The technological innovation and tuberculosis elimination: a Technology Foresight study

Ethics statement

Confidentiality and privacy were described to all participants prior to enrolment in the study. All data was anonymized, without possibility of participant’s identification. Such security is guaranteed by the software itself. All participants provided consent by the enrolment in the study (See supplementary information – Introduction page). Participation was voluntary and was not stimulated by any gains. Participants were free to give up at any time and an information contact was made available for any questions or issues. The application of the questionnaire did not involve any direct contact between researchers and the participants.

Respondents’ characteristics

Of the 29,988 invited researchers, 2,711 (9.0%) agreed to participate in the survey and answered the questionnaire. We obtained 2,567 valid answers², of which 69.2% were complete.

Regarding the characteristics of the respondents, the vast majority (95.5%) reported having a post-graduate degree, with 76.4% holding a doctorate. Considering areas of knowledge, 35% of respondents reported having knowledge in basic science areas - microbiology, genetics, immunology; 24% in clinical areas - infectiology, pneumology - and 20% in public health and epidemiology.

Regarding professional activity, respondents were mainly affiliated with teaching and research institutions - only 2% reporting working in industry. About 33% reported working in basic research, 17% in public health, 14% in clinical research, 12% in epidemiology, and 6% in translational research. About 2/3 of the respondents declared professional experience of more than 10 years. Finally, regarding geographical origin, 31% reported being from Asia, 23% from Europe, 16% from South America, 15% from Africa, and 14% from North America.

Strategies with the greatest potential impact for eradicating tuberculosis by 2050

The TB specialists who answered the survey considered strategies oriented towards improving diagnosis and treatment as those with the greatest potential impact on the TB epidemic. Moreover, the intensification of actions targeting the most vulnerable populations was ranked as the second most important factor. It was observed that interventions of a preventive nature - reducing the prevalence of TBI (36.6%), population screening to identify individuals at high risk for the development of TB disease (36.0%) and increasing population immunity (26.8%) - were the least valued strategies.

This response profile reinforce that those researchers are highly driven by the biomedical paradigm. This paradigm is represented by the seminal idea that “TB prevention begins with cure” (STYBLO & BUMGARNER (9). It assumes that the proper diagnosis and treatment of patients (cure), would reduce mortality, prevalence and, on an ongoing basis, incidence (39).

This same paradigm is present in several scientific and technical documents of literature specialized in TB (4). In the guidelines of the End TB Strategy (2015), for example, the first pillar - Patient-Centred: Care and Prevention, specifically expresses more than the importance of care and cure, but additionally emphasizes prevention within appropriate diagnosis and treatment of cases (WHO, 2015). In the same sense, Lönnroth et al. (2009) reinstate that the main challenge of TB control is to ensure globally adequate access to diagnosis and treatment (9). They point out that options to include preventive efforts have not yet been fully considered.

Thus, the priorities selected by the specialists appear to lie within the biomedical paradigm, in line with KUHN’s (1998) premise: “The commitments that govern normal science specify not only what sorts of entities the universe does contain, but also, by implication, those that it does not.” 40 (26)

It is worth emphasizing that the biomedical paradigm does not deny the existence of social factors that interfere with contracting TB. Only, such conditions are treated as externalities and not as part of a complex process of social production. Therefore, it is not paradoxical that the strengthening of control actions aimed at the most socially vulnerable population groups has been listed among the most prioritized actions. Notably, no specific epidemiological studies were found on the potential effects of this strategy, nor on new drugs for the treatment of MDR-TB for the elimination of TB.

Moreover, prospective studies based on Technology Forecast (31), have indicated opposite expectations regarding specialists’ responses about the potential impact of new strategies and technologies for TB elimination.

Dye et al (2013) estimate that reducing the time between diagnosis and the beginning of treatment would have an impact on the order of 28% in reducing the incidence of TB over a 35-year time span (3). Regarding the high prevalence of TBI, the authors point out that even if TB transmission were totally interrupted in 2015, the existing pool of infected persons would still generate more than 110 cases per million population in 2050 (hypothetical case of 1100 new cases per million/year). On the other hand, if only 8% of individuals infected with M. tuberculosis were permanently protected from the disease each year, the incidence could drop to 90 new cases per million. If the intervention reached 14% of those infected, the incidence could drop to 20 cases per million by 2050.

