2. Paediatric tuberculosis
Beate Kampmann FRCPCH PhD
A/Professor in Paediatric Infection & Immunity
Consultant Paediatrician
Imperial College London, UK
and
Institute of Infectious Diseases and Molecular Medicine
University of Cape Town, RSA

3. Overview What is special about TB in children?
Epidemiology- who are our patients?
Clinical presentations
How can we make the diagnosis?
New Immunological Tools-how helpful are they?
Issues in TB therapy in children
Future research topics

4. Tuberculosis in Children…. the problem
Significant Morbidity and Mortality
1.4 million cases annually (95% developing countries)
450,000 Deaths
estimated 10-15% of global burden related to childhood TB
Different clinical spectrum of disease
5-10% < 2 yr meningitis
disseminated disease more common
Co infection with HIV- clinically very difficult to distinguish
Remains a diagnostic challenge
paucibacillary, rarely culture confirmed :
Sputum smear positive in 10.3% (10-14yr), 1.8% (5-9) and1.6% (<5)
Cultures positive 21% (10-14), 5% (5-9) and 4.2% (<5),

5. Tuberculosis in Children differs from adults
Immune responses are
Age-dependent: Following infection 40% < 2 yr, 25% 2-5 yr and 5-15% of older children will develop disease within 2 years
Majority of disease results from progression of primary infection rather than reactivation
might affect detectable immune responses
More likely to be extrapulmonary and disseminated, particularly in infants
Newton, Kampmann The Lancet Infectious Diseases, August 2008; Vol 8:

6. Percentage of TB cases of foreign origin, 2006
Andorra
Malta
Monaco
San Marino
Not included or not reporting to EuroTB
0% – 4%
5% – 19%
20% – 49%
> 49%
Trends in incidence of TB in children under 15 years
by ethnic group in London,

7. UK: Tuberculosis rates in persons born abroad by age
Development of TB
in immigrant children
Sources: Enhanced Tuberculosis Surveillance, Labour Force Survey population estimates, Abubakar et al Arch. Dis. Child. 2008;93; ;

8. Children with TB at Imperial HCT:
Ethnicity and country of birth:

9. Children acquire TB from (household) contacts
If you ask yourself, does this child have TB,
ask yourself:
is there TB in this family?
Or: if you see adults with TB, ask yourself if they have children

10. Presentation of PAEDIATRIC TB
Case 1
-14 month Asian girl
-previously well, no F/H of TB
-3 weeks cough and unwell
-admitted to local hospital
-low grade fever
-normal chest examination
WHAT INVESTIGATIONS WOULD YOU DO?

12. PAEDIATRIC TB Gastric washings;
Case 1
Mantoux test: 2mm
Gastric washings;
- microscopy and (later) culture negative
-Rx. Erythromycin and Augmentin
-no improvement on antibiotics
-bronchoscopy planned

13. PAEDIATRIC TB - afebrile, less cough, looking well
Case 1
1 day before bronchoscopy:
- afebrile, less cough, looking well
-continued improvement,
- discharged home, no ∆

14. PAEDIATRIC TB out-patient review 6 weeks later:
Case 1
out-patient review 6 weeks later:
-completely well, thriving, no cough
“Grandfather admitted to local hospital with pulmonary TB”
-repeat Mantoux: now 25mm
-TB treatment commenced

15. PAEDIATRIC TB Primary TB in children:
Case 1
Discussion Points
Primary TB in children:
- spontaneous recovery is possible
- diagnosis is difficult
- no visible AFB
- cultures usually negative
- tuberculin test negative
- F/H is often the clue to diagnosis

16. LATE REACTIVATION OF PULMONARY DISEASE
CHILDHOOD EXPOSURE
PRIMARY
PULMONARY
INFECTION
WELL
ADULT
IMMUNITY
(live MTB)
Successful
immune
response
LATE REACTIVATION OF PULMONARY DISEASE
FORMS
CAVITY

17. PRIMARY PULMONARY INFECTION PROGRESSIVE DISEASE Lympho/ haematogenous
CHILDHOOD EXPOSURE
PRIMARY
PULMONARY
INFECTION
PROGRESSIVE
DISEASE
MILIARY TB or
EXTRA-PULMONARY
Inadequate
immune
response
Lympho/
haematogenous
spread
Self healing??

