TY - JOUR
TI - Sepsis severity is the major determinant of circulating thrombopoietin
levels in septic patients
AU - Zakynthinos, SG
AU - Papanikolaou, S
AU - Theodoridis, T
AU - Zakynthinos,
AU - EG
AU - Christopoulou-Kokkinou, V
AU - Katsaris, G
AU - Mavrommatis, AC
JO - Pediatric Critical Care Medicine
PY - 2004
VL - 32
TODO - 4
SP - 1004-1010
PB - Lippincott, Williams & Wilkins
SN - 1529-7535
TODO - 10.1097/01.CCM.0000121433.61546.A0
TODO - sepsis; septic shock; thrombopoietin; platelets; disseminated
intravascular coagulation; sepsis markers
TODO - Objective: To measure serum thrombopoietin levels and to investigate
their relationship with platelet counts and other potential determinants
in septic patients.
Design: Prospective study comparing septic patients and healthy
volunteers.
Setting: General intensive care units in two tertiary university
hospitals.
Patients: A total of 152 consecutive septic patients (69 with sepsis, 24
with severe sepsis, and 59 with septic shock). Twenty-two healthy
volunteers served as control subjects. Sepsis severity was determined by
grading septic patients in those having sepsis, severe sepsis, and
septic shock.
Interventions. None.
Measurements and Main Results. After blood sampling, platelet counts,
and serum thrombopoietin, interleukin-6 and C-reactive protein levels
were measured. Platelets did not decrease in patients with sepsis, but
they significantly decreased in patients with severe sepsis and septic
shock (p <.01 vs. controls and sepsis). In contrast, thrombopoietin
levels (median [range]) increased in patients with sepsis (159
[34-1272] pg/mL) compared with controls (57 [33-333] pg/mL, p
<.001), exhibiting further significant increase in patients with severe
sepsis and septic shock (461 [73-1550] and 522 [45-2313] pg/mL,
respectively, p <.001 vs. sepsis). In multiple regression analysis,
thrombopoietin levels were independently related only to sepsis severity
(higher in patients with increased sepsis severity, p <.001) and
platelet counts (higher in patients with lower platelet counts, p
=.004). Sepsis severity accounted for most of the variance explained by
the model. Thrombopoietin was significantly related to interleukin-6
(r(2) =.26) and C-reactive protein (r(2) =.37, p <.001 for both). In
serial measurements, interleukin-6 peak values constantly preceded those
of thrombopoietin, whereas peaks in thrombopoietin levels coincided with
clinical episodes of septic shock.
Conclusions: Sepsis severity is the major determinant of elevated
thrombopoiefin levels in septic patients, whereas platelet count is a
secondary determinant. Thrombopoietin represents a potential marker of
sepsis severity.
ER -