- Email: [email protected]
Monocyte phagocytosis as a reliable parameter for predicting early-onset sepsis in very low birthweight infants
Early Human Development 67 (2002) 1 – 9 www.elsevier.com/locate/earlhumdev
Monocyte phagocytosis as a reliable parameter for predicting early-onset sepsis in very low birthweight infants $ Ulrike Hallwirth a,*, Gerhard Pomberger a, Daniela Zaknun b, Zsolt Szepfalusi b, Ernst Horcher a, Arnold Pollak b, Erich Roth c, Andreas Spittler c a
Division of Pediatric Surgery, Department of Surgery, University Medical School of Vienna, AKH Waehringer Guertel 18-20, 1090 Vienna, Austria b Department of Pediatrics, University Medical School of Vienna, Vienna, Austria c Surgical Research Laboratories, University Medical School of Vienna, Vienna, Austria Accepted 26 October 2001
Abstract Background: Septic complications lead to a high mortality rate in very low birthweight infants (VLBWI). Therefore, prognostic markers for the development of sepsis attach importance to start an efficient therapy as early as possible. Aims: Functional and phenotypical variables of blood monocytes in the cord and peripheral blood were investigated to evaluate the parameters for predicting early-onset and late-onset sepsis (nosocomial infections). Study design: In a prospective study, 25 VLBWI were investigated. Methods: In the cord blood taken immediately after birth, the capacity of the monocytes to phagocytose non-opsonized E. coli bacteria by flow cytometry and the ex-vivo production of tumor necrosis factor (TNF)-a, interleukin (IL)-1h and IL-6 (enzyme-linked immunoassay (ELISA)) after lipopolysaccharide (LPS) stimulation were measured. At the third day, the HLA-DR expression on the monocytes (flow cytometry) and the LPS-induced cytokine production were measured from the peripheral blood. Results: Five VLBWI already developed early septic complications after 24 – 72 h, while the other three infants had late-onset sepsis 10 – 18 days after birth. The prognostic significance for early-onset sepsis was highest for the decreased monocyte phagocytic capacity (sensitivity and specificity: 100%) and for the LPS-induced formation of TNF-a and IL-1h in cord blood. Moreover, in septic VLBWI, the HLA-DR expression on the monocytes was lowered on day 3 after birth. The prognostic significance for late-onset sepsis was highest for TNF-a and IL-1h levels in the peripheral $
This study was conducted under the auspices of the Ethical Committee of the University of Vienna, and informed consent was obtained from the patients’ relatives. * Corresponding author. Tel.: +43-140-400-6830; fax: +43-140-400-6782. E-mail address: [email protected] (U. Hallwirth). 0378-3782/02/$ - see front matter D 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 8 - 3 7 8 2 ( 0 1 ) 0 0 2 4 5 - 6
2
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
blood on the third day after birth. Conclusions: The determination of phagocytosis in the cord blood seems to be a reliable parameter for predicting early-onset sepsis and offers the possibility for a forward start of an antibiotic therapy. D 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Monocyte phagocytosis; Early-onset sepsis; Low birthweight infants
1. Introduction Neonatal infections remain a major cause of morbidity in very low birthweight infants (VLBWI) and despite modern intensive care and antibiotic treatment, the mortality of septic complications reaches up to about 50% [1,2]. The increased susceptibility of infections is the result of a variety of factors such as a delayed maturation of the specific humoral and cellular immune response and an incomplete reactivity of the complement system [3 –5]. The phagocytic capacity of the monocytes representing the first line immune response is known to be decreased during sepsis in adult patients. In preterm neonates, polymorphonuclear leukocytes (PMNL) are deficient both in numbers and in function [6,7]. These cellular effects result in the well-recognized susceptibility of such infants to infections. Furthermore, monocytes from healthy individuals express high levels of major histocompatibility complex (MHC) class II molecules, which play a crucial role in the antigen (Ag) presentation to CD4 + T cells [8]. Several investigations have shown that lowered levels of HLA-DR expression in patients with infection and sepsis are linked to recovery and mortality rates [9– 11]. The role of the tumor necrosis factor (TNF)-a as a central mediator associated with bacteriemia and sepsis syndrome is widely documented [12,13]. Other cytokines, such as interleukin-1 (IL-1), act synergistically with TNF-a, stimulating the release of mediators like IL-6, IL-8, IL-10, IL-13 and transforming growth factor-h (TGF-h) which participate in the pathogenesis of sepsis [14 –16]. Previous studies in preterm infants assessed IL-1 receptor antagonist and IL-6 concentrations as markers of neonatal sepsis with high sensitivities and specificities [17]. Because of the rapid progress of bacterial infection to severe sepsis with multiple organ dysfunction, predictive and early recognition of these infections is mandatory. The first clinical signs of neonatal sepsis are non-specific in the early course of sepsis, and laboratory parameters are often not sensitive enough [18 – 20]. Moreover, the exact determination of the microorganisms responsible for the sepsis needs at least 72 h. However, for this purpose, our aim was to find a test system which is easy and rapid to perform in the clinical situation. In our study, we therefore investigated in VLBWI the monocyte phagocytic capacity from the cord blood, the phagocytic properties as well as the expression of HLA-DR on peripheral blood monocytes at the third day after birth and the production of LPS-induced TNF-a, IL-1h and IL-6 at both time points.
