Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 71
6
Addressing TB and Drug-Resistant
TB in Vulnerable Populations
Key Messages
• he difficulty of obtaining positive cultures in children complicates
T
the treatment of MDR TB and has obscured the extent of drug
resistance in pediatric TB cases.
• any children are infected by the adults with whom they live, but in
M
a significant fraction of cases, the infection comes from elsewhere.
• n one study from Peru, children living in households with an
I
MDR TB patient had 10 times the risk of infection of the general
population.
• oinfection with HIV and drug-resistant TB is a serious threat to
C
TB control.
• sia has a history of major refugee movements, and refugee status
A
can lead to displacement and overcrowding, which lead in turn to
the spread of infection.
Some vulnerable populations—including children, people coinfected
with HIV, and refugees—are at higher risk of contracting TB and are more
difficult to treat than others. Speakers at the workshop addressed each of
these three populations, discussing MDR TB among children in India, Peru,
and globally; data from India and elsewhere linking HIV infection to drug-
resistant TB; the occurrence of drug-resistant TB among Tibetan refugees
living in India; and case studies dealing with vulnerable populations (chil-
dren in Cambodia and MDR TB patients in Ethiopia, many coinfected with
71
OCR for page 72
72 DRUG-RESISTANT TUBERCULOSIS IN INDIA
HIV). All of the speakers emphasized the particular difficulties of reaching
vulnerable populations and the steps that must be taken to identify, diag-
nose, and treat MDR TB among these groups.
DRUG-RESISTANT TB IN PEDIATRIC POPULATIONS1
A major risk for pediatric TB is contact with an infected adult, observed
Soumya Swaminathan, Head, Division of Clinical Research, National Insti-
tute for Research in Tuberculosis. Rates of infection among adults aged
25–44 are highest in the African region, followed by Southeast Asia (Figure
6-1), and the risk of TB in children is likely to be correspondingly high in
these regions. Other risk factors include large household size, severe mal-
nutrition, exposure to household smoke, having a female index case, and
in some cases, being a member of certain minorities.
Like adults, children tend to go through several phases after infection
with M.tb. After an initial phase marked by hypersensitivity responses and
skin test conversion, which typically occur in the first 6–8 weeks, the pri-
mary disease follows. Most of the disseminated disease tends to occur in the
first 2–4 months after infection. Lymph node disease in younger children
and pleural disease in older children can occur at 6–8 months. The adult
form of the disease, which generally is seen in older children, can occur
several years after infection.
Swaminathan explained that more children than adults with TB are
smear-negative, although this varies with the population under study. In
one study of 1,098 children seen at the LRS Institute of Tuberculosis and
Respiratory Diseases in New Delhi, 414 children were smear-positive, 404
were smear-negative, and sputum status was not known for 280 patients
(Sharma et al., 2008). The smear-positivity rate was higher among older
children—about 60 percent—but even among children younger than 6 years
old, 30 percent were smear-positive.
Unknown Burden of Pediatric MDR TB
Data on MDR TB in children are virtually nonexistent. WHO does
not include children in drug resistance surveys, and most countries have
not collected these data systematically. A plan to gather data on children is
urgently needed, said Swaminathan.
1 This section is based on the presentation of Soumya Swaminathan, who was Coordinator
for Neglected Priorities Research with the WHO Special Programme for Research and Training
in Tropical Diseases (TDR) at the time of the workshop. Since the workshop, Swaminathan
has rejoined the National Institute for Research in Tuberculosis as Head, Division of Clinical
Research.
OCR for page 73
73
TB AND DRUG-RESISTANT TB IN VULNERABLE POPULATIONS
Incidence per 100,000 population
180
160
140
120
100
80
60
40
20
0
0–14 15–24 25–34 35–44 45–54 55–64 65+
African Region American Region
Eastern Mediterranean Region European Region
Southeast Asian Region Western Pacific Region
FIGURE 6-1 TB incidence rates are highest in young adults in the African and
Southeast Asian regions.
SOURCE: Swaminathan, 2011.
In India, an estimated 7 percent of the reported 1.3 million new TB
cases annually are in children. This percentage ranges from 2 percent of
new smear-positive cases to 15 percent of new cases of extrapulmonary
TB. The percentage of drug-resistant TB is more difficult to estimate. In
the Western Cape Province of South Africa, Schaaf and colleagues (2009)
found isoniazid resistance to be 7.7 percent among pediatric TB cases and
the MDR TB rate to be 6.7 percent, which represented an increase since the
1990s. These levels are higher than in adults in the general TB population
in South Africa. Previously treated children had significantly higher rates of
drug resistance than new TB cases, and HIV infection was not associated
with drug resistance in children, which is also the case in India.
