Translational Interventional Radiology

By Tao Liu and Sanjeeva P. Kalva

 Translational Interventional Radiology, a volume in the Handbook for Designing and Conducting Clinical and Translational Research series, covers the principles of evidence-based medicine and applies these principles to the design of translational investigations in Interventional Radiology. The reader will come to fully understand important concepts including case-control study, prospective cohort study, randomized trial, and reliability study. Medical researchers will benefit from greater confidence in their ability to initiate and execute their own investigations, avoid common pitfalls in Interventional Radiology, and know what is needed for successful collaboration. Further, this reference is an indispensable tool in grant writing and funding efforts. The practical, straightforward approach helps aspiring investigators navigate challenging considerations in study design and implementation. This book provides valuable discussions of the critical appraisal of published studies in Interventional Radiology, elucidating the evaluation of the quality with respect to measuring outcomes and making effective use of all types of evidence in patient care. In short, this practical guide will be of interest to every medical researcher and interventional radiologist who has ever had a good clinical idea but not the knowledge of how to test it.

• Focuses on the principles of evidence-based medicine and applies these principles to the design of translational investigations within interventional radiology
• Provides a practical, straightforward approach that helps investigators navigate challenging considerations in study design and implementation
• Details discussions of the critical appraisal of published studies in interventional radiology, supporting evaluation with respect to measuring outcomes and making effective use of all types of evidence in patient care

• Language: English
• Publisher: Academic Press
• Release date: Apr 5, 2023
• ISBN: 9780128230558

Book preview

Preface

Translational research is essential to the advancement of Interventional Radiology (IR). Written for clinicians, scientists, students, and biotech/medtech entrepreneurs, this IR-specific guide to translational research serves as a practical, step-by-step roadmap for designing and conducting basic, translational, and clinical research, including basic regulatory considerations for new technology. Translational Interventional Radiology aims to help bridge the gap between clinical problems, research, and practice. Comprehensively spanning from preclinical research, clinical research, clinical implementation, to public health, the book provides a clear process for understanding, designing, executing, and analyzing clinical and translational research.

The reader will learn how to more critically evaluate the quality of published studies with respect to measuring outcomes and to make effective use of all types of evidence in patient care. This book covers the principles of evidence-based medicine and applies these principles to the design of translational investigations.

The book’s straightforward approach aims to help the aspiring investigator navigate challenging considerations in study design and implementation. Researchers will benefit from greater confidence in their ability to initiate and execute investigations along the translational spectrum, gain valuable insights on funding, avoid common pitfalls, and know what is needed in collaborators.

Whether you are a basic scientist with a promising therapeutic agent interested in moving from laboratory to clinical experimentation, an engineer needing to complete a pilot study for a new biomedical device, a student learning the principles of statistics or epidemiology, a clinician orchestrating and preparing grants for a randomized clinical trial, or an entrepreneur seeking insight on regulatory considerations, Translational Interventional Radiology is a practical guide for effectively advancing and translating an IR research question.

Adam E. M. Eltorai

April 2023

Part I

Introduction

Outline

Chapter 1. Translational process

Chapter 2. Scientific method

Chapter 3. Basic science research

Chapter 1: Translational process

Priyam Choudhury, and Ashutosh Mohapatra     Department of Radiology, Mayo Clinic, Jacksonville, FL, United States

Abstract

This chapter introduces the concept of Translational Research (TR) and its importance. It starts by defining the benefits of TR and the term itself. It goes on to classify the various stages of TR and what challenges investigators face when involved in TR and how it can be mitigated. By the end of the chapter, emphasis is given on the importance of incorporating TR training programs and how the trainees and the research outcome can be evaluated.

Keywords

Benefits and challenges in TR; NCAT; Translational research; UW ICTR

Key points

• Translational Research (TR) aims to build on basic scientific research to create new therapies, medical procedures, and diagnostics for the betterment of public health.

• Collaboration between multidisciplinary investigators is crucial for TR success.

• TR results in public policy changes and is also beneficial for the community's economy.

Translational research (TR) aims to improve public health by applying the outcomes of multidirectional and multidisciplinary research based on basic, clinical, patient, population, and policy-based science. It promotes collaboration among various disciplines, translating knowledge from one science to another (for example, from basic science to clinical science) ¹ to mitigate health concerns and generate novel ideas. TR provides public health benefits, clinical benefits, and economic benefits by adopting new tools and procedures into practice settings and amending the policy-making process formally.

While the definition of TR has been less clear and was first associated with cancer research in the 1990s, today, various fields have attempted to define the term. The NIH defines TR, while applying for the CTSA (Clinical and Translational Science Award) program, as encompassing two areas of translation: applying discoveries generated through research in the laboratory and preclinical studies to the development of trials and studies in humans and enhancing and adopting best practices in the community through research. Treatment strategies and cost-effectiveness in preventing diseases are also essential parts of translational science. ¹

In the continuum, T1 refers to the first part of research where knowledge is transferred from basic research to clinical research, and T2 refers to the second part of research that transfers findings from clinical research to practice, aiding in improving the community's health.

Successful TR examples include cancer therapy, requiring collaboration between basic researchers, clinicians, and industry to generate target compounds with reduced toxicity and enhanced efficacy, ² the introduction of insulin, and the advent of antibiotics.

The National Center for Advancing Translational Sciences (NCATS) describes translational research broadly as the process in which observations in the laboratory, clinic, and community are translated into interventions to improve the people's health in the community. It involves implementing diagnostics and therapeutics in medical procedures and behavioral changes in the community. ³

The translational science is not unidirectional, but rather, each stage is built on information acquired from other stages. It consists of stages starting from the basics of health and disease to interventions in improving public health. Patient involvement in all stages is crucial to develop new approaches, demonstrate their usefulness, and disseminate those findings in treatment (Fig. 1.1).

In June 2013, the Institute of Medicine (IOM), which supplies major investment to University of Wisconsin Institute for Clinical and Translational Research (ICTR) and more than 60 CTSA sites nationwide, made recommendations to change research classification definitions. The T0–T4 classification replaced the T1 and T2 categories. ⁴

T0 Research includes basic biomedical research, including preclinical and animal studies, not including interventions on human subjects.

