ORIGINAL ARTICLE Infect Dis Clin Microbiol 2024; 6(3): 206-15

Changes of Procalcitonin Kinetics According

to Renal Clearance in Critically Ill Patients
with Primary Gram-Negative Bloodstream
Infections
Hasan Selçuk Özger1 , Şeref Kerem Çorbacıoğlu2 , Nazlıhan Boyacı-Dündar3 , Mehmet Yıldız1 , Özant Helvacı4 ,
Fatma Betül Altın1 , Melda Türkoğlu3 , Gülbin Aygencel3 , Murat Dizbay1

1 Infectious Disease and Clinical Microbiology, Gazi University School of Medicine, Ankara, Türkiye
2 Emergency Departments, Atatürk Sanatoryum Training and Research Hospital, Ankara, Türkiye
3 Department of Internal Medicine, Division of Intensive Care Medicine, Gazi University School of Medicine, Ankara, Türkiye
4 Department of Nephrology, Gazi University School of Medicine, Ankara, Türkiye

ABSTRACT
Objective: This study aimed to investigate the relationship between procalcitonin (PCT)
kinetic and estimated glomerular filtration rates (eGFR) in critically ill patients who had
Gram-negative primary bloodstream infection (GN-BSI) and responded to the antimicrobial
therapy.
Materials and Methods: This single-centered study was retrospective and observational.
Critically ill GN-BSI patients over 18 years old who had clinical and microbiological re- Corresponding Author:
sponses to antibiotic treatment were included in the study. Patients were divided into two Hasan Selçuk Özger
groups according to eGFR (eGFR <30 mL/min/1.73m2 and ≥30 mL/min/1.73m2) and com-
E-mail:
pared for PCT kinetic at seven different measurement points as initial, first, third, fifth,
[email protected]
seventh, tenth, and fourteenth days.
Results: The study included 138 patients. Initial PCT levels were higher in patients with Received: May 03, 2024
eGFR <30 mL/min/1.73m2 (4.58 [1.36-39.4] ng/mL) than in eGFR ≥30 mL/min/1.73m2 (0.91 Accepted: July 26, 2024
Published: September 26, 2024
[0.32-10.2]) (p<0.001). This elevation was present at all measurement points (p<0.05). The
decrease in PCT values by ≥30% (26.0% vs 47.9%; p=0.024) on the third day and ≥50% (69.2% Suggested citation:
vs 76.6%; p=0.411) on the fifth day was less in the low eGFR (<30 mL/min/1.73m2) group. The Özger HS, Çorbacıoğlu ŞK,
effect of low GFR on serum PCT kinetic was present in both fermenter and non-fermenter Boyacı-Dündar N, Yıldız M,
GN-BSIs but was more prominent in the fermenter group. Helvacı Ö, Altın FB, et al.
Changes of procalcitonin kinetics
Conclusion: Serum PCT levels during therapy were higher in patients with low eGFR. Early
according to renal clearance in
PCT (<5 days) response was not obtained in non-fermenter GN-BSI patients with low eGFR. critically ill patients with primary
Antibiotic revision decisions should be made more carefully in patients with low eGFR due Gram-negative bloodstream
to high initial PCT levels and slow PCT kinetic. infections. Infect Dis Clin
Microbiol. 2024;3:206-15.
Keywords: procalcitonin, kinetics, renal clearance, critically ill patients
DOI: 10.36519/idcm.2024.363

This work is licensed under a Creative Commons Attribution-Non Commercial 4.0 International License. 206
Procalcitonin Kinetics in Gram-Negative Bloodstream Infections