According to Abu-Raddada et al (2009), reducing the treatment time for TB-DS, would also have a limited effect on incidence, with an estimated reduction of 8.3%; 19.3%; 22.9% with shortening of the therapeutic regimen to 4 months, 2 months, and 10 days, respectively (41).. On the other hand, the authors estimate that neonatal vaccination, considering the portfolio of vaccines listed in the study, could decrease the incidence of tuberculosis by 39% to 52% by 2050. They point out that a vaccine targeting adolescents and young adults would have a greater and faster epidemiological impact than vaccination of neonates, an estimated 80% reduction in incidence by 2050 (41).

On the other hand, based on the model of social determinants of health (SDH), derived from the Public Health paradigm, it is possible to unveil an explanatory network, organized in layers of intertwined causal processes, involving from macro elements of the social production of the disease to its “natural” history, at the level of each individual.

It is noteworthy that TB has benefited nowadays from the negative effects on health, stemming from the urbanization process (agglomerations) that favor transmission and from the modern lifestyle (increase in chronic non-communicable diseases, including mental illness and drug addiction) that increase the risk of contracting TB (42).

The slow reduction in the incidence and increase in the proportion of MDR-TB has demonstrated that current control actions are progressively losing effectiveness and that there is a great need for new interventions primarily oriented towards prevention of illness, early detection, and reduction of the pool of infected people.

Schrager et al. (2020) reaffirm that only a new vaccine with the ability to prevent tuberculosis, targeted at adolescents and young adults, could effectively contribute to reducing incidence. If introduced by 2025, such a new vaccine could result in a drop in incidence of nearly 17% per year, effectively paving the way to eradication (43). In addition, for several authors, immunization would be one of the most effective strategies to address bacterial resistance, since the genetic sites of immunity are distinct from those that generate drug resistance (38).

In a prospective epidemiological study on the potential impact of a new vaccine, Harris et al. (2020), estimated that the expected reduction of incidence (IRR) in this period would be 51%, 52%, and 54% respectively for China, South Africa, and India, using 10 years of massive vaccination campaign, directed to adolescent population and an efficacy of 70% in preventing the disease as parameters (26). According to the authors, preventive action targeted only at the infected population would have a greater and faster impact, considering the goal of eradicating the disease by 2050.

On the other hand, in closeness with the public health paradigm, researchers such as Zenteno-Cuevas (2017), Schrager (2018,2020), Keshavjee (2019), Ole F Olesen (2019), Roland Diel (2013), Philipe Glaziou (2013), Holmes (2017); Barry Bloom and Ranif Atun (2013), Helen McShane (2011), advocate the need for new strategies, in particular aimed at reducing the huge contingent of infected (TBI) and preventing the active disease in the general population. However, even today a minority of experts support this view.

A second line of explanation for these results may be related to the tendency observed in TF exercises, that respondents should be captured by current problems, rather than theoretical reflection on future possibilities. The empirical, tangible reality sometimes imposes itself over an abstract, future-oriented logic. Thus, the focus on interventions with long-term epidemiological impact are displaced by the need for short-term care solutions.

Bonaccorsi et al, from the cognitive psychology and decision theory paradigm, poses the possibility of the occurrence of cognitive bias in technology foresight exercises conducted with experts only (44). According to this paradigm, competence to make judgments and decision making processes depend on memory capacity and the nature of the information being processed. “People do not perform exhaustive search and analysis of alternatives, but rather use well-functioning (but not optimal), rules of thumb (heuristics), and are subject to systematic distortions (biases) in their judgments (particularly, probability judgments) and decisions. (44) (2)

The participation of several stakeholders in foresight exercises is a seminal recommendation as a strategy to prevent biases. For technical-operational reasons, in this study, the exercise was conducted only with TB experts.

