Feng Yee Chang

Methicillin-Resistant Staphylococcus aureus: The Other Emerging Resistant Gram-Positive Coccus among Liver Transplant Recipients

We undertook a study of the characteristics and clinical impact of infections due to methicillin-resistant Staphylococcus aureus (MRSA) after liver transplantation. Of 165 patients who received liver transplants at our institution from 1990 through 1998, 38 (23%) developed MRSA infections. The predominant sources of infection were vascular catheters (39%; n=15), wound (18%; n=7), abdomen (18%; n=7), and lung (13%; n=5). A significant increase in MRSA infections (as a percentage of transplant patients infected per year) occurred over time (P=.0001). This increase was greater among intensive care unit patients (P=.001) than among nonintensive care unit hospital patients (P=.17). Cytomegalovirus seronegativity (P=.01) and primary cytomegalovirus infection were significantly associated with MRSA infections (P=.005). Thirty-day mortality among patients with MRSA infections was 21% (8/38). Mortality was 86% in patients with bacteremic MRSA pneumonia or abdominal infection and 6% in those with catheter-related bacteremia (P=.004). Thus the incidence of MRSA infection has increased exponentially among our liver transplant recipients since 1990. These infections have unique risk factors, time of onset, and a significant difference in site-specific mortality; deep-seated bacteremic infections, in particular, portend a grave outcome.

Thrombocytopenia in liver transplant recipients: predictors, impact on fungal infections, and role of endogenous thrombopoietin.

BACKGROUND Thrombocytopenia is a frequent and potentially serious complication in liver transplant recipients. The role of endogenous thrombopoietin level in posttransplant thrombocytopenia, has not been fully defined in liver transplant recipients. Additionally, there is accumulating evidence to suggest that platelets play a important role in antimicrobial host defense. METHODS There were 50 consecutive liver transplant recipients studied. Variables predictive of thrombocytopenia, its impact on infectious morbidity and outcome, and serial thrombopoietin (TPO) serum concentration were assessed. RESULTS The median pretransplant platelet count was 67 x 10(3)/cmm. After the liver transplantation, the median nadir platelet count was 33 x 10(3)/cmm and was reached a mean of 6 days after the transplant. A lower pretransplant platelet count (r= +.068, P=.0001), lower serum albumin before the transplants (r=+0.39, P=.014), longer operation time (r=0.27, P=.05), higher intraoperative packed red cells (r=0.28, P=.049) and fresh frozen plasma transfusions (r=0.42, P=.004), higher bilirubin at Day 7 (r=-.386, P=.005), and higher serum creatinine at Day 7 after the transplants (r=-.031, P=.025) correlated significantly with a lower nadir in platelets after the transplant. Nadir in platelet count was significantly lower in nonsurvivors compared with survivors (16 vs. 36 x 10(3)/cmm, P=.0001). Forty-three percent (9 of 21) of the patients with nadir platelet counts of < or =30 x 10(3)/cmm had a major infection within 30 days of the transplant compared with 17% (5 of 29) with nadir platelet counts > 30 x 10(3)/cmm (P=.04). Fungal infections occurred in 14% of the patients with nadir platelet counts of < or =30 x 10(3)/cmm versus 0% in those with nadir platelet counts of > 30 x 10(3)/cmm (P=.06); all patients with fungal infections had nadir platelet counts of < or =30 x 10(3)/cmm before fungal infection. Nadir in platelet count preceded the first major infection by a median of 7 days. Pretransplant TPO level did not differ between survivors (mean 103 pg/ml) or nonsurvivors (mean 144 pg/ml). After the transplantation, TPO levels increased in both groups. TPO level peaked at Day 7 and subsequently declined in survivors. Nonsurvivors had persistent thrombocytopenia despite a progressive rise in TPO level; TPO level was significantly higher at Day 7 (P=.02), Day 9 (P=.0019), and Day 14 (P=.04) in nonsurvivors compared with survivors. CONCLUSION Persistent thrombocytopenia portended a poor outcome in liver transplant recipients and was not related to low TPO levels. Thrombocytopenia preceded infections and identified a subgroup of liver transplant patients susceptible to early major infections; its precise role in fungal infections warrants validation in larger studies.

Fever in Liver Transplant Recipients: Changing Spectrum of Etiologic Agents

Febrile episodes in liver transplants were prospectively evaluated. Fever was due to infections in 78% of the episodes (35 of 45) and due to noninfectious causes in 22% (10 of 45). The predominant sources of fever were bacterial infections (62%; 28 of 45) and viral infections (6%; 7 of 45), whereas rejection accounted for only 4% of the episodes (2 of 45). Forty percent of the infections were unaccompanied by fever; fungal infections were significantly less likely to be associated with fever than were viral or bacterial infections (P = .001). Eighty-six percent (6) of the 7 febrile viral infections were due to viruses other than cytomegalovirus, of which human herpesvirus-6 was the predominant pathogen (71%; 5 of 7). Eighty percent (four) of the five febrile episodes with leukopenia were due to human herpesvirus-6. Episodes of fever were most likely to occur within 12 weeks (58%) or 1 year (29%) after transplantation; 100% of the latter episodes were in patients with recurrent hepatitis due to hepatitis C virus, malignancy, or chronic hemodialysis. In conclusion, cytomegalovirus and rejection were no longer the predominant etiologies of fever in liver transplant recipients, and viruses other than cytomegalovirus (e.g., human herpesvirus-6) are emerging as a significant cause of febrile viral illnesses in these patients.

