Вы находитесь на странице: 1из 6

Egyptian Journal of Chest Diseases and Tuberculosis (2016) 65, 499–504

H O S T E D BY
The Egyptian Society of Chest Diseases and Tuberculosis

Egyptian Journal of Chest Diseases and Tuberculosis


www.elsevier.com/locate/ejcdt
www.sciencedirect.com

ORIGINAL ARTICLE

Platelet count: Is it a possible marker for severity


and outcome of community acquired pneumonia?
Ashraf A. ElMaraghy a,*, Eman B. AbdelFattah a,1, Mostafa S. Ahmed b,2

a
Chest Department, Ain Shams University, Cairo, Egypt
b
Ain Shams University Specialized Hospital, Cairo, Egypt

Received 5 August 2015; accepted 14 September 2015


Available online 9 October 2015

KEYWORDS Abstract Background: Platelets have been increasingly recognized as an important component of
Community acquired innate and adaptive immunity. Platelet response in antimicrobial host defense is similar, in many
pneumonia; ways, to the leukocyte response: both cell types contain antimicrobial peptides that act against a
Platelet count; broad range of pathogens.
Ain Shams University Objective: The aim of this study was to detect whether platelet count could be used as a marker
Hospital; for severity of community acquired pneumonia or not.
Respiratory intensive care Subjects and methods: The study included 40 cases of community acquired pneumonia admitted at
unit Chest Department and Respiratory ICU at Ain Shams University Hospital, as well as, Chest Depart-
ment and Respiratory ICU at Ain Shams University Specialized Hospital. All cases were subjected to
the following: full history taking, thorough clinical examination including general and local examina-
tion, arterial blood gases, electrocardiography, radiological work up, routine laboratory investiga-
tions including compete blood picture and CURB-65 score as a marker for severity of pneumonia.
Results: The results showed that there was a significant relation between the occurrence of respi-
ratory complications and both thrombocytopenia and thrombocytosis. Also, there was a significant
relation between both thrombocytopenia and thrombocytosis and the CURB-65 score as a score
for severity of CAP. There was a significant relation between both thrombocytopenia and thrombo-
cytosis and mortality among the patients with CAP. Finally, platelet count is considered better pos-
itive than a negative predictor value to the outcome.
Conclusion: A better understanding of the role of platelets in the outcomes of patients with com-
munity acquired pneumonia may generate new prognostic and therapeutic modalities for patients
with severe disease.
Ó 2015 The Authors. Production and hosting by Elsevier B.V. on behalf of The Egyptian Society of Chest
Diseases and Tuberculosis. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).

* Corresponding author. Tel.: +20 1001770702.


E-mail addresses: ashrafelmaraghy@yahoo.com (A.A. ElMaraghy), Introduction
emanbadawy2006@yahoo.com (E.B. AbdelFattah), drmostafa_
shawki@yahoo.com (M.S. Ahmed).
1
Tel.: +20 1001770703. Platelets have been increasingly recognized as an important
2
Tel.: +20 1001452044. component of innate and adaptive immunity. Platelet response
Peer review under responsibility of The Egyptian Society of Chest in antimicrobial host defense is similar, in many ways, to the
Diseases and Tuberculosis.
http://dx.doi.org/10.1016/j.ejcdt.2015.09.001
0422-7638 Ó 2015 The Authors. Production and hosting by Elsevier B.V. on behalf of The Egyptian Society of Chest Diseases and Tuberculosis.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
500 A.A. ElMaraghy et al.

