Secondary Logo

Journal Logo

Your Diagnosis, Please

Fever and Limp in a 10-Year-old Girl With Congenital Heart Disease

Rodrigues, Jorge MD*; Pinto, Margarida MD; Brito, Maria João MD*; Martins, José Diogo MD, PhD; Gouveia, Catarina MD*,§; on behalf of The Group

Author Information
The Pediatric Infectious Disease Journal: November 2021 - Volume 40 - Issue 11 - p 1055-1057
doi: 10.1097/INF.0000000000003119
  • Free


A 10-year-old girl was admitted with a 7-day history of fevers up to 40°C (every 4–6 hours), right upper thigh and lumbar pain and inability to walk for 2 days. She also had 2 episodes of vomiting and watery diarrhea. She denied other symptoms and had no history of recent infection or trauma. Her medical history was significant for a diagnosis of tetralogy of Fallot at 3 years of age, requiring corrective surgeries at age 3 and 6, including placement of a right ventricle-to-pulmonary artery (RV-PA) conduit.

Her family history was unremarkable. Born in Cape Verde, she had resided in Portugal since she was 3 years old. At the time of illness, she had returned from rural Cape Verde where she had stayed for 3 months, had contact with various livestock (including unvaccinated dogs, pigs, cows and poultry) and had drank unsafe fresh water. She had not attended a traveler’s medicine appointment before the trip, and her immunization schedule was incomplete, lacking a diphtheria-tetanus booster dose and the first dose of the human papillomavirus vaccine.

On admission, the patient was ill-appearing with fever of 38.8°C (tympanic), heart rate of 85 beats/min, blood pressure of 114/71 mm Hg, respiratory rate of 20 breaths per minute and oxygen saturation of 98% in ambient air. Her physical examination was also notable for a grade IV/VI mitral holosystolic murmur without signs of cardiac dysfunction, including no hepatosplenomegaly. In addition, she had right hip tenderness, positive right Flexion Abduction External Rotation test and lumbar pain when moving the hip, without significant limitation. Several insect bites were noted on both legs.

Her complete blood count revealed a hemoglobin of 9.5 g/dL; hematocrit of 28.2%; white blood count of 20.5 × 109/L with differential of 81.3% neutrophils, 12.2% monocytes, 5.5% lymphocytes, 0.8% eosinophils and 0.2% basophils and platelet count of 173 × 109/L. Coagulation panel showed prothrombin time 14.3 seconds (10.2–12), activated partial thromboplastin time 25.6 seconds (26–37), fibrinogen 8.8 g/L (1.54–4.88) and D-dimers 781 μg/L (< 230). Her chemistry panel revealed urea 64 mg/dL (15–36), creatinine 0.77 mg/dL (0.44–0.69), aspartate transaminase 54 U/L (18–36), alanine transaminase 26 U/L (9–25) lactate dehydrogenase 616 U/L (157–272), C-reactive protein 373 mg/L (< 5.0), procalcitonin 5 ng/mL (< 0.06), erythrocyte sedimentation rate 76 mm/h (< 16) and urine leukocyte count 18/μL (< 25). The remaining values were within normal range. Hip pelvic radiography and hip ultrasonography were normal. Pelvic magnetic resonance imaging showed T2-weighted signal hyperintensity in the right sacroiliac joint line and adjacent muscles and subcutaneous tissue, with enhancement of the subchondral regions and fluid within the sacroiliac joint, without interline irregularities, edema or erosions, suggesting the presence of sacroiliitis (Fig. 1). There were no signs of arthritis of the hip joint, osteomyelitis of the femur and iliac bones nor any intramuscular collections.

Figure 1.
Figure 1.:
Pelvic magnetic resonance imaging exhibiting T2-weighted signal hyperintensity in the right sacroiliac joint line and adjacent muscles and subcutaneous tissue, with enhancement of the subchondral regions and fluid within the sacroiliac joint, suggestive of sacroiliitis.

