Malaria-Induced Splenic Infarction

Abstract

Splenic infarction is a rare complication of malaria. We report two recent cases of splenic infarction after Plasmodium vivax infection. No systematic review of malaria-induced splenic infarction was available, therefore we conducted a systematic review of the English, French, and Spanish literature in PubMed and KoreaMed for reports of malaria-associated splenic infarction from 1960 to 2012. Of the 40 cases collected on splenic infarction by Plasmodium species, 23 involved P. vivax, 11 Plasmodium falciparum, one Plasmodium ovale, and five a mixed infection of P. vivax and P. falciparum. Of the 40 cases, 2 (5.0%) involved splenectomy and 5 (12.5%) were accompanied by splenic rupture. The median time from symptom onset to diagnosis was 8.5 days (range, 3–90 days). Improved findings after treatment were observed in 8 (88.9%) of 9 patients with splenic infarction on follow-up by computed tomography or ultrasonography. All patients survived after treatment with the exception of one patient with cerebral malaria. Clinicians should consider the possibility of splenic infarction when malaria-infected patients have left upper quadrant pain.

Introduction

Malaria is a serious parasitic infection that affects both residents and travelers in tropical climates. Approximately 300 to 500 million malaria cases and 1.5 to 3.5 million deaths are reported annually.¹ However, according to a 2013 World Health Organization (WHO) report, 207 million cases of malaria and an estimated 627,000 deaths occurred in 2012 and the malaria mortality rate has fallen by 42% globally since 2000.² Malaria in humans is transmitted by the Anopheles mosquito vector and is caused by five species: Plasmodium falciparum (P. falciparum), Plasmodium vivax (P. vivax), Plasmodium ovale (P. ovale), Plasmodium malariae (P. malariae), and Plasmodium knowlesi (P. knowlesi).^1,3 Almost all severe forms of malaria are caused by P. falciparum, but serious complications such as severe anemia, respiratory distress, splenic complications, shock, and multiple organ dysfunction also develop with P. vivax infection.⁴ Splenic complications identified in cases of malaria are splenic infarction, spontaneous splenic rupture, hyperreactive malarial syndrome, hypersplenism, ectopic spleen and splenic torsion, and splenic cysts.⁵ Splenic infarction is not usually noted and is likely underdiagnosed in many cases of complicated malaria.⁶ Reports of malaria-associated splenic infarctions are rare, however, more recently, PubMed reports of cases have appeared almost annually. We recently encountered two cases of malaria with splenic infarction caused by P. vivax. However, systematic reviews of the clinical characteristics of splenic infarction resulting from malaria have not been conducted. Therefore, we performed a literature survey of splenic infarction in malaria in addition to summarizing these two cases. Our focus was on clinical outcomes, associated factors, and pathogenesis.

Case Reports

Case 1.

A 46-year-old man visited the emergency room because of a fever that had commenced 8 days previously. On admission, his body temperature was 40.6°C, blood pressure 110/70 mm Hg, heart rate 90/min, and respiratory rate 18/min. Physical examination revealed tenderness in the right upper quadrant of the abdomen, but the liver and spleen were not detectable. His complete blood count was white blood cell count 2.5 × 10³/μL, hemoglobin (Hb) 11.2 g/dL, and platelet count of 15 × 10³/μL. His chemistry profile was serum aspartate aminotransferase 76 U/L, alanine aminotransferase 24 U/L, total bilirubin 2.5 mg/dL, direct bilirubin 1.6 mg/dL, alkaline phosphatase 328 IU/L, and gamma-glutamyl transpeptidase level 218 IU/L. Features of abdominal computed tomography (CT) were consistent with splenic infarction, including multiple low attenuations in the spleen (Figure 1A). A Wright-Giemsa stain of a peripheral blood smear revealed a ring-form P. vivax trophozoite among the erythrocytes.

Figure 1.

Contrast-enhanced computed tomography of two patients, case 1 (A) and case 2 (B), showing low attenuation density in the spleen.

The patient was given oral chloroquine (25 mg/kg over 48 hours) on the day of admission and his fever subsided after 3 days. Right upper quadrant pain also decreased and additional pain in the left upper quadrant did not appear. At discharge, a 14-day regimen of oral primaquine (15 mg/day) was prescribed. One month later, the patient was stable and symptom free.