In data from India, MDR TB rates among children with TB were found
to be 2 percent about two decades ago (Ramachandran and Prabhakar,
1992).2 A few years later, a multicenter study on children with pulmonary
TB found a rate of MDR TB of 3.5 percent. These rates depend on the
population under study and whether patients are coming from the com-
munity or a hospital. In general, said Swaminathan, drug-resistant TB in
2 These low pediatric MDR TB rates were observed prior to the implementation of DOTS
in India.
OCR for page 74
74 DRUG-RESISTANT TUBERCULOSIS IN INDIA
children mirrors that in the adult population because adults are the source
of their infections.
Challenges in the Diagnosis and Treatment
of Pediatric Drug-Resistant TB
Bacterial confirmation of MDR TB in children is not always possible.
Instead, drug-resistant TB in a child must be suspected when the child
has been in contact with a known case of drug-resistant TB, the child’s
adult contact has been on chronic irregular treatment and continues to
be sputum-positive, the adult contact dies after irregular treatment, or the
child shows initial improvement with anti-TB treatment and then deterio-
rates clinically and radiologically. Contact investigation is vital to detecting
pediatric cases of TB and especially MDR TB.
Diagnosis of drug-resistant TB in children requires specimens, such as
gastric aspirate, induced sputum, nasopharyngeal aspirate, and extrapul-
monary specimens, along with good specimen processing, transport, and
testing. The yield of culture in various studies in children has ranged from
10 to 40 percent, depending on patient selection criteria and laboratory
methods. The MODS method has been shown to be faster and more sensi-
tive than Löwenstein-Jensen medium. Another study found that MODS has
a slightly higher sensitivity than the MGIT method while also being faster
(Ha et al., 2009).
GeneXpert has the highest sensitivity in smear-positive culture-posi-
tive adults, providing an assessment of rifampicin resistance within about
2 hours, but these data are not available for children. (See the section
Case Studies in Cambodia and Ethiopia later in this chapter.) Preliminary
data from South Africa show that the sensitivity of GeneXpert is prob-
ably around 70-80 percent of culture-positive cases, which represents only
10–40 percent of all pediatric TB cases. More research will be necessary
before recommendations can be made regarding the use of GeneXpert with
different samples from children.
The treatment approach in children is largely the same as that in adults.
Factors to consider include the following:
• The child should receive treatment that is consistent with that of
the adult source case if no isolate is obtained from the child.
• At least three or, preferably, four or more drugs to which the isolate
is susceptible should be used.
• The child’s growth and development need to be monitored, and
drug dosages need to be adjusted for weight gain.
• The caregiver needs to receive counseling about adherence, treat-
ment duration, and adverse effects at every visit.
OCR for page 75
75
TB AND DRUG-RESISTANT TB IN VULNERABLE POPULATIONS
• It has been suggested that early primary pulmonary MDR TB,
because of its paucibacillary nature, can be treated for 12–15
months rather than 18–24 months; however, this needs to be vali-
dated in clinical trials. Microbiological monitoring is important,
but follow-up cultures often are difficult to obtain and more often
are negative. Clinical and chest radiographic monitoring during
follow-up is helpful.
In one study, the delay in initiating treatment for MDR TB in children
was just 2 days if the MDR TB source case was taken into account, but
the delay was 246 days if the source case was not considered (Schaaf et
al., 2003). The correlation between the DST results for the child and adult
was 68 percent in this study, although this correlation can vary greatly from
place to place. Obtaining a detailed contact history is essential since a delay
in initiating appropriate MDR TB treatment can have serious consequences.
In very small cohorts from Peru, second-line TB treatment was well
tolerated by children, even though they had high rates of malnutrition and
anemia (Drobac et al., 2006). Even when children showed resistance to
as many as five drugs, sputum conversion occurred in a majority of those
receiving individualized treatment. However, significant residual sequelae
and morbidity occurred, including 24 percent with airway obstruction and
40 percent with restrictive lung disease.
Swaminathan highlighted several challenges in the diagnosis and treat-
ment of MDR TB in children:
• The definition of MDR TB in children is different from that in
adults because a culture is not always available.
• Rapid molecular tests need to be studied in the context of pediatric
MDR TB.
• More data are needed on the burden of disease in children, includ-
ing drug resistance.
• More information is needed early in the process of drug develop-
ment on how new drugs work in children.
• Shorter regimens are needed for MDR TB in children.
• Preventive therapy is needed for MDR TB contacts of children.
THE BURDEN OF PEDIATRIC TB IN HOUSEHOLDS
OF PATIENTS WITH MDR TB3
An ongoing study in Peru, presented by Mercedes Becerra, Assistant
Professor, Harvard Medical School, is examining the extent of MDR TB
3 This section is based on the presentation by Mercedes Becerra, Assistant Professor, Har-
vard Medical School.
OCR for page 76
76 DRUG-RESISTANT TUBERCULOSIS IN INDIA
in the household contacts, including children, of patients with MDR TB.
