Artificial intelligence enabled parabolic response surface platform identifies ultra-rapid near-universal TB drug treatment regimens comprising approved drugs.

BACKGROUND: Shorter, more effective treatments for tuberculosis (TB) are urgently needed. While many TB drugs are available, identification of the best regimens is challenging because of the large number of possible drug-dose combinations. We have found consistently that responses of cells or whole animals to drug-dose stimulations fit a parabolic response surface (PRS), allowing us to identify and optimize the best drug combinations by testing only a small fraction of the total search space. Previously, we used PRS methodology to identify three regimens (PRS Regimens I-III) that in murine models are much more effective than the standard regimen used to treat TB. However, PRS Regimens I and II are unsuitable for treating drug-resistant TB and PRS Regimen III includes an experimental drug. Here, we use PRS methodology to identify from an expanded pool of drugs new highly effective near-universal drug regimens comprising only approved drugs. METHODS AND FINDINGS: We evaluated combinations of 15 different drugs in a human macrophage TB model and identified the most promising 4-drug combinations. We then tested 14 of these combinations in Mycobacterium tuberculosis-infected BALB/c mice and chose for PRS dose optimization and further study the two most potent regimens, designated PRS Regimens IV and V, consisting of clofazimine (CFZ), bedaquiline (BDQ), pyrazinamide (PZA), and either amoxicillin/clavulanate (AC) or delamanid (DLM), respectively. We then evaluated the efficacy in mice of optimized PRS Regimens IV and V, as well as a 3-drug regimen, PRS Regimen VI (CFZ, BDQ, and PZA), and compared their efficacy to PRS Regimen III (CFZ, BDQ, PZA, and SQ109), previously shown to reduce the time to achieve relapse-free cure in mice by 80% compared with the Standard Regimen (isoniazid, rifampicin, PZA, and ethambutol). Efficacy measurements included early bactericidal activity, time to lung sterilization, and time to relapse-free cure. PRS Regimens III-VI all rapidly sterilized the lungs and achieved relapse-free cure in 3 weeks (PRS Regimens III, V, and VI) or 5 weeks (PRS Regimen IV). In contrast, mice treated with the Standard Regimen still had high numbers of bacteria in their lungs after 6-weeks treatment and none achieved relapse-free cure in this time-period. CONCLUSIONS: We have identified three new regimens that rapidly sterilize the lungs of mice and dramatically shorten the time required to achieve relapse-free cure. All mouse drug doses in these regimens extrapolate to doses that are readily achievable in humans. Because PRS Regimens IV and V contain only one first line drug (PZA) and exclude fluoroquinolones and aminoglycosides, they should be effective against most TB cases that are multidrug resistant (MDR-TB) and many that are extensively drug-resistant (XDR-TB). Hence, these regimens have potential to shorten dramatically the time required for treatment of both drug-sensitive and drug-resistant TB. If clinical trials confirm that these regimens dramatically shorten the time required to achieve relapse-free cure in humans, then this radically shortened treatment has the potential to improve treatment compliance, decrease the emergence of drug resistance, and decrease the healthcare burden of treating both drug-sensitive and drug-resistant TB.

Ultra-rapid near universal TB drug regimen identified via parabolic response surface platform cures mice of both conventional and high susceptibility.

As current treatment of tuberculosis is burdensomely long, provoking non-adherence and drug resistance, effective short-course treatments are needed. Using the output-driven parabolic response surface (PRS) platform, we have identified drug regimens that treat tuberculosis more rapidly in mice than the current Standard Regimen used in humans. We show that PRS Regimen III, comprising clofazimine, SQ109, bedaquiline and pyrazinamide, rapidly sterilizes the lung both in conventionally studied BALB/c mice and in C3HeB/FeJ mice, highly susceptible mice that develop massive necrotic granulomatous lung lesions akin to those in humans, achieving relapse-free cure in only 4 weeks (p<0.0001 versus Standard Regimen). In contrast, the Standard Regimen required 16 weeks to attain lung culture negative status and 20 weeks to achieve relapse-free cure. Thus, PRS Regimen III dramatically cuts by ~80% the time to relapse-free cure in mouse tuberculosis models. PRS Regimen III, with three nonstandard drugs, can potentially treat both drug-sensitive and most drug-resistant tuberculosis.

Optimizing drug combinations against multiple myeloma using a quadratic phenotypic optimization platform (QPOP).

