Another Whipple’s triad? Pericardial, myocardial and valvular disease in an unusual case presentation from a Canadian perspective

Background Whipple’s disease is a clinically relevant multi-system disorder that is often undiagnosed given its elusive nature. We present an atypical case of Whipple’s disease involving pan-valvular endocarditis and constrictive pericarditis, requiring cardiac intervention. A literature review was also performed assessing the prevalence of atypical cases of Whipple’s disease. Case presentation A previously healthy 56-year-old male presented with a four-year history of congestive heart failure with weight loss and fatigue. Notably, he had absent gastrointestinal symptoms. He went on to develop pan-valvular endocarditis and constrictive pericarditis requiring urgent cardiac surgery. A clinical diagnosis of Whipple’s disease was suspected, prompting duodenal biopsy sampling which was unremarkable, Subsequently, Tropheryma whipplei was identified by 16S rDNA PCR on the cardiac valvular tissue. He underwent prolonged antibiotic therapy with recovery of symptoms. Conclusions Our study reports the first known case of Whipple’s disease involving pan-valvular endocarditis and constrictive pericarditis. A literature review also highlights this presentation of atypical Whipple’s with limited gastrointestinal manifestations. Duodenal involvement was limited and the gold standard of biopsy was not contributory. We also highlight the Canadian epidemiology of the disease from 2012 to 2016 with an approximate 4% prevalence rate amongst submitted samples. Routine investigations for Whipple’s disease, including duodenal biopsy, in this case may have missed the diagnosis. A high degree of suspicion was critical for diagnosis of unusual manifestations of Whipple’s disease.

Cardiac manifestations involving Whipple's disease are atypical but are often described as endocarditis, considered the most useful clinical finding [3,4]. Several of the case reports involving endocarditis have a geographic predominance from France [5] to southern Germany [6]. It has been speculated that prevalence of T. whipplei is higher in these so-thought 'endemic' areas or perhaps clinical suspicion of Whipple's is greater. As the bacterium is ubiquitous within the soil environment, there has been speculation of an oral infectious route [4]. Amongst healthy individuals, T. whipplei has been found in 1-11% of stool samples, and up to 26% of sewage plant workers [7]. Pericardial involvement, namely constriction, is a rare clinical finding in patients with Whipple's disease. To our knowledge, this is the first report of confirmed Whipple's disease with constrictive pericarditis in concordance with multi-valvular pancarditis.

Case presentation
A 56-year-old Caucasian male presented with a diagnosis of recurrent heart failure. Previously, he was given a diagnosis of palindromic rheumatism based on a four- year history of lower extremity migratory joint pain and swelling that failed to respond to NSAIDs and hydroxychloroquine. During this time, he also developed unintentional weight loss, fatigue and cognitive impairment. Past medical history was significant for a renal cell carcinoma, treated with a left nephrectomy 4 years prior to presentation. The patient had a 45-pack year smoking history. He had not travelled abroad except for New Zealand and Hawaii 25 years ago. The patient had been previously evaluated over 4 years by numerous services including rheumatology, hematology, cardiology and dermatology with an extensive work-up, including blood cultures that were negative on repeat occasions (Table 1). Eleven months prior to current presentation, the patient developed recurrent acute congestive heart failure exacerbation, requiring multiple admissions with treatment involving diuretics and chest tubes, providing limited short-term symptomatic relief.
On this presentation, the patient had symptoms consistent with an exacerbation of congestive heart failure. The patient stated he had lost approximately 40 pounds over 6 months. He denied any GI symptoms, fevers, chills or night sweats. Physical examination revealed the patient as pale, afebrile and hemodynamically stable. He was noted to be cachectic with a weight of fifty-one kilograms (BMI: 17.8). Pitting edema of both lower extremities to the proximal shin with hyperpigmentation was noted. As observed on the initial CT scan, there was diffuse lymphadenopathy and right epitrochlear lymphadenopathy in particular was noted. Jugular venous distension was noted to 8 cm above the sternal angle with a positive Kussmaul's sign on inspiration. The patient had a 3/6 systolic ejection murder without a pericardial knock. There was no oculomasticatory myorhythmia or supranuculear vertical gaze palsy observed.
Transthoracic echocardiogram imaging was reviewed from 2014 to the 2016 (Figs. 1, 2, Additional files 1 and 2). In 2014 (Figs. 1a, 2a and Additional file 1), subtle tethering of the mitral valve was noted with thickening and cord calcification. The posterior leaflet had reduced excursion and the mitral valve apparatus was apically displaced. There was the observation of subtle diastolic doming over the anterior mitral leaflet. Progressive echocardiogram findings in 2016 (Figs. 1b, 2b and Additional file 2) showed the anterior mitral valve to be thickened with reduced excursion, the posterior leaflet to be fixed and cords with shortening. Overall, this was in keeping with pseudoprolapse. Over the aortic valve, there was systolic doming of the right coronary cusp and calcification from the commissural aspect of the aortic valve with sparing of the base. Notably, there was the presence septal shuttering in early diastole with early inspiratory septal bounce/shift, indicative of abnormal heart and lung interaction. Overall, this was suggestive of early constrictive physiology.
A repeat transthoracic echocardiogram on admission showed numerous interval changes including; a reduction in left ventricle systolic function (ejection fraction of 48%), severe mitral regurgitation, moderate tricuspid regurgitation, moderate aortic regurgitation and worsening pulmonary hypertension (RSVP 78 mmHg). Calcific aortic valve changes were noted that were felt to be post-inflammatory as opposed to degenerative changes. The mitral valve had an unusual appearance suggestive of prior valvulitis, in particular rheumatic in nature. The pericardium was thickened with features of exaggerated heart and lung interactions, consistent with constrictive physiology. Based on these findings, pericardial stripping and valvular surgery were proposed.
Perioperative findings consisted of thick pericardium and densely adherent to the surrounding anatomy. The mitral valve was grossly abnormal with fibrosis, in keeping with a rheumatic type process and was subsequently replaced with a bioprosthetic valve. The tricuspid valve had thickened leaflets with slightly thickened papillary muscle and was repaired with a ring annuloplasty. The thickened aortic valve was addressed with primary repair.
The above constellation of symptoms, in the context with prior work-up as unremarkable, lent to a strong clinical suspicion of an atypical presentation of Whipple's disease. Two days post-op, duodenal biopsies were obtained (Fig. 3a, b), and surprisingly were not consistent with Whipple's disease. On the request of the clinician, the cardiac valves and pericardium pathology specimens were examined for PAS staining and PCR. The cardiac specimens demonstrated numerous foamy macrophages filled with PAS positive material compatible with Whipple's disease (Fig. 4). Tropheryma whipplei was identified by direct 16S rDNA PCR on the mitral valve with > 99% sequence match. The pathological changes were consistent with active endocarditis, myocarditis, and pericarditis, caused by T. whipplei. The patient started treatment with IV Penicillin G for 14 days and subsequently maintained regimen of oral sulfamethoxazoletrimethoprim for long-term treatment. Six months following diagnosis, the patient has had no complications and is clinically improving.

