 1000 ng/mL were 33.9% and 26.9%, respectively (P=0.356). The rates of proven or probable IFD in patients with MBL levels below and above 1000 ng/mL were 11.9% and 15.1%, respectively (P=0.579). MBL levels <1000 ng/mL were not predictors of death (P=0.233). As expected, IFD was associated with death (P<0.0001). Our findings indicate that MBL levels <1000 ng/mL were not associated with an increased risk of developing IFD or overall survival.

Effects of lipids on the functional and metabolic recovery from global myocardial stunning in isolated rabbit hearts.

OBJECTIVES: High concentrations of free fatty acids may increase myocardial ischaemic damage. However, the administration of lipid emulsions during reperfusion improves the functional recovery of stunned myocardium. From this apparent controversy we hypothesise that the effect of lipids is related to the time of its administration: we compared the effects of pre- and post-ischaemic administration of Intralipid((R)) on stunned myocardium. We also examined the role of fatty acids and phospholipids, respectively, in the effect of lipid emulsions on stunned myocardium. METHODS: Myocardial stunning was produced by 15 min of ischaemia and 90 min of reperfusion in isolated blood perfused rabbit hearts. Intralipid((R)) was administered either prior to ischaemia or during reperfusion. Left ventricular pressure (LVP) and its first derivative (LVdP/dt) were measured to assess functional recovery. High energy phosphates were measured with HPLC. The effects of linoleic acid, phosphatidylcholine and their combination were also studied. RESULTS: Only when Intralipid((R)) was administered during reperfusion, it improved recovery from contractile function and increased high energy phosphate content in globally stunned myocardium. Both linoleic acid and phosphatidylcholine significantly improved myocardial function in stunned myocardium. CONCLUSIONS: The effect of lipids on the contractile performance and metabolic state of stunned myocardium depends mainly on the timing of its administration with regard to the ischaemia/reperfusion event. Both free fatty acids and phospholipids contribute to the beneficial effect of lipid emulsions on functional recovery of stunned myocardium.

Drug interactions in the hematopoietic stem cell transplant (HSCT) recipient: what every transplanter needs to know.

Pharmacokinetic drug interactions among hematopoietic stem cell transplant recipients can result in either increases in serum concentrations of medications, which may lead to enhanced toxicity; or reduced serum concentrations, which can lead to treatment failure and the emergence of post transplant complications. The use of drugs that have a narrow therapeutic index, such as cyclosporine/tacrolimus (calcineurin inhibitors), increases the significance of these interactions when they occur. This report will review the clinical data evaluating the drug interactions of relevance to HSCT clinical practice, focusing on the pharmacokinetic interactions, and provides recommendations for managing these interactions to avoid both toxicity and treatment failure.

Late onset of invasive aspergillus infection in bone marrow transplant patients at a university hospital.

Despite new antifungal treatment strategies, invasive aspergillosis (IA) remains a principal cause of infectious mortality after bone marrow transplantation (BMT). We reviewed the medical records of 93 allogeneic and 149 autologous transplant recipients during a 20 month period, with attention to cases of proven or probable IA. No autologous transplant recipient developed IA, whereas IA was seen in 15.1% of allogeneic recipients (including two of five patients with a prior history of IA despite prophylaxis), for an overall incidence of 5.8%. The median time to occurrence was 92 days post transplant, with no de novo cases developing prior to engraftment. Survival 100 days from diagnosis was 29%. Risk factors for the development of IA included > or = 21 days of corticosteroid therapy of >or= 1mg/kg/day and post-transplant cytomegalovirus (CMV) infection. These two risk factors were statistically linked. Our data illustrate a shift toward a later occurrence of post-transplant IA, suggesting a need for close, prolonged surveillance in the outpatient environment. The contributory role of protracted corticosteroid use is also highlighted. These data have important implications in an era of alternate donor transplants and more intense immunosuppression. Established strategies implementing newer, less toxic antifungal agents as prophylaxis in high-risk patients are needed.