T1 Research encompasses translation to humans, including proof of concept studies, Phase 1 clinical trials, and focus on new methods of diagnosis, treatment, and prevention in highly controlled settings.

Figure 1.1  Stages in translational research.

T2 Research entails translation to patients, including Phase 2 and 3 clinical trials, and controlled studies leading to clinical application and evidence-based guidelines.

T3 Research includes translation to practice, including comparative effectiveness research, postmarketing studies, clinical outcomes research, as well as health services, and dissemination and implementation research; and

T4 Research translates to communities, including population level outcomes research, monitoring of morbidity, mortality, benefits and risks, and impacts of policy and change.

TR programs are trying to design an effective training program where trainees are expected to develop certain skills and knowledge that are not traditionally taught in training. Although it's challenging to track outcomes as every individual program and trainee has different needs and expectations, it can be documented whether the preestablished goals were met or not. The Institute for Clinical and Translational Sciences (ICTS) Tracking and Evaluation Team developed the Translational Science Benefits Model (TSBM), which can be used as a tool to assess and evaluate the outcome to measure community health benefits and clinical benefits from TR. ⁵ They plan to form a resource library which can be freely accessed by academicians, evaluators, and scientists who wish to evaluate effort.

TR does not come without barriers despite aiming at improving patient's health. Cultural and training differences between groups of investigators, lack of communication between the groups, lack of resources and the complex regulatory issues that encompass tissue banking, FDA, ethics involved in human research, and Medicines and Healthcare Regulatory Agency (MHRA) approvals—all these factors can be intimidating to the researchers. ⁶ These issues can be overcome with support from regulatory bodies, as well as by the building of and collaboration between interdisciplinary forums. In the United States of America, the CTSA scheme in 2006 and the NIH Roadmap in 2003 were designed to support TR. ⁷

References

1. Rubio D.M, Schoenbaum E.E, Lee L.S, et al. Defining translational research: implications for training. Acad Med. 2010;85(3):470–475. doi: 10.1097/ACM.0b013e3181ccd618.

2. Goldblatt E.M, Lee W.H. From bench to bedside: the growing use of translational research in cancer medicine. Am J Transl Res. 2010;2(1):1–18 Published 2010 January 1.

3. Ncats.nih.gov. 2021 [online] Available at:. https://ncats.nih.gov/files/NCATS_Factsheet_508.pdf [Accessed 17 March 2021].

4. What Are the T0 to T4 Research Classifications? ICTR; 2021. https://ictr.wisc.edu/what-are-the-t0-to-t4-research-classifications/ [Accessed 17 March 2021].

5. Luke D.A, Sarli C.C, Suiter A.M, et al. The translational science benefits Model: a new framework for assessing the health and societal benefits of clinical and translational sciences. Clin Transl Sci. 2018;11(1):77–84. doi: 10.1111/cts.12495.

6. Homer-Vanniasinkam S, Tsui J. The continuing challenges of translational research: clinician-scientists' perspective. Cardiol Res Pract. 2012;2012:246710. doi: 10.1155/2012/246710.

7. Zerhouni E. Medicine. The NIH Roadmap. Science. 2003;302(5642):63–72. doi: 10.1126/science.1091867.

Chapter 2: Scientific method

Joshua Katz ¹ , and Zachary T. Smith ²       ¹ Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ, United States      ² Robert Larner, M.D. College of Medicine at the University of Vermont, Burlington, VT, United States

Abstract

The scientific method is the fundamental means by which humanity makes reliable discoveries. At its core are six steps: Observation, Question, Hypothesis, Experiment, Analysis, and Conclusion. Understanding these components and their limitations is crucial to the successful and accurate uncovering, reporting, and application of knowledge. In medicine, the rigor of this process of inquiry enables researchers to identify pathologies and develop new technologies in an organized, verifiable manner.

Keywords

Analysis; Conclusion; Experiment; Hypothesis; Observation; Question; Scientific method

Key points

• The scientific method is an organized process of reproducible discovery.

• It consists of Observation, Question, Hypothesis, Experiment, Analysis, and Conclusion.

• Proper understanding of these steps is important to produce reliable research results.

Why it matters?

The scientific method is the fundamental means by which humanity makes reliable discoveries. It consists of six core steps, as illustrated in Fig. 2.1. Understanding these components and their limitations is crucial to successfully and accurately uncover, report, and apply knowledge.

In medicine, the rigorous process of inquiry enabled by the scientific method allows researchers to identify pathologies and develop new technologies in an organized and verifiable manner. This process ensures that the knowledge gained is reliable and can be used to make informed decisions.

Figure 2.1  Steps of the scientific method.

Observation

The scientific method begins with observation, which kickstarts the entire process. While this step may seem simple relative to the ones to come, its success is critical to the entire process. Only by paying attention to the world around them can researchers recognize the surprising or unknown and find inspiration to ask novel questions or solve interesting problems.

For instance, a clinician may notice that a certain subset of her patients experiences fewer complications than another or may observe that patients on a particular medication regimen have decreased mortality. Observation is fundamental to question generation and is the initial step in making important discoveries.

Question

The research question forms the basis for the remainder of the scientific process (as shown in Fig. 2.2). A quality research question can be a novel question that has not been asked before, or a question that has been previously investigated, but the current literature does not provide a definitive answer. By asking novel or unanswered questions, the investigator can ensure that the research is relevant and worthwhile.

Figure 2.2  Considerations when asking a research question.

It is essential to note that research questions are always evolving, and discovering answers to previously asked questions leads to the development of new and novel questions, allowing for a better understanding of the subject matter. For example, Breen et al. decided to investigate the efficacy of percutaneous cryoablation for renal tumors. ¹ Following the advent of percutaneous ablation for renal cell carcinoma, Lourenco et al. questioned and compared outcomes of percutaneous ablation versus nephrectomy in patients with T1 renal cell carcinoma. ² New discoveries always lead to the development of new and novel research questions.

Before asking the research question, it is crucial to be well-versed in the current published literature. Developing a thorough understanding of the literature allows for a more thoughtful and meaningful research question to be asked. Additionally, current NIH-funded projects underway can be searched online using the NIH Research Portfolio Online Reporting Tools at report.nih.gov.