Case 2.

A 27-year-old man visited the emergency room because of fever and severe chills that had begun 5 days before admission. The patient was previously admitted to another hospital because of 4 days of fever and chills. However, the patient gradually worsened and he was referred to our hospital for further evaluation and to treat the fever of unknown origin.

Past medical history revealed that the patient had served military duty in a malaria-endemic area of Korea 3 years previously. During mandatory military service, he intermittently used chloroquine and had not experienced a malarial infection. When the patient arrived at our hospital he had a temperature of 39.0°C. The spleen was palpable in the abdomen and the patient complained of abdominal left upper quadrant tenderness. Initial laboratory results included white blood cell count 4.8 × 10³/μL, Hb 12.7 g/dL, platelet count 60 × 10³/μL, aspartate aminotransferase 47 IU/L, alanine aminotransferase 85 IU/L, and total bilirubin 0.9 mg/dL. Abdominal CT revealed splenomegaly of ∼15 cm and a low-attenuated wedge-shaped region that was consistent with splenic infarction (Figure 1B). A Wright-Giemsa stained blood smear performed within 24 hours of hospitalization revealed erythrocytes infected with P. vivax that ranged from ring-form to merozoites. Oral chloroquine was administered (25 mg/kg over 48 hours) in addition to primaquine (15 mg/day for 14 days). One month later the patient was stable and symptom free.

Methods

Literature search.

We examined relevant literature published between 1960 and 2012 that discussed malaria-associated splenic infarction using the keywords “spleen infarction, malaria” or “spleen rupture, malaria” in a PubMed Medline search. The results were limited to human studies published in English, French, and Spanish. Domestic cases that were not found in PubMed were searched in KoreaMed (http://koreamed.org/SearchBasic.php) using the keywords “spleen infarction, malaria,” “splenic infarction, malaria,” or “spleen rupture, malaria.”^7–9 Reports were in Korean or English. Two investigators independently reviewed the articles and discordance was resolved by consensus.

Selection criteria and data extraction.

Diagnostic tools used to diagnose malaria were peripheral blood smear, enzyme-linked immunosorbent assay for antibody detection, nested polymerase chain reaction using 18s ribosomal RNA, and immunochromatography for Plasmodium lactate dehydrogenase. Patients had a positive result for at least one of the tests. After a diagnosis of malaria was confirmed, cases were selected that used at least one of three methods for diagnosing splenic infarction: CT, ultrasonography (US), or gross findings or histopathology. Patients with hemorrhage from splenic rupture accompanied by splenic infarction in examinations were considered eligible. Data collected from each study were age, sex, symptoms (fever, left upper quadrant pain, and right upper quadrant pain), vital signs (hypotension, tachycardia, and respiratory distress), laboratory findings (anemia, thrombocytopenia, and parasitemia), disease state (acute or chronic), type of malaria, order of malarial infection (first or second), chemoprophylaxis (performed or not performed), splenic rupture (presence or absence), splenectomy (undergone or not undergone), and outcome (living or dead). This study was approved and performed according to the guidelines of the Institutional Review Board of Chonbuk National University Hospital (IRB no.: CUH 2014-01-026).

Definition.

Hypotension was defined as systolic pressure ≤ 80 mm Hg, tachycardia as ≥ 100 beats/min, respiratory distress as respiratory rate > 30 breaths/min, severe anemia as Hb < 7.0 g/dL, thrombocytopenia as < 1.5 × 10⁵/μL, and severe thrombocytopenia as < 6.0 × 10⁴/μL.^10,11

Results

Selected articles.

Figure 2 shows the process of identifying eligible papers. We discovered 126 references from PubMed (N = 119) and KoreaMed (N = 7) and screened the titles, abstracts, and the full text of these publications. Reasons for exclusion of cases were no case report form, report of only spleen rupture, no malarial cause for spleen infarction or spleen rupture, previous splenectomy and duplication. We included 23 articles and excluded 103.

Figure 2.

Identification and selection of eligible papers.

Clinical characteristics.

