Can active and past TB disease phenotypes provide insight into TB natural history and immunity?
Emily B. Wong, M.D.1,2,3,4, Stephen Olivier, M.A.1, Resign Gunda, Ph.D.1,5, Dickman Gareta, M.Sc.1, Theresa K. Smit, Ph.D.1, Sashen Moodley, B.Sc.1, Ngcebo Mhlongo M.B.,Ch.B., D.P.H.1, Alison D. Grant, M.B., B.S., Ph.D.1,6,7,8, Kobus Herbst, M.B., Ch.B., M.Sc.,1,7, Mark J. Siedner, M.D., M.P.H.1,2,8, Thumbi Ndung'u, Ph.D.1,4,9,10,11, Willem A. Hanekom, M.B.,Ch.B.1,4 and the Vukuzazi Team*

1Africa Health Research Institute, KwaZulu-Natal, South Africa
2Division of Infectious Diseases, Massachusetts General Hospital, Boston, MA, USA; 3Division of Infectious Diseases, University of Alabama, Birmingham, AL, USA; 4Division of Infection and Immunity, University College London, London, UK; 5School of Nursing and Public Health, College of Health Sciences, University of KwaZulu-Natal, Durban, South Africa; 6London School of Hygiene and Tropical Medicine, London, UK; 7School of Public Health, University of the Witwatersrand, Johannesburg, South Africa; 8School of Clinical Medicine, College of Health Sciences, University of KwaZulu-Natal, Durban, South Africa; 9HIV Pathogenesis Programme, Doris Duke Medical Research Institute, Nelson R. Mandela School of Medicine, University of KwaZulu-Natal, Durban, South Africa; 10The Ragon Institute of MGH, MIT, and Harvard, Harvard Medical School, Cambridge, MA
11Max Planck Institute for Infection Biology, Berlin, Germany

Background: Improved strategies to control TB transmission in endemic areas require a better understanding of the spectrum and natural history of TB disease in real-world settings. Defining human TB phenotypes based on active and past disease may provide insight into these and the nature of protective TB immunity.

Methods: Between May 2018-Mar 2020, in a health and demographic surveillance area in rural KwaZulu-Natal, we conducted a multimorbidity survey that offered all adolescent and adult residents testing for HIV, TB and diabetes. The TB screening algorithm consisted of symptom assessment, digital chest x-ray and, in a sub-set of individuals, sputum microbiology (MTB/RIF GeneXpert Ultra and MGIT culture). These allowed the definition of TB disease phenotypes that distinguished clinically-diagnosed active and past disease (Active-C and Past-C) from disease states that were detected only as a result of the population-based sputum and chest x-ray survey (Active-P and Past-P). We determined the risk factors associated with each of these TB disease phenotypes and calculated their sex- and age-specific distribution across the adult lifespan.

Results:
Among the 18,028 adolescents and adults (46.2% of the eligible population) who participated in multi-disease screening, 245 (1.36%) had active TB at the time of the survey: 71 (0.39% of the population, 29.0% of active disease) had clinically-diagnosed active disease and were on TB treatment (Active-C) and an additional 174 (0.96% of the population, 71.0% of active disease) had previously-undetected microbiologically-proven active TB (Active-P). The majority of people with Active-P disease reported no symptoms (81.69%) and were triaged for sputum microbiology due to chest x-ray abnormality (80.24%). Active-P disease was associated with increasing age (RR=1.02 (95CI 1.01-1.03) with each year of increasing age, p<0.001), male sex (RR 1.74 (1.22-2.47), p=0.002), current tobacco smoking (RR 1.78 (1.10-2.87), p=0.018) and HIV (ART-treated with suppressed viremia (RR=1.65 (1.16-2.35), p=0.005 and viremic HIV (RR=2.98 (1.88-4.73), p<0.001).

Among those who did not have active TB disease, evidence of past TB disease was present in 3,415 (18.94%): 1745 (9.68% of the population, 51.1% of past disease) had clinically-diagnosed and treated past TB (Past-C) and 1670 (9.26% of the population, 48.9% of past disease) had no history of clinical TB but had radiographic signs consistent with past disease (Past-P). Both states of past TB were associated with increasing age (Past-C, RR 1.11 (95CI 1.11-1.11), with each year of increasing age p<0.001) vs. Past-P, RR 1.04 (1.04-1.04), p<0.001) and male sex, though the magnitude of association was far greater for Past-C (RR 13.5 (13.5-13.5), p<0.001) compared to Past-P (RR 1.19 (1.07-1.32), p=0.001). Current smoking (RR 33.5 (33.5-33.5), p<0.001), history of household contact with an active TB case (RR 19.9 (19.9-19.9), p<0.001), HIV with suppressed viremia (RR 17.7 (17.7-17.7), p<0.001) and viremic HIV (RR 5.4 (5.4-5.4), p<0.001) were all strongly associated with Past-C and showed no association with Past-P.

Conclusions: In a HIV- and TB-endemic population in rural KwaZulu-Natal, we found high rates of active TB, most of which was asymptomatic and would have gone undetected in the absence of population-based chest x-ray screening. Further work is required to determine the nature of the Active-P phenotype, which may represent subclinical disease, and to determine its contribution to TB transmission. Nearly one-fifth of the overall and one-third of the older population (>50 years of age) had evidence of past TB disease. Interestingly, half of these had the Past-P phenotype with radiographic evidence of "missed" past TB that had resolved without treatment. The age- and sex-distribution and distinct risk-factor profile of the Past-P phenotype suggest that this state may be associated with protective TB immunity. Further study of the outcomes and immunological responses of these active and past TB phenotypes may shed light on the clinical and immunological events associated with resolution vs. progression of minimally-symptomatic active TB.

Emily B. Wong MD

Africa Health Research Institute & University of Alabama Birmingham

Appearances