18. PRIMARY PULMONARY INFECTION PROGRESSIVE DISEASE Lympho/ haematogenous
CHILDHOOD EXPOSURE
PRIMARY
PULMONARY
INFECTION
PROGRESSIVE
DISEASE
MILLIARY,
EXTRA-PULMONARY
TB MENINGITIS
Inadequate
immune
response
Lympho/
haematogenous
spread
-most serious form of TB, fatal if untreated
-most common in < 2 year-olds
-worst prognosis in < 2 year-olds
-insidious onset;

19. inflammatory response
TB MENINGITIS
Primary focus
in lung
Brain focus
(Rich focus)
Meninges
CSF
Severe granulomatous
inflammatory response
VI NERVE
PALSY
BRAIN
INFARCTION
CSF protein 
 ICP
Thick gelatinous
exudate forms,
envelopes base
of brain
Cranial nerve palsies
Vascular occlusions
Hydrocephalus

20. TB MENINGITIS
CSF
-lymphocytes, low sugar, high protein, AFB may be visible
-but often -polymorphs initially
-protein normal initially
-no visible organisms
-sugar normal initially
So: repeat LP, neuro-imaging, CXR, contacts

21. Ring enhancing
tuberculomata
CT scan; enhanced

22. MRI > SENSITIVITY THAN CT
ENHANCED CT SCAN GADALLINIUM ENHANCED MRI
MRI > SENSITIVITY THAN CT

23. HYDROCEPHALUS

24. SUCCESS of Rx DEPENDS ON
TB MENINGITIS
SUCCESS of Rx DEPENDS ON
EARLY DIAGNOSIS

25. TB MENINGITIS TREATMENT with quadruple therapy Drugs:
-? CSF penetration
- duration
- sensitivity
Adjunctive therapy:
- steroids
- SIADH
- acetazolamide
- surgery

26. Diagnostic tests
Microbiological
Organism
smear culture DNA

27. The “gold-standard”
Appearance in sputum
Appearance in culture
‘cording’

28. PAEDIATRIC TB: Implications of bacterial load
Paediatric TB: 106 bacteria Adult TB: >109 bacteria
- children less infectious
- difficulty in confirming diagnosis (< 30%)
- difficulty in detecting resistance
PAEDIATRIC TB: Implications of bacterial load

29. skin test antigen-specific
Diagnostic tests
Microbiological
Organism
smear culture DNA
Immunological
Host response
skin test antigen-specific
production of IFNγ

30. IGRA Acknowledgement & Thanks IGRA and the diagnosis of active TB
Travel
Signs and
symptoms
Contact history
Active TB
Radiology
Microbiology
TST
IGRA

31. TUBERCULIN SKIN TEST (used since 1890)
-technically difficult in children
-UK : 2 units of SSI tuberculin (PPD)
- > 200 antigens,
Read-out:
-degree of hypersensitivity to PPD
TUBERCULIN SKIN TEST (used since 1890)

32. *In children a negative TST does not exclude TB
Problems with the TST:
-poor specificity,does not distinguish between;
-TB disease
-TB infection
-BCG
-non-typical mycobacteria
poor sensitivity, can be falsely negative in;
-early infection
-disseminated disease
-severe malnutrition
-other acute infections (measles, pertussis)
-live vaccines
-immunocompromised (HIV)
*In children a negative TST does not exclude TB

33. T cell tests (interferon-g) that distinguish M. tuberculosis
Gene deletions and the origin of BCG
major antigens
ESAT6 and CFP10
M. tuberculosis
10 deletions
64 genes
RD1 region
M. bovis
4/5 deletions
30/40 genes
T cell tests (interferon-g)
that distinguish M. tuberculosis
infection from BCG vaccination
BCG substrains