Infections following hematopoietic stem cell transplantation.

Numerous advances have been made in the management of infection in HSCT recipients. With increasing knowledge the authors are able to prevent several serious infections from occurring, and reduce the severity of infections once they occur. Despite these advances, several previously unrecognized pathogens have emerged and pose risks to this population. Ongoing surveillance and reporting of atypical infections are warranted. Transplant and infectious disease clinicians alike must be vigilant to the shifts in infectious syndromes as a consequence of various prophylaxis and preemptive strategies, and be ready to modify empiric strategies to meet the changing microbiologic milieu. As we increase our understanding of the HSCT process, and use the immune system rather than relying on high-dose chemotherapy, the authors are likely to reduce toxicities and improve patient outcomes.

Treatment of refractory acute leukemia with timed sequential chemotherapy using topotecan followed by etoposide + mitoxantrone (T-EM) and correlation with topoisomerase II levels.

A phase I/II clinical study evaluated 17 patients with refractory/recurrent acute leukemia treated with 1.5 mg/m2/day topotecan on days 1-3 followed by etoposide (100 mg/m2/day)+mitoxantrone (10 mg/m2/day) on days 4, 5 and 9, 10. Timed sequential chemotherapy using the topoisomerase I-inhibitor topotecan before the topoisomerase II-inhibitors, etoposide+mitoxantrone (T-EM) treatment is proposed to induce topoisomerase II protein levels and potentiate the cytotoxic activity of the topoisomerase II-directed drugs. Fourteen patients had refractory and three had recurrent acute leukemia. The majority of patients were heavily pre-treated with greater than three re-induction chemotherapy regimens. Ten patients responded to T-EM treatment (59%). Four of seventeen (24%) had a complete remission and one had a partial remission. Four additional patients (24%) who scored complete leukemia clearance had no evidence of disease with complete white and red blood cell recovery but with platelet counts less than 100,000. The lack of platelet recovery in one patient having a partial response was scored as a partial leukemia clearance. The toxicity profile included major non-hematological toxicity including grade 3 mucositis (29%) and neutropenic fever (65%). Paired measurements of intracellular levels of topoisomerase II isoforms alpha and beta in leukemia blast cells (bone marrow) collected before (day 0) and after topotecan treatment (day 4) showed that a relative increase of topoisomerase IIalpha (Topo IIalpha) > or = 40% strongly correlated with response after T-EM treatment. Increased Topo IIalpha levels also corresponded to increased DNA fragmentation. Two patients who had an increase of Topo IIalpha of 20-25% had either a PR or PLC while patients with a < 10% increase showed no response to T-EM treatment. We conclude that timed sequential chemotherapy using topotecan followed by etoposide+mitoxantrone is an effective regimen for patients with refractory acute leukemia, and demonstrate Topo IIalpha protein level increases after topotecan treatment.

Empiric antifungal therapy for the neutropenic patient.

One of the major challenges facing oncologists today is invasive fungal infection. Difficult to diagnose and deadly when missed, invasive fungal infection--primarily by Candida and Aspergillus organisms--is the major infectious cause of death associated with chemotherapy-induced myelosuppression. In this review, the problem will be described and evidence-based approaches to management, including assessment for risk factors and empiric antifungal therapy, will be discussed. Finally, the future of diagnostic and therapeutic strategies for protecting the immunocompromised patient will be considered.

Clinical care for patients receiving autologous hematopoietic stem cell transplantation in the home setting.

