Electronic Supplementary Material: Intensive Care Medicine

Low monocyte human leukocyte antigen-DR is independently associated with nosocomial infections after septic shock

Caroline LANDELLE, Alain LEPAPE, Nicolas VOIRIN, Eve TOGNET, Fabienne VENET, Julien BOHE, Philippe VANHEMS, Guillaume MONNERET

Corresponding author:

Dr. Guillaume MONNERET

Laboratoire d’Immunologie - Hôpital E. Herriot – Hospices Civils de Lyon

Pavillon E – 5 place d’Arsonval - 69437 Lyon Cedex 03, France

Tel: (33) 4 72 11 97 58 - Fax: (33) 4 72 11 97 53

e-mail:

MATERIAL AND METHODS

Setting

This prospective cohort study was conducted in 2 French ICUs, 1 medical and 1 surgical, in a 900-bed university hospital over a 41-month period. The 2 ICUs have 28 beds and treat over 1,100 patients per year. The protocol was reviewed by the institutional ethics committee, which waived the need for informed patient consent because mHLA-DR expression was measured in residual blood after completing routine follow-up.

A total of 209 consecutive septic shock patients were included from December 1, 2001 through April 30, 2005. These patients were described previously in detail regarding epidemiological aspects of NI after septic shock[S1]. The relationship between mHLA-DR and mortality was assessed earlier in a subgroup of patients [S2].

Study population

The diagnosis of septic shock was based on classical criteria of the American College of Chest Physicians/Society of Critical Care Medicine [S3]. It was defined by an identifiable infection site and evidence of systemic inflammatory response syndrome (SIRS), documented by at least 2 of the following criteria: (1) body temperature 38°C or 36°C; (2) heart rate 90 beats per min; (3) respiratory rate 20 breaths per min; (4) white blood cell count 12,000/mm3 or 4,000/mm3; and systolic blood pressure below 90 mm Hg despite fluid resuscitation and requiring vasopressor therapy (more than 5 μg/mn of epinephrine and/or norepinephrine and/or more than 10 μg/kg/mn body weight of dopamine). The timing of septic shock onset was defined by the start of vasopressor treatment. Septic shock treatment was in adherence with current international guidelines [S4, S5]. In the two participating units, recommendations existed even before the onset of the Surviving Sepsis Campaign, concerning fluid resuscitation, use of vasopressors (preferentially norepinephrin), and early use of antimicrobials according to a protocol after blood and sepsis source sampling. Exclusion criteria were patients younger than age 18 years or subjects with aplasia.

During follow-up, from admission to ICU discharge, clinical and biological data were collected from 2 sources: a trained research nurse collected specific data on all septic shock patients, and trained physicians in charge of the patient collected data based on the national ICU NI surveillance network[S6]. For all variables, standardized definitions were provided in an operating manual.Data collection comprised demographic characteristics (age, gender); admission category (medical, emergency or scheduled surgery, trauma); co-morbidities (AIDS, haematological malignancies, other immuno-compromised states, diabetes mellitus, metastatic cancer, chronic heart, pulmonary, renal, or hepatic failure); place of septic shock acquisition (community, hospital or ICU); microbiological findings of septic shock (diagnosis documentation, origin and identified microorganisms); interventions (steroid treatment, renal replacement therapy, intubation, urinary tract and central venous catheterization (CVC) after septic shock) and their duration; antimicrobial therapy and its adequacy, according to predetermined rules adapted from Kumar et al.[S7]; and outcome (death or survival) at 28 days as well as at ICU and hospital discharge.

Three scores were recorded: the severity of underlying medical conditions by the McCabe and Jackson scale (range 1-3)[S8]; initial severity and at the onset of septic shock, assessed by Simplified Acute Physiology Score II (SAPS II, range 0-163)[S9]; Sepsis-related Organ Failure Assessment (SOFA, range 0–24) score at day 1 or 2 after admission [S10].

All patients were screened daily by a clinical approach for 4 NI sites after septic shock onset during ICU stay: microbiologically-documented pulmonary (PI), urinary tract (UTI), bloodstream (BSI) and catheter-related infections (CRI). For each NI, the onset date and maximum 2 microorganisms were reported. Definitions of NI were those used by the European Network for Surveillance of NI in ICUs[S11].

PI is defined according to clinical criteria (X-rays, fever 38ºC, leukocytosis12000WBC/mm3 or leukopenia <4000WBC/mm3, purulent sputum, etc.) and further sub-categorized in three entities according to the level of microbiological confirmation:

- minimally contaminated lower respiratory tract sample with quantitative culture (104 CFU/ml for broncheoalveolar lavage, 103 CFU/ml for protected brush samples or distal protected aspirate).