34. IFN-γ responses to specific antigens by ELISPOT or Whole Blood Assay
Unable to distinguish between TB and BCG effect
Clear difference between acute TB and BCG vaccination
Van Pinxteren Clin Diagn Lab Immunol 2000

35. IGRA and National TB guidelines
UK: NICE Guidelines 2006
So,
Children with evidence of TB infection should receive chemoprophylaxis
But how do we establish this?
We used to do it with the TST which can diagnose infection with mycobacteria.
Unfortunately, the TST lacks specificity, as BCG vaccination can give a false-positive result
We now have blood based assays, which appear specific for infection with M.TB and measure the immunological footprint of M.Tb infection in the form of secreted interferon-gamma- the so-called IGRA.
The Nice guidelines in 2006 recommended the use of IGRA to guide the assignment of chemoprophylaxis, as seen in this algorithm.
Only children with positive IGRA should now receive prophylaxis

36. 2 commercially available assays
Antigens used:
ESAT-6
CFP10 +/- TB7.7
mitogen
negative control
In principal: can both distinguish
between BCG vaccination and
M.tuberculosis infection
But:
Paucity of data in children
Confusion about use of IGRA

37. Principal of Quantiferon Gold

38. ELISPOT assay In vitro blood test: Coating antibody PBMC+antigen
IFN-γ production
Biotinylated 2nd antibody
Avidin-peroxidase
Each spot is an individual T cell that has released IFNγ

39. Spot the Difference Interferon-γ release assays (IGRA)
IGRA versus TST: our own research
Interferon-γ release assays (IGRA)
in paediatric active and latent tuberculosis in London
- a side-by-side comparison with TST
Kampmann B, Whittaker E, Williams A, Walters S,
Gordon A, Martinez-Alier N, Williams B, Crook AM,
Hutton AM, Anderson ST.
Interferon- gamma release assays do not identify
more children with active TB than TST.
Eur Respir J Jun;33(6):1374-8

40. IGRA missed between 20-40% of definite active TB
Acknowledgement & Thanks
IGRA and the diagnosis of active TB
IGRA missed between 20-40% of definite active TB

41. Acknowledgement & Thanks
Combining IGRA and TST in the diagnosis of active TB
A combination of TST and IGRA increases sensitivity to above 93%

42. Acknowledgement & Thanks
IGRA and the diagnosis of active TB- other studies
Bamford et al, Arch Dis Child 2009 Oct 8 (Epub ahead of print)
UK-wide study of 333 children from 6 large UK centers, 49 with culture-confirmed TB:
TST had a sensitivity of 82%, Quantiferon-Gold in tube (QFT-IT) had a sensitivity of 78% and T-Spot.TB of 66%.
Neither IGRA performed significantly better than a TST with a cut-off of 15 mm. Combining results of TST and IGRA increased the sensitivity to 96% for TST plus T-Spot.TB and 91% for TST plus QFG-IT in the definite TB cohort.
Nicol et al; Pediatrics 2009,Jan; 123(1): 38-43
Comparison of T-SPOT.TB assay and TST for the evaluation of young children at high risk for tuberculosis
Sensitivity of the T-SPOT.TB was no better than that of the tuberculin skin test (>10mm) for culture-confirmed tuberculosis (n=10) (50% T-Spot and 80% TST) and was poorer for the combined group of culture-confirmed and clinically probable tuberculosis (n=58)
(40% T-Spot and 52% TST).
Bianchi et al, PIDJ 2009, Jun;28(6):510-4
IGRA was positive in 15 of 16 (93.8%) children with active pulmonary TB
Connell et al, Thorax 2006, Jul;61(7):616-20
The whole blood IFN-gamma assay was positive in all 9 children (100%) with TB disease.

