Challenges of treating two
infections (HIV and TB):
the ART of HIV/TB management
William Burman MD
Denver Public Health
University of Colorado
Case 1
27 y/o man from Ethiopia,
admitted with cough, fevers,
and 20 lb. weight loss over
one month
Sputum - rare AFB
HIV-positive, CD4 - 18, viral
load > 1,000,000
Dramatic initial improvement
with IRZE
Case 1
1. Should antiretroviral therapy be started during TB
treatment?
2. When during TB therapy should antiretroviral
therapy be started?
3. What regimens can be used for co-treatment of HIV
and TB therapy?
Survival of persons with HIV-related TB in the
pre-HAART era – San Francisco
N Engl J Med 1991; 324: 289-94
Complicating factors: antiretroviral
therapy during TB therapy


Need for coordination between TB and HIV
treatment programs
Challenge of adherence to multidrug therapy
for both diseases

Overlapping drug toxicity profiles

Drug interactions

Immune reconstitution (paradoxical)
reactions
SAPiT: Starting Antiretroviral therapy
(ART) in three Points in TB
Primary Objective:
 To determine the optimal time to initiate ARVs in TB
patients
Inclusion Criteria:
 Smear pulmonary TB
 HIV positive with CD4 count < 500 cells/mm3
 Effective contraception (efavirenz)
Endpoint
 10 – all-cause mortality
Karim S, et al. N Engl J Med. 2010;362:697-706
Initiation of ART during vs. after TB
treatment: SAPIT
Abdool Karim S, et al. New Engl J Med 2010; 362: 697-706
Mortaltiy rate
(per 100 person-years)
Effects of timing of ART on mortality, by
baseline CD4 cell count: SAPIT
18
16
14
12
10
8
6
4
2
0
21/137
22/281
6/86
2/186
< 200
Integrated
200-500
Sequential
Abdool Karim S, et al. New Engl J Med 2010; 362: 697-706
Mortaltiy rate
(per 100 person-years)
Effects of timing of ART on mortality, by
baseline CD4 cell count: SAPIT
18
16
14
12
10
8
6
4
2
0
21/137
All patients with HIV-TB
should receive ART
during TB treatment
22/281
6/86
2/186
< 200
Integrated
200-500
Sequential
Abdool Karim S, et al. New Engl J Med 2010; 362: 697-706
Competing risks in the timing of ART
during TB treatment
Immediate (< 2 wks)
Early (2 months)
Benefits:
• ↓ risk of other OIs
Benefiits:
• ↓ risk of IRD
Risks:
• ↑ adverse effects
• ↑ incidence of IRD
Risks:
• ↑ incidence of OIs
• feasibility
Mortality
General schema for CAMELIA, STRIDE, and
integrated arms of SAPIT
Primary
endpoint
“Immediate ART” TB treatment
(within 2 weeks) ART
HIV+
TB
TB treatment
“Early ART”
(2-3 months)
ART
0
8
24
Study week
Key characteristics of trials of timing of
ART during TB treatment
Study
Setting
Key
enrollment
criteria
Median CD4
(IQR)
Primary
endpoint
Cambodia
Smear +,
CD4 < 200
25 (10 - 56)
Death
STRIDE
Multinational
Clinical TB,
CD4 < 250
77 (36 – 145)
AIDS or
death
SAPIT
South
Africa
Smear +,
CD4 < 500
150 (77 –
254)
AIDS or
death
CAMELIA
N Engl J Med 365; 2011; 1471-1501
Effect of ART timing on death (CAMELIA) or
death/AIDS (STRIDE, SAPIT)
18
16
14
12
10
8
6
4
2
0
34% ↓
p=0.004
19% ↓
p=0.45
11% ↓
p=0.73
CAMELIA
STRIDE
Immediate
N Engl J Med 2011; 1471-1501
SAPIT
Early
Relationship between median baseline CD4 count and
the effect of immediate ART on death (CAMELIA) or
death/AIDS (STRIDE, SAPIT)
35
160
P = 0.004
140
30
120
P = 0.45
25
100
20
80
15
P = 0.73
60
10
40
5
20
0
0
CAMELIA
N Engl J Med 2011; 1471-1501
STRIDE
SAPIT
Median baseline CD4 cell count
% decrease in death/AIDS
with immediate ART
40
Effects of ART timing on outcomes in
CAMELIA and patients with CD4 < 50 in
STRIDE and SAPIT
30
42% ↓
p=0.02
68% ↓
p=0.06
STRIDE
SAPIT
25
20
15
34% ↓
p=0.004
10
5
0
CAMELIA
Immediate
N Engl J Med 2011; 1471-1501
Early
Effects of ART timing on death/AIDS among
patients with CD4 > 50 in STRIDE and SAPIT
14
AIDS or death
12
p=0.67
10
p=0.34
8
6
4
2
0
STRIDE
Immediate
SAPIT
Early
N Engl J Med 2011; 1471-1501
IRD
Effects of ART timing on Immune Reconstitution
Disease among patients with CD4 > 50 in
STRIDE and SAPIT
18
16
14
12
10
8
6
4
2
0
p=0.02
p=0.009
STRIDE
Immediate
SAPIT
Early
N Engl J Med 2011; 1471-1501
Effect of ART timing on survival of patients
with TB meningitis




