The Story of Glioblastoma: History and Modern Correlates

Glioblastoma remains a formidable pathologic entity. Without the tools necessary to make a neurologic diagnosis, early medical and surgical clinicians were likely not only puzzled but also appalled by the downward trajectory of patients affected by this tumor. The development of neurosurgery and neuropathology served as the first necessary steps in understanding glioblastoma. Despite major diagnostic and therapeutic advances, clinicians continue to struggle in providing favorable outcomes for patients with glioblastoma. This chapter provides a historical context for glioblastoma and outlines the evolution of the diagnosis and management of this tumor.

History of glioblastoma: from trephination to World Health Organization classification

Although evidence of trephination or trepanation has been found from ancient cultures dating back to the late Paleolithic period from around the world, the first medical description of the procedure was documented by Hippocrates. Practitioners performed trephination for traumatic injuries (to elevate depressed skull fractures) and for a variety of other reasons, including to alleviate seizures, headaches, superficial growths, and psychiatric maladies ( Fig. 1.1 ). The ancient physician and surgeon, Abu al-Qasim Al-Zahrawi (Latinized to Albucasis) of Andalusia, developed and described numerous surgical instruments and procedures in the volumes of Kitab al-Tasrif (The Method of Medicine), including operations to treat neurosurgical disorders. These disorders included tumors of the central nervous system.

The acceptance and formalization of the scientific method, advancements in clinical and laboratory medicine, and the development of equipment like the light microscope brought forth new knowledge and opportunities to study human disorders. Between 1856 and 1865, Rudolph Ludwig Carl Virchow first described neuroglia, defined the gliomas and separated them into what are now considered low-grade and high-grade disorders, and developed the foundation for pathologic study. At the time, tumors found in the brain at autopsy were named according to the presumed (normal) cellular counterpart. Given the available staining and visualization techniques, many of the cell types that are now considered glial cells were not clearly defined; they were thought of as elements of connective tissue without a cellular origin. Based on Camilo Golgi’s work in identifying foot processes along neurons (1873), and Michael von Lenhossék’s description of the astrocyte (1891), the idea that glial cells provided support for neurons was propagated. Santiago Ramón y Cajal and Pío del Río-Hortega contributed much to the understanding of glial cells and the cellular architecture of the brain using gold-based (labeling glial fibrillary acidic protein) and platinum-based (labeling oligodendrocytes) agents.

In 1926, Percival Bailey and Harvey Cushing, both American Midwesterners by birth and upbringing, collaborated to publish A Classification of the Tumors of the Glioma Group on a Histogenetic Basis with a Correlated Study of Prognosis . Bailey used histologic staining techniques to study 254 gliomas from Cushing’s series. These and an additional 160 gliomas were used in classifying 13 groups according to the cellular configuration. The classification scheme was later condensed. In addition, tumors were grouped according to patient survival. Before this work, central nervous system tumors were generally considered gliomas; prognosis was not clearly linked to histopathologic diagnosis. Because of the important and practical implications of this work, it was received with great interest worldwide ( Figs. 1.2–1.4 ).

Cushing and Bailey identified spongioblastoma multiforme as a distinct tumor with a specific cell of origin based on its histologic appearance, which appeared different from the other gliomas (see Fig. 1.2 ). Despite heterogeneity (prompting the term multiforme) when visualized under the microscope, patients with these tumors uniformly experienced rapid declines in their clinical trajectories. By the 1940s, spongioblastoma multiforme became better known as glioblastoma multiforme.

The German neuropathologist Hans Joachim Scherer first hypothesized the concept of primary versus secondary glioblastoma in 1938 and published a series of articles substantiating this into the 1940s. His ideas were ahead of his time in both a theoretic and scientific sense. He noted that patients with secondary glioblastoma had long clinical courses compared with those with primary glioblastoma, emphasizing that they could be distinguished from biological and clinical perspectives. Although this was noted by both Scherer and Cushing, both were in disagreement over systems of classification. Penfield, Kernohan, Sayre, Schaffer, Bergstrand, Purdy, and Olivecrona proposed differing systems of classification based on novel clinical and histopathologic findings; these systems were introduced with considerable bias.

A consensus system for classification was needed. From 1956 to 1979, the World Health Organization (WHO) recruited 23 centers worldwide consisting of approximately 300 pathologists to provide microscopic samples of brain tumors for classification. In 1979, the first edition of the WHO Classification of Tumors of the Central Nervous System was published. Subsequent editions were published in 1982, 2000, and 2007. At the time of this writing, a fifth edition is due for release (See Chapter 4 for details).