Additionally, interferon-gamma release assay for tuberculosis, serologic testing for HIV, Q fever, bartonellosis and brucellosis and stool and urine culture were negative. Chest radiography and abdominal ultrasonography were also normal.

The patient was treated with intravenous ceftriaxone and clindamycin.

An additional result revealed the diagnosis.

For Denouement see P. 1056.


(Pediatr Infect Dis J 2021;40:1056)

Continued from P. 1055.

On day 3, Corynebacterium diphtheriae was isolated from 2 consecutive blood cultures. A nontoxigenic strain was identified by the National Health Institute through the Elek test. Susceptibility testing revealed the bacteria was sensitive to both penicillin and cephalosporins, which prompted the switching of therapy from ceftriaxone to penicillin, while continuing clindamycin because of concern of bacterial coinfection. Serial transthoracic echocardiograms failed to reveal cardiac involvement. Cutaneous and oropharyngeal cultures were negative.

On day 4, the patient developed signs of arthritis on her right ankle and underwent arthrocentesis. Synovial fluid showed glucose of 98 mg/dL, total protein of 4.2 g/dL and lactate dehydrogenase of 641 U/L. There was insufficient sample for cytologic analysis, but culture and 16s polymerase chain reaction of the synovial fluid were negative.

The patient became afebrile by day 8 of hospitalization and was discharged after completing 14 days of intravenous antibiotics with slight residual pain on internal rotation and abduction of the right leg. On discharge, her immunization schedule was updated with the sixth dose of diphtheria-tetanus toxoid, and she was prescribed an additional 6 weeks of oral amoxicillin to complete at home for presumed osteoarticular infection from C. diphtheriae.

At her 4-month follow-up, the patient complained of dyspnea with moderate exertion, her sedimentation rate increased to 58 mm/h and a transthoracic echocardiogram now revealed an obstacle in the RV-PA distal conduit, with severe right ventricle dysfunction and hypertrophy and paradoxical movement of the interventricular septum. Transesophageal echocardiogram documented the presence of vegetations in the RV-PA conduit. Moreover, this conduit, which had previously been functioning normally, was now severely stenotic. All these findings were consistent with endocarditis. A diagnosis of subacute bacterial endocarditis because of C. diphtheriae was made, although blood cultures collected at the time were negative. She was treated with intravenous penicillin for a total of 10 weeks and gentamicin for 14 days and underwent uneventful surgical replacement of the dysfunctional conduit 5 months after the first admission. Culture of the vegetation and the conduit were both negative, but they were not sent for 16s polymerase chain reaction. She has remained asymptomatic ever since.

C. diphtheriae, a Gram-positive, nonmotile, nonsporulating, aerobic, pleomorphic coccobacillus, first described in 1884 as the cause of diphtheria.1 Humans are the most frequent host, and transmission usually occurs via droplets and only rarely through skin lesions or fomites, although some reports suggest skin carriers to be more infectious than previously assumed.1 The pathogenesis of the classic respiratory form of diphtheria is mediated by toxigenic strains of C. diphtheriae, which are typically noninvasive but can induce severe local inflammation of the mucosa. Since the introduction of effective immunization in mid-20th century, cases of diphtheria have been steadily declining, with around 5000 cases diagnosed worldwide per year between 2010 and 2015.2 The burden of disease lies mainly in several developing countries, where it remains endemic.2

The significant decrease in the incidence of diphtheria after the introduction of the vaccine has paved the way for the emergence of nontoxigenic C. diphtheriae (NTCD) strains, as diphtheria toxoid immunization only protects against toxin-mediated disease.3 In fact, at the end of the 1990s, NTCD strains started to emerge in Europe and America,4 and multiple outbreak-associated clones have since been reported in Europe, mostly from Germany, France, United Kingdom, Switzerland and Poland.4–8 Wagner et al9 undertook a multicenter European screening study in populations with presumed high vaccination coverage in 2008 and found a widespread pattern of NTCD in circulation, pointing to a potential reservoir for future epidemics. While nontoxigenic strains do not contain the toxin, they do have the ability to adhere to host cells and stimulate cytokine production by the host immune system, influencing disease severity.8 In addition to the virulence, host factors may contribute to invasive infection, and the main predisposing factors for colonization and infection are intravenous drug use, poor socioeconomic conditions, dental caries, alcoholism and homelessness.4–9

