The development of
tuberculosis (TB) after the initiation of antiretroviral therapy is associated
with a low CD4 cell count, injecting drug use, and non-white race,
investigators from the US and Canada report in the September 15th
edition of the Journal of Infectious
“Our study highlights
the possible benefits of screening for tuberculosis prior to or shortly after
antiretroviral initiation even in low incidence areas,” write the
“screening should be focused on persons with a CD4 cell count below 200
cells/mm3, or increased HIV-1 RNA, persons of nonwhite race…history
of injecting drug use, and possibly male sex.”
Worldwide, TB is the
leading cause of serious illness and death in patients with HIV. Like other
AIDS-defining illnesses, treatment with antiretroviral drugs reduces the risk
of the infection.
conducted in Africa has found a high incidence of TB in patients shortly after
the initiation of HIV therapy. Possible reasons could include previously
undiagnosed disease, the unmasking of sub-clinical TB, recent infection, or
immune reconstitution inflammatory syndrome.
Regular screening for
TB is recommended for HIV-positive patients in the US, but there are no
recommendations about the screening for the infection prior to the initiation
of HIV therapy.
An understanding of
the risk factors for TB after starting HIV treatment could help inform screening
strategies and the targeting of resources at those most of this.
investigators examined the medical records of 37 845 individuals who started
antiretroviral therapy in the US and Canada between 1995 and 2009.
The patients had a
median age of 39 years, 77% were men, and 19% had a history of injecting drug
user. The study population was racially diverse: 43% were white, 38% were black
and 15% Hispanic.
Median CD4 cell count
at the time HIV therapy was started was 207 cells/mm3 and median
viral load was 48,312 copies/ml.
A total of 145
patients (0.4%) were diagnosed with TB.
This included 20
patients who were diagnosed with the infection within three months of starting
HIV therapy. These patients were significantly older than individuals who developed
TB after three months (41 vs. 37 year; p = 0.03); a lower baseline CD4 cell
count (61 vs. 134 cells/mm3; p = 0.04); and a higher median viral
load (217,344 vs. 64,069 copies/ml; p = 0.05).
The incidence of TB
was highest in the first month after HIV therapy was started (254 per 100,000
person years), and was still high after three months (196 per 100,000 person
years). Incidence then fell steadily and was 43 per 100,000 person year five
years after HIV therapy was started. The investigators note this was still
“8-fold higher than the overall tuberculosis incidence in the United States and
Canada during the study period.”
A set of analyses were
then performed to identify the risk factors for TB after starting HIV therapy.
These included black
race (in US only; p = 0.002), Hispanic ethnicity (p = 0.003), a history of
injecting drug use (p = 0.01), and a CD4 cell count below 200 cells/mm3
(p = 0.008).
Each increase log10
in baseline viral load significantly increased the risk of TB during the first
three months of treatment (p < 0.001), but not subsequently.
There was also a
non-significant relationship between male sex and an increased risk of TB
during the early months of treatment.
TB recurred in 1.5% of
patients. This low rate suggested to the investigators “the beneficial of HAART
[highly active antiretroviral therapy], which would be expected given the
decrease of recurrent tuberculosis with increasing CD4+ lymphocyte count.”