Neoehrlichia mikurensis-associated secondary capillary leak syndrome and septic shock: a case report

Abstract

Background

Candidatus Neoehrlichians mikurensis is the agent of an emerging tick-borne zoonosis responsible for systemic inflammatory syndrome mainly in immunocompromised patients and in subjects with underlying hematological or immunological disorders. The reported clinical manifestations include fever, chills, night sweats, weight loss, skin rash, myalgia/arthralgia, and edema limited to the ankles, and the infection is associated with a high incidence of vascular events. We report an unusual case of neoeherlichiosis, diagnosed by means of broad-range eubacterial polymerase chain reaction assay from a sample of spontaneous psoas muscle hematoma, which presented with septic shock and secondary capillary leak syndrome.

Case presentation

A 58-year-old Caucasian Swiss woman undergoing treatment of rituximab presented after 8 weeks of fever, chills, night sweats, and diffuse edema resistant to recently introduced loop diuretics. The patient developed septic shock with anasarca, possibly due to secondary capillary leak syndrome. Extensive microbiological investigations identified Candidatus Neoehrlichia mikurensis from a sample of spontaneous psoas muscle hematoma using eubacterial polymerase chain reaction assay and in subsequent blood samples. The patient fully recovered after 3 weeks of antibiotic therapy with doxycycline.

Conclusion

This case broadens the known clinical spectrum of the disease and may suggest a pathogenic role in hematoma formation, likely related to endothelial tropism and intravascular dissemination of the bacteria. Recognizing such atypical presentations is key to avoiding diagnostic delays. This case also highlights the diagnostic value of eubacterial polymerase chain reaction in emerging tick-borne infections.

Introduction

Neoehrlichia mikurensis is a strict Gram-negative intracellular coccoid bacterium within the Anaplasmataceae family, transmitted by Ixodes ticks, with rodents as the primary reservoir [1, 2]. Previously referred to as Candidatus Neoehrlichia mikurensis due to its uncultivable state, it was successfully cultured in 2019, prompting a taxonomic revision [3]. This pathogen causes neoehrlichiosis, a rare tick-borne disease that predominantly affects immunocompromised patients, although some cases were reported in immunocompetent hosts [4–9]. Clinical features are often nonspecific, including fever, weight loss skin rash, myalgia/arthralgia, and edema limited to ankles, and infection is associated with high incidence of thromboembolic and vascular events [3, 4, 10]. Laboratory findings are also nonspecific, thus resulting in delayed diagnosis (median number of 60 days from onset to diagnosis) [10–12]. Diagnosis requires polymerase chain reaction (PCR) detection of bacterial DNA. Treatment with doxycycline is effective [2, 7, 12]. Since 2009, more than 38 cases have been reported in Europe, with 6 in Switzerland [2, 4, 10–15]. This report describes a unique presentation of neoehrlichiosis with septic shock, probable secondary capillary leak syndrome, and spontaneous psoas hematomas, features not previously associated with the infection. The diagnosis was made using eubacterial PCR in a sample of psoas muscle hematoma. Widespread anasarca due to probable secondary capillary leak syndrome, a rare but severe complication of systemic inflammation, has not been reported in relation to N. mikurensis infection until now [16–19].