A total of 40 patients with malaria-associated splenic infarction were reported from 1960 through 2012. We confirmed the age and sex of 26 patients. The median age was 32 years (range, 3–65 years) with 17 men and 9 women. Left upper quadrant pain developed 2-fold more before antimalarial treatment than after or during antimalarial treatment. Laboratory findings and vital signs are in Table 1. The median time from onset of symptoms to diagnosis of splenic infarction was 8.5 days (range, 3–90 days). Of the 40 patients, 9 had follow-up US or CT and 8 (88.9%) of the 9 with splenic infarction improved. Follow-up periods after antimalarial treatment of the 8 patients that showed improvement were 1, 2, 3, 4, and 10 weeks for a single patient each, two patients that improved after 12 weeks, and one that was not checked during the follow-up period. However, one patient did not have improved findings on follow-up imaging after 1 week of treatment. The reviewed cases had limited details about the clinical situation related to splenic infarction.

Table 1.

Demographic and clinical features in patients with splenic infarction in malaria (N = 40)^*

Characteristics	Malaria type				No. (%)
	P. vivax (N = 23)	P. falciparum (N = 11)	Mixed infection in P. vivax and P. falciparum (N = 5)	P. ovale (N = 1)
Age, y, median (IQR)	38 (25.5–49.5)	28 (11–36)	28 (12.5–53.5)	NA	32 (21–41.25)†
Male sex	6	8	2	1	17 (65.4)‡
Fever	10/10	10/10	5/5	1/1	26/26 (100)
LUQ pain	10/10	8/8	5/5	1/1	24/24 (100)
Before anti-malarial drug	6/10	7/8	3/5		16/24 (66.7)
After or during anti-malarial drug	4/10	1/8	2/5	1/1	8/24 (33.3)
Hypotension	0/2	0/2	3/3	NA	3/7 (42.9)
Tachycardia	1/2	2/2	3/3	NA	6/7 (85.7)
Respiratory distress	1/2	0/1	0/2	0/1	1/6 (16.7)
Severe anemia	2/9	2/6	0/4	0/1	4/20 (20.0)
Thrombocytopenia	7/9	4/4	3/3	1/1	15/17 (88.2)
Severe thrombocytopenia	5/9	3/4	2/3	1/1	11/17 (64.7)
Time from the onset of symptom to diagnosis of splenic infarction, d, median (IQR)	10 (6–12)	9 (6–20)	7 (5.25–7.25)	NA	8.5 (6.25–10.75)§
Splenic infarction and splenomegaly improvement after anti-malarial treatment	7/7	0/1	1/1	NA	8/9 (88.9)

^* The denominator indicates the available number out of total 40 cases.

^† Available in 26 cases.

^‡ Available in 26 cases.

^§ Available in 24 cases.

IQR = interquartile range; LUQ = left upper quadrant; NA = not available.

Clinical outcomes.

Of the 40 cases, 23 (57.5%) were caused by P. vivax and 11 (27.5%) were caused by P. falciparum. One (2.5%) case was caused by P. ovale and five (12.5%) were caused by mixed infections of P. vivax and P. falciparum. All reported patients survived with the exception of a cerebral malaria patient. Only two (5.0%) patients received a splenectomy and five had cases that were accompanied by splenic rupture (12.5%) (Table 2). All patients were in the acute stage of malaria except for one patient who had concurrent visceral leishmaniasis. Twelve cases had available parasitemia results; of these, four had parasitemia > 10⁵/μL.

Table 2.

Splenic infarction in Plasmodium vivax, Plasmodium falciparum, Plasmodium ovale, and mixed infection with P. vivax and P. falciparum (N = 40)

Species	Reference (number of patients)	Parasitemia (/μL)	Splenomegaly	Splenic rupture	Splenectomy	Survival
P. vivax	7 (1)	1.25 × 105	Yes	No	No	Live
	8 (1)	2.44 × 103	Yes	No	No	Live
	9 (1)	NA	NA	No	No	Live
	13 (1)	1.88 × 103	Yes	No	No	Live
	48 (1)	NA	Yes	Yes	Yes	Live
	49 (2)	1.28 × 104	Yes	No	No	Live
		4.60 × 103
	50 (13)	NA	NA	No	No	Live
	51 (1)	NA	Yes	Yes	No	Live
	52 (1)	NA	Yes	No	No	Live
	53 (1)	NA	Yes	No	No	Live
P. falciparum	6 (1)	2.50 × 104	Yes	No	No	Live
	49 (1)	1.20 × 105	Yes	No	No	Live
	54 (2)	NA	Yes	No	No	Live
	55 (1)	NA	Yes	No	Yes	Live
	56 (1)	NA	Yes	No	No	Live
	57 (1)	1.50 × 104	NA	No	No	Live
	58 (1)	NA	Yes	No	No	Live
	59 (1)	NA	Yes	No	No	Live
	60 (1)	NA	Yes	Yes	No	Live
	61 (1)	1.25 × 107	Yes	No	No	Death
Mixed infection in P. vivax and P. falciparum	49 (1)	9.20 × 104*	Yes	No	No	Live
	52 (1)	NA	Yes	No	No	Live
	62 (1)	5.00 × 105†	Yes	No	No	Live
		3.50 × 107‡
	63 (1)	NA	Yes	Yes	No	Live
	64 (1)	NA	Yes	Yes	No	Live
P. ovale	12 (1)	5.00 × 10	Yes	No	No	Live