2. Materials and methods Our study was performed at the University of Vienna/Department of Neonatology and Pediatric Surgery. It was conducted under the auspices of the Ethical Committee of the
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
3
University of Vienna, and informed consent was obtained from the patients’ relatives. Twenty-five preterm neonates aged between the 24th and 30th week of gestation with a birthweight under 1200 g (VLBWI) were enrolled into the study. The medical records of all patients and their mothers were evaluated retrospectively. Clinical data were recorded by a standard protocol including gestational age and birthweight, type of delivery, maternal history, antibiotic treatment, chorioamnionitis and rupture of membranes. Additionally, Apgar score, age at time of diagnosis of sepsis, respiratory status, bacteriological findings and radiological findings were evaluated. Sepsis was defined when the blood culture was positive. Parental approval based on thorough information was obtained for every patient before entry in the study. The study was approved by the local ethics committee. Directly after birth ( < 15 min), 4 ml of blood was obtained aseptically by squeezing or puncturing the cord into endotoxin-free vacutainer tubes (Endo Tube ET, Chromogenix, Moelndal, Sweden) with a heparin concentration of 30 IU/ml. At the third day of life, 800 Al of peripheral blood was obtained by venipuncture or by drawing from an indwelling venous line. Samples were collected in plastic tubes containing 15 IU heparin (Sarstedt, Nuembrecht, Germany) so that the final heparin concentration was 18 IU/ml. There was no additional venipuncture for study purposes. Monocyte phagocytosis and the expression of TNF-a, IL-1h and IL-6 were determined from the cord blood and peripheral blood at the third day of life. HLA-DR expression was determined from peripheral blood monocytes at the third day after birth. To determine phagocytosis, 100 Al heparinized whole blood was incubated for 15 min at 37 jC with non-opsonized FITC-conjugated E. coli bacteria (Orpegen, Heidelberg, Germany). Subsequently, the cells were stained with anti-CD14 mAb (My4 RD1, Coulter, Hialeah, FL), and red blood cells were lysed (Multi-Q-Prep, Coulter). Cells were washed three times in Hanks’ balanced salt solution (HBSS). The mean channel fluorescence intensity (MCF) of 2.5 103 CD14 + phagocytic cells was determined by FACS analysis. Fifty microliters of peripheral blood was stained with anti-CD14 mAb (My4 Cy5; Coulter), and anti HLA-DR (PE; Becton Dickinson, San Jose, CA). Fifty microliters of whole blood was stained with isotype control mAb (Immunotech, Marseille, France). After incubation for 30 min on ice, the red blood cells were lysed. Subsequently, the cells were washed three times in HBSS and the expression of HLA-DR on CD14 + cells was measured by FACS analysis. At least 2.5 103 of the CD14 + cells were counted. Results are given as MCF. TNF-a, IL-1h and IL-6 production was induced with 1 Ag/ml lipopolysaccharides (LPS) for 3 h by culturing 100 Al of cord blood and 100 Al peripheral blood both diluted in 900 Al of RPMI 1640 medium. Centrifuged supernatants were frozen immediately at 70 jC until analysis. The concentrations of the cytokines were measured by the use of commercially available enzyme-linked immunoassay (ELISA) kits (Amersham, Southampton, UK). Assays were performed in duplicates and analyzed in a plate reader (Dynatech, Chantilly, VA) [21]. The differences between the median values of phagocytic capacity, HLA-DR expression and plasma cytokine levels were analyzed by the Kruskal – Wallis test (one-way ANOVA). Analyzed data were performed using the SPSS software package (SPSS for Windows, Release 8.0.0, Chicago, IL). Differences were considered significant at p < 0.05. Furthermore, we calculated the optimum cut-off values, i.e. those that combined maximum
4
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
Fig. 1. E. coli phagocytosis of CD14 + monocytes from cord blood (A) and from peripheral blood (B) at the third day of age. Phagocytosis is given as mean channel fluorescence intensity (MCF). Statistical significance: * p < 0.001 difference between septic VLBWI and non-infected infants.
Fig. 2. HLA-DR expression on CD14 + peripheral blood monocytes at the third day of age. HLA-DR expression is given as mean channel fluorescence intensity (MCF). Statistical significance: * p < 0.04 difference between septic VLBWI and non-infected infants.
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
5
specificity and sensitivity for the diagnosis of early-onset sepsis from cord blood. As the positive separator variable, we used all individual values with proven early-onset sepsis measured from the cord blood. All individual values from the nonseptic VLBWI were used as the negative separator variable. Specificity was calculated as the percentage of the negative test results from all the values obtained from the nonseptic VLBWI.
3. Results Five of the neonates developed severe septic symptoms within the first 24 – 72 h after birth (early-onset sepsis). They were classified as septic neonates according to the inclusion criteria with respect to the defined clinical features. Three preterm neonates developed late-onset sepsis 10– 18 days after birth (nosocomial infections). According to the maternal records, none of the mothers of these eight neonates received an antibiotic treatment, and no chorioamnionitis was detected.
Table 1 Phagocytic capacity of monocytes and LPS-induced production of IL-1h, IL-6 and TNF-a from cord blood and from peripheral blood at the third day of age Cord blood None Phagocytosis (MCF) Median Maximum Minimum Number of samples p TNF-a (pg/ml) Median Maximum Minimum Number of samples p IL-1h (pg/ml) Median Maximum Minimum Number of samples p IL-6 (pg/ml) Median Maximum Minimum Number of samples p
Peripheral blood (day 3) Early-onset
Late-onset
4.8 7.4 1.81 5 0.001*
– – – – –
6772 17 387 1663 17 –
31 937 60 798 11 638 5 0.008*
29 133 32 807 21 966 3 0.022*
3260 9352 857 17 –
21 296 26 907 17 190 5 0.001*
23 714 35 622 3449 20 –
25 777 39 305 15 027 5 ns
83.4 176.6 22.3 20 –
None
Early-onset
Late-onset
8.6 13.4 2.0 5 0.001*
– – – – –
19 835 33 404 5165 17 –
11 683 14 806 7297 5 ns
41 630 57 326 32 494 3 0.032**
19 516 28 454 11 555 3 0.016*
8804 27 018 1748 17 –
4179 6481 1448 5 ns
27 659 35 130 21116 3 0.024**
25 703 29 108 19 323 3 ns
27 629 31 712 20 812 17 –
29 016 38 060 22 864 5 ns
28 354 29 098 27 050 3 ns
53.0 175.0 12.4 20 –
* p value calculated by the comparison of group-specific differences vs. non-infectious complications (cord blood). ** p value calculated by the comparison of group-specific differences vs. early onset (day 3).
