Abstract

New viral drug development can take up to 15 to 20 years to reach the clinic, while drugs during development that display intense adverse drug reactions or signs of drug ineffectiveness can be discontinued at any time. This unreliability of developing new drugs explains the costly and long process of drug development, which is an unsuitable process for tackling acute onset diseases. As a result, several outbreaks of viral disease have resulted in millions of deaths, especially in developing countries. Drug development for these outbreaks have been only partially successful, but there is a growing need for effective anti-viral drugs to counter outbreaks. One way this may be achieved is through repurposing previously approved drugs which target similar aspects of the virus. This strategy has the advantage of pre-clinical work, with factors such as dosage and adverse drug reaction previously identified and studied. The purpose of this paper is to identify why to repurpose viral therapies, how to repurpose them, and future implications of repurposing.

Introduction

Repurposing drugs is a strategy where scientists allocate an already-approved drug for reasons other than its original purpose. Discovering a novel drug from initiation to approval can take up to 25 years, proving to be a costly and inefficient way to allocate drugs to illnesses (McNamee et al.) The time-consuming process of drug discovery and approval results in only a few anti-viral drugs being approved over the past decade, making it a costly way for pharmaceutical companies to seek profit (De Clercq and Li). Due to the pharmaceutical industry’s profit-driven incentives, companies have pursued a more efficient path: repurposing drugs (Cha et al.).

Repurposing drugs is essential in today’s drug market in aiding the pharmaceutical industry to assign drugs to illnesses which require timely and inexpensive drug-approval processes. Such events usually involve an acute viral outbreak, in which a reliable drug must be available for treatment as early as possible. Production of a novel drug may be too time consuming during a viral outbreak. The extensive supporting studies used to determine the safety of the drug are time consuming and by comparison, repurposing a drug has a shorter timeline because fewer supporting studies and less research are required (Hernandez et al.). Further, repurposed drugs may be more reliable, since vigorous studies have been conducted previously to determine the safety of the drug and the drug is likely to have been tested in the public domain (Pan et al.). Repurposed drugs may prove beneficial in orphan viral illnesses (diseases found in less than 200,000 patients) as well. Due to the rare occurrence of orphan viral illnesses, pharmaceutical companies are unincentivized to discover novel drugs due ratio of cost of development to patient number and the high levels of expertise required. Therefore, repurposed drugs may be the only available treatment (GJ) and thus, repurposing a drug may be favored.

Developing countries may also find repurposing drugs useful. Due to a lack of available funding needed to start drug programs nor sufficient money to attain drugs from pharmaceutical companies, developing countries lack the ability for new-drug discovery. Thus, developing countries may find it plausible to repurpose drugs. Since the prevalence of viruses such as Zika, Ebola, or Malaria in developing countries is relatively higher, it is essential for a timely anti-viral to control virus spread since acute viruses spread dramatically without interference, thus requiring repurposing drugs.

Viral Illnesses

Viruses are those that generally cause acute illnesses, often resulting into serious problems or death. Common viral illnesses include HIV/AIDS and COVID-19, which both begin with acute symptoms, but can result into extreme symptoms or death. Viruses are also more difficult to treat, and unlike bacteria, they cannot be directly eliminated by broad-spectrum antibiotics. Rather, targeted anti-virals are required for viral illnesses, targeting specific life parts of the viral life cycle.

As obligate intracellular parasites, viruses can only reproduce by infecting a host cell and utilizing the reproduction machinery in the host cell to produce copies of itself.

Viral life cycles vary according to specific viruses (Ryu), but usually follow a pattern of:

1. Attachment
2. Penetration
3. Uncoating
4. Release
5. Spread

1. Attachment

The virus contacts the surface of the host cell, involving attachment factors and viral receptors. Attachment factors facilitates the attachment between the virus and the surface of the cell, while viral receptors facilitate penetration of the virus.

2. Penetration

After attachment, viruses penetrate into the cytoplasm of the host cell using one of two mechanisms: direct fusion or receptor mediated endocytosis. Direct fusion is when the cell membrane of a virus fuses with the membrane of the host cell, while receptor-mediated endocytosis utilizes membrane receptors, forming a coat pit and forming an endosome

3. Uncoating

In order for the virus to perform gene expression, the genetic material of the virus needs to be exposed to the host cell, a processed named uncoating. The uncoating is generally done by an increase in cations in the cytoplasm or the acidic state of the endosome.