INTRODUCTION clearance. Besides renal failure, other confounding

factors such as patients’ comorbidities, source of in-

P
rocalcitonin (PCT) is a biomarker frequently fection, infectious agents, and severity of infection
used in the diagnosis and prognosis of infec- also affect PCT kinetics and increase the uncertainty
tion (1-4). The lack of optimum cut-off values between renal clearance and PCT kinetics (13).
and the high serum PCT levels in some non-infec-
tious conditions restrict its clinical use (3-5). There- This study evaluated the relationship between
fore, PCT use is not recommended regardless of PCT kinetic and estimated glomerular filtration
clinical parameters (6). This limitation in clinical rates (eGFR) in critically ill patients with primary
use has led to PCT kinetic evaluations rather than Gram-negative primary bloodstream infections
predictions based on single PCT measurements. (GN-BSIs) who had clinical and microbiological re-
The clinicians use PCT kinetics to predict the op- sponses to antibiotic therapy.
timal duration of antibiotic therapy without worse
clinical outcomes (5). However, due to heteroge- MATERIALS AND METHODS
neous patient groups and confounding factors on
PCT kinetics, there are different published results This study was a single-center, retrospective, and
and recommendations for the effect of PCT on the observational-analytic study conducted at Gazi Uni-
optimal antibiotic therapy duration (7, 8). versity Hospital between January 2019 and Decem-
ber 2022. It included all adult patients (≥18 years)
Acute or chronic renal failure is one of the main with GN-BSIs who had clinical and microbiological
confounders of PCT kinetics (9, 10). In renal failure, responses to appropriate antibiotic therapy. Patients
altered immune responses to infection, ongoing in- with breakthrough bloodstream infections (BSIs)
flammation, and altered PCT clearance affect serum during antibiotic therapy, patients who had an anti-
PCT levels. Compared to patients with normal renal biotic change due to a proven or possible secondary
clearance, the PCT values ​​accepted for the diagnosis infection in any other site or therapy failure, and pa-
of infection are higher than the standard PCT lim- tients without follow-up PCT were excluded.
it. However, the relationship between renal clear-
ance, PCT kinetic, and its clinical effects remains Renal function was assessed according to eGFR lev-
unknown. The acceptable PCT decreases cut-off els. eGFR was calculated based on creatinine, age,
and time interval for PCT response in these patients gender and race via the Chronic Kidney Disease
are unclear (11, 12). This uncertainty and more vari- Epidemiology Collaboration (CKD-EPI) 2009 calcu-
able PCT kinetic may lead to unnecessary antibiot- lator (14). Patients were divided into two groups ac-
ic revisions and therapy extensions in renal failure cording to their eGFR levels. Clinical response was
patients compared to patients with normal renal defined as the resolution of infection signs (such
as fever, hypotension, and leukocytosis). Microbial
response was defined as sterilization of blood cul-
ture under antibiotic therapy (15). Breakthrough
HIGHLIGHTS BSIs were defined as isolating the same or different
• Baseline, follow-up and final serum procalcitonin microorganisms in blood culture during antibiotic
(PCT) levels were higher in patients with low esti- therapy. Therapy failure was defined as the need
mated glomerular filtration rates (eGFR). to revise antibiotic therapy due to a lack of clinical
• PCT kinetic was slow in patients with low GFR and microbiological response. According to antimi-
levels, especially in non-fermenter Gram-nega- crobial susceptibility testing, appropriate antibiotic
tive primary bloodstream infections. treatment was defined as using at least one effec-
• Early PCT evaluation was not suitable for PCT re- tive antibiotic as empirical or definitive treatment.
sponse in patients with low GFR levels.
• PCT response should be delayed until the fifth Age, gender, comorbid diseases, Charlson comorbidi-
day of antibiotic therapy in these patients. ty index (CCI), systemic inflammatory response syn-
drome (SIRS), sequential organ failure assessment

Özger HS et al. 207

 Infect Dis Clin Microbiol 2024; 6(3): 206-15

Table 1. Demographic and clinical characteristics of patients.

eGFR <30 eGFR <30

mL/min/1.73m2 mL/min/1.73m2 p
n=61 n=77

Age, median (IQR: 25-75%) 75 (62-82) 65 (53-77) 0.011

Gender (male), n (%) 27 (44.3) 41 (53.2) 0.294

Comorbidities n (%)

Chronic hypertension 26 (33.8) 38 (62.3) 0.001

Diabetes mellitus 12 (15.6) 22 (36.1) 0.006

Coronary artery disease 21(34.4) 11(14.3) 0.005

Chronic heart failure 10 (16.4) 7 (9.1) 0.195

Chronic pulmonary disease 9 (14.8) 11 (14.3) 0.938

Malignancy 27 (44.3) 31 (40.3) 0.636

CCI, median (IQR: 25-75%) 6 (4.0-7.5) 4 (2-6) 0.001

SIRS, median (IQR: 25-75%) 2 (1-3) 2 (1-3) 0.737

SOFA, median (IQR: 25-75%) 6 (4-9) 5 (2-7) <0.001

Intermittent hemodialysis, n (%) 36 (59) - n/a

Continuous renal replacement therapy, n (%) 8 (13.1) - n/a

Type of BSIs, n (%)

Hospital-acquired 22 (36.1) 18 (23.4)

0.103
Community-acquired 39 (63.9) 59 (76.6)

Microorganism, n (%)

Fermenter GNs a 39 (63.9) 46 (59.7)

0.615
Non-fermenter GNs b 22 (36.1) 31 (40.3)

MDR-GNs, n (%) 15 (26.2) 28 (36.4) 0.205

eGFR: Estimated glomerular filtration rates, IQR: Interquartile range, CCI: Charlson comorbidity index,
SIRS: Systemic inflammatory response syndrome, SOFA: Sequential organ failure assessment score,
BSI: Bloodstream infections, GN: Gram-negative, MDR: Multidrug resistance.

a
Klebsiella spp., n=42; E. coli, n=26; Enterobacter spp., n=7; Proteus spp., n=7; Morganella spp., n=1, Rautella spp.,
n=1; Serratia spp., n=1.
b
Acinetobacter baumannii, n=25; Pseudomonas aeruginosa, n=14; Stenothrophomonas maltophilia, n=5; Ralstonia
paucula, n=3; Delfia avidovorans, n=3; Burkholderia cepacia, n=2; Sghingomonas spp., n=1.

