Cytomegalovirus

Objective: To determine the prevalence, associated findings, and consequences of cytomegalovirus (CMV) antigenemia in critically ill patients.

Design: A retrospective, case-control clinical study.

Setting: A 12-bed university hospital medical-surgical ICU.

Patients: Two hundred thirty-seven patients with fever for > 72 h, without proven evidence of bacteriologic and/or fungal origin, and whose pp65 antigenemia assays were studied. Patients with HIV infection and transplant recipients were excluded.

Interventions: None.

Measurements and results: CMV antigenemia was diagnosed within 20 [+ or -] 12 days (mean [+ or -] SD) after ICU admission in 17% patients in whom the pathology was suspected. The 40 patients in the CMV group were matched with 40 other patients in the control group. CMV infection was linked to renal failure (58% vs 33%, respectively; p = 0.02) and steroid use (55% vs 33%, respectively; p = 0.04). Patients with CMV had a significantly longer stay in the ICU (41 [+ or -] 28 days vs 31 [+ or -] 22 days, respectively; p = 0.04), a longer duration of mechanical ventilation (35 [+ or -] 27 days vs 24 [+ or -] 20 days, respectively; p = 0.03), a higher rate of nosocomial infection (75% vs 50%, respectively; p = 0.04), and a higher mortality (50% vs 28%, p = 0.02).

Conclusions: CMV antigenemia is not an uncommon diagnosis in critically ill ICU patients with unexplained prolonged fever after 10 days of hospitalization, regardless of their immune system status. Although associated with a higher morbidity and mortality, the clinical significance of CMV is unknown. Further prospective studies should evaluate the impact on ICU outcome and whether CMV is truly a pathogen or simply another indicator of immunosuppression.

Key words: critical illness; cytomegalovirus disease; ICU; mechanical ventilation; prolonged fever; sepsis

Abbreviations: CMV = cytomegalovirus; PCR = polymerase chain reaction; SAPS = simplified acute physiology score

**********

Cytomegalovirus (CMV) infection is an important cause of morbidity and mortality among immunosuppressed patients in contexts such as organ transplantation, malignant hematologic disease, and AIDS. Infection with CMV is usually asymptomatic in the immunocompetent host, (l,2) or merely causes a viral-like syndrome with fever. After a primary infection, which usually occurs during childhood, (3) CMV (as with all herpes viruses) resides in the host throughout life. Therefore, 60 to 100% of all individuals have positive CMV serology results by adulthood, depending on geographic and social factors. (1) CMV persists in a latent state in the nucleus of polymorphonuclear cells, and could therefore contaminate blood products and transplants. (l,4,5) The reactivation of CMV is a consequence of cell-mediated impaired immunity (transplantation patients with immunosuppressive treatment or patients with AIDS). (6,7) The manifestations of CMV infection in immunosuppressed patients range from asymptomatic virus shedding to severe organ disease. Patients who have not been treated with immunosuppressive drugs and who are HIV-negative are usually not considered immunocompromised, and are therefore not at high risk for CMV infection.

However, some reactivations have been observed in other populations, for example, after trauma, (8) as well as in patients with cirrhosis. (9) Critically ill patients frequently demonstrate profound immunity abnormalities as a result of either their illness or its treatment. (10) In the ICU, systemic inflammatory response syndrome develops in a high number of admitted patients, and shock develops in some of them, which induces the release of catecholamines and glucocorticoids into the systemic circulation. (11) During systemic inflammatory response syndrome and the compensatory anti-inflammatory response syndrome, the immune function may be impaired. (12-15) In such a way, CMV reinfection or reactivation could be an important indicator of the "ICU-acquired immunosuppression." (15) Moreover, some authors (10,16,17) have shown that tumor necrosis factor-[alpha], which is one of the most important central mediators in the pathophysiology of sepsis, can induce activation of CMV in septic patients. Therefore, it may be reasonable to consider CMV infection in the differential diagnosis of critically ill patients with prolonged hospital stay, multiple organ dysfunction, and unexplained prolonged fever.

However, clinical trials investigating the frequency of CMV infection and/or antigenemia in ICU patients have not shown uniform results (Table 1). Some studies (10,18-20) report an incidence ranging from 25 to 35%, whereas others (21-24) found very low or no CMV infection. These discrepancies may be explained by differences in study design, patient populations, methods, and delay of CMV infection diagnosis. Moreover, the clinical significance of the CMV infection is unknown, and its influence of morbidity and mortality is still debated.

