Immature platelet fraction predicts coagulopathy-related platelet consumption and mortality in patients with sepsis

    Immature platelet fraction predicts coagulopathy-related platelet consumption and mortality in patients with sepsis Tomohiro Muronoi, Kansuke Koyama, Shin Nunomiya, Alan Kawarai Lefor, Masahiko Wada, Toshitaka Koinuma, Jun Shima, Masayuki Suzukawa PII: DOI: Reference:

S0049-3848(16)30411-X doi: 10.1016/j.thromres.2016.06.002 TR 6362

To appear in:

Thrombosis Research

Received date: Revised date: Accepted date:

19 February 2016 6 May 2016 2 June 2016

Please cite this article as: Muronoi Tomohiro, Koyama Kansuke, Nunomiya Shin, Lefor Alan Kawarai, Wada Masahiko, Koinuma Toshitaka, Shima Jun, Suzukawa Masayuki, Immature platelet fraction predicts coagulopathy-related platelet consumption and mortality in patients with sepsis, Thrombosis Research (2016), doi: 10.1016/j.thromres.2016.06.002

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1 Full Length Article Immature platelet fraction predicts coagulopathy-related platelet consumption and

IP

T

mortality in patients with sepsis

SC R

Tomohiro Muronoia,b,*, Kansuke Koyamaa, Shin Nunomiyaa, Alan Kawarai Leforc, Masahiko Wadaa, Toshitaka Koinumaa, Jun Shimaa, Masayuki Suzukawab

Division of Intensive Care, Department of Anesthesiology and Intensive Care Medicine, Jichi

NU

a

b

MA

Medical University School of Medicine, Tochigi, Japan

Department of Emergency Medicine, Jichi Medical University School of Medicine, Tochigi,

Japan

D

Department of Surgery, Jichi Medical University School of Medicine, Tochigi, Japan

CE P

*Corresponding author

TE

c

Tomohiro Muronoi, MD

AC

Division of Intensive Care, Department of Anesthesiology and Intensive Care Medicine, Jichi Medical University School of Medicine, 3311-1 Yakushiji, Shimotsuke, Tochigi 329-0498, Japan

Tel.: +81-285-44-2111; Fax: +81-285-44-8845 Email: [email protected]

-1-

ACCEPTED MANUSCRIPT 2 Abstract Introduction: The diagnostic and prognostic value of immature platelet fraction (IPF) in

T

sepsis has not been determined. This study aimed to assess whether IPF is an early predictor

IP

of platelet decline due to coagulopathy and is associated with mortality in patients with sepsis.

SC R

Materials and Methods: In total, 149 patients with a platelet count of ≥80×103/L on intensive care unit admission (101 with sepsis, 48 controls without sepsis) were prospectively evaluated. We measured IPF on admission and observed for development of subsequent

NU

platelet count decline (defined as a >30% decrease or <80×103/L) in 5 days, and mortality at

MA

28 days. The absolute immature platelet count (AIPC) was calculated to evaluate thrombopoiesis.

D

Results: Forty-seven patients with sepsis subsequently developed a decrease in platelet count.

TE

The IPF was highest in patients whose platelet count decreased, followed by patients without a decrease in platelet count and controls (median, 4.3% [3.1%–8.1%] vs. 3.7% [2.6%–4.6%]

CE P

vs. 2.1% [1.6%–3.5%], respectively; P<0.0001). The AIPC was similar in patients with and without a decrease in platelet count (7.6 [4.2–10.0] vs. 5.9 [4.2–8.7] ×103/μL, respectively;

AC

P=0.32). Coagulation derangement was more severe in patients who did than did not subsequently develop a decreased platelet count. Cox regression and receiver operator characteristic curve analysis revealed that IPF was a strong independent predictor of mortality, with accuracy similar to a standard prognostic scoring system. Conclusions: The admission IPF in septic patients predicts a subsequent decrease in platelet count, indicating platelet consumption with ongoing coagulopathy and risk of poor prognosis.

Keywords: immature platelet fraction; thrombocytopenia; sepsis; coagulopathy; disseminated intravascular coagulation; mortality -2-

ACCEPTED MANUSCRIPT 3

Abbreviations

T

ICU, intensive care unit

IP

DIC, disseminated intravascular coagulation

SC R

IPF, immature platelet fraction AIPC, absolute immature platelet count

SOFA, Sequential Organ Failure Assessment

NU

APACHE, Acute Physiology and Chronic Health Evaluation

MA

PT-INR, prothrombin time-international normalized ratio FDP, fibrin degradation products AT, antithrombin

AC

CE P

AUC, area under the curve

TE

D

TAT, thrombin-antithrombin complex

-3-

ACCEPTED MANUSCRIPT 4 Introduction Platelets play a pivotal role in antimicrobial host defense, linking the processes of

T

inflammation and coagulation [1-3]. Substantial evidence suggests that platelets detect and

IP

respond to bacterial infections with specific receptors such as Toll-like receptors and release

SC R

cytokines and chemokines. In parallel, proinflammatory cytokines produced by host responses against infection activate the coagulation cascade, leading to massive generation of thrombin, which is the most potent platelet agonist [4]. Activated platelets interact with leukocytes and

NU

endothelial cells, provide a phospholipid surface for coagulation, and are consumed by being

MA

captured on a microvascular fibrin meshwork. Thrombocytopenia is therefore common in patients with severe infection or sepsis and indicates a serious pathophysiological status

TE

D

associated with coagulopathy [1, 5].

