L S Wilson1, J Lightwood. 1. Department of Pharmacy, University of California, San Francisco, CA 94143, USA. lwilson@itsa.ucsf.edu
Abstract
OBJECTIVE: Most studies show protective effects of non-steroidal anti-inflammatory drugs (NSAIDs) against polyps and colorectal cancers (CRCs) of up to 50%. Current models are unable to directly estimate changes in effects of chemoprevention on CRCs. The purpose is to develop a model to examine effects of changes in growth rates of polyps on surveillance intervals and risk of CRC. METHODS: The growth model simulates 500 people after polypectomy, estimating number and size of polyps annually over 10 years. Each polyp is assigned a random growth rate consistent with distributions of empirically observed growth assumed to follow a log linear model. Rates of CRC were calculated from largest polyps distributed to people. RESULTS: Simulated distributions of polyps and CRCs closely match empirical estimates which confirms the usefulness of the model. If polyp growth is 25% of normal, the number of cancers by year 10 after index colonoscopy decrease from 146 to only 57/100 000 for those in risk group 0 (no polyps at index colonoscopy) and from 840 to 124/100 000 for those of risk group 3 (4 or more polyps). CONCLUSIONS: This is the first model based on polyp growth rates. The CRC rates suggest that for those with no polyps on index colonoscopy, surveillance may be as for people of average risk (7-10 years), whereas those with one polyp or more need more surveillance (2-5 years). The use of the model is the indication that surveillance intervals could be increased by as much as 2-10 years if the growth rates of polyps are slowed.
OBJECTIVE: Most studies show protective effects of non-steroidal anti-inflammatory drugs (NSAIDs) against polyps and colorectal cancers (CRCs) of up to 50%. Current models are unable to directly estimate changes in effects of chemoprevention on CRCs. The purpose is to develop a model to examine effects of changes in growth rates of polyps on surveillance intervals and risk of CRC. METHODS: The growth model simulates 500 people after polypectomy, estimating number and size of polyps annually over 10 years. Each polyp is assigned a random growth rate consistent with distributions of empirically observed growth assumed to follow a log linear model. Rates of CRC were calculated from largest polyps distributed to people. RESULTS: Simulated distributions of polyps and CRCs closely match empirical estimates which confirms the usefulness of the model. If polyp growth is 25% of normal, the number of cancers by year 10 after index colonoscopy decrease from 146 to only 57/100 000 for those in risk group 0 (no polyps at index colonoscopy) and from 840 to 124/100 000 for those of risk group 3 (4 or more polyps). CONCLUSIONS: This is the first model based on polyp growth rates. The CRC rates suggest that for those with no polyps on index colonoscopy, surveillance may be as for people of average risk (7-10 years), whereas those with one polyp or more need more surveillance (2-5 years). The use of the model is the indication that surveillance intervals could be increased by as much as 2-10 years if the growth rates of polyps are slowed.
Authors: Jane V Carter; Henry L Roberts; Jianmin Pan; Jonathan D Rice; James F Burton; Norman J Galbraith; Maurice R Eichenberger; Jeffery Jorden; Peter Deveaux; Russell Farmer; Anna Williford; Ziad Kanaan; Shesh N Rai; Susan Galandiuk Journal: Ann Surg Date: 2016-10 Impact factor: 12.969
Authors: Matthew D Rutter; Amit Chattree; Jamie A Barbour; Siwan Thomas-Gibson; Pradeep Bhandari; Brian P Saunders; Andrew M Veitch; John Anderson; Bjorn J Rembacken; Maurice B Loughrey; Rupert Pullan; William V Garrett; Gethin Lewis; Sunil Dolwani Journal: Gut Date: 2015-06-23 Impact factor: 23.059
Authors: Gustavo Hernandez-Suarez; Maria Carolina Sanabria; Marta Serrano; Oscar F Herran; Jesus Perez; Jose L Plata; Jovanny Zabaleta; Albert Tenesa Journal: Eur J Hum Genet Date: 2014-02-12 Impact factor: 4.246