In our transplantation ICU, after becoming aware of the CMV pathologies in patients receiving liver transplantation, we investigated symptoms of CMV infection in the nonimmunosuppressed patients on admission. An analysis of these results aimed to identify prevalence, associated factors, and consequences of CMV antigenemia in patients who were not treated with immunosuppressive drugs, who were HIV-negative, and who were not considered immunocompromised on ICU admission.

MATERIALS AND METHODS

The study protocol fulfilled the ethical standards of the French Society of Intensive Care Medicine. Because of the noninterventional nature of our study, ethical approval and patient consent were judged unnecessary. This retrospective, ease-control study of a matched-paired design was conducted in a medical-surgical and transplantation unit comprised of 12 beds at the Montpellier University Hospital, Saint-Eloi (France). The study period lasted 6.5 years, from January 1, 199.5 to June 30, 2001.

We selected HIV-negative, immunocompetent patients who had not been treated with immunosuppressive drags and who underwent at least one antigenemia assay during their ICU stay. Patients are tested for CMV based on the clinical judgement of the attending physician.

The diagnosis of CMV antigenemia was defined by a positive CMV pp65 antigenemia assay result. (25-27) The high sensitivity and specificity of this technique for diagnosing active CMV infection were demonstrated previously. (25-27) Patients were excluded from the study for any of the following reasons: AIDS, pregnancy, organ or bone marrow transplantation, immunosuppressive therapy, long-term treatment with corticosteroids (> 3 months), cancer, or hematologic diseases with previous anticancer radiotherapy or chemotherapy.

The matching procedure was identical to that used in previous studies. (28,29) Patients with a diagnosis of CMV antigenemia (CMV group) were matched to patients in whom the diagnosis was suspected but not confirmed by antigenemia assay (control group). The two groups were matched according to the following five criteria: age [+ or -] 10 years, gender, simplified acute physiology score (SAPS) II of [+ or -] 7, (26) admission date in ICU [+ or -] 12 months, and type of admission (critical postoperative, postoperative complications, medical, trauma). We obtained two identical cohort groups (CMV and control); each pair was matched for at least four of the predefined criteria (see above).

Analyzed Parameters

On Admission to the ICU: The following parameters were determined for matching pairs: existence and duration of previous ICU admission, neoplasia, and sepsis defined according to the criteria of Bone (13) (ie, the presence of at least two criteria from among body temperature > 38[degrees]C or < 36[degrees]C, heart rate > 90 beats/min, respiratory rate > 25 breaths/rain or receiving mechanical ventilation, infectious focus proven by bacteriologic screening, bacteremia, and WBC count > 12,000/ [micro]L or < 4,000/[micro]L).

During ICU Stay: The following parameters were determined: mechanical ventilation, renal failure defined as creatininemia > 160 mmol/L, need for dialysis, use of vasoactive amines, treatment by steroids for at least 48 h before the first pp65 antigenemia, type of nutrition (enteral, parenteral), and transfusion of blood products once admitted (RBCs, platelets, fresh-frozen plasma). The level of hemoglobin (grams per deciliter), leukocyte count (per milliliter), platelet count (per milliliter), liver enzymes (aspartate aminotransferase, alanine aminotransferase, and lactate dehydrogenase) were collected until a positive CMV antigenemia assay result was found in the CMV group and until the last negative antigenemia assay result in the control group, aiming to determine the likelihood of eventual predictability of CMV occurrence.

On Discharge From the ICU or the Hospital: The following parameters were determined: the incidence of nosocomial infection, expressed as the ratio of nosocomial infections to the number of days exposed to the risk, and to the number of patients having a nosocomial infection (ventilator-acquired pneumonia); at least one organism isolated by BAL at a concentration [greater than or equal to] 10 (4) cfu/mL; colonization of central venous catheters (at least one organism at a concentration [greater than or equal to] 10 (3) cfu/mL identified by a culture of the catheter tip via the Brun-Buisson technique (27); urinary catheter-related infection (the association of a leukocyturia at a concentration of [greater than or equal to] 10 (4) mL with the presence of a organism at a concentration of 10 (5) cfu/mL); and bacteremia (a positive hemoculture finding with the isolation of a organism, or at least two positive hemoculture findings for a coagulase-negative Staphylococcus); the duration of mechanical ventilation; ICU and hospital length of stay; and ICU and hospital mortality.