Numerous studies have shown that thrombocytopenia is associated with a poor

CE P

prognosis in critically ill patients [6, 7]. Most studies have evaluated the baseline platelet count as a prognostic factor, but the development of thrombocytopenia after intensive care

AC

unit (ICU) admission may hold greater prognostic significance. Even a 30% decrease in platelet count during the ICU stay has been shown to be associated with increased mortality [8]. In addition, thrombocytopenia is a relevant marker of severe coagulopathy or disseminated intravascular coagulation (DIC) in patients with sepsis, and DIC itself is an independent risk factor for mortality [5, 9]. Hence, a marker that detects an initial drop in platelet count may help to stratify patients at risk for severe coagulopathy and a poor prognosis in the early course of sepsis.

During the last several years, new technology has appeared in the form of the commercial automatic analyzer, which has allowed for the measurement of immature platelet -4-

ACCEPTED MANUSCRIPT 5 fraction (IPF) in daily clinical practice [10]. The IPF has been shown to be a reliable alternative to reticulated platelet as an indicator of increased platelet turnover, and it can

T

compensate for the limitation that the reticulated platelet is difficult to standardize [11]. The

IP

major advantage of IPF measurement is that it is quantified automatically by means of a

SC R

reproducible laboratory test and can be established as part of a routine complete blood count.

The IPF is commonly reported as the percentage IPF (the percentage of platelets with

NU

above-threshold RNA), but it can also be expressed as the absolute immature platelet count

MA

(AIPC), which is the actual number of immature platelets per unit volume (%IPF × platelet count). A high IPF indicates platelet consumption or recovering thrombocytopenic disorders, while a low IPF or low AIPC is seen in an aplastic state. Several studies have demonstrated

TE

D

the clinical utility of IPF in laboratory diagnosis of thrombocytopenia due to peripheral platelet destruction, particularly autoimmune thrombocytopenic purpura and thrombotic

CE P

thrombocytopenic purpura [12, 13]. The IPF was recently introduced as a novel diagnostic biomarker for infection and developing sepsis [14, 15]. To our knowledge, however, no

AC

studies have evaluated the clinical value of IPF for coagulopathy or its prognostic value in patients with sepsis.

In this study, we evaluated IPF in association with a subsequent decrease in platelet count and coagulopathy and assessed the prognostic value of a single determination of IPF in patients with sepsis. We hypothesized that IPF may identify impending thrombocytopenia in septic coagulopathy and therefore be associated with mortality in patients with sepsis.

-5-

ACCEPTED MANUSCRIPT 6 Materials and Methods Study design and setting

T

This was a single-center, prospective observational study of patients with sepsis admitted to a

IP

medical-surgical ICU at a tertiary hospital (Jichi Medical University Hospital) from October

SC R

2013 to February 2015. The study was strictly observational, and all interventions and laboratory tests were part of our routine practice. Clinical decisions were made at the discretion of attending ICU physicians. All patients were managed in accordance with the

NU

Surviving Sepsis Campaign Guideline with the goal of initial resuscitation and infection

MA

control [16] and received mechanical thromboprophylactic treatment without concomitant low-dose heparin until the absence of active bleeding or severe coagulopathy was established. Patients at risk of bleeding or complications were transfused with platelet concentrates or

TE

D

fresh frozen plasma at the discretion of the ICU physician. The conduct of the study was

Patients

CE P

approved by the Institutional Research Ethics Committee of Jichi Medical University.

AC

Patients were enrolled at the time of ICU admission if they fulfilled the criteria for sepsis, had a platelet count of ≥80×103/µL, and were studied for 5 consecutive days (days 1–5) during their ICU stay. Sepsis was diagnosed according to the criteria of the 2001 International Sepsis Definitions Conference [17]. A control group was also evaluated and comprised patients admitted to the ICU after elective surgery for noninfectious diseases. No patients in the control group had signs of infection in the 4 weeks before surgery.

The exclusion criteria included an age of <18 years, missing laboratory data, presence of hematologic disorders, decompensated liver cirrhosis (Child–Pugh class B or C), a history of chemotherapy, and a history of a blood transfusion during the preceding 4 weeks. -6-

ACCEPTED MANUSCRIPT 7

Data collection

T

Clinical and demographic data, including age, sex, and comorbidities, were recorded on ICU

IP

admission. The Acute Physiology and Chronic Health Evaluation (APACHE) II score, a

SC R

standard severity and prognostic score for critically ill patients, was also calculated on admission [18]. The Sequential Organ Failure Assessment (SOFA) score [19] excluding coagulation (platelet count) and the overt DIC score according to the International Society on

NU

Thrombosis and Haemostasis (ISTH) criteria were determined daily. Clinical variables and

MA

treatments, such as transfusion of blood products including red blood cells, fresh frozen plasma, and platelet concentrates, were also recorded daily. Clinical outcomes were assessed

Measurement of IPF

TE

D

by the number of ICU-free days during the first 28 days [20] and all-cause 28-day mortality.

CE P

The IPF was measured automatically with a hematology analyzer (XE-5000; Sysmex, Kobe, Japan) when the first complete blood count was obtained on day 1 (the day of ICU admission)

AC

and on days 2 to 5. Measurement of IPF was performed in the dedicated platelet channel of the hematology analyzer by flow cytometry, using a proprietary fluorescent dye containing polymethine and oxazine. These dyes penetrate the cell membrane, staining the DNA and RNA in platelets. Platelets are divided into two groups—mature and immature—according to the intensity of cell fluorescence, which correlates with the RNA content and is consequently higher in immature platelets. The IPF corresponds to the fraction (%) of immature platelets in the total platelet population [14, 15]. The AIPC was calculated as the absolute number of immature platelets per unit volume (%IPF × platelet count).