Discussion and conclusions
The diagnosis of T. whipplei valvular involvement is elusive as there is the absence of clear diagnostic criteria. Often the diagnosis hinges on 16S rDNA PCR following surgical removal of valvular tissue coinciding with high clinical suspicion from physician vigilance. Currently, limited serologic assays only can distinguish between The diagnostic criterion for Whipple's disease includes two of three tests to be positive from PAS staining, PCR or immunohistochemistry [8]. The gold standard tissue is considered to be histology from small bowel biopsies, which classically present as PAS-positive foamy macrophages within the lamina propria [9]. However, PASpositive macrophages are not specific and have been observed in patients with intestinal infections (Mycobacterium avium complex, Rhodococcus. sp., Bacillus cereus, Corynebacterium spp., and Histoplasma spp.,), Crohn's disease, histiocytosis and Waldenstrom's macroglobulinemia [9][10][11].
We also sought to assess the epidemiology of Whipple's disease in Canada, which previously was unknown. A consortium of data collected from the National Microbiology Laboratory in Canada was done from 2012 to 2016 and highlighted in Table 2. Notably, only 4% of samples were positive from those collected; of which 13.5% were elucidated from the gold standard of duodenal biopsy, with 86.2% of other tissue types being found to be positive for T. whipplei. A limitation of this should be noted in that multiple samples were present from the same patient. However, this highlights the diagnostic utility of other samples apart from the previously denoted 'gold standard' of duodenum.
Our patient presented with heart failure due to pericardial constriction with multiple cardiac valve involvement in the context of a systemic process including cachexia and neurological impairment with reduced cognition. He did not display classical Whipple's symptoms including gastrointestinal manifestations and had few PAS-positive macrophages within the intestinal wall. Given the current diagnostic guidelines, this may substantially underrepresent the true incidence of Whipple's disease in the absence of overriding strong clinical suspicion. The accurate clinical evaluation of the patient described in this case is likely multifactorial and may represent improved modern medical care to bridge to surgery, improved current diagnostic imaging and high clinical acumen. PAS staining is not routinely performed on cardiac specimens at our institution. Furthermore, this patient represents the first description in the literature of constrictive pericarditis with pan-valvular findings. Constrictive pericarditis, involving cardiac valvular pathology, in the absence of common etiologies should promote clinical suspicion as a manifestation of Whipple's disease even in the absence of gastrointestinal manifestations.
This case further highlights as a key example of blood culture-negative endocarditis (BCNE). BCNE are thought to account for 2.5-70% of all cases of endocarditis based upon standard culture-dependent laboratory techniques, with a differential diagnosis including both infectious and non-infectious etiologies (Table 3) [13]. Among culture-negative endocarditis, T. whipplei was found to be the fourth most common cause as demonstrated by both histological and molecular techniques [14]. A recent study from France demonstrated that by adding real-time PCR assays in patients with BNCE, not only did the diagnostic efficiency improve by 24.3%, but detection of atypical organisms including T. whipplei also increased, thus improving diagnostic certainty [15]. In Canada, epidemiological rates of BCNE are currently unknown, but given the prevalence from prior case series from other countries is likely underreported.
Ultimately, as this case highlights, repeated negative blood cultures should prompt the clinician to further evaluate atypical causes of BCNE, including Whipple's disease. This case further reinforces prior studies in which addition of molecular studies may aid in diagnostic certainty by increasing positive yield. Further research is required in order to optimize medical care and highlighting a likely under-diagnosed disease; both with regards to Whipple's disease and BCNE. *GI tract includes fresh or paraffin embedded tissue samples from small and large intestine and samples labelled as "gastric" **Multiple positive sample types may be submitted from the same patient