Statistical Analysis: The results are expressed as mean [+ or -] SD, or number of positive results. The results of the CMV group were compared to those of the control group using the Student t test for quantitative Gaussian variables, the Wilcoxon rank-sum test for the non-Gaussian variables, and the MacNemar test for qualitative variables. The Symutrie test was used for variables consisting of several subcategories (cause of admission, type of nutrition). A multivariate analysis (logistic regression) was performed in a stepwise manner with the aim of searching for risk factors linked to CMV antigenemia and to account for confounding factors. The variables having a level of significance < 0.2 in the univariate analysis were included in the model. The validity of the model was evaluated by the degree of concordance and the Hosmer and Lemeshow test. We also performed a multivariate analysis on the entire group tested for CMV to determine whether CMV could be identified as an independent risk factor for mortality. The level of significance was set at 0.05. The statistical analysis was performed by the medical statistical department of the Montpellier University Hospital with the help of statistical software (SAS, version 6.12; SAS Institute; Cary, NC).

RESULTS

A total of 2,016 admissions were recorded during the study period. At least one pp65 antigenemia assay was performed in 237 patients who had not been treated with immunosuppressive drugs and who were HIV-negative, of which 40 patients formed the CMV group. Among the 197 patients with negative CMV antigenemia assay results, 40 patients were selected using the matching-pairing method to form the control group. Table 2 summarizes the effectiveness of matching, patient by patient; 92% (184 of 200 criteria) were correctly matched. Each pair was correctly matched for at least four of the five criteria. No significant difference was observed in the number of CMV antigenemia assays performed prior to obtaining the first positive antigenemia assay result for the CMV group and during ICU stay for the control group, as shown in Figure 1.

[FIGURE 1 OMITTED]

The main characteristics of patients from each group are illustrated in Table 3. The rates of renal failure and steroid therapy were significantly higher for the CMV group than the control group (Table 3).

The precise cause of admissions to ICU are detailed in Table 4. Transfusion of blood products is described in Table 5. The results of biological parameters studied in the two groups are shown in Table 6. As shown in Table 7, the number of patients with a nosocomial infection was 'also significantly higher in the CMV group than in the control group. The duration of mechanical ventilation and the ICU length of stay were also significantly longer in the CMV group than the control group (Table 7). We detected CMV only in blood, and we did not detect it in any other organ, ie, in the CNS or in the lungs, although we performed 28 BAL procedures in the 40 CMV patients.

The mean delay from admission to diagnosis of CMV infection was 20 [+ or -] 12 days after direct ICU admission, and 30 [+ or -] 16 days for CMV group patients hospitalized before being admitted to the ICU. Figure 2 shows the delay of hospital admission to diagnosis of CMV antigenemia expressed in the range of 10 days. The majority of patients (n = 30) acquired CMV antigenemia after 20 days of hospitalization. In terms of biological and clinical variables and outcomes, no significant differences were observed between patients who had CMV antigenemia early during hospitalization (before 20 days) and those who acquired late CMV antigenemia (after 20 days) [data not shown].

[FIGURE 2 OMITTED]

No independent parameters were linked to the outcome of CMV antigenemia in the logistical regression tests. However, a strong relationship was observed with renal failure (odds ratio, 2.50; p = 0.066) and corticosteroid therapy (odds ratio, 3.61; p = 0.082). In the multivariate analysis, CMV could not be identified as an independent risk factor for mortality on the entire group tested for CMV for different model of logistic regression tested. The mortality rate was significantly higher in the CMV group than in the control group (50% vs 28%, p = 0.02).

DISCUSSION

The main results of this study are that CMV antigenemia is not an uncommon pathology in selected patients who have not been treated with immunosuppressive drugs and who were HIV-negative on admission to the ICU, and that it is associated with higher morbidity and mortality (Table 7). In addition, in unexplained prolonged fever in severely ill ICU patients, CMV antigenemia could be present.

CMV antigenemia can also be encountered among patients admitted for medical as well as surgical causes. The CMV group had a ratio of two surgical patients to one medical patient.

