LAUSR

To the Editor: The authors should be commended for trying to resolve the controversy around the management of Bosniak 3 cysts, whose malignant potential approaches 50%. They have tried to define the factors that can help us in grouping the management of Bosniak 3 cysts based on their classification into the categories of 3s (septated enhancing cysts) and 3n (cysts with wall or septation-only nodularity). However, we observed several issues that need to be addressed. All images underwent individual repeat review by urologists, and each cyst was assigned a Bosniak class and measured. In case of discrepancy the imaging was also reviewed by a blinded radiologist. It is not clearly mentioned whether the urologist was also blinded to the outcome. If not, this implies that for an index patient if a greater number of images (average 6.8 in the study) are available for review, it can be safely assumed that the patient is being followed for a benign cyst, and thus the urologist will tend to rate the cyst as 3s due to ascertainment bias. In the Discussion section the authors themselves state that there is high interobserver variability for characterizing Bosniak cysts. The authors also state that each Bosniak cyst was classified by a urologist, and only images of controversial complex cysts were reviewed by an experienced radiologist. It would have been helpful to provide the number of images reviewed by the blinded radiologist and the distribution of assignment. An analysis of interobserver variability using Cronbach a or interclass correlation would have been useful. A minor point is that it is not clearly stated how the images were retrieved, ie whether imaging involved use of a picture archiving computerized system or the authors used the films of cross-sectional imaging for assessment. This issue also factors in since the duration of study is longer than a decade and the quality of imaging will affect the study outcome. We fully agree with the statement by Reese in the accompanying editorial comment that cystic lesions are indolent and surveillance should be considered strongly for all cystic lesions.