Call for ultrasound in paediatric tuberculosis work‐up: A case report from Germany

Diagnosing childhood tuberculosis (TB) remains challenging. Microbiologic tests have low sensitivity due to the paucibacillary nature of paediatric disease and the difficulty in obtaining adequate samples; confirmation of TB disease is therefore often not achieved in children.1 Chest radiography (CXR) is the standard imaging modality but is limited by poor specificity and reader inconsistency in children.2 Extra-thoracic extra-pulmonary TB (EPTB) is not primarily investigated for in children despite a higher risk thereof. Focused ultrasound to detect human immunodeficiency virus (HIV)-associated EPTB is routinely applied in adults in settings with high HIV/TB prevalence to support the diagnosis of TB where confirmation is restrained by limited diagnostic infrastructure or impeding host conditions such as HIV-infection.3 Data from South African children with presumptive pulmonary tuberculosis (PTB) showed that a positive TB focused ultrasound correlated with PTB.4 Abdominal lymphadenopathy and focal splenic lesions (micro-abscesses) correlated with HIV/TB co-infection but were also prevalent in HIV-negative children with PTB.4 Previous ultrasound studies in South African children with PTB had also reported that related abdominal lymphadenopathy and focal splenic lesions were frequent5 and a recent multi-centre study from Asian and African countries suggested a diagnostic algorithm including ultrasonography for abdominal lymph nodes to guide antituberculosis treatment decision in HIV-infected children with suspected TB.6 To date the potential diagnostic yield of systematically investigating children with presumptive TB for (concurrent) EPTB in addition to PTB has, however, not been studied in settings with low HIV/TB prevalence and ultrasound is commonly not systematically performed during TB work-up. Here we report on a paediatric case from Germany where TB focused abdominal ultrasound during the diagnostic TB work-up was the determining diagnostic tool to establish a timely diagnosis of active TB. The boy's parents consented to publication of his case. A 20-month-old boy was referred to the Department of Paediatrics at a tertiary care centre in Germany, to rule out active TB due to a tuberculin skin test (TST) conversion despite 3 months of anti-tuberculous prophylaxis. The boy's family had immigrated from Somalia to Germany some years before. The child's HIV-negative mother had been diagnosed with drug-sensitive PTB 4 months before. At contact screening 4 weeks after the mother's diagnosis, the boy was asymptomatic and a TST was negative. The child was started on a two-drug TB prophylaxis. Follow-up TST after 3 months of prophylaxis showed an ulcerating induration exceeding 10 mm and diagnostic TB workup was initiated. TB-related symptoms were absent and clinical examination revealed no pathology. The patient was on the 21st weight-related percentile. Routine laboratory analyses were normal. The parents reported good compliance with TB prophylaxis. First-line diagnostic TB work-up consisted of interferon-gamma-release assay (positive), anterior–posterior CXR (no TB-specific findings) and three morning gastric aspirates examined by microscopy (Ziehl-Neelsen stain), polymerase chain reaction (PCR) and liquid culture. No acid-fast bacilli were detected by microscopy, but two out of three PCR tests detected faint traces of nucleic acid of Mycobacterium tuberculosis complex. The PCR results were communicated as possibly false-positive by the microbiology department. Given the inconclusive results of first-line tests, a TB focused abdominal bedside ultrasound was performed by the clinician in charge of the patient to assess for abdominal lymphadenopathy and focal splenic lesions suggestive of active EPTB. The clinician performing the bedside ultrasound initiated this point-of-care examination because of his personal preceding positive experience with TB focused ultrasound in presumptive TB patients in India.7 Ultrasound revealed multiple hypoechoic splenic lesions sized 5–15 mm characteristic of TB micro-abscesses (Fig. 1a). Consecutively, the diagnosis of concurrent PTB and EPTB was established and a four-drug anti-tuberculous regimen with rifampicin, isoniazid, pyrazinamide and ethambutol was initiated. The patient was discharged 14 days later in a good condition. Several weeks later, liquid culture grew drug-sensitive M. tuberculosis. At follow-up 2 and 4 months after treatment initiation, the patient continued to be asymptomatic and in good clinical condition; compliance was reported good and routine laboratory blood tests remained normal, as did a repeat CXR. Medication was switched to a two-drug regimen (rifampicin, isoniazid) only after 4 months, since after 2 months of treatment no reduction of splenic lesions was seen on follow-up ultrasound. Treatment was extended beyond 6 months in view of persistent splenic lesions. After a 12-month course of anti-tuberculous treatment, the patient had reached the 51st weight-related percentile, the splenic lesions had disappeared (Fig. 1b) and TB treatment was stopped. Our case reflects the challenges of establishing a diagnosis of TB in children. First-line diagnostic tests were inconclusive and clinical presentation did not suggest active TB. In our case, the additional diagnostic value of TB focused abdominal ultrasound was decisive in establishing the diagnosis of active TB at the time of presentation and hence several weeks before cultures confirmed the diagnosis. This allowed a timely initiation of TB treatment which is important to reduce the risk of progression to severe disease and of disease transmission. Moreover, sonographic detection of spleen involvement allowed an accurate delineation of disease extent (PTB and concurrent EPTB). Per German national paediatric guidelines this prompted a four-drug regimen in the intensive phase as opposed to a three-drug regimen recommended for smear negative uncomplicated PTB in children.1 During follow-up ultrasound was an effective monitoring tool. Since the child was asymptomatic and CXR was not suggestive of TB, neither resolution of symptoms nor CXR findings could serve as marker for treatment response. Microbiologic follow-up was not performed. Resolution of sonographic spleen findings after 12 months of TB treatment indicated therapy success and allowed for discontinuation of medication. More data are needed to better understand the evolution and resolution of sonographic TB findings during anti-tuberculous treatment. Sonographic features compatible with TB are not confirmatory and need to be interpreted within the clinical and epidemiological context. Differential diagnoses for abdominal lymphadenopathy and splenic micro-abscesses in children are broad and include many different infectious diseases (e.g. brucellosis, bartonellosis) as well as non-communicable diseases (e.g. lymphoma). As pointed out above, evidence on the utility of focused ultrasound for EPTB is reported from settings with moderate and high TB/HIV prevalence.4, 6, 7 Recently the utility of lung ultrasound for PTB has also been reported and discussed for children and adults8-12; however, more data are needed to better understand the diagnostic potential of lung ultrasound in discriminating PTB from other respiratory infections. The utility of systematic ultrasound investigation in settings with low TB/HIV-prevalence has not been studied yet. Irrespective of TB endemicity children have a higher risk for extra-thoracic TB even in the absence of HIV-infection and may therefore benefit from abdominal ultrasound during TB work-up. Depending on the setting, ultrasound may be performed either as a focused point-of-care examination by the treating clinician or as formal ultrasound examination by the radiology department. In settings with a low TB case load, formal ultrasound should be performed by radiologists or equally trained sonographers. In institutions with regular TB cases, however, for example institutions with specialised TB or infectious disease clinics, clinicians can quickly acquire sufficient expertise in TB focused ultrasound which will provide them with a point-of-care tool to improve timely TB diagnosis and follow-up. In summary, available evidence and our case report suggest that the diagnostic utility of (TB focused) ultrasound as a standard diagnostic examination during paediatric and adult TB work-up should also be investigated in settings with an overall low TB burden. S Bélard is currently a participant in the BIH-Charité Clinician Scientist Program funded by Charité-Universitätsmedizin Berlin and the Berlin Institute of Health.