Available online http://ccforum.com/content/13/2/127
Page 1 of 2
(page number not for citation purposes)
In a recent study conducted in a cohort of 52 septic patients,
Monserrat and coworkers found that profound failure of peripheral
T cells to convert from a naïve phenotype to an activated pheno-
type has positive predictive value in identifying patients who do not
recover. These data support the hypothesis that failure of the
innate immune system to engage the T-cell compartment contri-
butes to sepsis mortality and provides motivation for the develop-
ment and clinical evaluation of immunostimulatory therapies for
patients with sepsis.
Clinical scoring systems stratify patients on a population
basis, but they do not have sufficient sensitivity and
specificity to answer fundamental questions about individuals.
For instance, is the patient infected, responding to therapy, or
likely to recover? In this issue, Monserrat and coworkers 
report their examination of surface marker expression on peri-
pheral T cells over a 28-day period (intensive care unit [ICU]
admission through follow up) in 52 patients with septic shock
and 36 age-matched and sex-matched healthy control individ-
uals. Consistent with previous reports, this cohort of septic
patients had significantly depressed T-cell numbers on ICU
admission that persisted for weeks [2,3]. Monserrat and
coworkers retrospectively classified patients as either sur-
vivors or nonsurvivors and showed that, with the exception of
injury severity assessments (Acute Physiology and Chronic
Health Evaluation II, Multiple Organ Dysfunction Score,
Sepsis-related Organ Failure Assessment), these two groups
of patients were similar on enrollment.
Although all septic patients had similarly depressed numbers
of CD4+and CD8+lymphocytes, serial profiling of the
absolute abundance of naïve lymphocytes with different
phenotypic markers revealed significant differences between
survivors and nonsurvivors. The number of cells expressing
surface markers of the naïve phenotype was significantly
higher in nonsurvivors than in survivors.
Other biomarker studies of ICU populations have examined
additional patient populations (critically ill nonseptic patients,
and patients with sepsis and severe sepsis) as well as healthy
control individuals and patients with septic shock. The
comparison between septic shock and health reveals the
most profound differences, and it is easiest to power a study
with just those two groups. However, it would be valuable to
know whether Monserrat and coworkers’ leukocyte
phenotype has utility outside the diagnosis of septic shock.
No standard critical care protocol was followed for this study,
and it is possible that current patient management regimens
do not effectively treat the nonsurvivor patient cohort.
These findings add to the body of leukocyte phenotyping
studies that aim to differentiate patients with syndromic
diagnoses that are common in the ICU setting [2-9]. Results
from single-center studies have illustrated that there are
multiple leukocyte phenotypes with potential clinical utility for
either diagnosis or prognostication. Monserrat and coworkers
 found that the best receiver operator characteristic (ROC)
curve based on a single variable was obtained with data
obtained from ICU admission, and a cut-off of 136
CD8+CD28+T-cells per microliter exhibited 70% sensitivity
and 100% specificity for predicting death (area under the
curve [AUC] = 0.84).
In earlier work, Saenz and coworkers  identified a monocyte
phenotype (total number of monocytes per microliter [= 445]
+ percentage of monocytes that were CD13/HLA-DR
positive [32%]) that distinguished survivors from non-
survivors in patients with severe sepsis (AUC = 0.918). By
combining two informative variables, those investigators
extracted additional power from their data. Tschaikowsky and
coworkers  have already shown that lymphocyte and
monocyte markers exhibit concordant changes in a cohort of
patients with severe sepsis. It would be valuable to know
Leukocyte phenotyping to stratify septic shock patients
Jonathan E McDunn and Richard S Hotchkiss
Department of Anesthesiology, Washington University School of Medicine, 660 South Euclid Avenue, Saint Louis, MI 63110, USA
Corresponding author: Richard S Hotchkiss, firstname.lastname@example.org
See related research by Monserrat et al., http://ccforum.com/content/13/1/R26
Published: 3 April 2009
This article is online at http://ccforum.com/content/13/2/127
© 2009 BioMed Central Ltd
Critical Care 2009, 13:127 (doi:10.1186/cc7748)
AUC = area under the curve; ICU = intensive care unit; ROC = receiver operating characteristic.
Critical Care Vol 13 No 2McDunn and Hotchkiss
Page 2 of 2
(page number not for citation purposes)
whether combining the positive predictors from the study by
Monserrat and coworkers , or whether combining
lymphocyte and monocyte phenotyping in the same study,
would improve the ROC.
