| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
150,000/µL. This study supplements previous literature on the hematologic effects of malaria and helps define those alterations for a semi-immune population. Thrombocytopenia is identified as a key indicator of malaria in these febrile patients. |
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Hematologic changes associated with malaria infection are well recognized, but specific changes may vary with level of malaria endemicity, background hemoglobinopathy, nutritional status, demographic factors, and malaria immunity.1 This study examines the hematologic effects of acute malaria on adults in Tak Province. Specifically, the hematologic profiles of persons infected with P. falciparum or P. vivax are compared with expected normal values, as well as with the profiles of otherwise similar febrile persons without microscopically detectable parasitemia. Additionally, the clinical symptoms and hematologic parameters most predictive of malaria in this population are identified.
|
SUBJECTS AND METHODS |
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
20 years; and at least one of the following: an oral temperature 38°C, headache, or a history of fever within the past 72 hours. Severely ill patients were referred to district hospitals; therefore, complicated malaria cases were excluded in practice. Each patient was assigned a unique patient identifier. All enrolled patients were interviewed in their own vernacular language, Thai, Burmese, or Karen, for information on current symptoms and previous malaria episodes and treatments. Venous blood was drawn into EDTA-filled tubes to be used for preparation of blood film slides and automated complete blood counts (CBCs). Two slides for study purposes were promptly prepared at the field site, each with a thin and thick smear. A third smear was prepared and provided immediately to the local clinic staff to be stained and interpreted according to routine clinic procedures for therapeutic purpose. Tubes were transported on ice within two hours to the field laboratory, where cell counts were performed using a Coulter T-890 automated cell counters (Beckman-Coulter, Inc, Fullerton, CA). Daily quality assurance checks were performed and recorded, and commercial standards were used in accordance with the manufacturer’s recommendations. The cell counters provided data on the white blood cell count (WBC), red blood cell count (RBC), hemoglobin level (Hb), hematocrit, mean corpuscular volume, mean corpuscular hemoglobin, mean corpuscular hemoglobin concentration, platelet count, lymphocyte count, and lymphocyte percent.
Blood smear slides were stained with Giemsa. One slide from each study participant was examined independently by two experienced microscopists. They determined the presence or absence of Plasmodia parasites, the species, and the number of asexual parasites per 200 WBCs. If fewer than 10 asexual parasites per 200 WBCs were observed, counting continued to 500 WBCs. Parasite densities were calculated as parasites per microliter of blood (parasites/WBCs counted x total WBCs in 1 µL of blood), and the results of the two readings were averaged. Concordance between the two blinded microscopists’ interpretations was assessed and cases of species discordance from the two readings, mostly associated with very low parasitemia, were excluded from this analysis.
Data were analyzed using SAS version 8.1 (SAS Institute, Cary, NC). Patients were categorized based on microscopy results: no malaria (control group), P. falciparum detected, P. falciparum at densities greater than 1,000 parasites/µL, P. vivax detected, P. vivax at densities greater than 1,000 parasites/µL, mixed P. falciparum/P. vivax infection, or P. malariae detected. Mean hematologic parameters were calculated by sex for the control group (who also presented with fever, headache, or a history of fever). Hematologic parameters were compared between the control group and the specific malaria groups using independent-sample t-tests. Parasite densities for single-species infections were divided into quartiles, and their associations with the hematologic parameters were detected using analysis of variance.
To test for an association between malaria and symptoms or specific hematologic values, categorical variables were created for platelets, RBCs, and WBCs. Bivariate associations were examined between these variables or clinical symptoms and various outcomes: any malaria, P. falciparum, P. vivax, mixed infection, P. malariae, P. falciparum > 1,000 parasites/ µL, P. vivax > 1,000 parasites/µL, or the lower densities for either species. Variables significant in any of these analyses were fit into logistic regression models to identify factors that were collectively predictive of malaria. Each malaria outcome was modeled separately, and stepwise selection was used to identify the variables with the most predictive value for each outcome.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The mean values of selected hematologic parameters for the negative (control) group were determined for this population, stratified by sex (Table 1
). T-tests showed that RBCs, Hb, platelet count, and lymphocyte count were significantly different between men and women without malaria, so subsequent analysis of continuous variables was stratified by sex. For comparison, Table 1 also includes published normal values, although derived from a population from the United States and with relatively wide ranges.3 All the values from this study population fall within the acceptable normal limits.