Systemic infections by NTCD have been increasingly reported and include bacteremia and sepsis, endocarditis, meningitis, skin infection and osteoarticular infection.8,9 These are more common in predisposed individuals, such as those with congenital heart disease or prosthetic valvular disease. Infection is also increased by unsanitary, crowded conditions,10 which predispose to skin infection and colonization. This serves as an important reservoir for transmission to susceptible patients in the community. We suspect that our patient became colonized through cutaneous lesions by a nontoxigenic strain while in Cape Verde and that her artificial conduit provided an ideal nidus for proliferation of the organism.

Endocarditis is a common complication of systemic disease and one of the most frequent causes of morbidity and mortality.10 Other types of cardiac complications such as myocarditis, dysrhythmias and conduction abnormalities result from toxin-mediated C. diphtheriae.11 The most comprehensive review of C. diphtheriae endocarditis included 76 patients (of which 36 were caused by NTCD), most of whom were males under the age of 30 years and about one-third with underlying valvular disease.12 According to Schnell et al,10 vegetations caused by NTCD are typically large and prone to embolization, and valvular lesions are particularly destructive, resulting in the need for surgery in up to 43% of cases and a high mortality rate (41%). In contrast, Muttaiyah et al12 reported that only 18% of their cases required surgery, yet their observed mortality rate was similarly high at 38%. In fact, our patient required surgical intervention with conduit replacement because of severe stenosis and disease progression.

To date, 16 cases of osteoarticular involvement caused by C. diphtheriae have been reported, 11 with NTCD strains and 9 occurring concomitantly or secondary to endocarditis.12–15 Six cases were reported in children, 2 of whom had associated endocarditis.12–15 The first case was an 11-year-old female with mitral valve disease and septic arthritis of the knee described by Hogg et al.13 The second was a 5-year-old female with a ventricular septal defect and osteomyelitis of the calcaneum reported by Muttaiyah et al.12 There has been only 1 case of C. diphtheriae sacroiliitis previously described in literature, a 22-year-old woman with a prosthetic mitral valve.14

At present, there are no consensus guidelines for treatment of endocarditis caused by C. diphtheriae.10 Intravenous penicillin is usually effective and generally recommended. Endocarditis and septic arthritis can be treated with either a β-lactam alone or in combination with an aminoglycoside for 4–8 weeks.1,12 Muttaiyah et al12 reviewed 11 cases treated solely with β-lactam antibiotics and 25 patients who had received β-lactam and aminoglycoside, with similar outcomes and no relapses.

In our case, the diagnosis of endocarditis was not initially recognized. Sequential IV to oral antimicrobial treatment (2 weeks intravenous plus oral regimen maintained for 6 weeks) was provided for presumed osteoarticular infection but was inadequate treatment for the unrecognized endocarditis.16 In hindsight, our patient’s initial presentation with 2 separate foci of infection was likely a manifestation of septic embolization to sacroiliac and ankle joints. This finding combined with her previous cardiac surgery for tetralogy of fallot, including an artificial conduit, and her bacteremia with NTCD, a pathogen associated with destructive endocarditis, should have prompted strong consideration of endocarditis despite initially unremarkable transthoracic echocardiograms. This case raises awareness of NTCD as significant pathogen that can cause invasive disease, particularly in predisposed hosts.