(SOFA) scores, blood culture results, empiric and tar- according to their lowest eGFR level within 14 days.
geted antibiotics, duration of antibiotic treatments, These two groups were compared for PCT kinetic.
and clinical outcomes (intensive care unit stay and The procalcitonin test was measured quantitative-
mortality) were recorded. The first day of the appro- ly in serum samples using the electrochemilumi-
priate antibiotic therapy was considered day zero. nescence immunoassay “ECLIA” (Elecsys® BRAHMS
Blood PCT, C-reactive protein (CRP), creatinine lev- PCT; Roche Diagnostics Inc., USA) method.
els, and eGFR values were measured and recorded
for 14 days from day zero. Patients were divided into The study primarily aimed to evaluate the effect of
eGFR <30 mL/min/1.73m² and ≥ 30 mL/min/1.73m², renal clearance on PCT kinetics in patients with GN-

This work is licensed under a Creative Commons Attribution-Non Commercial 4.0 International License. 208
Procalcitonin Kinetics in Gram-Negative Bloodstream Infections

Table 2. PCT kinetics in patients with GN-BSIs according to eGFR levels.

eGFR <30 eGFR <30

mL/min/1.73m2 mL/min/1.73m2 p
n=61 n=77

PCT, ng/mL, median (IQR: 25-75%)

Day 0 (n=55/68) 4.58 (1.36-39.4) 0.91 (0.32-10.2) <0.001

Day 1 (n=54/66) 8.73 (1.52-39.4) 1.3 (0.38-13) 0.001

Day 3 (n=53/63) 4.18 (1.20-19.9) 0.9 (0.23-5.25) <0.001

Day 5 (n=55/65) 2.98 (1.06-13) 0.58 (0.17-2.71) <0.001

Day 7 (n=50/63) 1.38 (0.68-3.68) 0.47 (0.14-1.07) <0.001

Day 10 (n=39/59) 0.85 (0.47-2.31) 0.27 (0.11-0.67) <0.001

Day 14 (n=36/53) 0.56 (0.29-1.41) 0.21 (0.09-0.71) 0.009

ΔPCT, between baseline and day 3ª

13 (26) 23 (47.9) 0.024
(≥30% decrease), (n=50/48), n (%)

ΔPCT, between baseline and day 5a,b

36 (69.2) 36 (76.6) 0.411
(≥ 50% decreases), (n=52/47), n (%)

Final PCT level, ng/mL, median (IQR: 25-75%) 0.62 (0.29-1.63) 0.21 (0.09-0.7) 0.001

GN-BSI: Gram-negative bloodstream infections, PCT: Procalcitonin, ΔPCT: Delta procalcitonin, eGFR: Estimated
glomerular filtration rates, IQR: Interquartile range.

ªThe highest PCT values ​​on days 0-1 and above the normal limit value (>0.5 ng/mL) were included in the analyses.
b
Insufficient power to assess the statistical significance of the difference between groups (Power: 0.12 with 0.05
type 1 error and 0.20 type 2 error)

BSIs who had clinical and microbiological respons- RESULTS

es. Its second aim was to assess the changes in PCT In the study, a total of 138 patients with eGFR <30
kinetics depending on eGFR between fermenting and ≥30 ml/min/ 1.73 m² were compared for their
and non-fermenting Gram-negatives (GNs). demographic and clinical characteristics (Table 1,
Figure 1).
Statistical Analysis
The statistical analyses were performed using Of the patients, 71% (n=98) received empirical treat-
the Statistical Package for Social Sciences (SPSS) ment, and 29% (n=40) received definitive appropriate
25.0 (IBM Corp., Armonk, NY, USA). The normali- antibiotic therapy. PCT levels on day zero of antibiot-
ty of the data distribution was determined by the ic therapy in patients with eGFR <30 mL/min/1.73m²
Shapiro–Wilk test, histograms, and Q-Q plots. The and eGFR ≥30 mL/min/1.73m² were 4.58 (1.36-39.4)
categorical variables of the patients were present- ng/mL and 0.91 (0.32-10.2) respectively (p<0.001).
ed as numbers and percentages and compared by PCT level changes in both groups during therapy are
using the chi-square test. Parametric variables presented in Figure 2 and Table 2.
were presented as mean standard deviation (SD).
Non-parametric variables were presented with PCT levels on day zero of antibiotic therapy in BSIs
median values and an interquartile range (IQR) of with fermenter GNs were 1.68 (0.43-12.9) ng/mL
25–75% and compared using the Mann-Whitney U in GFR ≥30 and 13.0 (1.45-48.7) ng/mL in GFR <30
test. A p value of 0.05 was considered statistically (p=0.001). PCT levels on day zero of antibiotic thera-
significant. Graphs were created with the Graph- py in BSIs with non-fermenter GNs were 0.57 (0.18-
Pad Prism 8 (GraphPad Software, Boston, USA) pro- 5.15) ng/mL in eGFR ≥30 and 2.21 (0.59-10.1) ng/
gram. mL in GFR <30 (p=0.040). Comparison of PCT levels