|
. For P. falciparum and P. vivax, the mean values of WBCs, RBCs, platelet counts, and lymphocyte counts were significantly lower for the infected groups than for the uninfected groups among both men and women. The exclusion of persons with parasite densities less than 1,000 parasites/µL did not yield substantially different results. Although sample sizes for P. malariae and mixed infections were small, similar trends were observed in these groups.
|
and Figure 1).
|
|
. A WBC < 5,000/µL, a RBC < 4,000,000/µL, a platelet count < 150,000/µL, an oral temperature 38.5°C, a history of fever in the past 72 hours, malaise, and chills were all significantly more likely to occur in persons with any malaria, P. falciparum, or P. vivax, than in persons with a negative blood film. Low WBC, RBC, and platelet counts were also significant for persons with P. malariae, despite the small sample size for this group. Vomiting was a significant predictor in persons with any malaria or P. falciparum. Myalgias and lack of cough were significant for any malaria and P. vivax. Analysis of persons with high levels of parasitemia produced very similar results to any parasitemia; the lower parasitemia categories showed similar tendencies but with reduced strength and significance.
|
. The three hematologic parameters were significant predictors for the presence of malaria, with the exception of RBCs in P. vivax infections. Platelet count showed the strongest association and had the greatest predictive power in each of the models; persons with platelet counts < 150,000/µL were 12.5–14.7 times more likely to have any of these malaria outcomes than persons with platelet counts 150,000/µL. Fever and being ill for less than three days were also indicators in all the final models. Vomiting was a significant predictor in the P. falciparum models; malaise and lack of cough were predictors in the P. vivax models. A history of fever in the past 72 hours was a significant predictor of any malaria model only. For P. malariae infections, only low platelet and RBC counts were significant when controlling for the other factors, and only low platelet count and fever were significant for mixed infections.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Low platelet counts have been consistently found for both P. falciparum4–6,8,9 and P. vivax.4,7,10–13 A substantial proportion of cases in this study had very low platelet counts, especially among those with P. falciparum infection, 39% of whom had platelet counts less than 100,000/µL. Plasmodium falciparum cases with platelet levels below this count have also been documented in other populations of semi-immune4,5 and nonimmune patients.6
The trend of decreasing platelet count with increasing levels of parasitemia observed in this study has been previously noted for P. falciparum,4,6 but to our knowledge, has not been documented in the literature for P. vivax. Decreased platelet production has been ruled out.5,11,14 Thrombocytopenia is a result of peripheral destruction and consumption. Immune complexes generated by malarial antigen lead to sequestration of the injured platelets by macrophages in the spleen.15–18 Platelet consumption in disseminated intravascular coagulation contributes to thrombocytopenia in complicated P. falciparum malaria; however, this process is not relevant to the present study population. Platelet dysfunction resulting in hyperaggregation is another alteration occurring in association with malaria resulting in hyperaggregation.19 During malaria infection, there are several factors that activate platelets, among which are formation of immune complexes and damage to endothelial cells. Surface contact of platelet membrane to with parasitized RBCs is another stimulator.20 Intravascular lysis of the activated platelets may also occur.9 Trends between increasing parasite density with a decrease in the level of hematologic parameters other than platelet count were not observed in this study, nor have they been consistently noted in the literature.
