Preface

1

Tuberculosis (TB) has been an unrelenting public health problem for decades mainly due to the prominent gap between the estimates of incident case load and the actual number of cases reported.^1–3 This gap is largely due to the limitations of the existing diagnostic tests among other considerations, prompting calls from WHO for efforts to reduce missed cases with TB under its “FIND. TREAT. ALL. #ENDTB” call for action.^4,5 Ocular tuberculosis (OTB) is a relatively unrecognized form of extra-pulmonary TB that can be the first or only presentation of TB infection.⁶ The disease causes significant ocular morbidity with recurrent episodes of ocular inflammation and can be potentially blinding if left untreated. However, there is a significant delay in the initiation of specific therapies due to considerable heterogeneity in disease phenotypes combined with lack of diagnostic criteria and consensus to treat with anti-tuberculosis therapy.⁷

This is largely attributed to a lack of robust evidence for the clinical phenotypes of this condition, whereby current studies are limited by isolated descriptions of phenotypic subgroups, small population size, or single-center configurations. These problems are further compounded by possible regional variations in disease expression, complicating the interpretation of these studies.⁷ The Collaborative Ocular Tuberculosis Study (COTS) group sought to address these uncertainties in several phases, beginning with a cloud-computing-enabled retrospective analysis of big data on ocular TB, followed by expert consensus through a modified Delphi technique.

Collaborative Ocular Tuberculosis Study (COTS)-1

The COTS group supported by the International Ocular Inflammation Society (IOIS) and the International Uveitis Study Group (IUSG) and Foster Ocular Immunology Society (FOIS) was conceived in 2015 to address the knowledge gap related to OTB through international collaboration amongst uveitis experts. COTS-1 involved pooled data analysis of 945 patients with tubercular uveitis (TBU) from twenty-five tertiary eye care centres. To facilitate data collection, a novel encrypted web-based platform (Cognito Form, Columbia, South Carolina, USA) allowing real-time regular updates of data entry was created.^6–17 Information on clinical findings, investigations, treatment regimens, and outcomes over two years was then analysed for a better understanding of OTB in the real world.^6–17 These results highlighted an unmet clinical need for an evidence-based approach in the diagnosis and management of OTB. In particular, global variations in clinical practices for investigations and therapeutic regimens based on geographic distribution, TB endemicity, local protocols and personal experience emerged worldwide, with no internationally accepted standards for management protocols.^6–17

COTS Consensus (COTS CON)

The Collaborative Ocular Tuberculosis Study Consensus (COTS CON) was conceived with the idea of bridging the gap between unmet clinical needs and available medical evidence, to guide clinicians in the treatment of OTB. COTS CON was a five-point Likert scale-based survey, consolidating the expertise of 81 uveitis specialists across the globe on the management of OTB using a 2-step modified Delphi technique.^18–20 Four hundred and eighty-six questions related to initiation of ATT in tubercular anterior uveitis, intermediate uveitis, panuveitis, retinal vasculitis and choroiditis and 46 questions related to concomitant use of corticosteroids and immunosuppressive agents in different phenotypes of tubercular choroiditis were deliberated. Experts provided their opinion based on anecdotal experience and review of the scientific evidence supporting the plethora of clinical scenarios discussed. These scenarios were extracted based on descriptions in the literature of specific clinical phenotypes associated with the presence or absence of corroborative evidence for TB from immunologic tests and radiologic tests in both endemic and non-endemic countries.^18–20 The resulting evidence- and experience-based algorithms for the initiation of ATT in OTB have been consolidated in the simplified algorithmic flow charts.^18–20

In the current supplement issue, we have compiled together the data from COTS-1 as well as the proceedings of COTS CON.

The aspirational goal of ending the TB epidemic will require global multidisciplinary collaboration including partnership with regulatory authorities and public health agencies. A reinvigorated research effort driven by AI is essential for achieving this goal. This trans-disciplinary collaboration will conceive efficient and robust disease models with machine learning constructs to aid the conduct of trials, research and development. The COTS group aims to implement an aggressive research agenda using big data from the ocular TB registry as a foundation on which to build a comprehensive AI-driven point-of-care CDSS that will enable transformative changes in the standard of care for ocular TB diagnosis and management, resulting in improved prevention and treatment of latent TB.

Disclosure statement

Dr Rupesh Agrawal is supported by a grant from the National Medical Research Council (NMRC), Singapore for the Clinician Scientist Award (CSA) from 2020-2023. (CSAINV19nov-0007: To establish a predictive artificial intelligence (AI) based model using immune-phenotype correlation for disease stratification and prognosis in patients with ocular tuberculosis (OTB)). He has not received funding for his work in this publication.

Additional information

Funding

COTS group would like to thank Foster Ocular Immunology Society (FOIS) for supporting publication charges for this supplement. COTS group would also like to acknowledge the funding support from Santen for conduct of COTS consensus meeting in Chandigarh, India.