REPLY: In his letter commenting on our article (1), Dr. Seto says that he believes the numbers in Table 2 are incorrect. From the table, he calculates a total increase in group A of 28% and an 11% increase in group B. As reported in the footnote of our table, 4 patients were excluded. Therefore, we reported the total increase of only the included patients. This results in a 38% increase of the maximum standardized update value (SUV) for group A and a decrease of 11% for group B.
Dr. Seto's comments focus on 2 issues: first, the calculation of the averages; second, the reasons for excluding patients.
Regarding the first of these issues, the average increase in maximum SUV as presented in Table 2 of our article is the average increase in maximum SUV of the separate lesions per patient. Indeed, the increase in maximum SUV could also have been presented the way Dr. Seto suggests, as the increase between the averages of the initial maximum SUV and the averages of the maximum SUV at follow-up. Calculating the increases in this way would result in a 31% increase for group A and a 14% decrease for group B. Clearly this would not change our overall conclusion, that is, an increase in maximum SUV for the patients in group A, compared with the decrease in maximum SUV for the patients in group B.
Regarding the second issue, reports show that serial measurements of lesions with high initial 18F-FDG uptake tend to be inaccurate because of rapid growth and the development of central necrosis (2). For this reason, we described in detail the exclusion of patients in whom most metastatic lesions showed SUVs greater than 10 on the initial 18F-FDG PET scan. Dr. Seto suggests also excluding those patients who showed SUVs greater than 10 on the second 18F-FDG PET scan. We do not completely understand this suggestion since these final scans do not need any serial measurement afterward.
Furthermore, Dr. Seto suggests that, rather than excluding patient 9, we selectively count those lesions with SUVs less than 10 at the initial measurement. In our opinion, it is more reasonable to exclude this patient than to selectively analyze only the 2 lesions with SUVs less than 10. If we were to include only these 2 lesions, the increase in maximum SUV for one lesion would be 15% (initial maximum SUV, 9.7; maximum SUV at follow-up, 10.2) and the increase in maximum SUV for the other lesion would be 16% (initial maximum SUV, 7.8; maximum SUV at follow-up, 9.1).
Therefore, we do not understand Dr. Seto's questioning the overall conclusions on the basis of the exclusion of patient 9, as inclusion of the 2 lesions of patient 9 is in line with our conclusions.
Moreover, Dr. Seto questions the exclusion of patient 17 in group B. We excluded this patient because a second liver metastasis had developed directly adjacent to the first metastasis. It proved to be impossible to reliably determine the separate SUVs on 18F-FDG PET. Therefore, we believe that excluding this patient is valid.
Although significant differences between both groups were observed, we did not overestimate the results: our conclusions were qualified by the admission that only a limited number of patients were studied. Thus, we are well aware that our findings may not be definitely conclusive for primary tumor–induced growth inhibition of metastatic disease. However, our results are in concordance with earlier reports (3,4) in which an increase in vascularization and growth of liver metastases was reported after resection of primary colorectal carcinoma.
Footnotes
-
COPYRIGHT © 2009 by the Society of Nuclear Medicine, Inc.