Median CD4 ~ 40 (16 – 100)
60% + CSF culture
KM survival estimates at 9
months
 35.2% in immediate arm
 40.3% in deferred arm
Hazard ratio 1.1 (95% CI
0.8 – 1.6), p = 0.52
Early ART
Immediate ART
Similar in per protocol
analysis
Török et al, 41st Union World Conference on Lung Health, Berlin Nov 2010
Percentage of patients
Effect of ART timing on risk of adverse
events in patients with TB meningitis
p = 0.04
100
90
80
70
60
50
40
30
20
10
0
Grade 3 or 4
Immediate
Grade 4
Early
Török et al, 41st Union World Conference on Lung Health, Berlin Nov 2010
Timing of ART in patients with TB

Advanced AIDS (CD4 < 50): immediate ART (within 2
weeks) improves survival




Markedly increased risk of IRIS, including fatal IRIS
events
Overall survival benefit despite IRIS
CD4 > 50: early ART (~ 2 months) provides good
balance of competing risks of death/AIDS vs. IRIS
Caveats

CNS involvement – no benefit to immediate therapy,
and there may be increased risk (Clin Infect Dis.
2011;52:1374-83)

Programmatic complexities of early ART
Programmatic challenges of immediate
ART during TB treatment

Rapid HIV diagnosis

Rapid provisional diagnosis of TB



Rapid way to identify those in need of
immediate ART: CD4 cell count, BMI, clinical
status
ART available in settings where TB is
diagnosed (hospital or clinic)
Training in diagnosis and management of IRD
events
Adverse events during treatment of HIV-TB
 54% (99/167) had adverse events,
 34% interrupted TB or HIV therapy
 Common adverse events
 Peripheral neuropathy (21%) - more common
with use of stavudine
 Skin rash (17%) - TB drugs (16), co-trimoxazole
(7), nevirapine (2), other drugs (4)
 hepatitis (6%) - TB drugs (6), unknown (5)
AIDS 2002;16:75-83
Example of drug-drug interactions in HIVTB care: atazanavir with rifampin
HIV Medicine 2007;8:131-4
Effect of rifampin on exposure (AUC)
of NNRTIs
% of normal AUC
100
80
78
69
60
40
20
0
Efavirenz
Nevirapine
Effect of EFV dose (600 vs. 800 mg) on
mid-dose levels, patients on RIF
Outcomes at 48 wks
On EFV
600 mg – 81%
800 mg – 74%
VL < 50
600 mg – 91%
800 mg – 87%
AIDS 2005;19:1481-6, AIDS 2006;20:131-2
%with viral load > 400
Virological failure of efavirenz-based ART, among
patients with and without rifampin for TB
14
11.5
12
10
8
8.1
8
6.4
10.4
5.7
6
4
2
0
6
12
18
Months of co-treatment
TB
No TB
JAMA 2008; 300: 530-9
Case #2 - Intubated in the ED
38 year old man sent from jail – 1 wk of fevers, cough,
dyspnea
 BP – 85/36
P – 100 T – 38.8
 ABG – pH – 7.21, PCO2 – 29, PO2 - 38
 Intubated for CV instability, acidosis, hypoxia
PMH –
 Meds – trim/sulfa, azithro, acyclovir
 AIDS CD4 – 2, VL – 10,200
 Crack cocaine abuse, frequent incarcerations
 PPD negative 3 mos. prior
Hospital course