Measurement of inflammatory and coagulation biomarkers -7-

ACCEPTED MANUSCRIPT 8 The following parameters were measured from day 1 (the time of ICU admission) to day 3 in patients with sepsis: routine biochemistry and inflammatory markers such as C-reactive

T

protein and procalcitonin and coagulation test parameters including prothrombin time,

IP

prothrombin time-international normalized ratio (PT-INR), fibrinogen, fibrin degradation

SC R

products (FDP), antithrombin (AT), plasminogen, and thrombin-AT complex (TAT). All assays were performed in the clinical laboratories of Jichi Medical University Hospital.

NU

Data analysis

We defined a subsequent decrease in platelet count as a decrease of >30% and/or platelet

MA

count of <80×103/µL within 5 days of ICU admission. The study population was grouped according to the presence or absence of a subsequent decrease in platelet count. The control

TE

D

group included patients without sepsis admitted after elective surgery. Statistical differences between the groups were analyzed using Student’s t-test, the Wilcoxon rank-sum test, and the

CE P

χ² or Fisher’s exact test as appropriate. Differences in IPF among the three groups were compared with one-way analysis of variance, followed by the Tukey–Kramer post-hoc test.

AC

Receiver operating characteristic curve analysis and the derived area under the curve (AUC) statistic provided prognostic accuracy of a marker or composite score for 28-day mortality. The prognostic value of the variables was also evaluated with univariate and multivariate analysis in the Cox regression model. All P-values reported are two-tailed, and P<0.05 was considered statistically significant. Data were analyzed using JMP version 12 (SAS Institute, Tokyo, Japan).

-8-

ACCEPTED MANUSCRIPT 9 Results

T

Patient characteristics and outcomes

IP

One hundred eighty-four patients with sepsis were admitted to the ICU during the study

SC R

period. Eighty-three patients fulfilled one or more of the exclusion criteria, and the data of the remaining 101 patients were subject to analysis. Forty-eight patients without sepsis admitted

NU

to the ICU after elective surgery were enrolled in the control group.

MA

The baseline characteristics and outcomes of the study cohort are shown in Table 1 and Table S1. Of the 101 patients with sepsis, 47 (46.5%) exhibited a subsequent decrease in platelet count within the first 5 days of the ICU stay (Table 1). Patients who developed a

TE

D

subsequent decrease in platelet count were more severely ill with higher APACHE II and SOFA scores on ICU admission than patients in whom the platelet count did not subsequently

CE P

decline. Platelet concentrate was administered to patients with a decreased platelet count significantly more often than to those without a decrease in platelet count (7.9% vs. 0.0%,

AC

respectively; P=0.0016). There were no significant differences between the two groups in the proportion of patients transfused with red blood cells or fresh frozen plasma (12.9% vs. 8.9%, P=0.23; 5.9% vs. 3.0%, P=0.30, respectively).

Elevation of IPF in septic patients with a subsequent decrease in platelet count The changes in IPF and platelet count from days 1 to 5 in patients with and without a subsequent decrease in platelet count are shown in Figure 1. The IPF rose on days 1 to 5 in patients in whom the platelet count subsequently fell (Figure 1a). In contrast, IPF was stable throughout days 1 to 5 in patients in whom the platelet count did not decrease (Figure 1b).

-9-

ACCEPTED MANUSCRIPT 10 The IPF on the day of ICU admission was highest in patients in whom the platelet count subsequently declined and slightly less elevated in patients without a subsequent

T

decrease in platelet count, as compared with controls (median 4.3% [range 3.1%–8.1%] ,

IP

3.7% [2.6%–4.6%], 2.1% [1.6%–3.5%], respectively; P <0.0001; Figure 2a). The AIPC was

SC R

significantly higher in both groups with sepsis compared with controls (4.2 [3.0–6.4] ×103/L); however, there was no significant difference in AIPC between septic patients with versus without a subsequent decrease in platelet count (median, 7.6 [4.2–10.0] vs. 5.9 [4.2–

NU

8.7] ×103/L; P=0.32; Figure 2b). These data suggest that platelet production was normal or

MA

even increased in the acute phase of sepsis; hence, the elevated IPF may have been due to

D

platelet consumption.

TE

Association of deranged coagulation and fibrinolytic biomarkers with a subsequent decrease in platelet count in patients with sepsis

CE P

We measured indices of coagulation and fibrinolytic biomarkers on day 1 (the day of ICU admission) to day 3 and assessed their relationship with a subsequent decrease in platelet

AC

count. On the day of admission, PT-INR, a global coagulation marker, and TAT, the marker of thrombin generation, were higher in patients with than without a subsequent decrease in platelet count (Figure 3). Other biomarkers, such as AT and plasminogen, which represent anticoagulant and fibrinolytic activity, respectively, were lower on day 1 in patients with than without a subsequent decrease in platelet count. Furthermore, the patients with a subsequent decrease in platelet count exhibited more severe coagulopathy, as indicated by a higher PT-INR and TAT and lower plasminogen, until day 3.

