Utility of Biomarkers in the Evaluation of Fever in the Intensive Care Unit After Brain Injury
Utility of Biomarkers in the Evaluation of Fever in the Intensive Care Unit After Brain Injury
Umer Mukhtar 1
Umer Shoukat 1
M. Kamran Athar 1
Fred Rincon 0 1
0 Department of Neurology, Thomas Jefferson University , Philadelphia, PA , USA
1 Department of Neurological Surgery, Thomas Jefferson University , Philadelphia, PA , USA
Fever is frequent in patients with neurologic injury. Differentiating infectious fever from central fever can be challenging. It is important to diagnose the cause of fever in the neurological intensive care unit because of the detrimental effects of fever on brain injured patients. This is a comprehensive review of the role of the two commonly available biomarkers, C-reactive protein and procalcitonin in differentiating the central fever from infectious fever. Fever is frequently seen in the neurologic intensive care unit (NICU). Incidence rates of up to 70% have been reported in various studies.1-5 Fever can help host defenses by local activation of the coagulation cascade, cytokine-mediated T-cell activation, as well as neutrophil and macrophage recruitment to injured tissues. In brain injured patients, after the initial insult, secondary neuronal injury is speculated to be caused by several processes including mitochondrial dysfunction, inflammatory response, free radical generation, and excitatory neurotransmitter release. Fever has been shown to exacerbate secondary neuronal injury and physiologic dysfunction after traumatic brain injury (TBI), subarachnoid hemorrhage (SAH), and major neurosurgery.6
Proclacitonin; C reactive protein; Subarachnoid hemorrhage; Intracerebral hemorrhage
INTRODUCTION AND BACKGROUND
CAUSES OF FEVER IN NEUROLOGICALLY INJURED PATIENTS
Infections are the most common cause of fever in the NICU population, accounting
for at least half of the febrile episodes.2,4-7 A significant percentage of patients in the
NICU have central fevers. Central fever results from loss of physiological regulation of
body temperature by the hypothalamus.7 The diagnosis of central fever is challenging
as there is no clear definition or diagnostic criteria. In addition, the prevalence of
systemic inflammatory response syndrome (SIRS) and leukocytosis may be similar in
patients with both central and infectious fevers.8 When there is clinical suspicion of
infection, it is common practice to initiate broad spectrum antibiotics after obtaining
appropriate culture specimens in ICU patients. This practice results in another
challenge for neurointensivists as many patients with brain injury have central fever and
antibiotics may be continued unnecessarily in these patients. This may result in the
emergence of antibiotic resistant organisms, increase side effects of antibiotics as well
as healthcare costs. A better understanding of predictors of central fever is important
as it will help in antibiotic stewardship and may also allow for earlier discontinuation
of antibiotics in patients with central fever.2 Rabinstein et al identified a number of
variables that are predictive of central fever in the neurocritical care population. Their
study showed that in patients with onset of fever within 72 hours of admission, patients
with persistent fevers, negative cultures and lack of infiltrates on chest X-ray were
more likely to have central fever. This was especially true if their primary diagnosis was
SAH, tumor or they had intraventricular
In addition to these variables, a biomarker
predictive of sepsis can be helpful in
differentiating infectious from central
fever and can also aid in antibiotic
Overview of Biomarkers
Ideally, in order to guide antibiotic use,
clinicians need a valid, reliable and
readily available test that would
correlate well with their clinical suspicion and
also help them in differentiating between
central and infectious fever. A number of
biomarkers have been studied in patients
with sepsis, some of which have also
been evaluated in brain injured patients.
One of the more promising ones is
procalcitonin (PCT). Other biomarkers
include C reactive protein (CRP) and
Interleukin-6 (IL-6). These biomarkers
and their validity, efficacy and availability
in intensive care unit are discussed under.
Procalcitonin (PCT) is a glycopeptide
consisting of 116 amino acids produced
under normal conditions in the C cells of
the thyroid gland as the precursor
molecule of calcitonin. Several studies have
demonstrated that PCT levels are raised
in severe invasive bacterial infections and
decrease rapidly after appropriate
antibiotic therapy.9-14 In contrast, PCT levels
are normal or only slightly increased in
localized bacterial infections, viral
infections, and inflammatory reactions of
noninfectious origin.9,15-17 There is recent
evidence suggesting that PCT can
distinguish sepsis from non-infectious SIRS in
general critical care patients, allowing
clinicians to make better diagnostic and
Procalcitonin vs. Other Biomarkers
A number of studies have evaluated the
role of PCT in comparison with CRP
and IL-6 as a diagnostic and prognostic
biomarker. In a prospective cohort study
Ville Pettilä et al evaluated the predictive
value of PCT and IL-6 in patients with
suspected sepsis. PCT and IL-6 values on
day 2 of suspected sepsis were
independently predictive of hospital mortality.22
Simon et al conducted a systemic
review and meta-analysis. They studied
the relation between serum PCT and
CRP levels as a marker of inflammation.
