Diagnosis of Mycobacterial Infections in Rwanda

Jessica Crothers, MD

Fellow, Medical Microbiology, Brigham and Women’s Hospital

PGY 7

May 8, 2019

Entrance to Butaro Hospital in northern Rwanda

Tuberculosis is the leading cause of death from a single infectious agent worldwide. The WHO estimates that 10 million people developed active tuberculosis in 2017 and that 1.5 million of them died. Compounding these staggering statistics, Mycobacterium tuberculosis (MTB), the causative organism of tuberculosis, can be one of the most challenging diagnoses in clinical medicine. Notoriously difficult to culture, MTB often takes 4-6 weeks to grow in laboratory conditions and false negatives are common. Histologic visualization of the organism is similarly challenging, requiring special stains, time and experience as infections are often impressively pauci-bacillary and difficult to establish. Additional diagnostic strategies have been implemented to enhance timely diagnosis and therapeutic intervention, particularly in resource-limited settings in which patient follow-up can be difficult. These include AFB smear review and molecular (PCR) testing directly from patient specimens (GeneXpert, Cephied).  The multimodal and time-consuming approach needed for accurate diagnosis, often pushes physicians in high prevalence regions towards presumptive diagnosis and treatment as an important tact for both patient care and epidemiologic containment.

Land of one thousand hills, Rwanda is lush and agrarian

Rwanda suffers high rates of tuberculosis, particularly in the context of HIV-infection. Diagnosis often includes clinical assessment, PCR testing with GeneXpert, histologic evaluation and less frequently culture, which is only available at the national reference laboratory. Not uncommonly, surgical pathology specimens are consistent with a tuberculosis infection, revealing granulomatous or suppurative inflammation, but are unexpected and thus no culture/PCR or AFB smear was performed. In illustration of this, a young woman presented to CHUK with primary infertility. An exploratory laparotomy was performed and she was found to have bilateral hydrosalpinx. Histology revealed diffuse granulomatous inflammation and while MTB was suspected, it could not be confirmed. She was initiated on 9+ months of antituberculousis therapy, no small task. In truth, she could have MTB. But she also may not. Her disease could be due to gonorrhea, chlamydia or immunologic dysfunction. But without the appropriate diagnostic tests, she must be treated statistically: granulomas in a high-burden region = tuberculosis.

As countries like Rwanda continue to build their medical infrastructure and capacity, scenarios like this can be focused. Without too much technical burden, special chemical stains (AFB/Kinyoun/Ziehl-Neelsen) can be performed on surgical specimens to enhance diagnostic precision. Visualization of acid-fast bacilli in the context of granulomatous inflammation is much more likely to represent true tuberculosis infection and warrant intensive therapy. To that end, we plan to help develop a protocol and procedure for AFB staining of paraffin-embedded histologic tissue sections in the CHUK laboratory while here in Rwanda. I brought supplies in my luggage (which luckily didn’t result in tie-dyed clothing!) and we began staining. It was challenging, not straightforward and of course, took ten times longer than expected. Working in new environments with language and cultural barriers can often be frustrating, and I felt it today, but these experiences also offer the promise of a special kind of success. When you overcome small differences, find common ground, and work towards a common goal, the collective success you feel is so much greater than if it had been easy.  The first day, it took us 2 hours to even start staining, just finding the proper glass wear and reagents was challenging.  But, by the end of the week, a new tool was added to the local armamentarium.

Of course, the disease encountered in this particular young woman could be MTB, it could be gonorrhea, but it could also represent a different type of mycobacterial infection; like M. abcessus or M. chelonea. Such diagnoses, however, require identification by culture or genetic sequencing. We discussed and taught about the molecular and culture differences that can be encountered in these infections, but test implementation may be a task for another day!