Elevated IPF as a strong predictor of 28-day mortality

-10-

ACCEPTED MANUSCRIPT 11 To evaluate the impact of IPF on mortality, we compared IPF on admission and APACHE II scores, an established prognostic score for critical illness, in patients with sepsis who died

T

within 28 days with those in survivors. We found that nonsurvivors displayed a significantly

IP

higher IPF and APACHE II score at admission than did survivors (IPF: 8.2% [5.6%–10.9%]

SC R

vs. 3.8% [2.7%–5.7%], P=0.0003; APACHE II: 30.5 [27.3–34.0] vs. 19.0 [16.0–26.0], P=0.0008). Multivariate Cox regression analysis including the APACHE II score revealed that IPF was an independent predictive marker for 28-day mortality in patients with sepsis (Table

NU

2). Next, we conducted a receiver operating characteristic analysis to describe the AUC for

MA

prognostic accuracy. We found that IPF showed a high AUC for 28-day mortality (AUC 0.886 [95% confidence interval 0.754–0.952], P=0.0002), similar to that of the APACHE II score (0.857 [0.748–0.924], P=0.0017) (Figure 4). Moreover, the combination of IPF and APACHE

TE

D

II scores increased the AUC for 28-day mortality to 0.912 [0.790–0.966] (vs. IPF, P=0.0029; vs. APACHE II scores, P=0.032). In contrast, markers of systemic inflammation, such as

CE P

C-reactive protein (0.523 [0.320–0.718], P=0.74) and procalcitonin (0.455 [0.285–0.636], P=0.87), and the coagulation biomarkers (PT-INR, 0.614 [0.323–0.841], P=0.035; FDP, 0.723

AC

[0.546–0.850], P=0.36; AT, 0.549 [0.304–0.773], P=0.56; plasminogen, 0.581 [0.289–0.825], P=0.36; and TAT, 0.589 [0.313–0.819], P=0.092) on ICU admission showed a low AUC or results that did not reach statistical significance.

-11-

ACCEPTED MANUSCRIPT 12 Discussion In this observational study, we demonstrated that IPF could predict a subsequent decrease in

T

platelet count in association with severe alterations in coagulation. We also found that the

IP

calculated AIPC was not different between patients with versus without a decline in platelet

SC R

count. These findings suggest that the underlying mechanism of IPF elevation in the acute phase of sepsis was mainly coagulopathy-related platelet consumption rather than increased thrombopoiesis. Considering that an increased IPF was observed before an overt decrease in

NU

platelet count, IPF could be an early indicator of septic coagulopathy. Furthermore, an

MA

elevated IPF was a strong independent predictor of 28-day mortality in patients with sepsis, with prognostic accuracy similar to that of routinely used composite scoring systems such as

TE

D

the APACHE II score.

In recent years, studies have revealed substantial evidence that IPF and its derivative,

CE P

AIPC, are useful tools for evaluating patients with thrombocytopenia. In particular, the clinical utility of IPF is well established in the diagnosis of peripheral thrombocytopenia due

AC

to platelet consumption, such as autoimmune thrombocytopenic purpura and thrombotic thrombocytopenic purpura. Barsam et al. [21] reported that IPF was increased in patients with autoimmune thrombocytopenic purpura, reflecting peripheral platelet destruction, and that calculation of AIPC was useful in assessing the effect of treatment on platelet production. Bat et al. [22] showed that AIPC was not affected by platelet transfusion and thus a reliable marker for bone marrow production.

In our study, an increased IPF was observed in association with a later decline in platelet count and increased severity of coagulopathy in patients with sepsis. Microvascular thrombus formation and platelet adhesion occur in patients with sepsis [23], and a decrease in -12-

ACCEPTED MANUSCRIPT 13 platelet count has been documented in DIC in both humans and animal models of sepsis [24, 25]. Given that AIPC, which is a marker of thrombopoiesis, was not different between the two

T

groups of patients with sepsis, our findings suggest that the decline in platelet count in septic

IP

coagulopathy is not due to bone marrow suppression but is a consequence of platelet

SC R

consumption. This is the first study to use IPF to evaluate platelet production and consumption in the acute phase of sepsis, and these results provide additional evidence to support the theory that thrombocytopenia in sepsis is mainly caused by consumption

MA

NU

coagulopathy [1, 2, 26].

The AIPC at the time of ICU admission were higher in patients with sepsis than in controls in this study. The mechanism of AIPC elevation without an associated decrease in

TE

D

platelet count remains unclear; however, the finding suggests that platelet production is increased, possibly from ongoing inflammatory response to sepsis. Recent studies have

CE P

reported that IPF could be a predictive or diagnostic marker for the development of sepsis. De Blasi et al. [14] reported that IPF was increased before sepsis became clinically overt and was

AC

a predictor of sepsis when procalcitonin, C-reactive protein, and the white blood cell count were not. In other studies, an increased IPF was related to suspected bacterial infection and the severity of sepsis [15, 27]. Although none of those studies examined the coagulation profile of participants in detail, IPF elevation suggested the onset of severe infection or sepsis. Cytokines such as interleukin-6 are known to have thrombopoietic properties [28], and activated platelets themselves release thrombopoietin [29], which stimulates the proliferation and differentiation of megakaryocytes in the course of sepsis. Another possibility is that AIPC was increased as a result of reactive platelet production against mild consumption. In our study, patients with sepsis in whom platelet count did not decline exhibited mild coagulation

-13-

ACCEPTED MANUSCRIPT 14 abnormalities (increased PT-INR, FDP, and TAT), which may reflect a small amount of

T

platelet consumption due to mild coagulopathy.

IP

In our study, IPF was significantly associated with mortality in patients with sepsis.