Aiming to mitigate the occurrence of this type of bias, Bolger and Wright (2017) suggest the application of multiple techniques - a process called expert knowledge elicitation (EKE) that makes it possible to confront, at different analytical stages, information from several sources (45). In this research, forecast studies and literature review data were used as analytical counterpoints to the research findings.

Such limitations, however, do not invalidate the method, nor its results. It only indicates that the future may result from this dominant thinking, in this case, linked to a biomedical view of the problem.

It is important to stress that the specialists’ choices may be able to solve important problems in clinical practice, which, on its own, is remarkable. Nevertheless, bearing in mind the expected reduction in the incidence of TB, we believe that such priorities will have little impact on the elimination of TB.

Most promising technological solutions to reach the market by 2035

Most respondents reported that point-of-care diagnostic tests and new drugs for MDR-TB would be the most promising technologies to reach the market by 2035. Given the very close scores (39.6% and 39.2%), the results suggested that the experts’ highest innovation expectations were also focused on diagnosis and treatment. New drugs for treating TB-DS was the third technological solution with the greatest potential to enter the market.

There was also a rather skeptical attitude toward the emergence of new TB vaccines and prognostic tests for individuals with TBI. Indeed, as discussed in a later section on new diagnostic technologies, significant technological advances in the last 10 years have made the field quite dynamic (42). NAAT Technology-based diagnostics have incorporated incremental innovations that bring new products closer to the ideals of point-of-care diagnostics and operational simplification (46). A technological path that adds to PCR-based nucleic acid amplification other technologies such as microfluidics, biosensors, and nanotechnology is already feasible.

This phenomenon has not been observed in the vaccine area, except for the use of technological platforms focused on the production of protein subunits, the most used in the development of several vaccine candidates. Other technological platforms such as recombinant or inactivated vaccines have also not been able to open a path of innovations. In the drug area, the most recent technological development also shows promising results with new molecules from new chemical classes. At least 18 new drugs from this category are in phase III of clinical trials (47).

This evidence, in the light of the postulates of the Economics of Innovation, suggests that the resources, the interests of public and private innovation agents, and the processes of interaction and learning have proven to be more effective in the diagnostic area. It is worth remembering, in the neo-Schumpeterian line of thought, that technological innovation is:

1. a continuous and cumulative process, involving not only incremental and radical innovation, but also its diffusion, appropriation, and use (48);

2. a process involving interactive learning and accumulation of knowledge, which takes place in connection with ongoing activities in production and sales (48);

3. institutions, in different innovation systems, affecting the nature of these interactions between people and customs, which, in part, may explain the accumulation or non-accumulation of learning processes (49)

4. intensive learning processes can be converted into technological innovation avenues or technological pathways (49, 50)

Thus, as an economic phenomenon, it is likely that technological innovation in diagnosis has been recognized by innovation agents in industry, governments, research organizations, and the financial system as more timely and promising -all this with the proper scientific backing of the biomedical paradigm.

It should be noted that technology is itself a body of knowledge about certain classes of events and activities, and not a mere application of knowledge brought from another field. The normal situation in the past, and to a considerable degree also in the present, has been that technological knowledge precedes scientific knowledge (e.g. BCG vaccine) (51).

Most promising vaccine technologies (to 2035)

The technological platform most valued by respondents for developing a new tuberculosis vaccine corresponds to the approach proposed by Zenteno-Cuevas (26.8%), defined as follows: This new concept of vaccinology encompasses the utilization and integration of diverse fields and techniques of study, including reverse vaccinology, immunogenomics, immunogenetics, systems biology, immune profiling, proteomics, whole genome sequencing, transcriptomics, host–pathogen interaction, bioinformatics, and computational modelling. 38 (5)

However, the scattering of the values assigned to the listed alternatives was observed. This, even considering the integrated technological platform proposed by Cuevas obtained 8 percentage points above the second option (new routes of administration - 17.9%) and double the score received by the inactivated vaccine technology (11%), considered the least promising of the technologies.