Initial treatment – trim/sulfa and prednisone
Sputum DFA – negative for PCP
Sputum AFB – strongly positive
Started on parenteral INH, RIF, levo,
amikacin
Extubated, switched to oral IRZE
Culture – susceptible M. tuberculosis
In the ID Clinic





3 weeks into TB treatment – first ID Clinic
visit since TB diagnosis
Current TB treatment - IRZE 5 days/wk by
DOT
Living situation – SRO provided by TB
program
Drug use – clean and sober
Interested in ART, but very worried about
side effects and being experimented on
ART history


6 years ago – brief multidrug regimen, no
records, patient unable to identify meds
18 months ago – tenofovir / 3TC / EFV
 Initial suppression to < 50 copies/ml
 CD4 from 4 to 24
 Subsequent virological and immunological
failure 2o nonadherence
 Genotype: L100I, K103N (EFV), M184V
(3TC)
Patients who cannot be treated with
EFV-based ART




Efavirenz intolerance
Resistance to efavirenz (other 1st-generation
NNRTIs)
Pregnancy (at least for the first 1-2
trimesters)
Very young children (< 3 years)
Comparison of the effects of RIF vs. RBT
on trough concentrations of boosted PIs
% of noormal trough
concentration
200
175
150
120
117
113
100
50
1
2.5
ND
ND
0
LPV/r
ATZ/r
rifampin
DRV/r
FPV/r
rifabutin
AAC 2204;48:1553-60, AAC 2006; 50:3336-42, AAC 2010;54:4440-5, AAC 2008;52:534-8,
Effect of protease inhibitors on serum
concentrations (AUC) of rifamycins
PI
Rifabutin
Rifampin
Ritonavir
 400%
unchanged
Indinavir
 270%
unchanged
Nelfinavir
 200%
NR
Amprenavir
 400%
NR
Lopinavir/ritonavir
 300%
NR
Atazanavir
 250%
NR
Clin Infect Dis 1999; 28: 419-30
Clinical relevance of increased rifabutin
concentrations due to ritonavir
Adverse effect
Arthralgia
Joint stiffness
Uveitis
Leukopenia
% of patients on % of patients
ritonavir + rifabutin on ritonavir
9.2
4.1
4.7
38
0.6
0
0.6
19
11th International Conference on AIDS; abstract Mo.B171
Rifabutin PK with lopinavir/R in TB
patients (n = 16)
PK parameter
RBT 300
mg/day
RBT 150 mg
QOD + LPV/r
RBT 150
mg/day+
LPV/r
Median AUC
(exposure)
3026
2307
5010
Median Cmax
(peak)
297
168
311
Naiker S, et al. 2011 CROI, abstract 650
Rifabutin and TB therapy





Rifabutin is as active as rifampin
No dose adjustments of ART needed for
commonly-used drugs (ATZ, lopinavir/R)
Decrease RBT from 300 mg daily to 150 mg
daily when given with boosted PIs
Give remainder of TB drugs daily
Caution – RBT dose would be inadequate if
patient stopped PI
HIV, TB drug interaction - summary