PURPOSE/OBJECTIVES: To undertake a pilot study of autologous hematopoietic stem cell transplantation (HSCT) in the patient's home to improve satisfaction of care, reduce financial costs, and relieve pressure on inpatient accommodation. DESIGN: Descriptive, cross-sectional, qualitative. SETTING: Patients' homes within the metropolitan area of Perth, Australia. SAMPLE: 25 Caucasian adults with recurrent multiple myeloma, non-Hodgkin's lymphoma, or Hodgkin's disease requiring autologous HSCT. METHODS: A program was developed to use the bone marrow transplant team from a major tertiary hospital to permit home visiting, treatment with cytotoxic chemotherapy at home, treatment of complications at home, and an integrated home/hospital caring facility to expedite hospital admission if complications developed. FINDINGS: The program was practical to administer, improved overall patient satisfaction, and was significantly less costly than inpatient transplantation. Fifteen patients (60%) of the total study group of 25 required hospital admission for a median of five days (range 1-13 days) for management of complications, predominantly febrile neutropenia. Nineteen (76%) of the 25 patients received i.v. antibiotic therapy at home during the period of neutropenia. Two patients died of transplant-related complications--one from respiratory syncytial virus infection and one from veno-occlusive disease of the liver. These complications were not attributable to the home setting. IMPLICATIONS FOR NURSING PRACTICE: This program increased the responsibility and sense of autonomy of advanced practice nurses and developed their counseling skills as well as their ability to participate more actively in the decision-making process of those involved. MAIN RESEARCH VARIABLES: Participation in the home transplant program, patient satisfaction, nursing development, and cost-effectiveness of the program.

Multifocal idiopathic fibrosis presenting as facial pain and trismus.

An unusual case of idiopathic sclerosing fibrosis is described in which there was involvement of the face, neck, mediastinum and lungs with sparing of the thyroid and retroperitoneal tissues. In contrast to previous cases, facial pain and limitation of jaw movements were the presenting complaints and major causes of disability. Prednisolone treatment failed to arrest the disease.

Reduced-intensity conditioning using fludarabine with either antithymocyte globulin and BU, or low-dose TBI allowing allogeneic hematopoietic SCT.

In a single-center study, we analyzed the outcomes of 66 patients with advanced hematological malignancies receiving two reduced-intensity conditioning regimens for allogeneic transplantation: fludarabine and low-dose TBI (flu/TBI, n=25), or fludarabine, antithymocyte globulin and BU (flu/ATG/BU, n=41). The selection criteria were based on the hypothesis that flu/TBI patients were expected to achieve autologous recovery in the event of non-engraftment. Sixty-three patients (95%) engrafted. Regimen-related mortality at day 100 and 1 year was 6 and 15%, respectively. With median follow-up of 50.4 months, survival did not differ by regimen. Multivariate analysis confirmed that the type of regimen did not affect relapse. In patients achieving full donor chimerism by day 30, those conditioned with flu/TBI showed greater overall survival (P=0.02). Engraftment failure occurred in two patients (3%), both of whom received flu/TBI. We conclude that conditioning with flu/TBI or flu/ATG/BU yields comparable survival and remission outcomes. By contrast to our hypothesis, patients receiving flu/TBI who subsequently failed engraftment did not achieve autologous recovery. Yet, rapid attainment of full donor chimerism after flu/TBI is associated with greater survival than after flu/ATG/BU. Further, larger prospective randomized studies are required to define the advantage of one regimen over the other.

Performance characteristics of the platelia Aspergillus enzyme immunoassay for detection of Aspergillus galactomannan antigen in bronchoalveolar lavage fluid.

We have evaluated the Platelia Aspergillus enzyme immunoassay for detection of galactomannan in bronchoalveolar lavage (BAL) specimens in solid organ transplant patients with aspergillosis. The precision and reproducibility in serum or BAL to which galactomannan was added were similar. Sensitivity was 81.8% in patients with aspergillosis, and specificity was 95.8% in lung transplant patients who underwent BAL for surveillance for infection or rejection. Among transplant controls, positive results were more common in patients (i) who underwent diagnostic BAL performed for evaluation of symptoms or chest computed tomographic abnormalities, (ii) who had undergone lung transplantation, or (iii) who were colonized with Aspergillus. Galactomannan testing in BAL is useful for diagnosis of aspergillosis in transplant patients. The significance of positive results in patients without confirmed aspergillosis requires further evaluation.