6
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
Table 2 Sensitivity and specificity of phagocytic capacity of monocytes and LPS-induced TNF-a, IL-1h and IL-6 production from cord blood from VLBWI Cut-off value
Phagocytosis TNF-a IL-1h IL-6
Sensitivity
Specificity
Early-onset
Late-onset
Early-onset
Late-onset
Early-onset
Late-onset
8 MCF 20 000 pg/ml 15 000 pg/ml 30 000 pg/ml
– 20 000 pg/ml 15 000 pg/ml 30 000 pg/ml
100% 100% 100% 40%
– 33.3% (1/3) 66.7% (2/3) 33.3% (1/3)
100% 100% 95% 35%
– 100% (20/20) 95% (19/20) 35% (7/20)
(5/5) (5/5) (5/5) (2/5)
(20/20) (20/20) (19/20) (7/20)
In the five VLBWI developing early-onset sepsis, we found a significantly decreased median monocyte phagocytosis (MCF 4.8) in comparison to the infants who showed no infectious complications (MCF 83.4; p < 0.001; Fig. 1A). At day 3 of age in these five septic VLBWI, the phagocytic capacity of the monocytes from the peripheral blood remained significantly diminished (MCF 53.0 vs. MCF 8.6; p < 0.01; Fig. 1B). The expression of HLA-DR on CD14 + monocytes of the VLBWI was significantly decreased in the septic neonates (MCF 7.6) as compared to the nonseptic neonates (MCF 14.5; p = 0.04; Fig. 2). As can be seen in Table 1, the five VLBWI developing early-onset sepsis and the three neonates developing late-onset sepsis showed a significantly increased TNF-a (early-onset sepsis: p < 0.008; late-onset sepsis: p = 0.022) and IL-1h (early-onset sepsis: p < 0.001; late-onset sepsis: p = 0.016) expression as compared to the nonseptic neonates. The LPS-induced TNF-a and IL-1h levels from the peripheral blood were significantly elevated in the three VLBWI developing late-onset sepsis (TNF-a: p = 0.032; IL-1h: p = 0.024) as compared to the neonates developing early-onset sepsis. No significant differences in the IL-6 plasma levels could be found either in the early-onset or in the lateonset septic VLBWI when compared with the neonates without infectious complications. Sensitivity and specificity for early-onset sepsis were highest for phagocytosis and for LPS-induced TNF-a production (Table 2).
4. Discussion Very low birthweight infants have an immature immune system, and the development of infectious complications in these infants is associated with high morbidity and mortality. Since the outcome depends on an early and effective anti-bacterial therapy, the development of a method for precise and rapid predictive diagnosis of infections is a major challenge for neonatologists. Within 2 h after birth, we had an exact and extensive information about an early-onset sepsis at a time point before severe clinical symptoms appeared. This was indicated by a significantly reduced phagocytic capacity of the monocytes from the cord blood. Determination of phagocytosis by flow cytometry can be very rapidly performed with tiny samples of the blood so that the results are already available 1.5 h after sampling. Moreover, our results show that preterm neonates have a decreased HLA-DR expression on peripheral blood monocytes on day 3 when clinical
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
7
signs of infections were already detectable. Determination of LPS-induced cytokine production in the cord blood seems to be of a more prognostic value than the quantification of cytokines in the peripheral blood. Several investigators studied the role of polymorphonuclear leukocytes in newborns and showed that these cells are deficient both in numbers and in function [6,7]. Only few studies are available investigating the role of monocytes in preterm infants. The results of these studies are contradictory because both lowered phagocytic capacity and mature phagocytosis have been described [22 – 26]. However, none of these studies compared the phagocytic properties with infectious complications. In contrast to other investigators who determined phagocytosis from isolated cells, we used in our test system a whole blood technique with non-opsonized and FITC-labelled E. coli bacteria. Again, already 2 h after birth, the phagocytosis of the E. coli bacteria from the cord blood in the neonates developing early-onset sepsis was reduced to 94% when compared with non-infected VLBWI. At the third day when septic signs were already present, a significantly diminished phagocytosis from the monocytes of the peripheral blood was still detectable. The expression of HLA-DR on human monocytes correlates with the potency of antigen presentation [8,27]. Several studies demonstrated that the HLA-DR expression may be used as a parameter for the clinical management of immunosuppressed patients with sepsis [28]. Our results are in accordance with these studies since very low birthweight infants with septic complications had a significantly reduced monocyte HLA-DR expression when compared to unobtrusive neonates. To our knowledge, this is the first study investigating monocyte HLA-DR expression in preterm neonates. Earlier studies have revealed that pro-inflammatory cytokines are useful early predictors of neonatal sepsis [29 – 32]. IL-1ra and IL-6 can already predict neonatal sepsis 2 days before clinical manifestation [17]. However, the determination of these two cytokines takes at least 5 h until the results are available. Thus, it is more time-consuming than the determination of phagocytosis. This may be a disadvantage in the treatment of early-onset sepsis when effective antibiotic therapy at the earliest stage of the disease may be lifesaving. In our study, we stimulated the cord blood and peripheral blood taken 3 days after birth and determined the capability of white blood cells to produce pro-inflammatory cytokines. This test system seems to have an advantage over the measurement of cytokine concentrations in the plasma because, as earlier results have demonstrated, the capacity of the leukocytes to produce pro-inflammatory cytokines in whole blood is downregulated in septic patients [33]. Indeed, following LPS stimulation, a discrimination between the neonates developing early-onset sepsis and those who did not could be established. Interestingly, the ability to synthesize and secrete TNF-a and IL-1h in this exvivo test system was significantly enhanced in the cord blood and in the peripheral blood from neonates who developed early-onset or late-onset sepsis. However, in these patients who already showed sepsis signs, a tendency of reduced TNF-a and IL-1h could be observed. The release of IL-6 following LPS stimulation was without any predictive value. This result is in contrast to earlier results obtained by measuring plasma concentrations of IL-6 [29,31]. In conclusion, our data indicate that the phagocytic capacity of the cord blood gives predictive information about the development of early-onset sepsis. Since only small
8
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
amounts of cord blood are required and the test yields results within 1.5 h, this method is very suitable for very low birthweight infants. The early diagnosis made possible by this test allows early supportive and protective treatment in the preterm neonate.
Acknowledgements This work was partly supported by the ‘‘Jubila¨ums-Fond’’ National Bank Vienna, Austria (Project No. 7452).
