Abbreviations
APTT, Activated partial thromboplastin time
CRP, C-reactive protein
EKG, Electrocardiogram
ES, Eisenmenger syndrome
IL, Interleukin
PA, Pulmonary artery
PE, Pulmonary embolism
PH, Pulmonary hypertension
PT, Prothrombin time
INTRODUCTION
Pulmonary embolism (PE) is often termed the "silent killer", because of its fatal yet non-specific clinical manifestations. Patients with PE may present with dyspnea, chest pain, cough, and hemoptysis. However, in patients with cardiovascular disease or pulmonary hypertension (PH), PE becomes more difficult to identify due to overlapping symptoms. Moreover, patients with Eisenmenger syndrome (ES) have a high prevalence of pulmonary artery (PA) thrombosis.1 Although fever is a non-specific and under-recognized symptom of PE, it is often overlooked during differential diagnosis. Notably, previous studies have reported that fever occurs in approximately 25% of acute PE cases, particularly among younger patients.2 Therefore, PE should be included in the differential diagnosis of unexplained fever, especially in patients with low-grade fever when the infectious disease work-up results are negative. Here, we present a case of ES with intermittent fever that was eventually diagnosed as PE.
CASE
A 32-year-old woman with ES secondary to patent ductus arteriosus was admitted due to worsening dyspnea on exertion. She had been regularly followed up at a cardiology outpatient clinic and was receiving combination therapy with sildenafil and bosentan for pulmonary arterial hypertension.3,4
On admission, her vital signs showed body temperature of 37 °C, pulse rate of 96 beats/min, respiratory rate of 16 breaths/min, and blood pressure of 82/64 mmHg. On room air, oxygen saturation measured by pulse oximetry was 85% in the fingers and 65% in the toes. Physical examination revealed left parasternal heave, an accentuated S2 with a grade II/VI systolic murmur at the left lower sternal border, and a grade III/VI diastolic murmur at the left upper sternal border. The lower legs showed mild pitting edema. Cyanosis and clubbing were more prominent in the toes than in the fingers (Figure 1A). Laboratory testing showed a mildly elevated C-reactive protein (CRP) level of 27.5 mg/L (normal: < 5 mg/L) without leukocytosis. D-dimer was 0.41 μg/L (normal: < 0.5 μg/L). Hemoglobin was 15.6 g/dL (normal: 12.0-16.0 g/dL). Platelet count was 240 K/μL (normal: 150-400 K/μL). Prothrombin time (PT) was 15.3 sec. (normal: 10-13 s), with an international normalized ratio (INR) of 1.4, and activated partial thromboplastin time (APTT) was 28.5 sec (normal range: 24-31 s). Blood and urine bacterial cultures showed no growth. Chest radiography revealed bilateral pulmonary arterial dilatation and increased cardiac shadow (Figure 2A). Electrocardiogram (EKG) showed right axis deviation and right ventricular hypertrophy. A minor ST-T changes with ST-segments depression and smaller T-waves at precordial leads compared with previous EKG (Figure 2B).
Figure 1.
(A) Differential clubbing and cyanosis of the digitals, with more cyanosis and clubbing in the toes compared with the fingers. (B) Lung perfusion scan using Tc-99m MAA showing multiple non-wedge-shaped perfusion abnormalities in both lungs (left). A significant right-to-left shunt is demonstrated with the ratio of systemic (shunt) tracer activity versus lung tracer activity estimated as 47% versus 53% (right). (C) Chest computed tomography showing a hypodense filling defect in right pulmonary artery. (D) Dilatation of the pulmonary arteries: MPA 51 mm, RPA 36 mm, and LPA 25 mm; RPA/LPA ratio 1.44 compare to no thrombosis 1.06.9 Ao, aorta; E, pericardial effusion; LPA, left pulmonary artery; MPA, main pulmonary artery, RPA, right pulmonary artery; T, thrombus; Tc-99m MAA, Technetium-99m macro-aggregated albumin.
Figure 2.

(A) Chest radiography showing increased cardiac shadow on admission (bottom) compared with 6-month ago (top). (B) EKG showing a minor ST-T changes with ST-segments depression and smaller T-waves at precordial leads V2-6 on admission (bottom) compared with 6-month ago (top). (C) Clinical course and body temperature of the patient. BNP, B-type natriuretic peptide; CRP, c-reactive protein; EKG, electrocardiogram; WBC, white blood cell.