Peru, with a population of almost 30 million, about 10 million of whom
live in Lima, recently was removed from the list of the 22 countries with
the highest TB burden. It now has an annual TB case notification rate of
just over 100 per 100,000 population. Eighty percent of TB cases in the
country are found in the capital, and the disease is concentrated in the
capital’s poorer districts.
In the last national drug resistance survey, which was conducted in
2006, MDR TB was found in 5 percent of new TB patients and in 23
percent of those with a history of TB treatment. HIV coinfection still is
relatively rare in MDR TB patients, with an estimated seroprevalence of
around 0.5 percent and a rate of about 1.4 percent.
Study Environment and Design
In 1996, Partners In Health, in collaboration with the government of
Peru, launched a program to treat patients with confirmed MDR TB, lead-
ing to a cure for about two-thirds (Mitnick et al., 2003, 2008). In 2004,
the program conducted a retrospective cohort study to determine what had
happened to the family members of these MDR TB patients, including chil-
dren. First, each patient who had started the MDR TB treatment regimen
between 1996 and 2002 was identified. A study team visited the households
of these index patients and interviewed them and others in the household,
asking specifically about TB treatment in any of the household members.
The study team then reviewed the medical charts at public health centers
of each household member who had reported TB treatment to obtain the
dates of the regimen received and other details.
With these data, the study team confirmed the list of individuals who
had been living with patients when they started their MDR TB regimen.
The team also identified those who had been treated for TB after the index
patients started their regimen. In January 2011, the team published a first
report about the TB disease burden in this cohort of almost 5,000 house-
hold contacts (Becerra et al., 2011).
Study Results and Findings
Unpublished data being prepared for publication as of the time of the
workshop cover 1,299 children in the households studied, 70 percent of
whom had at least 4 years of follow-up from the time the index patient
started the MDR TB regimen. Child household contacts were defined as
those less than 15 years of age. Three key results emerged from these data.
First, 67 children in the households were treated for TB disease during
this 4-year retrospective observation period, so that pediatric TB accounted
OCR for page 77
77
TB AND DRUG-RESISTANT TB IN VULNERABLE POPULATIONS
for 20 percent (67/343) of all the TB cases reported in the household con-
tacts. Thus, by the end of this period, more than 5 percent (67/1,299) of the
children in the households had been treated for active disease. Furthermore,
because 30 percent of the children had less than 4 years of follow-up, this
proportion may be an underestimate.
Second, of the 67 children treated for TB, only 8 had DST results in
their medical chart, because in children it is difficult to obtain adequate
sputum specimens for testing. Of these eight children, seven had MDR TB
isolates. Six of the eight had TB isolates available for genotyping, and all six
were found to be an identical genotypic match to the isolate from the index
patient. They were most likely either infected directly by the patient or part
of the same chain of transmission, so that both were infected by some other
source patient. This result strongly suggests that the great majority of the
observed TB disease in these children was due to MDR TB transmitted
from the index case.
Third, among this population of child contacts, the window to estimate
the prevalence of treated TB was defined as the period up to 6 months
before and up to 1 month after the date that the index patient started the
MDR TB treatment regimen. The prevalence of treated TB was almost
1,800 per 100,000 children. This prevalence was highest among 1- and
2-year-olds—more than 2,500 cases per 100,000 children, which is 10
times the prevalence in the general population. The TB prevalence in 1- and
2-year-olds was similar to that observed in the group of adults. Similarly,
the incidence rate of treated TB in all children during the first year of
follow-up exceeded 2,200 per 100,000 child-years.
“These disease rates are certainly alarming and should rightly give us
pause,” said Becerra. “These rates among children [in the prevalence win-
dow and in year 1 of follow-up] are squarely in the range of the TB case
rates that were observed in the jails and prisons of Russia in the 1990s.”
The only optimistic finding is that the high-risk window is within the
2-year period that is required to complete an MDR TB treatment regimen.
If a patient is being visited by a health worker or treatment supporter dur-
ing that period, that individual, with enough training and support, can
observe others in the household. But programs need clear guidance about
how to evaluate the household contacts efficiently over time.
The results of this study are an important reminder, according to
Becerra, that the household contacts of MDR TB patients, including chil-
dren, are a likely source of more MDR TB cases. Children who are living
with a patient starting MDR TB treatment are at high risk for having TB.
In this study, the estimated prevalence of TB disease at baseline was about
2 percent, which means that at least 50 children would need to be screened
to find each TB case. This amounts to roughly 10 times the TB case rates
reported in the general population.
OCR for page 78
78 DRUG-RESISTANT TUBERCULOSIS IN INDIA
Finally, the results underline the need for appropriate pediatric treat-
ment regimens: in the small number of children for whom DST results were
available, as in the larger number of adults tested, approximately 90 percent
had MDR TB. To optimize their chance of being cured, children with MDR
TB require regimens designed to treat drug-resistant disease.