It is common for studies to ask more than one research question. Having secondary research questions can add complexity to the study by introducing multiple variables and complicating data analysis, but this will be discussed in later chapters on study design.

Hypothesis

Once good research questions have been formulated, it is crucial for the researcher to consider the possible answers they may expect. This is important as generating hypotheses forces a more in-depth consideration of the exact questions to be asked and clarifies the means by which answers will be found during experimentation.

For example, when a clinician generates a differential diagnosis list for a given patient's concern, such as new-onset calf pain and swelling, they are generating a set of hypotheses to answer the question, What is going on here? Potentially, there was trauma to the limb, or the patient may have a deep vein thrombosis (DVT), or there could be a superficial infection. These hypotheses then guide the clinician's next steps: to ask about trauma and look for signs of injury, to inquire about risk factors for DVT and perform a doppler ultrasound, and to observe for signs of infection.

Hypothesis generation, therefore, helps determine the specifics of the next step, experimentation. It allows researchers to identify the variables to be tested, the methods for data collection, and the statistical analyses to be performed. Overall, hypothesis generation is a critical component of the scientific process that ensures the validity and reliability of research findings.

Experiment

Experimentation can take many forms depending on the research question, ranging from observation of risk factors or disease incidence in a specified population to randomized manipulation of a single variable between two groups (as shown in Fig. 2.3). In this step, it is essential to consider extraneous factors that may affect the collected data. The most powerful studies examine the effect of changing one variable between large sample groups and control for other differences between the groups through study design adjustments.

An illustrative example of good experimental design is seen in Granger et al.'s trial, which compared apixaban versus warfarin for patients with atrial fibrillation. The study found that apixaban is superior to warfarin in preventing stroke or systemic embolism, caused less bleeding, and resulted in lower mortality. ³ This was an effective trial as it was randomized, double-blinded, had a large sample of 18,201 patients, and was focused on assessing one variable's effect on outcomes.

Overall, experimentation is a critical step in the scientific method as it allows researchers to test their hypotheses and generate new knowledge. However, it is essential to ensure that the experimental design is rigorous and well-controlled to minimize bias and ensure the validity and reliability of the findings.

Analysis

Once the experiment has been conducted and the data has been collected, the next step is to perform an analysis of the results. This is generally done through a series of statistical tests that allow the investigator to objectively understand the results.

The analysis process is highly variable and depends heavily on the design of the study, as certain statistical tests are preferred for different experimental designs. For instance, parametric tests, such as t-tests and ANOVA, are commonly used for normally distributed continuous data, while non-parametric tests, such as Wilcoxon rank-sum and Kruskal-Wallis tests, are used for non-normally distributed or ordinal data.

In addition to selecting the appropriate statistical test, it is crucial to consider potential confounding variables, control for them through statistical adjustments, and interpret the results with caution. Various statistical methods of analyzing results of a study will be discussed in detail in later chapters, including the advantages and limitations of different statistical tests and strategies to minimize bias and improve the accuracy and generalizability of the findings.

Figure 2.3  Types of experiments with examples. DVT, deep vein thrombosis. PAD, peripheral artery disease.

Conclusion

After the data has been analyzed, and the results are available for interpretation, the investigator needs to return to the initial hypotheses made before implementing the experiment. Based on the findings of the study, the researcher makes an informed decision on how the study's results relate to the initial hypotheses. If the study's results support the hypotheses, the researcher can consider them validated. However, if the results contradict the hypotheses, the researcher needs to re-evaluate their assumptions and develop new hypotheses that can explain the findings.

Interesting findings may serve as the observations that spark continued iteration of the scientific method. The researcher may use the findings to develop new research questions or to refine existing hypotheses, leading to further studies that build upon the previous work. The scientific method is, therefore, an iterative process that relies on the constant re-evaluation of assumptions and the refinement of research questions to generate new knowledge and improve our understanding of the natural world.

Get started

Now that the fundamentals of the scientific method have been reviewed, it is time to begin refining the research question and establish the study population. Please reference the following chapters for further discussion regarding these next steps:

Chapter 19 – The question: Types of research questions and how to develop them

Chapter 20 – Study population: Who and why them?

Potential pitfalls

• Poorly defined variables

• Before jumping to experimental design, it is crucial to clearly define the variables under investigation and the metrics by which they will be measured. For instance, a researcher may wonder, Is percutaneous biliary drainage safer than cholecystectomy? However, they must define the safety metrics, such as all-cause mortality, hospital readmission rate, complication rate, and others, before beginning the experiment. Failure to do so may result in data being collected or analyzed incorrectly, leading to inaccurate conclusions.

• Not controlling for variation between study groups

• During experimentation, it is essential to control for as many variables as possible to avoid confounding factors. For example, a researcher planning to compare the safety of two procedures must ensure that the patient populations receiving each procedure are similar in terms of illness severity and risk factors. Failure to do so may lead to inaccurate conclusions and flawed study design.

• Understanding limitations

• Researchers must accurately understand the generalizability of the results of any study. It is crucial to recruit a large, representative sample of the population under investigation to extrapolate the findings to the larger population of interest. Additionally, researchers must manage expectations and accurately report the impact of any study. For example, a study examining the life expectancy of patients diagnosed with pancreatic cancer in a single urban tertiary care center in the United States may not be applicable to rural patients in another country, or even to rural patients in the same state. Understanding the study's limitations and reporting them accurately is crucial for scientific integrity.

Real-world examples

Question

Prologo et al asked if percutaneous cryoneurolysis is a feasible and safe treatment option for patients with refractory phantom limb pain (PLP). ⁴ Safety and efficacy was then tested in a pilot cohort of 21 patients with refractory PLP.

Hypothesis

Sacks et al. hypothesized that interventional radiologists (IRs) and neurointerventional (NI) physicians have similar outcomes of endovascular stroke thrombectomy (EVT), which could improve the availability of thrombectomy treatment. ⁵ They tested this hypothesis by retrospectively analyzing outcomes between IRs and NI physicians across eight institutions.

