Culture-negative endocarditis: “whatever remains, however improbable, must be the truth”; the importance of asking the right questions!

Abstract

The complicated case of a patient with recurrent culture-negative endocarditis with a history of repeated mitral valve replacement is described. Investigations disqualified common pathology but serology revealed a diagnosis of Q fever endocarditis. The ongoing problematic management of this patient is described, followed by a brief review of the clinical features, investigations and treatment of Q fever.

BACKGROUND

Q fever is a rare disease in the UK. It highlights the need to exclude common causes before more complex conditions are investigated. Further to this, it is important that patients with repeat admission with similar problems are fully investigated to rule out all possible causes.

This case confirms that once Q fever endocarditis has been diagnosed it is very complicated to treat, requiring concurrent medical and surgical intervention and with a low probability of cure

CASE PRESENTATION

A 67-year-old retired mining engineer with hypertension presented in 1999 with a persistent cough and dyspnoea, and was found to have grade III mitral regurgitation (MR). Although a former smoker 30 years previously, his chest x ray was normal; it was believed that the mitral regurgitation caused his symptoms. Transoesphageal echocardiography (TOE) revealed mild leaflet thickening, no annular calcification and a flail posterior mitral valve leaflet (PMVL), causing severe MR. He gave no history of rheumatic fever. Further investigation confirmed three-vessel coronary heart disease. He underwent an uncomplicated three-vessel coronary artery bypass graft (CABG) and mosaic porcine mitral valve replacement (MVR) with an uneventful recovery. Local guidelines indicated that a tissue valve was recommended as he was over 65 and did not want long-term anticoagulation, which would have been required with a metallic valve.

Follow-up routine transthoracic echocardiography (TTE) in 2002 and 2004 showed the prosthesis to be functioning satisfactorily with no significant residual MR. In October 2005, he underwent elective left total hip arthroplasty for osteoarthritis, which was complicated by formation of localised haematoma with colonisation with faecal Streptococcus and coagulase-negative Staphylococcus. He had manipulation under anaesthetic for dislocation of the hip joint, and long-term antibiotic treatment with coamoxiclav. However, he became unwell with progressive dyspnoea and subsequent haemoptysis. A CT pulmonary angiogram (CTPA) was negative but further TTE revealed significant MR, with negative blood cultures, but no vegetations were identified on TOE.

Repeat cardiac surgery occurred in October 2006. Macroscopic examination by the surgeon of the MV prosthesis described a left-sided leaflet tear, with severe MR with appearance of possible endocarditis. A new mosaic bioprosthesis was implanted, and the notes documented the plan for the resected prosthesis to be sent to microbiology; unfortunately there was no evidence of the sample or PCR findings. His postoperative recovery was complicated in December 2006 by pneumonia and atrial fibrillation. Follow-up TOE (April 2007) suggested the new valve was well positioned with mild to moderate paravalvular MR, unchanged from the perioperative TOE. In his latest admission he reported lethargy, rigors, sweats and an unintentional weight loss of 6.4 kg over 18 months. Clinical examination revealed a pansystolic murmur and splenomegaly.

INVESTIGATIONS

Transthoracic echocardiography at this stage identified a small mobile echogenic structure suggestive of vegetation on the posterior MV strut (fig 1)

Figure 1.

Transthoracic echocardiography image illustrating small echogenic structure on posterior mitral valve (MV) strut.

Multiple venous blood cultures (prior to antimicrobial therapy) proved negative, but culture-negative endocarditis was considered. Investigations showed non-specific elevation of immunoglobulins (IgG 32.7, IgA 10.80, IgM 8.15), paraprotein of 7 g, normocytic, normochromic anaemia with a blood film showing monocytosis, toxic neutrophils and rouleaux formation.

Results were negative for HACEK (for Haemophilus, Actinobacillus, Cardiobacterium, Eikenella and Kingella) organisms, lyme and brucellosis, but positive Coxiella burnetii titres were found; IgG >1280, IgM >1280.

CTPA in 2006 was negative for thromboembolic disease and CT of the abdomen, pelvis and thorax in 2007 showed no malignancy or lung disease but evidence of mild cardiac failure due to small bilateral plural effusions and cardiomegaly. An autoantibody screen to assess for vasculitis was negative.

DIFFERENTIAL DIAGNOSIS

Subacute bacterial endocarditis due to Coxiella. Multiple myeloma was excluded.

TREATMENT

Suspecting subacute bacterial endocarditis, the patient was treated with vancomycin and gentamycin. Serology was sent for HACEK organisms, lyme, brucellosis and C burnetii. Surprisingly, the C burnetii titre returned positive (IgG >1280, IgM >1280) indicating active Q fever. Given the lack of clinical or biochemical improvement, he was switched to oral doxycycline and ciprofloxacin,¹ after microbiological advice. He clinically improved with decreasing inflammatory markers (table 1). His treatment was complicated by a marked drug-induced photosensitivity rash, successfully treated with steroids and sunscreen. A recent TOE (fig 2) showed no vegetations but a perforation in the posterior leaflet of the mitral valve resulting in a regurgitation jet with an area of 8.1 cm², therefore he has been referred for a repeat MVR.

Table 1.

C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) results over time

	November 1999	October 2002	November 2005	December 2005	October 2006	April 2007	September 2007	Current
CRP (n<5)		62	37	0	20	22	95	28
ESR (n<10)	40	60	59	44	74	87	141	125

Figure 2.