PCT had higher accuracy than CRP for
discriminating bacterial infections from
non-infectious causes of inflammation.
In addition PCT was also significantly
better than CRP in differentiating
bacterial from viral infections.23
Choi et al conducted a prospective study
to evaluate the predictive performance
of serum PCT as a differentiating marker
between postoperative bacterial
meningitis (PBM) and postoperative aseptic
meningitis (PAM). For the diagnosis of
PBM, PCT level ≥ 0.15 ng/mL had a
specificity of 80.0% and sensitivity of 50%.
The combined criteria of a CRP level
≥2.5 mg/dL, WBC count ≥ 9,500/mm3,
and PCT level ≥ 0.15 ng/mL had the
highest specificity of 92.6% and higher
sensitivity of 85.7%. They concluded that
PCT alone has a limited performance for
the diagnosis of PBM, but has improved
diagnostic value when used as an adjunct
test with other inflammatory markers.24
Procalcitonin as a biomarker for diagnosis of sepsis in brain injured patients
Fever is common in brain injured
patients. As stated earlier, a significant
proportion of these patients have
central fever and differentiating central
from infectious fever is challenging. In a
prospective observational study, Festic et
al showed that PCT may be a useful tool
when managing SIRS in a patient with
aneurysmal subarachnoid hemorrhage
(aSAH). They found that in these patients,
serum PCT values have high specificity,
high negative predictive value and good
overall predictive utility for infections,
particularly for major infections. Patients
with infection were > 25 times more likely
to have an elevated PCT values compared
to those with no infection. For patients
with a major infection, the odds ratio was
even higher (>33).25
Early diagnosis of bacterial or viral
meningitis is important so that antibiotic
treatment can be started without delay.
For immediate diagnosis of bacterial
meningitis, the sensitivity of direct
cerebrospinal fluid (CSF) examination or the
detection of bacterial antigens in CSF is
Viallon et al conducted a prospective
study to determine the ability of
inflammatory biomarkers commonly used
for the diagnosis of acute meningitis
to differentiate between bacterial and
viral meningitis, in adult patients with a
negative CSF examination. Out of 254
patients with meningitis with a negative
direct CSF examination, 35 had bacterial
meningitis and 181 had viral meningitis.
Serum PCT was a highly discriminative
biomarker and had a sensitivity of 95%, a
specificity of 100%, a negative predictive
value of 100%, and a positive predictive
value of 97% at a diagnostic cut-off level
of 0.28 ng/ml .30
In another prospective study, Berger et
al measured daily PCT levels in patients
requiring temporary external ventricular
drains (EVD). They showed that PCT
levels were significantly higher (4.7 vs 0.2
ng/ml) in patients with proven cerebral
ventriculitits. CSF cell count could not
differentiate bacterial infections from
Martinez et al measured serum PCT in
15 consecutive patients with ventriculitis
in which an EVD had been inserted and
compared the data with ten patients
who had bacterial meningitis. Four out
of fifteen patients had
microbiologically proven bacterial ventriculitis with
positive bacterial cultures. PCT value of
1.0 ng/ml showed a specificity of 77%
and sensitivity of 68% in patients with
ventriculitis with positive CSF bacterial
In a prospective case series Schwarz et al
compared serum PCT levels in patients
with bacterial meningitis to those with
abacterial meningitis. At admission, PCT
levels were significantly higher in patients
with bacterial meningitis as compared
with those with abacterial meningitis (p
< .001). The specificity of PCT was 100%
for bacterial infections, but there were
false-negative findings in five patients
with bacterial meningitis (a sensitivity of
69%). Persistently elevated or increasing
PCT levels after 2 days were associated
with an unfavorable clinical course.33
In a case series comparing 7 patients
with Neuro-Behcet disease to 3 patients
with bacterial meningitis, Suzuki et al
showed that serum PCT levels were
increased in patients with bacterial
meningitis, but not in those with
NeuroBehcet disease. Therefore, serum PCT
may be a useful marker for
discrimination between Neuro-Behcet disease and
septic meningitis, especially in cases of
the meningeal form of Neuro-Behcet
Schwarz et al. >0.5 (33)
Martinez et al. 1 (32) 0.2 0.28
However, only CRP elevation remained
associated with mortality in the
multivariate model, after adjusting for multiple
The review of the studies related to
CRP show that it is raised universally
in all kinds of inflammatory responses,
whether infectious or noninfectious. It
may be useful in addition to PCT, but it
cannot be solely used to differentiate
between infectious and noninfectious
fever. As a result its utility for antibiotic
stewardship in NICU is limited.
