Introduction
Due to the wide array of available accepted therapies, as well as the increasing number of experimental treatments undergoing clinical trials for the management of glial neoplasms, the task of the radiologist to make appropriate interpretations can seem daunting. In this chapter, we review the clinical, pathologic, and imaging findings associated with these therapies and offer recommendations on the imaging approach in these scenarios.
Surgical and Therapeutic Options: Overview
The ideal surgical treatment for glial tumors is maximal total resection with minimal neurologic side effects. However, this objective is achieved only in a relative minority of cases, due to the deep location of some tumors or their close proximity to eloquent cortical areas and other vital structures. Chemoradiation is a well-established step in the treatment of high-grade tumors, with partial brain fractionated radiotherapy combined with temozolomide (TMZ) being the accepted standard treatment after maximal cytoreductive surgery. Antiangiogenic treatment, particularly bevacizumab, is usually reserved for tumor recurrence after chemoradiation. Newer therapies include immunotherapy and checkpoint inhibitors, but their effectiveness has not been consistently proven in clinical trials.
Intraoperative Imaging
Leakage of contrast along the surgical margins can confuse the radiologic assessment of residual tumor, particularly after a repeat dose of gadolinium contrast agents or when there is significant delay after contrast administration ( Fig. 15.1 ). Hyperacute hemorrhage can be difficult to identify due to similarities in its signal characteristics to cerebrospinal fluid (CSF) ( Fig. 15.2 ). Therefore a high index of suspicion for hyperacute hemorrhage should be present when large unexplained areas of fluid signal are seen within the surgical cavity or surrounding brain parenchyma.
Postsurgical Changes
Immediate Postoperative Period: Pearls and Pitfalls
The aims of imaging in the postoperative period include detection of residual tumor and assessment of postoperative complications. One potential imaging pitfall is the presence of hemorrhage. In the setting of subacute hemorrhage shortly after surgery, the presence of intracellular methemoglobin demonstrates T1 shortening which may be mistaken for enhancement on gadolinium-enhanced T1-weighted images. The acquisition of a precontrast T1-weighted sequence is helpful for clarifying this because residual enhancing tumor would show relative T1 shortening between the precontrast and postcontrast images ( Fig. 15.3 ).
A variety of nontumoral processes can cause apparent enhancement, including postsurgical changes. For example, postoperative ischemic changes presenting with diffusion restriction along the margins of the surgical site may have associated enhancement in the subacute period. As such, it is generally suggested that contrast-enhanced magnetic resonance imaging (MRI) should be performed within 48 hours after surgery.
One potential pitfall is patients who have received electrocoagulation at the surgical site. Solid parenchymal enhancement that can be indistinguishable from residual tumor may appear even on intraoperative imaging.
Laser-Induced Thermotherapy
Laser-induced thermotherapy (LITT) allows for both biopsy and laser-guided ablation of lesions. A laser probe is placed into the core of the lesion and induces thermal coagulation, creating a region of coagulative necrosis. The typical early postsurgical change after this procedure is characterized by five concentric areas, including a peripheral rim of abnormal enhancement and perilesional edema ( Fig. 15.4 ).
Local Chemotherapy or Brachytherapy
Local delivery of chemotherapeutic agents and radiation therapy (RT) are also options in the treatment of glial tumors. Examples of these therapeutic options include Gliasite for localized RT ( 125 I) into the surgical site and Gliadel wafers for intracavitary chemotherapy. A significant proportion of cases treated with Gliadel wafers can experience transient worsening of imaging findings during the first 4 weeks after placement ( Figs. 15.5 and 15.6 ).
Foreign Body Reactions and Textilomas
Packing or hemostatic material (e.g., Gelfoam, Surgicel) left inside a surgical cavity during the resection of a glial tumor can trigger a foreign body reaction in some patients. The resulting pseudotumoral lesion is often called gossypiboma (when triggered by cotton material), textiloma, and gauzoma, among other terms. Importantly, this reaction can exhibit imaging features that overlap with tumor recurrence, such as contrast enhancement. However, in contrast to high-grade gliomas, follow-up imaging should demonstrate relatively stable mass effect and adjacent T2 prolongation except in the setting of complications (i.e., secondary infections) ( Fig. 15.7 ).
Radiation-Related Abnormalities
Early Postradiation Changes (Pseudoprogression)
Definition and Epidemiology
Pseudoprogression is the apparent worsening of abnormal enhancement, T2-hyperintense areas, or local mass effect that is often seen in patients with glial tumors after radiation treatment. This phenomenon usually occurs weeks to 3 months after treatment, whereas radiation necrosis often occurs 18 to 24 months to years afterward ( Table 15.1 , Fig. 15.8 ).
| Time Course | Imaging Findings | Outcomes | |
|---|---|---|---|
| True progression | Anytime |
|
|
| Pseudoprogression (radiation therapy) | ≤3 months |
|
|
| Radiation necrosis | ≥6 months |
|
|
| Pseudoresponse (antiangiogenic therapy, i.e., bevacizumab) | Weeks to months |
| |
| Pseudoprogression (immunotherapy, i.e., checkpoint inhibitors) | ≤6 months |
|
|
Although pseudoprogression can occur after isolated RT, it has also been described after combined RT and TMZ. There is also an association with methylguanine methyltransferase (MGMT) gene promoter methylation. Although the incidence is thought to be between 20% and 25% among all patients, those with MGMT gene promoter methylation have a significantly higher incidence, at approximately 35% to 40%. Along these lines, MGMT methylation predicts pseudoprogression in approximately 90% of cases.