Longitudinal but not circumferential deformation reflects global contractile function in the right ventricle with open pericardium.

The clinical evaluation of right ventricular (RV) contractility is problematic because instantaneous RV volumetry is difficult to achieve. Our aim was to test whether global RV contractility can be assessed by using regional indexes in the longitudinal and/or circumferential axis. Six anesthetized adult ewes were instrumented with a RV conductance catheter and four RV free wall sonomicrometry crystals (interrogating the longitudinal and circumferential axes). Global and regional preload recruitable stroke work (PRSW) were measured by using acute vena cava occlusions at baseline, during esmolol and dobutamine infusion, and during stable low-preload and high-afterload conditions. The agreement between regional and global PRSW was assessed with regression and Bland-Altman analysis. Both regional PRSW indexes correlated well with global PRSW in baseline conditions, during inotropic modulation (R(2) = 0.83 and 0.74 for longitudinal and circumferential regional PRSW, respectively), and during preload reduction (R(2) = 0.62 and 0.83, respectively), but only longitudinal regional PRSW correlated with global PRSW in increased afterload conditions (R(2) = 0.59 and 0.13 for longitudinal and circumferential regional PRSW, respectively). We conclude that in the open-chest, open-pericardium animal model, deformation in the longitudinal axis accurately reflects global RV contractile function in baseline conditions and during acute load modulation, whereas circumferential motion is influenced by changes in afterload.

Acute pulmonary hypertension causes depression of left ventricular contractility and relaxation.

BACKGROUND AND OBJECTIVE: The haemodynamic effects of acute pulmonary hypertension can be largely attributed to ventricular interdependence during diastole. However, there is evidence that the two ventricles also interact during systole. The aim of the present study was to examine the effects of acute pulmonary hypertension on both components of left ventricular systole, i.e. contraction and relaxation, using load-independent indices. METHODS: Ten pigs were instrumented with biventricular conductance catheters, a pulmonary artery flow probe and a high-fidelity pulmonary pressure catheter. Haemodynamic measurements were performed in baseline conditions and during stable pulmonary vasoconstriction induced by the thromboxane analogue U46619. Contractility was quantified using the end-systolic pressure-volume and preload recruitable stroke work relationships. The tau-end-systolic pressure relationship was used to assess load-dependency of relaxation. RESULTS: Acute pulmonary hypertension caused a decrease in the slope of the left ventricular preload recruitable stroke work relationship (from 6.64 +/- 1.7 to 5.19 +/- 1.9, mean +/- SD; P < 0.05), a rightward shift of the end-systolic pressure-volume relationship (P < 0.05), and an increase in the slope of the tau-end-systolic pressure relationship (from -0.15 +/- 0.5 to 0.35 +/- 0.17; P < 0.05). The diastolic chamber stiffness constant of both ventricles increased during pulmonary hypertension (P < 0.05). CONCLUSIONS: In the present model, acute pulmonary hypertension impairs left ventricular contractile function and relaxing properties. The present study provides additional evidence that, besides the well-known diastolic ventricular cross talk, systolic ventricular interaction may play a significant role in the haemodynamic consequences of acute pulmonary hypertension.

Hemodynamic effects of different lung-protective ventilation strategies in closed-chest pigs with normal lungs.

