Acepromazine

Objective: The feasibility of high resolution CT imaging for evaluating experimentally-induced brain tumors in rats was assessed.

Methods: The gliosarcoma cell line (9 L/lacZ) was inoculated in 34 male Fischer 344 rats and CT studies were performed prior to and at 4, 7, 9, 12 and 14 days post-tumor cell implantation. Brain imaging pre- and post-contrast was performed and correlated with autopsy findings.

Results: Tumors were identified by CT in 19 of the 34 animals after contrast administration and their presence was confirmed at autopsy. Tumors were present at autopsy and not identified by CT scanning in eight additional animals and in the remaining seven animals, the CT scan was normal and no tumors were present at autopsy. The sensitivity and specificity of CT scanning with contrast in detecting tumors in this rat model of gliosarcoma was 70 and 100%, respectively.

Conclusion: The improved CT technology currently available can be used to identify and follow tumor burden in a rat model of gliosarcoma, and be a good tool to utilize in determining treatment outcomes experimentally, especially when MR imaging is not available. [Neurol Res 2005; 27: 57-59]

Keywords: Computed tomography; 9L glioma

INTRODUCTION

Despite the improved spatial resolution brought forth by new generation computed tomography (CT) units, the popularity of CT to identify brain tumors in small animals (mice, rats) has declined. This is due largely to the high resolution anatomic detail provided by magnetic resonance imaging (MRI), which is able to detect brain tumors in rats, permitting investigators to accurately evaluate the efficacy of therapy on tumor growth, displacing the CT as the pre-eminent imaging modality in experimental neuro-oncology1'2.

In the past few decades, further refinement in CT technology has rekindled interest in this technique for small animal research, but few reports3'4 validating the usefulness of new generation scanners to detect brain tumors exist. This study was therefore undertaken to evaluate the feasibility of using a new generation, high resolution CT scanner to identify brain tumors in rats, the animal used most commonly in brain tumor research.

MATERIALS AND METHODS

All experiments were performed with the approval of the Institutional Animal Care and Use Committee of the University of Texas Southwestern Medical Center. Male Fischer 344 rats weighing 450-550 g (age 60-75 days) were anesthetized with a mixture of ketamine, acepromazine, xylazine and sterile water administered by intraperitoneal injection.

A 10 µl suspension of 10^sup 6^ 9 L/lacZ gliosarcoma cells (CRL-2200, ATCC, Manassas, VA) were injected, with the aid of a guide screw, into the right caudate nucleus of 34 rats utilizing previously described methods5 (Figure 7) CT studies were performed before, and on 4, 7, 9, 12 and 14 days post-tumor cell implantation in all rats, which were placed on their sides to dramatically reduce screw artifact and allow better visualization of intracranial contents. The brains were imaged prior to and 30 minutes after intraperitoneal administration of 10 ml lohexol contrast media.

Axial slices of 1 mm thickness were obtained using a multi-slice detector system scanner (Toshiba America Medical Systems, Tustin, CA). Scans were reconstructed over a 10-cm diameter reconstruction circle with a 512 × 512 pixel matrix. The following parameters produced the best CT images: scan time 2 seconds, 100 mA, 100 KVP, head mode, 1 mm slice thickness, zoom factor 4.5 and 10 ml intraperitoneal contrast administration followed by scanning 30 minutes postcontrast administration. For interpretation purposes, the scan was expanded by the computer approximately five-fold.

The animals were killed after scanning (day 14) by decapitation under deep anesthesia, and the brain removed and immersed in 10% formalin. Axial sections corresponding as closely as possible to the plane of the scan were made, stained with hematoxyln and eosin (H&E), reviewed by light microscopy and findings correlated with CT results.

RESULTS

The average weight of the brains was 2.61 g (range 2.36-2.97 g). Each brain measured 1.8-2.5 cm in length (antero-posterior dimension) and 0.9-1.4 cm in width. The intracranial cavity and brain were easily identified on CT scans of every rat. In one case, a marked hypodense area not suggestive of a brain tumor encompassed the entire right hemisphere, which cut sections revealed was agenesis of that hemisphere. The ventricles were not visualized on any scans, and the grey and white matter junction was not discernable.

