INADVERTENT intravascular injection of large amounts of local anesthetic solutions intended for epidural anesthesia can result in potentially lethal cardiovascular or central nervous system toxicity. 1,2Thus, administration of an epidural test dose containing 10–15 μg epinephrine through an epidural catheter is common practice to diagnose intravascular injection of a local anesthetic solution. 3
For detecting intravascular injection of an epinephrine-containing test dose, hemodynamic criteria have been used commonly. 3More recently, changes in T-wave amplitude in lead II electrocardiography have been implicated in intravascular injection in adult and pediatric patients. 4,5In this report, we describe a case of an elderly woman who had a decrease in T-wave amplitude with essentially unaltered systolic blood pressure (SBP) and heart rate (HR) after unintentional intravascular injection of unknown amount of the epinephrine test dose during combined epidural–general anesthesia.
A 77-yr-old woman (American Society of Anesthesiologists physical status II; weight, 62 kg; height, 140 cm) presented for revision of total hip arthroplasty. Except for hypertension treated with 10 mg oral nifedipine four times daily and obesity (body mass index = 32), she was healthy, and her activities of daily life were not restricted. At admission, her blood pressure (BP) and HR were 130/70 mmHg and 66 beats/min, respectively. Physical and laboratory examinations, including serum electrolytes, yielded otherwise normal results, and electrocardiography showed regular sinus rhythm without significant ST–T-segment abnormalities.
She was administered premedication with 10 mg oral diazepam and 20 mg famotidine 90 min before induction of general anesthesia. A radial arterial catheter was placed after local anesthetic infiltration for subsequent BP measurements. With the patient in the right lateral decubitus position, a 20-gauge epidural catheter with multiple orifices was placed at the L4–L5 interspace using an 18-gauge Tuohy needle and a loss of resistance to saline technique. The tip of the catheter was advanced 5 cm into the epidural space without resistance or paresthesia. The patient then was turned to the supine position. After an aspiration test for blood and cerebrospinal fluid that yielded negative results, 3 ml lidocaine, 1.5%, with 15 μg epinephrine (1:200,000) wasinjected through the catheter. After confirming negative responses in SBP, HR, and electrocardiographic morphology in lead II, 3,4an additional 10 ml of the same lidocaine–epinephrine solution was administered in divided doses. BP was initially 140/60 mmHg and gradually decreased to 110/54 mmHg, with a concomitant HR increase from 64 to 69 beats/min 15 min after injection, when sensory analgesia determined by a pin-prick method was below T10 bilaterally. General anesthesia then was induced with 100 μg intravenous fentanyl and 250 mg thiopental, and tracheal intubation was facilitated with 8 mg intravenous vecuronium. Her lungs were ventilated mechanically with 0.5–1.2% (inspired) sevoflurane and 67% nitrous oxide in oxygen, while end-tidal carbon dioxide tension was maintained between 28 and 34 mmHg thereafter. The surgery commenced with the patient in the right lateral decubitus position and was uneventful until a reinforcing dose of the identical lidocaine–epinephrine solution was necessary, approximately 60 min after the initial dose. The aspiration test using a 20-ml syringe again yielded negative results for blood. Her BP and HR were 111/63 mmHg and 63 beats/min before injecting 3 ml of the test dose through the epidural catheter and remained 105–117/59–65 mmHg and 63–67 beats/min within the next 5 min after injection, respectively. However, a decrease in T-wave amplitude on a strip-chart was noted on the lead II electrocardiogram from 30 to 60 s after the test dose injection, and the T-wave morphology returned to the preinjection level in 2 min (fig. 1). Using a 1-ml syringe, frank blood could be aspirated freely. Arterial blood gas analysis and electrolytes after this event were unremarkable. Inspiratory concentration of sevoflurane was 1% at this time. Epidural anesthesia was halted then, and anesthesia was maintained using intravenous fentanyl, sevoflurane, and nitrous oxide, without event during the rest of the planned surgical procedure.
Unrecognized intravascular injection of large amounts of local anesthetic solution is a significant hazard of epidural blockade. As seen in the current patient, vessel entry of the epidural catheter tip may occur not only during initial insertion, but also at any time during ongoing epidural therapy. 6Because reported frequency of vessel entry ranges from 0.2% to 11%, 6the earliest possible detection of the catheter migration is essential for preventing life-threatening cardiovascular or central nervous system complications. Although typical HR increases of 20 beats/min or more after the test dose injection have been regarded as a testing threshold in awake subjects, 3SBP increases of 15 mmHg or more seem to be more reliable during sevoflurane anesthesia. 4Similarly, a recent study has shown that a decrease in T-wave amplitude of 25% or more in lead II is at least as sensitive as the SBP criterion but more sensitive than the HR criterion when a fractional dose of the epinephrine-containing test dose is injected intravenously. 7Although the use of T-wave morphology has been regarded only as an adjuvant diagnostic tool, virtually absent hemodynamic alterations and an easily detectable decrease in T-wave amplitude in the current patient indicate potential clinical usefulness of the T-wave criterion for the diagnosis of intravascular injection of the test dose during general anesthesia.
To the best of our knowledge, concomitant changes in T-wave morphology with considerably attenuated or virtually absent hemodynamic responses associated with the intravascular test dose have not been reported previously in the literature. In our case, several factors may have contributed to the absence of significant hemodynamic changes. Advanced age and volatile anesthetics may have played a role in diminished HR response to the intravenous epinephrine. 8,9Even though we cannot exclude the possibility that preoperative oral nifedipine might have modified cardiovascular responses to the intravenous test dose, previous studies have shown that intravenous nifedipine did not alter epinephrine-induced BP and HR responses significantly in cats 10and that BP increase in response to psychologic stimuli associated with increased plasma epinephrine concentration was not suppressed by nifedipine in hypertensive humans. 11Although the reliability of HR response for detecting intravascular injection of the test dose is known to decrease in acutely ß-blocked subjects, 3effects of other antihypertensive medications on hemodynamic changes to and effectiveness of the intravenous test dose have not been addressed previously. In our case, a more likely explanation for the absence of SBP response may be that an amount of epinephrine inadvertently administered into the epidural vein was not sufficient to cause the plasma epinephrine concentration that produces α-adrenoceptor–mediated vasoconstriction and resultant BP increase. Because the current patient had a significant, detectable decrease in T-wave amplitude, it is of important clinical relevance to determine the minimum effective epinephrine dose that should elicit reliable changes in T-wave morphology in various clinical circumstances.
In conclusion, our reported case indicates that close attention should be paid to electrocardiographic morphology immediately after the epidural test dose injection, especially when the patient is associated with factors that depress hemodynamic responses, such as advanced age and administration of concomitant general anesthetic.