SC R

This result may reflect the finding that IPF was an early predictor of platelet count decline and its possible underlying mechanisms of septic coagulopathy. The pathophysiology of coagulopathy or DIC is characterized by systemic intravascular activation of coagulation and

NU

microvascular endothelial injury with impaired anticoagulation and insufficient fibrinolysis,

MA

which leads to thrombin generation and widespread thrombosis in the microvasculature. Excessive thrombin and subsequent fibrin deposition exacerbate inflammation and ischemia, contributing to organ damage [26]. A number of studies have reported on the association

TE

D

between DIC and organ failure and found that DIC is an independent risk factor for mortality in patients with sepsis [9, 30]. Intervention to treat coagulopathy in the initial phase may

CE P

therefore be a key factor in improving outcomes in patients with sepsis.

AC

Accepted criteria for the diagnosis of DIC have some potential weaknesses in the early phase of the disorder [31, 32], mainly because they use the platelet count as one of the key components for establishing the diagnosis of DIC. Platelets are consumed over time in the presence of ongoing septic coagulopathy, and it takes several days for the platelet count to fall sufficiently for the diagnostic criteria to be fulfilled. In our study, an increased IPF was observed at the time of ICU admission, prior to any decline in platelet count in patients with sepsis. The IPF can be determined as part of a routine complete blood count, meaning that the results are available quickly; therefore, IPF has the potential to be a useful diagnostic marker of impending DIC in routine clinical practice.

-14-

ACCEPTED MANUSCRIPT 15 This study had some limitations. First, it was a single-center study and the patient population was relatively small. We elected to include patients with a relatively low platelet

T

count (≥80×103/L) on ICU admission to ensure that we recruited a sufficient number of

IP

patients with a subsequent decline in platelet count. Second, we excluded patients with

SC R

hematologic disorders, chronic liver failure, or a history of myelosuppressive therapy to eliminate possible causes of bone marrow dysfunction, thrombocytopenia, or coagulopathy other than sepsis. Additionally, admission AIPC was higher in patients with sepsis, compared

NU

with controls, and the reticulocyte count in peripheral blood was within the normal range in

MA

patients with sepsis (data not shown), suggesting no evidence of bone marrow dysfunction or decreased thrombopoiesis, at least on ICU admission. Further research will be needed to

D

examine the clinical utility of IPF in these patient subgroups. Third, we had to exclude some

TE

study patients because of missing laboratory data. Characteristics of the excluded patients, however, were not significantly different compared with those of included patients, as to age,

CE P

sex, severity of illness, proportion of patients with platelet count decline, or prognosis. This exclusion could have introduced small bias into the study. Fourth, some patients received

AC

platelet concentrate transfusion after measurement of IPF on day 1, which might alter the time course of changes in IPF to some extent. Bat et al. reported that IPF decreased significantly owing to the dilutional effect in transfused patients. In our study, however, IPF increased over time, reflecting increased platelet production in patients with sepsis (Figure 1). Additionally, our rough indication for platelet transfusion was platelet count <20×103/L. The impact of transfusion on IPF might be small in this study. Finally, we could not measure the serum thrombopoietin concentration, despite the fact that it is a key regulator of thrombopoiesis. In addition to its role in platelet production, thrombopoietin directly modulates the hemostatic potential of mature platelets, enhancing platelet activation and platelet-leukocyte adhesion [33, 34]. An elevated thrombopoietin concentration has been reported in healthy volunteers after -15-

ACCEPTED MANUSCRIPT 16 endotoxin infusion [35] and in patients with sepsis [36]. Zakynthinos et al. [36] showed that the serum thrombopoietin concentration was higher in patients with sepsis than in controls

T

and was more closely correlated with sepsis severity than was the decrease in platelet count

IP

alone. Bjerre et al. [37] reported in their small study that the thrombopoietin concentration

SC R

was elevated in patients with meningococcal septicemia, whereas the circulating platelet count was low. These results appear to concur with our finding that IPF was higher in patients

AC

CE P

TE

D

MA

was a subsequent decline in platelet count.

NU

with sepsis than in controls and was further increased in patients with sepsis in whom there

-16-

ACCEPTED MANUSCRIPT 17 Conclusions The IPF was elevated in the early phase of sepsis owing to platelet consumption associated

T

with coagulopathy. As a simple and readily available cellular marker, IPF could be a useful

IP

tool for illuminating the underlying pathophysiologic mechanisms of thrombocytopenia in

SC R

sepsis, and identifying patients with impending septic coagulopathy and poor prognosis in

AC

CE P

TE

D

MA

NU

routine clinical practice.