- non-protected sample (endotracheal aspirate, ETA) with quantitative culture (106 CFU/ml).

- alternative microbiological criteria (e.g. positive blood culture).

UTI is defined as:

-urinary cultures ≥105 CFU/ml for patients who received a urinary catheter during the last 7 preceding days.

-urinary cultures were 105 CFU/ml or ≥103 CFU/ml with leukocyturia ≥104 CFU/ml and patients presented the following clinical features: fever >38°C and/or pollakiuria and/or dysuria and/or suprapubic pain.

BSI is defined as a positive blood culture for a recognised pathogen or the combination of clinical symptoms (fever 38ºC, chills, hypotension) and two positive blood cultures for a common skin contaminant from 2 separate blood samples drawn within 48 hours.

CRI is defined as:

-local CVC-related infection (no positive blood culture): quantitative CVC culture ≥103 CFU/ml and pus at the insertion site or tunnel.

-general CVC-related infection (no positive blood culture): quantitative CVC culture ≥103 CFU/ml and clinical signs improve within 48 hours after catheter removal.

When a patient experienced more than 1 episode per NI site after septic shock, the analysis was restricted to the first episode of NI per site. For patients who acquired device-associated NI, the duration of device exposure was censored at the first NI onset. All NI after septic shock were reviewed concomitantly by 2 investigators blinded to mHLA-DR results, and in cases of dissension, agreement was reached after re-discussion when necessary.

mHLA-DR measurement by flow cytometry

The expression of cell surface HLA-DR on monocytes was assessed at D3-4 (days 3-4) and D6-9 (days 6-9) by standardized flowcytometry in peripheral whole blood collected in EDTA anticoagulant tubes. As planned, blood samples were not collected on Saturdays and Sundays during which the laboratory did not operate; a fact that accounts for most of the missing values. Staining and cell acquisition were performed within 2h after blood sampling and undertaken as described previously [S2, S12]. For patients who acquired NI, the expression of mHLA-DR was censored at the first NI onset. The results are expressed as percentages of HLA-DR-positive monocytes out of the total monocyte population. Control values are 90% [S12]. The data are also expressed as means of fluorescence intensities (MFI) related to the entire monocyte population, reflecting HLA-DR density per cell.

Statistical analysis

Baseline characteristics were described by frequencies, median, and interquartile range (IQR). Groups were compared by the Mann-Whitney U-test and Pearson’s χ2 test or Fisher’s exact test as appropriate for continuous and categorical variables, respectively. The mHLA-DR results obtained at two time periods (D3-4 and D6-9) were analyzed; the lowest levels per period were selected, and measures after NI were excluded. mHLA-DR expression was stratified according to the best threshold indicated by receiver operating characteristic (ROC) curves at D3-4 and D6-9 (i.e., maximized sensitivity and specificity). For the analysis of NI, follow-up was stopped at 28 days.The incidences of NI among patients according to the thresholds were compared using cumulative incidence curves [S13, S14, S15, S16].The cumulative incidence depends on both the risk of failure attributable to the event of interest (presence of NI) and the risk of failure attributable to 2 competing events (discharge alive and death). A proportional subdistribution hazards model was used to compute the time to the onset of NI and to investigate risk factors for NI. The Fine and Gray extension [S17, S18] of the Cox model was adopted to take the 2 competing risks into account. Adjustments were made for gender, SAPS II, SOFA, intubation and central venous catheterization.Among scores, SAPS II and SOFA were analyzed (Wilcoxon signed-ranks test <0.001). Among device exposures, only intubation and central venous catheterization were analyzed because of high co-linearity of intubation or central venous catheterization with urinary tract catheterization(Mc Nemar test >0.5). Time 0 was the day of septic shock onset.

The model was determined by a backward procedure based on the Wald test. Variables with p≤0.20 were conserved in the model. Adjusted hazard ratios (aHRs) of NI and 95% confidence intervals (95% CI) were calculated.All statistical tests were 2-tailed, and P0.05 was considered to be statistically significant. Analyses were performed by SPSS software (version 11.5, SPSS, Inc., Chicago, IL) and R software [S19].

