This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Bagley, C. M. Articles by Sorensen, A. G. Search for Related Content PubMed PubMed Citation Articles by Bagley, C. M., Jr Articles by Sorensen, A. G. Social Bookmarking                            What's this?

Measurement of Brain Tumor Volumes by the Perimeter Method

Puget Sound Cancer Centers, Seattle, WA

To the Editor:The careful analysis by Sorensen et al¹ on measurement of brain tumor volumes by the perimeter method clearly establishes this to be a more accurate and probably simpler method of brain tumor volume measurement. The authors consider a partial response to be a 50% or more decrease in tumor volume.

However, for decades a 50% decrease in tumor cross-sectional area has been the standard used by medical oncologists, including the World Health Organization Handbook² that Sorensen cited, for determining that a tumor has reached a partial response. True, measurement of glioblastoma shrinkage by careful perimeter volume measurements may be reasonable and more accurate for that kind of tumor. However, this has not been and should not be standard for measurement for tumors elsewhere in the body, because the vast oncology literature on tumor response to chemotherapy has been based on a 50% reduction in area, and area measurements are easily performed by the oncologist on chest x-rays, computed tomography scans, and lymph nodes.

The volume criterion for response is less stringent than area; a 50% reduction in tumor volume represents only a 37% reduction in cross-sectional area. Conversely, to reach the 50% reduction in area, a tumor has to reduce in volume by approximately 65%. If the oncology community is going to change from using cross-sectional area to volume as the standard for tumor shrinkage in describing responses to chemotherapy, it should be done explicitly by international consensus and not by nibbling away at it, disease by disease.

REFERENCES

1. Sorensen AG, Patel S, Harmath C, et al: Comparison of diameter and perimeter methods for tumor volume calculation. J Clin Oncol 19: 551-557, 2001[Abstract/Free Full Text]

2. World Health Organization: Handbook for Reporting Results of Cancer Treatment. Geneva, Switzerland, World Health Organization, 1979

Response

Massachusetts General Hospital, Charlestown, MA

In Reply:We appreciate Dr Bagley’s interest in our article. We agree with Bagley that a 50% change in volume differs from a 50% change in cross-sectional area, and we concur that a 50% decrease in cross-sectional area has been a standard for response definition in the past. In our article, the choice of a 50% change in volume (rather than, for example, a 65% reduction in volume) was somewhat arbitrary and chosen for convenience only. Indeed, our results demonstrating the potential impact of the greater variability of the cross-sectional method compared with the volumetric method remain intact and relevant no matter what threshold is chosen to determine response.

We purposely hoped to avoid the ongoing discussion (some might suggest controversy) surrounding how the oncology community chooses to define various response parameters. Our aim was instead to illustrate that efficient methods now exist to reduce the variance of these measures, and that this reduction in variance is large enough to improve estimates of tumor response and thereby potentially allow smaller patient groups to be used to test novel therapies. This reduction in variance is not a result of choosing a volumetric method instead of a cross-sectional method, but rather a result of computer assistance in the segmentation.

However, Bagley may imply in his final comments about the accessibility of film images to the oncologist that the simpler measurement is also preferable in clinical trials. We believe one of the implications of our results is that an oncology trial that uses semi-automated tumor quantitation could require significantly fewer subjects, and that this added benefit could offset the issue that fewer investigators could perform the analysis on-site. We thank Dr Bagley for agreeing that this reduction in variance seems useful and apologize for any confusion that our choice of measurement change may have caused.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Bagley, C. M. Articles by Sorensen, A. G. Search for Related Content PubMed PubMed Citation Articles by Bagley, C. M., Jr Articles by Sorensen, A. G. Social Bookmarking                            What's this?