References [1] Stoll BJ, Gordon T, Korones SB, et al. Early-onset sepsis in very low birthweight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network. J Pediatr 1996; 129:72 – 80. [2] Stoll BJ, Gordon T, Korones SB, et al. Late-onset sepsis in very low birthweight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network. J Pediatr 1996; 129:63 – 71. [3] Splawski JB, Jelinek DF, Lipsky PE. Delineation of the functional capacity of human neonatal lymphocytes. J Clin Invest 1991;87:545 – 53. [4] Wilson CB, Westall J, Johnston L, et al. Decreased production of interferon-gamma by human neonatal cells. J Clin Invest 1986;77:860 – 7. [5] Zach TL, Hostetter MK. Biochemical abnormalities of the third component of complement in neonates. Pediatr Res 1989;26:116 – 20. [6] Ruef P, Bo¨hler T, Linderkamp O. Deformability and volume of neonatal and adult leukocytes. Pediatr Res 1991;29:128 – 32. [7] Strauss RG, Snyder EL. Neutrophils from human infants exhibit decreased viability. Pediatr Res 1981;15: 794 – 800. [8] Spittler A, Winkler S, Go¨tzinger P, et al. Influence of glutamine on the phenotype and function of human monocytes. Blood 1995;86:1564 – 9. [9] Hershman MJ, Cheadle WG, Wellhausen SR, et al. Monocyte HLA-DR antigen expression characterizes clinical outcome in the trauma patient. Br J Surg 1990;77:204 – 7. [10] Van den Berk JMM, Oldenburger RHJ, van den Berg AP, et al. Low HLA-DR expression on monocytes as a prognostic marker for bacterial sepsis after liver transplantation. Transplantation 1997;63:1846 – 8. [11] Wakefield CH, Carey PD, Foulds S, et al. Changes in major histocompatibility complex class II expression in monocytes and T cells of patients developing infection after surgery. Br J Surg 1993;80:205 – 9. [12] Spooner CE, Markowitz NP, Saravolatz LD. The role of tumor necrosis factor in sepsis. Clin Immunol Immunopathol 1992;62:11 – 7. [13] Cerami AC, Beutler B. Role of cachectin TNF in endotoxic shock and cachexia. Immunol Today 1988; 9:28 – 31. [14] Fong Y, Lowry SF. Tumor necrosis factor in the pathophysiology of infection and sepsis. Clin Immunol Immunopathol 1990;55:157 – 70. [15] Tracey KJ, Fong Y, Hesse DG, et al. Anti-cachectin/TNF monoclonal antibodies prevent sepsis shock during lethal bacteriemia. Nature 1987;330:662 – 4. [16] Doherty GM, Lange JR, Langstein HN, et al. Evidence for IFN-g as a mediator of the lethality of endotoxin and tumor necrosis factor-a. J Immunol 1992;149:1666 – 70. [17] Ku¨ster H, Weiss M, Willeitner AE, et al. Interleukin-1 receptor antagonist and interleukin-6 for early diagnosis of neonatal sepsis 2 days before clinical manifestation. Lancet 1998;352:1271 – 7. [18] Rodwell RL, Leslie AL, Tudehope DI. Early diagnosis of neonatal sepsis using a hematologic scoring system. J Pediatr 1988;112:761 – 7.
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
9
[19] Da Silva O, Ohlsson A, Kenyon C. Accuracy of leukocyte indices and C-reactive protein for diagnosis of neonatal sepsis: a critical review. Pediatr Infect Dis J 1995;14:362 – 6. [20] Mathers NJ, Pohlandt F. Diagnostic audit of reactive protein in neonatal infection. Eur J Pediatr 1987; 146:147 – 51. [21] Spittler A, Reissner CM, Oehler R, et al. Immunomodulatory effects of glycine on LPS-stimulated human blood monocytes: reduced TNF-a production and accelerated IL-10 expression. FASEB J 1999;13:563 – 71. [22] Mills EL. Mononuclear phagocytes in the newborn: their relation to the state of relative immunodeficiency. Am J Pediatr Hematol/Oncol 1983;5:189 – 201. [23] Speer ChP, Wieland M, Ulbrich R, Gahr M. Phagocytic activities in neonatal monocytes. Eur J Pediatr. 1986;145:418 – 21. [24] Kretschmer RR, Stewardson PB, Papierniak CK, Gotoff SP. Chemotactic and bactericidal capacities of human newborn monocytes. J Immunol 1976;117:1303 – 7. [25] Colny ME, Speer DP. Human neonatal monocyte-derived macrophages and neutrophils exhibit normal nonopsonic and opsonic receptor-mediated phagocytosis and superoxide anion production. Biol Neonate 1991;60:361 – 6. [26] Orlowski JP, Sieger L, Anthony BF. Bactericidal capacity of monocytes of newborn infants. J Pediatr 1976; 89:797 – 801. [27] Kurt-Jones E, Virgin HW, Unanue ER. Relationship of macrophage Ia and membrane IL1 expression to antigen presentation. J Immunol 1985;135:3652 – 4. [28] Docke WD, Randow F, Syrbe U, et al. Monocyte deactivation in septic patients: restoration by IFN-gamma treatment. Nat Med 1997;3:678 – 81. [29] Berner R, Niemeyer CM, Leititis JU, et al. Plasma levels and gene expression of granulocyte colonystimulation factor, tumor necrosis factor-a, interleukin (IL)-1h, IL-6, IL-8 and soluble intercellular adhesion molecule-1 in neonatal early onset sepsis. Pediatr Res 1998;44:469 – 77. [30] Peters AMJ, Bertram P, Gahr M, Speer CP. Reduced secretion of interleukin-1 and tumor necrosis factor-a by neonatal monocytes. Biol Neonate 1993;63:157 – 66. [31] Meister B, Herold M, Mayr A, et al. Interleukin-3, interleukin-6, granulocyte-macrophage colony-stimulating factor and erythropoietin cord blood levels of preterm and term neonates. Eur J Pediatr 1993;152:569 – 73. [32] Strugo I, Berger A, Lapidot Z, et al. Interleukin-1 secretion by blood monocytes of septic premature infants. Infection 1991;19:150 – 3. [33] Ertel W, Kremer JP, Kenney J, et al. Down-regulation of proinflammatory cytokine release in whole blood from septic patients. Blood 1995;85:1341 – 7.