This study raises numerous questions. If survivors are
responding to therapy and recovering, then why was the
ROC performance best at ICU admission? Is it because of
loss of individuals from the nonsurvivor cohort, reducing
statistical power at subsequent time points, or is the immune
compartment in nonsurvivors responding to therapy even
though these patients die? Additional demographic informa-
tion would be valuable to assist in interpreting these data,
especially regarding existing co-morbid conditions and the
admitting diagnosis. Based on the results of interventional
trials, it is clear that the type of patient and ICU can influence
the efficacy or utility of a test or intervention (for example,
intensive insulin therapy has different effects on outcome
depending on whether it is instituted in a surgical or medical
ICU setting [10,11]). If leukocyte phenotyping conclusively
identifies a group of patients who cannot be rescued by
current care, then how should physicians use this information?
From a mechanistic standpoint, these data add support to the
hypothesis that failure of the innate immune system to engage
the T-cell compartment plays a significant role in sepsis
mortality  and suggest that immune-supportive or immuno-
stimulatory therapies may be of value in the management of
the septic patient . Further development of a leukocyte
phenotyping-based test for ICU patient stratification beyond
clinical scoring could aid in both diagnosis and prognosis and,
if validated in a prospective patient cohort, could be
incorporated into both clinical practice and clinical trial design.
The authors declare that they have no competing interests.
1. Monserrat J, De Pablo R, Reyes E, Diaz D, Barcenilla H, Zapata
MR, De La Hera A, Prieto A, Alvarez-Mon M: Clinical relevance of
the severe abnormalities of the T cell compartment in septic
shock patients. Crit Care 2009, 13:R26.
2.Lin RY, Astiz ME, Saxon JC, Rackow EC: Altered leukocyte
immunophenotypes in septic shock. Studies of HLA-DR,
CD11b, CD14, and IL-2R expression. Chest 1993, 104:847-
3. Cheadle WG, Pemberton RM, Robinson D, Livingston DH,
Rodriguez JL, Polk HC Jr: Lymphocyte subset responses to
trauma and sepsis. J Trauma 1993, 35:844-849.
4. Sáenz JJ, Izura JJ, Manrique A, Sala F, Gaminde I: Early progno-
sis in severe sepsis via analyzing the monocyte immunophe-
notype. Intensive Care Med 2001, 27:970-977.
5. Tschaikowsky K, Hedwig-Geissing M, Schiele A, Bremer F, Schy-
walsky M, Schuettler J: Coincidence of pro- and anti-inflamma-
tory responses in the early phase of severe sepsis:
longitudinal study of mononuclear histocompatibility leuko-
cyte antigen-DR expression, procalcitonin, C-reactive protein,
and changes in T-cell subsets in septic and postoperative
patients. Crit Care Med 2002, 30:1015-1023.
6. Holub M, Klucková Z, Helcl M, Príhodov J, Rokyta R, Beran O:
Lymphocyte subset numbers depend on the bacterial origin
of sepsis. Clin Microbiol Infect 2003, 9:202-211.
7. Roth G, Moser B, Krenn C, Brunner M, Haisjackl M, Almer G,
Gerlitz S, Wolner E, Boltz-Nitulescu G, Ankersmit HJ: Suscepti-
bility to programmed cell death in T-lymphocytes from septic
patients: a mechanism for lymphopenia and Th2 predomi-
nance. Biochem Biophys Res Commun 2003, 308:840-846.
Faivre V, Lukaszewicz AC, Alves A, Charron D, Payen D, Haziot A:
Accelerated in vitro differentiation of blood monocytes into
dendritic cells in human sepsis. Clin Exp Immunol 2007, 147:
Schwulst SJ, Muenzer JT, Chang KC, Brahmbhatt TS, Cooper-
smith CM, Hotchkiss RS: Lymphocyte phenotyping to distin-
guish septic from nonseptic critical illness. J Am Coll Surg
10. van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyn-
inckx F, Schetz M, Vlasselaers D, Ferdinande P, Lauwers P, Bouil-
lon R: Intensive insulin therapy in critically ill patients. N Engl J
Med 2001, 345:1359-1367.
11. van den Berghe G, Wilmer A, Hermans G, Meersseman W,
Wouters PJ, Milants I, Van Wigngaerden E, Bobbaers H, Bouillon
R: Intensive insulin therapy in the medical ICU. N Engl J Med
12. Marshall JC, Charbonney E, Gonzalez PD: The immune system
in critical illness. Clin Chest Med 2008, 29:605-616.
13. Lyn-Kew K, Standiford TJ: Immunosuppression in sepsis. Curr
Pharm Des 2008, 14:1870-1881.