To our knowledge, this is the largest series reporting the association of CMV antigenemia with higher morbidity and mortality in critically ill patients considered "immunocompetent" on their admission to ICU (Table 1). This pathology is somewhat unknown and insufficiently diagnosed in the context of unexplained prolonged fever in ICU patients. Indeed, the significance of CMV detection in critically ill intensive care patients is unknown. The differentiation between CMV detection and CMV disease represents a difficult diagnosis dilemma. However, our study and others (Table 1) demonstrated that patients with detectable CMV tend to have a higher morbidity and mortality compared with patients in whom the virus remained undetectable.

CMV patients were severely ill in the ICU with a mean SAPS II of 51 [+ or -] 16, 73% of which had a SAPS II > 40 (Table 3). The average delay to diagnosis of CMV antigenemia was 20 [+ or -] 12 days after admission to ICU. Figure 2 shows that the pathology appears more often after a delay of 10 days after hospital admission. Renal failure and corticosteroid therapy were significantly associated with CMV infection in the univariate analysis, as was dialysis and the amount of blood transfused (Table 3). The role of CMV in patients with renal failure or dysfunction and those receiving dialysis is well established. (2,30,31) Moreover, some studies (31-33) reported the clinical benefits of antiviral agents and polyclonal pooled enriched CMV IgG. Blood transfusion is considered a contamination risk when blood products were not leukocyte depleted. In our practice, all blood products used were leukocyte depleted. But this process usually does not eliminate all leukocytes, and few leukocytes are required for a CMV antigenemia to be positive. However, no independent risk factor was found in the multivariate analysis. Perhaps a multivariate analysis with a greater number of patients would have allowed identification of independent risk factors for the development of CMV antigenemia.

A significantly longer duration of mechanical ventilation was found for CMV patients, as well as a significantly higher rate of nosocomial infection. A significantly higher mortality rate was obtained in CMV patients (50% vs 28%, respectively, p = 0.02). Death always occurred in the ICU.

Excessive morbidity and mortality was observed despite SAPS II matching. Our study cannot in any way prove that the cause of higher mortality is due to CMV antigenemia, but it strongly underlines the association between the two. Further studies are needed to determine whether there is a causal relationship or simply a correlation between CMV and mortality.

Clinical trims examining the frequency of CMV infection in ICU patients have not shown uniform results (Table 1). Some studies (11,19,20,22) report an incidence ranging from 25 to 35%, while others (21,34-36) indicate very low or no CMV infection. These discrepancies may be explained by study design differences, patient groups, and CMV infection diagnosis methods and timing. The studies that failed to find CMV infection (Table 1) often did not use specific inclusion criteria (21,35,36) and/or a small number of cohorts. (21) Moreover, these trials (21,35,36) performed the CMV research in the first days of hospitalization in the ICU (2 to 5 days after admission), contrary to the studies (19,20,22) reporting CMV infection that researched CMV later (> 15 days) in the course of ICU stay. Cook et ala (34) included in their retrospective study a large number of patients with specific criteria and researched CMV late in the course of ICU stay (Table 1), but reported a lower prevalence of CMV infection (8%) than the other studies (19,20,22) and our present study (17 to 35%). This could be explained in part by the fact that the method of diagnosis used (viral culture) was less sensitive than the others (polymerase chain reaction [PCR] and/or antigenemia) [Table 1].

In a prospective study, Heininger et al (22) assessed the occurrence of CMV infection and evaluated potential risk factors in immunocompetent ICU patients after major surgery or trauma. Patients were screened weekly for CMV infection using PCR method. Infection was found in 20 patients (36% of the population studied). Patients with infection were compared to 36 patients in whom viral screening proved negative. A significant correlation was found between sepsis on admission, neoplasia, and CMV infection. CMV IgG seropositivity on admission in ICU permitted the exclusion of primary infection. Heininger et a1 (22) only investigated surgical patients with CMV infection, contrary to our study, which concerned surgical-medical patients acquiring CMV infection. The study by Heininger et a1 (22) and our study complement each other in that they illustrate the relation between CMV antigenemia with a severe pathology on admission. Contrary to the study by Heininger et al, (22) we did not find sepsis and malignant disease as predictive factors for CMV disease, but it must be noted that matching on a gravity score tends to erase this tendency. However, we had a cohort group with a high incidence of sepsis (63%). In the studies by Heininger et a1 (22) and Papazian et al, (20) the incidences of malignant disease in the CMV group were 40% and 25%, respectively. These results were comparable to our present study (25%). CMV is well known to be an important cause of morbidity and mortality in AIDS or immunosuppressed organ transplantation recipients, but the pathogenicity of CMV in septic patients or cancer patients remains unclear. The study by Heininger et a1 (22) reported that patients with active infection tended to have a higher mortality rate and required significantly longer ICU treatment, similar to the findings of our present study.