CONCLUSIONS AND FUTURE
The development of fever in critically ill
patient needs immediate attention and
action to rule out infection. This situation
is more challenging in the NICU patient
population due to the high rate of
noninfectious fever. On one hand there is high
morbidity and mortality associated with
sepsis, but on the other hand use of
broad spectrum antibiotics in patients
who do not have infection results in high
rates of antibiotic resistance, infections,
and adverse drug reactions. In order to
help diagnosis, a test is definitely needed
that can differentiate between
infectious and noninfectious fever. If a test
helps us rule out infection and diagnose
central fever correctly, then in spite of
unnecessary use of antibiotics, we can
use novel techniques such as surface or
intravascular cooling measures like artic
sun to treat hyperthermia in brain injured
We conclude from the above review
of literature that PCT may be a more
sensitive and specific test that can help
us differentiate between infectious and
central fever. In addition it can also be
a useful test to differentiate between
primary bacterial CNS infections vs other
types of CNS infections or non-infectious
CNS inflammatory processes. However,
data is very limited in neurocritical and
neurosurgical patient populations.
More clinical studies and clinical trials
are needed that can validate the use of
PCT as a diagnostic test to differentiate
between infectious and noninfectious
fever. In addition, the diagnostic cut-off
levels of PCT also need to be validated.
1. Occurrence of potentially detrimental
temperature alterations in hospitalized
patients at risk for brain injury. Albrecht
RF 2nd, et al. Mayo Clin Proc. 1998
2. Hyperthermia in the neurosurgical intensive
care unit. Kilpatrick MM, et al. Neurosurgery.
3. Pyrexia in head-injured patients admitted to
intensive care. Stocchetti N, et al. Intensive
Care Med. 2002 Nov;28(
5. Elevated body temperature independently
contributes to increased length of stay in
neurologic intensive care unit patients.
Diringer MN, et al. Crit Care Med. 2004
6. Patterns of increased intracranial pressure
after severe traumatic brain injury. O’Phelan
KH, et al. Neurocrit Care. 2009;10(3):280-6.
7. Non-infectious fever in the neurological
intensive care unit: incidence, causes
and predictors. Rabinstein AA, et al.
J Neurol Neurosurg Psychiatry. 2007
8. Indicators of central fever in the neurologic
intensive care unit. Hocker SE, et al. JAMA
Neurol. 2013 Dec;70(
17. Cytokines, nitrite/nitrate, soluble tumor
necrosis factor receptors, and procalcitonin
concentrations: comparisons in patients with
septic shock, cardiogenic shock, and bacterial
pneumonia. de Werra I, et al. Crit Care Med.
18. Efficacy of procalcitonin in the early
diagnosis of bacterial infections in a critical
care unit. Nakamura A, et al. Shock. 2009
Monthly Epilepsy Support Groups
919 Walnut Street, Suite 700
Philadelphia, PA 19107
4. Risk factors for fever in the neurologic intensive care unit . Commichau C, et al. Neurology. 2003 Mar 11 ; 60 ( 5 ): 837 - 41 .
9. High serum procalcitonin concentrations in patients with sepsis and infection . Assicot M, et al. Lancet. 1993 Feb 27 ; 341 ( 8844 ): 515 - 8 .
10. Elevated serum procalcitonin levels in patients with melioidosis . Smith MD, , et al. Clin Infect Dis . 1995 Mar; 20 ( 3 ): 641 - 5 .
11. Procalcitonin as a marker for the early diagnosis of neonatal infection . Gendrel D, et al. J Pediatr . 1996 Apr; 128 ( 4 ): 570 - 3 .
12. Procalcitonin in patients with and without immunosuppression and sepsis . al -Nawas B , et al. Infection . 1996 Nov-Dec; 24 ( 6 ): 434 - 6
13. Reliability of procalcitonin concentrations for the diagnosis of sepsis in critically ill neonates . Chiesa C, et al. Clin Infect Dis . 1998 Mar; 26 ( 3 ): 664 - 72 .
14. Procalcitonin and C-reactive protein levels in neonatal infections . Monneret G, et al. Acta Paediatr . 1997 Feb; 86 ( 2 ): 209 - 12 .