Case description

In March 2022, a 58-year-old Caucasian Swiss woman with seropositive rheumatoid arthritis treated with rituximab since 2016 was admitted to the Yverdon hospital (Switzerland) after 8 weeks of fever, chills, night sweats, and diffuse edema resistant to recently introduced loop diuretics. She reported a 17 kg weight gain in the last 2 months (usual weight: 75 kg, weight at admission: 92.6 kg). The patient had not noticed any insect bites or skin rash. Physical examination revealed a patient with diffuse pitting anasarca of the entire body (face included) and a temperature of 39.5 °C. Blood pressure was 117/87 mmHg and heart rate was elevated at 120 beats per minute. Oxygen saturation was measured above 95% without supplemental oxygen. Cardiac examination was unremarkable except for the presence of anasarca. No skin or joint abnormalities were found. Laboratory tests showed leukocytosis [11.1 G/l; reference range (RR) 3.8–10 G/l], hyporegenerative normocytic normochromic anemia (70 g/l; RR 117–157 g/l), thrombocytopenia (90 G/l; RR range 150–350 G/l), decreased hematocrit (0.21 l/l; RR for female patients 0.36–0.48 l/l), slightly diminished kaliemia (3.3 mmol/l; RR 3.5–4.9 mmol/l), natremia (135 mmol/l; RR 136–145 mmol/l), and osmolality (271.4 mOsm/kg; RR 275–295 mOsm/kg), with normal corrected calcemic profile, elevated C-reactive protein (CRP) level (54.3 mg/l; RR < 5 mg/l), elevated hepatic parameters (aspartate aminotransferase 92 U/l; RR 10–35 U/l, alanine aminotransferase 45 U/l; RR 10–35 U/l), alkaline phosphatase 482 U/l (RR 35–104 U/l), gamma-GT 242 U/l (RR < 40 U/l), total bilirubin 19 µmol/l (RR < 15 µmol/l), and conjugated bilirubine 10.1 µmol/l (RR < 5 µmol/l), hypoalbuminemia (31 g/l; RR 35–52 g/l), diminished prothrombin time (53%; RR 80–127%), normal factor V (110%; RR 70–180%), and elevated ferritin (1056 µg/l; RR 20–150 µg/l) associated with normal transferrin saturation coefficient (25.3%; RR 20–40%). Lactate dehydrogenase was within range (211 U/l; RR < 250 U/l). N-terminal pro b-type natriuretic peptide (NT-pro-BNP) level was slightly elevated at 328 ng/l (RR < 300 ng/l). Renal function was normal, and there was neither proteinuria nor albuminuria. Chest radiograph showed no signs of congestive heart failure. Nasopharyngeal PCRs for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), respiratory syncytial virus, and influenza were negative. Human immunodeficiency virus serology was negative. Cytomegalovirus immunoglobulin (Ig)G serology was suggestive of a past infection. Antinuclear antibodies (ANA, $\le$ 1/80; normal range $\le$ 1/80), antineutrophil cytoplasmic antibody (ANCA) PR3 and MPO (< 1 UI/ml; RR, respectively, < 5 UI/ml and < 6 UI/ml), and rheumatoid factor (< 11 UI/ml; RR < 20 UI/ml) and anti-cyclic citrullinated peptide (anti-CCP, 8 U; RR < 20 U) were not detected. Serum protein electrophoresis revealed hypoproteinemia (alpha-2-globulin 4.2 g/l; RR 5.1–8.5 g/l, beta-1-globulin 1.6 g/l; RR 3.4–5.2 g/l, beta-2-globulines 1.4 g/l; RR 2.3–4.7 g/l) and hypogammaglobulinemia (3.7 g/l; RR 8–13.5 g/l). Her basal cortisol concentration was elevated (584 nmol/l; RR 133–537), and her adrenocorticotropic hormone (ACTH) stimulation test was normal (cortisol 725 nmol/l; RR > 500). Serum tryptase concentration was 2.41 µg/l (RR < 13.5 µg/l). The situation deteriorated the next day, as the patient became severely hypotensive with persistent fever and strong chills. She was admitted to the intensive care unit (ICU) for suspected septic shock of unknown origin. Blood cultures were drawn and the patient was started on large-spectrum empiric antibiotic therapy of piperacillin/tazobactam as well as aggressive fluid resuscitation and vasopressors. The initial blood cultures were sterile. Apart from one episode of catheter-related contamination by Staphylococcus hominis, all subsequent blood cultures remained sterile. During the next days, the patient remained tachycardic and hypotensive. Her body temperature stayed very high, up to 40 °C. Intensive fluid administration worsened