Accidental Intraarterial Drug Injections via Intravascular Catheters Placed on the Dorsum of the Hand

Abstract

Accidental intraarterial injections of IV substances used during anesthesia can result in significant morbidity secondary to local ischemia and subsequent tissue necrosis. Such accidents are relatively rare, but can contribute to a disastrous outcome if not recognized early (1,2). Intraarterial injections are usually caused by accidental injection into an arterial line. We describe two cases of cannulation of aberrant arterial vessels located on the dorsum of the hand that were thought to be veins. We also review the literature on the pathogenesis and various therapies used to treat intraarterial injections with thiopental and propofol. Case Reports Patient 1 A right-handed, 37-yr-old man was scheduled for debulking of a posterior fossa brain tumor. The patient had no other medical history, and no anesthetic history. His vital signs on admission were: heart rate 68 bpm, blood pressure 125/72 mm Hg, temperature 36.9°C, and respiratory rate 18 breaths/min. Preoperatively, an IV catheter was established by a nurse for magnetic resonance imaging-guided stereotactic frame placement. A 20-gauge plastic catheter was easily inserted into a superficial vessel on the dorsum of the left hand just posterior to the middle finger approximately 3 in. from the wrist (see Fig. 1). The IV seemed to be functioning well and the patient was given 450 mL of lactated Ringer’s solution and 100 μg of fentanyl through the catheter without difficulty or pain.Figure 1: Intraarterial catheterization of the radial artery on the dorsum of the hand into a vessel appearing as a vein.After the magnetic resonance imaging procedure, the patient was taken to the operating room and was alert and cooperative. Standard monitors were placed, and anesthesia was induced with 100 μg of fentanyl, 350 mg of thiopental, and 120 mg of succinylcholine through the same 20-gauge catheter. After endotracheal intubation, a triple-lumen right internal jugular catheter was placed without difficulty. The patient was then rotated 180°, and an additional 16-gauge IV line was established in the left saphenous vein on the foot. Finally, a left radial arterial line was placed near the wrist without difficulty. Approximately 45 min into the procedure, it was noticed that the 20-gauge left hand IV was functioning poorly. Initially, it was thought to be positional, but flushing the line with saline produced an instantaneous and marked increase in arterial blood pressure (as measured by the arterial line on the same hand) of >200 mm Hg. Closer inspection of the line with the fluid bag placed near the floor revealed pulsatile retrograde blood flow into the plastic tubing. A sample of blood was aspirated from the catheter, and its analysis revealed a pHa of 7.45, Paco2 of 38 mm Hg, and a Pao2 of 100 mm Hg with an Fio2 of 32%, confirming arterial cannulation. The catheter was discontinued, and immediately postoperatively, the patient developed 2+ pitting edema and bluish discoloration of the left hand. The arm was elevated for 72 h, and regularly massaged by the ward nurse. Because it showed steady improvement in appearance, no specific therapy was instituted, and within 4 days, all sequelae had completely resolved. Patient 2 A previously healthy, 23-yr-old woman was brought to the emergency room after a traffic accident involving an impact of faster than 50 miles per hour. Upon arrival in the emergency room, she was obtunded and hypotensive, with an initial systolic blood pressure of 60 mm Hg and a heart rate of 140 bpm. An 18-gauge IV had been placed on the dorsum of the right hand by paramedics, and seemed to be functioning well. The location of the catheter was approximately 2 cm posterior to the index finger. The trachea was intubated after administration of 100 mg of succinylcholine through this catheter. After confirming equal breath sounds and CO2, 2 additional 16-gauge antecubital IV catheters were started (1 on each arm), and a radial artery catheter was also introduced near the left wrist without difficulty. The patient received a total of 3 L of lactated Ringer’s solution through the 18-gauge right hand IV, 100 μg of phenylephrine, 5 mg of succinylcholine, and 1 U of packed red blood cells when it was noticed that there was retrograde, pulsatile blood flow in the tubing. Flushing of this catheter resulted in increased blood pressure as measured by the radial artery catheter on the same hand. An aspirated blood sample from the presumed IV catheter was submitted for gas analysis and revealed a pHa of 7.34, Paco2 of 35 mm Hg, and a Pao2 of 490 mm Hg with an Fio2 of 100%, confirming its arterial origin. The catheter was immediately discontinued. Postoperatively, the hand appeared mildly mottled, but there was no evidence of ischemia 24 h later. The patient died on the fourth postoperative day, because of the development of respiratory distress syndrome, but there was no necrosis noted in the hand attributable to the previous intraarterial injections. Discussion Initial reports of accidental intraarterial injections of thiopental appeared more than 50 years ago (3,4). Fortunately, this is an infrequent event, estimated to occur between 1 in 3440 to 1 in 56,000 cases involving the parenteral administration of medications (1,5). The incidence is difficult to determine accurately because it is an infrequent event. In both cases presented herein, accidental arterial cannulation was performed by nonphysician personnel, but this clearly occurs even in the hands of experienced