Journal of Neuro-Oncology 45: 111–116, 1999. © 2000 Kluwer Academic Publishers. Printed in the Netherlands. Laboratory Investigation Brain tumor invasion rate measured in vitro does not correlate with Ki-67 expression Sami Khoshyomn 1 , Sean Lew 1 , Joseph DeMattia 2 , Elan B. Singer 3 , and Paul L. Penar 1 1 Division of Neurosurgery, Department of Surgery, University of Vermont College of Medicine, Burlington, VT, USA; 2 Department of Neurosurgery, New York Medical College, Valhalla, NY, USA; 3 Department of Surgery, Mt. Sinai Medical Center, New York City, NY, USA Key words: glioma, meningioma, proliferation makers, invasion Comments As the authors state, the literature on measures of the proliferative potential of various tumor types is not consistent and, practically speaking, these measures add little, if anything, to standard histological review when considering patient management. The insight gained from the current investigations as well as those cited in the discussion will lead to better understanding of the mechanisms of tumor growth and invasion and thereby, hopefully, patient management. Dr. J. Rock (NewYork City, NY, USA) Intuitively, mitosis and cell migration/invasion are thought to be temporally exclusive phenomena, as a cell is unlikely to be able to commit its cytoskeletal proteins and biochemical machinery for both cell division and motility concurrently – such is the basis for the ‘go or grow hypothesis’. In their article, Khoshyomn et al. present data that support this premise, as there was no observed correlation between the proliferative and invasive index of the cells derived form patient tumor specimens. As both proliferation and invasion contribute to tumor growth and affect patient outcome, assessment of one factor in exclusion of the other likely offers a highly inaccurate predictor of patient prognosis. The question remains as to whether a combined analysis of both phenotypes (i.e., proliferation and invasion) may be better correlated with patient outcome. Furthermore, while Ki-67 is a marker of a cell that is in cycle, it does not distinguish between the various phases of the cell cycle. Cell cycle regulatory proteins such as cyclins and their respective kinases may reflect better markers than Ki-67. Dr. R. Sawaya/Joon H. Uhm Summary The need for more accurate prediction of the biological behavior of brain tumors has lead to the use of immunohisto- chemical methods for assessment of proliferating cell nuclear antigens such as Ki-67 . There is a variable association of glioma Ki-67 labeling index with patient survival. Brain invasion by individual tumor cells also defines biologi- cal aggressiveness, and can be assessed in vitro. Further, proliferation and migration seem to be mutually exclusive behaviors for a given cell at a point in time. We studied the relationship between Ki-67 labeling index and invasion rate in a group of 10 gliomas, and 2 meningiomas. Human tumor spheroids obtained from operative speciments were co-cultured with fetal rat brain aggregates, and invasion rate was measured by confocal microscopic observation. There was no correlation between two measures of invasion and Ki-67 labeling. This finding supports the dichoto- mous nature of glioma proliferation and invasion, and may in part explain the limited usefulness of proliferation marker labeling.