STAPHYLOCOCCUS AUREUS NASAL COLONIZATION AND ASSOCIATION WITH INFECTIONS IN LIVER TRANSPLANT RECIPIENTS

BACKGROUND Staphylococcus aureus has emerged as a leading cause of bacterial infections after liver transplantation. However, the role of nasal colonization in the development of S aureus infections has never been explored in liver transplant recipients. The objectives of this study were to determine whether nasal carriage of S aureus was a risk factor for S aureus infections in liver transplant recipients. METHODS Over a 2-year period, 30 consecutive liver transplant recipients were studied. Beginning when the recipients were transplant candidates, nasal cultures were performed at each admission and monthly thereafter until discharge or death. RESULTS Overall, 67% (20/30) of the patients were nasal carriers, 70% of the carriers had methicillin-resistant S aureus (MRSA), 15% had methicillin-sensitive S aureus, and 15% had both MRSA and methicillin-sensitive S aureus. Infections were significantly associated with the carrier state; 100% (9/9) of the infected patients were carriers as compared with 50% (11/21) of the noninfected patients (P=0.01). All infections were a result of MRSA, and 56% (5/9) of the infections were bacteremia. Median time to the onset of S aureus infections was 16 days after transplant. Pulse field gel electrophoresis (with digestion of S aureus with SmaI restriction enzyme) in seven infected patients demonstrated that the isolates from the anterior nares matched the invasive isolates in all cases. A total of 43% (3/7) of these infected patients shared the same restriction pattern. CONCLUSION MRSA colonization of the anterior nares was a significant predictor of MRSA infections in liver transplant recipients. Infections occurred only in those colonized with MRSA and were a result of the endogenously colonizing S aureus strains in all cases.

Fever in liver transplant recipients in the intensive care unit

Whether febrile illnesses in the intensive care unit (ICU) have unique spectrum, etiologies, and outcome has not been determined in liver transplant recipients. We studied 78 consecutive febrile patients over a 4‐yr period; 49% (38/78) were in the ICU and 51% (40/78) were in the non‐ICU setting. Of febrile patients in the ICU, 87% (33/38) had infection and 13% had non‐infectious etiology for fever. Seventy‐nine percent (26/33) of the infections associated with fever in the ICU were bacterial, 9% (3/33) were viral, and 9% (3/33) were fungal in etiology. Pneumonia (30%), catheter‐related bacteremia (15%), and biliary tree (9%) were the predominant sources of infections associated with fever in the ICU. Bacteremia was documented in 45% of the patients with fever in the ICU. Fifty‐three percent (20/38) of the febrile episodes in the ICU occurred during the initial post‐transplant stay, and 47% (18/38) during a subsequent readmission. Pneumonia accounted for 41% of all febrile infections during the first 7 d of ICU stay, but only 14% of those after 7 d. Febrile patients in the ICU had higher APACHE II scores (p=0.001), higher APS scores (p=0.0001), higher bilirubin (p=0.001), lower cholesterol (p=0.019), higher prothrombin time (p=0.001), were more tachycardiac (p=0.002), and were more likely to have abnormal blood pressure (p=0.001) than those in the non‐ICU setting. Twenty‐three percent of all infections in the ICU were unaccompanied by fever and 9% were accompanied by hypothermia. Mortality at 14 d (24 versus 0%, p=0.001) and at 30 d (34 versus 5%, p=0.001) was significantly higher in febrile patients in the ICU, as compared to the patients in the non‐ICU setting. These data have implications for diagnostic evaluation and management of critically ill febrile liver transplant recipients.

Predicting bacteremia and mortality in bacteremic liver transplant recipients

Predictors of bacteremia and mortality in bacteremic liver transplant recipients were prospectively assessed. One hundred eleven consecutive episodes of fever or infections were documented in 59 patients over a 4-year period. Forty-nine percent (29 of 59 patients) of the patients had bacteremia, 39% (23 of 59 patients) had nonbacteremic infections, and 12% (7 of 59 patients) had fever of noninfectious cause. Primary (catheter-related) bacteremia (31%; 9 of 29 patients), pneumonia (24%; 7 of 29 patients), abdominal and/or biliary infections (14%; 4 of 29 patients), and wound infections (10%; 3 of 29 patients) were the predominant sources of bacteremia. Diabetes mellitus (odds ratio, 6.9; P =.03) and serum albumin level less than 3.0 mg/dL (odds ratio, 0.14; P =.02) were independently significant predictors of bacteremia compared with nonbacteremic infections. Mortality at 14 days was 28% (8 of 29 patients) in those with bacteremia compared with 4% (1 of 23 patients) in those with nonbacteremic infections and 0% (0 of 7) in patients with fever of noninfectious cause (P =.03). Intensive care unit stay at the time of bacteremia (100% v 47%; P =.005), absence of chills (0% v 53%; P =.005), lower temperature at the onset of bacteremia (99.2 degrees F v 101.5 degrees F; P =.009), lower maximum temperature during the course of bacteremia (99.3 degrees F v 102 degrees F, P =.008), greater serum bilirubin level (7.6 v 1.5 mg/dL; P =.024), presence of abnormal blood pressure (80% v 16%; P =. 0013), and greater prothrombin time (15.6 v 13.3 seconds; P =.013) were significantly predictive of greater mortality in the bacteremic patients. These data have implications for discerning the likelihood of bacteremia and initiation of empiric antibiotics pending cultures. Lack of febrile response in bacteremic liver transplant recipients portended a poorer outcome.