The relation between CMV infection and sepsis is complex, and tumor necrosis factor-[alpha], which is a central mediator of sepsis, may influence CMV reactivation. (11,16-18) We found a significant higher rate of nosocomial infections in the CMV group (Table 7), which correlates with the prospective longitudinal study of 34 septic ICU patients by Kutza et al. (11) These authors (11) observed an incidence of CMV infection of 32.4% in a median of 4 days when diagnosed using PCR, or 11 days when diagnosed using pp65 antigenemia assay.

Domart et al (19) diagnosed CMV infection in 23 patients with mediastinitis among 115 immunocompetent individuals following cardiac surgery. Only 8 of these 23 patients were seropositive for CMV. The authors (19) found that the perioperative blood transfusion was the main factor responsible for CMV seroconversion in the absence of leukocyte-depleted blood.

Papazian et al (20) demonstrated the possibility of organ damage due to CMV in immunocompetent ICU patients. Of 86 patients receiving mechanical ventilation with ARDS or signs of nonproven chest infection, 25 patients showed specific histologic signs of CMV pneumonia on lung biopsy or autopsy. The average delay between admission and diagnosis was 22 [+ or -] 9 days. The results of our study are consistent with those reported by Papazian et al. (20)

Our study certainly has limitations. We do not know the serologic CMV status of our patients before admission to the ICU. Positive serology results eliminate primary infection, nonspecific to immunosuppressed patients. (2) This limitation is notable, given that 60 to 80% of adults aged [greater than or equal to] 40 years are CMV-seropositive. Furthermore, the diagnostic value of CMV serology is limited by the presence of false-positive diagnosis due to the passive transfer of anti-CMV antibodies during blood transfusions. Because we did not have the serologic CMV status of patients before admission to ICU, it is difficult to distinguish a primary CMV infection and a recurrent infection resulting from reactivation of latent virus (endogenous) or reinfection (exogenous).

We have defined and classified different parameters observed, based on analyzed data from previous studies and knowledge related to CMV infection. We have therefore analyzed the transfusion of blood products as a risk factor for CMV infection, while it also could be a factor for comorbidity. We found an incidence of 17% of CMV antigenemia among patients in whom the pathology was suspected. This number is subject to scrutiny, as CMV antigenemia assay was not systematically performed, and it represents only 2.3% of the patients who were immunocompetent on their admission to the ICU. Indeed, CMV antigen testing was only performed in a selected group of patients, presumably those in whom there was a high index of suspicion. A prospective study should be performed to address these aspects. Another limitation of our study is related to the antiviral treatment of patients with CMV. In fact, 30 of 40 patients had treatment with IV ganciclovir for 2 to 3 weeks. The other 10 patients were not treated with ganciclovir because of either clinical improvement or rapid death. However, there was no significant difference in morbidity and mortality between those treated and not treated: 57% (17 of 30 patients) vs 30% (3 of 10 patients) [p = 0.27].

A complex interaction exists between CMV and the immune system. Even though we could evoke the existence of an acquired immunosuppression during ICU stay (13,14,37,38) in the CMV patient group, it is impossible for us to determine whether CMV is responsible for or the consequence of immunosuppression. Only prospective studies with immunologic exploration can determine this. Moreover, the clinical significance of the viral infections, especially CMV infection, is unknown and remains debated, although CMV patients have significantly higher morbidity and mortality rates than the control group. However, there is actually no information on the management of CMV infection or disease in immunocompetent patients. Proving causality in CMV antigenemia can be particularly difficult. Further studies are needed to define the pathogenicity of CMV, and to evaluate the effect of preemptive antiviral treatment on morbidity and mortality.

