Diagnosis of tuberculosis in children H Simon Schaaf Desmond Tutu TB Centre Department of Paediatrics and Child Health, Stellenbosch University, and Tygerberg Children s Hospital (TCH)
Estimated TB incidence rates 2014 WHO Global TB Report 2015
Epidemiology ~320 000 notified TB cases in South Africa/year (0.6-0.7% of population) Children make up 15-20% of the national TB burden The MDR-TB estimate (overall) for 2012-2014 was of 2.8%; (95%CI: 2.0%-3.6%); remained stable from previous survey in 2001-02 of 2.9% (95% CI: 2.4%-3.5%). However, increase in RIF resistance as well as INH mono-resistance Of concern is a rising trend in both RIFmonoresistant and INH-monoresistant TB
Is diagnosing TB in children difficult? Case 1: 3-year-old, coughing 3 weeks, LOW, fever Father recently diagnosed with smear-positive PTB, living in same house
Is diagnosing TB in children difficult? Case 2: Baby-girl 14 weeks cough >4 weeks. Coughing bouts turns red says mom. Was seen at clinic several times. Amoxicillin given. Mom told to stop smoking Fever for last two weeks. Unbooked mom: Baby born at home. Breastfed since birth, but on clinic card not thriving. No BCG given (out of stock) No-one known with TB at home. Mom says she has a normal cough O/E Decreased air entry LUL; hepatosplenomegaly Irritable baby, no neck stiffness or bulging fontanelle, bilateral femoral hernias and low grade fever
Case from dr S. Mukherjee Case 3: 7-year-old male, progressive swelling both lower limbs (pedal oedema), fatigue and occasional fever. HIVinfected on ART x 6 months. No history of TB contact. Severe pallor. No heart problem. No history of any decreased urine output or history of jaundice. Abdomen: hepatomegaly 7cm (non-tender, firm consistency, surface smooth, margins sharp) and splenomegaly of 4cm. Abdomen soft, non-tender. No ascitis. Generalized lymph-adenopathy. Mantoux TST 12mm. No AFB was isolated from the sputum or the gastric aspirate. Chest X-ray was normal. Hemoglobin of 5g/dL, WCC 7000 (N=55; L=42), LFT and renal function test was normal. U/S of abdomen - multiple intra-abdominal lymph nodes (mesenteric, para-aortic, retroperitoneal lymph nodes)
Answers to cases Case 1 = Hilar nodes, culture-confirmed TB Case 2 = Culture-confirmed TB. Mother cavitary TB 3+ smear positive INFECTION RISK in ward! Case 3 = A large lymph node (20mmx17mm) was compressing the IVC. Biopsy was taken from a retroperitoneal lymph node. It revealed histological features of non-hodgkin s lymphoma.
Conventional approach to TB diagnosis A constellation of the following: History of chronic symptoms and TB contact Clinical examination (incl. growth assessment) Tuberculin skin testing (or IGRAs?) Chest radiography Bacteriological confirmation Histology (especially EPTB) HIV testing (high prevalence areas or patients at risk) (Scoring systems and diagnostic algorithms)
Age-related risk 50 40 Immune compromised % 30 20 PTB Disseminated 10 0 <1 1to2 2to5 5to10 10to15 Age in years
History of symptoms of chronic disease Symptoms may overlap with other diseases, but welldefined symptoms give improved yield: - Chronic cough = unremitting cough >3 weeks - Fever = >7-14 days after excluding common causes - Weight loss or FTT = preferably documented on growth chart (RTHC) - Fatigue (tiredness) not keeping up with others In some children TB presents as an acute pneumonia; both in HIV-infected and -uninfected
Symptom characteristics 1) Acute with delayed recovery 2) Recurrent Intensity of Cough 3) Persistent, non-remittent 1w 2w 3w 4w Duration in weeks
History of contact A close contact = living in the same household (or in frequent contact) with source case. Sputum smear-positive TB case > infectious than smear-negative source cases, but still infectious! Screen all children (especially <5 years or HIVinfected) in HH contact with PTB cases for TB Contact with source case is found in only 40-70% of cases. May be infected outside of household Often undiagnosed or other TB cases in the family - in infants, may be worthwhile to screen mother Find out about DST of adult source cases!