-17-

ACCEPTED MANUSCRIPT 18 References

AC

CE P

TE

D

MA

NU

SC R

IP

T

[1] Z. Li, F. Yang, S. Dunn, A.K. Gross, S.S. Smyth, Platelets as immune mediators: their role in host defense responses and sepsis, Thromb Res 127(3) (2011) 184-8. [2] J.N. Katz, K.P. Kolappa, R.C. Becker, Beyond thrombosis: the versatile platelet in critical illness, Chest 139(3) (2011) 658-68. [3] S.S. Smyth, R.P. McEver, A.S. Weyrich, C.N. Morrell, M.R. Hoffman, G.M. Arepally, P.A. French, H.L. Dauerman, R.C. Becker, P. Platelet Colloquium, Platelet functions beyond hemostasis, J Thromb Haemost 7(11) (2009) 1759-66. [4] M. Levi, T. van der Poll, Inflammation and coagulation, Crit Care Med 38(2 Suppl) (2010) S26-34. [5] H. Ten Cate, Trombocytopenia: one of the markers of disseminated intravascular coagulation, Pathophysiol Haemost Thromb 33(5-6) (2003) 413-6. [6] S. Vanderschueren, A. De Weerdt, M. Malbrain, D. Vankersschaever, E. Frans, A. Wilmer, H. Bobbaers, Thrombocytopenia and prognosis in intensive care, Crit Care Med 28(6) (2000) 1871-6. [7] S. Akca, P. Haji-Michael, A. de Mendonca, P. Suter, M. Levi, J.L. Vincent, Time course of platelet counts in critically ill patients, Crit Care Med 30(4) (2002) 753-6. [8] D. Moreau, J.F. Timsit, A. Vesin, M. Garrouste-Orgeas, A. de Lassence, J.R. Zahar, C. Adrie, F. Vincent, Y. Cohen, B. Schlemmer, E. Azoulay, Platelet count decline: an early prognostic marker in critically ill patients with prolonged ICU stays, Chest 131(6) (2007) 1735-41. [9] S. Zeerleder, C.E. Hack, W.A. Wuillemin, Disseminated intravascular coagulation in sepsis, Chest 128(4) (2005) 2864-75. [10] Y. Abe, H. Wada, H. Tomatsu, A. Sakaguchi, J. Nishioka, Y. Yabu, K. Onishi, K. Nakatani, Y. Morishita, S. Oguni, T. Nobori, A simple technique to determine thrombopoiesis level using immature platelet fraction (IPF), Thromb Res 118(4) (2006) 463-9. [11] C. Briggs, S. Kunka, D. Hart, S. Oguni, S.J. Machin, Assessment of an immature platelet fraction (IPF) in peripheral thrombocytopenia, Br J Haematol 126(1) (2004) 93-9. [12] Y. Koike, K. Miyazaki, M. Higashihara, E. Kimura, M. Jona, K. Uchihashi, A. Masuda, K. Kandabashi, M. Kurokawa, Y. Yatomi, Clinical significance of detection of immature platelets: comparison between percentage of reticulated platelets as detected by flow cytometry and immature platelet fraction as detected by automated measurement, Eur J Haematol 84(2) (2010) 183-4. [13] I. Pons, M. Monteagudo, G. Lucchetti, L. Munoz, G. Perea, I. Colomina, J. Guiu, J. Obiols, Correlation between immature platelet fraction and reticulated platelets. Usefulness in the etiology diagnosis of thrombocytopenia, Eur J Haematol 85(2) (2010) 158-63. [14] R.A. De Blasi, P. Cardelli, A. Costante, M. Sandri, M. Mercieri, R. Arcioni, Immature platelet fraction in predicting sepsis in critically ill patients, Intensive Care Med 39(4) (2013) 636-43. [15] R.M. Enz Hubert, M.V. Rodrigues, B.D. Andreguetto, T.M. Santos, M. de Fatima Pereira Gilberti, V. de Castro, J.M. Annichino-Bizzacchi, D. Dragosavac, M.A. Carvalho-Filho, E.V. De Paula, Association of the immature platelet fraction with sepsis diagnosis and severity, Sci Rep 5 (2015) 8019. [16] R.P. Dellinger, M.M. Levy, A. Rhodes, D. Annane, H. Gerlach, S.M. Opal, J.E. Sevransky, C.L. Sprung, I.S. Douglas, R. Jaeschke, T.M. Osborn, M.E. Nunnally, S.R. Townsend, K. Reinhart, R.M. Kleinpell, D.C. Angus, C.S. Deutschman, F.R. Machado, G.D. Rubenfeld, S. Webb, R.J. Beale, J.L. Vincent, R. Moreno, S. Surviving Sepsis Campaign Guidelines Committee including The Pediatric, Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012, Intensive Care Med 39(2) (2013) 165-228. [17] R.P. Dellinger, M.M. Levy, J.M. Carlet, J. Bion, M.M. Parker, R. Jaeschke, K. Reinhart, D.C. Angus, C. Brun-Buisson, R. Beale, T. Calandra, J.F. Dhainaut, H. Gerlach, M. Harvey, J.J. Marini, J. Marshall, M. Ranieri, G. Ramsay, J. Sevransky, B.T. Thompson, S. Townsend, J.S. Vender, J.L. Zimmerman, J.L. Vincent, Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008, Intensive Care Med 34(1) (2008) 17-60. [18] W.A. Knaus, E.A. Draper, D.P. Wagner, J.E. Zimmerman, APACHE II: a severity of disease classification system, Crit Care Med 13(10) (1985) 818-29. [19] J.L. Vincent, R. Moreno, J. Takala, S. Willatts, A. De Mendonca, H. Bruining, C.K. Reinhart, P.M. Suter, L.G. Thijs, 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 Med 22(7) (1996) 707-10. [20] L. National Heart, N. Blood Institute Acute Respiratory Distress Syndrome Clinical Trials, H.P. Wiedemann, A.P. Wheeler, G.R. Bernard, B.T. Thompson, D. Hayden, B. deBoisblanc, A.F. Connors, Jr., R.D. Hite, A.L. Harabin, Comparison of two fluid-management strategies in acute lung injury, N Engl J Med 354(24) (2006) 2564-75. -18-