4. Release

Once replication is completed, the virus exists the host cell by exocytosis, budding, or cell lysis. Budding allows the virus to use the host cell’s membrane, forming an envelope for the virus. Cell lysis bursts the cell membrane, releasing the virus but killing the host cell. Exocytosis is when vesicles are excreted out of the cell.

5. Spread

The infectious ability of the virus is a large determinant of the magnitude of spread. Contrary to times when travel was restricted to walking distance, modern mass transport has allowed viruses to spread from continent to continent. To determine contagiousness, R0 is used as a scale to show how easily a virus may spread from person to person. Thus, a virus with high infectious ability, or R0, can be extremely easy to spread, given mass transport.

Prevention and Treatment of Viruses

The most common and effective way of preventing a virus is by immunization. Over the past decade, more than one billion children have been immunized, preventing almost three million deaths each year (world health). Although vaccines are extremely effective, with some effective to 95% of recipients (world health), the development of a long-term effective vaccine is dependent on the category of virus. There are two main subcategories of viruses important to the development of vaccines: DNA versus RNA viruses.

DNA viruses such as smallpox or polio are more stable and unsusceptible to change, which allows targeted and effective vaccines. However, RNA viruses are very unstable, and mutate rapidly due to a lack of proofreading of RNA virus polymerase. This error changes the hemagglutinin and neuraminidase surface proteins of the virus, a process named the “antigenic drift.” Antibodies developed from the previous vaccine would have recognized specific antigens, rendering the previous vaccination obsolete with even a small alteration in the surface proteins. Thus, the development process of vaccinations for RNA vaccines is much more difficult, as they have to. To maximize effectiveness, vaccines may be re-engineered after each time period. For example, influenza, an RNA virus, requires a new vaccine each year due to the mutations in the coat protein of the virus.

However, the method of developing a new vaccine each year is costly, especially for illnesses that may not have the magnitude of impact the influenza virus does. Due to the spread of influenza, which may infect up to one billion people each year an, there is an influx of funding is given for research and development of a vaccine (Clayville). Conversely, viruses that impact a small percent of society will have less funding towards treatment, and the development of a new vaccine is less plausible.

In instances where a vaccine is unavailable or ineffective, antivirals can be used as a form of treatment after infection. Antivirals are compounds that interfere with the replication of a virus, stopping the further spread of the virus in the body. However, because viruses live inside the host cell in order to replicate, antivirals which interferes reproduction in viruses also interferes with reproduction of the host cell. Thus, there may be severe adverse drug reactions to different forms of antivirals.

How to Repurpose?

As RNA viruses are those that are difficult to raise a vaccine against, and thus have no concrete preventative measures, rapid repurposing of viral therapies is imperative for novel RNA viruses. As novel viruses emerge, there is a rapid need for viral therapies in order to combat the spread of a virus. However, the process of developing of a new drug to FDA approval can take up to ten years, proving as an ineffective way to quickly contain the spread of a virus (Department of Health and Human Services). Thus, as a novel RNA virus emerges, repurposing drugs that have not only shown efficacy in similar viruses but also have manageable adverse drug reactions is the most effective way of treatment.

In interfering with the replication of viruses, antivirals target the stages in the life cycle of a virus. There are 3 major areas where antivirals effectively stop a viral infection:

1. Entrance
2. Uncoating
3. Replication

1. Entrance

Antivirals such as Camostat Mesylate and Nafamostat Mesylate are serine protease inhibitors, which bind to viral proteases. Thus, viruses are unable to bind to viral proteases and are unable to entrance the host cell. This prevents endocytosis, and thus prevents replication of the virus.

2. Uncoating

Viruses need to have genetic information exposed to cellular machinery in order to replicate, and thus the virus needs to be uncoated. However, anti malaria drugs such as chloroquine and hydroxychloroquine interfere with the pH of the endocytic vacuole. Being basic , at high enough concentrations they can raise the pH of the vacuole environment thereby preventing viral uncoating which requires an acidic pH.

3. Replication

Even after the virus enters the cell and goes through uncoating mechanisms, there is one last option to halt the spread of the virus. Remdisivir is an example of an antiviral that inhibits replication of the virus and has even been repurposed as treatment for COVID-19. It works by mimicking a nucleotide and is incorporated into the growing RNA chain. Once incorporated it inhibits the addition of new nucleotides thus stopping replication.

The Future of Repurposing:

As novel viruses emerge, more and more antivirals need to be repurposed in order for a timely control in the magnitude of spread. Although it is plausible for new viral therapies or vaccines to be developed, the time frame in which new drugs are developed is too long for the containment of a virus.

