Apparent Diffusion Coefficient and Hematoma: “Dose-Dependent” Relationship between Hemorrhage and Edema
===========================================================================================================

* Robert D. Zimmerman

In this issue of *AJNR*, Carhuapoma et al report on the relationship between the size of acute intracerebral hematomas and the severity of vasogenic edema in adjacent brain. The authors measured changes in the apparent diffusion coefficient (ADC) in brain tissue adjacent to hematomas and found a correlation between ADC and hematoma size that indicates a “dose-dependent” relationship between hemorrhage and edema. They also discovered that ADC increased in contralateral homologous brain.

Why is this important? It is old news to neuroradiologists that peripheral edema develops around acute hematomas and that the extent of edema correlates with hematoma size. One simple answer is that while we “know” from clinical experience that this relationship exists, we have not collected data that quantify this relationship. Radiologic studies have been criticized, often unjustly, for their lack of quantitative rigor; therefore, it is edifying when intuitions based on clinical experience are verified by more rigorous scientific analysis. Carhuapoma et al, however, surpass simple verification of a long-standing clinical observation. These investigators are asking a new question by quantifying severity rather than extent of edema. The authors did not measure the volume of edematous brain but rather the change in ADC within the edematous brain (as defined by hyperintensity on T2-weighted images). Although the measurement of volume of edematous brain can be performed with either CT or MR imaging, measurement of severity of edema within affected brain can only be accomplished with MR imaging. The same is true for the notable (that is to say unexpected) finding in this study; namely, the presence of increased ADC in homologous brain tissue contralateral to the intracerebral hematoma. The authors believe that this phenomenon represents a reaction of the brain to this distant insult.

Therefore, diffusion-weighted MR imaging provides information concerning hematomas that is unavailable with CT. Does that mean that all acute hemorrhages should be imaged with MR? Most radiologists and clinicians are justifiably reluctant to adopt this strategy. Improvements in availability of MR systems, study time, and patient monitoring have been matched by improvements in CT scanners. Despite enormous advances in MR technology, it is still faster and easier to perform CT scanning in a critically ill patient, and CT scans are easier to interpret than are MR images. Nonetheless, I must confess to a fondness for and fascination with the MR imaging features of intracranial hemorrhage. I have always thought that the complexity of MR findings in hemorrhage reflect, more accurately than CT, the complex events occurring in or around hematomas. But is the increased effort required to perform and interpret MR studies worth it? Carhuapoma and colleagues suggest that it might be. The authors encourage us to think of blood as a neurotoxin with a dose-dependent deleterious effect on brain, an effect that might be prevented or at least ameliorated by medical or surgical intervention. This will require a cognitive shift for clinicians and neuroradiologists.

We tend to think of hemorrhage as an undesirable end result rather than as a dynamic process whose course can be changed. Hemorrhage is what we don’t want to see on an image obtained in someone with an acute neurologic event, because this finding exposes the patient to the possibly disastrous complication of medical or surgical therapy. Currently, limited treatment options are available, such as stereotactic hematoma removal, but we do not know their efficacy. Medical treatments have yet to be devised and tested. Carhuapoma and colleagues point to animal data indicating that thrombin is the major culprit in the inflammatory reaction that produces perihematoma edema. Medical treatment aimed at reducing thrombin’s effects is just one of many possible treatment options. An understanding of the biochemical and physiologic changes associated with hemorrhage will be required if new effective therapies are to be devised. The ability of MR imaging to display the complex processes occurring in and around hematomas offers us a unique opportunity to study the effects of hemorrhage in vivo.

As the authors state, quantitative data from MR images will probably turn out to be critical to research of the efficacy of therapeutic options and clinical decision making. Treatment of acute vascular neurologic events will always be stuck between the rock of infarction and the hard place of hemorrhage. Procedures that limit infarction increase the risk of hemorrhage and vice versa. Qualitative assessment of images will not be accurate enough to make decisions about whom, when, and how to treat acute vascular events. We will need hard data, and MR imaging is the technique most likely to provide these data. Assessment of a variety of parameters including relaxation times, susceptibility effects, perfusion, ADC, tissue anisotropy, magnetization transfer, and brain metabolites can be obtained with MR imaging. It is unclear which of these parameters singly or in combination will prove most effective, but if we are to make headway in the treatment of this common cause of stroke, we must begin to study hemorrhage in a systematic way.

Carhuapoma et al have presented us with data from a small number of patients with acute hematomas. They have assessed a single variable, ADC, at a single and somewhat variable time point. As with any good science, the study raises more questions than it answers. What is the effect of ADC change on prognosis? Which is more important for outcome: the volume of the edematous brain or the severity of edema within that brain? How does edema progress during the initial hours and days of ictus? Are treatment options time-sensitive (as they are in acute infarction)? Does the cause of hemorrhage determine the extent of edema? Hypertension and cerebral amyloid angiopathy produce focal hematomas that displace normal brain, whereas hemorrhagic infarction and contusions produce hemorrhage into and extensive destruction of brain tissue. Does this alter the time course and severity of edema? Could it affect treatment options?

The most fascinating finding in this study is the increase in ADC in the brain contralateral to the hematoma. Hematomas are quintessentially focal lesions, yet we are now told that they have distant effects. The authors speculate that this represents a remote response to neurotoxic effects of blood products. This is probably true, but we don’t how this response is mediated and what effect this action at a distance has on brain function. The data in this study do point to another advantage that MR imaging offers us in the assessment of pathologic processes. We can measure global and regional changes with parameters such as ADC with relative ease. In this study, the authors compared the ADC in the contralateral homologous brain. Were the effects limited to this region or were they global? The answer to this question might assist us in determining the cause of this phenomenon and its functional consequences.

This study is important, because it points us in a new direction. We now can think of hemorrhage in a new way. This will inevitably lead to new challenges. I am certain that many of the questions I have posed will turn out to be unimportant, but I am equally certain that important questions will persist, and the answers will aid in developing effective therapies for intracranial hemorrhage. MR imaging, with its high intrinsic information content, will be our tool for investigation, and eventually, for clinical decision making.

*   Copyright © American Society of Neuroradiology