latent TB: when should we treat it?

 I bumped into an article from 2018 which challenged the current approach of testing/treating latent tuberculosis (see TB LTBI if treat BMJ2018 in dropbox, or doi: 10.1136/bmj.k2738). This article is replete with references on the many cited studies, though some of them are very old and not accessible for my review. the article is open access and should be downloadable easily

Details:

-- natural history of TB infection, from the pre-antibiotic era:

    -- there were 3 longitudinal studies of TB acquisition and progression

        -- a 1950 article by Poulsen in the Acta Tuberc Scand Journal on the incubation period of tuberculosis, when he was working at the Faroe Islands from 1939 to 1947 (isoniazid, or INH, became available in 1952), at which time he could pinpoint the time of exposure to TB within a two week period, and often in a single day. He found that new tuberculosis reactivity (ie positive TB skin test) to TB-exposed people was under six weeks; and the incubation period to developing active TB was between three and nine months and almost always under two years

        -- a 1948 Swedish study by Wallgren in the journal Tubercle also found that active TB typically develops within 1 to 2 years of exposure

        -- a 1952 article by Gedde-Dahl in the journal Am J Hygiene monitored people regularly with tuberculosis skin testing (TST) conversion, finding that the development of active disease had an incubation of 3 to 9 months and rarely beyond two years after the development of newly documented tuberculosis reactivity

            -- this was all consistent with our current conception, which is that the development of active TB is disproportionately higher in the first 1 to 2 years of a TB test conversion to positive,

 

-- TB incubation studies from the post-antibiotic era:

    -- a 1970 article by SH Ferebee in the journal Bibl Tuberc reviewed the isoniazid chemoprophylaxis trials in the US between 1956 and 1966, finding that those individuals assigned to the placebo groups developed active TB particularly in the first year after a conversion of a TB test to being positive. and this development of TB dropped precipitously after that. 

       -- as noted in the graph below, those given placebo had a five-fold reduction in the development of active TB after the first year, with equalization to those on active treatment with INH for the rest of the 10 year study:

 

  

-- similar results were found in studying isoniazid prevention in South African gold miners (https://www.nejm.org/doi/full/10.1056/NEJMoa1214289), noting that in the very large number of people in the study, "during the first 9-month period, the incidence of tuberculosis was 58% lower in the isoniazid cohort than in the control cohort, with adjusted rate ratio, 0.42 (0.20 to 0.88; P=0.03) but subsequently was similar in the two cohorts (P=0.02 for the interaction between the cohort and follow-up period)": ie., the only benefit for isoniazid therapy was within the first nine months after treatment, with essentially no benefit subsequently vs placebo

        -- so, it was notable from these studies that INH protection happened specifically in that one to two year time period after exposure to TB, tracking along with the first pre-therapy studies finding that the vast majority of cases of TB after exposure were in the first 1 to 2 years, and treatment after that time was not effective. In fact there was no statistically significant difference in INH protection versus placebo in the 9-to-18-month period after initiating therapy. 

            -- and, this result is especially striking given that coal miners in general have a much higher risk of developing active tuberculosis, given the typically concomitant silica exposure (which alters the lung's immunologic response to pathogens), their close proximity of their work environment, and likely other factors)

    -- there is also evidence that radiologic fibrosis consistent with tuberculosis does not change the time intervals noted above, even though these patients are considered to be at a higher risk for reactivation

    -- and, of course, the requisite animal study: in the pre-antibiotic era 96%-98% of visible calcified tuberculosis lung lesions in patients who died from causes other than TB did not reveal viable tuberculosis mycobacteria in these calcified lung lesions when implanting these lesions into guinea pigs (guinea pigs are extremely sensitive to TB: even one TB bacillus will kill them)

 

-- "reactivation TB", the argument for not treating latent TB in older people who have waning immunity:

    -- the above guinea pig study suggested that calcified granulomas for example do not seem to contain viable TB mycobacteria

