|
Development of a long-acting direct-acting antiviral system for hepatitis C virus treatment in swine
 
|
|
|
Download the PDF here  
PNAS June 2, 2020  
Malvika Vermaa,b,c,1, Jacqueline N. Chub,d,1, John A. F. Salamab,1, Mohammed T. Faizb,1, Feyisope Ewejeb,e,f, Declan Gwynneb,g, Aaron Lopesb,g, Kaitlyn Hessb,g, Vance Soaresb,g, Christoph Steigerb,f,g, Rebecca McManusb,g, Ryan Koeppenb,e, Tiffany Huab,g, Alison Haywardb,g,h, Joy Collinsb,g, Siddartha M. Tamangb,g, Keiko Ishidab,g, Jonathan B. Milleri, Stephanie Katzj, Alexander H. Slocumc,e, Mark S. Sulkowskij, David L. Thomasj, Robert Langera,b,c,e,g,2, and Giovanni Traversob,c,e,f,2  
Significance  
Direct-acting antiviral (DAA) therapy is highly effective against hepatitis C virus (HCV). Despite this, the burden of chronic HCV remains high, particularly in populations for whom daily medications can be challenging. The generation of long-acting modes of DAA delivery could thus expand HCV treatment. We have developed a long-acting DAA system (LA-DAAS) that can support multigram-level drug depots of HCV DAAs and provide controlled release of those drugs over the course of weeks. We demonstrate initial safety and efficacy of the LA-DAAS in a swine model. Expanding the range of delivery options of DAAs to patients and healthcare providers will further efforts toward global elimination of HCV infection.  
Abstract  
Chronic hepatitis C virus (HCV) infection is a leading cause of cirrhosis worldwide and kills more Americans than 59 other infections, including HIV and tuberculosis, combined. While direct-acting antiviral (DAA) treatments are effective, limited uptake of therapy, particularly in high-risk groups, remains a substantial barrier to eliminating HCV. We developed a long-acting DAA system (LA-DAAS) capable of prolonged dosing and explored its cost-effectiveness. We designed a retrievable coil-shaped LA-DAAS compatible with nasogastric tube administration and the capacity to encapsulate and release gram levels of drugs while resident in the stomach. We formulated DAAs in drug-polymer pills and studied the release kinetics for 1 mo in vitro and in vivo in a swine model. The LA-DAAS was equipped with ethanol and temperature sensors linked via Bluetooth to a phone application to provide patient engagement. We then performed a cost-effectiveness analysis comparing LA-DAAS to DAA alone in various patient groups, including people who inject drugs. Tunable release kinetics of DAAs was enabled for 1 mo with drug-polymer pills in vitro, and the LA-DAAS safely and successfully provided at least month-long release of sofosbuvir in vivo. Temperature and alcohol sensors could interface with external sources for at least 1 mo. The LA-DAAS was cost-effective compared to DAA therapy alone in all groups considered (base case incremental cost-effectiveness ratio $39,800). We believe that the LA-DAA system can provide a cost-effective and patient-centric method for HCV treatment, including in high-risk populations who are currently undertreated.  
We assembled drug-polymer pills for multiple DAAs, including sofosbuvir, daclatasvir, ledipasvir, and ribavirin (SI Appendix, Table S2) (30↓↓↓-34). Drug release occurred via diffusion through channels in the polymer matrix (17, 27, 35). Several factors contributed to tuning the drug-release rate from the polymer matrix, including the molecular weight of PCL, coating of the pill, the drug-loading percentage in the polymer matrix, and the surface-area-to-volume ratio of the pill (Fig. 2 B-E and SI Appendix, Fig. S7) (17, 28, 36↓-38). The rate of sofosbuvir release increased as the molecular weight of PCL increased from 25,000 to 80,000. More crystallinity (low amorphous regions) is associated with lower PCL molecular weight (38, 39). Therefore, the higher molecular weight has a faster degradation rate due to the presence of more amorphous regions. Increasing the surface-area-to-volume ratio of the pill also increased the sofosbuvir release rate. Daclatasvir release was modulated by varying the molecular weight of PCL and coating the pill to reduce the burst release of drug, resulting in a linear cumulative release versus time profile. Ribavirin release increased at higher PCL molecular weight and with a higher drug-loading percentage. Increasing the drug-loading percentage of ledipasvir resulted in a faster release rate. Overall, the drug-polymer pills provided a method to tune the release rate of four DAAs for 1 mo.  
In Vivo Applications from the LA-DAAS.  