Monocyte phagocytosis as a reliable parameter for predicting early-onset sepsis in very low birthweight infants $ Ulrike Hallwirth a,*, Gerhard Pomberger a, Daniela Zaknun b, Zsolt Szepfalusi b, Ernst Horcher a, Arnold Pollak b, Erich Roth c, Andreas Spittler c a
Division of Pediatric Surgery, Department of Surgery, University Medical School of Vienna, AKH Waehringer Guertel 18-20, 1090 Vienna, Austria b Department of Pediatrics, University Medical School of Vienna, Vienna, Austria c Surgical Research Laboratories, University Medical School of Vienna, Vienna, Austria Accepted 26 October 2001
Abstract Background: Septic complications lead to a high mortality rate in very low birthweight infants (VLBWI). Therefore, prognostic markers for the development of sepsis attach importance to start an efficient therapy as early as possible. Aims: Functional and phenotypical variables of blood monocytes in the cord and peripheral blood were investigated to evaluate the parameters for predicting early-onset and late-onset sepsis (nosocomial infections). Study design: In a prospective study, 25 VLBWI were investigated. Methods: In the cord blood taken immediately after birth, the capacity of the monocytes to phagocytose non-opsonized E. coli bacteria by flow cytometry and the ex-vivo production of tumor necrosis factor (TNF)-a, interleukin (IL)-1h and IL-6 (enzyme-linked immunoassay (ELISA)) after lipopolysaccharide (LPS) stimulation were measured. At the third day, the HLA-DR expression on the monocytes (flow cytometry) and the LPS-induced cytokine production were measured from the peripheral blood. Results: Five VLBWI already developed early septic complications after 24 – 72 h, while the other three infants had late-onset sepsis 10 – 18 days after birth. The prognostic significance for early-onset sepsis was highest for the decreased monocyte phagocytic capacity (sensitivity and specificity: 100%) and for the LPS-induced formation of TNF-a and IL-1h in cord blood. Moreover, in septic VLBWI, the HLA-DR expression on the monocytes was lowered on day 3 after birth. The prognostic significance for late-onset sepsis was highest for TNF-a and IL-1h levels in the peripheral $
This study was conducted under the auspices of the Ethical Committee of the University of Vienna, and informed consent was obtained from the patients’ relatives. * Corresponding author. Tel.: +43-140-400-6830; fax: +43-140-400-6782. E-mail address: [email protected] (U. Hallwirth). 0378-3782/02/$ - see front matter D 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 8 - 3 7 8 2 ( 0 1 ) 0 0 2 4 5 - 6
2
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
blood on the third day after birth. Conclusions: The determination of phagocytosis in the cord blood seems to be a reliable parameter for predicting early-onset sepsis and offers the possibility for a forward start of an antibiotic therapy. D 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Monocyte phagocytosis; Early-onset sepsis; Low birthweight infants
1. Introduction Neonatal infections remain a major cause of morbidity in very low birthweight infants (VLBWI) and despite modern intensive care and antibiotic treatment, the mortality of septic complications reaches up to about 50% [1,2]. The increased susceptibility of infections is the result of a variety of factors such as a delayed maturation of the specific humoral and cellular immune response and an incomplete reactivity of the complement system [3 –5]. The phagocytic capacity of the monocytes representing the first line immune response is known to be decreased during sepsis in adult patients. In preterm neonates, polymorphonuclear leukocytes (PMNL) are deficient both in numbers and in function [6,7]. These cellular effects result in the well-recognized susceptibility of such infants to infections. Furthermore, monocytes from healthy individuals express high levels of major histocompatibility complex (MHC) class II molecules, which play a crucial role in the antigen (Ag) presentation to CD4 + T cells [8]. Several investigations have shown that lowered levels of HLA-DR expression in patients with infection and sepsis are linked to recovery and mortality rates [9– 11]. The role of the tumor necrosis factor (TNF)-a as a central mediator associated with bacteriemia and sepsis syndrome is widely documented [12,13]. Other cytokines, such as interleukin-1 (IL-1), act synergistically with TNF-a, stimulating the release of mediators like IL-6, IL-8, IL-10, IL-13 and transforming growth factor-h (TGF-h) which participate in the pathogenesis of sepsis [14 –16]. Previous studies in preterm infants assessed IL-1 receptor antagonist and IL-6 concentrations as markers of neonatal sepsis with high sensitivities and specificities [17]. Because of the rapid progress of bacterial infection to severe sepsis with multiple organ dysfunction, predictive and early recognition of these infections is mandatory. The first clinical signs of neonatal sepsis are non-specific in the early course of sepsis, and laboratory parameters are often not sensitive enough [18 – 20]. Moreover, the exact determination of the microorganisms responsible for the sepsis needs at least 72 h. However, for this purpose, our aim was to find a test system which is easy and rapid to perform in the clinical situation. In our study, we therefore investigated in VLBWI the monocyte phagocytic capacity from the cord blood, the phagocytic properties as well as the expression of HLA-DR on peripheral blood monocytes at the third day after birth and the production of LPS-induced TNF-a, IL-1h and IL-6 at both time points.