Anemia has frequently been associated with malaria. In regions of sub-Saharan Africa with stable, high malaria transmission, severe anemia is common among children or pregnant women infected with P. falciparum.21 Studies among nonimmune or semi-immune populations outside Africa have also found statistically significant levels of mild anemia in falciparum malaria patients.4,5,22,23 Plasmodium vivax has been associated with mild anemia in some studies,7,10,12,24,25 but not in others.4 Two possible causes of this anemia are increased hemolysis or a decreased rate of erythrocyte production.26,27 Despite the extensive documentation of anemia in malaria, only mild decreases in Hb were observed in this study. This discrepancy may be related to the multifactorial etiology of anemia. The impact of malaria anemia is greatest in regions of sub-Saharan Africa where underlying anemia and poor nutrition are common. Although a few cases of very low Hb were observed among our study participants, only 16% of all participants had even mild anemia (Hb level < 11.0 g/dL). Furthermore, some observers have suggested that malaria-related anemia is more severe in areas of intense malaria transmission and in younger children rather than in older children or adults.21,27 While this study and others in southeastern or eastern Asia have noted Hb decreases or mild anemia among malaria cases,4,7,10,23 the small degree of Hb change observed in this study population may reflect a lower prevalence of underlying anemia, better nutritional status, and/or better access to treatment.
Fever and history of fever are fairly sensitive measures of malaria, but they lack specificity or positive predictive value, especially in regions where malaria is less prevalent. The identification of signs or constellations of signs indicative of malaria could also help to improve appropriate treatment in areas where most people are parasitemic, but malaria diagnosis is still based solely on fever in the presence of any parasitemia.28
Studies from various areas have failed to identify factors that can substantially improve upon current measures. The lack of specificity of malaria-associated signs and symptoms has led some researchers to conclude that only fever and history of fever have any real diagnostic value.29,30 A history of fever in the absence of another major symptom was the best predictor of malaria-positive blood slides in adults in Papua New Guinea,31 and a study in Zimbabwe concluded that only microscopy could improve diagnostic capabilities beyond those "unknown decision rules" used by health care workers.29 Combinations of factors have proved slightly better at identifying malaria. In Ethiopia, fever, previous malaria, and pallor was the measure with the best combined sensitivity and specificity.32 The identification of additional criteria helpful in diagnosing malaria would be of use in this field.
The symptoms found in this study to be independently associated with the presence of malaria parasites have been well recognized: oral temperature 38.5°C, history of fever in the past 72 hours, malaise, chills, and (for P. falciparum) vomiting. Associations between low WBC, RBC, or platelet counts and malaria infection are not as widely acknowledged, in part because they are not routinely obtained at malaria clinics. These parameters, however, appear to have the strongest independent measures of association besides fever; crude odds ratios as high as 23 are observed for low platelet count and the presence of P. falciparum. Even when controlling for other clinical factors, platelet count still has a higher predictive value for malaria infection than any symptoms. A person from this study population infected with P. falciparum is almost 15 times more likely to have a low platelet count (< 150,000/µL) than a malaria-negative one, adjusting for fever, vomiting, days ill, RBCs, and WBCs.
Platelet count may be of limited diagnostic applicability in settings where health care workers must operate without even the benefits of a microscope. However, in Thailand, capability of a routine CBC is available at all district-level hospitals. Interestingly, malaria diagnosis at those hospitals is sometimes less accurate than at malaria clinics because of reliance on only thin blood films or the limited experience of hospital microscopists in detecting malaria parasites or in reading a thick smear compared with specially-trained malaria clinic microscopists. Also, patients presenting to district hospitals often have prior self-treatment with ineffective or subtherapeutic doses of antimalarials. Such treatment may suppress the parasites or distort their morphology, but not fully eliminate them. Therefore, in this malaria-endemic area, an acutely febrile patient with low platelet count and a reduced WBC count, irrespective of a malaria smear report, should always be thoroughly re-evaluated for malaria. A recent study also identified a prognostic value for thrombocytopenia in African children, with a lower platelet count associated with more severe illness or outcome.33 While prognosis could not be evaluated in this study, platelets might play a similar role in this population.
This study used very rigorous microscopy procedures, which should minimize misclassification of malaria status. Additionally, the large sample size and high proportion of malaria cases give this analysis substantial power to detect differences between cases and controls. Daily quality control checks of the automated cell counters maintained the accuracy and precision of the hematologic measurements.