Özger HS et al. 209

 Infect Dis Clin Microbiol 2024; 6(3): 206-15

Table 3. PCT kinetics according to eGFR in patients with fermenter and non-fermenter GN-BSIs.

eGFR <30 eGFR <30

mL/min/1.73m2 mL/min/1.73m2 p
n=61 n=77

PCT, ng/mL, median (IQR: 25-75%), levels in patients with fermenter GN-BSIs

Day 0 (n=35/40) 13 (1.45-48.7) 1.68 (0.43-12.9) 0.001

Day 1 (n=33/41) 19.1 (3.5-77.1) 2.31 (80.40-19.2) 0.001

Day 3 (n=33/40) 5.19 (1.69-39.4) 1.15 (0.22-6.38) 0.002

Day 5 (n=35/40) 3.14 (1.116-13) 0.70 (0.16-2.89) <0.001

Day 7 (n=34/38) 1.23 (0.55-2.79) 0.58 (0.17-1.15) 0.010

Day 10 (n=29/36) 0.95 (0.54-2.2) 0.27 (0.12-0.75) 0.004

Day 14 (n=27/31) 0.5 (0.29-1.29) 0.19 (0.10-0.72) 0.035

ΔPCT, between day 0-1 and day 3 a, b

11 (33.3) 18 (56.3) 0.062
(≥ 30% decrease), (n=33/32), n (%)

ΔPCT, between day 0-1 and day 5a, c

28 (80) 27(87.1) 0.44
(≥ 50% decrease), (n=35/31), n (%)

Final PCT level, ng/mL, median

0.61(0.29-1.54) 0.19 (0.1-0.7) 0.003
(IQR: 25-75%), (n=32/39)

PCT, ng/mL, median (IQR: 25-75%, levels in patients with non-fermenter GN-BSIs

Day 0 (n=20/28) 2.21 (0.59-10.1) 0.57 (0.18-5.15) 0.04

Day 1 (n=21/25) 1.75 (0.94-12.4) 0.96 (0.31-3.45) 0.062

Day 3 (n=20/23) 2.78 (1.02-9.18) 0.49 (0.25-2.34) 0.032

Day 5 (n=20/25) 2.1 (0.96-11.6) 0.43 (0.18-2.71) 0.008

Day 7 (n=16/25) 2.14 (0.79-4.07) 0.39 (0.12-1.33) 0.003

Day 10 (n=10/23) 0.84 (0.20-2.33) 0.23 (0.08-0.53) 0.034

Day 14 (n=9/22) 0.89 (0.30-2.18) 0.28 (0.08-0.75) 0.098

ΔPCT, between day 0-1 and day 3 a, d

2 (11.8) 5 (31.3) 0.166
(≥ 30% decrease), (n=20/19), n (%)

ΔPCT, between day 0-1 and day 5a, e

8 (47.1) 9 (56.3) 0.598
(≥ 50% decrease), (n=32/35), n (%)

Final PCT level, ng/mL, median

0.89 (0.15-2.92) 0.36 (0.09-0.85) 0.127
(IQR: 25-75%), (n=11/24)

GN-BSI: Gram-negative bloodstream infections, PCT: Procalcitonin, ΔPCT: Delta procalcitonin, eGFR: Estimated
glomerular filtration rates, IQR: Interquartile range.

ªThe highest PCT values ​​on days 0-1 and above the normal limit value (>0.5 ng/mL)
were included in the analyses.
b
Insufficient power to assess the statistical significance of the difference between groups
(Power: 0.46 with 0.05 type 1 error and 0.20 type 2 error).
c
Insufficient power to assess the statistical significance of the difference between groups
(Power: 0.11 with 0.05 type 1 error and 0.20 type 2 error).
d
Insufficient power to assess the statistical significance of the difference between groups
(Power: 0.32 with 0.05 type 1 error and 0.20 type 2 error).
e
Insufficient power to assess the statistical significance of the difference between groups
(Power: 0.11 with 0.05 type 1 error and 0.20 type 2 error).