2. Materials and methods Our study was performed at the University of Vienna/Department of Neonatology and Pediatric Surgery. It was conducted under the auspices of the Ethical Committee of the
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
3
University of Vienna, and informed consent was obtained from the patients’ relatives. Twenty-five preterm neonates aged between the 24th and 30th week of gestation with a birthweight under 1200 g (VLBWI) were enrolled into the study. The medical records of all patients and their mothers were evaluated retrospectively. Clinical data were recorded by a standard protocol including gestational age and birthweight, type of delivery, maternal history, antibiotic treatment, chorioamnionitis and rupture of membranes. Additionally, Apgar score, age at time of diagnosis of sepsis, respiratory status, bacteriological findings and radiological findings were evaluated. Sepsis was defined when the blood culture was positive. Parental approval based on thorough information was obtained for every patient before entry in the study. The study was approved by the local ethics committee. Directly after birth ( < 15 min), 4 ml of blood was obtained aseptically by squeezing or puncturing the cord into endotoxin-free vacutainer tubes (Endo Tube ET, Chromogenix, Moelndal, Sweden) with a heparin concentration of 30 IU/ml. At the third day of life, 800 Al of peripheral blood was obtained by venipuncture or by drawing from an indwelling venous line. Samples were collected in plastic tubes containing 15 IU heparin (Sarstedt, Nuembrecht, Germany) so that the final heparin concentration was 18 IU/ml. There was no additional venipuncture for study purposes. Monocyte phagocytosis and the expression of TNF-a, IL-1h and IL-6 were determined from the cord blood and peripheral blood at the third day of life. HLA-DR expression was determined from peripheral blood monocytes at the third day after birth. To determine phagocytosis, 100 Al heparinized whole blood was incubated for 15 min at 37 jC with non-opsonized FITC-conjugated E. coli bacteria (Orpegen, Heidelberg, Germany). Subsequently, the cells were stained with anti-CD14 mAb (My4 RD1, Coulter, Hialeah, FL), and red blood cells were lysed (Multi-Q-Prep, Coulter). Cells were washed three times in Hanks’ balanced salt solution (HBSS). The mean channel fluorescence intensity (MCF) of 2.5 103 CD14 + phagocytic cells was determined by FACS analysis. Fifty microliters of peripheral blood was stained with anti-CD14 mAb (My4 Cy5; Coulter), and anti HLA-DR (PE; Becton Dickinson, San Jose, CA). Fifty microliters of whole blood was stained with isotype control mAb (Immunotech, Marseille, France). After incubation for 30 min on ice, the red blood cells were lysed. Subsequently, the cells were washed three times in HBSS and the expression of HLA-DR on CD14 + cells was measured by FACS analysis. At least 2.5 103 of the CD14 + cells were counted. Results are given as MCF. TNF-a, IL-1h and IL-6 production was induced with 1 Ag/ml lipopolysaccharides (LPS) for 3 h by culturing 100 Al of cord blood and 100 Al peripheral blood both diluted in 900 Al of RPMI 1640 medium. Centrifuged supernatants were frozen immediately at 70 jC until analysis. The concentrations of the cytokines were measured by the use of commercially available enzyme-linked immunoassay (ELISA) kits (Amersham, Southampton, UK). Assays were performed in duplicates and analyzed in a plate reader (Dynatech, Chantilly, VA) [21]. The differences between the median values of phagocytic capacity, HLA-DR expression and plasma cytokine levels were analyzed by the Kruskal – Wallis test (one-way ANOVA). Analyzed data were performed using the SPSS software package (SPSS for Windows, Release 8.0.0, Chicago, IL). Differences were considered significant at p < 0.05. Furthermore, we calculated the optimum cut-off values, i.e. those that combined maximum
4
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
Fig. 1. E. coli phagocytosis of CD14 + monocytes from cord blood (A) and from peripheral blood (B) at the third day of age. Phagocytosis is given as mean channel fluorescence intensity (MCF). Statistical significance: * p < 0.001 difference between septic VLBWI and non-infected infants.
Fig. 2. HLA-DR expression on CD14 + peripheral blood monocytes at the third day of age. HLA-DR expression is given as mean channel fluorescence intensity (MCF). Statistical significance: * p < 0.04 difference between septic VLBWI and non-infected infants.
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
5
specificity and sensitivity for the diagnosis of early-onset sepsis from cord blood. As the positive separator variable, we used all individual values with proven early-onset sepsis measured from the cord blood. All individual values from the nonseptic VLBWI were used as the negative separator variable. Specificity was calculated as the percentage of the negative test results from all the values obtained from the nonseptic VLBWI.
3. Results Five of the neonates developed severe septic symptoms within the first 24 – 72 h after birth (early-onset sepsis). They were classified as septic neonates according to the inclusion criteria with respect to the defined clinical features. Three preterm neonates developed late-onset sepsis 10– 18 days after birth (nosocomial infections). According to the maternal records, none of the mothers of these eight neonates received an antibiotic treatment, and no chorioamnionitis was detected.
Table 1 Phagocytic capacity of monocytes and LPS-induced production of IL-1h, IL-6 and TNF-a from cord blood and from peripheral blood at the third day of age Cord blood None Phagocytosis (MCF) Median Maximum Minimum Number of samples p TNF-a (pg/ml) Median Maximum Minimum Number of samples p IL-1h (pg/ml) Median Maximum Minimum Number of samples p IL-6 (pg/ml) Median Maximum Minimum Number of samples p
Peripheral blood (day 3) Early-onset
Late-onset
4.8 7.4 1.81 5 0.001*
– – – – –
6772 17 387 1663 17 –
31 937 60 798 11 638 5 0.008*
29 133 32 807 21 966 3 0.022*
3260 9352 857 17 –
21 296 26 907 17 190 5 0.001*
23 714 35 622 3449 20 –
25 777 39 305 15 027 5 ns
83.4 176.6 22.3 20 –
None
Early-onset
Late-onset
8.6 13.4 2.0 5 0.001*
– – – – –
19 835 33 404 5165 17 –
11 683 14 806 7297 5 ns
41 630 57 326 32 494 3 0.032**
19 516 28 454 11 555 3 0.016*
8804 27 018 1748 17 –
4179 6481 1448 5 ns
27 659 35 130 21116 3 0.024**
25 703 29 108 19 323 3 ns
27 629 31 712 20 812 17 –
29 016 38 060 22 864 5 ns
28 354 29 098 27 050 3 ns
53.0 175.0 12.4 20 –
* p value calculated by the comparison of group-specific differences vs. non-infectious complications (cord blood). ** p value calculated by the comparison of group-specific differences vs. early onset (day 3).