1

REFERENCES

S1. Landelle C, Lepape A, Francais A, et al. (2008) Nosocomial infection after septic shock among intensive care unit patients. Infect Control Hosp Epidemiol29:1054-1065

S2. Monneret G, Lepape A, Voirin N, et al. (2006) Persisting low monocyte human leukocyte antigen-DR expression predicts mortality in septic shock. Intensive Care Med32:1175-1183

S3. Bone RC, Balk RA, Cerra FB, et al. (1992) Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest101:1644-1655

S4. Wheeler AP, Bernard GR (1999) Treating patients with severe sepsis. N Engl J Med340:207-214

S5. Dellinger RP, Carlet JM, Masur H, et al. (2004) Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Intensive Care Med30:536-555

S6. Vanhems P, Lepape A, Savey A, Jambou P and Fabry J (2000) Nosocomial pulmonary infection by antimicrobial-resistant bacteria of patients hospitalized in intensive care units: risk factors and survival. J Hosp Infect45:98-106

S7. Kumar A, Roberts D, Wood KE, et al. (2006) Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med34:1589-1596

S8. McCabe WA, Jackson GG (1962) Gram negative bacteremia. Arch Intern Med110:847-855

S9. Le Gall JR, Lemeshow S and Saulnier F (1993) A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. Jama270:2957-2963

S10. Vincent JL, Moreno R, Takala J, et al. (1996) The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med22:707-710

S11. HELICS (2005) Surveillance of Nosocomial Infections in Intensive Care Units. Available online at: Accessed Jun 15, 2009

S12. Monneret G, Elmenkouri N, Bohe J, et al. (2002) Analytical requirements for measuring monocytic human lymphocyte antigen DR by flow cytometry: application to the monitoring of patients with septic shock. Clin Chem48:1589-1592

S13.Kalbfleish JD, Prentice RL (1980) Multivariate failure time data and competing risks. In:

Kalbfleish JD, Prentice RL, eds. The statistical analysis of failure time data. New York: Wiley 163-188

S14. Benichou J, Gail MH (1990) Estimates of absolute cause-specific risk in cohort studies. Biometrics 46:813-826

S15. Korn EL, Dorey FJ (1992) Applications of crude incidence curves. Stat Med 11:813-829.

S16. Gray RJ (1988) A class of K-sample tests for comparing the cumulative incidence of a competing risk. Annals of Statistics 16:1141-1154

S17. Alberti C, Metivier F, Landais P, et al. (2003) Improving estimates of event incidence over time in populations exposed to other events: application to three large databases. J Clin Epidemiol 56:536-545

S18. Fine J, Gray RJ (1999) A proportional hazards model for the model for the subdistribution of a competing risk. J Am Stat Assoc 94:496-509

S19. R Development Core Team (2009). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL

Table S1. Baseline characteristics of 153 septic shock patients by development of nosocomial infection. Characteristics were described as frequencies and percentages (%) for categorical data and median and interquartile range (IQR) for continuous variables.