In the Pipeline Report Tuberculosis Vaccines 2021, there are 15 vaccine candidates listed, which are in clinical trials, distributed from the technological point of view in: a) inactivated/fragments (3); b) attenuated MTB (1); c) protein subunits (5); d) viral vector (3), e) BCG (2) (52). No candidates in this portfolio have been developed based on an integrated or mRNA/DNA platform (52).

Mike Frick, in an introductory text to this report, offer an interesting reflection on the dynamics of technological development in the area of tuberculosis vaccines, compared to the results achieved against COVID19.

While the technological development time was measured in months to obtain 20 vaccine candidates during the COVID19 pandemic, the technological development time used for TB – a disease that kills more than 1 million people per year – is measured in centuries. Nowhere to be found is the tool that would reduce TB incidence in time to realize the World Health Organization’s vision of eliminating TB by 2030: a new vaccine (or multiple vaccines) that protects adolescents and adults against TB. The confluence of the BCG centenary and the advent of COVID-19 vaccines begs the question: why are there no new vaccines against TB? 52 (1).

The author points out the large volume of resources applied (about 11 times what was applied in TB in the last decade) and the regulatory changes, which enabled the parallel achievement of some stages of technological development as main factors related to the success of this technological innovation process.

In addition, Frick highlights the key role of governments and funders in the proper direction of resources in the promotion of this process. This explanation brings us back to the innovation ecosystem approach, where innovation is the result of the push of demands arising from public health policies into the market. According to LI, GARNSEY (2014) “This requires new sources of funding and new relationships between public sector organizations, philanthropic trusts, medical foundations and entrepreneurial innovators” 53 (1). In the history of technological development of TB vaccines, we highlight some lessons learned:

• - Immunity produced by the classical T-lymphocyte induction pathway is not sufficient to produce protection against TB. Vaccines with a single dominant antigen, as a booster to BCG, do not seem to be as effective as expected. This has somewhat called into question technologies based on protein subunits and viral vector (Bloom and Lambert, 2016).

• - Bacillus’ containment is a more tangible goal from a technological perspective than preventing primary infection or sterilizing the granuloma (54).

• - Adolescents and young adults should be the preferred population group for a new vaccine, considering the goal of TB elimination (26).

• - Several vaccine schemes should be considered in the process of developing a new vaccine (55).

• - New vaccines are probably the best strategy for tackling MDR-TB (27).

The M72/AS01E (GSK) vaccine, protein subunit-based (containing Mtb32a-Mtb39a dormancy phase antigens), is the closest candidate to reach the market by 2035 (27). The efficacy observed in phase II studies is 54%. This may mean that we are still far away from a new widely effective TB vaccine, in accord to expectation of the respondents.

Diagnosis and prognosis of tuberculosis

The technologies that had the highest scores from experts for point-of-care TB diagnosis were those based on nucleic acid amplification (38.4%) by polymerase chain reaction (PCR) and new generation of gene sequencing (WGS or T-NGS) (32.9%).

Two issues should be considered when analyzing these results. First, incremental innovations in NAAT technology have been running for more than 20 years, with a number of new products becoming available on the market, including LoC devices that do not require a laboratorial environment or even electrical power (56).

In fact, NAAT is one of the most promising technologies to become an effective point-of-care diagnostic tool, progressively more accessible, simultaneously meeting the demand for diagnosing tuberculosis and mapping the profile of bacterial resistance.

At present, several products are available on the market and arebeing recommended by the WHO: Gene Xpert ULTRA, Truenat MTB plus, Easy NAAT MTBC, FluoroType® MTB (RT-PCR), BD Max MDR-TB (57). The VereMTB Test, in Lab-on-a-chip device by Veredus (Singapore), is under validation within research laboratories (42). The rise in market competition for the new tests should contribute to cost reduction and further diffusion of the technology to poorer countries.

