OBJECT: Absorbable topical hemostatic agents are commonly used in neurosurgery. In this study the authors examine the longitudinal relaxation time (T1) of blood in contact with these agents over time, measured in vitro, to determine if their presence could affect the interpretation of postoperative magnetic resonance (MR) images. METHODS: Coagulated and anticoagulated blood were used, both oxygenated and deoxygenated. The effects of a collagen-based agent (Collastat) and a cellulose-based agent (Surgicel) on the pH and T1 values of blood and on those of saline (used as a control) were investigated. The T1 was measured as a function of magnetic field strength and time by using a field-cycling relaxometer. This instrument measures 1/T1, the rate of T1, from which the T, value is computed. The T1 values of blood were compared with those of hemostat-induced blood clots and with those of both gray and white matter of the brain. Signal changes on T1-weighted MR images were predicted on the basis of altered T, values in vitro. Postoperative images were visually examined for the predicted changes. With the addition of Surgicel, blood had decreased pH and significantly shortened T1 at all fields, essentially within minutes, although it affected the T1 of saline only minimally. The effect of Surgicel increasingly shortened the T1 for 4 days in oxygenated blood. Collastat had no significant effect. The presence of some paramagnetic methemoglobin in Surgicel-induced clots was demonstrated using the relaxometer at a time when diamagnetic oxyhemoglobin would be present in naturally occurring blood clots. A bright signal that could mimic residual tumor on contrast-enhanced images was predicted and confirmed on postoperative T1-weighted MR images obtained in patients in whom Surgicel lined the tumor bed. It was not present in cases in which Surgicel was not used. CONCLUSIONS: Surgicel alters the appearance of early postoperative MR images. To avoid misinterpretation, clinicians should be aware of this phenomenon.
OBJECT: Absorbable topical hemostatic agents are commonly used in neurosurgery. In this study the authors examine the longitudinal relaxation time (T1) of blood in contact with these agents over time, measured in vitro, to determine if their presence could affect the interpretation of postoperative magnetic resonance (MR) images. METHODS: Coagulated and anticoagulated blood were used, both oxygenated and deoxygenated. The effects of a collagen-based agent (Collastat) and a cellulose-based agent (Surgicel) on the pH and T1 values of blood and on those of saline (used as a control) were investigated. The T1 was measured as a function of magnetic field strength and time by using a field-cycling relaxometer. This instrument measures 1/T1, the rate of T1, from which the T, value is computed. The T1 values of blood were compared with those of hemostat-induced blood clots and with those of both gray and white matter of the brain. Signal changes on T1-weighted MR images were predicted on the basis of altered T, values in vitro. Postoperative images were visually examined for the predicted changes. With the addition of Surgicel, blood had decreased pH and significantly shortened T1 at all fields, essentially within minutes, although it affected the T1 of saline only minimally. The effect of Surgicel increasingly shortened the T1 for 4 days in oxygenated blood. Collastat had no significant effect. The presence of some paramagnetic methemoglobin in Surgicel-induced clots was demonstrated using the relaxometer at a time when diamagnetic oxyhemoglobin would be present in naturally occurring blood clots. A bright signal that could mimic residual tumor on contrast-enhanced images was predicted and confirmed on postoperative T1-weighted MR images obtained in patients in whom Surgicel lined the tumor bed. It was not present in cases in which Surgicel was not used. CONCLUSIONS: Surgicel alters the appearance of early postoperative MR images. To avoid misinterpretation, clinicians should be aware of this phenomenon.