Of the 34 rats injected with tumor, 19 were identified by CT scan after contrast administration (day 14), which was confirmed at necropsy; four identified prior to contrast administration (day 14); eight were present at autopsy and were not identified by CT; seven were absent on CT and no tumor was present at autopsy. Of the four identified prior to contrast administration (day 14), a circumscribed area slightly more hyperdense than the surrounding brain was identified. Moreover, two of these had punctate areas of hyperdensity, which related to either calcification or intratumoral bleeding (Figure 2a) as confirmed by autopsy. Tumor necrosis was evident in eight tumors at autopsy not identified by CT scanning (Figure 2b) and, in most cases, the brain tumor interface was easily discernible at autopsy, but not on CT scan (Fieure 2c).

The tumors visualized were large, generally not well demarcated and had heterogeneous contrast enhancement (Figure 3). In five animals, the tumors were suspected due to asymmetry in enhancement between the two hemispheres, and/or difference in density between tumor and surrounding tissue. Homogenous enhancement was not present on any scan.

Tumors could not be detected on day 4, 7 or 9 days in any animal after contrast administration, but three were detected on day 12 and again on day 14, and confirmed at autopsy. CT tumor dimensions were 1.2 mm less than the fixed, sectioned autopsy specimens. Average tumor size on CT scan was 7.3 mm in diameter (range 4.2-11.5 mm) and average size on stained sections was 6.1 mm in diameter (range 3.4-10.9 mm). Review of the eight lesions seen at necropsy and not identified by CT scanning revealed tumors smaller than 4 mm in diameter were more likely to be missed.

DISCUSSION

Groothius et al6, previously described the appearance of experimentally induced brain tumors by CT in dogs. Subsequently, Reith ef al7, accurately localized brain tumors in monkeys by contrast enhanced scans. Kumar ef al.8 showed that CT is useful for documenting the efficacy of chemotherapy in a rabbit model and Kabuto ef al.9 demonstrated that a C6 glioma possessed similar CT characteristics as human gliomas in an adult mongrel cat. Others10 have also demonstrated the feasibility of scanning larger animals, but only few reports exist, characterizing the usefulness of this technique in rats3,4, the animal used most commonly in experimental neuro-oncology.

Casanueva ef al.3 demonstrated the usefulness of CT to identify pituitary tumors in rats injected with estrodiol valerate, a hormone known to lead to the development of pituitary tumors, if given in high doses. All pituitary tumors weighing more than 54 mg were detected with the exception of one and the authors concluded that CT scanning was helpful in identifying experimentally-induced pituitary tumors in rats.

Kapp and Holla4 utilized an Ohio Nuclear 2010 scanner to detect tumors in newborn rats inoculated with Rous sarcoma virus. Five tumors were demonstrated by CT scans and their presence confirmed at autopsy, four tumors were present at autopsy, but not diagnosed by scanning and in nine animals, the CT scan was considered normal and no tumor was present at autopsy. Although not specified in the article, the sensitivity and specificity of CT scanning for detecting brain tumors was 56 and 100%, respectively. The high incidence of false negative scans in this article precluded the exclusion of a tumor on the basis of scanning alone and compares unfavorably to reports utilizing MR imaging to identify tumors in small animals1,2.

The likely reason that the percentage of false-negative scans was reduced in our report, as compared with others4, was probably related to improved CT technology and the various factors used to improve image quality. Specifically, we attempted to operate under conditions that optimized spatial resolution and signalto-noise ratio. Specifically, the highest spatial resolution recommended (512 × 512 matrix) was obtained by reducing pixel size and improving our zoom factor. Additionally, to minimize partial volume averaging, the slice thickness was reduced to 1 mm and, to reduce noise, we optimized the scanner output. The use of higher doses of contrast media11, thinner CT slices12 and optimal post-contrast scan timing13 may have resulted in more reliable tumor enhancement and detection of less obvious lesions. Sensitivity and specificity was 70 and 100%, respectively, a modest improvement from prior reports4.

Despite the refinement in CT technology over the past few decades, this technique still plays a secondary role in detecting and evaluating brain tumors in rats. The superior anatomic detail and physiological information yielded during a single examination by MRI makes this technique more promising than CT scanning for small animal research1,2.