Drug interactions in HIV-TB are regrettably
complex, but should not prevent HIV-TB cotreatment
Co-treatment regimen of choice: rifampinbased TB treatment + efavirenz-based
(standard dose) ART
Drug interactions should be managed, not
avoided – use a rifamycin-based regimen
New drug interaction guidelines at
http://www.cdc.gov/tb/TB_HIV_Drugs
Case 1 – Chest x-ray response to therapy
Diagnosis
2 months
Case 3 – Chest x-ray response to therapy - II
3 months
Started antiretroviral
therapy at 8 weeks of
TB therapy
Developed fever,
cough, left pleuritic
chest pain 10 days
after starting HAART
Types of immune reconstitution inflammatory
syndrome (IRIS) events in HIV-TB
• Hectic fever
• New or worsening adenitis - peripheral or central
nodes
• New or worsening pulmonary infiltrates, including
respiratory failure
• New or worsening pleuritis, pericarditis, or ascites
• Intracranial tuberculomas, worsening meningitis
• Disseminated skin lesions
• Epididymitis, hepatosplenomegaly, soft tissue
abscesses
Association between timing of ART and
risk of IRIS event (SAPIT)
42% hospitalized
22% hospitalized
5% hospitalized
Ann Intern Med 2012; 157:313-24
Association between timing of ART and
risk of IRIS event (SAPIT)
Median duration - 71 days
Median duration - 34 days
Median duration – 24 days
Ann Intern Med 2012; 157:313-24
IRIS in the CAMELIA study (median CD4
of 25)




Immediate ART increased risk of IRIS (33%
vs. 14% for early ART)
Similar timing of IRIS events (14 vs. 16 days
after starting ART
6 deaths, all in the immediate arm, were
attributed to IRIS events
However, immediate ART was associated with
a lower risk of death (8% vs. 14%)
N Engl J Med 2011; 365: 1471-81
IRIS events - implications for use of
antiretroviral therapy (ART)




Those who need ART the most (patients with low
CD4 cell counts) have higher risk for an IRIS
event and for a serious IRIS event
Delaying ART decreases risk of severe paradoxical
reactions, but increases risk of another OI or
death
Anticipate IRIS events – discuss beforehand with
patient and other care providers
Schedule early follow-up after starting ARV detect and manage IRIS events
Management of IRIS



Anticipate IRIS events – warn patients and other
care providers
Rule out other possible causes – bacterial
infections, a 2nd OI, inadequate Rx for OI, drugresistant pathogen
For relatively severe manifestations, prednisone
is reasonable

1 mg/kg (1.5 mg/kg with rifampin), tapering over
4-6 weeks
What’s happening in the clinic? Starting ART
in TB patients in London, 1998-2007)

British recommendations (at that time)
CD4 < 100 – at 2 weeks
CD4 100-200 – at 2 months
CD4 > 200 – after TB treatment

83 patients eligible to start ART

20 patients (24%) started ART at the
recommended point in TB treatment
Thorax 2008;63:935
Reasons for the delay in starting ART
among patients with CD4 < 100


Patient-related reasons:
 Refused to start
 Fear of side effects of ART
 Poor adherence
Physician-related reasons:
 Serious side effect of TB treatment
 Concern about ART side effects / IRIS
 Presence of another illness
 Seriousness of the manifestations of TB
7 (21%)
2 (6%)
3 (9%)
8
6
4
5
(24%)
(18%)
(12%)
(15%)
Thorax 2008;63:935
Starting ART during TB treatment –
summary of the steps required







Start TB therapy, deal with initial side effects
Help patient deal with the diagnosis of two
stigmatizing diseases
Start cotrimoxazole, deal with initial side effects
Assess readiness for HAART
Coordinate start of ART (~ 2 weeks for CD4 < 50,
~ 2 months for CD4 > 50)
Use DOT visits to  adherence with HAART
Anticipate and manage immune reconstitution
events
Summary – treatment of HIV-related TB:
issues with antiretroviral therapy

Should antiretroviral therapy be used during TB
treatment?


What regimens can be used for co-treatment of HIV
and TB?



Yes, for all patients
Preferred: efavirenz-based HAART + rifampin-based TB
treatment
Alternative: PI-based HAART + rifabutin-based TB
treatment
When should HAART be started?

2 weeks (CD4 < 50 to 2 months after starting TB
treatment
Two infections;
one patient
GLOBAL PARTNERSHIP TO STOP TB