43. A negative IGRA does not exclude active TB
Acknowledgement & Thanks
IGRA and the diagnosis of active TB
A negative IGRA does not exclude active TB
IGRA is not a “rule-out”- test,
but can add value to additional investigations

44. IGRA Acknowledgement & Thanks IGRA and the diagnosis of latent TB
Travel/Immigration
Contact history
Absence of clinical
signs and symptoms
Latent TB
negative
Radiology
TST
IGRA

45. IGRA and the diagnosis of latent TB
No “gold standard” for LTBI
Acknowledged discrepancy of TST and IGRA results
- due to poor specificity of TST
(Kampmann ERJ 2009, Connell PlosOne 2008, Bianchi PIDJ 2009)
Which IGRA is better?
- Good agreement between 2 IGRAS (92%, k=0.82)
(similar to Connell et al, PLoS One Jul:
agreement between QFT-IT and T-SPOT.TB 93%, k=0.83).
Currently: ? “over-treatment” by paediatricians
but: to date no studies of negative predictive value of IGRA in children
Performance in very young children- conflicting messages
Increased sensitivity in immuno-compromised hosts

46. Conclusion 2: Latent TB Good agreement between 2 IGRAS (92%, k=0.82)
“over-treatment” by Paediatricians,
compared to NICE recommendations
How many children will develop TB if TST> 15, but untreated with chemopro as IGRA negative?
How many children have neg TST but would have had pos IGRA at screening
Longitudinal survey required

47. IGRA should not currently replace the TST in children
Conclusions
IGRA should not currently replace the TST in children
we should not forget the many additional challenging question in childhood TB
IGRA detect immune memory but do not confirm the
presence or absence of M. tuberculosis- active or latent
higher specificity than the TST
designed to test for evidence of TB infection, not TB disease
can be used as a rule-in test for active TB in children,
but not as a rule-out test
higher sensitivity in immunocompromised patients compared to TST
better microbiological diagnostics
better biomarkers than IFNγ
better vaccines
improved understanding of primary TB

48. TB treatment in children
Treatment regimens are adopted from adult schemes
Children respond very well to treatment, incl DOTS
Dosages need to be adjusted for weight:
Pharmakokinetics in children differ from adults
- INH mg/kg, rapid acetylators (1)
- Ethambutol mg/kg (2)
Problems with treating TB in the HIV-infected child on ART
1: Schaaf et al, Arch Dis Child 2005; 90:614
2: Donald et al, Int J Tuberc Lung Dis 2006; 10:1318

49. IF YOU DON’T TAKE THE DRUGS,
Drugs and ADHERENCE
IF YOU DON’T TAKE THE DRUGS,
THEY WON’T WORK

50. PAEDIATRIC TB POOR ADHERENCE Support -hospital TB clinic -community
-health care workers
-social services
-DOT (Directly Observed Therapy)
-accurate record of treatment
-successful treatment
-prevention of resistance
-different adult
-different location

51. Take Home messages:
Think of the diagnosis, especially in the epidemiological context
TB is a family disease
The diagnosis of active TB in children is based on a jigsaw of findings
IGRA can be an additional piece in the jigsaw, but a negative IGRA does not exclude active TB
TB therapy needs a lot of support for families

52. founded in April 2009, branch of TBNET
to date: members from 15 European
countries, incl. Eastern Europe
includes clinicians, epidemiologists and
laboratory scientists
Aims
enhance the understanding of the pediatric aspects of tuberculosis
facilitate collaborative research studies
for childhood TB in Europe
provide expert opinion through excellence in science and teaching
establish a better evidence base for diagnosis and treatment of TB
in children

53. Research questions
immunological mechanisms of age related susceptibility to TB
Epidemiology of childhood TB
MDR /XDR TB in children
Performance/evaluation of new diagnostics
New drugs
New vaccines

54. Thank you Any questions? E-mail: b.kampmann@imperial.ac.uk
Website: www1.imperial.ac.uk/medicine/people/b.kampmann/