^* Adding up the parasitemia of P. vivax and P. falciparum.

^† P. vivax.

^‡ P. falciparum.

NA = not available.

Discussion

Frequency of splenic infarction in malaria.

Splenic infarction has been described as a rare complication of malaria, but the exact frequency of malaria-associated splenic infarction remains unclear because of underdiagnosis and underreporting.^12,13 Our study found that reports of splenic infarction were primarily from three countries, France, India, and Korea; this finding suggests the possibility of reporting bias by region.

Imaging can usually be performed to evaluate splenic complications in patients that complain of left upper quadrant and when splenic infarction is suspected. However, having clinicians routinely order US or CT images to identify splenic infarction is unreasonable. In addition, US or CT might not be conducted on patients with splenic infarction that is asymptomatic or accompanied by mild symptoms. Even if symptoms develop, splenic infarction might not be diagnosed because US or CT imaging is not available.

Splenic rupture resulting from malaria was reported more frequently with P. vivax than with P. falciparum infection.¹⁴ A recent study reported that malaria-associated splenic infarction is primarily caused by P. falciparum in contrast to splenic rupture.⁶ However, based on our results, P. vivax is as likely to cause splenic infarction as P. falciparum.

Malaria-associated splenic infarction prognosis.

Clinical aspects of splenic rupture can be severe and life-threatening.⁵ In one study, the case fatality rate of splenic rupture was 18.2%: among 55 patients, 12 died, 10 by splenic rupture and two from pneumonia that developed after splenectomy.¹⁴ Of these 55 patients, 33 (60%) underwent splenectomy,¹⁴ which was higher than the 5% of splenic infarction patients in our study. In addition, more than 50% of the patients in the previous study presented with hemodynamic collapse at the time of splenic rupture diagnosis, indicating the severity of this complication.¹⁴ No patient deaths from malaria-associated splenic infarctions have yet been reported. Almost all patients in our study received conservative management without splenectomy and all survived after treatment, suggesting that splenic infarction in malaria that is not accompanied by complications, such as rupture, can have a benign course and that splenic infarction is not an indication for operation.¹⁵

Most splenic infarctions were described during the acute state of malaria infection (Table 2). The median time from the beginning of symptoms to the diagnosis of splenic infarction was 8.5 days (range, 3–90; available in 24 cases), which was longer than the median time from symptoms to splenic rupture, which was 5 days (range, 0–37; available in 49 cases).¹⁴ This could be because diagnostic images were performed relatively late and patients' vital signs during splenic infarction were generally more stable than during splenic rupture. The key presentation and management features of splenic infarction and splenic rupture are presented in Table 3. Parasitemia values, the first event of malaria infection, and chemoprophylaxis protocols were not available in most cases.

Table 3.

Comparison of splenic infarction and splenic rupture in clinical presentation and outcome^*

	Splenic infarction (N = 35)†	Splenic rupture (N = 53)‡	Mixed splenic infarction and splenic rupture (N = 5)
Clinical findings
LUQ pain or diffuse abdominal tenderness	19/19 (100%)	23/52 (44%)	5/5 (100%)
Fever	21/21 (100%)	36/52 (69%)	5/5 (100%)
Splenomegaly	20/20 (100%)	20/52 (38%)	5/5 (100%)
Median time from fever onset to splenic complications (no. of days, range)	9 (5–90)§	5 (0–37)	5 (3–11)
Management
Only medical treatment	34 (97%)	13 (24%)	4 (75%)
Medical treatment with splenectomy	1 (3%)	3 (6%)	1 (25%)
Only splenectomy	0	29 (55%)	0
None	0	8 (15%)	0
Death	1 (3%)	12 (23%)	0

^* The denominator indicates the available number.

