Editor:
Pickhardt et al. provided further evidence that “flat lesions…do exist in a typical US screening population.” We agree. However, we disagree with the authors’ conclusion that “flat lesions are detectable on [computed tomographic colonography] in a proportion relative to polypoid lesions that is similar to recent endoscopic experiences.” They did not present data to support their conclusions.
Pickhardt et al. used optical colonoscopy as the reference standard because current detection and diagnosis of nonpolypoid colorectal neoplasm (NP-CRN) require optical visualization and pathologic confirmation. Concerns regarding the quality of their standard have been published, raising questions regarding the validity of their results and conclusion . Herein, we raise additional concerns. The detection of NP-CRN requires endoscopists to search for abnormal patches of mucosa , in addition to looking for protruding lesions. NP-CRN may be detected as a subtle patch of reddish mucosa, which is devoid of vascular pattern or innominate grooves. NP-CRN may cause friability and wall deformity. In general, prior to the public awareness of NP-CRN in March 2008 , gastroenterologists were not focused on the importance of NP-CRN or on their detection method. The data included by Pickhardt et al. were largely collected before 2008, and there is no indication that the participating colonoscopists practiced the detection methods required to find NP-CRN. In fact, the low adenoma detection rate reported (2% [106 of 5107]) for flat lesions ≥ 6 mm (the best scenario assuming that all lesions were neoplastic) provides numerical evidence of the limitation of their study. In our 1-year cross-sectional prevalence study, 24% of patients had at least one flat neoplasm ≥ 6 mm. Others have reported a similar prevalence rate . Pickhardt et al.’s colonoscopic image examples of NP-CRN illustrate, in fact, a suboptimal bowel preparation for the adequate detection of NP-CRN. Furthermore, the reported high rate of perforation (1 in 500 colonoscopies) is among the highest published in the recent literature. Such colonoscopic quality indicators reflected in their study suggest that careful search for subtle mucosal findings was not the priority.
It is convention for an epidemiologic study to present results in tabular form for easy understanding, comparison, and assessment of the completeness and accuracy of the data. In epidemiologic studies of colorectal neoplasms, the inclusion of prevalence data is standard. Pickhardt et al. did not systematically describe the prevalence of either polypoid or nonpolypoid neoplasms in text or tabular form. We are unclear as to which “recent endoscopic experiences” were referred to by the authors in their conclusions. If they meant their reference colonoscopy, Pickhardt et al. indeed reported that “flat lesions are detectable on [computed tomographic colonography] in a proportion relative to polypoid lesions that is similar to” suboptimal colonoscopic experiences. Reviewers, associate editors and editors, and ultimately readers should ensure and insist that data be analyzed and presented systematically. Otherwise, a paper is less scientific and more rhetorical.
References
1. Pickhardt P.J., Kim D.H., Robbins J.B.: Flat (nonpolypoid) colorectal lesions identified at CT colonography in a US screening population. Acad Radiol 2010; 17: pp. 784-790.
2. Regula J., Polkowski M.: CT colonography versus colonoscopy for the detection of advanced neoplasia. N Engl J Med 2008; 358: pp. 88-89.
3. Soetikno R.M., Kaltenbach T., Rouse R.V., et. al.: Prevalence of nonpolypoid (flat and depressed) colorectal neoplasms in asymptomatic and symptomatic adults. JAMA 2008; 299: pp. 1027-1035.
4. Lieberman D.A.: Nonpolypoid colorectal neoplasia in the United States: the parachute is open. JAMA 2008; 299: pp. 1068-1069.
5. Lieberman D.A., Holub J.L., Moravec M.D., et. al.: Prevalence of colon polyps detected by colonoscopy screening in asymptomatic black and white patients. JAMA 2008; 300: pp. 1417-1422.
6. Kim D.H., Pickhardt P.J., Taylor A.J., et. al.: CT colonography versus colonoscopy for the detection of advanced neoplasia. N Engl J Med 2007; 357: pp. 1403-1412.