Variable / Total
n=153 / Patient
with NI
n=37 / Patients
Without NI
n=116 / p-value
Baseline descriptors
Gender male / 99 (64.7) / 23 (62.2) / 76 (65.5) / .7
Age (years) / 64.9(49.0-75.5) / 63.3 (52.0-73.8) / 65.4 (47.8-75.8) / .9
Admission categories: / .6
Medical / 84(54.9) / 17 (45.9) / 67 (57.8)
Emergency surgery / 55(35.9) / 16 (43.2) / 39 (33.6)
Scheduled surgery / 7(4.6) / 2 (5.4) / 5 (4.3)
Trauma / 7 (4.6) / 2 (5.4) / 5 (4.3)
At least 1 co-morbidity / 77(50.3) / 18 (48.6) / 59 (50.9) / .8
McCabe and Jackson: / .1
No vital prognosis / 99 (64.7) / 21 (56.8) / 78 (67.2)
Within 5 years / 36 (23.5) / 8 (21.6) / 28 (24.1)
Within 1 year / 18 (11.8) / 8 (21.6) / 10 (8.6)
Septic shock characteristics:
SAPS II at onset of septic shock / 51.0(41.5-62.0) / 54.0 (46.5-65.5) / 49.0(39.0-61.0) / .04
SOFA score atdays 1-2 after onset of shock / 10(8-12) / 11 (9.5-12) / 9(8-12) / .04
Place of acquisition: / .6
Community / 82 (53.6) / 21 (56.8) / 61 (52.6)
Hospital / 59 (38.6) / 12 (32.4) / 47 (40.5)
ICU / 12 (7.8) / 4 (10.8) / 8 (6.9)
Main sites of septic shock: / .6
Pulmonary / 67 (43.8) / 15(40.5) / 52 (44.8)
Intra-abdominal / 52 (34.0) / 15 (40.5) / 37 (31.9)
(continued)
Other / 34 (22.2) / 7 (18.9) / 27 (23.3)
Clinically-suspected / 3 (2.0) / 1 (2.7) / 2 (1.7) / .6
Surgically- or radiologically-documented: / 27 (17.6) / 12 (32.4) / 15 (12.9) / .007
clinical + radiological documentation / 14 (9.2) / 6 (16.2) / 8 (6.9) / .1
clinical + surgical documentation / 13 (8.5) / 6 (16.2) / 7 (6.0) / .08
Microbiologically-documented / 123 (80.4) / 24 (64.9) / 99 (85.3) / .006
Microorganisms
GPC* / 57 (28.5) / 13 (31.7) / 44 (27.7) / .6
GNB* / 87 (43.5) / 18 (43.9) / 69 (43.4) / .9
Anaerobes* / 14 (7.0) / 4 (9.8) / 10 (6.3) / .5
Other bacteria* / 8 (4.0) / 2 (4.9) / 6 (3.8) / .7
Fungi* / 32 (16.0) / 3 (7.3) / 29 (18.2) / .09
Viruses* / 2 (1.0) / 1 (2.4) / 1 (0.6) / .4
Total number of microorganisms / 200 / 41 / 159
mHLA-DR measurements (% of positive monocytes):
Days 3-4 / 37.6 (20.2-52.7) / 26.6 (15.8-44.3) / 39.5 (23.7-54.6) / .01
Days 6-9 / 45.3 (26.7-61.0) / 34.5 (23.3-58.2) / 49.9 (30.3-61.2) / .04
mHLA-DR measurements (means of fluorescence intensities):
Days 3-4 / 59.0 (30.1-89.8) / 39.0 (27.0-61.9) / 65.3 (34.5-93.9) / .008
Days 6-9 / 60.0 (35.8-105.0) / 49.3 (28.0-72.0) / 67.3 (42.2-111.3) / .01
Interventions
Antimicrobial therapy: / .7
Adequate within 24 h / 107 (69.9) / 25 (67.6) / 82 (70.7)
Adequate for more than 24 h / 39 (25.5) / 11 (29.7) / 28 (24.1)
(continued)
Inadequate / 7 (4.6) / 1 (2.7) / 6 (5.2)
Low dose hydrocortisone treatment / 68 (44.4) / 19 (51.4) / 49 (42.2) / .3
Drotrecogin alpha activated treatment / 5 (3.3) / 1 (2.7) / 4 (3.4) / .9
Renal replacement therapy / 35 (22.9) / 14 (37.8) / 21 (18.1) / .01
Intubation / 133 (86.9) / 36 (97.3) / 97 (83.6) / .04
Duration of mechanical ventilation(days) / 10.0 (5.0-16.3) / 12.0 (7.0-20.5) / 9.0 (4.5-15.5) / .06
Central venous catheterization / 149 (97.4) / 37(100) / 112 (96.6) / .6
Duration of central venous catheterization (days) / 12.0 (8.0-21.0) / 21.0 (15.0-28.5) / 10.0 (6.0-16.8) / <.001
Urinary tract catheterization / 142 (92.8) / 35 (94.6) / 107 (92.2) / .9
Duration of urinary tract catheterization (days) / 11.0 (6.0-20.0) / 19.0 (10.0-27.0) / 10.0 (6.0-16.0) / .001
Outcomes
Length of hospital stay after septic shock onset (days) / 29.0(14.5-35.0) / 29.0 (24.5-51.0) / 29.0 (12.3-30.0) / .07
Length of ICU stay after septic shock onset (days) / 13.0(8.0-26.0) / 27.0 (16.5-39.0) / 11.0 (6.3-19.0) / <.001
ICU death rate within 28 days of septic shock onset / 51 (33.3) / 14 (37.8) / 37 (31.9) / .5
ICU crude death rate / 58 (37.9) / 20 (54.1) / 38 (32.8) / .02
Hospital crude death rate / 70 (45.8) / 22(59.5) / 48 (41.4) / .06
*Percentages by number of isolates.

NOTES. NI: Nosocomial infections; SAPS II: Simplified Acute Physiology Score II; SOFA: Sepsis-related Organ Failure Assessment; mHLA-DR: monocyte human leucocyte antigen-DR; ICU: Intensive care unit; GPC: Gram-positive cocci; GNB: Gram-negative bacilli.

Groups were compared by Pearson’s χ2 test or Fisher’s exact test as appropriate and the Mann-Whitney U-test for categorical and continuous variables, respectively.

1