Letters MR
U
to the
Imaging
of Firearm
Editor
Projectiles
From: Robert
B. Lufkin,
MD
Department
of Radiological Sciences University of California, Los Angeles School of Medicine Los Angeles, CA 90024 Editor: Our right to bear arms combined with the increasing violence in our society means that an increasing number of individuals undergoing imaging studies will harbor firearm projectiles within their bodies. Several excellent studies have addressed the safety and utility of imaging these patients with magnetic resonance (MR) imaging (1-3). There is consensus that when the composition of the projectile is known to be nonferromagnetic (by means of history or identification of the shell casing), patients can be safely imaged. Ferromagnetic bullets are uncommon and are usually jacketed and of foreign manufacture. The situation regarding the safety of imaging patients with retained shotgun pellets is less clear-cut, especially in the radiology literature. Common lead pellets create little artifact and are usually safe to study with MR imaging. However, steel pellets are ferromagnetic and pose a potential hazard to patients examined with MR imaging.
This
is extremely
important
information
for radiologists,
es-
pecially given the fact that recent federal legislation will ban the use of lead shot on all federal land by the year 1992 (4). This will force the use of steel shot to avoid the lead toxicity to the environment that has occurred with the lead shot. Awareness
of this
tial injuries
during
situation
by
radiologists
may
avoid
poten-
MR examination.
References 1.
2.
3.
Teitelbaum GP, Yee CA, Van Horn DD, Kim HS, Coletti PM. Metallic ballistic fragments: MR imaging safety and artifacts. Radiology 1990; 175:855-859. Zheutlin JD, Thompson JT, Stewart Shofner R. The safety of magnetic resonance imaging with the intraorbital metallic objects after retinal reattachment or trauma. Am J Opthalmol 1987; 103:831. Schiebler ML, Robertson L, Mauro MA, Koomen MA, Whaley RA. Torque of projectiles at 1.5T (abstr). Magn Reson Imag 1990; 8(suppl 1):120.
4.
Williams
T.
U
Treatment
Let them
eat steel.
of Variceal
Audubon
1988; 90:22-33.
Hemorrhage
of Colorado study were compared with ours, significant differences in technique exist. Their approach was limited, in most patients, to mechanical occlusion of the esophagogastnic varices which, except for the improved delivery system (emergency mmnilaparotomy), is similar to transhepatic portal embolization. Because of the long-term ineffectiveness of transhepatic portal embolization (3), we attempted to improve survival by expanding our procedure to create the radiologic equivalent of surgical devascularization. Great success has been reported for surgical devascularization procedures in Japan (4). Survival should be further improved by eliminating the stress of major surgery, since there is a high mortality rate when a patient with cirrhosis undergoes such operations. A devascularization
tnic,
Volume
179
Number
#{149}
1
left
gastroepiploic
arteries
technique and by sclerosing during minilaparotomy with
might
radiologic
embolization
by
using
the
standard
Se!-
the entire variceal netuse of absolute alcohol and
have
been
improved.
We agree with Durham et al that mmnilaparotomy for variceal sclerosis with or without arterial embolization represents a valuable treatment for variceal hemorrhage. How this technique should be modified to best control bleeding and improve survival still requires more investigation. We hope that more centers will now consider such embolization procedures for treatment of variceal hemorrhage.
MD,t
Durham
New
U.
surgical
hemorrhage.
York 2.
in the October it is to treat varto bleed despite patients has been of the University
JD, Kumpe
Rothbarth
bined
results of the study by Durham et al (1) issue of Radiology show just how difficult iceal hemorrhage in patients who continue sclerotherapy. Our experience with similar more encouraging (2). Although the results 1990
on
Disparity in our results and those described by Durham et a! may be partly due to differences in technique and the addition of arterial embolization. Emergency midsplenic artery embolization alone temporarily controlled bleeding in most patients, thus allowing time to prepare the patient for minilaparotomy. Although semiurgent performance of the minilaparotomy for variceal sclerosis was still necessary to prevent recurrent bleeding, the better-stabilized patient had a lower operative mortality and a decreased risk of hepatic failure. The combination of absolute alcohol and coil embolization improved the effectiveness of variceal occlusion over coils alone (5). Left gastric and left gastroepiploic artery embolization decreased the risks and severity of recurrent bleeding. By incorporating some of these techniques into their approach, the long-term survival rates of the study of Durham et a!