Multiple myeloma is an incurable hematological malignancy that relies on drug combinations for first and secondary lines of treatment. The inclusion of proteasome inhibitors, such as bortezomib, into these combination regimens has improved median survival. Resistance to bortezomib, however, is a common occurrence that ultimately contributes to treatment failure, and there remains a need to identify improved drug combinations. We developed the quadratic phenotypic optimization platform (QPOP) to optimize treatment combinations selected from a candidate pool of 114 approved drugs. QPOP uses quadratic surfaces to model the biological effects of drug combinations to identify effective drug combinations without reference to molecular mechanisms or predetermined drug synergy data. Applying QPOP to bortezomib-resistant multiple myeloma cell lines determined the drug combinations that collectively optimized treatment efficacy. We found that these combinations acted by reversing the DNA methylation and tumor suppressor silencing that often occur after acquired bortezomib resistance in multiple myeloma. Successive application of QPOP on a xenograft mouse model further optimized the dosages of each drug within a given combination while minimizing overall toxicity in vivo, and application of QPOP to ex vivo multiple myeloma patient samples optimized drug combinations in patient-specific contexts.

Drug regimens identified and optimized by output-driven platform markedly reduce tuberculosis treatment time.

The current drug regimens for treating tuberculosis are lengthy and onerous, and hence complicated by poor adherence leading to drug resistance and disease relapse. Previously, using an output-driven optimization platform and an in vitro macrophage model of Mycobacterium tuberculosis infection, we identified several experimental drug regimens among billions of possible drug-dose combinations that outperform the current standard regimen. Here we use this platform to optimize the in vivo drug doses of two of these regimens in a mouse model of pulmonary tuberculosis. The experimental regimens kill M. tuberculosis much more rapidly than the standard regimen and reduce treatment time to relapse-free cure by 75%. Thus, these regimens have the potential to provide a markedly shorter course of treatment for tuberculosis in humans. As these regimens omit isoniazid, rifampicin, fluoroquinolones and injectable aminoglycosides, they would be suitable for treating many cases of multidrug and extensively drug-resistant tuberculosis.

Individualizing liver transplant immunosuppression using a phenotypic personalized medicine platform.

Posttransplant immunosuppressive drugs such as tacrolimus have narrow therapeutic ranges. Inter- and intraindividual variability in dosing requirements conventionally use physician-guided titrated drug administration, which results in frequent deviations from the target trough ranges, particularly during the critical postoperative phase. There is a clear need for personalized management of posttransplant regimens to prevent adverse events and allow the patient to be discharged sooner. We have developed the parabolic personalized dosing (PPD) platform, which is a surface represented by a second-order algebraic equation with experimentally determined coefficients of the equation being unique to each patient. PPD uses clinical data, including blood concentrations of tacrolimus--the primary phenotypic readout for immunosuppression efficacy--to calibrate these coefficients and pinpoint the optimal doses that result in the desired patient-specific response. In this pilot randomized controlled trial, we compared four transplant patients prospectively treated with tacrolimus using PPD with four control patients treated according to the standard of care (physician guidance). Using phenotype to personalize tacrolimus dosing, PPD effectively managed patients by keeping tacrolimus blood trough levels within the target ranges. In a retrospective analysis of the control patients, PPD-optimized prednisone and tacrolimus dosing improved tacrolimus trough-level management and minimized the need to recalibrate dosing after regimen changes. PPD is independent of disease mechanism and is agnostic of indication and could therefore apply beyond transplant medicine to dosing for cancer, infectious diseases, and cardiovascular medicine, where patient response is variable and requires careful adjustments through optimized inputs.

Output-driven feedback system control platform optimizes combinatorial therapy of tuberculosis using a macrophage cell culture model.

Tuberculosis (TB) remains a major global public health problem, and improved treatments are needed to shorten duration of therapy, decrease disease burden, improve compliance, and combat emergence of drug resistance. Ideally, the most effective regimen would be identified by a systematic and comprehensive combinatorial search of large numbers of TB drugs. However, optimization of regimens by standard methods is challenging, especially as the number of drugs increases, because of the extremely large number of drug-dose combinations requiring testing. Herein, we used an optimization platform, feedback system control (FSC) methodology, to identify improved drug-dose combinations for TB treatment using a fluorescence-based human macrophage cell culture model of TB, in which macrophages are infected with isopropyl ß-D-1-thiogalactopyranoside (IPTG)-inducible green fluorescent protein (GFP)-expressing Mycobacterium tuberculosis (Mtb). On the basis of only a single screening test and three iterations, we identified highly efficacious three- and four-drug combinations. To verify the efficacy of these combinations, we further evaluated them using a methodologically independent assay for intramacrophage killing of Mtb; the optimized combinations showed greater efficacy than the current standard TB drug regimen. Surprisingly, all top three- and four-drug optimized regimens included the third-line drug clofazimine, and none included the first-line drugs isoniazid and rifampin, which had insignificant or antagonistic impacts on efficacy. Because top regimens also did not include a fluoroquinolone or aminoglycoside, they are potentially of use for treating many cases of multidrug- and extensively drug-resistant TB. Our study shows the power of an FSC platform to identify promising previously unidentified drug-dose combinations for treatment of TB.

