CNS tuberculosis: an update · TB meningitis. Arnold Rich. 1893-1968. Bull John Hopkins Hosp. 1933;52:5 -37. Bacteraemia ↑Inflammation. DEATH (25%) ↑bacteria

Guy ThwaitesDirector, Oxford University Clinical Research Unit

Wellcome Trust Africa Asia ProgrammeViet Nam

CNS tuberculosis: an update

@Thwaitesguy @oucru_vietnam

TB meningitis

Arnold Rich1893-1968

Bull John Hopkins Hosp. 1933;52:5-37.

Bacteraemia ↑Inflammation

DEATH(25%)

↑bacteria

TB meningitis: pathology• Basal meningitis• Hydrocephalus• Infarcts• Tuberculomas

Time from start of treatment (days)

3002001000

Prop

ortio

n aliv

e

1.0

.9

.8

.7

.6

.5

.4

.3

.2

.1

.0

HIV negative

HIV positive

Log rank P<0.001

Why study TB meningitis?545 adults with TBMConservative estimate: 100,000 cases each year

TBM is a medical emergency

Treatment before the onset of coma is the greatest benefit a physician

can give a patient with TBM

The problem: rapid diagnosisClinical algorithms CSF ZN stain, microscopy, culture

GeneXpertCan we get to this?

Is a ZN stain/culture of the CSF useful?

Modern confirmation of the method• 132 consecutive adults

with TBM• AFB seen in CSF of 77

(58%)• AFB seen or cultured in 94

(71%)

J Clin Microbiol. 2004 Jan;42(1):378-9.

1 2 3 4 5

Volume of CSF examined (mls)

0

25

50

75

100

M.tb

isola

ted f

rom

CSF

(%)

40

5762

80 78

0-1.9 2.0-3.9 4-5.9 6-7.9 >8

Odds ratio CI P-value

Duration of symptoms (days) 1.05 1.00-1.10 0.050

Volume of CSF (ml) 1.36 1.06-1.75 0.017CSF neutrophils (% of white cell count)

1.03 1.01-1.05 0.008

CSF lactate (mmol/liter) 1.42 1.16-1.73 0.001CSF/blood glucose ratio 0.03 0.02-0.62 0.023

Predictors of positive bacteriology

Lancet Infect Dis. 2018 Jan;18(1):68-75

• 129 HIV infected adults, suspected TBM• 23 definite/probable TBM• Sensitivity:

Ultra: 70% (47-87)Xpert: 43% (23-66)Culture: 43% (22-66)

• Likelihood of positive test ++ if >6mls CSF tested

A randomised diagnostic study comparingson of GeneXpert Ultra MTB/RIF and GeneXpert MTB/RIF for the diagnosis of tuberculous meningitis

Ultra Xpert ZN smear MGIT culture

Reference standard: definite, probable and possible TBMPositive tests 25/53* 21/53 77/108 45/94Sensitivity (95% CI) 47·2%

(34·4-60·3%)

39·6%

(27·6-53·1%)

71·3%

(62·5-79.0%)

47·9%

(38·0-57·9%)Specificity (95% CI) 100%

(92·0-100%)

100%

(92·6-100%)

100%

(96·1-100%)

100%

(95·6-100%)Reference standard: definite and probable TBMPositive tests 25/43 21/43 77/88 45/75Sensitivity (95% CI) 58·1%

43·3-71·6%

48·8%

34·6-63·2%

87·5%

79·0-92·9%

60·0%

48·7-70·3%Specificity (95% CI) 100%

93·4-100%

100%

93·8-100%

100%

96·8-100%

100%

96·4-100%

Ultra Xpert

HIV negative HIV positive HIV negative HIV positiveReference standard: definite and probable TBMPositive tests 14/29 9/11 8/27 10/12Sensitivity (95% CI)

48·3%31·4-65·6%

81·8%52·3-94·9%

29·6%15·9-48·5%

83·3%55·2-95·3%

TBM management

Enhance bacterial killing

Control intra-cerebral inflammation

INTENSIVE CARE

Intra-cerebral drug penetration

INH

PZARifampicinETHSM

Blood-brain barrier

New agentse.g. bedaquilinedelaminid

?

LevofloxacinMoxifloxacin

Lancet Infect Dis. 2013 Jan;13(1):27-35.