    -- there is essentially no reactivation in those studies having 10 years of follow-up, which likely included some people with immunologic dysfunction. For example the South African gold miners do typically have local immunologic dysfunction within the lung tissue itself

    -- a 1951 (pre-INH) study done by K Terplan in the journal Bibl Tuberc studied people who died from tuberculosis, finding that 90% of the 51 patients over age for 40 actually died from TB involvement from new exogenous foci and not from the previous focus of infection

    -- G Canetti et al in a 1972 study in the Bull Int Union Tuberc, assessed the isolates of TB having drug resistance (at that time, drug resistance likely occurred because of use of isoniazid, streptomycin or aminosalicylic acid), making the assumption that if reactivation of remote TB were responsible, then the older people should have a much lower frequency of drug-resistant TB (since they got the active TB before the mycobacterium was likely to be treated and developing drug resistance) as opposed to the younger cohort who were more likely to be exposed to resistant bugs. they compared people aged 15 to >60 years old finding that the difference in was minimal (9.2% of all 9456 isolates from younger patients were resistant to these antibiotics , vs 7.6% of 1996 isolates from those age >60). This supports the concept that people with TB later on actually were reinfected instead of having old bacilli reactivated

        -- i could not find several of these references, but these were likely populations with high exposure rates to tuberculosis

 

--Other issues:

    1. has the incubation period of TB changed over time? 

        -- this does not seem to be the case. In a study in Amsterdam of TST positive contacts who did not take isoniazid prophylaxis for 10 years, 75% of active TB cases occurred within one year of diagnosis and 97% within two years

        -- two other studies also supported this result using molecular fingerprinting of the TB or other genetic markers

    2. is there a late spike in TB disease that would support a significant increase in TB infections in older people?

        --  in a 20-year study following TST positive and TST negative adolescents in England and Wales in a BCG vaccine trial but were in the control arm from 1951 to 1970, there was a sharp decline in TB incidence. This finding was also true in the intervention group for the rest of the study

        -- another 20-year study of Asian migrants to London, stratified by whether they remained in the UK versus having returned to Asia to visit friends or relatives, found that for those who never returned to Asia, the majority of TB cases occurred in the first two years after their arrival to the UK with a steady steep decline afterwards and no late peaks; in contrast those who visited their home countries and had no known contacts with TB in the UK before their diagnosis of TB had continued active TB cases over the next 20 years

        -- a Norwegian study analyzed the incidence of TB over 20 to 30 years in 10-year birth cohorts from 1879-1888 up to 1959-68, finding a decreased incidence over the 10-year observation periods in all age cohorts

   3. what does TB "immunoreactivity" really mean? (the authors' definition is when a person has a positive TST or IGRA but is asymptomatic):

        -- a study in hospital employees found that those with positive TST tests who were recent converters (<1 year), 5 of 20 people remained TST positive, the rest reverted to TST negative or were positive to a smaller extent; but of all of the individuals who had been TST positive for more than a year remained so after a year of INH treatment (ie, the length of exposure to TB was important in leading to longterm TST reactivity)

    -- a naval study found similar results

    -- in a study of 38 patients who had active TB, TST positivity was retained in all 38 patients even though they had completed a treatment regimen associated with <3% recurrence at five years

 

-- Background guidelines in the US:

    -- Latent TB treatment guidelines from the CDC, 2020: http://gmodestmedblogs.blogspot.com/2020/02/new-ltbi-treatment-guidelines.html

    -- USPSTF guidelines from 2023: http://gmodestmedblogs.blogspot.com/2023/05/latent-tb-infection-screening-uspstf.html. This blog reviews the recommendations for screening Individuals for TB, as well as the various treatment regimens

 

 -- several issues are still unclear as to whether there are groups of patients who should be tested and treated for latent tuberculosis:

    -- for example, there are immunosuppressives where latent TB should be treated. Are there specifics of what particular doses and length of time on these medications indicate necessary therapy

    -- for patients who are on corticosteroids, what dose and what length of exposure are truly associated with TB reactivation (and what is the quality of the data on this?)