An important clinical implication is that the presence of radiologic pseudoprogression is associated with improved 1- and 2-year progression-free survival. Conversely, patients with MGMT-unmethylated tumors have relatively higher rates of progressive disease.
Pathophysiology
The pathophysiology of pseudoprogression is not well understood, which is partly due to the complex changes that result from radiation and chemotherapy. It is hypothesized that vascular injury and treatment-related cellular effects are the primary contributors. RT is known to disrupt the blood-brain barrier (BBB), which leads to contrast enhancement and vasogenic edema. Along with tumoricidal effects, inflammatory changes have been reported on histopathology.
There is debate as to whether pseudoprogression represents a similar process as radiation necrosis. One distinction is that pseudoprogression primarily reflects treatment changes in tumor, whereas radiation necrosis can affect residual/recurrent tumor and adjacent normal parenchyma.
As noted, MGMT promoter methylation is strongly linked to pseudoprogression. The exact cause is not clear but may relate to increased inflammation and BBB disruption, potentiating therapeutic effects. In particular, tumor cells that show decreased expression of MGMT have apparent increased sensitivity to TMZ. As a result, patients with MGMT methylation particularly benefit from TMZ adjuvant therapy.
Imaging Findings
On conventional MRI, patients with pseudoprogression show increasing contrast-enhancing lesion size and surrounding T2-hyperintense signal abnormality, typically within 3 months after completing RT. This is followed by improvement or stabilization of the imaging findings (decreased mass effect, T2/fluid-attenuated inversion recovery [FLAIR] signal abnormality, and abnormal enhancement) that can be observed within weeks to months ( Fig. 15.9 ).
The challenge lies in the fact that it is difficult to prospectively determine if abnormal enhancement that develops shortly on the initial post exam represents pseudoprogression or rapid true progression. Although certain imaging patterns such as the presence of subependymal enhancement favor progressive disease, they are not sufficiently specific to distinguish between pseudoprogression and true tumor progression. In 2010 the Response Assessment in Neuro-Oncology (RANO) working group proposed that within the first 12 weeks after completion of RT, true disease progression can be determined only if it occurs outside of the radiation field or if there is histopathologic confirmation. However, it is important to note that pseudoprogression can occur outside of this 12-week window. As such, barring a biopsy, follow-up imaging and clinical assessments remain the current standard for distinguishing pseudoprogression from true progression.
In light of these limitations, there has been interest in the use of advanced imaging methods. Diffusion-weighted imaging (DWI) is one method that is now widely available. It has been reported that pseudoprogression tends to have higher ADC values and ADC ratios (the ratio of ADC values from enhancing tissue to normal appearing white matter) compared with tumor recurrence. Unfortunately, the interpretation can be confounded by superimposed areas of necrosis and vascularity.
MR perfusion imaging with either dynamic contrast enhancement (DCE) or dynamic susceptibility contrast (DSC) techniques has also shown promise for distinguishing pseudoprogression from true progression. Tumor progression has increased relative cerebral blood volume (rCBV) compared with normal parenchyma, whereas pseudoprogression generally does not show significantly increased rCBV.
MR spectroscopy (MRS) is another advanced MR method that can be used alternatively or in conjunction with the previous techniques. Increased choline (Cho)/creatinine (Cr) and Cho/N-acetylaspartate (NAA) ratios have been consistently reported in recurrent tumors compared with RT changes. From a technical standpoint, multivoxel acquisitions are more helpful in capturing the presence of regional differences in the proportion of posttreatment changes and progressive tumor.
Delayed Postradiation Changes
Definition and Epidemiology
Delayed changes from RT, such as radiation necrosis, occur between 3 months to years after therapy. The incidence of radiation necrosis reportedly ranges from 3% to 24%. A direct relationship between the incidence of these changes with the total radiation dose, treatment duration, and irradiated tissue volume has also been described.
Pathophysiology
Delayed radiation-related injury is thought to occur through a combination of demyelination and vascular changes. Within the white matter, it is thought that oligodendrocytes are primarily targeted. In addition, endothelial cells lining vessel walls are injured, which leads to BBB breakdown and the subsequent enhancement seen on imaging. The vascular injury also leads to reduced microvasculature that contributes to chronic ischemia and oxidative stress, which initiates the process of coagulative and liquefactive necrosis. In addition, histologic studies show calcification, vascular hyalinization, and fibrinoid deposition. Vascular endothelial growth factor (VEGF) expression is increased, which further contributes to small vessel permeability and subsequent edema.
Radiation Necrosis
As noted previously, radiation necrosis occurs months (typically 18–24 months) to years after RT. Clinically, patients may be asymptomatic or present with focal neurologic deficits. On MRI, it can present as an enhancing mass lesion with associated abnormal T2 prolongation and mass effect. This presents a diagnostic challenge because it can closely resemble recurrent brain tumor.
However, in certain instances, a distinction can be made with greater confidence based on conventional MRI. New areas of enhancement that develop in a previously nonenhancing tumor is often the result of RT rather than progression to higher-grade tumor ( Fig. 15.10 ). A “soap bubble” or “swiss cheese” pattern of enhancement, as well as involvement of the septum pellucidum, also suggests radiation necrosis, although recurrent neoplasm could have a similar imaging appearance.