Editor: The
and
dinger work coils.
1.
R. M. DelGuercio, Surgery,
based
References
From: Howard L. Berman, MD,* Louis and Stuart C. Katz, MD* Departments of Radiology* and Medical College Valhalla, NY 10595
procedure
would hopefully increase patient survival. Surgical devascularization procedures include splenectomy, ligation of the arterial blood supply to the upper half of the stomach, and occlusion of the esophagogastric varices. Similar results are created with embolization of the splenic, left gas-
3.
4.
DA,
Direct
and
Van
Stiegmann
catheterization
radiologic
Radiology
JS, Subber
C, Goff of the
approach
to the
mesenteric
SW,
vein:
treatment
com-
of variceal
1990; 177:229-233.
Berman HL, DelGuercio LRM, Katz SC, Hodgson WJ, Savino JA. Minimally invasive devascularization for variceal bleeding that could not be controlled with sclerotherapy. Surgery 1988; 104: 500-506. Benner KG, Keeffe EB, Keller FS, Rosch J. Clinical outcome after percutaneous transhepatic obliteration of esophageal varices. Gastroenterology 1983; 85:146-153.
Sugiura in the 1321.
M, Shunji treatment
F.
Further
of esophageal
evaluation varices.
of the Sugiura Arch
Surg
1977;
Radiology
procedure 112:1317-
#{149} 285
5.
Yune HY, O’Connor KW, Klatte EC, Olson EW, Becker Strickler SA. Ethanol thrombotherapy of esophageal ther experience. AJR 1985; 144:1049-1053.
Drs Durham
and
Van Stiegmann
GJ, varices:
fur-
reply:
We thank Berman et al for their letter regarding our recent artide (1). They appropriately point out the striking difference in long-term survival-62% 1-year survival for patients with Child-Pugh class C cirrhosis in their series (2) compared with 27%
in
ours.
In our series, minilaparotomy for variceal embolization was performed only as a last resort after endoscopic therapy had failed. Such patients represent only 8% of all patients with variceal hemorrhage treated during the time of our study. Despite the similarity in the Child-Pugh classification in both studies, advanced hepatic decompensation and late intervention most likely explain differences in survival between the two trials. We wonder at what point devascularization was performed in the patients of Berman et al and what percentage of patients with variceal hemorrhage treated at New York Medical College this group represents. Could the improved survival be a result of earlier and more aggressive utilization of this procedure? We agree that the addition of left gastric, left gastroepiploic, and splenic arterial embolization to variceal embolization should be equivalent to a surgical devascularization procedure and, hence, result in better control of bleeding; however, additional embolization should have little effect on survival when death results from liver failure following successful control of bleeding. Only four of 1 1 deaths in the 15 patients in our series resulted from bleeding, two within 24 hours and two at 74 and 365 days. The majority of deaths (64%) occurred as a result of liver failure (five patients) or compounding disease (two patients). The ease with which minilaparotomy and variceal embolization can be performed, its relative safety compared with transhepatic procedures or surgery, and the possibility of combining variceal embolization with intrahepatic shunts should
encourage
gation
earlier
of this
intervention
and
continued
investi-
technique.
References 1.
Durham
JD, Kumpe
Rothbarth bined
U.
Direct
surgical
hemorrhage. 2.
DA,
and
Van
Stiegmann
catheterization
radiologic
Radiology
C, Goff
JS, Subber
of the mesenteric
approach
to the
treatment
SW,
vein:
1990; 177:229-233.
Berman HL, DelGuercio LRM, Katz 5G. Hodgson WJ, Savino Minimally invasive devascularization for variceal bleeding could not be controlled with sclerotherapy. Surgery 1988; 104:500-506.
