Extravasation is an unintentional injection or leakage of fluid in the perivascular or subcutaneous space. Extravasation may or may not be associated with tissue injury. Extravasation injury results from a combination of factors, including solution cytotoxicity, osmolality, vasoconstrictor properties, infusion pressure, regional anatomical peculiarities, and other patient factors (1–3). Extravasation injury has been frequently described in patients undergoing cancer chemotherapy; thus, oncologists and related medical personnel are well aware of the severe untoward effects when cytotoxic drugs leak into the perivascular space. In contrast to the oncologic setting, extravasation in the perioperative period often involves nontoxic fluids. Because of this, although extravasation is fairly common, it is usually without serious consequences. Furthermore, the potentially injurious effects of extravasated hyperosmolar solutions and concentrated electrolytes are not well known. This has led clinicians in this area to underestimate the potentially serious consequences of extravasation.
There are surprisingly few data regarding the incidence, consequences, and management of extravasation in the anesthesia and intensive care literature. Also, information regarding the potential of any given drug to cause extravasation injury is not widely available.
To assess the morbidity associated with extravasation, we retrospectively reviewed the hospital files of five patients who had sustained a significant extravasation injury in the perioperative setting and had been treated at the University Hospital of the Friedrich-Schiller University (Jena, Germany) or at the Zentralklinik (Bad Berka, Germany). Data examined included age, the substance implicated in the extravasation injury, the site of extravasation injury, the method of treatment, and long-term results.
For coronary revascularization surgery in a 73-yr-old man, a 7F triple-lumen catheter was placed into the right internal jugular vein. The cannulation procedure was performed without any problems, and the catheter was carefully fixed at 15 cm by suturing it to the skin. On the second postoperative day, the patient became confused and agitated. The dressing of the central line was found to be soaked; upon its removal, the neck was found to be swollen at the insertion site, and the catheter was displaced. The distal lumen had been used for continuous central venous pressure measurement, and the proximal lumen had been used only for potassium chloride infusion. The infusion of potassium chloride 7.45% (1 mmol/L) at 20 mL/h was stopped after 35 mL had been infused, and the catheter was removed immediately. However, the exact amount of potassium chloride extravasated could not be determined. A surgeon was consulted, who initially proposed conservative management with dry dressings. One week after extravasation, skin necrosis became apparent, and local treatment with topical antiseptic ointment was started; this led to no substantial improvement. Four weeks after the incident, swelling developed at the site of extravasation and extended to the tissue overlying the parotid gland. An infection was suspected, and IV treatment with antimicrobials (amoxicillin, clavulanic acid, and gentamicin) was initiated. Vancomycin was added when culture of the wound swab revealed methicillin-resistant Staphylococcus aureus. Blood cultures were negative. Five weeks after the incident, magnetic resonance imaging of the neck revealed deep necrosis of the neck soft tissue with imminent erosion of the carotid artery. The patient was stable at that stage, the lesion was painless, and there was no increase in inflammatory variables such as leukocytes or C-reactive protein. Neck dissection was performed with débridement of necrotic tissue and partial resection of the sternocleidomastoid muscle. The common carotid artery was saved without damage. Two further wound débridements were necessary. Secondary wound healing was supported by adjuvant hyperbaric oxygen therapy (2.5–3 atmospheres [atm] absolute; 15 treatments, each last-ing for 60–90 min). Eight weeks after extravasation, the affected areas were surgically covered with a vascularized pectoralis flap.
A 64-yr-old man underwent cervical spine surgery. Because the operation was performed transorally, nil per os was ordered for the subsequent 5 days, and a single-lumen central venous catheter was inserted via an antecubital vein. On the third postoperative day, parenteral nutrition (Aminomix-3®; Fresenius Kabi, Bad Homburg, Germany; osmolarity, 1149 mOsm/L; 680 kcal/L) was started at 2000 mL/d through the central venous catheter. To allow physiotherapy, the infusion was disconnected but was then erroneously reconnected to a peripheral cannula on the dorsum of the right hand. During a 6-h infusion, a large but unknown amount (maximum of 500 mL) of the parenteral infusion leaked into the extravascular space. When this was discovered, the hand was swollen, red, and blistered. The involved area measured 8 × 5 cm. The cannula was removed, and local cooling was applied. Over the next 7 days, a deep necrosis involving the skin and subcutaneous tissue developed over the right hand. Wound débridement and two operations involving split skin grafting were subsequently performed. The patient remained hospitalized for 8 wk because of the incident.
A 40-yr-old paraplegic man was found unconscious in his hospital room. During the subsequent cardiopulmonary resuscitation, the anesthesiologist placed a peripheral cannula in the right femoral vein. A total of 1 mg of epinephrine and 200 mL of sodium bicarbonate solution (8.4%) were injected through the cannula. Hemodynamic stability was achieved, and the patient was transferred to the intensive care unit. Here extravasation of fluids in the vicinity of the femoral venous cannula was noticed; the cannula was removed, and a dry dressing was applied. Within a week, extensive skin and subcutaneous tissue necrosis (approximately 15 × 8 cm) with infectious complications developed. Multiple surgical débridements and, 2 mo after the injury, skin grafting was necessary to treat and cover the injury. The patient survived with diffuse cerebral injury and was later transferred to a nursing home.
A 71-yr-old patient presented for repair of a thoracic aortic aneurysm. In the course of stabilizing the circulation after cardiopulmonary bypass, 100 mL of sodium bicarbonate solution (8.4%) was injected through a cannula placed in the dorsum of the left hand. After the patient arrived in the intensive care unit, extravasation was discovered. The dorsum of the hand was incised, and drains were inserted, but no flushout was performed. The extravasation injury progressed to severe soft tissue necrosis and required multiple sessions of surgical débridement. Finally, a large defect was closed with a vascularized latissimus dorsi flap (Figs. 1–3).