During the discussion period, Edward Nardell, Harvard Medical
School, noted that inhaled drugs, particularly kanamycin and capreomy-
cin, could offer promise in the treatment of children. Preliminary studies in
guinea pigs reveal that inhaled capreomycin can achieve therapeutic levels
in the blood and high levels in the lung (see, for example, Fiegel et al.,
2008; Garcia-Contreras, 2007). Although very young children might have
difficulty with inhalation, many children already receive asthma medication
by that route.
Becerra noted the possibility of using a shorter treatment regimen with
children because their disease is being detected in an earlier phase relative
to adults. Nardell also cited the possibility of treatment with experimental
drugs because the drugs could be used under controlled circumstances to
look for an effect in a very short time.
There was some discussion about the use of prophylactic therapy in
contacts of TB and MDR TB patients. One participant asked whether
children should be considered a high-risk population and be offered DST
even if they are not smear-positive, despite limited laboratory resources.
Salmaan Keshavjee, Harvard Medical School, responded that this practice
would represent a major policy change in the developing world, although
it is seen in the developed world.
DRUG RESISTANCE IN HIV-INFECTED POPULATIONS4
Coinfection with HIV and drug-resistant TB is a serious threat to TB
control, said Digambar Behera, Director, LRS Institute of Tuberculosis
and Respiratory Diseases. Kawai and colleagues (2006) found that more
than 50 percent of HIV-MDR TB patients in Peru died within 2 months
of diagnosis. Studies with longer follow-up observed death rates ranging
from 72 to 89 percent (Coker, 2004). Authors of a study in the United
Kingdom estimated that MDR TB patients who are immunocompromised
are nine times more likely to die than those who are not immunocompro-
mised (Drobniewski et al., 2002). In a 2005–2006 study in the province
of KwaZulu-Natal, South Africa, 98 percent (52 of 53) of coinfected XDR
TB and HIV patients died, with a median survival time of 16 days from
the XDR TB diagnosis (Gandhi et al., 2006). A follow-up, retrospective,
4This section is based on the presentation of Digambar Behera, Director, LRS Institute of
Tuberculosis and Respiratory Diseases.
OCR for page 79
79
TB AND DRUG-RESISTANT TB IN VULNERABLE POPULATIONS
observational study in South Africa conducted from 2005 to 2007 revealed
that while 1-year mortality for MDR and XDR TB patients had improved,
the majority of deaths—40 percent of MDR TB cases and 51 percent of
XDR TB cases—still occurred in the first 30 days after diagnosis (Gandhi
et al., 2010). Among the 272 diagnosed MDR TB patients and 382 XDR
TB patients in this study, HIV coinfection rates were 90 and 98 percent,
respectively (Gandhi et al., 2010).
Although HIV infection has been associated with MDR TB outbreaks
in institutional settings, such as hospitals and prisons, whether HIV infec-
tion also is associated with MDR TB outbreaks in community settings
remains unclear. In Thailand, data collected prospectively on pulmonary TB
cases treated in public clinics showed that HIV is common among MDR
TB patients but is not an independent risk factor for MDR TB (Akksilp et
al., 2009). Nevertheless, populations at high risk for HIV—including young
adults, men, and injection drug users—should be a priority for DST, said
Behera.
A systematic meta-analysis summarizing the evidence from 32 studies
found no clear association between MDR TB and HIV infection across time
and geographic location (Suchindran et al., 2009). Comparisons of MDR
TB prevalence to HIV status ranged from 0.21 to 41.45. Assessment by
geographic region or study period did not reveal noticeable patterns. The
summary prevalence ratios for acquired and primary MDR TB were 1.17
and 2.72, respectively. While this meta-analysis could not demonstrate an
overall association between MDR TB and HIV or acquired MDR TB and
HIV, it does suggest that HIV infection is associated with primary MDR TB.
In general, well-designed studies and surveillance in all regions of the world
are needed to better clarify the relationship between HIV and MDR TB.
Specifically with respect to India, Deivanayagam and colleagues (2002)
found that about 60 percent of 1,000 TB patients were culture-positive, and
34 percent had MDR TB. The HIV seropositivity in the MDR TB group
was around 4.4 percent. Swaminathan and colleagues (2005) found that
the MDR TB rate in both new and previously treated TB cases was not
substantially different in HIV-positive and HIV-negative patients. However,
a study from Pune (Pereira et al., 2005) found that 10 percent of HIV-pos-
itive patients and only 2.5 percent of HIV-negative patients had MDR TB.
S. Singh and colleagues (2007) determined that of 54 patients with AIDS,
12 were resistant to first-line drugs, and 4 of these were also resistant to
second-line drugs.
Reports of XDR TB in India have been surfacing. However, most of
these reports do not include information about HIV status (Table 6-1).