^† The number except for 5 cases⁴⁵,⁴⁸,⁵⁷,⁶⁰,⁶¹ with splenic infarction and splenic rupture from a total of 40 patients.

^‡ The number except for 2 cases⁴⁵,⁶⁰ with splenic infarction and splenic rupture from total 55 patients; see Reference ¹².

^§ Available in 19 cases.

LUQ = left lower quadrant.

Splenic infarction in malaria: pathogenesis.

Splenic infarction is a result of parenchymal ischemia from the occlusion of the arterial or venous circulation.¹⁵ However, the mechanism by which malaria results in splenic infarction is not clearly understood.⁶ For our investigation of the literature on malaria, we hypothesized that the following factors would contribute to a hypoxic state attributable to splenic infarction: 1) hypercoagulable state; 2) intrasplenic structural change by adhesion of malaria-infected red blood cells (iRBCs) to endothelial cells, with rosetting of iRBCs and non-iRBCs and splenic cellular hyperplasia; and 3) anemic hypoxia.

Hypercoagulable state.

During the acute stage of malaria, alteration of the coagulation system is observed with P. falciparum and P. vivax. Levels of antithrombin III, protein C, and protein S decrease remarkably in P. falciparum infection.^16,17 An increase in von Willebrand factor (vWF) and plasminogen activator inhibitor and a decrease in ADAMTS13 and tissue plasminogen activator were observed in P. vivax and P. falciparum infections.^16,18 The reductions in protein C, protein S, and antithrombin III levels strongly suggest the consumption of clotting factors caused by microvascular thrombosis, which is sufficient evidence for the generation of thrombi in malaria.¹⁶ Elevated vWF and reduced ADAMTS13 might be associated with intravascular platelet aggregation and microvascular disease.¹⁸ These factors are more likely to contribute toward ineffective fibrinolysis and the hypercoagulable state in acute malaria infection by P. falciparum than P. vivax.¹⁹

In postmortem studies on patients infected with P. falciparum, however, the formation of thrombin and fibrin, which is a final result of the hypercoagulable state, is not clearly observed in the spleen.^20,21 For patients who received antimalarial drugs, which induce a non-natural state during treatment, the rare finding of thrombin and fibrin formation might result from a return to normal condition from a hypercoagulable state after administration of antimalarial drugs.^17,22 In contrast with P. falciparum infection, altered thrombostasis and intravascular coagulation have not yet been characterized in P. vivax infection.^4,23 However, a study of acute kidney injury that developed with P. vivax infection found that the characteristics of renal biopsies were thrombotic microangiopathy, arterioles filled with an intraluminal fibrin thrombus or platelets, and endothelial injury.²⁴ Local splenic infarction caused by thrombus formation was also observed in splenic rupture caused by P. vivax.²⁵ Therefore, splenic microvascular obstruction by thrombin formation, fibrin deposition, and platelet aggregation, followed by reduction and congestion of local blood flow in the intrasplenic vascular system might partially contribute to local tissue hypoxia in the spleen. Further studies on splenic pathological findings in the absence of antimalarial treatment are needed to identify the role of coagulation disorders in malarial splenic infarction.

Vascular congestion and occlusion: cytoadhesion and rosetting of iRBCs and splenic cellular hyperplasia.

Elevation of cytokines such as tumor necrosis factor (TNF)-α and interferon (INF)-γ, in conjunction with the cytoadhesion of malaria-iRBCs to endothelial cells, can cause endothelial injury and activation, followed by expression of tissue factors and hypercoagulability.^22,26,27 Microvascular endothelial adhesion of P. falciparum-iRBCs appears to develop throughout the body, and distribution of iRBCs and extent of sequestration varies by organ system and is not even uniform within the same organ.^{20,21,28–32} Elevation of TNF-α and IFN-γ/IL-10 and increase in endothelial activation markers such as thrombomodulin, intracellular adhesion molecule (ICAM)-1, vascular adhesion molecule (VCAM)-1, and E-selectin are also observed in P. vivax infection.^33,34 Previously, P. vivax was not seen to become sequestered in the deep microvasculature of inner organs.³⁵ However, recent studies suggest the cytoadherence and sequestration of P. vivax-iRBCs.^36–39 Further research is required to identify the role of cytoadhesion and sequestration in P. vivax pathogenesis.