Experiment

In an effort to explore the potential applications of histotripsy, Swietlik et al designed an experiment to evaluate the feasibility and safety of in-vivo histrotripsy of subcutaneous fat in a porcine model. ⁶ Subcutaneous fat volume reduction was measured through serial MRI at 7, 28, and 56 days post-treatment.

References

1. Breen D.J, Bryant T.J, Abbas A, et al. Percutaneous cryoablation of renal tumours: outcomes from 171 tumours in 147 patients. BJU Int. October 2013;112(6):758–765. doi: 10.1111/bju.12122 Epub 2013 Apr 12. PMID: 23581293.

2. Lourenco P, Bilbey N, Gong B, Bahrabadi A, Halkier B. Percutaneous ablation versus nephrectomy for small renal masses: clinical outcomes in a single-center cohort. Cardiovasc Intervent Radiol. December 2018;41(12):1892–1900. doi: 10.1007/s00270-018-2050-9 Epub 2018 Aug 7. PMID: 30088062.

3. Granger C.B, Alexander J.H, McMurray J.J, et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2011;365(11):981–992. doi: 10.1056/NEJMoa1107039.

4. Prologo J.D, Gilliland C.A, Miller M, et al. Percutaneous image-guided cryoablation for the treatment of phantom limb pain in amputees: a pilot study. J Vasc Interv Radiol. January. 2017;28(1):24–34.e4. doi: 10.1016/j.jvir.2016.09.020 Epub 2016 Nov 23. PMID: 27887967.

5. Sacks D, Dhand S, Hegg R, et al. Outcomes of stroke thrombectomy performed by interventional radiologists versus neurointerventional physicians. J Vasc Interv Radiol. June. 2022;33(6):619–626.e1. doi: 10.1016/j.jvir.2021.11.018 Epub 2022 Feb 9. PMID: 35150837.

6. Swietlik J.F, Knott E.A, Longo K.C, et al. Histotripsy of subcutaneous fat in a live porcine model. Cardiovasc Intervent Radiol. September. 2022 12 doi: 10.1007/s00270-022-03262-4 Epub ahead of print. PMID: 36097074.

Additional reading

Reiff-Cox R. Exchanging the myth of a step-by-step scientific method for a more authentic description of inquiry in practice. In: McComas W, ed. Nature of Science in Science Instruction. Science: Philosophy, History and Education. Cham: Springer; 2020. https://doi-org.proxy.libraries.rutgers.edu/10.1007/978-3-030-57239-6_6.

Gauch J.H.G. Scientific Method in Brief. ProQuest Ebook Central; 2013. https://ebookcentral-proquest-com.proxy.libraries.rutgers.edu.

Further reading

1. InformedHealth.org [Internet]. Cologne, Germany: Institute for Quality and Efficiency in Health Care (IQWiG); 2006 What types of studies are there? 2016 Jun 15 [Updated 2016 Sep 8]. Available from:. https://www.ncbi.nlm.nih.gov/books/NBK390304/.

Chapter 3: Basic science research

Amanda Bronte Balon     Wake Forest School of Medicine, Winston-Salem, NC, United States

Abstract

This chapter introduces the idea of basic science research as the foundation of medical progression. The chapter starts by introducing the concept of basic science research and defining the modern practice of scientific exploration. The chapter then goes on to explain methodology of basic science research and the way in which this research translates to clinical practice. Lastly, the chapter concludes by describing the barriers to research and the limitations within research.

Keywords

Barriers to research; Basic science research; Scientific method of inquiry; Translational research

Introduction to basic science research

Basic science research is a fundamental study incorporating the principles of biological sciences with the central aim of discovering, understanding, and refining scientific theories and mechanisms. ¹ This type of research encompasses all forms of experimentation that are directed toward increasing the fundamental knowledge of observable scientific phenomena. This research includes a multitude of subjects, including biochemistry, microbiology, physiology, pharmacology, and pathology extending to social and behavioral sciences. The rationale for basic research in medical and information science is best summarized by Horst Kunge: Learning to master theoretically and in practical application, the ground rules of research creates the best foundation for continuing growth in a profession. ² This rationale is especially applicable within the field of medicine, a field in which the structure of the profession relies on providing advice to others derived from a body of generalized and systemic knowledge that comprises its theoretical core.

The scientific method of inquiry

The process of research involves systematic collection and analysis of data to facilitate organized exploration of scientific principles. Basic science research can be categorized into quantitative and qualitative methods, which are highly structured and rely on purposeful measurement, analysis, and evaluation. ³ The observation of scientific phenomena leads to inductive reasoning, contributing to the development of the scientific method of inquiry. ⁴ This method of scientific discovery and research gained support in the 16th century and has remained a valid method for answering questions and solving problems in modern research. ⁴ The scientific method of inquiry generally starts with identifying a problem or question. The scientific method initially involved identifying a problem or interest as a lead point to theory construction, resulting in the derivation of theoretical hypotheses. ³ Frankfort-Nachmias and Nachmias describe the research process as consisting of seven principal stages: problem, hypothesis, research design, measurement, data collection, data analysis, and generalization. ⁵ These stages outline modern scientific exploration and experimentation.

Hypothesis development is a critical stage in this process. Farrugia et al. state that a well-formulated hypothesis is based on a good research question and is developed before the start of the study, serving as the driving focus behind data collection for the study. The authors also highlight the importance of the hypothesis as the central element of the study, serving as the basis for testing and establishing the statistical and clinical significance of the research. ⁶

Translating basic science research

An important component of creating impactful basic science research is translating the discoveries of that research into clinical trials. Eventually, the findings from these clinical trials can be applied to diverse populations in an effort to develop tools and knowledge for implementing well-developed scientific changes into healthcare delivery systems. ⁷ This process of applying data from a clinical trial to large populations is known as translational research and is the ultimate goal of clinically relevant basic science research. Surgeons and clinicians rely on updated studies founded in basic science research to guide their clinical practice. The process of translating research into actual patient care guidelines is heavily prioritized, and the National Institutes of Health (NIH) works towards minimizing natural barriers to the utilization of updated data in daily clinical practice. These barriers include improving the applicability of research, motivating clinicians to increase adherence, and improving the availability of data to a broader group of healthcare workers. ⁸ The importance of minimizing barriers to the utilization of basic science research as a means of informing healthcare cannot be overstated, as it contributes to the mission of medicine by improving patient health outcomes and the scientific basis for the practice of medicine.