anesthesiologists. Moreover, in both cases, significant time elapsed before arterial catheterization was discovered, and medications had already been administered. Signs suggestive of intraarterial cannulation include: 1) anatomical location in areas where an artery is expected, 2) presence of pulsatile movement of blood, 3) manual palpation of an arterial pulse proximal to the catheter, 4) distal signs of ischemia (e.g., nail bed anemia), 5) bright red appearance of blood, and 6) a more intense than expected level of pain at the catheter location. None of these alone or in combination is absolutely diagnostic of intraarterial cannulation. However, transducing the pressure and/or blood gas analysis will typically confirm arterial cannulation in most patients. Unless the patient is hypotensive, an arterial waveform should demonstrate a characteristic tracing, possible dicrotic notch indicating aortic valve closure, and a pressure reading consistent with arterial catheterization. However, in hypotensive patients, or in patients with a catheter blockage (e.g., kink), pressure alone may be an unreliable indicator. A blood gas analysis can also be performed, and in some cases a very high Pao2 with low inspired Fio2 generally establishes arterial blood sampling. Again, in some circumstances, arterialization of the venous blood is possible (e.g., arteriovenous fistula), so this alone is not completely diagnostic. In summary, an index of suspicion combined with the presence or absence of the aforementioned observations should alert health care personnel to the possibility of arterial cannulation. If the suspicion is high, and medications are to be administered, it is wise to repeat IV catheterization unless the situation is extremely dire and IV access is not readily obtainable. Studies involving therapy of accidental intraarterial drug injections of anesthetics are extremely limited. In fact, almost all case reports and investigations only address those injections involving thiopental and propofol. Although the patients received additional substances (phenylephrine, succinylcholine, red blood cells, and thiopental), it cannot be determined what effect each (if any) of these substances had individually. However, it is worthwhile to discuss the literature with regard to thiopental, because this drug results in a well defined clinical syndrome and is very commonly used to induce anesthesia. After the intraarterial injection of thiopental (and rarely propofol), severe pain radiates along the distribution of the vessel into which it has been administered. This clinical sign however, is not always elicited or recognized, because anesthesia and unconsciousness are being induced, and propofol can cause exquisite pain even with IV injection. In the first several hours, the clinical presentation is variable, but if severe, skin pallor, hyperemia, and cyanosis may occur. In awake patients, hypesthesia, muscular weakness, paralysis, and anesthesia may occur. Severe cases are associated with profound edema and, ultimately, gangrene. The location of tissue necrosis can be proximal and distal to the site of injection. The risk factors associated with accidental intraarterial injections are very difficult to assess because of the rarity of the event. Potential risk factors based on studies are summarized in Table 1.Table 1: Risk Factors Associated with Inadvertent Arterial CannulationAccording to Wood et al. (6), superficial radial arteries in the forearm and hand are found in 1% of all cases. Bilateral superficial radial arteries, like the ones described in this case, are even less common. In a pattern similar to that of primates and some prosimiae, the radial arteries of both hands passed superficial to the lateral tendons of the anatomical snuff-box. In addition, the left hand showed a median artery originating distally from the radial artery and coursing ventral to the flexor retinaculum. In humans, a high origin of the radial artery is the most common arterial anomaly of the upper limb. One result of this is the presence of a palmar arch formed completely by the ulnar artery and the incomplete formation of the radial artery. In the styloid process of the radius, the radial artery was present only as a thin superficial palmar branch that arrests in the thenar region instead of forming the palmar arch (7). In another case, the complete absence of the radial artery was observed. Codominant median and ulnar arteries were present to supply the forearm with blood (8). In an anomaly known as “antebrachialis superficialis dorsalis,” a condition present in 1% of patients, the radial artery bifurcates in the forearm. The main branch of the radial artery then passes dorsally and radialward and ends in an incomplete deep palmar arch. The anomalous superficial branch supplies blood to the radial side of the index finger and the ulnar side of the thumb (9). When the anomalous artery is present, the hand probably receives less blood from the ulnar artery (10). Variations of the superficial palmar arch include cases in which the ulnar artery is the primary blood vessel forming the arch (66%) and cases in which the superficial radial palmar artery is responsible for the formation of the arch (30%). In the first case, the seven different types of blood vessel formations vary from the superficial blood supply being completely provided by the ulnar artery to being entirely supplied by the radial artery. The most common variation is one in which the radial artery forms the two volar collaterals of the thumb and the collateral radial of the index finger whereas the ulnar artery terminates