Endometrial TB-PCR positivity in women with unexplained infertility with no visible, microbiological or histopathological evidence of disease in a high TB burden area: whether to ignore or treat

Sir, We read the paper entitled ‘Favourable infertility outcomes following anti-tubercular treatment prescribed on the sole basis of a positive polymerase chain reaction test for endometrial tuberculosis’ (Jindal et al., 2012) with a sense of growing alarm. To come to a conclusion such as this, based on a study with a flawed methodology where intervention and non-intervention groups are so noticeably unequal, is not just illogical but also dangerous. The authors’ argument that it was unethical to leave women testing positive for tuberculosis (TB)-PCR untreated and hence their inability to have TB-PCR positive individuals in the non-intervention group seems to be based on a very flawed premise. First of all, how do the authors explain the pathogenesis of infertility in women with latent TB? Where is the evidence to support their claim that molecular mechanisms are responsible for infertility in such women who have novisible, laparoscopic, histopathological, radiologic or microbiologic evidence of the disease? The studies that the authors cite to support their claim of molecular mechanisms causing infertility in latent TB (Gurgan et al., 1996; Kumar and Rattan, 1997; Malik, 2003; Singh et al., 2011) in fact deal with mechanisms of infertility in women with active TB. On the contrary, Singh et al. (2011) found no difference in pregnancy rates or endometrial blood flow in women with or without genital TB based on TB-PCR. Also, the prevalence of latent TB in India is estimated to be between 35 and 50% depending on the methods used for testing (Aggarwal, 2005; Pai et al., 2005). There are no published studies yet that give us the prevalence of latent genital TB in fertile versus infertile couples or suggest that latent TB is sequestered more in the infertile population. There are published studies that have found the presence of mycobacillary DNA in the endometrium of couples with unexplained infertility, who suffer recurrent implantation failures (Dam et al., 2006), but whether this is the cause for either infertility or recurrent implantation failure is arguable. Until we have studies to prove otherwise, it seems logical to conclude that mycobacillary DNA is found in equal frequency in women with and without infertility in a TB endemic zone such as India. So the question of ‘with-holding treatment in individuals confirmed to have latent genital tuberculosis because it is un-ethical’ should not arise, since the association between latent TB and infertility is still intangible. Hence we come back to our opinion that this was a study that was majorly flawed in methodology. The other thing that shakes us up is the confidence the authors seem to place on PCR as a test for confirming or rejecting latent TB. For them, endometrial TB-PCR positivity seems to be synonymous with TB. It is true that most studies they quote including their own have found it to be a highly specific but not such a sensitive test for picking up TB (92% specificity as opposed to 59% sensitivity as per the authors’ own 2010 study) (Jindal and Bala, 2010). Fifty-nine percent sensitivity means that 41% of patients who had latent TB have been labelled by the test as not having TB. This in numbers comes to 118 individuals (169 × 100\59–169), who were pushed into the control group because of a false negative TB-PCR. This population that had TB but tested negative on PCR now forms 43% (118/274) of the control group. This group did not receive treatment and yet they fared just as well as the intervention group (pregnancy rates over 3 years 60.9 versus 59.8%, respectively). So the pertinent question to be answered by the authors is whether the so-called study group really benefited from this really toxic, prolonged, chemotherapy that is not without serious gastro-intestinal, hepatic, ophthalmic and auditory adverse effects (Yee et al., 2003). The only apparent ‘benefit’ that the authors claim of treatment (shortening of median time to conception from 10 to 7 months) hardly seems to be any benefit at all if one has to deal with such serious aforementioned complications. The social implications of being labelled a tubercular patient in a country like India are immense. Also, indiscriminate and incorrect use of antitubercular treatmentbasedonaflimsyandflawedpremise, suchaspositive TB-PCR alone, leads to a very real and present danger of the emergence of multi-drug resistant strains of TB, which remains the biggest challenge for Revised National Tuberculosis Control Program (RNTCP) health personnel battling active TB in India today.