During hospitalization, the patient exhibited intermittent fever accompanied by dry cough, dyspnea, and chest pain. The fever showed a diurnal pattern, rising in the evening to a maximum of 38.3 °C at around 20:00 and gradually subsiding without antipyretics (Figure 2C). Empirical antibiotic therapy with piperacillin and tazobactam was administered, and furosemide was given for right heart failure and leg edema; however, the fever persisted. Because moderate pericardial effusion was detected on echocardiography, and an inflammatory scan revealed trivial pericardial 18F-fluorodeoxyglucose uptake, the patient was treated for acute pericarditis with colchicine (0.5 mg once daily) and ibuprofen (400 mg four times daily). She was discharged after the fever subsided following treatment.
However, the fever recurred one week later, accompanied by worsening heart failure, and she was readmitted. With suspicion of PE, a lung perfusion scan was performed, showing multiple non-wedge-shaped heterogeneous perfusion abnormalities in both lungs and a significant right-to-left shunt with a shunt fraction of 53% (Figure 1B). Due to the progressive decompensation of heart failure, the medical team considered lung transplantation.
During transplantation work-up, computed tomography revealed a filling defect in the proximal right PA, consistent with PE (Figure 1C). Peripheral venous ultrasound showed no evidence of deep vein thrombosis in the lower limbs, indicating PA thrombosis in situ. Subcutaneous enoxaparin 45 mg every 12 h was initiated for PE treatment, and the fever subsided over the following days without recurrence (Figure 2C). Follow-up echocardiography revealed that the estimated systolic PA pressure had decreased from 136 to 118 mmHg, with resolution of the pericardial effusion. The patient was discharged with apixaban 5 mg twice daily.
DISCUSSION
Fever in acute PE is well documented.1 Previous studies have reported that 14-25% of the patients with PE present with fever, often low grade, with no significant difference between patients with or without pulmonary hemorrhage or infarction.2,5 When thrombus occludes pulmonary vasculature, ischemia and tissue injury occurs in lung parenchyma leading to sterile inflammation, which triggers the release of cytokines such as Interleukin (IL)-1β, IL-6, and tumor necrosis factor-α. These cytokines stimulate the production of prostaglandin E2 in the hypothalamus, which increases the body temperature setpoint and ultimately results in fever.6,7 IL-6 stimulates the liver to produce CRP during the acute-phase response to inflammation, which is compatible with the elevated CRP in this case.
A diurnal fever pattern, with body temperature rising in the evening (approximately 19:00), peaking at 20:00, and returning to normal in the morning, was observed in the patient. Although the mechanism of the diurnal fever pattern is not fully understood, circadian regulation of cytokines has been reported. IL-6 levels are lowest at 08:05 and highest at 20:05; IL-6 is particularly important during chronic or low-grade inflammation.8
PA thrombosis in ES is not uncommon, with a reported prevalence of 21-29%.1,9 The formation of thrombi in situ is multifactorial. Chronic hypoxemia and progressive PH cause endothelial injury in the pulmonary arteries and compensatory changes in coagulation. Although platelet function abnormalities have been reported in patients with ES, no significant differences in platelet function or coagulation factors were found between patients with or without thrombus formation.9-11 Instead, a mildly prolonged PT INR with normal APTT and platelet count was observed in this case. Elevated pressure may traumatize the pulmonary vessels causing endothelial damage, while small breaks in the intimal layer of the PA may induce blood clot formation.9
PA dilatation with significant variations in vessel diameter may induce turbulent blood flow and facilitate blood cell aggregation, representing another possible mechanism of thrombus formation. PA thrombosis was found in patients with larger PA diameters and higher right-to-left PA diameter ratios. This is in line with this case has dilated PA and the ratio of right to left PA is 1.44 (mean ratio 1.06 ± 0.18 in no thrombosis) (Figure 1D).9
In this case, the minor ST-T changes on EKG may provide a clue of increasing right ventricular strain during PE (Figure 2B). However, the lung perfusion scan may not have helped in the diagnosis of PE because of severe PH. Additionally, right-to-left shunts may have compromised the sensitivity of the test in the diagnosis of PE because the radiotracer injected for perfusion imaging bypasses the pulmonary capillary and enters the systemic circulation, resulting in insufficient tracer uptake in the lungs.12 Therefore, both the physiology of ES and the technical limitations of the imaging must be considered when evaluating PE in such patients.
In conclusion, we report a case of PE in a patient with ES secondary to patent ductus arteriosus. Patients with PE may present with intermittent fever and exacerbate preexisting symptoms. ES increases the risk of PA thrombosis, and PE should be considered in the differential diagnosis of patients with "unexplained" fever.