The authors thank Dr. Paula Lavado for the help in strain identification and assessment of toxin in our isolate. Members of the group: Pedro Jordão Å, Pediatric Orthopedic Unit, Hospital de Dona Estefânia, CHULC - EPE, Lisbon, Portugal; Alexandra Vasconcelos, Infectious Diseases Unit, Hospital de Dona Estefânia, CHULC - EPE, Lisbon, Portugal.


1. Stechenberg BW. Feigin RD, Cherry JD, eds. Diphtheria. In: Textbook of Pediatric Infectious Diseases. 2018:8th ed. Elsevier; 931–938.e2.
2. Clarke KEN. Review of the Epidemiology of Diphtheria (2000-2016). 2017.US Centers for Disease Control and Prevention;
3. Zakikhany K, Efstratiou A. Diphtheria in Europe: current problems and new challenges. Future Microbiol. 2012;7:595–607.
4. Funke G, Altwegg M, Frommelt L, et al. Emergence of related nontoxigenic Corynebacterium diphtheriae biotype mitis strains in Western Europe. Emerg Infect Dis. 1999;5:477–480.
5. Dangel A, Berger A, Konrad R, et al. Geographically diverse clusters of nontoxigenic Corynebacterium diphtheriae infection, Germany, 2016-2017. Emerg Infect Dis. 2018;24:1239–1245.
6. Zasada AA, Baczewska-Rej M, Wardak S. An increase in non-toxigenic Corynebacterium diphtheriae infections in Poland–molecular epidemiology and antimicrobial susceptibility of strains isolated from past outbreaks and those currently circulating in Poland. Int J Infect Dis. 2010;14:e907–e912.
7. Zakikhany K, Neal S, Efstratiou A. Emergence and molecular characterisation of non-toxigenic tox gene-bearing Corynebacterium diphtheriae biovar mitis in the United Kingdom, 2003-2012. Euro Surveill. 2014;19:20819.
8. Sharma NC, Efstratiou A, Mokrousov I, et al. Diphtheria. Nat Rev Dis Primers. 2019;5:81.
9. Wagner KS, White JM, Neal S, et al.; Members of the Diphtheria Surveillance Network. Screening for Corynebacterium diphtheriae and Corynebacterium ulcerans in patients with upper respiratory tract infections 2007-2008: a multicentre European study. Clin Microbiol Infect. 2011;17:519–525.
10. Schnell D, Beyler C, Lanternier F, et al. Nontoxinogenic Corynebacterium diphtheriae as a rare cause of native endocarditis in childhood. Pediatr Infect Dis J. 2010;29:886–888.
11. Washington CH, Issaranggoon na ayuthaya S, Makonkawkeyoon K, et al. A 9-year-old boy with severe diphtherial infection and cardiac complications. BMJ Case Rep. 2014;2014:bcr2014206085.
12. Muttaiyah S, Best EJ, Freeman JT, et al. Corynebacterium diphtheriae endocarditis: a case series and review of the treatment approach. Int J Infect Dis. 2011;15:e584–e588.
13. Hogg GG, Strachan JE, Huayi L, et al. Non-toxigenic Corynebacterium diphtheriae biovar gravis: evidence for an invasive clone in a south-eastern Australian community. Med J Aust. 1996;164:72–75.
14. Tiley SM, Kociuba KR, Heron LG, et al. Infective endocarditis due to nontoxigenic Corynebacterium diphtheriae: report of seven cases and review. Clin Infect Dis. 1993;16:271–275.
15. Holthouse DJ, Power B, Kermode A, et al. Non-toxigenic Corynebacterium diphtheriae: two cases and review of the literature. J Infect. 1998;37:62–66.
16. Hirata R Jr, Pereira GA, Filardy AA, et al. Potential pathogenic role of aggregative-adhering Corynebacterium diphtheriae of different clonal groups in endocarditis. Braz J Med Biol Res. 2008;41:986–991.

sacroiliitis; osteoarticular infection; endocarditis; Corynebacterium diphtheria

Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.