6
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
Table 2 Sensitivity and specificity of phagocytic capacity of monocytes and LPS-induced TNF-a, IL-1h and IL-6 production from cord blood from VLBWI Cut-off value
Phagocytosis TNF-a IL-1h IL-6
Sensitivity
Specificity
Early-onset
Late-onset
Early-onset
Late-onset
Early-onset
Late-onset
8 MCF 20 000 pg/ml 15 000 pg/ml 30 000 pg/ml
– 20 000 pg/ml 15 000 pg/ml 30 000 pg/ml
100% 100% 100% 40%
– 33.3% (1/3) 66.7% (2/3) 33.3% (1/3)
100% 100% 95% 35%
– 100% (20/20) 95% (19/20) 35% (7/20)
(5/5) (5/5) (5/5) (2/5)
(20/20) (20/20) (19/20) (7/20)
In the five VLBWI developing early-onset sepsis, we found a significantly decreased median monocyte phagocytosis (MCF 4.8) in comparison to the infants who showed no infectious complications (MCF 83.4; p < 0.001; Fig. 1A). At day 3 of age in these five septic VLBWI, the phagocytic capacity of the monocytes from the peripheral blood remained significantly diminished (MCF 53.0 vs. MCF 8.6; p < 0.01; Fig. 1B). The expression of HLA-DR on CD14 + monocytes of the VLBWI was significantly decreased in the septic neonates (MCF 7.6) as compared to the nonseptic neonates (MCF 14.5; p = 0.04; Fig. 2). As can be seen in Table 1, the five VLBWI developing early-onset sepsis and the three neonates developing late-onset sepsis showed a significantly increased TNF-a (early-onset sepsis: p < 0.008; late-onset sepsis: p = 0.022) and IL-1h (early-onset sepsis: p < 0.001; late-onset sepsis: p = 0.016) expression as compared to the nonseptic neonates. The LPS-induced TNF-a and IL-1h levels from the peripheral blood were significantly elevated in the three VLBWI developing late-onset sepsis (TNF-a: p = 0.032; IL-1h: p = 0.024) as compared to the neonates developing early-onset sepsis. No significant differences in the IL-6 plasma levels could be found either in the early-onset or in the lateonset septic VLBWI when compared with the neonates without infectious complications. Sensitivity and specificity for early-onset sepsis were highest for phagocytosis and for LPS-induced TNF-a production (Table 2).
4. Discussion Very low birthweight infants have an immature immune system, and the development of infectious complications in these infants is associated with high morbidity and mortality. Since the outcome depends on an early and effective anti-bacterial therapy, the development of a method for precise and rapid predictive diagnosis of infections is a major challenge for neonatologists. Within 2 h after birth, we had an exact and extensive information about an early-onset sepsis at a time point before severe clinical symptoms appeared. This was indicated by a significantly reduced phagocytic capacity of the monocytes from the cord blood. Determination of phagocytosis by flow cytometry can be very rapidly performed with tiny samples of the blood so that the results are already available 1.5 h after sampling. Moreover, our results show that preterm neonates have a decreased HLA-DR expression on peripheral blood monocytes on day 3 when clinical
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
7
signs of infections were already detectable. Determination of LPS-induced cytokine production in the cord blood seems to be of a more prognostic value than the quantification of cytokines in the peripheral blood. Several investigators studied the role of polymorphonuclear leukocytes in newborns and showed that these cells are deficient both in numbers and in function [6,7]. Only few studies are available investigating the role of monocytes in preterm infants. The results of these studies are contradictory because both lowered phagocytic capacity and mature phagocytosis have been described [22 – 26]. However, none of these studies compared the phagocytic properties with infectious complications. In contrast to other investigators who determined phagocytosis from isolated cells, we used in our test system a whole blood technique with non-opsonized and FITC-labelled E. coli bacteria. Again, already 2 h after birth, the phagocytosis of the E. coli bacteria from the cord blood in the neonates developing early-onset sepsis was reduced to 94% when compared with non-infected VLBWI. At the third day when septic signs were already present, a significantly diminished phagocytosis from the monocytes of the peripheral blood was still detectable. The expression of HLA-DR on human monocytes correlates with the potency of antigen presentation [8,27]. Several studies demonstrated that the HLA-DR expression may be used as a parameter for the clinical management of immunosuppressed patients with sepsis [28]. Our results are in accordance with these studies since very low birthweight infants with septic complications had a significantly reduced monocyte HLA-DR expression when compared to unobtrusive neonates. To our knowledge, this is the first study investigating monocyte HLA-DR expression in preterm neonates. Earlier studies have revealed that pro-inflammatory cytokines are useful early predictors of neonatal sepsis [29 – 32]. IL-1ra and IL-6 can already predict neonatal sepsis 2 days before clinical manifestation [17]. However, the determination of these two cytokines takes at least 5 h until the results are available. Thus, it is more time-consuming than the determination of phagocytosis. This may be a disadvantage in the treatment of early-onset sepsis when effective antibiotic therapy at the earliest stage of the disease may be lifesaving. In our study, we stimulated the cord blood and peripheral blood taken 3 days after birth and determined the capability of white blood cells to produce pro-inflammatory cytokines. This test system seems to have an advantage over the measurement of cytokine concentrations in the plasma because, as earlier results have demonstrated, the capacity of the leukocytes to produce pro-inflammatory cytokines in whole blood is downregulated in septic patients [33]. Indeed, following LPS stimulation, a discrimination between the neonates developing early-onset sepsis and those who did not could be established. Interestingly, the ability to synthesize and secrete TNF-a and IL-1h in this exvivo test system was significantly enhanced in the cord blood and in the peripheral blood from neonates who developed early-onset or late-onset sepsis. However, in these patients who already showed sepsis signs, a tendency of reduced TNF-a and IL-1h could be observed. The release of IL-6 following LPS stimulation was without any predictive value. This result is in contrast to earlier results obtained by measuring plasma concentrations of IL-6 [29,31]. In conclusion, our data indicate that the phagocytic capacity of the cord blood gives predictive information about the development of early-onset sepsis. Since only small
8
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
amounts of cord blood are required and the test yields results within 1.5 h, this method is very suitable for very low birthweight infants. The early diagnosis made possible by this test allows early supportive and protective treatment in the preterm neonate.
Acknowledgements This work was partly supported by the ‘‘Jubila¨ums-Fond’’ National Bank Vienna, Austria (Project No. 7452).