The new generation of gene sequencing (WGS) used to study DNA from isolated M. tuberculosis although highly valued by respondents, applies more to Mtb genomic surveillance than to clinical practice. On the other hand, the use of T-NGS in clinical samples has an advantage because other than the potential of miniaturization and low operational cost, with the use of metal-oxide semiconductors and PCR reagents in very few quantities, the use of T-NGS can allow for TB-MDR/XDR diagnosis in 2 to 3 weeks (58).

However, none of these technologies are applicable to the diagnosis of TBI, remaining a serious obstacle to clinical practice and TB epidemiological control measures because they do not extend the diagnostic capacity in people with immunodeficiency, chronic diseases and in children by using sputum as a biological sample (46). Prior TBI testing is recommended for HIV-uninfected contacts over five years of age, as TPT is more likely to benefit those with positive tuberculin skin tests (TST). TST with PPD Rt-23 has low specificity in BCG-vaccinated populations and chain distribution limitations. Interferon-gamma release assays (IGRA) are acceptable alternatives, but the costs and the need for laboratory infrastructure are a matter of concern. New TSTs, using recombinant ESAT6-CFP10 antigen proteins, which include C-Tb (AJVaccines, Denmark, to be transferred to Indian Serum Institute), Diaskin Test® (Diaskin, Generium, Russian Federation), EC-test (Anhui Zhifei Longcom, China) have been developed to increase specificity in laboratory-free settings and recently recommended by WHO (59).

Host molecular biomarkers appear as the third most promising technologies (21.1%). The recent discovery of the ability of these markers to discriminate individuals with active TB from healthy individuals and with TBI (60–62), probably opens new routes of technological development (63).

MacGregor-Fairlie et al. (2020) note that TB diagnosis has been through several transformations, including the use of part of the bacillus (rather than the whole bacillus) in direct diagnosis and, most importantly, the introduction of indirect methods related to the host immune response profile (46).

Nevertheless, these technologies were considered less promising by the respondents. Point-of-care testing has the possibility of radically changing the method of laboratory production, specifically through means of decentralized approach and without a large technical equipment. The central laboratory may be replaced by a chip in the future, distributed throughout the primary health care network.

Two explanatory hypotheses can be raised in relation to these results. First, the medical tradition in the area of TB reinforces the need to identify the bacillus in order to make a diagnosis. Either due to the complexity of the disease or the need for differential diagnosis, it is plausible to assume a bias in this perception. A distorted perception that distinguishes tuberculosis from other infectious diseases with regard to technical requirements to the diagnostic and treatment processes.

The second aspect refers to the relevance given more to the diagnosis of clinically active disease than to the detection of TBI or increased risk of disease (prognosis). The biomedical paradigm prevails in the experts’ expectations.

The use of host biomarkers for the diagnosis of TBI and risk of illness (TBI/TB) could contribute to the control of the epidemic insofar as they make it possible to direct preventive treatment or the use of a therapeutic vaccine. This aspect is highlighted in the WHO report on research and technological development (2020), reiterating the need for increasing investments in this sector (WHO, 2020).

It is crucial to note that such strategic directing would face significant trade barriers due to competition and eventual replacement of NAAT-based technology.

Treatment of Tuberculosis

Most respondents considered new drugs from new chemical classes as most promising for TB treatment, regardless of the clinical form (TBI or active TB) or the drug resistance status (DS or DR/MDR). In the case of TBI, however, host direct therapy, by means of vaccines, was considered the second most promising option and with very close value to new drugs in new chemical bases.

For TB-DS, drug repositioning still appears as the third most promising technological option. For MDR-TB, new drugs in new chemical classes was the proportionally most valued alternative, with 40% of the responses scoring (5).

Currently there are 30 drugs in various stages of preclinical and clinical development (47). Of this list, twelve are new drugs from known chemical classes, with or without a lead drug. Eighteen constitute new chemical structures with new mechanisms of action. Of the drugs in phase 2-3 of technology development, seven are new drugs from new chemical classes and three from known chemical classes, which demonstrates the strength of the technology development arena⁴.

These data support the experts’ choice of drugs from new chemical classes as the most promising solution for the treatment of TB. For TBI, the immunotherapy strategy through vaccines still constitutes a technological challenge.