Historical perspectives on the diagnosis of glioblastoma

Imaging technologies such as radiograph (Roentgen, 1895), ventriculography (Dandy, 1919), angiography (Lima and Moniz, 1927), computed tomographic imaging (Hounsfield and Cormack, 1971), and magnetic resonance (MR) imaging (Lauterbur, Mansfield, and Damadian, 1977) significantly changed the diagnosis and subsequent management of glioblastoma. At the time of this writing, T1-weighted and T2-weighted imaging sequences on MR imaging, with and without gadolinium administration, serve as standard imaging sequences for newly diagnosed glioblastoma. Advanced imaging sequences developed in the last 15 to 20 years can be used to ascertain recurrence and differentiate this from radiation-related changes. The current contrast agent of choice (gadolinium) results in enhancement on T1-weighted imaging within the major cellular components of a tumor based on breakdown of the blood-brain barrier (BBB).

Even at present, many clinicians consider areas of contrast enhancement on MR imaging to equate to borders of neoplastic disease in the brain. It is now understood that the genetic and phenotypic heterogeneity of glioblastoma is spatial and may account for neoplastic cells identified in the brain outside of enhancing regions on MR imaging. When Walter Dandy performed right hemispherectomy procedures on patients with this tumor in 1928, relapses were noted ( Fig. 1.5 ). Several decades later, Patrick Kelly performed stereotactic biopsies outside of regions of contrast enhancement on imaging, showing neoplastic disease in nonenhancing tissue ( Fig. 1.6 ). Over time, through these and numerous other studies, clinicians developed an understanding that glioblastoma is a microscopic and not a macroscopic disease.

The many scientific and biomedical advancements made in the twentieth century contributed to the present understanding of the genetic and environmental basis for disease. In 1974, the p53 tumor suppressor gene was discovered. Its role in the pathogenesis of glioblastoma was later shown through a large body of scientific evidence. It is now understood that primary glioblastoma is hallmarked by epidermal growth factor ( EGFR ) overexpression and that p53 mutations characterize secondary tumors. Collaboration among centers and data from integrated genomic profiling have identified subtypes of glioblastoma typified by genetic alterations and variability in expression. In the last 10 years, isocitrate dehydrogenase ( IDH ) mutations were found with a high frequency in secondary glioblastomas, and patients with these mutations have a longer survival than patients with wild-type IDH status. In the early 2000s, The Cancer Genome Atlas group defined classic, proneural, neural, and mesenchymal subtypes. Earlier studies showed 3 dominant subtypes (mesenchymal, proneural, and proliferative. The genetic and expression variation between these groups, along with common clinical manifestations between these groups, are discussed later because they are beyond the scope of this chapter, but their establishment is an example of the use of advanced computing and genomic sequencing to better understand a disease first recognized almost a century ago.

Clinicians continue to improve on their ability to diagnose glioblastoma. Developing methods to detect circulating tumor cells or circulating elements of a glioblastoma (eg, vesicles or other cell membrane components) in the circulation may improve the ability to screen for this condition or monitor for recurrence. Advancements made in molecular imaging will, in the future, allow clinicians to more precisely visualize neoplastic cells or tissues radiologically and in the operating room.

Historical perspectives on the management of glioblastoma

Amid changes in diagnosis, the nonsurgical and surgical management of glioblastoma has improved with technology. In the 1970s and 1980s, Judah Folkman and others helped establish the angiogenesis concept in tumors. Antiangiogenesis-based therapies such as bevacizumab (a vascular endothelial growth factor–neutralizing antibody) remain in the armamentarium with specific indications (recurrence) at the time of this writing. Other targeted pathway inhibitors (eg, mTOR (mechanistic target of rapamycin), PDGF ) have been developed and are under investigation in the modern era. As mentioned previously, mutations of EGFR were identified in a subset of patients with glioblastoma. In the past 5 to 10 years, targeted therapeutic agents have come to light; genetic factors, such as coexpression of EGFRvIII, phosphatase and tensin homolog ( PTEN ), and others may alter host susceptibility to therapeutic agents. In the early 2000s, the methylguanine methyltransferase ( MGMT ) excision repair enzyme was linked to tumor resistance to alkylating agents; thus methylation/inactivation was shown to predict outcome/survival in patients with glioblastoma receiving temozolomide chemotherapy. Genetic factors are no longer solely prognostic; they now guide therapeutic strategies. Although the concepts of immunotherapies directed at cancer are not novel, human glioblastoma research using immunotherapy is in its infancy and shows promise. Challenges in terms of the antigens used, humoral versus cytotoxic mechanisms, and delivery with respect to the BBB and tumor remain.

Neurosurgical treatment has changed with improvements in operative technology and visualization, especially with the introduction of the microscope and loupe magnification in the operating theater. M. Gazi Yaşargil is an important figure in neurosurgery because of his development of novel techniques and tools, his relentless devotion to surgical nuance, and his teaching ( Fig. 1.7 ). Influenced by Hugo Krayenbühl and R.M. Peardon Donaghy, his work advanced the practice of neurosurgery not only in patients with cerebrovascular disorders but also in those with brain tumors.