leukocyte response: both cell types contain antimicrobial All cases were subjected to the following:
peptides that act against a broad range of pathogens. After
platelets and neutrophils are activated, they accumulate at 1. Full history taking.
the site of infection to produce direct contact between their 2. Thorough clinical examination including general and local
antimicrobial peptides and invading bacteria. Leukocytes need examination.
to phagocytize bacteria to achieve interaction with intracellu- 3. Arterial blood gases.
lar peptides; platelets can also internalize microorganisms into 4. Electrocardiography.
phagosome-like vacuoles, enhancing pathogen clearance. 5. Radiological work up (conventional CXR, CT scan and/or
Antimicrobial peptides from both cells exert a rapid, potent, chest ultrasound).
and direct antimicrobial effect that contributes to limiting 6. Routine laboratory investigations including compete blood
the infection [1]. picture.
Clinicians have always evaluated the degree of leucocytosis 7. CURB-65 score tool.
in patients with pneumonia as an indication of systemic
inflammatory response and severity of disease. Thrombocy- CURB-65 Mortality Prediction Tool for Patients with
topenia is also a recognized marker of poor outcomes in Community-Acquired Pneumonia [5].
patients with pneumonia due to the association of low platelet
counts with disseminated intravascular coagulation and severe
sepsis. Abnormalities in the coagulation system can be due to
Prognostic variables*
low as well as high platelet count. However, the association
Confusion
between thrombocytosis and clinical outcomes in adult
Blood urea nitrogen level >20 mg per dL (7.14 mmol per L)
patients with community-acquired pneumonia (CAP) has not Respiratory rate P30 breaths per minute
been investigated [2]. Blood pressure (systolic <90 mmHg or diastolic 660 mmHg)
Considering that platelets play a crucial role in antimicro- Age P65 years
bial host defenses and the coagulate system, it is hypothesized Score Inpatient vs. outpatient 30-day mortality (%)
that an abnormal platelet count may be an important marker 0 or 1 point Treat as outpatient 0.7–2.1
to assess severity of disease in patients with CAP [3,4]. 2 points Treat as inpatient 9.2
P3 points Treat in intensive care unit 15–40
*
Assign 1 point for each variable
Subjects and methods

This was a prospective study of the relation between platelet


count abnormality and the severity and outcome of
Study definitions
community-acquired pneumonia cases. The study included
40 cases of CAP admitted at the Chest Department and Respi-
ratory ICU at Ain Shams University Hospital, as well as, CAP is defined as the presence of a new pulmonary infil-
Chest Department and Respiratory ICU at Ain Shams Univer- trate on chest radiograph at the time of hospitalization
sity Specialized Hospital. associated with at least one of the following: (1) new or
increased cough, (2) an abnormal temperature (<35.6 °C
or > 37.8 °C), (3) an abnormal serum leukocyte count
(leucocytosis, left shift, or leucopenia defined by local labo-
Table 1 Distribution of the studied patients as regards the
ratory values). CAP severity was measured using CURB-65
general data.
scoring tool.
Variables Number % Hypotension is defined as a systolic blood pressure
Gender <90 mmHg or diastolic blood pressure <60 mmHg.
Male 30 75 Alteration of gas exchange is defined as Pa O2 <60 mmHg
Female 10 25 or Pa O2/FIO 2 < 300 or O2 saturation <90%.
Site Thrombocytopenia and thrombocytosis are defined as pla-
Lobar 32 80 telet counts <150,000/L or >400,000/L, respectively.
Multilobar 8 20
Patients with immunosuppression, neoplasms, active TB or
Respiratory failure
hematologic diseases were excluded.
Non 8 20
Type 1 12 30
Type 2 20 50 Statistical methodology
Outcome
Alive 17 42.5 Analysis of data was done by IBM computer using SPSS (sta-
Died 23 57.5 tistical program for social science version 16) as follows:
Complications
No 19 47.5
 Description of quantitative variables as mean, SD and
Respiratory 12 30.0
Non respiratory 7 17.5 range.
Both 2 5.0  Description of qualitative variables as number and
percentage.
Length of hospital stay 11.5 ± 5 3–25
 Fisher exact test was used instead of chi-square test when
Age 60 ± 17 17–84
one expected cell or more less than or equal 5.
Platelet count as a marker for severity of community acquired pneumonia 501