This strategy can be seen in the rush to develop treatments for COVID-19. Even with unprecedented measures- over 23 pharmaceutical companies participating in vaccine research and 4 phase-3 trials in the working, scientists still predict a possible treatment to be available more than 1 year after the discovery of COVID-19 (Department of Health and Human Services). However, there are promising results in repurposing viral treatments such as remdisivir and hydroxychloroquine during the wait for a vaccine. Thus, future large-scale pandemics should learn from this pandemic and not only work to develop a vaccine but work to repurpose approved antivirals.

Conclusion

One of the largest inequities in the world is healthcare and its implications can be seen worldwide. Wealthy countries have seen major developments in drug research in healthcare while underdeveloped countries suffer with poor healthcare services and almost no drug development. Unfortunately, these underdeveloped countries are ones that suffer with viruses that harvest acute viral diseases due to poor sanitation and lack of funding.

Not only are underdeveloped countries unable to afford research for novel drug development, but global pandemics such as the spread of COVID-19 are in need of drugs to combat viral illnesses. Thus, repurposing viral therapies should be concerned as an alternative to novel drug development.

Works Cited

Cha, Y, et al. “Drug Repurposing from the Perspective of Pharmaceutical Companies.” British Journal of Pharmacology, John Wiley and Sons Inc., Jan. 2018, www.ncbi.nlm.nih.gov/pmc/articles/PMC5758385/.

Clayville, Lisa R. “Influenza Update: a Review of Currently Available Vaccines.” P & T : a Peer-Reviewed Journal for Formulary Management, MediMedia USA, Inc., Oct. 2011, www.ncbi.nlm.nih.gov/pmc/articles/PMC3278149/.

De Clercq, Erik, and Guangdi Li. “Approved Antiviral Drugs over the Past 50 Years.” Clinical Microbiology Reviews, American Society for Microbiology, July 2016, www.ncbi.nlm.nih.gov/pmc/articles/PMC4978613.

“Fourth Large-Scale COVID-19 Vaccine Trial Begins in the United States.” National Institutes of Health, U.S. Department of Health and Human Services, 23 Sept. 2020, www.nih.gov/news-events/news-releases/fourth-large-scale-covid-19-vaccine-trial-begins- united-states.

GJ;, Brewer. “Drug Development for Orphan Diseases in the Context of Personalized Medicine.” Translational Research : the Journal of Laboratory and Clinical Medicine, U.S. National Library of Medicine, pubmed.ncbi.nlm.nih.gov/19931198/.

Hernandez, J Javier, et al. “Giving Drugs a Second Chance: Overcoming Regulatory and Financial Hurdles in Repurposing Approved Drugs As Cancer Therapeutics.” Frontiers in Oncology, Frontiers Media S.A., 14 Nov. 2017, www.ncbi.nlm.nih.gov/pmc/articles/PMC5694537/.

“How Are Drugs Approved for Use in the United States?” Eunice Kennedy Shriver National Institute of Child Health and Human Development, U.S. Department of Health and Human Services, www.nichd.nih.gov/health/topics/pharma/conditioninfo/approval.

McNamee, Laura M, et al. “Timelines of Translational Science: From Technology Initiation to FDA Approval.” PloS One, Public Library of Science, 8 May 2017, www.ncbi.nlm.nih.gov/pmc/articles/PMC5421779/).

Pantziarka , Pan, et al. “The Repurposing Drugs in Oncology (ReDO) Project.” Ecancermedicalscience, U.S. National Library of Medicine, 2014, pubmed.ncbi.nlm.nih.gov/25075216/.

Pantziarka P;Bouche G;Meheus L;Sukhatme V;Sukhatme VP;Vikas P; “The Repurposing Drugs in Oncology (ReDO) Project.” Ecancermedicalscience, U.S. National Library of Medicine, pubmed.ncbi.nlm.nih.gov/25075216/.

About the author

Nicole Zhou

Nicole is a junior at the Shanghai American School.

The post The Past, Present, and Future of Repurposing Viral Therapies appeared first on Exploratio Journal.

Abstract

Pharmaceutical Preparation Manufacturing U.S. industry comprises establishments primarily engaged in manufacturing in-vivo diagnostic substances and pharmaceutical preparations (except biological) intended for internal and external consumption in dose forms, such as ampoules, tablets, capsules, vials, ointments, powders, solutions, and suspensions.

