Pranav Suresh
Title of Project: The Accuracy of Tumor Identification
Date: 9 January 2014
Statement of Purpose:
The SDTM (Study Data Tabulation Model) method of identifying tumors clinically and lesions is widely used throughout the globe, but an increasing number of scientists are finding minor problems with it, specifically with the methodology used. (Litzsinger, 2014)The actual scientific process and equipment used vary from hospital to hospital, however. The Mayo Clinic is experimenting with a new method, the core lab method, with the goal of increasing accuracy of tumor detection. The question is, does it surpass the clinical (old) method? My job will be to help analyze data from usage of both methods of analyzing data (confusing, am I right?) and determine which is better.
Background:
I have no prior experience with lesions themselves, though my interest and study of biology and statistics should help. I am very interested in radiology as well, which will make this question quite interesting to research. I hope to expand both my biology and statistics knowledge from this project, specifically radiology and analysis of quantitative data. I will also hope to gain valuable internship experience, and to be able to help Dr. Hara and others where possible. If I am able to help the Mayo Clinic determine that this new method of detecting tumors is indeed more accurate, it will help patients in the future, and though I may not have been a big part in the research, I can still feel happy about it.
Prior Research:
There is some context to be provided on how tumors and/or lesions are normally identified before analyzing the way it is done.
Tumor markers are markers that can be produced by both cancerous cells and immune cells in response to cancer cells, as stated by the American Cancer Association. (ACA, 2011) These markers are easier to detect than the cancer cells themselves, so many methods to diagnose tumors involve them. To increase the overall accuracy of diagnosis, this may be combined with other tests, such as biopsies. (CRU, 2009)
Tumor markers are measured by taking a sample of possibly tumorous tissue, or blood in the case of leukemia, and analyzing them of a period of days at a laboratory. If the patient has already been diagnosed with cancer, these markers can find whether treatment has been working or failing. (Litzsinger, 2014)
Examples of types of tumor markers are ALK, AFP, B2M, beta-hCG, BRAF, Mutation V600E, and many others. (Mayo, 2014) While these are not completely accurate, many hospitals and laboratories are doing research to find methods and markers with increased accuracy, such as in Mayo Clinic. For example some are using a new type of marker known plainly as biomarkers, which are in theory able to detect malign tumors earlier, thus limiting mortality from a possible cancer. (Crinion, 2012)
But this all remains constant. What really changes between the two methods is the analysis of numbers gathered from it.
To start off, the clinical analysis uses multiple doctors over the course of several days as they become available. The core lab uses one person through the same period to increase accuracy, as different individuals may interpret data uniquely. The core lab also compares lesion progress to all previous records, rather than the clinical method, which compares only to the most recent record. In addition, the core lab has some parts of the data collection process automated, clearing human error in those areas.
Significance:
It goes without saying that body-internal cancers kill many every year, and therefore the screening techniques for finding and subsequently destroying tumors must be as accurate as possible. If the accuracy of even one diagnosis is improved through an implementation, or lack thereof, of Mayo’s new core lab method for scanning for tumors, the internship will have been worth it to be a part of that. The core lab is merely an experimental add-on for the time being, but if successful, can become part of mainstream medical practice.
Description:
I will be analyzing (compilation, comparison, and assessment, followed by a fully formulated hypothesis) data sets from tumor scans, in addition to doing minor work for doctors who need it. Depending upon the number of patients available, I will be comparing anywhere between 20 and 50 patients’ data over the course of 10 weeks. While my on-site mentor is Dr. Hara, I will be working with and helping a few others staff members as well. I will present my findings and a conclusion at the end of the year.
Methodology:
I would start by formulating a hypothesis regarding the two methods. Then, I would analyze data sets on a daily basis and alter my existing conclusion as appropriate. Finally, I would make a more lay-audience friendly PowerPoint and speech to explain my and the clinic’s findings and their implications.
As for the data analysis specifically, I would, in addition to data sets from the standard method and the new method, need a control group of sorts, showcasing the base values for each quantity in a data set for a tumor. Variables would include, but are not limited to: doctor, disease, lesion/tumor size, lesion/tumor number, etc.) On a weekly basis, I would need to compare all three data sets and their effectiveness. Results will be compiled to form a conclusion at the end of the internship.
Mayo Clinic has informed me that I will be given access to the clinical database for lesions and tumors. Using it, I can make charts for both the clinical method and the core lab.
Problems:
I don’t know anything about radiology specifically, so I will have to research the background behind the field in order to understand my role in Mayo Clinic’s experimental approach. In addition, I will need to find a way to summarize medical data into a meaningful conclusion that anyone will be able to grasp. I will also need to dress professionally being an intern and all. (Something I’ve had a problem with - and by that I mean I always forget - for a long time.)
Bibliography:
Study Data Tabulation Model (SDTM). (2014, November 30). Retrieved December 10, 2014, from http://www.cdisc.org/sdtm
Cancer screening. (n.d.). Retrieved December 18, 2014, from http://en.wikipedia.org/wiki/Cancer_screening
How is cancer detected? (2009, November 16). Retrieved December 18, 2014, from http://www.cancerresearchuk.org/cancer-info/cancerandresearch/all-about-cancer/what-is-cancer/detecting-cancer/detectingcancer
Tumor Markers. (2011, December 7). Retrieved December 18, 2014, from http://www.cancer.gov/cancertopics/factsheet/detection/tumor-markers
Cancer. (2014, February 22). Retrieved December 18, 2014, from http://www.mayoclinic.org/diseases-conditions/cancer/in-depth/cancer-diagnosis/art-20046459
Can brain and spinal cord tumors in adults be found early? (2014, March 5). Retrieved December 18, 2014, from http://www.cancer.org/cancer/braincnstumorsinadults/detailedguide/brain-and-spinal-cord-tumors-in-adults-detection
Kenny, S., & Litzsinger, M. (n.d.). Strategies for Implementing SDTM and ADaM Standards. Retrieved December 18, 2014, from http://www.pharmasug.org/2005/FC03.pdf http://www.pharmasug.org/2005/FC03.pdf
Radiology. (n.d.). Retrieved December 18, 2014, from http://en.wikipedia.org/wiki/Radiology
Interventional Treatments for Liver Disease. (n.d.). Retrieved December 18, 2014, from http://www.sirweb.org/patients/liver-disease/ (Not entirely related but still provided context for me.)
Crinion, J. (2012, January 1). Download PDFs. Retrieved January 11, 2015, from http://www.sciencedirect.com/science/article/pii/S1053811912007732