This work is licensed under a Creative Commons Attribution-Non Commercial 4.0 International License. 210
Procalcitonin Kinetics in Gram-Negative Bloodstream Infections

Figure 2. PCT kinetics in patients with GN-BSIs according to eGFR levels.

Figure 1. Flow diagram.

The literature shows an inverse relationship be-
during antibiotic therapy in patients with fermen- tween serum PCT and GFR levels (15). Compared
ter and non-fermenter GN-BSIs according to eGFR with healthy controls, serum PCT values are higher
(<30 and ≥30 mL/min/1.73 m²) are presented in Ta- and above the PCT normal limits in ​​ patients with
ble 3, Figure 3, and Figure 4. renal failure without infection (10, 15). Also, in in-
fected patients, the stage of renal failure directly
DISCUSSION affects serum PCT values (16). Therefore, higher
PCT cut-off values ​​are recommended to predict in-
In our study, GN-BSIs-related PCT levels were high- fection in patients with low GFR (9, 17, 18). As in
er at the beginning of antibiotic therapy in patients diagnostic use, higher cut-off values ​​are accepted
with low GFR levels. This significant difference in for prognostic use of PCT in these patients (19). De-
PCT levels associated with lower GFR levels per- creased PCT clearance in renal failure is one of the
sisted during antibiotic therapy. PCT elevation in possible causes of high serum PCT levels, but there
patients with low GFR was higher in BSIs associat- is still no consensus on this hypothesis (10, 19). In
ed with fermenter GNs. On day three of antibiotic their study, Meisner et al. found no relationship
therapy, PCT response was inadequate in patients between serum PCT kinetic and age, gender, SOFA
with low GFR levels. Due to the slow kinetics of PCT, score, or GFR levels. So, they reported that renal se-
most patients with low GFR had a PCT response by cretion did not affect PCT elimination (20).
day five. Final PCT values in
​​ patients with low GFR
levels were above normal serum PCT limits and On the other hand, in some studies, PCT clearance
higher than those with eGFR ≥ 30 mL/min/1.73m2. decreased due to renal failure (9, 10, 18, 21). In our

Özger HS et al. 211

 Infect Dis Clin Microbiol 2024; 6(3): 206-15

sis that renal clearance of PCT is reduced due to re-

nal failure. However, the low eGFR group had higher
CCI and SOFA scores. This may cause differences in
infection severity and immune response between
groups, decreasing the causal relationship between
renal clearance and high serum PCT levels. The
possible increased severity of infection with renal
failure confuses the relationship between PCT level
and renal clearance (9). Therefore, in our study, we
preferred to focus on PCT kinetics in patients with
infection control rather than possible causes of PCT
elevation.

Failure in PCT clearance is also a prognostic factor

in critically ill patients (22-24). According to a study
by Schuetz et al., 28th-day mortality was twofold
high in patients without an 80% decline in PCT by
day 4 (25). In other observational studies, an earlier
30-70% decrease in PCT values ​​was associated with
survival (12, 26). In a study by Giamarellos-Bour-
Figure 3. PCT kinetics in patients with fermenter GN-BSIs boulis et al., compared to baseline, a 30% decrease
according to eGFR levels. in PCT value on the second day, a 40% decrease on
the fourth day, or detection of a normal PCT level
(<0.5 ng/mL) on the fourth day was an early predic-
tor of GNs-related bacteremia control (27). Kim et
al. reported that the decrease in PCT levels on day
three was associated with survival and renal re-
covery in patients with sepsis-induced acute renal
failure. Every 10% decrease in PCT values ​​was as-
sociated with a 10% decrease in mortality risk (11).
Despite these studies, there is insufficient data for
optimal cut-off and time interval for PCT decrease
in patients with renal failure.

Our study showed that the PCT responses to an-

tibiotic therapy in patients with infection control
change according to eGFR levels. Patients with low
eGFR had lower and slower PCT responses. eGFR di-
rectly affected PCT kinetics in both fermenter and
non-fermenter GN bacteria-related BSI subgroups.
This result is inconsistent with the study of Meisner
et al., which suggested that the PCT serum clear-
ance rate does not depend on renal function (20).
Figure 4. PCT kinetics in patients with non-fermenter GN-BSIs According to our study result, the PCT response
according to eGFR levels.
achieved in less than half of the patients on day
three may lead to unnecessary PCT-based antibiot-
study, serum PCT level was significantly higher in ic therapy changes in patients with low eGFR. PCT
patients with low eGFR in GN-BSIs during antibiotic response should be evaluated on the fifth day in
treatment. This elevation may support the hypothe- patients with clinical responses to reduce unneces-