With a population of about 1.21 billion, India has about 480 million
people infected with TB (Figure 6-2). It also has an estimated 2.27 million
people infected with HIV. People coinfected with TB and HIV are estimated
OCR for page 80
80 DRUG-RESISTANT TUBERCULOSIS IN INDIA
to number around 1 million. Of the approximately 2 million new TB cases
annually, an estimated 100,000 will have both HIV and TB infection, given
that about 5 percent of incident TB cases are estimated to be HIV-positive
(Dewan et al., 2010). However, the TB epidemic in India is being driven pri-
marily by the 400 million people with TB who are not coinfected with HIV.
The proportion of registered TB patients who are HIV-positive is highly
variable, ranging from less than 1 percent to more than 10 percent in differ-
ent parts of India. Within some districts, as many as 46 percent of registered
TB patients are HIV-positive.
Overall, said Behera, studies in India have not demonstrated an asso-
ciation between HIV infection and MDR TB, a finding that contrasts with
results of studies conducted elsewhere. However, the National Laboratory
Committee has decided that all future drug resistance surveys should cap-
ture HIV status and that TB patients should routinely be referred for HIV
testing.
Great challenges remain in the areas of diagnosis and treatment of TB-
HIV coinfection, said Behera. For example, DOTS-Plus does not include
a separate program for HIV-positive patients, and drug interactions in
coinfected patients can be difficult to manage. However, a TB-HIV col-
laboration begun in 2001 is conducting joint training, intensified case
finding, and HIV testing of TB patients with HIV risk factors. It has scaled
up its activities to 14 states and has piloted routine referral of TB patients
for HIV testing. A phased expansion of the TB-HIV initiative will cover
the entire country by 2012. Operational guidelines for airborne infection
control, refinement of treatment, and development of training materials all
are under way.
DRUG-RESISTANT TB IN MIGRANT AND
REFUGEE POPULATIONS5
The United Nations (UN) High Commissioner for Refugees defines a
refugee as a “person who owing to a well-founded fear of being persecuted
for reasons of race, religion, nationality, membership of a particular social
group or political opinion, is outside the country of his nationality and is
unable to or, owing to such fear, is unwilling to avail himself of the protec-
tion of that country.” An internally displaced person is a person sharing the
characteristics of refugees but displaced within the boundaries of a country.
The UN estimates that in 2006 there were more than 32 million refugees,
internally displaced people, or similarly vulnerable people.
Asia has a history of major refugee movements. Since 1979, nearly
5This section is based on the presentation of Kunchok Dorjee, Director, Tibetan TB Control
Programme, Delek Hospital.
OCR for page 81
TABLE 6-1 Profile of XDR TB in India
No. of No. of
MDR TB HIV-Positive Prevalence of
Study Setting Cases Cases XDR TB (%) Reference
Mondal and Jain, 2007 Tertiary care center, 68 Not reported 5 (7.4) Emerging Infectious
Lucknow Diseases, 2007
Singh et al., 2007 Tertiary care center, 12 All HIV-infected 4 (33.3) AIDS, 2007
New Delhi
Sharma et al., 2009 AIIMS, New Delhi, tertiary 211 All HIV-negative 5 (2.4) Indian Journal of Medical
care hospital Research, 2009
Ramachandran et al., 2009 Gujarat, field study 216 Not reported 7 (3.1) International Journal of
(all previously Tuberculosis and Lung
treated cases) Disease, 2009
Myneedu et al., 2011 LRS Institute 223 Not reported 45 (20.17) International Journal of
Tuberculosis and Lung
Disease, 2011
SOURCE: Behera, 2011.
81
OCR for page 82
82 DRUG-RESISTANT TUBERCULOSIS IN INDIA
Total population of 1.21 billion
≥0.9 million TB/HIV coinfected
400 million+ TB infected
~2.27 million HIV-infected
1.9 million new TB cases
4.85% of TB cases HIV-infected (~95,240 cases)
FIGURE 6-2 The TB epidemic in India is being driven primarily by the approxi-
mately 400 million people infected with TB who are not coinfected with HIV.
Figure 6-2
SOURCE: Behera, 2011.
6 million Afghan refugees have moved in and out of Pakistan and Iran.
The 1947 partition of India and Pakistan created the largest movement of
people in history—15 million. And in 1959, when the Himalayan refugee
crisis occurred, thousands of Tibetan refugees fled into India, Nepal, and
Bhutan.
Refugee status is a driver of TB, said Kunchok Dorjee, Director, Tibetan
TB Control Programme, Delek Hospital, because it results in displacement,
a scarcity of shelter, and overcrowding, which in turn lead to the spread of
infections, including TB. Refugee status also can lead to delayed diagnosis
as a result of such factors as financial and personal hardships, reluctance
to visit a doctor for anything less than an urgent condition, language and
cultural barriers, and a lack of health education. Delayed diagnosis of TB
can in turn lead to increased spread in the community. These same factors
can result in poor treatment adherence, default, and the emergence of drug-
resistant strains.
