According to the World Health Organization (WHO) Global Tuberculosis Report released in 2016, TB remains one of the top 10 leading causes of morbidity and death worldwide, even surpassing mortality attributable to HIV annually [12]. This report states that in the year 2015, 10.4 million people were infected with TB globally, and 1.4 million deaths were attributed to this disease. WHO surveillance data shows that TB is a major problem in the African region. Although the continent claimed 26% of the world’s new cases in the same year, it alarmingly had the most severe burden relative to population, that is 275 cases for every 100,000 persons. This figure surpassed more than double the global average of 142 per 100,000 [12]. Our country Kenya is currently classified as one of the 30 high TB burden countries, which collectively account for 87% of all estimated incident cases worldwide. In the year 2015, Kenya’s estimated incidence rate of TB was 233 per 100,000 population, [12] and this disease has been noted to rank as one of the leading causes of mortality in the country [13, 14]. In view of the above, one recognises the importance and relevance of TB to clinicians and public health services in our local setting, regionally and globally at large.
This case report highlights a rare instance of primary splenic tuberculosis. The spleen is one of the organs involved in extrapulmonary TB. However, more common extrapulmonary sites include the lymph nodes, pleura, genitourinary tract, bones and joints, meninges, peritoneum and pericardium [15].
Haematogenous dissemination occurs frequently in HIV-infected individuals. Isolated splenic involvement is rare, and is usually reported in immunosuppressed individuals [1–3, 8]. Coley first described TB of the spleen in 1846, referring to an “enlarged spleen secondary to tuberculosis with absent or limited involvement of other organs” [16]. Our patient was atypical in that he was a healthy immunocompetent male without any risk factors for splenic infection, and in whom MTB diagnostic tests (WHO-approved) such as ZN staining for AFB and GeneXpert MTB/RIF were negative. The initial presentation as pyrexia of unknown origin is characteristic of splenic TB in literature [7, 17]. Apart from fever which is present in majority of patients with splenic TB, other modes of presentation reported include fatigue and weight loss, splenomegaly, [18] splenic rupture, [19] hypersplenism, portal hypertension with and without gastrointestinal bleed, and fulminant forms involving rapid progression [18]. Hematologic abnormalities include reduced cell counts but cases of polycythaemia have also been reported [20].
Isolated splenic tuberculosis frequently proves to be a diagnostic conundrum as demonstrated in our case report, mainly due to the vague non-specific nature of clinical presentation. In many case reports, imaging is important in initial identification of splenic pathology, and diagnosis further clinched by histopathological examination of a splenic fine needle aspirate, core needle biopsy or splenectomy specimen [4]. Abdominal ultrasound is a cost-effective non-invasive imaging modality in such cases and may show a miliary pattern, nodular TB, tuberculous spleen abscess, calcific TB or a combination of these findings [9]. Our patient demonstrated nodular TB. On the other hand, abdominal CT scan has added value in ruling out involvement of other organs, and has higher diagnostic accuracy than ultrasound which suffers from operator-dependence. CT scan of the abdomen would often show multiple rounded hypodense lesions which may be present in a variety of conditions other than splenic TB [1, 21]. Differential diagnosis of solitary splenic masses that ought to be considered include cysts, haematoma, fungal infection, abscesses, infarcts, vascular tumours, lymphoma and metastatic tumours [8].
Histological confirmation via splenic biopsy provides a more accurate diagnosis of splenic pathology. Aspiration cytology of splenic lesions is variable. Suri et al. [22] reported up to 88% sensitivity for fine needle aspiration cytology (FNAC) for diagnosing a tuberculous pathology in the spleen. On the other hand, microscopic observation of tissue sections allows for histological typing and staging of tubercle lesions, as well as differentiation between granulomatous lesions and other radiographically comparable lesions such as lymphomas. However literature suggests that fixing of tissues with formalin and xylene greatly reduces the sensitivity of acid-fast staining and could potentially lead to false negatives [23]. Pottakkat et al. [24] reported that AFB staining of splenic tissue sections was negative in over 339 patients who underwent splenectomy for indications other than trauma. Fukunaga et al. [23] reported that tubercle bacilli were frequently missed or underestimated with acid fast microscopy on formalin fixed, paraffin embedded tissues. In such cases, real-time polymerase chain reaction (PCR) of the tissue sections demonstrated markedly increased sensitivity over acid-fast staining.
CNB has a high diagnostic yield in splenic pathology [25] and has demonstrated superior diagnostic accuracy to FNAC in characterising splenic lesions [26, 27]. Tubercular infection is histologically confirmed by presence of typical caseation with granuloma of Langhan’s giant cells and epitheloid cells. In our case, classical granulomatous inflammation and acid fast bacilli were demonstrated upon examination of the CNB specimen and further confirmed by GeneXpert MTB/RIF.
Therefore, in apparent “TB negative” cases, performing more than one test improves yield. In our case, despite several ZN stain negative results, other diagnostic modalities were employed such as GeneXpert MTB/RIF as well as histological confirmation through tissue diagnostic procedures including FNAC and CNB. All this was done in addition to the conventional ZN staining technique to get the final diagnosis of TB spleen. It is important to note that culture is useful as well, as it is still considered the current reference standard for detecting MTB in spite of its drawbacks such as slow turnaround time (of up to 12 weeks) as well as dependence on well-equipped laboratories, technical expertise and resources which may be lacking in less developed localities [12]. In our resource-constrained setting, MTB culture is not performed regularly, although its contribution would have been invaluable in diagnostics.
Literature shows that anti-TB therapy alone may suffice in splenic TB diagnosed without splenectomy [17, 28]. Patients with splenic abscess respond well to anti-TBs [29] and a treatment duration of 6 months, similar to that for other extrapulmonary sites is recommended. A few controlled trials recommend a 12-month course of anti-TB treatment for splenic TB [19]. However, there are few reports that show inadequate or absence of response to anti-TB therapy without splenectomy [8, 30]. In such cases, splenectomy is a viable option.
The patient’s worsening symptoms on initiation of anti-TB therapy was attributed to paradoxical reaction (PR). This is defined as the “worsening of existing lesions or presentation of new lesions during anti-TB therapy” [31], or the “worsening of clinical or radiological findings following the initiation of appropriate therapy” [32]. PR is typically associated with exaggerated inflammatory symptoms including fever, [33] lymphadenitis, [34] and pulmonary manifestations [33]—as illustrated in our case by worsening fever and pulmonary disease (pleural effusion with mediastinal lymphadenopathy). It has been suggested that rapid killing of bacilli with antibiotics may lead to the release of large amounts of microbial components, which stimulate an exuberant inflammatory response, [35] and that higher baseline numbers of bacilli may potentiate this process. It has also been postulated that this is essentially a hypersensitivity reaction to persistent mycobacterial antigen [34]. Ultimately, the PR phenomenon is likely to be missed, and in such cases TB may be labelled as a misdiagnosis warranting discontinuation of therapy. Fortunately, PR is self-limiting and does not always require steroid therapy [36]. Our patient improved with supportive management and continuous anti-TBs without steroids. In summary, the chronological course of events as pertains to this case has been represented in a timeline graphic (see Additional file 1).