In conclusion, our study shows that CMV antigenemia is not an uncommon diagnosis in critically ill ICU patients with unexplained prolonged fever after 10 days of hospitalization, regardless of immune status. The acquisition of CMV antigenemia seems to be a marker of severity of illness in ICU patients. Its clinical significance is unknown, although it is associated with a higher morbidity and mortality. Further prospective studies are needed to evaluate CMV influence on ICU outcome and whether it is truly a pathogen or simply another indicator of immunosuppression.

REFERENCES

(1) Threlkeld M, Cobbs C. In: Mandell GL, Bennett J, Dolin R, eds. Principles and practice of infectious diseases. New York, NY: Churchill Livingstone, 1995; 1568-1599

(2) Sweet C. The pathogenicity of cytomegalovirus. FEMS Microbiol Rev 1999; 23:457-482

(3) Hamprecht K, Maschmann J, Vochem M, et al. Epidemiology of transmission of cytomegalovirus from mother to preterm infant by breastfeeding. Lancet 2001; 357:513-518

(4) Schrier R, Nelson J, Oldstone M. Detection of human cytomegalovirus in peripheral blood lymphocytes in a natural infection. Science 1985; 230:1048-1051

(5) Taylor-Wiederman J, Sissons J, Borysiewicz L, et al. Monocytes are a major site of persistence of human cytomegalovirus in peripheral blood mononuclear cells. J Gen Virol 1991; 72:2059-2064

(6) Son J, The T. Cytomegalovirus infection after organ transplantation: an update with special emphasis on renal transplantation. Transplant Int 1989; 2:147-164

(7) Jacobson M, Mills J. Serious cytomegalovirus disease in the acquired immunodeficiency syndrome (AIDS). Ann Intern Med 1988; 108:585-594

(8) de Jong M, Galasso G, Gazzard B, et al. Summary of the II International Symposium on Cytomegalovirus. Antiviral Res 1998; 39:141-162

(9) Heininger A, Vogel U, Aepinus C, et al. Fatal human cytomegalovirus disease after severe multiple trauma. Crit Care Med 2000; 28:563-566

(10) Tanaka S, Toh Y, Minagawa H, et al. Reactivation of cytomegalovirus in patients with cirrhosis: analysis of 122 cases. Hepatology 1992; 16:1409-1414

(11) Kutza A, Muhl E, Hackstein H, et al. High incidence of active cytomegalovirus infection among septic patients. Clin Infect Dis 1998; 26:1076-1082

(12) Tamion F, Richard V, Lyoumi S. Gut ischemia and mesenteric synthesis of inflammatory cytokines after hemorrhagic or endotoxic shock. Am J Physiol 1997; 273:G314-G321

(13) Bone R. The pathogenesis of sepsis. Ann Intern Med 1991; 115:457-469

(14) Bone R. Immunologic dissonance: a continuing evolution in our understanding of the systemic inflammatory response syndrome (SIRS) and the multiple organ dysfunction syndrome (MODS). Ann Intern Med 1996; 125:680-687

(15) Bone R. Sir Isaac Newton, sepsis, SIRS, and CARS. Crit Care Med 1996; 24:1125-1128

(16) Ho M. Cytomegalovirus infection in patients with bacterial sepsis. Clin Infect Dis 1998; 26:1083-1084

(17) Smith P, Saini S, Raffeld M, et al. Cytomegalovirus induction of tumor necrosis factor-[alpha] by human monocytes and mucosal macrophages. J Clin Invest 1992; 90:1642-1648

(18) Docke W, Prosch S, Fietze E. Cytomegalovirus reactivation and tumor necrosis factor. Lancet 1994; 343:268-269

(19) Domart Y, Trouillet J, Fagon J, et al. Incidence and morbidity of cytomegaloviral infection in patients with mediastinitis following cardiac surgery. Chest 1990; 97:18-22

(20) Papazian L, Fraisse A, Garbe L, et al. Cytomegalovirus: an unexpected cause of ventilator-associated pneumonia. Anesthesiology 1996; 84:280-287

(21) Stephan F, Meharzi D, Ricci S, et al. Evaluation by polymerase chain reaction of cytomegalovirus reactivation in intensive care patients under mechanical ventilation. Intensive Care Med 1996; 22:1244-1249