15. Usefulness of procalcitonin for differentiation between activity of systemic autoimmune disease (systemic lupus erythematosus/ systemic antineutrophil cytoplasmic antibodyassociated vasculitis) and invasive bacterial infection . Eberhard OK , et al. Arthritis Rheum . 1997 Jul; 40 ( 7 ): 1250 - 6 .
16. The potential role of procalcitonin and interleukin 8 in the prediction of infected necrosis in acute pancreatitis . Rau B , et al. Gut . 1997 Dec; 41 ( 6 ): 832 - 40 .
19. Diagnostic value of procalcitonin, interleukin-6, and interleukin-8 in critically ill patients admitted with suspected sepsis. Harbarth S, et al . Am J Respir Crit Care Med . 2001 Aug 1 ; 164 ( 3 ): 396 - 402
20. Calcitonin precursors are reliable markers of sepsis in a medical intensive care unit . Müller B , et al. Crit Care Med . 2000 Apr; 28 ( 4 ): 977 - 83 .
21. Use of procalcitonin for early prediction of lethal outcome of postoperative sepsis . Novotny A , et al. Am J Surg . 2007 Jul; 194 ( 1 ): 35 - 9 .
22. Predictive value of procalcitonin and interleukin 6 in critically ill patients with suspected sepsis . Pettilä V , et al. Intensive Care Med . 2002 Sep; 28 ( 9 ): 1220 - 5 .
23. Serum procalcitonin and C-reactive protein levels as markers of bacterial infection: a systematic review and meta-analysis . Simon L, et al. Clin Infect Dis . 2004 Jul 15 ; 39 ( 2 ): 206 - 17 .
24. Predictive performance of serum procalcitonin for the diagnosis of bacterial meningitis after neurosurgery . Choi SH,. Infect Chemother . 2013 Sep; 45 ( 3 ): 308 - 14 .
25. The utility of serum procalcitonin in distinguishing systemic inflammatory response syndrome from infection after aneurysmal subarachnoid hemorrhage . Festic E, et al. Neurocrit Care . 2014 Jun; 20 ( 3 ): 375 - 81 .
26. The spinal tap: a new look at an old test . Marton K , et al. Ann Intern Med . 1986 Jun; 104 ( 6 ): 840 - 8 .
27. Clinical features and prognostic factors in adults with bacterial meningitis . van de Beek D , et al. N Engl J Med . 2004 Oct 28 ; 351 ( 18 ): 1849 - 59 .
28. Differential diagnosis of acute meningitis. An analysis of the predictive value of initial observations. Spanos A, et al . JAMA. 1989 Nov 17 ; 262 ( 19 ): 2700 - 7 .
29. High sensitivity and specificity of serum procalcitonin levels in adults with bacterial meningitis . Viallon A , et al. Clin Infect Dis . 1999 Jun; 28 ( 6 ): 1313 - 6 .
30. Meningitis in adult patients with a negative direct cerebrospinal fluid examination: value of cytochemical markers for differential diagnosis . Viallon A , et al. Crit Care . 2011 ; 15 ( 3 ): R136 .
31. Serum procalcitonin in cerebral ventriculitis . Berger C, et al. Crit Care Med . 2002 Aug; 30 ( 8 ): 1778 - 81 .
32. Serum procalcitonin monitoring for differential diagnosis of ventriculitis in adult intensive care patients . Martínez R , et al. Intensive Care Med . 2002 Feb; 28 ( 2 ): 208 - 10 .
33. Serum procalcitonin levels in bacterial and abacterial meningitis . Schwarz S, et al. Crit Care Med . 2000 Jun; 28 ( 6 ): 1828 - 32 .
34. Procalcitonin might help in discrimination between meningeal neuro-Behçet disease and bacterial meningitis . Suzuki N , et al. Neurology. 2009 Feb 24 ; 72 ( 8 ): 762 - 3 .
35. Preoperative C- reactive protein predicts the need for repeated intracerebral brain abscess drainage . Neidert MC , et al. Clin Neurol Neurosurg . 2015 Apr; 131 : 26 - 30 .
36. Significance of C-reactive protein and transcranial Doppler in cerebral vasospasm following aneurysmal subarachnoid hemorrhage . Hwang SH , et al. J Korean Neurosurg Soc . 2013 Oct; 54 ( 4 ): 289 - 95 .
37. C-reactive protein and vasospasm after aneurysmal subarachnoid hemorrhage . Romero FR , et al. Acta Cir Bras . 2014 May; 29 ( 5 ): 340 - 5 .
38. C-reactive protein level measurement improves mortality prediction when added to the spontaneous intracerebral hemorrhage score . Di Napoli M , et al. Stroke . 2011 May; 42 ( 5 ): 1230 - 6 .