The Producer Price Index (PPI) measures the change over time in the prices received by domestic producers of goods and services. More detailed indexes are used as sources for industry analysis and contract escalation in the public and private sectors.

PPI data are critical inputs into the development of sensitive economic indicators, including estimates of gross domestic product and industrial productivity. 

Pharmaceutical companies have deep scientific knowledge gained from decades of experience with several viruses. Companies are researching vaccine candidates and undertaking inventories of research portfolio libraries to identify additional potential treatments for R&D to combat COVID-19.

 Some are exploring ways to use existing technologies that provide the ability to rapidly upscale production once a potential vaccine candidate is identified.

So, it becomes crucial to estimate the value of Producer Price Index of Pharmaceutical Industry: Preparation and Manufacturing in this pandemic.

1. Introduction

1.1  Producer Price Index

The Producer Price Index (PPI) program measures the average change over time in the selling prices received by domestic producers for their output. The prices included in the PPI are from the first commercial transaction for many products and some services.

The data of Producer Price Index of Pharmaceutical preparation and manufacturing industry is obtained from ‘FRED’. The obtained data has monthly frequency starting from June 1981. This particular data has been utilized for our project.

1.2  Analysing Full Series

The plot for the full series shows that there is a significant increase in producer price index in March 1998 from the previous month. 

The plot shows increasing trend overall indicating a non-stationary trend. We have used “ndiffs” function of “forecast” library  to estimate the number of differences required to make the time series stationary. The result was 2. 

1.3 Second Order Differencing

In March 1998, the Food and Drug Administration (FDA) approved usage of few drugs resulting in significant increase of Producer price index from 265.0 to 291.0

This led the ndiffs function to estimate the second order of differencing to make the series stationary.

Later, we broke the entire timeline into two parts to exclude March 1998 period. 

• First 10 years
• Last 10 years.

Then, the ndiffs function indicated first order of differencing for both the time periods.

2. First 10 Years

2.1  Auto Regressive Models

The data of first 10 years is converted to first order difference log data. We look for the best Auto Regressive model plotting the Akaike Information Criterion (AIC) for values up to 17.

We see that AR model 14 has lower AIC value. Increasing the autoregressive order from 10 improves the fit here. To keep the model consistent with constraints of what models can be fitted, we had to limit the autoregressive order to 12 and fitting the MA terms. 

2.2 Moving Average Models

Using the ARIMA function, we’ve plotted AIC values keeping the AR model value 12 i.e., constant and Moving Average values variable up to 14.

From the figure 3, we can see that for Auto Regressive model 12, Moving Average model 5 suits the best. 

2.3 ARIMA Models

We have plotted AIC values of about 130 models with order of the autoregressive model varying from 0 to 12, the order of the moving-average model varying from 0 to 9 and degree of differencing equal to 1.

From Figure 5 and Figure 6, we can conclude that for Auto regression model 12 and moving average model 5 are the best one.

The coefficients table xx and the previous observations suggest that there tends to be a 5 month reversal effect but annual changes year over year tend to repeat themselves. There might be 5 months reversal that is consistent with managing expectations of the stock holder of the companies.

3. Last 10 Years

3.1  Auto Regressive Models

The data of last 10 years is converted to first order difference log data. We look for the best Auto Regressive model plotting of the Akaike Information Criterion (AIC) for values up to 14.

We see that AR model 12 is the best due to its lower AIC value.

3.2 Moving Average Model

Using the ARIMA function, we’ve plotted AIC values keeping the AR model value 12 i.e., constant and Moving Average values variable up to 14.

From the figure 7, we can see that for Auto Regressive model 12, the Moving Average model 0 suits the best. So we have no MA terms.

3.3 ARIMA Models

We have plotted AIC values of about 168 models with order of the autoregressive model varying from 0 to 12, the order of the moving-average model varying from 0 to 12 and degree of differencing equal to 1.

From Figure 8 and Figure 9, we can conclude that Auto regression model 12 and moving average model 0 are the best one.

With no Moving average terms, the producer price index is less affected by recent variations in producer price index. The past errors of the prediction model are not predictive of future production levels. So, the fact that past prediction errors are not very helpful suggests that the production process is perhaps stationary with a shorter memory.

4. Comparing Parameters of Two Models

4.1  R-Square

From figure 6, For the ARMA model (12,1,5) of first 10 years, the r-square value is equal to 0.5451 

From figure 7, For the ARMA model (12,1,0) of last 10 years, the r-square value is equal to 0.6404

In figure 13, it does appear that the period has very different months. 