One limitation of this study is that the controls are symptomatic, febrile individuals, and therefore do not represent a healthy population. They are likely to have had a diverse collection of bacterial and viral ailments that could affect the measured hematologic variables in different ways. Nonetheless, the mean hematologic parameters for the control group fell within the standard normal ranges measured in other populations. Another caution in interpreting our data is that the statistically significant differences found between malaria cases and controls in many of the measured parameters do not necessarily indicate clinical significance.
Although these basic hematologic changes in association with malaria are not new to the subject, our data add more detailed information to the limited body of knowledge. The observations of thrombocytopenia and lymphopenia are in accordance with those reported in nonimmune individuals,6 and are unlike observations reported from sub-Saharan Africa. This study implies that malaria must always be a key differential diagnosis in acutely febrile patients with thrombocytopenia and leukopenia from this endemic area. This is the first documentation that the parallel trend in thrombocytopenia with parasitemia is not only unique for infection with P. falciparum, but also with P. vivax. Greater exploration of the strong, inverse relationship between platelet levels and malaria infection may afford means to improve diagnosis and alleviate the clinical severity of or accelerate recovery from this disease.
Received July 11, 2003. Accepted for publication September 20, 2003.
Acknowledgments: This work was presented at the 51st Annual Meeting of the American Society of Tropical Medicine and Hygiene (Denver, CO) on November 12, 2002 by Laura M. Erhart (Abstract No. 242, Scientific Session J, Malaria Epidemiology and Diagnosis). We thank Dr. J. Sirichaisinthop for his support, and to the Malaria Field Team of the Department of Immunology of the Armed Forces Research Institute of Medical Sciences for technical assistance.
Financial support: This study was supported by U.S. Army Medical Materiel Development Activity (Fort Detrick, Frederick, MD).
Disclaimer: The opinions or assertions contained herein are those of the authors and should not be construed as reflecting the official positions of the U.S. Army, U.S. Air Force, or U.S. Department of Defense.
Authors’ addresses: Laura M. Erhart, Department of Immunology and Medicine, Armed Forces Research Institute of Medical Sciences, 315/6 Rajvithi Road, Bangkok 10400, Thailand, and University of Michigan School of Public Health, Ann Arbor, MI 48109-2029. Kritsanai Yingyuen, Niphon Chuanak, Nilawan Buathong, Anintita Laoboonchai, and R. Scott Miller, Department of Immunology and Medicine, Armed Forces Research Institute of Medical Sciences, 315/6 Rajvithi Road, Bangkok 10400, Thailand. Chansuda Wongsrichanalai, U.S. Naval Medical Research Unit No. 2, Kompleks P2M/PLP-LITBANGKES, Jalan Percetakan Negara No. 29, Jakarta 10570, Indonesia, Telephone: 62-21-421-4457 extension 1121, Fax: 62-21-420-7854. E-mail: chansuda{at}namru2.med.navy.mil. Steven R. Meshnick, University of North Carolina School of Public Health, Chapel Hill, NC 27599-7435. Robert A. Gasser Jr., Walter Reed Army Institute of Research, 503 Robert Grant Avenue, Silver Spring, MD 20910.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. W. Lee, K. Jeon, B. R. Jeon, and I. Park Rapid Diagnosis of Vivax Malaria by the SD Bioline Malaria Antigen Test When Thrombocytopenia Is Present J. Clin. Microbiol., March 1, 2008; 46(3): 939 - 942. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Obaldia III Clinico-Pathological Observations on the Pathogenesis of Severe Thrombocytopenia and Anemia Induced by Plasmodium vivax Infections During Antimalarial Drug Efficacy Trials in Aotus Monkeys Am J Trop Med Hyg, July 1, 2007; 77(1): 3 - 13. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. J. Evans, D. S. Hansen, N. van Rooijen, L. A. Buckingham, and L. Schofield Severe malarial anemia of low parasite burden in rodent models results from accelerated clearance of uninfected erythrocytes Blood, February 1, 2006; 107(3): 1192 - 1199. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