Clinical examination Pulmonary TB: No specific features on clinical examination to confirm that presenting illness is due to pulmonary TB Physical signs highly suggestive of EPTB: Gibbus, especially recent onset (spinal TB) Non-painful cervical lymphadenopathy (with/without fistula formation) Meningitis: with sub-acute onset (or not responding to AB) Pleural effusion older children Distended abdomen with ascites or abdominal masses Signs of tuberculin hypersensitivity (phlyctenular conjunctivitis, erythema nodosum, PNT) Painful joint(s)
Tuberculin skin test (TST) Useful to identify children infected with TB Mantoux TST the recommended test Read in mm induration after 48-72 hours. Regarded as positive as follows: High-risk children: 5 mm (HIV-infected or severely malnourished children) All other children: 10 mm Negative TST does not rule out TB
Interferon-gamma release assays Based on release of IFN-γ produced by T-cells Two blood tests commercially available - T-Spot.TB (ELISPOT) - Quantiferon-TB Gold In-tube test Diagnose infection and not disease (as TST) More specific than TST not affected by BCG IGRAs possibly more sensitive than TST, but this has not been confirmed in high-burden areas Time to lab / amount of blood / cost are limitations May be specific indications for use such as young children (infants) after BCG
Chest Radiography Remains an important tool in diagnosis although: Quality of radiographs is important Chest radiographs often misread in children with TB. Intra- and inter-observer error for hilar lymphadenopathy common Suggestive pictures Enlarged hilar and/or mediastinal lymph nodes with or without (persistent) opacification in the lung Ghon focus/complex (uncommon) Miliary pattern in HIV-uninfected children Adolescents: adult type PTB or pleural effusion
HIV/TB: Chest Radiography Basic features the same. Often more difficult to interpret: HIV-associated lung conditions look similar to TB, e.g. bronchiectasis, recurrent pneumonias, LIP TB can occur concurrently with other diseases Reticulonodular (miliary) picture ± adenopathy could be miliary TB, LIP or other conditions Reports: (Madhi IJTLD 2000, Hesseling IJTLD 2002) more cavitation in HIV+ TB more miliary TB in HIV+ TB Most children with TB are not HIV-infected!
Hilar nodes / thymus?
Pulmonary vasculature
Subcarinal lymphnodes
Is this TB or lobar pneumonia?
Mother s CXR
Other imaging (remember EPTB) Ultrasound: Abdominal TB, pleural and pericardial effusions CT scans: TBM, chest mainly MRI scans: Spinal TB, TBM Bronchoscopy: lymph nodes and its effects
Airways compression
HIV-infected vs. non-hiv-infected There are still many more HIV-uninfected children with TB than HIV-infected Treat HIV-infected children as a high-risk group (infection is as important as disease). As immune status deteriorates (lower CD4%), risk for TB increases Early HAART will reduce risk between 3-10 times, but risk remains higher than in HIVuninfected children
Point Scoring System Feature 0 1 2 3 4 Score GENERAL Weeks ill < 2 2 4 > 4 WFA > 80% 60 80% < 60% Family history Tuberculin Malnutrition Fever LOCAL None Reported Sputum + No response Positive Not improving Lymph ad. Bone / joint Abd. mass CNS sg / CSF Gibbus TOTAL
Scoring systems or diagnostic approaches Critical review of these approaches shows that few have been tested and sensitivity and specificity has not been calculated (A Hesseling et al. IJTLD 2002;6:1038-45 ) In an area with high HIV prevalence the specifity was 25% (95% CI 16-37%) (P van Rheenen. Trop Med Int Health 2002;7:435-41) In developing countries scoring systems is all that is available stepwise approach more a screening tool than a diagnostic tool
Contact management algorithm Child close contact of infectious PTB case 0-59 months or HIV+ >60 months & HIV- Well Symptomatic Symptomatic Well 6H Consider TB* No treatment If becomes symptomatic If becomes symptomatic *Follow guidelines for diagnosis