OBJECTIVE: The benefits of lung-protective ventilation strategies used for acute respiratory distress syndrome in subjects with normal lungs are uncertain. The purpose of this study was to investigate the hemodynamic effects of conventional lung-protective ventilation (CLPV) and high-frequency oscillatory ventilation (HFOV) in a normal lung animal model. DESIGN: Prospective laboratory investigation. SETTING: Animal laboratory in a university medical center. SUBJECTS: Seven landrace pigs (mean weight 41 kg). INTERVENTIONS: Pigs were ventilated at random conventionally with positive end-expiratory pressure 2-3 cm H2O and tidal volume 10-12 mL/kg (control), with CLPV (positive end-expiratory pressure 10 cm H2O, tidal volume 6 mL/kg), or with HFOV. Hemodynamics were analyzed after insertion of biventricular conductance catheters and a pulmonary artery catheter. MEASUREMENTS AND MAIN RESULTS: The protective strategies led to higher mean airway pressures and severe hypercapnia with acidosis, which was only significant with CLPV. Compared with control, oxygenation was worse with CLPV and HFOV. With HFOV and CLPV, mean arterial pressure, cardiac output, and stroke volume decreased significantly; pulmonary arterial elastance increased. The slope of the end-diastolic pressure volume relationship for the left and right ventricle remained unchanged (preserved ventricular function), whereas the intercept increased with both protective strategies (augmented intrathoracic pressure); left and right end-diastolic volumes decreased significantly. CONCLUSIONS: In the absence of a fluid resuscitation strategy, CLPV and HFOV caused decreased mean arterial pressure, cardiac output, and stroke volume and worsened oxygenation in this normal lung animal model. This resulted primarily from a biventricular decrease in preload.

Oesophageal Doppler monitoring overestimates cardiac output during lumbar epidural anaesthesia.

Oesophageal Doppler monitoring (ODM) has been advocated as a non-invasive means of measuring cardiac output (CO). However, its reliance upon blood flow measurement in the descending aorta to estimate CO is susceptible to error if blood flow is redistributed between the upper and lower body. We hypothesize that lumbar epidural anesthesia (LEA), which causes blood flow redistribution, causes errors in CO estimates. We compared ODM with thermodilution (TD) measurements in fourteen patients under general anaesthesia for radical prostatectomy, who had received an epidural catheter at the intervertebral level L2-L3. Coupled measurements of CO by means of the TD and ODM techniques were performed at baseline (general anaesthetic only) and after epidural administration of 10 ml of 0.25% bupivacaine. The two methods were compared using Bland-Altman analysis: before LEA there was a bias of -0.89 litre min(-1) with limits of agreement ranging between -2.67 and +0.88 litre min(-1). Following lumbar sympathetic block, bias became positive (+0.55 litre min(-1)) and limits of agreement increased to -3.21 and +4.30 litre min(-1). ODM measured a greater increase in CO after LEA (delta=+1.71 (1.19) litre min(-1) (mean (SD)) compared with TD (delta=+0.51 (0.70) litre min(-1)). We conclude that following LEA, measurements with the Oesophageal Doppler Monitor II overestimate CO and show unacceptably high variability. Blood flow redistribution may limit the value of ODM.

Preload-adjusted maximal power of right ventricle: contribution of end-systolic P-V relation intercept.

To assess whether preload-adjusted maximal power (PAMP), which is calculated as W(max)/V (where W(max) is maximal power and V(ed) is end-diastolic volume with beta = 2) is an index of right ventricular (RV) contractility, we measured RV pressure (P) and volume (V) and pulmonary artery pressure and flow in 10 dogs at baseline and after inotropic stimulation. PAMP was derived from steady-state data, whereas the slope (E(es)) and intercept (V(d)) of the end-systolic P-V relationship were derived from data obtained during vena caval occlusion. Inotropic stimulation increased E(es) (from 0.96 +/- 0.25 to 1.62 +/- 0.28 mmHg/ml; P < 0.001) and V(d) (from -3.0 +/- 17.2 to 12.4 +/- 10.8 ml; P < 0.05) but not PAMP (from 0.24 +/- 0.10 to 0.36 +/- 0.22 mW/ml(2); P = 0.09). We found a strong relationship between the optimal beta-factor for preload adjustment and V(d). A corrected PAMP, PAMP(c) = W(max)/(V(ed) - V(d))(2), which incorporated the V(d) dependency, was sensitive to the inotropic changes (from 0.23 +/- 0.12 to 0.54 +/- 0.17 mW/ml(2); P < 0.001) with a good correlation with E(es) (r = 0.88; P < 0.001).