Validity parameters from ROC (receiver operator characteristic


Table 3 Correlation between mean of platelet count versus
curve)
age and length of hospital stay.
Variables Platelet count Significance
 Sensitivity = true ve +/true + ve + false –ve.
 = ability of the test to detect + ve cases. r P
 Specificity = true -ve/true-ve + false + ve. Age 0.14 0.56 NS
 = ability of the test to exclude negative cases. Length of hospital stay 0.009 0.54 NS
 PPV (positive predictive value) = true+/true + ve ± false
+ ve.
 =% of true + ve cases to all positive.
 NPV = true-/true-ve + false –ve. Table 4 shows that there was a significant statistical relation
 =% of the true –ve to all negative cases. between the occurrence of respiratory complications and both
 Accuracy = true + ve + true-ve/grand total [6]. thrombocytopenia and thrombocytosis by using chi-square
test.
P value > 0.05 insignificant. Table 5 shows that that there was a significant statistical
P 6 0.05 significant. relation between both thrombocytopenia and thrombocytosis
P < 0.01 highly significant. and the CURB65 score as a score for severity of CAP by using
one way ANOVA test.
Results Table 6 shows that there was a significant statistical relation
between both thrombocytopenia and thrombocytosis and mor-
Table 1 shows that 75% of the enrolled patients were males tality among the patients with CAP by using chi-square test.
and 25% were females, the age ranged from 17 to 84 years with Table 7 shows that platelet count is considered better pos-
a mean of 60 ± 17 years, the length of hospital stay ranged itive than a negative predictor value to the outcome by using
from 3 to 25 days with a mean of 11.5 ± 5 days, 80% of the the receiver operating curve (ROC curve).
patients presented with lobar consolidation while the other
20% presented by multilobar consolidation, 50% of the Discussion
included patients had type II respiratory failure, 30% had type
I respiratory failure, while only 20% of the patients did not Clinicians have always evaluated the degree of leucocytosis in
have respiratory failure, 47.5% of the patients did not develop patients with pneumonia as an indication of systemic inflam-
any complications, 30% developed respiratory complications, matory response and severity of disease. Thrombocytopenia
17.5% developed non-respiratory complications and 5% is also a recognized marker of poor outcomes in patients with
developed both respiratory and non-respiratory complications, pneumonia, due to the association of low platelet counts with
and finally, 42.5% of the patients survived while 57.5% of disseminated intravascular coagulation and severe sepsis [3].
them died. Platelets are important inflammatory cells that can undergo
Table 2 shows that there was no significant statistical rela- chemotaxis and are able to release numerous pro inflammatory
tion between the different variables (gender, site of pneumonia, molecules. In patients with CAP, there is an association
presence or absence of respiratory failure, the need for between levels of inflammatory cytokines and severity of dis-
mechanical ventilation) and the mean of platelet count by ease. It can be theorized that thrombocytosis may favor an
using one way ANOVA test. exaggerated systemic inflammatory response that, in turn, pro-
Table 3 shows that there was no significant statistical corre- duces poor outcomes in patients with CAP [7].
lation between the mean of platelet count versus age and This study was a prospective one about the relation
length of hospital stay by using Spearman correlation. between platelet count abnormality and the severity and out-
come of community-acquired pneumonia cases. The study
Table 2 Correlation between the mean of platelet count included 40 cases of CAP admitted at the Chest Department
versus different variables. and Respiratory ICU at Ain Shams University Hospital, as
well as, the Chest Department and Respiratory ICU at Ain
Variables Platelets count Mean ± SD t P Sig
Shams University Specialized Hospital.
Gender The results showed that 75% of the enrolled patients were
Male 195 ±55 0.2 0.67 NS
males and 25% were females, the age ranged from 17 to
Female 202 ±49
84 years with a mean of 60 ± 17 years, the length of hospital
Site
Lobar 189 ±80 0.9 0.37 NS stay ranged from 3 to 25 days with a mean of 11.5 ± 5 days,7
Multilobar 230 ±110 patients (17.5%) had thrombocytopenia (platelet count
Respiratory failure <150,000), 30 patients (75%) had normal platelet count
Non 221 ±110 1.8 0.17 NS (150,000–400,000) and only 3 patients (7.5%) had thrombocy-
Type 1 144 ±78 tosis (platelet count >400,000), 80% of the patients presented
Type 2 219 ±105.5 with lobar consolidation while the other 20% presented with
MV multilobar consolidation, 50% of the included patients had
Positive 185 ±87 0.79 0.33 NS type II respiratory failure, 30% had type I respiratory failure,
Negative 215 ±110.9
while only 20% of the patients did not have respiratory failure,
#one way ANOVA test. 47.5% of the patients did not develop any complications,
30% developed respiratory complications, 17.5% developed
502 A.A. ElMaraghy et al.