A primary role of CT scanning in small animals with brain tumors is to assist in treatment planning. With its superior definition of bone and its acceptable demonstration of tumors, CT has the ability to demonstrate the relationship of the underlying tumor to the bony landmarks of the skull, which are important in planning a craniotomy site or potential radiation ports in these small animals for experimental purposes. We are currently investigating the feasibility of using CT to perform image- guided radiosurgery in rats utilizing the Cyberknife radiosurgical unit, a frameless, robotic radiosurgical unit pioneered by Adler14 for the clinical setting.

CONCLUSION

Contrast-enhanced CT scanning can be used in experimental neuro-oncology with a high specificity and a higher sensitivity than previous reports. The improved CT technology currently available could be used not only to identify, but also follow tumor burden in a rat model of gliosarcoma and be a good tool to utilize in determining treatment outcomes experimentally, especially when MRI is not available.

ACKNOWLEDGEMENTS

The authors would like to thank Theodore Strauss and the Annette Strauss Center for Neuro-oncology, UTSW, for their support.

REFERENCES

1 Jacobs AH, Winkler A, Dittmar C, et al. Molecular and functional imaging technology for the development of efficient treatment strategies for gliomas. Technol Cancer Res Treat 2002; 1: 187-204

2 Rajan SS, Rosa L, Francisco J, Muraki A, Carvlin M, Tuturea E. MRI characterization of 9L-glioma in rat brain at 4.7 Tesla. Magnet Reson lmag 1990; 8: 185-190

3 Casanueva FF, Gordon WL, Friesen HC. Computed cranial tomography in the evaluation of pituitary tumours in rats. Acta Endocrinol 1983; 103: 487-491

4 Kapp JP, Holla PS. Detection of experimental brain tumor in rat by computed tomography. Surg Neurol 1981; 16: 455-458

5 LaI S, Lacroix M, Tofilon P, Fuller G, Sawaya R, Lang FL. An implantable guide-screw system for brain tumor studies in small animals. J Neurosurg 2000; 92: 326-333

6 Groothius DR, Mikhael MA, Fischer JM, ef al. Computed tomography of virally induced canine brain tumors: a preliminary report. J Comput Assist Tomogr 1981; 5: 538-543

7 Reith KG, Chiro GD, London WT, ef at. Experimental glioma in primates: a computed tomography model. J Comput Assist Tomogr 1980; 4: 285-290

8 Kumar AJ, Hassenbusch S, Rosenbaum AE, ef al. Sequential computed tomographic imaging of a transplantable rabbit brain tumor. Neuroradiol 1986; 28: 81-86

9 Kabuto M, Hayashi M, Nakagawa T, ef al. Experimental brain tumor in adult mongrel cat. No To Shinkei 1990; 42: 339-343

10 Morgan DF, Silberman AW, Bubbers JE, Rand RW, Storm FK, Morton DL. An experimental brain tumor model in rabbits. J Surg Oncol 1982; 20: 218-220

11 Brekenfeld C, Foert E, Hundt W, Kenn W, Lodeann KP, Gehl HB. Enhancement of cerebral diseases: how much contrast agent is enough? Comparison of 0.1, 0.2, and 0.3 mm kg^sup -1^ gadoteridol a 0.2 T with 0.1 mmol kg^sup -1^ gadoteridol at 1.5 T. Invest Radiol 2001; 36: 266-275

12 Sighvatsson V, Ericson K, Tommasson H. Optimizing contrast-enhanced cranial CT for detection of brain metastases. Acta Radiol 1998; 39: 718-722

13 Mathews VP, Caldemeyer KS, Ulmer LJ, Nguyen H, Yuh WT. Effects of contrast dose, delayed imaging, and magnetization transfer saturation on gadolinium-enhanced MR imaging of brain lesions. J Magnet Reson lmag 1997; 7: 14-22

14 Adler JR, Murphy MJ, Chang SD, Hancock SL. Image-guided robotic radiosurgery. Neurosurgery 1999; 44: 1299-1307

T. G. Psarros, B. Mickey and C. Ciller

Department of Neumsurgery, University of Texas Southwestern Medical Center, Dallas, Texas, USA

Correspondence and reprint requests to: Thomas G. Psarros, University of Texas Southwestern Medical Center, 5323 Harry Mines Blvd, Dallas, TX 75390-8855, USA. [tompsarros@aol.com] Accepted for publication May 2004.

Copyright Maney Publishing Jan 2005
Provided by ProQuest Information and Learning Company. All rights Reserved