In 1959, thousands of Tibetans followed the Dalai Lama into exile.
The government of India provides asylum to Tibetans who continue to flee
OCR for page 83
83
TB AND DRUG-RESISTANT TB IN VULNERABLE POPULATIONS
across the Himalayas into exile in India. Tibetans have resettled in various
locations throughout India. The Tibetan Government in Exile is seated at
Dharamshala, Himachal Pradesh.
In the early years of exile, a large number of Tibetans died from TB.
The Department of Health of the Tibetan Government in Exile estimates
that the TB prevalence at that time was 30 percent of the entire exiled
population, although the exact number is not known.
The total population of Tibetan refugees is about 150,000, with the
largest number residing in India. It is an extremely mobile population, with
people coming into India and returning to Tibet annually. People move
across the Indian, Nepalese, and Tibetan borders and within countries,
which makes individual case management challenging. Many Tibetans live
in closed and congregate settings, such as dormitories in schools, monas-
teries, nunneries, and reception centers for newly arrived refugees. This
situation makes community transmission very easy and delayed diagnosis
very costly, said Dorjee.
In the 1990s, the incidence of TB among Tibetan refugees in India was
about 835 per 100,000 population (Nelson et al., 2005). A study of Tibetan
immigrants from India and Nepal to Minnesota showed a positive tubercu-
lin skin test (TST) rate of almost 98 percent (Truong et al., 1997). Among
Tibetan refugee claimants in Toronto, almost 97 percent were TST-positive
(Marras et al., 2003).
The number of new TB cases detected in the Tibetan population in
South India was relatively stable from 2006 to 2010, ranging from 223
to 291. Of these, between 14 and 23 were MDR TB cases, with a surge
of cases, from 9 to 23, occurring in 2010. One reason for the surge may
be that sputum cultures were done in 2010 for every smear-positive and
relapsed case, leading to the detection of more MDR TB patients.
At the Tibetan Delek Hospital in Dharamshala, the number of TB cases
declined from 290 in 2007 to 171 in 2010. But the number of MDR TB
patients stayed roughly stable, ranging between 33 and 43. Many of the
MDR TB patients at the hospital are college students, who must withdraw
from school to take their treatment for 2 years. These patients also include
monks, nuns, businessmen, and the unemployed. Tibetans born in Tibet
tend to have fewer cases of MDR TB than of drug-sensitive TB, even though
acquired drug resistance is relatively high in Tibet, while Tibetans born in
India have more. The number of women with MDR TB is slightly higher
than the number of men, and the great majority are between 14 and 40
years old. Resistance to second-line drugs is variable, and some XDR TB
strains are beginning to emerge in the Tibetan community.
The size of the Tibetan population is only 6 million, so an epidemic of
MDR and XDR TB in the community could be disastrous. Given Tibetans’
highly charged political situation, nonpolitical issues such as health care
OCR for page 84
84 DRUG-RESISTANT TUBERCULOSIS IN INDIA
tend to be overlooked. “MDR and XDR are a risk to the entire genera-
tion,” said Dorjee.
CASE STUDIES IN CAMBODIA AND ETHIOPIA6
Projects carried out by the Global Health Committee/Cambodian
Health Committee, which began working in Cambodia in 1994 and has
cured approximately 25,000 people since then, illustrate some of the dif-
ficulties of diagnosing and treating vulnerable populations.
Novel Diagnostic Modalities Among Children in Cambodia
The Global Health Committee/Cambodian Health Committee has been
investigating novel approaches for TB diagnosis among children in Cam-
bodia. As noted previously, diagnosis of TB is children is challenging. It is
often difficult to obtain sputum specimens, and children frequently have
paucibacillary disease, which makes microbiological diagnosis uncommon.
With standard diagnostic criteria, diagnostic accuracy in children is poor.
At the same time, validation of novel diagnostic technologies is per-
formed primarily among adults. Data on pediatric prevalence and incidence
are limited for both TB and drug-resistant TB. No TB drug trials in children
are under way, and few pharmacokinetic studies have been conducted in
pediatric populations. The result is that little information is available about
the efficacy of control and prevention measures.
The project in Cambodia (a partnership between the Global Health
Committee/Cambodian Health Committee and the Aeras Global TB
Foundation, with funding from the Annenberg Foundation), which was
described by Anne Goldfeld, Cofounder, Global Health Committee/Cam-
bodian Health Committee, and Professor of Medicine, Harvard Medical
School, is designed to evaluate the performance of GeneXpert and the
urine lipoarabinomannan (LAM) assay in a pediatric cohort, evaluate the
diagnostic utility of stool specimens, and determine the prevalence of TB
disease and latent TB infection in a cohort of Cambodian children. The
study involves a cross-sectional survey in a province of eastern Cambodia
abutting Vietnam where the Global Health Committee/Cambodian Health
Committee has worked since 1994. The study enrolled household contacts
of index patients with TB, children attending 28 outpatient health centers,
and children admitted to 2 district hospitals, with enrollment being con-
ducted from July 2010 to February 2011.