(22) Heininger A, Jahn G, Engel C, et al. Human cytomegalovirus infections in nonimmunosuppressed critically ill patients. Crit Care Med 2001; 29:541-547

(23) Kusne S, Grossi P, Irish W, et al. Cytomegalovirus PP65 antigenemia monitoring as a guide for preemptive therapy: a cost effective strategy for prevention of cytomegalovirus disease in adult liver transplant recipients. Transplantation 1999; 68:1125-1131

(24) Bein G, Bitsch A, Hoyer J, et al. The detection of human cytomegalovirus immediate early antigen in peripheral blood leukocytes. J Immunol Methods 1991; 137:175-180

(25) Carton JA, Rodriguez-Guardado A, Melon S, et al. Cytomegalovirus antigenemia surveillance in the treatment of cytomegalovirus disease in AIDS patients. J Chemother 1999; 11:195-202

(26) Le Gall JR, Lemeshow S, Saulnier F. A new simplified acute physiology score (SAPS II) based on a European/North American multicenter study. JAMA 1993; 270:2957-2963

(27) Brun-Buisson C, Abrouk F, Legrand P, et al. Diagnosis of central venous catheter-related sepsis: critical level of quantitative tip cultures. Arch Intern Med 1987; 147:873-877

(28) Atkins PM, Mion LC, Mendelson W, et al. Characteristics and outcomes of patients who self-extubate from ventilatory support: a case-control study. Chest 1997; 112:1317-1323

(29) Bregeon F, Ciais V, Carret V, et al. Is ventilator-associated pneumonia an independent risk factor for death? Anesthesiology 2001; 94:554-560

(30) Ljungman P, Griffiths P, Paya C. Definitions of cytomegalovirus infection and disease in transplant recipients. Clin Infect Dis 2002; 34:1094-1097

(31) Morales E, Andres A, Gonzalez E, et al. Prophylaxis of cytomegalovirus disease with ganciclovir or anti-CMV immunoglobulin in renal transplant recipients who receive antilymphocytic antibodies as induction therapy. Transplant Proc 2002; 34:73-74

(32) Snydman D, Werner B, Heinze-Lacey B, et al. Use of cytomegalovirus immune globulin to prevent cytomegalovirus disease in renal-transplant recipients. N Engl J Med 1987; 22:1049-1054

(33) Snydman D, Falagas M, Avery R, et al. Use of combination cytomegalovirus immune globulin plus ganciclovir for prophylaxis in CMV-seronegative liver transplant recipients of a CMV-seropositive donor organ: a multicenter, open-label study. Transplant Proc 2001; 33:2571-2575

(34) Cook CH, Yenchar JK, Kraner TO, et al. Occult herpes family viruses may increase mortality in critically ill surgical patients. Am J Surg 1998; 176:357-360

(35) Desachy A, Ranger-Rogez S, Francois B, et al. Reactivation of human herpesvirus type 6 in multiple organ failure syndrome. Clin Infect Dis 2001; 32:197-203

(36) Razonable RR, Fanning C, Brown RA, et al. Selective reactivation of human herpesvirus 6 variant a occurs in critically ill immunocompetent hosts. J Infect Dis 2002; 185:110-113

(37) Faist E, Schinkel C, Zimmer S. Update on the mechanism of immune suppression of injury and immune modulation. World J Surg 1996; 20:454-459

(38) Zimmerli W. Impaired host defence mechanisms in intensive care unit patients. Intensive Care Med 1985; 11:174-178

* From the Intensive Care Unit and Transplantation Department (Drs. Jaber, Chanques, Borry, Souche, Perrigault, and Eledjam), Saint Eloi Hospital, University Hospital of Montpellier; and Department of Biostatistics (Dr. Verdier) Arnaud de Villeneuve Hospital, University Hospital of Montpellier, Montpellier, France.

Manuscript received January 16, 2004; revision accepted July 14, 2004.

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: permissions@chestnet.org).

Correspondence to: Samir Jaber, MD, PhD, Unite de Reanimation-Departement d'Anesthesie-Reanimation "B", University Hospital, Chu de Montpellier Hopital Saint Eloi 80 Ave Augustin Fliche, 34295 Montpellier Cedex 5, France; e-mail: s-jaber@chumontpellier.fr

COPYRIGHT 2005 American College of Chest Physicians
COPYRIGHT 2005 Gale Group