We next studied the long-term drug delivery of LA-DAAS in swine and evaluated its electronic-sensing capabilities. According to prior work we conducted in a swine model, the renally eliminated metabolite GS-331007 is the primary analyte of interest for clinical pharmacology studies with sofosbuvir (SI Appendix, Fig. S8) (40, 41). We prepared three LA-DAASs loaded with 11.2 g of sofosbuvir and administered them in swine. Each LA-DAAS was assembled to contain 300 pills using two different formulations, which released drug simultaneously (SI Appendix, Table S3). The serum concentration profile of a 400-mg single dose is shown in Fig. 3A. The drug was absorbed rapidly, and detectable concentrations were observed within 3 h. No drug was detectable after 1 d with the single-dose formulation. In contrast, drug was detectable for at least 28 d when sofosbuvir was dosed in the LA-DAAS (Fig. 3B). A 22.4-g LA-DAAS showed ongoing high levels of sofosbuvir release on day 30, suggestive of greater than 1-mo drug delivery capability (SI Appendix, Table S4 and Fig. S9).  
After 1 mo of gastric residence in vivo, the LA-DAAS was retrieved, and the amount of sofosbuvir remaining in the pills was studied (SI Appendix, Fig. S10). The sofosbuvir-PCL pills were completely dissolved, and the amount of sofosbuvir was quantified by using high-performance liquid chromatography (HPLC). These pills contained 15 to 50% of the initial sofosbuvir drug load prior to dosing in vivo. The addition of excipients, such as Pluronic P407 or poly(ethylene glycol), to the drug-PCL matrix increased the release rate of sofosbuvir (SI Appendix, Fig. S11) (28).  
The LA-DAAS was equipped with a temperature sensor linked via Bluetooth to a phone application, as well as an alcohol sensor transmitting analog data to an external microcontroller. Batteries were included at the end of the LA-DAAS to provide power to the sensors (Fig. 3C).  
Hepatitis C virus (HCV) infection is one of the main causes of chronic liver disease worldwide (1↓-3). One forecast predicted that mortality from chronic hepatitis would exceed the mortality caused by HIV, tuberculosis (TB), and malaria combined by 2040 (4). Accordingly, the World Health Organization has called for elimination of HCV infection by 2030 (5, 6).  
Encouragingly, the development of direct-acting antiviral (DAA) therapy for HCV has revolutionized the field. The high cure rate of DAA therapies, with rates of sustained virologic response (SVR) of greater than 95%, short duration of treatment, and tolerability, has greatly expanded the number of patients being cured of chronic HCV (1, 2, 7, 8). However, due to factors such as cost, limited access to medical care, and inconsistent follow-up, less than half of those infected are actually treated (3, 8↓-10). People who inject drugs (PWID) have especially low HCV treatment uptake, despite being a large reservoir for chronic HCV infection (50% of all PWID globally) and the leading population infected with acute HCV (9, 11).  
Particular challenges facing the PWID population include homelessness, which can make storage of oral medications difficult; comorbid psychiatric conditions; limited access to healthcare; and distrust of the medical system (3). Modeling studies have suggested that expanding treatment to PWID can reduce transmission of HCV, even with poor rates of adherence and SVR (12). Studies utilizing methadone programs to deliver HCV treatment show promise, with high rates of treatment completion and SVR (3). However, such programs require significant infrastructure and cost and do not capture PWID who do not engage in other opioid agonist therapies such as buprenorphine or who may engage only intermittently (3, 13). Challenges that all populations face with regard to oral therapies include forgetfulness, low priority, pill phobia, dysphagia, and pill burden (14); rates of adherence to chronic medications in general are, at most, 50%, though higher for short courses of therapy such as DAAs (14). Thus, there is a clear need for alternative strategies to deliver HCV treatment, and long-acting treatment delivery systems are one of the innovative strategies that have been endorsed, including recently by Unitaid (15, 16).
A long-acting DAA system (LA-DAAS) has the potential to safely and consistently deliver DAA therapy to patients with good adherence and quality of life by decreasing the frequency of dosing, with the goal of ultimately providing single dosing (14, 15). We have previously described a coil-shaped gastric resident system (GRS) which, due to its dimensions and mechanical properties, can remain resident in the gastric cavity and deliver multigram levels of drugs for TB, which has similar treatment challenges as HCV with frequent and daily dosing (17↓-19). Here, we describe the development of a LA-DAAS inspired by the multigram GRS with evaluation in a swine model of chronic HCV. We developed additional features, including integration of ethanol, temperature, and Bluetooth sensors, to enable patient engagement via wireless communication with the LA-DAAS. Such tools have improved patient adherence to HCV therapy in other preliminary devices (20, 21). We then evaluated the cost-effectiveness of the LA-DAAS in general and PWID populations.  