6 This section is based on the presentation of Anne Goldfeld, Cofounder, Global Health
Committee/Cambodian Health Committee, and Professor of Medicine, Harvard Medical
School.
OCR for page 85
85
TB AND DRUG-RESISTANT TB IN VULNERABLE POPULATIONS
Children were admitted to Svay Rieng District Hospital, where stan-
dardized and systematic TB screening data were collected, including med-
ical history, physical examination, chest x-ray, symptom screening, TB
exposure, HIV status, and HIV exposure. The laboratory workup included
two gastric aspirates, one induced sputum sample, a stool specimen, and a
urine specimen. The microbiological workup included acid-fast direct smear
microscopy, Löwenstein-Jensen and MGIT culture, GeneXpert, and urine
LAM. Among 876 children enrolled, the TB prevalence was 16 percent. But
the yield of microbiological confirmation using routine culture methods in
a well-established laboratory at the Institute of Cambodia in Phnom Penh
was low—just 1 percent. The TB in these children was being detected very
early, which probably increased the rate of overdiagnosis.
GeneXpert could be used for gastric aspirates and provided rapid
diagnosis of TB, but it yielded no incremental results compared with the
culture method. The results were obtained in 2 hours as opposed to weeks
or months, said Goldfeld, but the results demonstrate that other approaches
to diagnosing children with TB are needed.
During the discussion period, a workshop participant pointed out
that the advantage of GeneXpert and other diagnostic technologies is not
necessarily greater sensitivity but the decentralization of analysis. A major
difficulty with diagnosis is collecting quality specimens and transporting
them to a laboratory, and this difficulty is even greater with children and
other high-risk groups. Automated technologies can decentralize diagnosis
to such populations, the participant pointed out.
Filling the MDR TB Treatment Gap in Ethiopia
Ethiopia has the world’s fifteenth highest burden of MDR TB and
is one of the poorest countries in the world. In 2009, the Global Health
Committee/Cambodian Health Committee, working under the name Global
Health Committee in Ethiopia, initiated the countrywide MDR TB treat-
ment program in partnership with the Jolie-Pitt Foundation and the Ethio-
pian Ministry of Health.
Among a population of 80 million people, an estimated 130,000 new
TB cases occur each year in Ethiopia, including about 6,000 new MDR TB
cases, and these numbers are likely to be underestimates, said Goldfeld.
Before 2009, a major effort to build laboratory capacity, with assistance
from FIND, resulted in 221 MDR TB cases being documented by DST. In
August 2008, these 221 patients were waiting in the Addis Ababa area for
treatment. A GLC application initiated in 2007 had been approved for the
first 45 courses of treatment starting in October 2008.
The Global Health Committee/Cambodian Health Committee brought
a team to St. Peter’s Hospital in Addis Ababa to help initiate the Ethiopian
OCR for page 86
86 DRUG-RESISTANT TUBERCULOSIS IN INDIA
MDR TB program. It also brought the Ethiopian MDR TB team to Cam-
bodia to provide didactic and hands-on treatment. As of December 2008,
however, there still were no anti-MDR TB drugs in Ethiopia. Furthermore,
many other problems challenged the delivery of MDR TB treatment. Isola-
tion beds were not available, and the construction of a new ward at the
hospital had been delayed. Human resources were limited, and only partial
laboratory testing was available. A pharmacy for second-line drugs had not
been established, and an outpatient system did not exist.
Using a supply of capreomycin donated by Eli Lilly & Co. and funding
from the Jolie-Pitt Foundation, the Global Health Committee/Cambodian
Health Committee initiated MDR TB care, in partnership with the Ethio-
pian Ministry of Health, at St. Peter’s hospital in February 2009. Between
then and the arrival of the GLC drugs in September 2009, three cohorts
totaling 37 patients began treatment. When the 45 courses of treatment
arrived, the treatment team was already assembled, and other issues, such
as ancillary medications for side effects and management issues, were under
control. By the time the new MDR TB ward at the hospital was completed
in June 2010, five more cohorts of patients were receiving treatment. After
reviewing the project, the GLC sent another 245 treatment courses to
Ethiopia. Another program in Gondar, in northern Ethiopia, began with
three patients in August 2010.
As of the time of the workshop, 213 patients had been initiated on
therapy, including 183 in Addis Ababa and 17 in Gondar. Seven patients
had completed treatment. Eighteen had died, six within the first 30 days, an
indication of how sick this group of people was. One of the deceased was
a suspected XDR TB case. A total of 188 patients were currently on active
treatment as of the workshop—125 outpatients and 46 inpatients in Addis
Ababa and 17 in Gondar. Three patients who were presumed to have XDR
TB were being treated for it. Only one patient had interrupted treatment.
