Large Scale Diagnostics

There are about 265,000 Shiga Toxin E.Coli and 1.35 million Salmonella bacteria caused infections in the US each year. These two alone cost an estimated $4.5 Billion to the US in medical care and lost wages. Worldwide, diahhreoa causes over 1.5 million deaths.  There are about 10.6 million all over the world, leading to over 1.5 million deaths as well. 

Although when we think of diagnostic, we think of the sensing method (such as a Covid-19 strip test), a diagnostic is also the sample used. Most bacteriologically confirming diagnostics (such as microscopy or PCR) require sampling i.e. Taking a small amount of sample from the whole, often heterogeneous mix, and examining it through a ‘probing’ method. In order to select and contrast to enhance signal to noise ratio for the probing methods, usually at least one reagent is used if not more. Not only does this increase per test cost, but it also loses most of the product being examined by being limited to work with very small sample volumes. Working with small and often bad samples, therefore, not only increases the cost of the sampling method to get meaningful performance (e.g. PCR), but also limits the use of diagnostics as a field, to personalized as opposed to public diagnostic. 

The graph shows that to detect even high prevalence of bacterial contamination in lettuce, a large number of samples must be taken from the product, which is extremely cost prohibitive through good sensing methods such as PCR. 

Personal diagnostics are inherently not designed for scale. Governments worldwide have tried scaling them for diseases such as Tuberculosis, for example. And while treatment outcomes have become better, number of new cases has remained steady. In South Africa, for example, where microscopy was replaced with GeneXpert for about a decade, rates of new cases have not gone down. One major reason is that it is community transmission that leads to new cases and needs to be stopped. Better centralized diagnostics only help diagnose people who go seek help, not the actual ongoing transmission. 

We need a new model of diagnostics. Large Scale Diagnostics, that can test media such as air, water, food before our communities come in contact with them. Before they get infected or the disease transmits.

Medupi is a 2-part system that can handle large physical volumes of samples of various types, and can detect and identify bacterial pathogens, reagent-free and label-free. The first part (MagnaSlide) is a millifluidic polymer system that can selectively concentrate certain bacteria. The second part (Medupi) is a polarimetric system that can measure scattering from captured bacteria to detect and identify species, reagent and label-free. This system will be able to handle large samples such as those from flooded waters, or food processing plants without needing to rely on manual sampling techniques to sporadically test and certify diseases such as E.Coli and Salmonella. Being able to selectively concentrate large physical volumes, will enable not only large-scale environmental screening, but also enable the use of sample types such as saliva and urine (and processed air in buildings) as well as pooled testing regimes for human bacterial pathogens such as Tuberculosis. While genomic sequencing is a powerful tool, it has significant practical limitations for widespread real-time surveillance - high costs, complex workflows, need for specialized infrastructure and expertise. Our solution overcomes these barriers, providing an affordable, portable and user-friendly system that can be deployed at the point-of-need.

This also allows practical testing at-source, thereby reducing the spread of bacterial diseases - High costs, complex workflows, and need for specialized lab infrastructure required for omics based testing restricts their widespread deployment, especially for real-time environmental monitoring and in resource-limited regions where the need is greatest. While not aiming to replace genomic methods, we provide a powerful complementary tool for the bulk of routine large-scale surveillance. Our system can rapidly prescreen samples to catch threats early and focus sequencing only where needed for detailed characterization. The difference between 40 minutes and 4 minutes is huge in public health surveillance as well as global health diagnostics that’re both extremely time and cost sensitive.

Our team has established partnerships in India, which has the highest TB burden in the world, apart from South Africa. Apart from routine patients visiting primary health care clinics (including lower and middle income patients in rural, peri-urban and urban areas), we will focus on piggybacking on the government's existing active case finding efforts through CCWs, as well as CCWs in the field for routine screening of other diseases. Vulnerable and neglected populations include migrant workers, people living in slums, prisoners, HIV infected patients, contacts of diagnosed TB patients in Tier - 2, and Tier - 3 cities, as well as villages. We will also be testing water sources through government WASH initiatives in the state of Tamil Nadu and Uttar Pradesh.