References [1] Stoll BJ, Gordon T, Korones SB, et al. Early-onset sepsis in very low birthweight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network. J Pediatr 1996; 129:72 – 80. [2] Stoll BJ, Gordon T, Korones SB, et al. Late-onset sepsis in very low birthweight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network. J Pediatr 1996; 129:63 – 71. [3] Splawski JB, Jelinek DF, Lipsky PE. Delineation of the functional capacity of human neonatal lymphocytes. J Clin Invest 1991;87:545 – 53. [4] Wilson CB, Westall J, Johnston L, et al. Decreased production of interferon-gamma by human neonatal cells. J Clin Invest 1986;77:860 – 7. [5] Zach TL, Hostetter MK. Biochemical abnormalities of the third component of complement in neonates. Pediatr Res 1989;26:116 – 20. [6] Ruef P, Bo¨hler T, Linderkamp O. Deformability and volume of neonatal and adult leukocytes. Pediatr Res 1991;29:128 – 32. [7] Strauss RG, Snyder EL. Neutrophils from human infants exhibit decreased viability. Pediatr Res 1981;15: 794 – 800. [8] Spittler A, Winkler S, Go¨tzinger P, et al. Influence of glutamine on the phenotype and function of human monocytes. Blood 1995;86:1564 – 9. [9] Hershman MJ, Cheadle WG, Wellhausen SR, et al. Monocyte HLA-DR antigen expression characterizes clinical outcome in the trauma patient. Br J Surg 1990;77:204 – 7. [10] Van den Berk JMM, Oldenburger RHJ, van den Berg AP, et al. Low HLA-DR expression on monocytes as a prognostic marker for bacterial sepsis after liver transplantation. Transplantation 1997;63:1846 – 8. [11] Wakefield CH, Carey PD, Foulds S, et al. Changes in major histocompatibility complex class II expression in monocytes and T cells of patients developing infection after surgery. Br J Surg 1993;80:205 – 9. [12] Spooner CE, Markowitz NP, Saravolatz LD. The role of tumor necrosis factor in sepsis. Clin Immunol Immunopathol 1992;62:11 – 7. [13] Cerami AC, Beutler B. Role of cachectin TNF in endotoxic shock and cachexia. Immunol Today 1988; 9:28 – 31. [14] Fong Y, Lowry SF. Tumor necrosis factor in the pathophysiology of infection and sepsis. Clin Immunol Immunopathol 1990;55:157 – 70. [15] Tracey KJ, Fong Y, Hesse DG, et al. Anti-cachectin/TNF monoclonal antibodies prevent sepsis shock during lethal bacteriemia. Nature 1987;330:662 – 4. [16] Doherty GM, Lange JR, Langstein HN, et al. Evidence for IFN-g as a mediator of the lethality of endotoxin and tumor necrosis factor-a. J Immunol 1992;149:1666 – 70. [17] Ku¨ster H, Weiss M, Willeitner AE, et al. Interleukin-1 receptor antagonist and interleukin-6 for early diagnosis of neonatal sepsis 2 days before clinical manifestation. Lancet 1998;352:1271 – 7. [18] Rodwell RL, Leslie AL, Tudehope DI. Early diagnosis of neonatal sepsis using a hematologic scoring system. J Pediatr 1988;112:761 – 7.
U. Hallwirth et al. / Early Human Development 67 (2002) 1–9
9
[19] Da Silva O, Ohlsson A, Kenyon C. Accuracy of leukocyte indices and C-reactive protein for diagnosis of neonatal sepsis: a critical review. Pediatr Infect Dis J 1995;14:362 – 6. [20] Mathers NJ, Pohlandt F. Diagnostic audit of reactive protein in neonatal infection. Eur J Pediatr 1987; 146:147 – 51. [21] Spittler A, Reissner CM, Oehler R, et al. Immunomodulatory effects of glycine on LPS-stimulated human blood monocytes: reduced TNF-a production and accelerated IL-10 expression. FASEB J 1999;13:563 – 71. [22] Mills EL. Mononuclear phagocytes in the newborn: their relation to the state of relative immunodeficiency. Am J Pediatr Hematol/Oncol 1983;5:189 – 201. [23] Speer ChP, Wieland M, Ulbrich R, Gahr M. Phagocytic activities in neonatal monocytes. Eur J Pediatr. 1986;145:418 – 21. [24] Kretschmer RR, Stewardson PB, Papierniak CK, Gotoff SP. Chemotactic and bactericidal capacities of human newborn monocytes. J Immunol 1976;117:1303 – 7. [25] Colny ME, Speer DP. Human neonatal monocyte-derived macrophages and neutrophils exhibit normal nonopsonic and opsonic receptor-mediated phagocytosis and superoxide anion production. Biol Neonate 1991;60:361 – 6. [26] Orlowski JP, Sieger L, Anthony BF. Bactericidal capacity of monocytes of newborn infants. J Pediatr 1976; 89:797 – 801. [27] Kurt-Jones E, Virgin HW, Unanue ER. Relationship of macrophage Ia and membrane IL1 expression to antigen presentation. J Immunol 1985;135:3652 – 4. [28] Docke WD, Randow F, Syrbe U, et al. Monocyte deactivation in septic patients: restoration by IFN-gamma treatment. Nat Med 1997;3:678 – 81. [29] Berner R, Niemeyer CM, Leititis JU, et al. Plasma levels and gene expression of granulocyte colonystimulation factor, tumor necrosis factor-a, interleukin (IL)-1h, IL-6, IL-8 and soluble intercellular adhesion molecule-1 in neonatal early onset sepsis. Pediatr Res 1998;44:469 – 77. [30] Peters AMJ, Bertram P, Gahr M, Speer CP. Reduced secretion of interleukin-1 and tumor necrosis factor-a by neonatal monocytes. Biol Neonate 1993;63:157 – 66. [31] Meister B, Herold M, Mayr A, et al. Interleukin-3, interleukin-6, granulocyte-macrophage colony-stimulating factor and erythropoietin cord blood levels of preterm and term neonates. Eur J Pediatr 1993;152:569 – 73. [32] Strugo I, Berger A, Lapidot Z, et al. Interleukin-1 secretion by blood monocytes of septic premature infants. Infection 1991;19:150 – 3. [33] Ertel W, Kremer JP, Kenney J, et al. Down-regulation of proinflammatory cytokine release in whole blood from septic patients. Blood 1995;85:1341 – 7.