Barriers to research

Although basic science research is fundamental to progress healthcare and improve patient health outcomes, numerous barriers exist to designing research, accessing research opportunities, resources, or equipment, funding research, motivating scientists toward specific disciplines, and sharing knowledge globally, as well as limitations within studies. A study by Okoduwa et al. revealed that at a particular institution many more students were motivated or interested in research but had extreme barriers impeding their ability to participate. ⁹ This phenomenon has been described across multiple studies at many other institutions. ¹⁰–¹² Even within the medical field, AlGhamdi et al. found that although majority of participating medical students believe that research is an important component of the progression of medicine; however, only around half of the students participated in research during medical school due to barriers to conducting research. ¹³

Once accessing and designing research there are still limitations within the study. Limitations are defined as any potential weakness of the study that is often out of the researcher's control and are usually associated with the research design, statistical model constraints, funding constraints, or other predictable or unpredictable factors. ¹⁴ Limitations can ultimately affect the conclusions and results of the study. A common example of a basic science limitation is a study in which the researcher only has access to a small number of participants' experience as data points. This is a limitation of the study because it does not give proper representation of the average person within a larger, more diverse population. All research unavoidably has some limits. Therefore, it is extremely important for researchers to acknowledge and discuss the limitations to their study within their final publishing to create space for less limited research on the topic in the future.

Finding ways to overcome these obstacles to and within research is an ongoing challenge and a topic that is itself studied heavily. The improvement in these barriers is imperative at the individual and institutional level to guarantee enduring progression in health research.

Conclusion

Basic science research serves as the foundation of healthcare advancements across a multitude of disciplines and requires a methodical and organized approach to data collection and analysis. Conducting research with these fundamentals in mind creates a pathway to clinical relevancy and application. The translation of basic science research to clinical trials and eventually generalizing to large populations changes the way clinicians practice medicine. Despite the ongoing acknowledgement of the importance of basic science research, there are still many avoidable barriers to these types of studies, which continue to pose ongoing challenges for future researchers. Finding ways to overcome these barriers is imperative at both the individual and institutional level to ensure enduring progression in health research.

References

1. La Caze A. The role of basic science in evidence-based medicine. Biol Philos. 2011;26:81–98. doi: 10.1007/s10539-010-9231-5.

2. Mouly, George J. Educational Research: The Art and Science of Investigation. Boston: Allyn and Bacon; 1978 (Print).

3. Guthrie G. Basic Research Methods: An Entry to Social Science Research. New Delhi: SAGE Publications India Pvt Ltd; 2010 SAGE Knowledge, 26 September 2021. http://www.doi.org/10.4135/9788132105961.

4. Connaway L.S, Powell R.R. Basic Research Methods for Librarians. Santa Barbara, Calif: Libraries Unlimited; 2010 (Internet resource.

5. Andrews J. Book review – research methods in the social sciences, by Chava Frankfort-Nachmias, David Nachimas and Jack De waard (8th edition). Popul Ageing. 2019;12:195–198. doi: 10.1007/s12062-017-9191-5.

6. Farrugia P, Petrisor B, Farrokhyar F. Practical tips for surgical research: research questions, hypotheses and objectives". J Can Chir. 2010;53(4):278–281.

7. Hulley S.B, Cummings S.R, Browner W.S, Grady D.G, Newman T.B. Designing Clinical Research. 4th ed. Lippincott Williams and Wilkins; 2013.

8. Kahn K, Ryan G, Beckett M, et al. Bridging the gap between basic science and clinical practice: a role for community clinicians. Implement Sci. 2011;6:34. doi: 10.1186/1748-5908-6-34.

9. Okoduwa S.I.R, Abe J.O, Samuel B.I, Chris A. Attitudes, perceptions, and barriers to research and publishing among research and teaching staff in a Nigerian research Institute. Front Res Metr. 2018;3:26. doi: 10.3389/frma.2018.00026.

10. Baro E.E, Bosah G.E, Obi I.C. Research funding opportunities and challenges: a survey of academic staff members in Nigerian tertiary institutions. Bottom Line. 2017;30:447–644. doi: 10.1108/BL-07-2016-0027.

11. Aslam F, Shakir M, Ahad-Qayyum M. Why medical students are crucial to the future of Research in South Asia. PLoS Med. 2005;2:e322. doi: 10.1371/journal.pmed.0020322.

12. Dickson-swift V, James E.L, Kippen S, Liamputtong P. Blurring the boundaries in qualitative health research on sensitive topics. Qual Res. 2006;16:853–871. doi: 10.1177/1049732306287526.

13. AlGhamdi K.M, Moussa N.A, Alessa D.S, Alothimeen N, Al-Saud A.S. Perceptions, attitudes and practices toward research among senior medical students. Saudi Pharmaceut J. 2014;22(2):113–117. doi: 10.1016/j.jsps.2013.02.006.

14. Theofanidis D, Fountouki A. Limitations and delimitations in the researsch process. J Perioper Nurs. 2019;7(3):155–162. doi: 10.5281/zenodo.2552022 E-ISSN:2241-3634.

Part II

Pre-clinical: discovery and development

Outline

Chapter 4. Overview of preclinical research

Chapter 5. Defining the problem to solve

Chapter 6. Types of problems

Chapter 7. Types of interventions

Chapter 8. Drug discovery

Chapter 9. Drug safety in vitro, in vivo, harm, toxicity, dosing, good laboratory practices

Chapter 10. Device discovery

Chapter 11. Device prototyping: Iterative refinement, optimization process

Chapter 12. Device classification

Chapter 13. Device testing: Safety and efficacy

Chapter 14. Diagnostic testing

Chapter 15. Regulatory process

Chapter 16. Procedural technique development

Chapter 17. Behavioral interventions

Chapter 4: Overview of preclinical research

Siddharth Venkatraman ¹ , and Capt. Tej Ishaan Mehta ² , ³       ¹ The Johns Hopkins University School of Medicine, Baltimore, MD, United States      ² Department of Interventional Radiology, The Johns Hopkins Hospital, Baltimore, MD, United States      ³ United States Air Force Medical Corps, Falls Church, VA, United States

Abstract

Preclinical research generally falls into one of three categories; in silico (computer modeling or simulations), in vitro (cell culture or partial biological systems), and in vivo (animal studies or complete nonhuman biological systems). Preclinical research encompasses the important first steps before translational and, ultimately, human research. To design a successful preclinical study, there are certain guidelines that researchers must follow to enhance scientific rigor without restricting creativity or innovation. In interventional radiology, preclinical research has played critical roles in the development of multiple therapies and continues to be conducted for potentially novel therapies.

