HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Erturk, S. M. Search for Related Content PubMed PubMed Citation Articles by Erturk, S. M. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?

CT Evaluation of Gastrointestinal Stromal Tumors Treated with Imatinib Mesylate

Brigham and Women's Hospital Boston, MA 02120

I read with great interest the article by Choi et al. [1] in the December 2004 issue of the AJR in which they evaluated the response of gastrointestinal stromal tumors after imatinib mesylate treatment. The authors attempted to correlate changes in tumor density on CT with changes in the maximum standardized uptake value (SUV_max) on FDG PET; they used 4 x 4 contingency tables and chi-square tests to assess associations between the grading of the average percentage of change in tumor density and in tumor size. In the 29 patients who underwent both CT and FDG PET in their study, no significant association was found between the percentage of change in tumor density on CT and the SUV_max on FDG PET (p = 0.3088, chi-square test).

Because this study had a matched design and includes the data of repeated measurements in the same patient population, the use of the chi-square test for statistical evaluation is not appropriate. For correlated proportions, like those in the study of Choi et al. [1], the correct choice would be the McNemar test [2, 3]. However, the McNemar test is available only for 2 x 2 contingency tables, and, therefore, its use would necessitate combining the groups so that the data would fit into a 2 x 2 table.

In fact, regarding the study purpose and the symmetric four grade scales used to evaluate tumor response, the most proper statistical solution for this study would be to use the kappa statistic; the kappa statistic is a common way to measure interobserver agreement [4, 5]. With use of the linear weighting kappa statistic, the authors would have been able to show the associations between the grade of changes in CT tumor density, tumor size, and SUV_max more comprehensively. Based on the data given in Tables 3 and 4 in the article by Choi et al. [1], the kappa value for the agreement between the grading of the change in SUV_max on FDG PET images and in tumor density on CT images is 0.31 (95% confidence interval [CI] = 0.09–0.52), whereas the agreement between the change in SUV_max and in the tumor size is 0.04 (95% CI = 0.00–0.09). This information might have led the authors to additional conclusions.

References

Top References

 

1. Choi H, Charnsangavej C, de Castro Faria S, et al. CT evaluation of the response of gastrointestinal stromal tumors after imatinib mesylate treatment: a quantitative analysis correlated with FDG PET findings. AJR 2004; 183:1619 -1628[Abstract/Free Full Text]
2. Hawass NE. Comparing the sensitivities and specificities of two diagnostic procedures performed on the same group of patients. Br J Radiol 1997; 70:360 -366[Abstract]
3. Daly LE, Bourke GJ, McGilvray J. Interpretation and uses of medical statistics, 4th ed. Oxford, UK: Blackwell Scientific Publications, 2001: 130-132
4. Fleiss JL. Statistical methods for rates and proportions, 2nd ed. New York, NY: John Wiley & Sons,1981 : 19-49
5. Ichikawa T, Haradome H, Hachiya J, et al. Pancreatic ductal adenocarcinoma: preoperative assessment with helical CT versus dynamic MR imaging. Radiology 1997;202 : 655-662[Abstract/Free Full Text]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				R. L. Ehman, W. R. Hendee, M. J. Welch, N. R. Dunnick, L. B. Bresolin, R. L. Arenson, S. Baum, H. Hricak, and J. H. Thrall Blueprint for Imaging in Biomedical Research Radiology, July 1, 2007; 244(1): 12 - 27. [Full Text] [PDF]

This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Erturk, S. M. Search for Related Content PubMed PubMed Citation Articles by Erturk, S. M. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?