Discussion  
With the development of effective and well-tolerated DAA therapy, it is now possible to cure hepatitis C in all patients. In this study, we describe a drug delivery system that can safely and effectively provide at least month-long doses of DAA therapy for HCV. Long-acting approaches to the treatment of HIV and TB have been previously developed, and here we describe a long-acting enteral solution to the challenge of expanding HCV treatment (17, 44, 45). These data provide a basis for expanded research to adapt the technology to treatment of HCV-infected humans, including improving the system to ultimately achieve single-dose administration of HCV treatment.  
Treatment of HCV infection historically involved the use of a pegylated form of interferon (IFN) alfa that prolonged half-life and allowed weekly dosing. An albumin-conjugated form of IFN with an even longer half-life was also considered (46). However, since the advent of DAA formulations, all treatment requires daily oral administration. We have now developed a long-acting enteral solution to enable monthly HCV treatment.  
We anticipate that further development of the LA-DAAS will include further preclinical evaluation in additional animal models. Optimizing release kinetics from the drug-polymer pills is a critical next step, such that serum concentrations of the drug remain within the therapeutic window for efficacious treatment with maximal release from the pills. We will also want to more closely mimic the typical regimens used in humans, including pangenotypic DAA formulations such as glecaprevir/pibrentasvir and study of daclatasvir or velpatasvir with sofosbuvir. Furthermore, based on our current data, future evaluation of prolonged gastric residency (>1 mo) and single administration of pangenotypic drug combinations is warranted. Additionally, long-term safety and stability of the electronic sensors must be assessed in the gastric environment, as well as long-term presence of the LA-DAA, for anticipated retrieval failure. Other considerations include whether nasogastric delivery of the LA-DAAS may be a challenge; although inexpensive, patients may find this delivery system uncomfortable. In such cases, endoscopic delivery may be preferable and would also allow for greater volume of drug delivery and reduced dosing frequency.  
Our economic analysis found that the LA-DAAS can be cost-effective across various patient populations, including both the general population and PWID. Although the LA-DAAS was generally preferred over DAA therapy in most scenarios tested, we identified several factors that could make conventional DAA therapy preferable to LA-DAAS: patients who find nasogastric delivery very unpleasant; high cost of the LA-DAAS or of its administration; short DAA treatment courses; and patients unlikely to return monthly for repeat LA-DAAS administration. This latter limitation could be particularly challenging in some settings. In others, such as in the correctional system, the LA-DAAS might be preferred over DAA therapy. For patients for whom nasogastric delivery is markedly unpleasant, one could consider administration under sedation or endoscopically, although these costs were not analyzed in this model. Other populations that could benefit from this system include intubated and ventilated patients, such as posttransplant patients in the advent of HCV+ organ transplantation, particularly those requiring DAAs such as glecaprevir/pibrentasvir that cannot be crushed and administered through typical enteral access.  
There are a number of limitations to our cost-effectiveness study. As with any modeling analysis, the accuracy of our model was restricted by available data. In particular, adherence rates, the impact of adherence on SVR, and the likelihood of returning for a repeated LA-DAAS administration were challenging to estimate. Therefore, we performed wide sensitivity analyses for these parameters. Our model did not capture the utility of reduced HCV transmission among PWID, which likely underestimates the overall benefit and cost-effectiveness of the LA-DAAS (11). Additionally, we did not include probability of reinfection in our model. If patients in one of the treatment strategies were reinfected more often than those in the other strategy, that would likely increase overall treatment costs in that arm and affect our calculated ICER. However, we do not expect that this would differ significantly between patients treated with DAA therapy versus the LA-DAAS. Finally, our model used the same drug cost for conventional DAA therapy and the cost of DAA included within the LA-DAAS (separate from device and administration costs). This assumes that further lowering of drug cost for DAAs will be applicable to the LA-DAAS as well. If there are heavy discounts applied to oral DAA therapy that are not included in pricing of the LA-DAAS, then the LA-DAAS could become cost-ineffective.  
Ultimately, whether any given patient is more likely to adhere to or prefer daily pills versus monthly repeated nasogastric treatment will vary depending on the patient, and patient preference regarding modality of HCV treatment should guide choice of therapy. Further steps in development of the LA-DAAS should include engaging patients and healthcare providers in the hepatitis C community to evaluate the acceptability and feasibility of such an approach, as well as ongoing work to move toward single-dose administration for full HCV treatment. We believe that providing a range of delivery options to patients and healthcare providers will aid in maximizing deployment of recognized effective DAAs.
| |
|
|
|
|
|