The team has 2 full time co-founders, Bonolo and Digvijay.

Bonolo has a background in biomedical and electrical engineering. Digvijay has a background in materials science and industrial design. They attended  the bioengineering program (CBID) at Johns Hopkins for their masters, as part of which they've conducted first hand observations and interviews in Brazil, South Africa, US and India. Coming from countries with very high TB populations, both Bonolo and Digvijay have seen first hand, the conditions of patients as well as the healthcare systems serving them. They also understand the exact local context under which patients operate. For example, during a community health worker interview, they learned how upon suspicion of TB, people don't go to their nearest clinic for the fear of being outed. They go to a clinic further away, which affects care and time very uniquely in transmission dynamics. 

Despite carrying masters debts, they decided to work together on creating impact for TB stricken populations, and ultimately created products that're applicable for a variety of global health problems and challenge the problem in new ways. The team has collected and assimilated over 300 interviews with stakeholders at various levels across TB, WASH and Food Safety, and has incorporated inputs in design, strategy and impact of the products.

Medupi consists of 2 parts:

1. A sample concentration and prep cartridge (MagnaSlide)

2. A Real-Time Label Free Imager (Medupi)

MagnaSlide has been tested with real HIV co-infected TB samples in India in an IRB approved study, where it increased the sensitivity of microscopy from 55% to over 95% (ie PCR level performance for DS-TB diagnosis for 1/10 the cost). Various private diagnostic lab chains in India (with over 2500 centers across the country) are currently in process for pilot prior to purchase for MagnaSlide for TB diagnosis. We've also received Letters of Intent for 20,000 devices.

Medupi has been tested in the lab with contrived samples and demonstrated proof of concept, and is currently in the engineering phase, with a beta prototype being delivered at the end of Q2 2024.

We're above all, looking for partners and teammates to join us. 

As the scope of our systems has expanded beyond TB, commercialization, business development and distribution have become a priority to reach funding goals for development. We're looking for experts who believe in our mission, and are enthusiastic about changing the paradigm of diagnostics with our Large Scale Models. 

We're also looking for non-dilutive funding to continue R&D and enable access for our NGO partners to our devices and services.

We're looking at diagnostics as field in an entirely new way. One key reason is how much harder distribution of testing and care is AFTER people get infected, and how much costs systems incur to make that unsustainable without aid for governments.

Moving to a prevention based model has largely been through qualitative health practices in the community, with sporadic medication use (outside of some vaccine scope). With systems like ours that enable large-scale testing of food and water sources, air in high density buildings, pooled testing of human samples while also retrofitting into existing health infrastructure (such as India's 40,000 public and private TB microscopy labs), governments have a low cost quantitative way to reduce incidence of infectious disease. Our current bacteria agnostic approach means that key participants in TB industry, Typhoid and Diahhreoa industry, food infections industry and more, can all come together and benefit from both data and controlled number of cases. Ultimately, both data from Medupi and hardware are cost effective enough for even local communities to collect and gain insight from. Such real-time systems can be the world's first real pandemic/outbreak detection and prevention system.

On a more tactical level, 

Behavioral insights:

1. Time for patients to get into the TB treatment system - This takes 4-6 weeks. Most products are lab based and thus do not focus on this part.

2. Cost for each test - Cost per test needs to be low enough so that it is realistic for state and central governments to not just test, but 'over-test' over long periods.

3. Existing infrastructure (including personnel who've already been trained) - Governments invest in infrastructure over decades. Faster and wider deployment requires fitting into existing infrastructure.

4. Patients habits (delay in going to doctor, going to small facilities first) - Are not going to change. Non urban patients will not suddenly change the way they access healthcare and go to central facilities or bear severe symptoms waiting for test results, or not be satisfied with alleviated symptoms (antibiotics prescribed at primary centers). Instant enrollment is key.

Technical insights:

1. Sample quality enhancement without additional steps to increase performance of existing tests is an underexploited part of the diagnostic pathway. Medupi cartridge concentrates the sample, enabling low limits of detection.