ACCEPTED MANUSCRIPT 19

AC

CE P

TE

D

MA

NU

SC R

IP

T

[21] S.J. Barsam, B. Psaila, M. Forestier, L.K. Page, P.A. Sloane, J.T. Geyer, G.O. Villarica, M.M. Ruisi, T.B. Gernsheimer, J.H. Beer, J.B. Bussel, Platelet production and platelet destruction: assessing mechanisms of treatment effect in immune thrombocytopenia, Blood 117(21) (2011) 5723-32. [22] T. Bat, S.F. Leitman, K.R. Calvo, D. Chauvet, C.E. Dunbar, Measurement of the absolute immature platelet number reflects marrow production and is not impacted by platelet transfusion, Transfusion 53(6) (2013) 1201-4. [23] M. Gawaz, T. Dickfeld, C. Bogner, S. Fateh-Moghadam, F.J. Neumann, Platelet function in septic multiple organ dysfunction syndrome, Intensive Care Med 23(4) (1997) 379-85. [24] N.H. Whitworth, M.A. Barradas, D.P. Mikhailidis, P. Dandona, An investigation into the effects of bacterial lipopolysaccharide on human platelets, Eur J Haematol 43(2) (1989) 112-9. [25] F.B. Taylor, Jr., H. Wada, G. Kinasewitz, Description of compensated and uncompensated disseminated intravascular coagulation (DIC) responses (non-overt and overt DIC) in baboon models of intravenous and intraperitoneal Escherichia coli sepsis and in the human model of endotoxemia: toward a better definition of DIC, Crit Care Med 28(9 Suppl) (2000) S12-9. [26] F.B. Taylor, Jr., C.H. Toh, W.K. Hoots, H. Wada, M. Levi, T. Scientific Subcommittee on Disseminated Intravascular Coagulation of the International Society on, Haemostasis, Towards definition, clinical and laboratory criteria, and a scoring system for disseminated intravascular coagulation, Thromb Haemost 86(5) (2001) 1327-30. [27] A. Di Mario, M. Garzia, F. Leone, A. Arcangeli, L. Pagano, G. Zini, Immature platelet fraction (IPF) in hospitalized patients with neutrophilia and suspected bacterial infection, J Infect 59(3) (2009) 201-6. [28] E.M. Wolber, W. Jelkmann, Interleukin-6 increases thrombopoietin production in human hepatoma cells HepG2 and Hep3B, J Interferon Cytokine Res 20(5) (2000) 499-506. [29] C.C. Folman, G.E. Linthorst, J. van Mourik, G. van Willigen, E. de Jonge, M. Levi, M. de Haas, A.E. von dem Borne, Platelets release thrombopoietin (Tpo) upon activation: another regulatory loop in thrombocytopoiesis?, Thromb Haemost 83(6) (2000) 923-30. [30] J.F. Dhainaut, S.B. Yan, D.E. Joyce, V. Pettila, B. Basson, J.T. Brandt, D.P. Sundin, M. Levi, Treatment effects of drotrecogin alfa (activated) in patients with severe sepsis with or without overt disseminated intravascular coagulation, J Thromb Haemost 2(11) (2004) 1924-33. [31] D. Oh, M.J. Jang, S.J. Lee, S.Y. Chong, M.S. Kang, H. Wada, Evaluation of modified non-overt DIC criteria on the prediction of poor outcome in patients with sepsis, Thromb Res 126(1) (2010) 18-23. [32] H. Wada, T. Hatada, K. Okamoto, T. Uchiyama, K. Kawasugi, T. Mayumi, S. Gando, S. Kushimoto, Y. Seki, S. Madoiwa, T. Okamura, C.H. Toh, D.I.C.s. Japanese Society of Thrombosis Hemostasis, Modified non-overt DIC diagnostic criteria predict the early phase of overt-DIC, Am J Hematol 85(9) (2010) 691-4. [33] A. Oda, Y. Miyakawa, B.J. Druker, K. Ozaki, K. Yabusaki, Y. Shirasawa, M. Handa, T. Kato, H. Miyazaki, A. Shimosaka, Y. Ikeda, Thrombopoietin primes human platelet aggregation induced by shear stress and by multiple agonists, Blood 87(11) (1996) 4664-70. [34] H.E. Tibbles, C.S. Navara, M.A. Hupke, A.O. Vassilev, F.M. Uckun, Thrombopoietin induces p-selectin expression on platelets and subsequent platelet/leukocyte interactions, Biochem Biophys Res Commun 292(4) (2002) 987-91. [35] P. Stohlawetz, C.C. Folman, A.E. von dem Borne, T. Pernerstorfer, H.G. Eichler, S. Panzer, B. Jilma, Effects of endotoxemia on thrombopoiesis in men, Thromb Haemost 81(4) (1999) 613-7. [36] S.G. Zakynthinos, S. Papanikolaou, T. Theodoridis, E.G. Zakynthinos, V. Christopoulou-Kokkinou, G. Katsaris, A.C. Mavrommatis, Sepsis severity is the major determinant of circulating thrombopoietin levels in septic patients, Crit Care Med 32(4) (2004) 1004-10. [37] A. Bjerre, R. Ovstebo, P. Kierulf, S. Halvorsen, P. Brandtzaeg, Fulminant meningococcal septicemia: dissociation between plasma thrombopoietin levels and platelet counts, Clin Infect Dis 30(4) (2000) 643-7.

-19-

ACCEPTED MANUSCRIPT 20 Acknowledgements The authors thank the nursing staff of the intensive care unit and clinical laboratory

IP

T

technicians at Jichi Medical University Hospital for their invaluable assistance.

SC R

Authors’ contributions

TM and KK conceived and designed the study. TM prepared the data for analysis. TM and KK conducted the data analysis and drafted the article. AKL, SN, and MS supervised the

NU

study. All authors read and approved the manuscript. TM and KK take responsibility for the

MA

paper as a whole.

Conflicts of interest

AC

CE P

TE

D

None.

-20-

ACCEPTED MANUSCRIPT 21 Figure legends Figure 1. Changes in immature platelet fraction and platelet count over days 1 to 5 of

T

intensive care unit (ICU) admission in patients (a) with and (b) without a subsequent decrease

IP

in platelet count. Data are expressed as the mean with the 95% confidence interval shown by

SC R

the error bars.

