Gliomas: Survival, origin and early detection

Abstract

I did some of my training with Paul Bucy. He had a special interest in the surgical treatment of glioblastoma. Bucy believed, as did many “cancer surgeons” of his day, that tumors resulted from good cells becoming bad cells that formed a mass of tumor and that cells from the tumor’s periphery invaded the surrounding normal tissue. He correctly observed that malignant gliomas usually grew locally and rarely metastasized outside of the central nervous system. If there was any surgically curable “cancer”, he believed, it was a glioblastoma; all that was necessary for a cure was an aggressive enough resection with an adequate margin.[1] Of course, his patients died right on schedule, just like anybody else’s patients. Bucy believed that this was because we just could not identify the true margin of the neoplasm at surgery and that the resection was rarely sufficiently adequate to provide a cure. The advent of computed tomography (CT scanning), magnetic resonance imaging (MRI) and image-guided neuronavigation would change all of this by potentially allowing us to accurately resect as much of a glioma as we chose to resect. Well, I have spent a career trying to cure gliomas with high technology-based surgery – in particular, imaging-based stereotactically guided volumetric resections – but the long-term survival in the vast majority of these patients is not much better than it was 60 years ago! We just do not hurt these patients as badly as we did 60 years ago. And what do we do? We keep throwing more and more expensive surgical high technology at the problem with marginal improvements in survival, if any. To be sure, there are some gliomas that we can cure with modern surgical techniques, such as pilocytic astrocytomas, the occasional oligodendroglioma, neurocytomas, gangliogliomas, subependymomas and a few xanthoastrocytomas and protoplasmic astrocytomas. But this is not a credit to neurosurgeons and our modern surgical methods. It is a function of the growth pattern of these particular tumors that lend themselves to complete and curative surgical excision. These tumors have a distinct boundary where tumor stops and normal brain begins. All that a surgeon has to do in these cases is identify the plane between tumor and surrounding brain tissue, develop that plane and remove the tumor. Image guidance helps a bit. But I will point out that Donald Matson claimed a 50% surgical cure rate in pilocytic astrocytomas over 50 years ago – without any “high technology”. Nonetheless, the “curable” tumors listed above are relatively rare compared to the overwhelmingly more common “fibrillary astrocytomas”, oligodendrogliomas and mixed gliomas. How are we doing with these tumors? Not so great! To be sure, there are many reports in the literature which show that patients having “total resection” and adjuvant therapy do better and live longer than those undergoing biopsy and adjuvant therapy. Comparisons to historical controls attempt to demonstrate the benefit of modern surgical techniques over methods used by past generations. An example of just such an exercise is shown in Figure 1. Figure 1 Post surgical survival following resection in patients with grade IV astrocytomas (glioblastoma) in a recent unpublished series (Kelly 2000) compared to survival curves adapted from earlier studies in the literature (Kelly 1992,[3] Burger 1986,[2] Bucy ... These life table survival curves compare my own experience with cases compiled in the years 1992 (published)[3] and 2000 (not published) to the 1986 series published by Burger et al,[2] Jelsma and Bucy’s series from 1967[4] and Ringertz’s experience from 1950.[2] At first glance at Figure 1, it appears that over the years, we have made some progress with better median and 2-year survivals. Regrettably, these experiences are not really comparable for two important reasons. First, survival times are measured from when surgery is performed and histology is available. Modern imaging methods allow patients to be diagnosed much earlier – usually at the onset of the first symptoms – in contrast to patients from the 1950s and 1960s, who could go for weeks or months before a diagnosis was made and surgery performed. Modern series have the benefit of therapy being delivered earlier in the natural history of the disease and a survival starting point that could be weeks or months earlier than in the past decades. Secondly, modern patients have had the benefit of more effective radiation therapy with linear accelerators instead of cobalt units and more specific chemotherapy. Indeed the improved 2-year survival noted in my 1992 and 2000 series more likely represents the efficacy of carboplatin in the 1980s and temazolamide in the 1990s and not necessarily “better surgery”.