Keywords

Basic science; Biases; Experimental design; In vitro; In vivo; Preclinical

Key points

• Preclinical research is an essential starting point for the development and evaluation of nearly all translational, clinically relevant medical interventions.

• Understanding the underlying components of preclinical research is essential for designing robust experimental frameworks.

• Early consideration of experimental biases in preclinical research may help improve experimental design.

What is preclinical research?

While clinical research refers to studies in humans, preclinical research encompasses the body of research that lays the groundwork for ethical justification of human trials. In this chapter, we will focus on hypothesis-testing in preclinical studies, explain general concepts related to types of experiments and the design of in vivo experiments, and discuss steps to mitigate biases in preclinical research. ¹

Why it matters?

Preclinical research is an essential component of nearly all medical interventions. Preclinical research tests for safety and efficacy of therapeutics prior to human trials, ultimately reducing the risk of potential treatment-related sequelae and refining treatment delivery parameters.

Preclinical research models

There are three broad categories of preclinical research: in silico, in vitro, and in vivo.

In silico (within the silicon) studies broadly cover studies performed using computer simulation. In silico studies allow for rapid modeling of greatly varied biological systems including but not limited to the following: novel pharmaceutical applications with virtual drug screening, creation of new pharmaceutical compounds using machine learning, developing mechanistic models of action for drug–protein interactions, and establishing structure–activity relationships, radiotherapy comparative dosimetric studies, etc. In silico imaging has even recently emerged as a field, potentially allowing for rapid development of novel imaging techniques and optimization of current techniques. ² Although still in its infancy, in silico modeling has been applied within interventional radiology, such as patient-specific simulation of yttrium-90 dosimetry. ³ Results from in silico models may help guide study design for future in vitro and in vivo studies.

In vitro (within the glass) studies provide a generally well-controlled environment for hypothesis testing, often using biological molecules or cell cultures outside the body. Cell lines or biological molecules are often derived from either humans or nonhuman animals and can be introduced to a therapy in question to monitor a response. As compared to in vivo studies, in vitro studies are generally cheaper, higher throughput and do not require animal husbandry. Historically, results from in vitro studies guided in vivo studies, but in vitro results poorly extrapolated directly to in vivo results as in vitro models do not fully model the complex cellular environment and interactions in vivo. In vitro studies may be viewed as reducing the degrees of freedom compared to in vivo studies, allowing for more rapid and simplified modeling, but sacrificing many of the complexities of independent biological systems. To help bridge the gap between in vitro and in vivo systems, modern three-dimensional cell cultures are being implemented, which theoretically maintain cellular heterogeneity and tissue architecture. ⁴ If there is evidence of efficacy in in vitro experiments in vivo trials may be attempted to gather additional information in more complex biological systems.

In vivo (within the living) studies cover a broad range of topics but, generally speaking, encompass trials within independent, complex biological systems. In medical preclinical research, in vivo studies almost always occur in nonhuman animals. These studies are able to better model efficacy and sequelae of therapies in complex biological environments, which can improve its predictions of safety, https://www.news-medical.net/life-sciences/What-is-Toxicity.aspx toxicity, and efficacy as compared to in vitro studies. However, in vivo studies often require significantly more money, time, and attention to comply with ethical policies and regulatory laws governing animal testing. Often, scientists are required to demonstrate that no alternative methodology can be used to conduct the experiment and that the benefits of the study outweigh the suffering caused to the animals. This field is also expanding, as technologies like CRISPR allow scientists to generate animal models of specific pathologies cheaper and faster than before. ⁴

Within the groups of in silico, in vitro, and in vivo studies, there are several common preclinical study models that are conducted and worth describing:

1. Screening test: Simple and rapid initial screen to determine the presence or absence of a pharmacological property in the putative drug. Avenues to investigate this may include (but are not limited to) molecular modeling of drug binding sites to specific proteins, application of pharmaceuticals to cell cultures, or applications of pharmaceuticals to living animals.

Ex: Testing if an antibiotic has efficacy on a specific bacterial strain.

2. General observation test: Observation of the effects from a therapy, generally focused on identifying sequelae associated with the therapy. One experimental design may include a drug of interest being injected in triplicate into a series of animals at varying doses and observing the animals for sequelae.

Ex: Assess if a novel contrast medium causes renal damage and, if so, to describe the relationship between contrast dose and degree of renal damage.

3. Mechanism of action: Experiments are conducted to determine the underlying mechanism of action for the drug. Avenues of investigation may include (but are not limited to) molecular simulations of drug-protein interactions and a litany of biochemical investigative techniques.

Ex: Determine if a drug with an antihypertensive effect is acting on α-receptors or β-receptors.

4. Systemic pharmacology experiment: Determining effects on individual and major organ systems (cardiovascular, respiratory, neurological). This requires in vivo experiments in most cases. At this stage, scientists determine the absorption, distribution, metabolism, excretion, and toxicity profile of the therapy in question.

Ex: The drug cimetidine is used to reduce stomach acid excretion, but has various other GI, neurological, and genitourinary side effects that require in vivo tests to describe and cannot be adequately determined in silico or in vitro.

5. Quantitative tests and pharmacokinetics: examining dose-response relationship, maximal effects, and therapeutic range. This also includes the movement of the drug across body systems (absorption, distribution, metabolism, localization in tissues, and excretion). These parameters provide data to compare the new therapy to an existing standard of care, as well as a preferred safe dose and route of administration.