Identification and Optimization of Combinatorial Glucose Metabolism Inhibitors in Hepatocellular Carcinomas.

Hepatocellular carcinoma (HCC) is the fifth most common cancer worldwide. The expression of glucose transporter isoform 1, a key factor in transporting glucose into cancer cells, is overexpressed in several human cancers, including HCC. In addition, this has been shown to correlate with a higher proliferation index and more advanced stages in HCC, suggesting that inhibition of glucose metabolism is a promising therapeutic strategy. Our study used high-content screening (HCS) for compounds that target glucose metabolism and effect cell death in HCC cells. Specifically, we showed that a fluorescent 2-deoxyglucose analog, 2-[N-(7-nitrobenz-2- oxa-1,3-diazol-4-yl)amino]-2-deoxyglucose, and CellTrace Calcein Red-Orange AM can be used reliably as readouts for glucose uptake and proliferative index, respectively, to identify drug candidates that simultaneously reduce glucose uptake and induce cell death in HCC cells. Thus, fluorescent glucose uptake bioprobes can be implemented in HCS assays to identify previously unknown regulators of glucose metabolism in HCC. In addition, our study also employs the use of feedback system control (FSC.II), a platform that optimizes the combinations of drugs identified through HCS. The coordinated use of HCS and FSC.II can improve the development of drug combinations and uncover previously unidentified signaling pathways that govern HCC as well as other cancers.

Mechanism-independent optimization of combinatorial nanodiamond and unmodified drug delivery using a phenotypically driven platform technology.

Combination chemotherapy can mediate drug synergy to improve treatment efficacy against a broad spectrum of cancers. However, conventional multidrug regimens are often additively determined, which have long been believed to enable good cancer-killing efficiency but are insufficient to address the nonlinearity in dosing. Despite improved clinical outcomes by combination treatment, multi-objective combination optimization, which takes into account tumor heterogeneity and balance of efficacy and toxicity, remains challenging given the sheer magnitude of the combinatorial dosing space. To enhance the properties of the therapeutic agents, the field of nanomedicine has realized novel drug delivery platforms that can enhance therapeutic efficacy and safety. However, optimal combination design that incorporates nanomedicine agents still faces the same hurdles as unmodified drug administration. The work reported here applied a powerful phenotypically driven platform, termed feedback system control (FSC), that systematically and rapidly converges upon a combination consisting of three nanodiamond-modified drugs and one unmodified drug that is simultaneously optimized for efficacy against multiple breast cancer cell lines and safety against multiple control cell lines. Specifically, the therapeutic window achieved from an optimally efficacious and safe nanomedicine combination was markedly higher compared to that of an optimized unmodified drug combination and nanodiamond monotherapy or unmodified drug administration. The phenotypically driven foundation of FSC implementation does not require any cellular signaling pathway data and innately accounts for population heterogeneity and nonlinear biological processes. Therefore, FSC is a broadly applicable platform for both nanotechnology-modified and unmodified therapeutic optimizations that represent a promising path toward phenotypic personalized medicine.

When Medicine Meets Engineering-Paradigm Shifts in Diagnostics and Therapeutics.

During the last two decades, the manufacturing techniques of microfluidics-based devices have been phenomenally advanced, offering unlimited potential for bio-medical technologies. However, the direct applications of these technologies toward diagnostics and therapeutics are still far from maturity. The present challenges lay at the interfaces between the engineering systems and the biocomplex systems. A precisely designed engineering system with narrow dynamic range is hard to seamlessly integrate with the adaptive biological system in order to achieve the design goals. These differences remain as the roadblock between two fundamentally non-compatible systems. This paper will not extensively review the existing microfluidic sensors and actuators; rather, we will discuss the sources of the gaps for integration. We will also introduce system interface technologies for bridging the differences to lead toward paradigm shifts in diagnostics and therapeutics.