Janette D. Durham, MD,* and Greg Van Stiegmann, Departments of Radiology* and Surgery,t University Colorado Health Sciences Center 4200 East Ninth Avenue, Denver, CO 80262
U Reducing Examinations
Doses
com-
of variceal
of Glucagon
Used
JA. that
MDt of
in Radiologic
From: Philippe
M. Jehenson,
MD,
BSc
Service
Hospitalier
Frederic
Joliot,
CEA
4, Place
du General
Leclerc,
91406
Orsay,
France
Editor: In their interesting commentary in the October 1990 issue of Radiology, Chernish and Maglinte (1) review the possible side effects of glucagon as used for radiologic examinations of the abdomen (including computed tomography or magnetic resonance imaging) because of its hypotonic effect on the gastrointestinal tract. They suggest that 1 mg of glucagon be administered intravenously in most cases, with an acceptable dose of
286
#{149} Radiology
up to 2 mg. They suggest that this dose range be used because side effects (at least the most common one, nausea) double as the dose doubles, whereas the duration of the required response (atonicity or hypotonicity) does not double in this range (it only increases by about 30%). One may probably take advantage of this fact (proportionality of the side effects, but saturation of the required response) by fractionating a smaller total dose into repeated injections (or perfusion). The desired effect should occur after each injection, if reasonably spaced, in the same way as it occurs after the first one. This is certainly the case for glucagon-induced hyperglycemia with injections at 10-20 minute intervals. It also seems reasonable for other effects, since insulin, the hormone “antagonist” to glucagon, reaches its maximum serum levels in 6-8 minutes and then rapidly decreases (1). Because nearly all subjects respond to 0.5 mg or even 0.25 mg of glucagon when intravenously injected (1), we would suggest (from data in Tables 2 and 3 of the commentary) starting by placing a small venous catheter and injecting 0.25 or 0.5 mg (through the catheter). Then inject 0.25 mg about 10 minutes later. One or two further injections may be used if needed (this could be decided during the examination itself). Three
such
injections
(a total
of 0.75
or
1 mg)
should
produce
atonicity or moderate which is longer than the dose (2 mg) given
hypotonicity for at least 30 minutes, that produced with more than double in a single injection. There also should
be
if any,
fewer
side
effects,
since
serum
levels
remain
much
closer to physiological situations (increasing the duration of a single injection from 1-2 seconds to 1 minute already strongly decreases at least some side effects [1]). Perfusion of 0.25 or 0.5 mg over 20-30 minutes after a first injection of the same amount should provide even better results. Reference 1. Chernish
SM, Maglinte
tions-review
and
DDT.
Glucagon:
recommendations.
common Radiology
untoward 1 990;
reac-
177:145-
146.
Drs Chernish We
thank
and
thoughtful
Dr
and
Maglinte
respond:
Jehenson for his interest and comments on our article.
He suggests that moderate hypotonicity)
to obtain the desired of glucagon that
very
reactions the drug
perceptive (atonicity be given
or in-
travenously in multiple small doses at 10-minute intervals. Thus, the desired effect would be obtained continuously for the required interval without the side effects noted (nausea) for the sum of these doses where given as one dose. Indeed, to our knowledge, this study has not been done. It is possible this technique may reduce the side effects as indicated; however, we doubt it. We believe the side effects of nausea and vomiting occur as a direct effect of glucagon on the central nervous system when the drug is given rapidly intravenously. This is thought because patients become nauseated and vomit almost immediately. When the drug is given slowly intravenously, nausea and vomiting usually occur in 1-2 hours. The reason for this delay is not known. Glucagon is a hormone of the pancreas. The commercially produced material has the same amino acid sequence as that of the human hormone. When given intravenously, it is rapidly metabolized to its constituent amino acid. The half-life of glucagon in plasma is 3-6 minutes, which is similar to that of insulin. The side effects we discussed occur 1 -2 hours after the administration of the drug. Therefore, the side effects may be due to glucagon stimulation of other body substances and hormones, or to its degradation products. In either case, because side effects occur so long after glucagon is given, we think smaller doses given more frequently are probably additive. Thus, the frequency of side effects may be the same as those from a large dose given slowly intravenously or the same dose given in multiple small doses or as a perfusion. However, as we said earlier, this study has not been done and
April
1991