Get started

Different aspects of preclinical research generally have different levels of entry beyond technical knowledge. In silico modeling has relatively low barriers to entry, requiring only a standard computer, though increased processing capabilities, such as with advanced graphics processing units, generally improve processing times. Online access to third-party graphics processing units and tensor processing units, such as through the Google Colaboratory, can also provide access to increased processing capabilities without the need for local advanced processing capabilities.

In vitro models may (in principle) be conducted with simple equipment such as cell-culture media and a microscope. In modern practice, a wide-range of equipment is often required such as labeling tools (I.e., histopathology media), purification/extraction tools (I.e., ultracentrifuges) and safety equipment (I.e., fume hoods). Evaluation of specific molecules further requires advanced biochemical technologies such as nuclear magnetic resonance spectroscopy, high-performance liquid chromatography or mass spectrometry.

Akin to in vitro and in silico models, in vivo models may be greatly varied in size, scope and complexity. Accordingly, establishing models for in vivo testing vary greatly in complexity. In general, there is an inverse relation between phylogenetic distance from humans and complexity of starting in vivo testing. While plant and fungi models require horticultural skills, animal models require animal husbandry with associated testing, approval and funding above and beyond that of the therapeutics being tested. Additionally, matters of equipoise become increasingly important with reduced phylogenetic distance from humans, further escalating the complexity of in vivo models.

Designing in vivo experiments

Designing an in vivo experiment requires meticulous planning beforehand to create a robust experimental plan. Typically, the framework is divided into the following components:

Hypothesis and effect size

By convention in experimental design, there are generally two hypotheses for any given experiment, a null hypothesis and an alternative hypothesis. The null hypothesis states that the intervention or the drug had no statistically significant effect or change on the outcome. The alternative hypothesis states that the intervention produced a statistically significant effect on the outcome. In statistics, the P-value reports the probability of observing an effect as extreme or more extreme than what would be observed if the null hypothesis was true. In other words, the smaller the P-value, the more likely the null hypothesis may be rejected.

Crucial to this process is determining an effect of interest, which is the variable that is measured as an outcome of the intervention. Then, the scientist must decide on a minimum effect size or the smallest effect the experiment is designed to be able to detect (often the minimum difference that would be of biological relevance). This can then be used to determine an adequate sample size to ensure that there is enough experimental power to detect changes that are biologically relevant, not just statistically significant. While there is no easy way to ensure that results will translate to clinical relevance, it is suggested that the minimum effect size fulfill the following criteria: ⁵

- Cause a beneficial effect for individuals rather than large cohorts

- The difference must be experimentally testable and reasonable to achieve

- There should be a rationale for translation into patients in the long run

Experimental unit and sample size

In a comparative in vivo experiment, the animals are separated into groups and exposed to different interventions (drugs, dosages, delivery mechanisms, surgical procedures, etc.). The sample size is the number of experimental units per group. The experimental unit is the entity subjected to an intervention independently of all other units. Under this definition, a unit could refer (for example) to individual mice, or to an entire cage of mice. Once identified, the experimental units are separated and subjected to different interventions. It is generally good practice to randomize the distribution of experimental units to reduce the effects of random error and selection bias. It is important to note that one of these groups should be used as a control group that minimizes the effect of confounding variables. There are several types of control groups, including negative control, vehicle control, positive control, sham control, comparative control ,and naïve control, and it is important to know when to use each of these. ⁶ We refer readers elsewhere for full descriptions of these types of control groups.

Outcome data and measurements

Outcome measurements are made throughout the course of the experiment. These are also known as dependent variables. While multiple outcome measurements can be made, there should be a set of outcome measures that are of greatest importance to the experiment. In statistical analysis, these outcomes are classified as either continuous or categorical. Continuous variables are quantitative data that can include any real number, as well as any real value in between (ex. bodyweight, body temperature). Categorical variables can take on a fixed and limited number of possible values, separating outcomes into categories that are nominal (no intrinsic order to them; ex. male/female/other) or ordinal (intrinsic order present; ex. mild/moderate/severe).

Reducing biases in preclinical studies

Experimental biases can cause significant weakness in the design, conduct, and analysis of in vivo animal studies, producing misleading results and wasting valuable resources. A list of experimental biases have been summarized in Table 4.1 (adapted from Cochrane Handbook for Systematic Reviews of Interventions) ⁷ :

Real-world examples

Scientific and layman literature abounds with examples of pre-clinical research. As a brief recap of the pre-clinical research examples described in this chapter, recall the use of in silico models for patient-specific simulation of yttrium-90 dosimetry, the use of in vitro models to develop three-dimensional cell cultures, and the use of CRISPR technology to generate animal models to mimic specific pathologies. ³ , ⁴

Resources

Exploration of institutional and local resources is generally the recommended starting point for establishing and developing pre-clinical research programs. The National Institutes of Health preclinical research toolbox provides a wide-range of publicly available tools and repositories for preclinical research. Pending institutional and local policies, support from private industry may also be considered.

Table 4.1

References

1. Cressman E.N, Newton I, Larson A.C, et al. State of the research enterprise in IR and recommendations for the future: proceedings from the Society of Interventional Radiology Foundation Investigator Development Task Force. J Vasc Intervent Radiol. 2018;29(6):751–757.

2. Badano A. In silico imaging clinical trials: cheaper, faster, better, safer, and more scalable. Trials. 2021;22(1):1–7.

3. Roncali E, Taebi A, Foster C, Vu C.T. Personalized dosimetry for liver cancer Y-90 radioembolization using computational fluid dynamics and Monte Carlo simulation. Ann Biomed Eng. 2020;48(5):1499–1510.

4. Lorian V. Differences between in vitro and in vivo studies. Antimicrob Agents Chemother. 1988;32(10):1600–1601.

5. Huang W, du Sert N.P, Vollert J, Rice A.S. General principles of preclinical study design. Handb Exp Pharmacol. 2020;257:55.

6. Bate S.T, Clark R.A. The Design and Statistical Analysis of Animal Experiments. Cambridge University Press; 2014.

7. Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M. Cochrane Handbook for Systematic Reviews of Interventions Version 6.2. Cochrane; 2021 Available from:. www.training.cochrane.org/handbook.