This work is licensed under a Creative Commons Attribution-Non Commercial 4.0 International License. 212
Procalcitonin Kinetics in Gram-Negative Bloodstream Infections

sary antibiotic revisions. However, in the fermenter eGFR and did not measure actual GFR (29). All eGFR
BSI subgroup with a low eGFR level, the sharp PCT formulas known to us are validated in stable pa-
decrease on the third day may be helpful for an ear- tients without acute kidney injury (AKI), and they
ly response. tend to overestimate eGFR when AKI is present. To
tackle this issue, some researchers suggest using
In clinical practice, serum PCT is generally used for newer indicators of AKI such as kidney injury mol-
optimal antibiotic therapy duration apart from its ecule-1 (KIM-1), neutrophil gelatinase-associated
diagnostic and prognostic use. In their meta-anal- lipocalin (NGAL) and others, but their success and
ysis, Heilmann et al., PCT-guided antibiotic therapy triumph over eGFR is not well established (30, 31).
in patients with renal failure resulted in an approx- Furthermore, measuring the actual GFR is a cum-
imately two-day reduction in antibiotic use without bersome process which can only be employed in
increasing mortality risk. However, measurement studies with a prospective design and intent (32).
times of PCT and PCT cut-offs for antibiotic thera- However, to minimize the effect of eGFR’s well-
py optimization were quite different in the studies known constraints on our work, we demonstrated
included in the meta-analysis (28). This heteroge- creatinine trends, which are more valuable than a
neity in the meta-analysis does not allow recom- single estimate of kidney function (33). Another lim-
mendations for clinical practice. In our study, final itation related to the retrospective and descriptive
serum PCT levels at the end of antibiotic therapy study design is missing data in PCT serial measure-
were above normal limits ​​in patients with low eGFR ments. This may cause the sample size to change
in the fermenter and non-fermenter GNs-related in each comparison and increase the risk of type 2
BSIs. In a PCT-guided antibiotic therapy approach, errors. Another limitation of our study is that since
these high final PCT values may​​ lead to unneces- PCT values ​​were not normally distributed, pairwise
sarily longer antibiotic therapy. Therefore, in opti- comparisons could be made at the measurement
mizing antibiotic therapy duration in patients with points instead of mixed ANOVA between the study
low eGFR, changes in PCT kinetics should be eval- groups.
uated together with the resolution of clinical infec-
tion signs instead of the final PCT level. In conclusion, according to our study results, serum
PCT levels were higher in patients with low eGFR
Our study had some limitations. First of all, despite at baseline, during, and at the end of therapy. Early
the strict inclusion criteria, all confounding factors PCT response evaluation should not be performed,
that may affect PCT kinetics could not be exclud- especially in these patients with non-fermenter
ed due to the retrospective study design. Although GN-BSIs, as the PCT response is delayed until the
we determined some criteria to exclude other ac- fifth day. Considering high PCT levels and slower
tive infections that may affect PCT kinetics, these PCT kinetics in patients with low eGFR may reduce
infections may not have been wholly excluded. Our unnecessary antibiotic revisions and extensions.
study used the CKD-EPI 2009 formula to calculate

Ethical Approval: The Ethics Committee of Gazi University F.B.A., Ö.H.; Analysis and/or Interpretation – H.S.Ö., M.Y., F.B.A.,
approved the study on January 23, 2024, with decision number Ş.K.Ç; Literature Review – H.S.Ö., M.Y., F.B.A., Ö.H.; Writer – H.S.Ö.,
2024-196. Ş.K.Ç, N.B.D., G.A., Ö.H., M.T.; Critical Reviews – G.A., M.T., M.D.,
N.B.D., Ö.H.
Informed Consent: N.A.
Conflict of Interest: The authors declare no conflict of interest.
Peer-review: Externally peer-reviewed
Financial Disclosure: The authors declared that this study has
Author Contributions: Concept – H.S.Ö.., Ş.K.Ç., N.B.D.., Ö.H.;
received no financial support.
Design – H.S.Ö., Ş.K.Ç, M.D., G.A., M.T.; Supervision – H.S.Ö., Ş.K.Ç,
M.D., M.Y., F.B.A.; Data Collection and/or Processing – H.S.Ö., M.Y.,