In figure 14, we have red points representing the first 10 years and blue points representing the last 10 years.

For the last 10 years data, there are many months where the predictions are quite big. There is much greater variability in the last 10 years

4.2 Coefficients Table

For the first 10 years and last 10 years, t-values and p-values  are shown in the table below along with statistically significant coefficients in the overall model i.e., p-value < 0.05

4.3  Standard Deviation

For the first 10 years, the Standard Deviation i.e., square root of sigma^2 for logarithmic data is 0.004099652 and for actual data is 1.004206 on the original scale.

For the last 10 years, the Standard Deviation i.e., square root of sigma^2 for logarithmic data is  0.005545686 and for actual data is 1.005561 on the original scale.

The higher standard deviation for the last 10 years data is evident in figures 13 and 14.

5. Forecasting Values

5.1  First 10 Years

We have taken the data of the first 10 years which fit the best model ARIMA  identified by us and predicted the next 24 values i.e., next 2 years data and compared the forecasted values with the original values.

In the below figure, the blue highlighted graph shows the predicted Producer Price Index for the Pharmaceutical Industry.

Using the percentage error formula, We have calculated percentage error for the values. And the mean percentage error is 0.7728233 %

The above table shows the Actual and Forecasted values of Producer Price Index which are 24 months ahead of our data.

5.2  Last 10 Years

We have taken the data from September 2010 to August 2018 i.e., 8 years and fit the best model ARIMA identified by us and predicted the next 24 values i.e., till August 2020 and then compared the forecasted values with the original values.

In the below figure, the blue highlighted graph shows the predicted logarithmic data of Producer Price Index for the Pharmaceutical Industry.

The below table shows the Actual and Forecasted values. Using the percentage error formula, We have calculated percentage error for the values. And the mean percentage error is 1.723034%

Conclusion and Future Scope

The pharmaceutical industry has never before been called upon to solve rapid response issues on a global scale. Although during World War II the production of penicillin was truly an unprecedented engineering feat, it cannot be compared with meeting worldwide pandemic vaccine requirements. Right now is the time for novel solutions. 

Meeting the needs of the global community for vaccine during a pandemic outbreak will require exponential improvements in today’s capabilities. It is likely that such advances will have to be made across many fronts to meet the time and volume requirements. 

Predicting the Producer Price Index for the upcoming months is necessary for the times when the vaccine needs to be produced in large scale to estimate and to be prepared.

For future scope, the methodology can be extended to have an adaptive time series forecast of the index.

Further research would be focused on building the most effective adaptive models for forecasting and future work could address extending the methodology to build adaptive time series models and such extension would present interesting challenges. The adaptive model is important to allow for regime changes in the process and so the extension to adaptive models would be sensitive to the potential for changes over time but accommodate at the larger time periods if the data is consistent with stable regime over the analysis period.

References

• https://fred.stlouisfed.org/series/PCU325412325412
• https://www.bls.gov/ppi/ 
• https://www.ftc.gov/sites/default/files/documents/cases/2000/03/genevaabbpttanalysis.htm
• https://haz-map.com/Industries/119
• https://www.abpi.org.uk/medicine-discovery/covid-19/what-are-pharmaceutical-companies-doing-to-tackle-the-disease/
• nae.edu/7640/PharmaceuticalPreparednessforaPandemic
• https://www.rdocumentation.org/packages/forecast/versions/8.13/topics/ndiffs

Appendix I

Ndiffs: 

Number Of Differences Required For A Stationary Series.

Function to estimate the number of differences required to make a given time series stationary. Ndiffs estimates the number of first differences necessary.

Acknowledgements

The satisfaction in the execution of this work would be incomplete without expressing my sincere gratitude to all those people whose constant guidance and encouragement made it possible.

Foremost, I would like to express my deep sense of gratitude and august regards to my professor Dr. Peter Kempthorne, Department of Mathematics on financial mathematics and statistics, Massachusetts Institute of Technology, for his unflinching motivation and supervision throughout the coursework and project completion. His untiring efforts in completing the course has helped me understand the concepts and utilize them in our project.

About the author

Roshna Chaik

Roshna graduated from Osmania university in 2019. She is currently Working at J.P Morgan Chase as a Software Engineer.

The post Analysing and Predicting Producer Price Index by Industry: Pharmaceutical Preparation Manufacturing appeared first on Exploratio Journal.