60 Indonesian Adults. Oral rifampicin (450mg) vs IV 600mg for 1st 2 weeks

N Engl J Med. Jan 2016;374:124-134.

817 Vietnamese adults. Standard regimen vs rifampicin 15mg/kg + levofloxacin (1g/day)For 1st 2 months

PK/PD analysis from the trial

• No relationship between rifampicin exposure and outcome

• Strong independent relationship between isoniazid exposure and survival

• 38 deaths; 28 were fast metabolisers

Rifampicin PK Standard therapy Intensified therapy Day 14 CMAX plasma (mg/L) 10.6 (2.8-21.6) 18.2 (0.9-41.8)Day 14 AUC0-24 plasma (h∙mg/L) 48.2 (18.2-93.8) 82.5 (8.7-161.0)Day 14 CMAX CSF (μg/L) 189.3 (64.9-566.6) 330.8 (35.1-828.8)Day 14 AUC0-24 CSF (h∙mg/L) 3.5 (1.2-9.6) 6.0 (0.7-15.1)

Controlling intracerebral inflammation: adjunctive dexamethasone

N Engl J Med. 2004;351(17):1741-51

545 Vietnamese adults. 6-8 week dexamethasone vs placebo

How does dexamethasone save lives?

Infarcts

Hydrocephalus

Thwaites et al. Lancet Neurol. 2007Simmons et al. J Immunol. 2007

Green J et al. PLOS One. 2009

LTA4H Genotype determines survival and dexamethasone responsiveness

CC genotype (LTA4H low) CT genotype TT genotype

Survival in 182 HIV uninfected Vietnamese adults with TBM treated with or withoutAdjunctive dexamethasone

Dex Dex

Dex

Cell. 2012; 148, 434–446

Homozygotes

P<0.001

Heterozygotes

J Infect Dis. 2017 Apr 1; 215(7): 1020–1028.

All patients(n=764)

HIV-uninfected(n=439)

HIV-infected(n=325)

TTCT

CC

Will more directed, ‘intelligent’ host-directed therapies improve outcome?

- Aspirin- Thalidomide- Anti-TNF biologicals- Interferon-gamma- Developing list of ‘rational’ candidates

Sir John Vane FRSNobel Acceptance1982

Aspirin

J Neurol Sci. 2010 Jun 15;293(1-2):12-7. 2010.

P=0.03 P=0.18

J Child Neurol. 2011 Aug;26(8):956-62 J Neurol Sci. 2010 Jun 15;293(1-2):12-7

Mechanism of action

↑Resolvins↑Protectins↑Maresins↑Lipoxins

Hypotheses

Low dose aspirin (<200mg/day)

Inhibits TXA2 ↓ infarcts

↑ survivalHigh dose aspirin (>600mg/day)

Anti-inflammatoryPro-resolving

↓ inflammation

Randomised double blind placebo-controlled trial of aspirin 81mg or 1000mg for TBM

Elife. 2018 Feb 27;7. pii: e33478.

Safety (ITT population)

Placebo 81mg 1000mg Placebo 81mg 1000mg Placebo 81mg 1000mg

Cerebral bleed Any GI bleed Either

ITT, PP and Planned sub-group analysis Placebo vs aspirin 81mg

Note: P-value for interaction between diagnostic criteria and outcome. P=0.01

ITT, PP, and Planned sub-group analysisPlacebo vs 1000mg aspirin

Efficacy in those with definite TBM

P=0.03

New infarcts by day 60 Deaths by day 60 Deaths or infarcts

Dose-dependent impact of aspirin on CSF inflammatory mediators

Lancet Infect Dis. 2018 Jan 23. pii: S1473-3099.

Distinct TBM metabolome

CSF tryptophan concentrations predicts survival and are controlled by 11 genetic loci

Critical care and TBM

Sodium/hyponatraemia

Glucose/diabetes

Temperature

Mechanical VentilationNutrition Rehabilitation

Current and future clinical research priorities

Prevention• Vaccine: Adult/HIV-infected

Diagnosis• High sensitivity; resistance detection

Treatment• PK optimised regimens; old and new drugs• Can we improve upon corticosteroids?• Precision/personalised anti-inflammatory therapy?• Evidence-based critical care

OUCRU Vietnam TB group

Nguyen Thuy Thuong ThuongLalli Ramakrishnan

Jes DalliReinout van Crevel and friends in Bandung

Raph Hamers, Darma Imran and friends in Jakarata