Özger HS et al. 213

 Infect Dis Clin Microbiol 2024; 6(3): 206-15

REFERENCES
1 Tan M, Lu Y, Jiang H, Zhang L. The diagnostic accuracy of 14 Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd,
procalcitonin and C-reactive protein for sepsis: A systematic Feldman HI, Kusek JW, et al; CKD-EPI (Chronic Kidney Disease
review and meta-analysis. J Cell Biochem. 2019;120(4):5852-9. Epidemiology Collaboration). A new equation to estimate glo-
[CrossRef] merular filtration rate. Ann Intern Med. 2009;150(9):604-12. Er-
2 Huang YH, Chen CJ, Shao SC, Li CH, Hsiao CH, Niu KY, et al. ratum in: Ann Intern Med. 2011;155(6):408. [CrossRef]
Comparison of the diagnostic accuracies of monocyte distri- 15 Harris PNA, Tambyah PA, Lye DC, Mo Y, Lee TH, Yilmaz M, et
bution width, procalcitonin, and C-reactive protein for sep- al; MERINO Trial Investigators and the Australasian Society
sis: a systematic review and meta-analysis. Crit Care Med. for Infectious Disease Clinical Research Network (ASID-CRN).
2023;51(5):e106-14. [CrossRef] Effect of piperacillin-tazobactam vs meropenem on 30-day
3 Cong S, Ma T, Di X, Tian C, Zhao M, Wang K. Diagnostic value of mortality for patients with E. coli or Klebsiella pneumoniae
neutrophil CD64, procalcitonin, and interleukin-6 in sepsis: a bloodstream infection and ceftriaxone resistance: a random-
meta-analysis. BMC Infect Dis. 2021;21(1):384. [CrossRef] ized clinical trial. JAMA. 2018;320(10):984-94. Erratum in: JAMA.
2019;321(23):2370. [CrossRef]
4 Arora S, Singh P, Singh PM, Trikha A. Procalcitonin levels in
survivors and nonsurvivors of sepsis: systematic review and 16 Park JH, Kim DH, Jang HR, Kim MJ, Jung SH, Lee JE, et al. Clinical
meta-analysis. Shock. 2015;43(3):212-21. [CrossRef] relevance of procalcitonin and C-reactive protein as infection
markers in renal impairment: a cross-sectional study. Crit
5 Scott J, Deresinski S. Use of biomarkers to individualize anti-
Care. 2014;18(6):640. [CrossRef]
microbial therapy duration: a narrative review. Clin Microbiol
Infect. 2023;29(2):160-4. [CrossRef] 17 Jiang L, Shao X, Xing W, Sun Y. Biomarkers and risk factors for
sepsis in stage 5 chronic kidney disease: a retrospective case-con-
6 Evans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith
trol study. Int Urol Nephrol. 2019;51(4):691-8. [CrossRef]
CM, French C, et al. Surviving sepsis campaign: international
guidelines for management of sepsis and septic shock 2021. 18 El-Sayed D, Grotts J, Golgert WA, Sugar AM. Sensitivity and
Intensive Care Med. 2021;47(11):1181-247. [CrossRef] specificity of procalcitonin in predicting bacterial infections
in patients with renal impairment. Open Forum Infect Dis.
7 Peng F, Chang W, Xie JF, Sun Q, Qiu HB, Yang Y. Ineffectiveness
2014;1(2):ofu068. [CrossRef]
of procalcitonin-guided antibiotic therapy in severely critical-
ly ill patients: a meta-analysis. Int J Infect Dis. 2019;85:158-66. 19 Han S, Kim MJ, Ko HJ, Lee EJ, Kim HR, Jeon JW, et al. Diagnostic
[CrossRef] and prognostic roles of C-reactive protein, procalcitonin, and
presepsin in acute kidney injury patients initiating continuous
8 Papp M, Kiss N, Baka M, Trásy D, Zubek L, Fehérvári P, et al.
renal replacement therapy. Diagnostics (Basel). 2023;13(4):777.
Procalcitonin-guided antibiotic therapy may shorten length of
[CrossRef]
treatment and may improve survival-a systematic review and
meta-analysis. Crit Care. 2023;27(1):394. [CrossRef] 20 Meisner M, Schmidt J, Hüttner H, Tschaikowsky K. The natural
elimination rate of procalcitonin in patients with normal and
9 Sun Y, Jiang L, Shao X. Predictive value of procalcitonin for di-
impaired renal function. Intensive Care Med. 2000;26 Suppl
agnosis of infections in patients with chronic kidney disease: a
2:S212-6. [CrossRef]
comparison with traditional inflammatory markers C-reactive
protein, white blood cell count, and neutrophil percentage. Int 21 Grace E, Turner RM. Use of procalcitonin in patients with vari-
Urol Nephrol. 2017;49(12):2205-16. [CrossRef] ous degrees of chronic kidney disease including renal replace-
ment therapy. Clin Infect Dis. 2014;59(12):1761-7. [CrossRef]
10 Wu SC, Liang CX, Zhang YL, Hu WP. Elevated serum procalci-
tonin level in patients with chronic kidney disease without in- 22 Liu D, Su L, Han G, Yan P, Xie L. Prognostic value of procalci-
fection: A case-control study. J Clin Lab Anal. 2020;34(2):e23065. tonin in adult patients with sepsis: a systematic review and
[CrossRef] meta-analysis. PLoS One. 2015;10(6):e0129450. [CrossRef]