Figure 2. Box plot showing (a) immature platelet fraction and (b) absolute immature platelet

NU

count on the day of intensive care unit admission in patients with a subsequent decrease in

MA

platelet count, patients without a subsequent decrease in platelet count, and controls.

Figure 3. Changes in hemostatic biomarkers over days 1 to 3 of intensive care unit admission.

TE

D

(a) Prothrombin time-international normalized ratio (PT-INR), (b) fibrin degradation products, (c) thrombin-antithrombin complex, (d) antithrombin, and (e) plasminogen in patients with

CE P

and without a subsequent decrease in platelet count. Data are expressed as mean with 95%

AC

confidence intervals.

Figure 4. Receiver operator characteristic curves comparing the sensitivity and specificity of immature platelet fraction (IPF) at the time of intensive care unit admission, the Acute Physiology and Chronic Health Evaluation II (APACHE II) score, and the combination of these two markers in predicting 28-day mortality. The figure shows the area under the receiver operating characteristic curve for IPF (0.886 [95% CI, 0.754–0.952]) and APACHE II score (0.857 [95% CI, 0.748–0.924]). The combined IPF and APACHE II score was superior to either marker alone (0.912 [95% CI, 0.790–0.966]; vs. IPF, P=0.0029; vs. APACHE II score, P=0.032).

-21-

ACCEPTED MANUSCRIPT 22

AC

CE P

TE

D

MA

NU

SC R

IP

T

Fig. 1

-22-

ACCEPTED MANUSCRIPT 23

AC

CE P

TE

D

MA

NU

SC R

IP

T

Fig. 2

-23-

ACCEPTED MANUSCRIPT 24

AC

CE P

TE

D

MA

NU

SC R

IP

T

Fig. 3

-24-

ACCEPTED MANUSCRIPT 25

AC

CE P

TE

D

MA

NU

SC R

IP

T

Fig. 4

-25-

ACCEPTED MANUSCRIPT 26

Table 1: Baseline characteristics and outcomes of patients with sepsis No decrease

Subsequent decrease

in platelet count

in platelet count

(n = 54)

(n = 47)

All patients with sepsis P-value*

(n = 101)

IP

T

Demographics 67 (59-78)

65 (58-78)

69 (62-80)

0.14

Male sex (n, %)

64 (63.4%)

34 (63.0%)

30 (63.8%)

1.00

8 (14.8%)

13 (27.7%)

0.14

33 (61.1%)

23 (48.9%)

0.23

2 (3.7%)

2 (4.3%)

1.00

19 (18.8%)

11 (20.3%)

8 (17.0%)

0.80

1 (1.0%)

0 (0%)

1 (2.1%)

0.47

23 (22.8%)

16 (29.6%)

7 (14.9%)

0.09

7 (6.9%)

4 (7.4%)

3 (6.4%)

1.00

22 (21.8%)

13 (24.1%)

9 (19.2%)

0.63

14 (13.9%)

5 (9.3%)

9 (19.2%)

0.25

21.0 (16.0–27.0)

17.5 (14.0-24.3)

25.0 (18.0-30.0)

0.0006

6 (4–8)

4 (2-6)

8 (6-10)

<0.0001

21 (16–24)

23 (20-25)

19 (13-21)

<0.0001

8 (7.9%)

1 (1.9%)

7 (14.9%)

0.024

SC R

Age (years, range)

Pulmonary infection (n, %)

21 (20.8%)

Abdominal infection (n, %)

56 (55.5%) 4 (4.0%)

MA

Urinary tract infection (n, %) Soft tissue infection (n, %)

D

Central nervous system infection (n, %)

TE

Comorbidities Hypertension (n, %)

Diabetes Mellitus (n, %)

AC

Chronic Kidney Disease (n, %)

CE P

Chronic Heart Failure (n, %)

Severity of illness

NU

Source of sepsis

APACHE II (median, range) Organ dysfunction

SOFA score** (median, range) Prognosis ICU-free days (median, range) 28-day mortality (n, %)

Data are expressed as median (interquartile range), or number (%).

-26-

ACCEPTED MANUSCRIPT 27 Abbreviations: APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, Sepsis-related Organ Failure Assessment; ICU, intensive care unit. *Comparison of groups with and without subsequent decrease in platelet count.

AC

CE P

TE

D

MA

NU

SC R

IP

T

**SOFA scores do not include coagulation parameter (platelet count).

-27-

ACCEPTED MANUSCRIPT 28 Table 2: Cox regression analysis for IPF and APACHE II score at ICU admission as predictors of 28-day mortality Univariate analysis

Multivariate analysis

95% CI

P-value

Hazard Ratio

95% CI

P-value

IPF

1.38

1.18-1.61

0.0002

1.33

1.11-1.58

0.0007

APACHE II

1.10

1.04-1.18

0.0023

1.09

1.01-1.16

0.014

IP

T

Hazard Ratio

SC R

Abbreviations: APACHE, Acute Physiology and Chronic Health Evaluation; CI, confidence intervals;

AC

CE P

TE

D

MA

NU

IPF, Immature Platelet Fraction

-28-

ACCEPTED MANUSCRIPT 29 Highlights 

Immature platelet fraction is increased in the early phase of sepsis due to platelet

Increased immature platelet fraction predicts decrease in platelet count in septic

IP



T

consumption.

Immature platelet fraction indicates severe coagulopathy and poor prognosis in septic

CE P

TE

D

MA

NU

patients.

AC



SC R

coagulopathy.

-29-