    -- for those with other "indications" for more aggressive treatment, such as chronic kidney disease (at what stage?, a concern which becomes complicated as people get older and often have relatively advanced CKD), people with chronic stomach acid disruption (e.g. post gastrectomy, or long-term PPIs, the latter being shown to be associated with a 1.3-fold increase odds of developing pulmonary TB), people with diabetes (for diabetes types 1 and 2, stratified by their level of control), patients with silicosis (though the above-mentioned South African gold miners suggest silicosis is unlikely to be a significant risk factor for reactivation, though still could be reinfection), HIV (is that still a risk factor with excellent HIV control, since we do know there is still some immunologic dysfunction that persists), etc.

    -- and, perhaps the biggest question is whether there is a difference in the the relative relationship between the approach to individuals in areas of high TB prevalence vs low

        -- we do know that prior exposure to TB leads to some protection against reinfection: there was a review of 18 publications (from observational cohorts of healthcare providers from the 1920s through the 1950s) reporting later active tuberculosis incidence among 23 paired cohorts of individuals with and without latent infection (total n = 19,886). The weighted mean adjusted incidence rate of tuberculosis in the latent TB and the uninfected groups attributable to reinfection was:

            -- those with prior TB test positivity but never treated: 3.5 per 1000 person-years (5.0–26.2 per 1000 person-years)

            -- those not previously TB infected (ie primary infection): 60.1 per 1000 person-years (38.6–87.4 per 1000 person-years)

                -- ie, the adjusted incidence rate ratio for tuberculosis in the latent TB group compared with the previously uninfected group was 0.21 (0.14–0.30), suggesting a 79% lower risk of developing active tuberculosis after a primary TB infection than for previously uninfected individuals. see https://pmc.ncbi.nlm.nih.gov/articles/PMC3284215/ ): ie, it happens but not so frequently

        -- this all leads to the following  graph of TB incidence in the population and the likelihood of TB reinfection, the other group presumably by TB reactivation (which is also pretty uncommon: https://pmc.ncbi.nlm.nih.gov/articles/PMC2921602/)

 

 

see TB reactivation vs reinfection IntlJMycobact2016 in dropbox, or doi.org/10.1016/j.ijmyco.2016.09.017

 

 

so,

-- this brings up the risk-benefit analysis:

    -- it is pretty clear from the above compilation of studies that for the considerable majority of older people have latent TB but live in high prevalence TB areas, later developing TB is from reinfection. And, detecting/treating the latent TB in this population when first found (but after the high risk exposure of the first couple of years) is very unlikely to be effective in preventing subsequent active TB infection. this makes sense since these people would continue to be exposed to TB, and the immunologic protection from their initial exposure is only partial

    -- the medications used to treat "latent TB" are certainly not without potential severe complications (eg, death from INH-induced hepatic failure....), though some regimens (eg rifampin) are less toxic but have considerable drug-drug interactions with other meds

    -- but there may be more utility to test and treat those with latent TB who are in low TB prevalence communities, since reactivation TB may be more prevalent (it would be very useful to have a formal risk/benefit analysis here, since reactivation TB is still very uncommon and testing/treating does have its risks)

    -- one concern is that in countries with low prevalence of active TB overall, there may well be groups of people who are exposed to TB much more frequently over time, eg immigrants from high-risk countries who return to their countries periodically or interact a lot with people newly coming from their countries, others who travel/work in higher TB prevalence countries, etc. These people are much more likely to get TB reinfection, as noted in the London study above of Asian immigrants who visit their countries (and see the graphs below), and perhaps these individuals should be considered to be similar to those who inhabit high-prevalence countries and just be followed closely for reinfection.

 

To reinforce the last statement. Here is the graph from the London study of people developing active TB, with graph A being those Asian immigrants in the UK who entered the UK and did not leave the UK or have known contacts with TB in the UK; graph B is those who visited their home countries but had no contact with TB in the UK before; and graph C is the time to onset of active TB in the group going home to the subsequent development of TB confirming that in that group developing active B was still predominantly in the first few years after the new exposure):

 

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them