Conventional Imaging of DIPG

It is now generally accepted that there is a distinct subtype of brainstem tumor which can be identified on imaging as a diffuse intrinsic pontine glioma (DIPG). These invariably correspond to fibrillary astrocytomas on pathology and have an extremely poor prognosis, the worst of any type of brainstem tumor. They demonstrate many of the characteristics that have been outlined in the History of Imaging. Those characteristics include:

  • They generally arise from the pons, more specifically the ventral pons.
  • They are diffuse tumors, infiltrating greater than half the transverse diameter of the brainstem segment of origin and having indistinct margins.
  • They also tend to expand the segment of the brainstem from which they arise.
  • This expansion coupled with the common location in the ventral pons results in the tendency to engulf the basilar artery.
  • These masses are generally iso- to hypo-dense (equal to or dark) to normal adjacent brain on CT, hypointense (dark) to normal adjacent brain on T1 weighted MRI, and hyperintense (bright) to normal adjacent brain on T2 weighted MRI.
  • These imaging characteristics generally reflect the increased water content of these tumors.
  • The presence of enhancement after the administration of contrast, cysts or necrosis, and hemorrhage is variable in these tumors. Necrosis is generally described as areas of relative increased T2 signal and decreased T1 signal within the tumor with peripheral or “ring” enhancement.

These typical findings of DIPG, considered to represent fibrillary astrocytomas, are in contrast to focal tumors of the brainstem, which usually represent tumors of different histopathologies.

Several studies focusing on DIPG have attempted to determine if any conventional MRI characteristics at the time of tumor presentation are helpful in predicting prognosis. No findings on conventional MRI at tumor presentation have been found to be correlated to response to therapy, progression, or overall survival. Imaging characteristics that have been analyzed include: volume and extent of tumor, signal intensity on T1 and T2 images, appearance of borders, peritumoral edema, exophytic components, encasement of the basilar artery, necrosis or cysts, hemorrhage, gadolinium enhancement, and metastatic disease.

It is now thought that regardless of the histopathology of the DIPG at the time of diagnosis, even if the tumor is a low grade WHO II astrocytoma at diagnosis, most diffuse intrinsic pontine gliomas will progress to WHO III anaplastic astrocytoma or WHO IV glioblastoma multiforme at the time of death. Although a small percentage of DIPG will demonstrate areas of necrosis at presentation, almost all will develop areas of necrosis during the course of the disease. There is debate regarding whether the development of necrosis indicates treatment related changes (i.e. radiation necrosis), or reflects the natural malignant degeneration of low grade tumors (WHO II) to high grade tumors (WHO III and IV). Although positron emission tomography (PET) has been used to make this distinction with other brain tumors (see section below on advanced imaging), there has been difficulty in applying this to DIPG. Nevertheless, many consider the development of new areas of enhancement or necrosis a grave prognostic indicator that often precedes death.

In addition to developing areas of necrosis, most tumors will continue to infiltrate more portions of the brainstem during the course of the disease. They preferentially grow in a cranial direction through the midbrain into the cerebral peduncles and thalamus, and it is suggested that this finding may also be a poor prognostic indicator.

Despite the fact that conventional MRI has not been helpful in the development of prognostic criteria for DIPG at initial presentation, there continues to be a small subset of patients that have an unexpectedly long survival. It is uncertain if the lesions in these patients represent a more benign tumor type other than a fibrillary astrocytoma or even lesions other than tumors, such as inflammatory lesions that mimic DIPG on imaging. Furthermore, perhaps these tumors are less biologically active or are more sensitive to treatment. The advent of new advanced imaging techniques including magnetic resonance spectroscopy and magnetic resonance perfusion imaging may soon improve our understanding of the biology and natural history of DIPG and lesions that mimic them on conventional MRI.