Effects of propofol on the systolic and diastolic performance of the postischaemic, reperfused myocardium in rabbits.

BACKGROUND AND OBJECTIVE: The effect of propofol on myocardial dysfunction during ischaemia and reperfusion is controversial yet important because of its frequent use in cardiac anaesthesia. Although animal studies suggest a free radical-scavenging potential, the cardioprotective properties of propofol have not been demonstrated consistently in vivo. Previous studies focused on systolic function while diastolic function may be a more sensitive marker of ischaemic injury. The main aim was to document the effect of propofol on diastolic function in isolated, blood perfused rabbit hearts subjected to moderate global ischaemia and reperfusion. METHODS: Propofol 168 micromol L(-1), or the equivalent of its vehicle, Intralipid, was administered to 34 paced parabiotic Langendorff blood-perfused isolated rabbit hearts before and after 30 min of global normothermic ischaemia. Recovery of systolic function was quantified with the maximum rate of rise of left ventricular pressure. Diastolic performance was assessed using the time constant of the decline in left ventricular pressure (tau) and chamber stiffness (VdP/dV at 12 mmHg). RESULTS: Recovery of systolic function during reperfusion was comparable in the two groups. There was no difference in left ventricular pressure between the two groups at any time during the experiments. Chamber stiffness increased significantly during ischaemia and reperfusion in the control group (from 34 +/- 9 to 54 +/- 8 mmHg during ischaemia, and 43 +/- 5 mmHg after 30 min reperfusion; mean +/-95% confidence interval) but not in the propofol-treated group (29 +/- 5, 36 +/- 8 and 30 +/- 8 at baseline, ischaemia and 30 min reperfusion, respectively). CONCLUSIONS: Propofol has no protective effect on active relaxation or on systolic function in the present model, but it reduces ischaemic and postischaemic chamber stiffness.

GM-CSF versus G-CSF: engraftment characteristics, resource utilization, and cost following autologous PBSC transplantation.

BACKGROUND: G-CSF and GM-CSF have both been shown to decrease the time to hematopoietic recovery when administered after autologous BM or peripheral stem cell re-infusion. However, few studies have compared G-CSF and GM-CSF to determine which is the preferred myeloid growth factor. METHODS: This study compares a prospectively accrued cohort of 22 patients receiving GM-CSF with a historical cohort of patients who received G-CSF commencing Day + 6 after autologous PBSC transplantation. Patients were matched based on disease type and stage, CD34(+) cell dose/kg, conditioning regimen, and prior treatment. Time to myeloid engraftment, growth factor utilization, antibiotic utilization, fever incidence, and cost were compared. RESULTS: The median time to neutrophil and platelet engraftment was similar in the two groups (ANC > 500 /mm(3), GM-CSF 12 versus G-CSF 11, P = 0.69). There was a trend towards more days of temperature > 38.0 masculine C (six versus three, P = 0.05) and febrile neutropenia (three versus two, P = 0.06) in the GM-CSF arm. There was a trend towards increased use of i.v. antibiotics in the GM-CSF cohort (7.6 days versus 5.5 days, P = 0.06). More chest X-rays (1.5 versus 1.0, P = 0.03) were ordered, and more blood cultures drawn (4.2 versus 2.7, P = 0.05) as part of fever evaluation in the group treated with GM-CSF. Resource utilization based on actual wholesale pricing (AWP) favored the G-CSF cohort. Applying a sensitivity analysis, GM-CSF became cost-effective when priced below $94 per 250 micro g, despite greater resource utilization. DISCUSSION: This study suggests that engraftment characteristics are similar with GM-CSF and G-CSF following PBSC transplantation. Resource utilization for fever treatment and evaluation may be greater with GM-CSF. Determination of which agent is more cost-effective depends on institutional acquisition costs.