Table 4 Relation between type of platelet count and different complications.


Type of complications Type of platelet count X2 P Sig
Thrombocytopenia Normal platelet count Thrombocytosis
No 2(28.57%) 17(56.67%) 0
Respiratory 5(71.43%) 5(16.66%) 2(66.67%)
Non respiratory 0 6(20.00%) 1(33.33%) 13 0.008 S
Both 0 2(6.67%) 0

Table 5 Relation between type of platelet count and CURB65 score.


Variable Type of platelet count t P Sig
Thrombocytopenia Normal platelet count Thrombocytosis
CURB65 score 4.2 ± 1.5 3 ± 1.3 3.8 ± 1.1 4.2 0.002 S

Table 6 Relation between type of platelet count and outcome.


Variable Type of platelet count X2 P Sig
Thrombocytopenia Normal platelet count Thrombocytosis
Outcome
Alive 2(28.57%) 15(50.0%) 0 4.7 0.05 S
Died 5(71.43%) 15(50.0%) 3(100%)

(89%) had normal platelet count. Lower platelet counts were


Table 7 Validity of platelet count versus outcome. associated with an increase of severity of disease by PSI or
Variables % CURB65. Patients with thrombocytopenia presented more
Best cut off = 150,000
often with severe sepsis and intensive care unit admission,
Area under the curve 0.56 whereas those with thrombocytosis presented more frequently
Sensitivity 70 with respiratory complications such as empyema. No associa-
Specificity 44 tion was found between abnormal platelet count and increased
PPV 53 risk of mortality in the multivariable logistic regression model.
NPV 75 They concluded that abnormal platelet count is associated with
severity of disease and more frequently complications in
patients with CAP.
non-respiratory complications and 5% developed both On the other hand, the results of the present study showed
respiratory and non-respiratory complications, and finally, that the overall mortality rate among the enrolled patients was
42.5% of the patients survived while 57.5% of them died. 57.5%, among the patients with thrombocytopenia was
There was no significant relation between the different 71.43%, among the patients with normal platelet count was
variables (gender, site of pneumonia, presence or absence of 50% and among the patients with thrombocytosis was
respiratory failure, the need for mechanical ventilation) and 100%. By defining a platelet count 150,000 as a cut-off point,
the mean of platelet count. Also, there was no significant the results of our study found that platelet count is considered
correlation between the mean of platelet count versus age better positive than a negative predictor to the outcome. This
and length of hospital stay. partially matches with the results of Georges et al. [9], who
Importantly, there was a significant relation between the conducted a multicenter retrospective study showing, in 822
occurrence of respiratory complications (the majority was in patients admitted to an ICU for severe CAP, that severe
the form of acute respiratory distress syndrome) and both thrombocytopenia (< 50,000 cells/lL) was an independent
thrombocytopenia and thrombocytosis. Moreover, there was predictor of mortality. They looked at the impact of thrombo-
a significant relation between both thrombocytopenia and cytosis in their patients. The overall ICU mortality rate was
thrombocytosis and the CURB65 score as a score for severity 35.4%. Thrombocytosis was present in 70 (5.7%) patients.
of CAP. These results are similar to those of Martinez et al. [8], They did not find any difference in outcome compared with
who conducted a 12 month prospective multicenter and longi- patients with thrombocytosis. When considering the cause of
tudinal study in 10 hospitals of a Spanish Mediterranean area. death according to platelet numbers, they found it was essen-
They included hospitalized adult patients with CAP. They tially related to sepsis complications in patients with thrombo-
analyzed 1314 patients. Forty-three patients (3%) presented cytopenia (septic refractory shock, n = 18; multi organ
with thrombocytopenia, 107 (8%) thrombocytosis and 1164 failure, n = 17; ARDS, n = 11; nosocomial pneumonia,
Platelet count as a marker for severity of community acquired pneumonia 503