in the second intermetacarpal space. In the five different categories of the second case, the superficial branch of the radial artery participates in the formation of the superficial palmar arch to varying degrees (11). Other variations of the radial artery in the hand occur in the deep palmar arch. These include differences in the formation of the deep palmar arch by the interior interosseous artery and either the radial artery or the deep branch of the ulnar artery (11). Formation of the dorsal carpal arch varies as well. In approximately 58% of cases, the radial dorsal carpal branch forms the dorsal carpal arch by itself. In the next most frequent variation, the dorsal carpal arch is made up by the joining of the radial carpal arteries, ulnar carpal arteries, and corresponding arteries (11). The most common variation of the superficial volar arch is one in which the ulnar artery alone completes the arch (37%). In the second most common type, the superficial volar branch of the radial artery and the ulnar artery join together to form the arch (34.5%). Another variation that occurs in only 1.2% of patients is one in which the superficial volar arch is formed by the radial, median nerve, and ulnar arteries (11). Another common anomaly is the superficial brachial artery forming a common trunk for the radial artery and ulnar artery (11). Many factors are thought to be responsible for the clinical syndrome of pallor, ischemia, pain, and tissue necrosis that can result after intraarterial drug injections. These include: 1) arterial spasm caused by the local release of norepinephrine or directly mediated by the drug, 2) direct tissue destruction by the drug (which is known to occur in the case of thiopental, but probably not propofol) (12,13), 3) subsequent chemical arteritis leading to the destruction of endothelium, subendothelium, and/or the muscle layers of the vessel, and 4) the release of endogenous substances that mediate deleterious effects themselves (e.g., thromboxane leading to vasoconstriction and subsequent thrombosis) (14). There are no well controlled studies that clearly demonstrate the effectiveness of any given therapy over another. The few controlled animal investigations are still very limited because of extremely small sample size and, in some cases, results are conflicting. The following is a brief discussion of the therapies that have been advocated. Elevation of the Extremity Raising the involved extremity is intended to facilitate venous drainage which may be impaired (if the tissue edema is too great and interstitial pressure exceeds venous pressure). Intuitively, this sounds reasonable; however, it is also possible that arterial vascular compromise (if not complete) should be the focus of therapy. If partial thrombosis or spasm is the predominating factor in producing ischemia, then one might argue that the limb be placed in a dependent manner to improve circulatory flow. Local Anesthetic Infiltration Chemical (local) sympathectomy is advocated to prevent reflex vasospasm and lead to local vasodilation. This is generally benign and may be used, but attended risks are also possible, such as arterial laceration or damage worsening the injury and/or contributing to thrombosis. Extremity Sympatholysis Using Local Anesthetics Stellate ganglion blocks or lower-extremity sympathetic blocks are possible methods to produce sustained arterial and venous vasodilation, both of which are desirable. However, the advantage of performing blocks must be weighed against the risks of the procedure. For example, a stellate ganglion block by an inexperienced anesthesiologist in a person with an obese neck can be risky, leading to pneumothorax, spinal anesthesia, recurrent laryngeal nerve or phrenic nerve paralysis, local anesthetic toxicity, etc. Therefore, these therapies must not be regarded as benign. They may be used if there are no contraindications to the procedure, and as an alternative to more drastic measures in the face of potential tissue necrosis (e.g., surgical revascularization or amputation). Other Chemical Therapies Many substances have been tried in the hopes of preventing thrombosis, vasospasm, or edema. These can be classified into three groups: 1) arterial vasodilators (e.g., reserpine, tolazoline), 2) specific thromboxane inhibitors (e.g., methimazole, iloprost, aloe vera) and nonspecific thromboxane inhibitors (e.g., aspirin, methylprednisolone), and 3) high-molecular-weight dextrans (e.g., dextran 40). Unfortunately, these are all of unproven value. The few studies to date are animal models with small sample sizes. At this point, no specific therapy can be strongly advocated. A summary of investigations is presented in Table 2.Table 2: A Summary of Related Case Reports and InvestigationsIn conclusion, accidental intraarterial injections are rare, but can occur even in experienced hands. There are no well established therapies, and persistent evidence of tissue injury requires evaluation by a vascular surgeon. Thiopental is clearly a more damaging drug than propofol, and demonstrates direct toxicity to the vascular endothelium; however, both substances can lead to ischemic changes. Prevention by vigilance is the best therapy: 1) an indwelling catheter should never simply be assumed to be positional, 2) a tourniquet should not be tight to the point of occluding arterial blood flow, allowing for unrecognized arterial catheterization, 3) hypotension must be considered as a reason why arterial cannulation may be overlooked, and 4) aberrant arterial anatomy should always be kept in mind, especially in the hand.