2. Specificity of microscopy comes from differential staining (enough contrast) and technician's knowledge. Contrast can come from selective concentration and knowledge from peak analysis.

Our surveillance system will become stronger as more states, join the network. We anticipate that since our surveillance system will provide direct visibility to donors in the space, state governments are unlikely to forgo, and will enroll their state's CCWs in the program.

One of our goals is to contribute towards eradicating TB in India by 2025. This ambitious plan will be realized through a series of immediate and future goals that we have laid out.

However, it's crucial to understand that combating TB isn't a straightforward process. It's not just about testing and treating; it's a multi-layered issue that needs a holistic approach. Better TB drugs, treating latent TB infections (people who have TB bacteria but aren't yet sick), developing a TB vaccine, and introducing non-health interventions, like improved living conditions, are all part of the broader strategy.

To give you an idea of what we're aiming for, let's look at some numbers from India's National Strategic Plan to End TB. According to this plan, increasing active TB case finding (that is, actively looking for and diagnosing TB) by 25% by 2025 could reduce the number of new TB cases to 170 per 100,000 people. If we can increase active case finding even more - say by 30%, 70%, or 90% - we could bring the new cases down to 148, 120, or even just 44 per 100,000 people, respectively.

Now, let's take Rajasthan as an example. This is a state where we already have established partnerships. Rajasthan is home to approximately 81.2 million people and carries about 7% of India's total TB burden. Currently, about 89% of confirmed TB cases are diagnosed using traditional methods like sputum smear microscopy (SSM), X-rays, and TB culture.

Here's where our technology comes in. If we replace just 30% of the tests currently conducted at primary health centers (PHCs) with our rapid, accurate technology, we could correctly diagnose an estimated 44,954 TB patients in our first year of operation alone. This would be a substantial step towards our ultimate goal of a TB-free India.

The cartridge is powered by our proprietary polymer system that concentrates bacteria to one spot. The cationic polymer system, poly diallyldimethylammonium chloride (pDADMAC) has previously been shown to have a high affinity for mycobacterium, for example [1-4]. Our recent presentation at the Union Lung Conference in Paris included results of our study in India showing the efficacy of using this system [0].

The reader makes use of a polarization imaging technique to probe the microstructural and optical information of TB bacterial colonies [5]. The system processes and reads unique bacterial signatures from the cartridge, which are then processed and analyzed by comparing it to an existing library of bacterial signatures using ML/AI to return results.

[0] OA23-384-16 https://conf2023.theunion.org/...

[1] Albert H, Ademun PJ, Lukyamuzi G, et al. Feasibility of magnetic bead technology for concentration of mycobacteria in sputum prior to fluorescence microscopy. BMC infectious diseases 2011;11:125.

[2] Liu J, Sun ZQ, Pei H, et al. Increased case finding of tuberculosis from sputum and sputum deposits after magnetic bead concentration of mycobacteria. Journal of microbiological methods 2013;93:144-7.

[3] Wilson S, Lane A, Rosedale R, Stanley C. Concentration of Mycobacterium tuberculosis from sputum using ligand-coated magnetic beads. Int J Tuberc Lung Dis 2010;14:1164-8. 

[4] Tadesse M, Aragaw D, Rigouts L, Abebe G. Increased detection of smear-negative pulmonary tuberculosis by GeneXpert MTB/RIF(R) assay after bleach concentration. International journal of mycobacteriology 2016;5:211-8.

[5] Badieyan, Saeedesadat, et al. "Detection and discrimination of bacterial colonies with Mueller matrix imaging." Scientific reports 8.1 (2018): 1-10.