Transoesophageal echocardiography image illustrating severe mitral regurgitation (MR) through a perforation in the posterior mitral valve (MV) leaflet.

OUTCOME AND FOLLOW-UP

His future management consists of another (third) MVR with potential lifelong antibiotic therapy, with concomitant side effects. Surgery would be indicated for haemodynamic reasons, requiring 3 weeks of preoperative replacement antibiotic chemotherapy, to sterilise the blood to minimise risk to the new valve.²

DISCUSSION

It is of interest to note that the patient had previously worked on farmland as a mining engineer. This case highlights the need to exclude common causes but also have an open, inquisitive mind, with a detailed history, when investigations are negative in complicated cases. Indeed, culture-negative endocarditis including brucellosis, lyme disease and Q fever should be considered in such individuals. It is possible the initial endocarditis and its reoccurrence could have actually been persistent undiagnosed Q fever infection.

Q fever is a rickettsiosis due to C burnetii, causing acute or chronic disease. It is a zoonosis spread throughout the world, affecting mostly men over the age of 40.³ Q fever endocarditis has a 24% death rate. Furthermore, a study showed that 55% of people with Q fever have prosthetic valve involvement.⁴ Patients with mitral and aortic valves specifically are more susceptible to Q fever.

Typical symptoms of Q fever endocarditis include valve dysfunction, symptoms of heart failure, constitutional symptoms (commonly, fever, malaise, night sweats, chills, weakness, weight loss). Peripheral manifestations include hepatosplenomegaly, clubbing, purpuric rash, microscopic haematuria and embolic events.

Owing to a vague presentation with negative blood cultures and echocardiogram, diagnosis is serological with measurement of antibody levels of phase 1 (IgM) and phase 2 (IgG) antigens. Antibody levels are also used to monitor the effectiveness of treatment.

Anaemia and thrombocytopoenia are common concurrent finding as a result of haematuria in approximately 50% of patients.^3,4 Other serology abnormalities include elevation of serum aminotransferases, lactate dehydrogenase, globulins, anti-smooth muscle antibodies, anti-phospholipid antibodies, anti-mitochondrial antibodies and rheumatoid factor.⁵

Prosthetic cardiac valves need regular surveillance by a cardiologist or in a designated valve clinic. If there is evidence of any valvular dysfunction this requires careful and thorough clinical and investigative assessment by a specialist. A transoesophageal echocardiogram is crucial in this assessment, as was the case in our patient. Furthermore, if prosthetic valve endocarditis is suspected a comprehensive assessment should include exclusion of culture-negative organisms, especially if serial blood cultures remain negative as was the case in our patient; PCR analysis of the resected infected valve at the time of replacement may help confirm the diagnosis in such difficult cases and also guide further treatment.⁶

Relapses are common after medical and surgical treatment in patients with or without positive valve cultures.^2,3 Combination antibiotic therapy is required for the initial 18 months, but with relapse lifelong treatment maybe required.^3,4 A retrospective study showed tetracycline (doxycycline) and hydrocholoroquine is preferred, but other combinations such as doxycycline and ciprofloxacin or doxycyclin and rifampicin are suitable alternatives.⁷

There are approximately 70–100 published case reports of Q fever endocarditis (table 2). They focus on native valve endocarditis and those involving the aortic valve.

Table 2.

Published case reports of Q fever endocarditis

Author (year of publication)	Valve	Focus	Outcome
Siciliano et al (2008)8	Native aortic		Death despite specific antibiotic therapy and valve surgery treatment
Krol et al (2008)9	Biological prosthetic aortic	Difficult diagnosis	Successfully treated with doxycycline monotherapy; prosthetic valve endocarditis is rare compared with native valve
Ngatchou et al (2007)10	Bicuspid aortic	Endocarditis with left ventricular fistula	Replacement and fistula repair were performed; treated with doxycycline and chloroquine for 1 year
Fradi et al (2006)11	Native aortic	Mitral stenosis and abscess of the mitroaortic trigone	Doxycycline, chloroquine, ofloxacine for 18 months and valve replacement
Deyell et al (2006)12	Mechanical aortic	Non-specific presentation use of serology for diagnosis	Treated with ciprofloxacin and rifampin and replacement, healthy at 12 months; earlier diagnosis could significantly decrease the morbidity and death rate associated with this disease
Issartel et al (2002)6	Native mitral	PCR diagnosis	PCR of infected valve found q fever and advised routine serology of Coxiella burnetii
Ciudad et al (1994)13	Aortic Prosthetic and native mitral	Recurrent disease treated with doxycycline and valve replacement	Recurrent disease with repeated successful outcomes. Can Q fever be eradicated? What is the required duration of antibiotic therapy for this disease?
Pedoe (1970)14	Aortic native		Treated with tetracycline and valve replacement but recurrent illness lead to patient death

Of all the case reports published, only a small subset of those involve mitral valve prosthesis. In our instance, the majority of published cases concentrate on the aortic valve. However, the principles of therapy can be applied to both.

LEARNING POINTS

• This case highlights the need to having an open, inquisitive mind, with a detailed history, when investigations are negative in complicated cases.

• There is a need for a high index of suspicion, after common conditions have been excluded, in culture-negative endocarditis.

• Te difficultly in diagnosing and managing Q fever endocarditis using medical and surgical intervention is highlighted.

• With prosthetic valve leakage, negative blood culture endocarditis should be suspected and the resected sample sent to pathology and microbiology departments for PCR evaluation.