LEARNING POINTS
1. Unexplained fever as a cause of pulmonary thrombosis: In patients with ES, intermittent fever may be the primary or sole manifestation of pulmonary arterial thrombosis. Clinicians should maintain a high index of suspicion for PE when infectious workups are negative, even if typical respiratory symptoms are vague.
2. Diagnostic limitations of lung perfusion scans in ES with PE: In the presence of severe PH and a significant right-to-left shunt, the radiotracer bypasses the pulmonary capillary bed and enters the systemic circulation, causing inconclusive results of the scan in PE.
3. Mechanisms of pulmonary thrombosis formation in ES: Pulmonary thrombosis in ES often forms in situ. The leading causes include endothelial injury due to chronic pulmonary hypertension, hypoxemia, and turbulent blood flow within dilated pulmonary arteries.
DECLARATION OF CONFLICT OF INTERESTS
All the authors declare no conflict of interest.
INFORMED CONSENT
Written informed consent was obtained from the patient before their participation in the study.
FUNDING
The authors declare no financial support for this study.
REFERENCES
- 1.Silversides CK, Granton JT, Konen E, et al. Pulmonary thrombosis in adults with Eisenmenger syndrome. J Am Coll Cardiol. 2003;42:1982–1987. doi: 10.1016/j.jacc.2003.07.022. [DOI] [PubMed] [Google Scholar]
- 2.Saad M, Shaikh DH, Mantri N, et al. Fever is associated with higher morbidity and clot burden in patients with acute pulmonary embolism. BMJ Open Respir Res. 2018;5:e000327. doi: 10.1136/bmjresp-2018-000327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Chen IC, Dai ZK. Insight into pulmonary arterial hypertension associated with congenital heart disease (PAH-CHD): classification and pharmacological management from a pediatric cardiological point of view. Acta Cardiol Sin. 2015;31:507–515. doi: 10.6515/ACS20150424B. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tsai TH, Ho WJ, Wang CL, et al. Longitudinal assessment of TAPSE/PASP ratio and its correlation with clinical outcomes in pulmonary arterial hypertension. Acta Cardiol Sin. 2026;42:22–32. doi: 10.6515/ACS.202601_42(1).20250818F. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Stein PD, Afzal A, Henry JW, Villareal CG. Fever in acute pulmonary embolism. Chest. 2000;117:39–42. doi: 10.1378/chest.117.1.39. [DOI] [PubMed] [Google Scholar]
- 6.Bontekoe E, Brailovsky Y, Hoppensteadt D, et al. Upregulation of inflammatory cytokines in pulmonary embolism using biochip-array profiling. Clin Appl Thromb Hemost. 2021;27:10760296211013107. doi: 10.1177/10760296211013107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Santacroce L, Colella M, Charitos IA, et al. Microbial and host metabolites at the backstage of fever: current knowledge about the co-ordinate action of receptors and molecules underlying pathophysiology and clinical implications. Metabolites. 2023;13:461. doi: 10.3390/metabo13030461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Nilsonne G, Lekander M, Åkerstedt T, et al. Diurnal variation of circulating interleukin-6 in humans: a meta-analysis. PLoS One. 2016;11:e0165799. doi: 10.1371/journal.pone.0165799. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Broberg CS, Ujita M, Prasad S, et al. Pulmonary arterial thrombosis in Eisenmenger syndrome is associated with biventricular dysfunction and decreased pulmonary flow velocity. J Am Coll Cardiol. 2007;50:634–642. doi: 10.1016/j.jacc.2007.04.056. [DOI] [PubMed] [Google Scholar]
- 10.Soares R, Maeda NY, Bydlowski SP, Lopes AA. Markers of endothelial dysfunction and severity of hypoxaemia in the Eisenmenger syndrome. Cardiol Young. 2005;15:504–513. doi: 10.1017/S1047951105001381. [DOI] [PubMed] [Google Scholar]
- 11.Caramuru LH, Lopes AA, Maeda NY, et al. Long-term behavior of endothelial and coagulation markers in Eisenmenger syndrome. Clin Appl Thromb Hemost. 2006;12:175–183. doi: 10.1177/107602960601200205. [DOI] [PubMed] [Google Scholar]
- 12.Gühne F, Drescher R, Ndum F, et al. Right-to-left shunt in lung perfusion scintigraphy. Clin Nucl Med. 2021;46:e162–e164. doi: 10.1097/RLU.0000000000003286. [DOI] [PubMed] [Google Scholar]