Additional reading

1. Good research practice in non-clinical pharmacology and biomedicine. In: Handbook of Experimental Pharmacology. Springer International Publishing; 2020:55–69. doi: 10.1007/164_2019_277 (Chapters: Minimum Information and Quality Standards for Conducting, Reporting, and Organizing In Vitro Research; Minimum Information in In Vivo Research).

Chapter 5: Defining the problem to solve

Samyuktha Balabhadra     Department of Diagnostic and Interventional Radiology, University of Texas-Houston, Houston, TX, United States

Abstract

This chapter describes the steps necessary for creating a plan to pursue a translational research project. Translational research projects, while impactful and imperative to progress in medicine and interventional radiology can be time and resource consuming. It is, therefore, essential to create a plan of action and assess project impact before initiation. The plan should involve a five-step process from start to finish: asking broad questions, literature review, impact evaluation, study design, and finally, initiation.

Keywords

Ex vivo; Impact evaluation; In vitro; In vivo; Institutional review board; Literature review; Study design

Key points

• Create an initial project plan.

• Begin evaluating the impact of your research in the real world.

• Assess the benefit of investing resources into this project.

Why it matters?

Translational research projects, while impactful and imperative to progress in medicine and interventional radiology, can be time and resource consuming. It is, therefore, essential to create a plan of action and assess project impact before initiation.

Step 1

Ask broad questions. What are some broad, overarching topics or issues that you have observed in your work environment? What areas or disease processes could benefit from innovation? Examples may be derived from patients you have seen, cases you have read about in medical literature, or discussions with colleagues. Prepare a list of ideas and questions on topics that you are interested in pursuing.

Case 1: Prostate artery embolization

Problem: Benign prostatic hyperplasia (BPH) causes lower urinary tract symptoms in males; a major quality of life issue (Fig. 5.1). ¹

Flowchart outlining a five-step process in project planning and development.

What happens to patients who do not respond to medication? What happens to patients who are not surgical candidates? What about patients who cannot stop anticoagulants for surgery? How can we create a beneficial, novel therapy for enlarged prostates using IR tools?

Case 2: Inferior vena cava (IVC) filters

Problem: Deep vein thrombosis (DVT) and pulmonary embolism (PE) are a major cause of morbidity and mortality. ²

What happens to patients who do not qualify for anticoagulation therapy? What happens to patients who are not surgical candidates? How can we create a beneficial, novel therapy for PE using IR tools?

Case 3: Arterial gene transfer

Problem: Peripheral arterial disease (PAD) is a major cause of morbidity and mortality. ³ The progression of PAD is such that it requires multiple occasions of medical, surgical, and/or endovascular care.

How can we create a more sustainable therapy? Is there a way to alter the course of this disease at a molecular level? Is there scope for a molecular/genetic transcatheter therapy?

Step 2

Literature review. Gather and read preexisting articles that are relevant to your chosen topic. What have others already found? As you read relevant articles, you will find answers to some of your questions, but you will also continue to build on your list. Continue to narrow your questions and make them more specific. The literature review is also an opportunity to understand the basic pathophysiology of the problem and to understand how pervasive it is within the medical field.

Case 1: Prostate artery embolization

The accepted treatments for BPH are medication, transurethral resection of the prostate (TURP), and open prostatectomy. ⁴ Initially, internal iliac artery branch embolization was performed for massive hemorrhage. ⁵ Few case reports have discussed a benefit in this method reducing the size of prostates overall.

Is it possible to use internal iliac artery branch embolization for BPH? Would reducing blood supply to the prostate be beneficial, possibly reducing its size? If this method is successful in reducing prostate size, it can be high impact, given that this is a condition that affects a majority of middle-aged and elderly men.

Case 2: IVC filters

For many years, the widely accepted treatment for managing DVT has been anticoagulation or open thrombectomy, if possible. For patients who could not receive anticoagulant medication, inferior vena cava (IVC) ligation was practiced. The initial intent for developing the IVC filter was to create an alternative method to IVC ligation for preventing recurrent pulmonary embolism. ⁶

Can an endovascularly delivered mechanical barrier to prevent pulmonary embolism be developed? How can such a barrier be made temporary for the period when anticoagulation is contraindicated? If this method is successful, it may reduce mortality and morbidity rates secondary to pulmonary embolism and its high-risk treatments.

Figure 5.1  Flowchart demonstrating a five-step process using Case 1: Prostate artery embolization.

Case 3: Arterial gene transfer

Current therapy for these patients involves a combination of medication, endovascular therapy, and open surgery. ⁷ These are often in a graded process because PAD is a progressive disease, and it is difficult to develop a sustainable therapy in this context. Several molecular pathways have been identified as being involved with angiogenesis. ⁸ Vascular endothelial growth factor (VEGF) has been identified as one key signaling protein that promotes angiogenesis. Catheter-directed gene therapy is emerging as an arena with great potential.

Would it be possible to transfect the VEGF gene into vulnerable arteries using an endovascular approach? Could this method provide a long-term solution?

Step 3

Impact evaluation. This phase requires a more detailed assessment of the impact of pursuing this project. Establish the goals of the project. What patients or organizations will benefit? How much will they benefit? Is this therapy a sustainable option as compared to current therapies or those already in the pipeline?

Case 1: Prostate artery embolization

The goal of this project would be to develop a minimally invasive therapy to reduce symptoms associated with enlarged prostate with minimal side effects, both in the short term and long term, reduce prostate size, and decrease postprocedural recovery time. The outcomes would have to be comparable to the accepted standard of care; in this case medical management and prostatectomy. If the technique is successful, patients can benefit from a procedure that reduces immediate postprocedural recovery time. In addition, the procedure could preserve sexual and urinary function, which are reported postsurgical effects. ⁹

Case 2: IVC filters

IVC filters were initially developed to prevent recurrent pulmonary embolism in patients; a significant population with high morbidity and mortality. It was reasoned that a mechanical, endovascularly delivered barrier would reduce recovery time as compared to mechanical thrombectomy.

Over time, IVC filters underwent evolution in terms of design and indications. Landmark studies demonstrated that these devices were most beneficial in patients who cannot undergo currently accepted treatments. ¹⁰ Even with the current specific indications, a large subset of patients is receiving benefits from this