the anasarca and the patient soon became oxygen dependent. No signs of infection were clinically identified, and laboratory tests still revealed elevated inflammatory markers with procalcitonin level as high as 21 µg/l (RR < 0.05 µg/l). The hypoalbuminemia worsened from 31 g/l to 21 g/l (RR 35–52 g/l). A transthoracic echocardiogram showed normal cardiac cavities and normal global function (left ventricular ejection fraction 80%) without any outflow obstruction, normal heart valves, and no signs of vegetation nor pericardial effusion. As no signs of infectious foci were found, a contrast-enhanced thoracic and abdominal computed tomography (CT) scan was performed. This showed a size difference between both psoas muscles, with the right one being slightly larger but without signs of fluid collection. The liver was normal. Although the patient had neither pain nor symptoms related to the psoas muscles, lumbar magnetic resonance imaging (MRI) was performed to better evaluate them (Fig. 1), which showed, in T1-weighted fat-suppressed (FS) image, signs of intramuscular collections in both psoas muscles: one measuring 21 × 15 mm transversely with a height of 60 mm and the other measuring 8 × 8 mm transversally with a height of 27 mm. There was no sign of spondylodiscitis. At this point, investigations were completed by a fungal infection assessment in the context of the patient’s immunosuppression, by means of a dosage of beta-D-glucan and galactomannan, which were normal (respectively, < 31 pg/ml; RR 60–79 pg/ml, and 0.007; RR index 0.25–0.5). The patient subsequently underwent a radio-guided needle biopsy of the psoas muscles lesions; cytology revealed numerous histiocytes and rare polymorphonuclear neutrophils, and histopathology revealed atrophic striated muscle fibers with fibrous alterations associated with an abundant lymphoplasmocytary infiltrate and focal neutrophils and histiocytosis, with a few hemosiderophages and no granulomas. Grocott and Gram staining failed to reveal a germ. Bacterial, mycobacterial and fungal cultures were sterile, and the panmycobacterial as well as panfungal PCR were negative. The investigation involved an analysis using broad-range bacterial polymerase chain reaction technique on the psoas lesions biopsy. This diagnostic approach relies on amplifying the 16S rRNA gene, followed by sequencing of the resulting amplicons [20, 21]. The obtained sequence underwent rigorous examination employing the Basic Local Alignment Search Tool (BLAST, provided by the National Center for Biotechnology Information at https://blast.ncbi.nlm.nih.gov/Blast.cgi). This analysis revealed 100% similarity to the corresponding segment of the 16S rRNA encoding gene of Neoehrlichia mikurensis (733 out of 733 nucleotides). After a significant delay in diagnosis due to nonspecific symptoms, laboratory testing, and the organization of the radio-guided needle biopsy in a tertiary center, the infection was then confirmed by broad-range bacterial PCR in blood samples. The final diagnoses were secondary capillary leak syndrome and septic shock associated with Neoehrlichia mikurensis infection. The patient was started on doxycycline (100 mg twice daily) for 3 weeks with immediate resolution of the fever, clear clinical improvement, and rapid weaning off vasopressors. This allowed the reintroduction of loop diuretics, resulting in a total weight loss of 18 kg and complete resolution of the anasarca. The treatment was well tolerated without adverse effects. Lumbar MRI suggested subacute psoas hematomas, and follow-up exams showed complete resolution of the hematomas. From the patient’s perspective, she highlighted the burden of weight gain and the prolonged medical uncertainty she faced both prior and during the hospitalization. She emphasized anxiety related to her symptoms and the delay in diagnosis, along with her apprehension about the progression of her condition. The establishment of a formal diagnosis and the rapid resolution of symptoms following treatment brought her significant relief and reassurance.