In postmortem studies, mature P. falciparum-iRBC sequestration is observed in red pulp, suggesting cytoadherence of iRBCs to the sinusoidal endothelium and mechanical retention.^20,21 Malarial parasites express various types of Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1) that have different adhesive properties and bind to alternative endothelial receptors.⁴⁰ This determines the organ in which iRBCs are sequestered and the severity of disease.⁴⁰ The mechanism for PfEMP1 and endothelial receptor ligand-receptor interaction has not yet been identified, particularly in the spleen.⁴⁰ If mature P. falciparum-iRBCs enter rapidly from a splenic artery into a perifollicular zone, venous sinus and splenic vein where iRBCs are sequestered and adhere to endothelial cells, autoagglutination between other iRBCs, rosetting of iRBCs with non-iRBCs, and platelet-mediated clumping can develop partially or extensively in the fast, closed circulation of the spleen, forming a lesion that obstructs blood flow.^21,41 Compared with the closed circulation system that mainly causes congestion of mature-iRBCs, both mature-iRBCs and ring-iRBCs appear to be retained in the splenic cord in slow, open circulation.^21,42 In particular, a large amount of ring-iRBCs accumulate upstream of interendothelial slits.⁴² Macrophage hyperplasia, which occupies about half of the cord volume, and an increasing number of erythrocytes, lymphocytes, plasma cells, polymorphonuclear cells, and mononuclear cells also build up in the splenic cord, contributing to blood congestion.^21,41 In P. falciparum malaria, impairment of intrasplenic blood flow by obstructive lesions, increased oxygen requirement and structural change by cellular hyperplasia in both circulation systems are more likely to be attributable to regional ischemia, which might result in splenic infarction.⁴³

In two patients with P. vivax infection who had splenic infarction and splenic rupture, several structures were identified including a hyperplastic splenic cord, multiple erythrocytes, macrophages, and leukocytes that filled ectatic or poorly outlined sinusoids and veins.^25,44 Thrombus formation in the vein or sinusoid was also observed.^25,44 Recently, in an untreated ruptured spleen that was infected with P. vivax, a diffuse hypercellularity in red pulp that presented with many P. vivax-iRBCs in the cords, massive proliferating plasma cells, and striking intra-sinusoidal histiocytosis were observed by immunohistopathological analysis.⁴⁵ These findings suggest that intrasplenic structural changes by excessive hyperplastic growth can cause blood flow obstruction, contributing to local splenic hypoxia in P. vivax infection. However, whether sequestration and cytoadhesion of P. vivax-iRBCs in sinusoids are important in pathogenesis remains unclear.^35,43

Anemic hypoxia.

Investigating anemia associated with malaria is beyond the scope of this work. The pathogenesis of malarial anemia is multifactorial. The mechanisms of anemia associated with P. vivax and P. falciparum infections vary and are not completely understood.⁴⁶ As the removal site of extravascular erythrocytes, the spleen is important in the destruction of iRBCs and non-iRBCs.⁴⁷ Therefore, malarial anemia likely contributes to tissue hypoxia in the spleen.

In summary, malaria-associated splenic infarction is reported as a complication in P. falciparum, P. vivax, and P. ovale infection, but has not yet been identified in P. malariae and P. knowlesi infections. Splenic infarction is primarily detected in the acute phase of infection. Clinicians should consider left upper quadrant pain a predictive sign of splenic infarction and perform imaging such as US or CT if available. In general, the clinical response to medical treatment will likely be sufficient and a splenectomy might not be necessary. However, the pathogenesis of splenic infarction has not been sufficiently elucidated. Based on previous studies, hypercoagulability, intrasplenic structural change directed toward vascular congestion, obstruction caused by hypercellularity and sequestration, and anemia progression likely contribute to the complexity of splenic hypoxia that is followed by splenic infarction in vulnerable areas. Further studies are needed to better understand the pathogenesis of splenic infarction.

Disclaimer: Transparency declarations, none to declare.