n = 8), while in patients with thrombocytosis, the cause of cardiopulmonary and humoral immune function as well as
death was mostly related to complications of ICU stay or asso- specific pathogen data, the platelet count is necessarily an
ciated co morbidity. Thus, they believed that thrombocytope- imprecise prognostic indicator. In immunocompetent adults,
nia remains an important predictor of outcome in patients it would seem that thrombocytopenia and thrombocytosis,
with severe CAP. In these patients, thrombocytosis is not asso- rather than having a prognostic significance, may be more
ciated with worse outcome. The difference between the results important diagnostically in suggesting a specific CAP
of our study and those of Georges et al. might be due to the pathogen.
obvious discrepancy in the number of enrolled patients with
thrombocytosis in both studies. On the contrary, Prina et al. Conclusion
[10] prospectively analyzed 2423 consecutive, hospitalized
patients with CAP. They found that fifty-three patients (2%) A better understanding of the role of platelets in the outcomes
presented with thrombocytopenia, 204 (8%) with thrombocy- of patients with pneumonia may generate new prognostic and
tosis, and 2166 (90%) had normal platelet counts. Patients therapeutic modalities for patients with severe disease.
with thrombocytosis presented more frequently with respira-
tory complications, such as complicated pleural effusion and
Recommendations
empyema, whereas those with thrombocytopenia presented
more often with severe sepsis, septic shock, need for invasive
mechanical ventilation and ICU admission. Patients with 1. Future research is needed to assess whether abnormality in
thrombocytosis and patients with thrombocytopenia had platelet counts in patients with CAP is just a marker of the
longer hospital stays and higher 30-day mortality and readmis- inflammatory response or is in part responsible for
sion rates than those with normal platelet counts. Multivariate mortality. In particular, future research should focus on
analysis confirmed a significant association between thrombo- investigating the cause of death for patients with abnormal
cytosis and 30-day mortality. Adding thrombocytosis to the platelet counts.
confusion, respiratory rate, and BP plus age P65 years score 2. Also, future researches should focus on the relation
slightly improved the accuracy to predict mortality (area under between platelet count abnormality and the causative
the receiver operating characteristic curve increased from 0.634 pathogen of CAP.
to 0.654, P = .049). They concluded that thrombocytosis in 3. When evaluating a complete blood count report in patients
patients with CAP is associated with poor outcome, with CAP, platelet count may be informative for predicting
complicated pleural effusion, and empyema. The presence of patient poor outcomes in addition to the commonly used
thrombocytosis in CAP should encourage ruling out respira- leukocyte count.
tory complication and could be considered for severity
evaluation. References
The results of our study are supported with those of Mir-
saeidi et al. [11], who conducted a retrospective cohort study [1] J.R. Fitzgerald, T.J. Foster, D. Cox, The interaction of bacterial
of 500 consecutive patients admitted with CAP to the Veterans pathogens with platelets, Nat. Rev. Microbiol. 4 (6) (2006) 445–
Administration Medical Center of Louisville, Kentucky, 457.