How to investigate contacts Clinical assessment: History (Symptoms; closeness and duration of contact; drug resistance) Clinical examination Clinical assessment alone is sufficient to decide whether contact is well or symptomatic If available: TST (exposure prophylax even if TST negative) CXR (for diagnosis of disease)
Why do we need microbiological confirmation? The majority of child TB cases are diagnosed at primary care level without microbiological confirmation, as specimens from children are difficult to obtain and often those children have primary (paucibacillary) TB In >80% of culture-confirmed cases in hospital the diagnosis of TB was made before culture result was available If DST of adult source case is known, child contact should be treated according to adult isolate s DST result, as concordance between source case and child s isolates is between 78-90% in different studies
Why do we need microbiological confirmation? To confirm TB in difficult cases, e.g. uncertain lung pathology, HIV-infected children, extrapulmonary TB To determine drug susceptibility in children with unknown source cases, especially if they have poor response to first-line treatment To confirm drug resistance if a source case with DR-TB is identified in our experience 10-20% not the same DST pattern as source case, either because different source case or infected before amplification of resistance in the source case
Culture for M. tuberculosis Only 5-10% smear-positive yield in children Cultures positive in 30-40% of hospital-based cases; lower in community-based studies. Respiratory samples in children: Induced sputum ~ gastric aspirate NPA, tracheal aspirates or BAL FNA biopsies are useful for diagnosis of EPTB Any other body fluid/biopsy of tissue suspected of TB (e.g. CSF, bone/sinovial biopsy)
Robert Koch 1843-1910 Discovered M. Tuberculosis 1882
GeneXpert MTB/RIF GXP PCR-based diagnosis of both M.tuberculosis complex as well as RIF resistance. Becoming more of point-of-care test and replacing sputum smear microscopy, but culture still needs to be done to confirm results and, in case of RIF resistance, to confirm MDR-TB and do second-line DST Mainly done on sputum samples, but several studies have shown its value also with other specimens, e.g. GA, FNA from lymph nodes, CSF and even stool Xpert MTB/RIF Ultra? Improved yield?
From LJ medium to MGIT Automated culture system
DST in children Culture & DST takes longer but provides best yield (30-70% in symptomatic children). Phenotypic or genotypic DST can be done, the latter providing more rapid results Xpert MTB/RIF should in children be followed by culture and DST, because children usually have smear-negative disease, and only ~60% of smear-neg, culture-pos cases will be identified. Xpert MTB/RIF also does not provide further DST results other than RIF (currently mainly sputum used) With increasing RIF-monoresistant TB cases in adults and Xpert MTB/RIFresults only, managing child contacts becomes a problem
Line Probe Assays Line-probe assays, a family of DNA strip-based tests that use PCR and reverse hybridization methods for the rapid detection of mutations associated with drug resistance, are available as commercial kits. Confirms M. tuberculosis complex and provides drug susceptibility test (DST) results for INH and RIF GenoType MTBDRplus INH and RIF mutations GenoType MTBDRsl second-line drugs now approved, but still need phenotypic DST Problems: Need laboratory set-up, risk of cross contamination Advantage: identifies the mutation conferring resistance to INH. This could assist with choosing the correct drugs, e.g. high-dose INH vs ethionamide
Line Probe Assay result
Conclusion of present data There is very little new data that has dramatically changed our ability to diagnose TB in children, but microbiological confirmation has become more rapid The more proof, the more certain the diagnosis Scoring systems, despite its shortcomings may still have a role to play as a screening tool in certain settings Younger children are at greater risk for complications e.g. disseminated TB and TBM, therefore treat earlier Role of culture/dst confirming diagnosis and identification & confirmation of drug resistance