11 Kim IY, Kim S, Ye BM, Kim MJ, Kim SR, Lee DW, et al. Procal- 23 Patnaik R, Azim A, Mishra P. Should serial monitoring of pro-
citonin decrease predicts survival and recovery from dialysis calcitonin be done routinely in critically ill patients of ICU:
at 28 days in patients with sepsis-induced acute kidney inju- A systematic review and meta-analysis. J Anaesthesiol Clin
ry receiving continuous renal replacement therapy. PLoS One. Pharmacol. 2020;36(4):458-64. [CrossRef]
2022;17(12):e0279561. [CrossRef] 24 Pantelidou IM, Giamarellos-Bourboulis EJ. Can procalcitonin
12 Vitorio D, Nassar AP Jr, Caruso P. Procalcitonin clearance and monitoring reduce the length of antibiotic treatment in blood-
prognosis in sepsis: are there really an optimal cutoff and time stream infections? Int J Antimicrob Agents. 2015;46 Suppl
interval? Crit Care Med. 2017;45(10):e1097-8. [CrossRef] 1:S10-2. [CrossRef]

13 Thomas-Rüddel DO, Poidinger B, Kott M, Weiss M, Reinhart K, 25 Schuetz P, Birkhahn R, Sherwin R, Jones AE, Singer A, Kline JA,
Bloos F; MEDUSA study group. Influence of pathogen and fo- et al. Serial procalcitonin predicts mortality in severe sepsis
cus of infection on procalcitonin values in sepsis patients with patients: results from the multicenter procalcitonin MOni-
bacteremia or candidemia. Crit Care. 2018;22(1):128. [CrossRef] toring SEpsis (MOSES) study. Crit Care Med. 2017;45(5):781-9.
[CrossRef]

This work is licensed under a Creative Commons Attribution-Non Commercial 4.0 International License. 214
Procalcitonin Kinetics in Gram-Negative Bloodstream Infections

26 de Azevedo JR, Torres OJ, Beraldi RA, Ribas CA, Malafaia O. 31 Zdziechowska M, Gluba-Brzózka A, Poliwczak AR, Franczyk
Prognostic evaluation of severe sepsis and septic shock: pro- B, Kidawa M, Zielinska M, et al. Serum NGAL, KIM-1, IL-18,
calcitonin clearance vs Δ sequential organ failure assessment. L-FABP: new biomarkers in the diagnostics of acute kidney in-
J Crit Care. 2015;30(1):219.e9-12. [CrossRef] jury (AKI) following invasive cardiology procedures. Int Urol
27 Giamarellos-Bourboulis EJ, Kotsaki A, Routsi C, Graziano E, Nephrol. 2020;52(11):2135-43. [CrossRef]
Righi E, Bassetti M. A prognostic score for the resolution of bac- 32 Pelletier K, Lafrance JP, Roy L, Charest M, Bélanger MC, Cailhier
teremia by Gram-negative bacteria resistant to carbapenems. JF, et al. Estimating glomerular filtration rate in patients with
Eur J Clin Microbiol Infect Dis. 2018;37(11):2083-9. [CrossRef] acute kidney injury: a prospective multicenter study of diag-
28 Heilmann E, Gregoriano C, Wirz Y, Luyt CE, Wolff M, Chastre nostic accuracy. Nephrol Dial Transplant. 2020;35(11):1886-93.
J, et al. Association of kidney function with effectiveness of [CrossRef]
procalcitonin-guided antibiotic treatment: a patient-level me- 33 Seller-Pérez G, Herrera-Gutiérrez ME, Maynar-Moliner J, Sán-
ta-analysis from randomized controlled trials. Clin Chem Lab chez-Izquierdo-Riera JA, Marinho A, do Pico JL. Estimating kid-
Med. 2020;59(2):441-53. [CrossRef] ney function in the critically ill patients. Crit Care Res Pract.
29 Bragadottir G, Redfors B, Ricksten SE. Assessing glomerular fil- 2013;2013:721810. [CrossRef]
tration rate (GFR) in critically ill patients with acute kidney
injury--true GFR versus urinary creatinine clearance and esti-
mating equations. Crit Care. 2013;17(3):R108. [CrossRef]
30 Zhang B, Song Y, Ma Q, Yang J, Bai L. Expression and significance
of KIM-1, NGAL, and HO-1 in patients with acute kidney injury
after cardiac valve replacement. J Inflamm Res. 2023;16:2755-
61. [CrossRef]

Özger HS et al. 215