ClinicalRadiology (1990)
41, 139-140
Case Report: Delayed Cardiac Tamponade Following Accidental Injection of Non-Ionic Contrast Medium into the Pericardium M. J. G A L L A N T and *J. G. N. S T U D L E Y
Departments of Radiology and *Surgery, Northampton General Hospital, Northampton
A case of cardiac tamponade developing over 10 hours after accidental injection of 80 ml iohexol (350 mg i/ml) during an intravenous digital subtraction arteriography (DSA) examination is described.
CASE REPORT A 55-year-old m a n had an elective aorto-bifemoral graft for occlusive peripheral vascular disease. Twenty-four hours following operation the right foot was noted to be cool and no pulses could be felt in that leg. Absence of pulses was confirmed on Doppler examination. An intravenous D S A examination was requested. A 5 gauge D S A catheter (Kimal) was introduced into a left antecubital vein and passed without difficulty into the right atrium where its position was confirmed on fluoroscopy with a small hand-held injection of contrast. At this stage the catheter tip appeared to be free within the atrial cavity. The nine inch field of view image intensifier was therefore centred over the origin of the right superficial femoral artery and an injection of 40 ml iohexol 350 was made at a rate of 17 ml per second. No contrast medium was seen in the graft or in any other vessel and it was assumed that either a block was present at a higher level, or that the timing had been incorrect. The intensifier was therefore centred at the upper level of the graft and a further injection of 40 ml made. No contrast m e d i u m was seen following this and the heart was therefore screened. W h e n contrast medium was seen in the pericardium (Fig. 1) an ECG monitor recorded a normal trace, the patient's heart rate was unaffected and the patient denied any symptoms. The procedure was abandoned a n d the catheter removed. Subsequently the patient returned to the operating theatre where reexploration of the right groin showed that the graft was occluded with clot. A stenotic lesion was noted just distal to the origin of the profunda femoris artery and a profundaplasty performed. During the subsequent 3 hours the pulse rate slowly increased to 120 beats per minute and the central venous pressure (CVP) rose to + 25 cm H20. He became oliguric and his general condition deteriorated. An E C G confirmed an echo-free space in the pericardium which was considered to contain greater than 80 cm 3 of fluid. A further chest radiograph showed persistence of contrast medium in the pericardium (Fig. 2). The heart had a somewhat globular appearance and enlargement of the mediastinal shadow above the contrast medium was presumed to be due to venous distention. A diagnosis of cardiac tamponade was made and the patient was returned to the operating theatre. The upper end of his abdomir~al incision was reopened, the pericardium exposed and opened with release of 400 ml of clear, colourless fluid and immediate relief of the cardiac tamponade. The patient made an uneventful recovery and returned h o m e 16 days later.
contrast medium injected into the pericardial space of the patient described attracted sufficient volume of fluid over a period of 10 hours to cause tamponade. Extravasation of contrast medium from the right atrium or superior vena cava is a rare but well recognised complication of intra cardiac contrast medium injection. Pinto et al. (1984), recorded extravasation of contrast medium into the mediastinum in two of 2488 examinations. They described one case of pericardial effusion associated with right atrial extravasation, noting a small pericardial effusion initially on a cardiac echogram. Repeat examination 24 hours later showed that the pericardial effusion had reabsorbed. Extravasation following intra cardiac or superior vena cava (SVC) injection has been associated with the use of a straight catheter and is thought to be related to catheter recoil during injection. Pinto et al. (1984) attributed the small pericardial effusion recorded in their case to inadvertent injection into the coronary sinus and recommended that right atrial injection be avoided. They did not favour the use of a pigtail catheter because of increased difficulty in manipulating the tip and greater patient discomfort on introduction and withdrawal. Seeger and Carmody (1985) dis-
ii
•
DISCUSSION Delayed cardiac tamponade following inadvertent injection of contrast medium into the pericardium has not been previously reported. Non-ionic contrast media have an osmolality about one third that of conventional ionic media providing the same iodine content, Nonetheless, we presume that despite its low osmolality the volume of
Correspondence to: D r M. J. Gallant, Department o f Radiology, Northampton General Hospital, N o r t h a m p t o n NN1 5BD.
;i
~?~:~i~
71i
......~!iii,i~iiiii!i¸•~
.....
Fig. 1 - R a d i o g r a p h demonstrating contrast in the pericardial cavity immediately following D S A injection into the right atrium.
140
CLINICAL RADIOLOGY
Fig. 2. Radiographdemonstrating globular appearance of the heart and enlargement of the mediastinal shadow suggesting tamponade.
agreed with this view and felt that the use of a pigtail catheter reduced the incidence of complications o f this kind, observing only one case of extravasation in 4000 patients in whom pigtail catheters had been used, half of these having an SVC injection and the other half a right atrial injection. It is unclear why extravasation of contrast occurred in our case. The position of the tip of the catheter was clearly demonstrated by hand injection before it was connected to a pumpi The cathether was secured to the skin with tape and it is unlikely that it could have been dislodged. It is possible that the atrial wall was ruptured as a consequence of catheter recoil on injection, associated with the use of a straight catheter. If this is true, extravasation may not have occurred if a pigtail catheter had been used. The complication could also have been avoided if the tip of the catheter had been positioned in the superior vena cava above the pericardium. In our patient tamponade was relieved by a formal pericardotomy. In retrospect, the problem would have been relieved by simple pericardiocentesis. However, as delayed tamponade was hitherto an unknown complication following DSA, we thought that there must have
been haemorrhage from the atrial wall related to the heparin received during the profundaplasty. As clot cannot be removed satisfactorily through a pericardiocentesis needle a pericardotomy was deemed to be necessary. The subxiphoid approach was facilitated by opening the upper end of the patient's abdominal incision. If there had been persistent bleeding from within the pericardial cavity following evacuation of its contents a median sternotomy would have been performed to repair the damage. This did not prove to be the case.
Acknowledgement. We thank Mr D. B. Hamer, FRCS, Consultant
Surgeon, for agreeing to have details of this case published and Mrs B. C. Studley for typing the manuscript.
REFERENCES
Pinto, RS, Manuell, M & Kricheff, II (1984). Complications of digital intravenous angiography:Experience in 2488 cervicocranialexaminations. American Journal of Roentgenology, 143, 1295-1299. Seeger, JF & Carmody, RF (1985). Digital subtraction angiographyof the arteries of the head and neck. Radiologic Clinics of North America, 23, 193-210.