between June 2001 and March 2006. This study indicated that [2] M.R. Yeaman, A.S. Bayer, Antimicrobial peptides versus
thrombocytopenia and thrombocytosis were significantly asso- invasive infections, Curr. Top. Microbiol. Immunol. 306
ciated with mortality in patients with CAP. At the time of hos- (2006) 111–152.
pitalization, abnormalities in platelet count were better [3] B.D. Elzey, D.L. Sprague, T.L. Ratliff, The emerging role of
platelets in adaptive immunity, Cell. Immunol. 238 (1) (2005) 1–
predictors of clinical outcomes in patients with CAP when
9.
compared with abnormalities in leukocyte count. On the other
[4] L.A. Mandell, R.G. Wunderink, A. Anzueto, Infectious
hand, when Cunha et al. [12] revised the results of Mirsaeidi Diseases Society of America/American Thoracic Society
et al., they were not at all convinced that the platelet count consensus guidelines on the management of community
alone is a prognostic indicator in adults hospitalized with sev- acquired pneumonia in adults, Clin. Infect. Dis. 44 (Suppl. 2)
ere CAP. In both pieces, information was lacking regarding (2007) S27–S72.
pathogen distribution, a key determinant of the platelet count. [5] British Thoracic Society Standards of Care Committee, BTS
Because thrombocytopenia is not an acute-phase reactant, it Guidelines for the Management of Community Acquired
cannot be said that thrombocytopenia with severe CAP is an Pneumonia in Adults, Thorax 56 (Suppl. 4) (2001) 61–64.
indicator of severity and, thereby, is indirectly related to prog- [6] M.C. Miller, R.G. Knapp, Clinical Epidemiology and
Biostatistics, 3rd ed., Williams & Wilkins, Maryland, 1992.
nosis. Any CAP pathogen may present as severe CAP, depend-
[7] G. Antunes, S.A. Evans, J.L. Lordan, A.J. Frew, Systemic
ing on the cardiopulmonary and immune status of the host.
cytokine levels in community-acquired pneumonia and their
Severity of CAP is largely related to these factors rather than association with disease severity, Eur. Respir. J. 20 (4) (2002)
to pathogen virulence per se. They also stated that if thrombo- 990–995.
cytopenia is present with bacterial CAP, clinicians should look [8] I. Martinez, F. Sanz, E. Fernandez, A. Cervera, M.L. Briones,
for an alternate explanation (e.g., drug induced), or it may Platelet count is a marker of outcome in community-acquired
represent a complication (e.g., hemolytic uremic syndrome). pneumonia, in: European Respiratory Society Annual Congress
In normal hosts, most CAP pathogens are not associated with 2013 Abstract Number: 3914.
either thrombocytopenia or thrombocytosis. Normal platelet [9] H. Georges, N. Brogly, D. Olive, O. Leroy, Thrombocytosis in
counts are the rule among typical CAP pathogens, as well as Patients with Severe Community-Acquired Pneumonia, Chest
138 (5) (2010) 1279.
among atypical CAP pathogens. In the absence of details on
504 A.A. ElMaraghy et al.

[10] E. Prina, M. Ferrer, O.T. Ranzani, E. Polverino, C. Cillóniz, E. time of hospitalization predict mortality in patients with
Moreno, J. Mensa, B. Montull, R. Menéndez, R. Cosentini, A. community-acquired pneumonia, Chest 137 (2) (2010) 416–420.
Torres, Thrombocytosis is a marker of poor outcome in community- [12] B.A. Cunha, J.E. Hage, N.Y. Mineola, Community-acquired
acquired pneumonia, Chest 143 (3) (2013 Mar) 767–775. pneumonia diagnostic vs. prognostic significance of the platelet
[11] M. Mirsaeidi, P. Peyrani, S. Aliberti, G. Filardo, J. Bordon, F. count, Chest 139 (5) (2011).
Blasi, J.A. Ramirez, Thrombocytopenia and thrombocytosis at