Fig. 1.

Magnetic resonance imaging T1-weighted fat-suppressed image; bilateral psoas hematomas

Discussion and Conclusion

The patient was immunocompromised due to rituximab treatment for rheumatoid arthritis and had no known history of hematological conditions [13]. Although our patient was exposed to the Ixodes ricinus habitat, she did not recall any tick bites, which is often the case for such tick-borne illnesses. According to a review by Portillo et al., only 39% of patients recalled a tick bite (18 patients from 2008 to 2015) [11]. The reported symptoms and biological findings were similar to those found in the scientific Literature, except for widespread edema and spontaneous hematoma, which to our knowledge, has never been reported. Nonspecific clinical and laboratory findings resulted in delayed diagnosis of 70 days from onset to diagnosis, which is comparable to that reported in literature (median number of 60 days from onset to diagnosis) [10]. Addressing the causes of the anasarca in our patient, a cardiac origin was excluded because NT-pro-BNP levels were not significantly elevated and cardiac function was normal according to the echocardiography. There was no proteinuria or albuminuria. Finally, there were no signs of liver failure, even though the liver parameters were initially slightly elevated; this was interpreted as secondary to septic shock. Furthermore, slightly disturbed liver enzymes have been previously reported in relation to Neoehrlichians mikurensis infection [11]. Thus, the anasarca did not suggest other causative diseases, such as cardiac failure, nephrotic syndrome, or hepatic failure. Hypoalbuminemia secondary to chronic inflammation likely contributed to the presence of edema. However, it was mild at admission despite full body anasarca and swiftly worsened during clinical deterioration on the next day, making the diagnosis of secondary capillary leak syndrome (CLS) to the disseminated Neoerhlichia infection probable [22]. We consider that the capillary leak syndrome was secondary to the disseminated Neoehrlichia infection. However, since members of the Rickettsiales order are well known to have a tropism for endothelial cells, the capillary leak syndrome observed here in the setting of a Neoehrlichia infection may be due to the direct impact of the disseminated intravascular infection on endothelial cell permeability. The cytokine storm at play during the sepsis may of course have further exacerbated the capillary leak syndrome. Conversely, the capillary leak syndrome may have also further negatively impacted the intravascular volume, worsening further the hypotension due to sepsis. Apart from refractory hypotension requiring vasopressor support and hypoalbuminemia, the patient’s laboratory results did not reveal hemoconcentration, which is the last component of the primary CLS triad [16]. Some data suggest that hemoconcentration is less consistent in secondary forms of CLS [22]. Thus, we considered the CLS to be secondary to the disseminated Neoerhlichia infection and partially to the associated septic shock. Some cases of CLS secondary to treatment by rituximab have been described [23]. Since the patient was stably on this treatment for 6 years, without known side effects, we find it very unlikely that her presentation should be secondary to this treatment. In the diagnostic process, imaging revealed two bilateral spontaneous psoas hematomas. In our case, germ identification was initially performed via PCR on the fluid drawn from the hematoma/collection and then confirmed by a second PCR positive on to the blood sample. Although we lacked a precise macroscopic description of the aspect and an exact cellular count from the psoas collection sample, there were no clinical or radiological findings supporting a diagnosis of psoas muscle abscess. Indeed, the MRI scan was not compatible with pyomyositis nor an abscess. Furthermore, the treatment time was considered short (3 weeks of doxycycline) without a relapse and there was a spectacular improvement as soon as treatment with doxycycline was started, arguing against an abscess. Additionally, the ongoing fever and worsening of the patient up to diagnosis despite broad-spectrum antibiotics can be explained by treatments that did not target intracellular bacteria. Neoehrlichia mikurensis is associated with thromboembolic and vascular complications, such as pulmonary embolism, deep vein thrombosis, arterial aneurysms, and transient ischemic attacks [2, 11]. Due to common thrombocytopenia, hemorrhage may also occur. Wenneras et al. reported 54.5% of vascular and/or thromboembolic events in a European study of 11 patients diagnosed with neoehrlichiosis [2]. Although the exact pathophysiology is not fully understood, endothelial vascular tropism is postulated for several reasons: (i) the high incidence of vascular events, (ii) its genetic similarities with E. ruminantium and most Rickettsia species, all of which infect the vascular endothelium, and (iii) the identification of Neoehrlichia mikurensis in circulating endothelial cells from an infected patient, as well as (iv) its recent cultivation in human endothelial cells [3]. Our patient did not experience pain related to the hematoma, had no evidence of trauma, and was not receiving anticoagulation therapy. To our knowledge, no previous cases of Neoehrlichia mikurensis infection have been associated with spontaneous psoas hematomas. This lead us to hypothesize that Neoehrlichia mikurensis may be involved in the presence of spontaneous psoas hematomas, in which the pathogen was identified. However, the role of Neoehrlichia remain speculative, as this is the first reported case and there is a lack of experimental evidence. To the best of our knowledge, this is the first reported case of capillary leak syndrome secondary to an infection with Neoehrlichia mikurensis. Furthermore, these atypical clinical findings (secondary capillary leak syndrome and psoas hematomas) may provide a better understanding of the pathophysiology of this emerging disease. Finally, this case report highlights the added value of the eubacterial PCR.

Abbreviations

CLS

Capillary leak syndrome

PCR

Polymerase chain reaction

ICU

Intensive care unit

RR

Reference range

Author contributions

MDSD and CH contributed equally to the clinical management, the literature review, the redaction of the original draft of this paper. OO and LM contributed equally to the review, editing and supervision of this paper. NF and GG contributed to the critical revision of the manuscript. GG and OO contributed to the publication funding. All the authors agreed to submit the paper for publication.

Funding

The authors declare that no funding was necessary for the case report.

Data availability

All data generated or analyzed during this study are included in this published article.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Written informed consent was obtained from the patient for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.

Competing interests

The authors declare that they have no competing interests.