Digvijay Singh, MSE; CEO at Drizzle Health;7+ years of consumer product design, IVD development and consulting; 

Bonolo Mathekga, MSE; CTO at Drizzle Health; 4+ years of IVD development and consulting;

Nicola Peserico, PhD; Research Assistant Professor at University of Florida, 5+ years of optics design experience in industry; 

Nabeel Seedat, MS; Machine Learning Researcher at University of Cambridge,3+ years of Deep Learning, development & deployment;

Yukari Manabe, MD; Johns Hopkins University School of Medicine; 20+ years in infectious disease research in resource limited settings; 

Mack Schermer, BS; 30+ years of IVD development including R&D Director at Perkin Elmer;

+interns, contractors

4 years

As a team, we firmly believe in acknowledging our differences in race, color, sexual orientation, gender identity, religion, disability, age, genetic information, veteran status, ancestry, and national or ethnic origin. It is only by acknowledging what makes us different, that we are able to truly embrace and celebrate our diversity. These differences govern how we uniquely experience the world, and motivate our decisions on when and how we access healthcare, the type of diseases endemic in our communities, and the medications available to us. 

Our goal is to fight TB and the destruction it is causing millions of lives around the globe each year. We understand that this is a fight where we need all the help we can get and need to rely on the best of each other.

The team's co-founders have differing backgrounds. Digvijay is an Indian national, hailing from the national capital region. Bonolo is South African and named the product Medupi, a Sepedi word that means gentle rain. Medupi is considered a blessing, as it is pleasant rain that soaks parched land, yielding a plentiful harvest. This aligns well with what we imagine equitable healthcare to look like. Rain affects everyone, regardless of socio-economic status, cultural, and identity-based human attributes. Similarly, our goal is to make healthcare accessible to everyone, regardless of socio-economic status, cultural, and identity-based human attributes. 

This spirit is reflected in our current team and extended team, and will guide us as we expand. 

In the short term (12-15 months), our focus is to roll the products out for TB use case first in India.

Quite simply, we're charging per device sold for MagnaSlide ($2), per device sold for Medupi ($2,000), and subscription for our AI bacterial recognition software based on use across TB and food safety (including here in the US).

For TB in India:

There's 2 main customers for us in India : The private diagnostic lab chains, and the governments (central and state, who each run TB programs). Our first point of attack are private diagnostic lab chains, where the product will be sold as a lab use only device (LDT). This aligns with our regulatory submission to the Indian authorities (CDSCO), which we expect to complete by the end of this year. This way we can get revenue even as we wait for regulatory approval. Once we receive (our Class A) approval, we'll sell to state governments first (where we've already established contact with Rajasthan and Uttar Pradesh).

Currently, we've working to service validation pilot prior to purchase requests from several regional and national private diagnostic lab chains and expect to realize revenues starting Q4 this year.

The sales effort will be initiated in the following manner, beginning with use for smear microscopy : Private Diagnostic Labs (LDT, Q3 2024), State and Central Govts (Class A, Q1 2025), WHO-PQ (Class 1, Q1 2026). Current TB Market is about 19 million smear microscopies/year in India (up from 15 million the year before), and about 80 million worldwide. As an LDT (which does not require regulatory approval, only internal validation with each private lab chain), we're targeting about 5 million tests initially (which is consolidated amongst top 15 providers, national and regional) by 2026. Simultaneously, we're submitting our Class A (lowest regulatory burden, as confirmed by the regulatory audit our regulatory partner did) for government approval in India, as well as WHO-Pre Qualification. The 7 high burden states (including Rajasthan, Uttar Pradesh and Tamil Nadu where we did our study) conduct about 9 million tests, who'll be our primary targets. WHO-PQ opens up the global market of about 80m microscopy tests, including the consolidated orgs (such as the 5-20 million through PEPFAR program of the US Government/USAID, as well as STOP TB and Global Fund) - we have partners advising us on the process. Overall, private first is the fastest path to market and in the hands of people, and helps us have sustainable business.

R&D and initial pilot lot manufacturing for MagnaSlide has fully funded through non-dilutive capital including over $100,000 from the co-founders. R&D for Medupi has also been non-dilutive funded, including a recent $100,000 grant from Emergent Ventures, which will also be used for initial testing in largescale food and water environments.

We expect to be profitable by 2025, with an expected recurring revenue of $20 million by Q4 2026.