Of the 221 backlogged cases of MDR TB confirmed by DST, 66 had
started on therapy. Twenty percent of these cases were confirmed dead in
a house-to-house search, and 50 percent, many of whom presumably had
died, could not be located.
Of the 18 deaths among the 213 people treated, the mean time to
death was 79 days, with a range of 1 to 298 days. The mean age was 31.5,
with a range of 20 to 58. The mean number of prior treatments was 2.24.
Comorbidities included severe malnutrition (44.4 percent), HIV infection
(33.3 percent), diabetes (16.7 percent), cor pulmonale (16.7 percent), and
cirrhosis (5.5 percent). Most people died as a result of respiratory decom-
pensation associated with end-stage TB, including two with probable ten-
sion pneumothorax. One had TB pericarditis, another had worsening chest
x-rays, another died suddenly with cor pulmonale, and others had prob-
able superimposed pneumonias. At the time the six patients who had been
OCR for page 87
87
TB AND DRUG-RESISTANT TB IN VULNERABLE POPULATIONS
on therapy longer than 6 months died, three had culture-converted, and
three were persistently positive. None of the deaths was the direct result of
adverse events attributable to MDR TB.
Among all the patients, the mean age was 30.17 (with a range of 8 to
76). The mean number of prior treatments was 2.65, with a range of 1 to
8, and the HIV coinfection rate was 23 percent. The mean time to culture
conversion was 38 days.
The very low default rate has been achieved by engaging family and
other social resources in supporting the patient, said Goldfeld. All patients
are visited at least once a month in their homes and once a month in the
clinic. If their discharge orders include daily injections, they are seen every
day in the health center. Health care providers make sure patients are
taking their medications and deliver medications to patients who cannot
obtain them. Patients sign a contract stating that they will complete treat-
ment. Providers also bring food baskets to patients to support them and
help them deal with the gastrointestinal difficulties associated with MDR
TB treatment.
The direct collaboration that has been achieved among Ethiopian and
Cambodian physicians has been described as a “south-to-south transfer,”
or the sharing of expertise and best practices from one resource-limited
setting to another. Goldfeld explained that the approach of integrating
hospital- and community-based treatment has filled the gap in Ethiopia and
provides a model for expansion. A key challenge is to forecast the need for
drugs and make them available. Goldfeld contrasted the difficulties faced
in obtaining needed drugs with the speed and efficiency of retail systems,
noting that “we can get flowers from Holland to Winnipeg in 12 hours.
Why can’t we get life-saving drugs from drug stocks controlled by the GDF
to countries where patients are dying due to their lack in a timely fashion?”
Funds also are needed for clinical care, ancillary medications, basic labora-
tory tests, staff support, food, and outpatient monitoring. In Ethiopia, these
funds come from private sources, which Goldfeld said is surprising since
the government sets funds aside for treatment. However, in Ethiopia, it has
not been possible to directly access U.S. Agency for International Develop-
ment (USAID) funds to support treatment of MDR TB, as the substantial
funds allocated are directed to historical USAID partners not doing direct
care. Similarly, in Cambodia, the initiation and expansion of MDR TB care
in the country done by the Global Health Committee/Cambodian Health
Committee has been supported by a private donation from the Annenberg
Foundation and not by the USAID grants awarded to that country.
OCR for page 88
88 DRUG-RESISTANT TUBERCULOSIS IN INDIA
POTENTIAL INNOVATIONS AND ACTION ITEMS
Through the presentations provided in this session and the subsequent
discussions, individual workshop speakers and participants noted key inno-
vations and action items. They include the following:
• More research to obtain more data on children with MDR TB is
urgently needed. Such data could assist those attempting to develop
better diagnostics and treatment regimes for children.
• Expansion of ongoing efforts to address the challenges associated
with diagnosis and treatment of TB-HIV coinfected patients is key
to controlling the spread of MDR TB.
• “South-to-south transfer,” or the sharing of expertise and best
practices from one resource-limited setting to another, offers an
opportunity to learn from relevant experiences elsewhere in the
world.
Finally, since the workshop, a research network on pediatric drug-
resistant TB, The Sentinel Project on Pediatric Drug-Resistant TB, has been
launched by two workshop participants.7
7 Since the workshop, Mercedes Becerra, Assistant Professor, Harvard Medical School, and
Soumya Swaminathan, Head, Division of Clinical Research, National Institute for Research
in Tuberculosis, collaborated to launch a research network on pediatric drug-resistant TB. As
of April 2012, more than 140 individuals from more than 30 countries had come together
to collaborate on joint projects through the network, titled The Sentinel Project on Pediatric
Drug-Resistant Tuberculosis. More information is available at http://sentinel-project.org/
(accessed April 30, 2012).