A 71-yr-old man became hypotensive after the induction of general anesthesia for carotid endarterectomy. Approximately 15 mL of a norepinephrine solution (20 μg/mL) was infused by infusion pump via an IV cannula placed in the right antecubital fossa. When the anticipated pharmacologic effect failed to occur, the anesthesiologist inspected the cannulation site and discovered extravasation. Another cannula was placed, norepinephrine was reinjected, and the patient's arterial blood pressure was stabilized quickly. After completion of carotid endarterectomy (60 min after extravasation), a flushout with 500 mL of normal saline was performed, and drains were inserted for 24 h (Fig. 4). Four weeks later, the site of extravasation had healed, without appreciable soft tissue loss.
In the perioperative period, extravasation of hyperosmolar solutions, vasoconstrictive substances, or concentrated electrolyte solutions may cause significant tissue necrosis. Tragically, the extravasation injury can sometimes be worse than the condition that originally brought the patient to the hospital. Even amputation of an extremity, although rare, has been reported as a consequence of extravasation injury (2,3).
Tissue damage after extravasation may be slight and may involve a local and limited inflammatory response or may be large and involve necrosis of the skin and underlying soft tissues (4–7). The degree of damage depends on the localization of the extravasation (4), the physicochemical characteristics of the agent administered, and the duration of soft tissue exposure to the agent, as well as on the patient's general health (Table 1). Pain, swelling, or local hyperthermia are not reliable predictors of the degree of tissue damage (4). Furthermore, agents that cause pain even during intravascular injection may not necessarily cause tissue injury upon extravasation. As an example, small amounts of propofol may cause substantial pain upon intravascular injection, but even large amounts of propofol cause no damage upon extravasation (8,9). In contrast, extravasation of substances that subsequently lead to major tissue injury may cause only minor discomfort in the early stages of extravasation. The patient in Case 1 mentioned only “some tension” after extravasation of potassium chloride (Table 2). Paresthesia, local induration of the skin, and ulcers are late signs of extravasation that may take days to develop (4,10). However, an induration lasting longer than 24 hours is a reliable sign for the formation of necrosis (7). These peculiarities of extravasation injury (see Cases 1 and 2) show that the severity of extravasation injuries is easily misjudged and that inadequate early treatment may result in severe morbidity (Figs. 1–3).
Most of the available data on extravasation are related to peripheral intravenous lines. Brown et al. (11) investigated 16,380 children with peripheral venous infusions and reported an extravasation rate of 11%. Most of these extravasations caused no severe damage and required no further therapy. However, extravasations with serious consequences have been noted in premature infants and neonates (1,12). Local skin necrosis after extravasation of chemotherapy drugs is responsible for 0.5%–6% of all side effects during antineoplastic therapies (4,10). Even in implanted venous access devices, extravasations caused by needle dislocation have been reported with an incidence of 6.4% (13).
For central venous catheters, extravasation is less frequent but potentially more dangerous because of the vulnerable anatomical structures and because extravasation might easily escape attention. Depending on insertion depth, the extravasal position of the proximal port can occur when the catheter is inadvertently withdrawn just a few centimeters. In triple-lumen catheters for adults, the distance between the proximal and the distal port varies between 4 and 5.4 cm (14); this increases up to 8.75 cm in 5-lumen catheters (15). Not all clinicians are aware of this issue. In both central and peripheral lines, extravasation may go unnoticed for a long period when the area is covered by drapes during surgery.
Agents and solutions that can cause tissue destruction are called vesicants. In oncology, the mechanism of injury can be related to the cytotoxic nature of the drug (4,5). In the perioperative period, the mechanism of tissue necrosis can be summarized as follows:
- Vasoactive substances: these agents cause tissue injury by inducing ischemia through severe vasoconstriction (16).
- Concentrated electrolyte solutions: large concentration of electrolytes may cause prolonged depolarization and contraction of pre- and postcapillary smooth muscle sphincters (2), thus leading to ischemia and tissue injury.
- Hyperosmolar solutions: these solutions exert osmotic pressure and may lead to a compartment syndrome (17).
The most important measure for minimizing complications caused by extravasation is prevention by absolute prohibition of infusing vesicants via IV catheters located outside the visual sight and observation limits of the anesthesiologist.
Multiple punctures of the same vein, a high infusion pressure, a tourniquet effect, and peripheral access sites in close proximity to tendons, nerves, or arterial vessels should be avoided. Vesicants should be injected by reliable peripheral venous access: the forearm is preferable to the dorsum of the hand. Whenever a central venous catheter is in place, all vesicants should be applied through the most distal port of this line. All venous accesses should be visible and checked regularly, and patients should be educated to recognize abnormalities in connection with venous access cannulas to allow early treatment of extravasation.
After an extravasation has occurred and has been recognized, a systematic approach (Fig. 5) may help to prevent extensive tissue injury. The infusion must be stopped immediately. The type of substance, the amount, the particular region, and the length of contact with the extravasated agent all affect the correcting procedures. Commonly recommended conservative procedures such as elevating the involved extremity or applying heat or cold have not shown any benefit (5,6,11). Also, mere aspiration of the extravasated liquid alone is unlikely to confer any benefit. Gault (3) reviewed 44 patients who were treated within 24 hours of the extravasation injury (thus, before skin necrosis had occurred) and found that saline flushout is an easy and practical procedure. He also found that saline flushout alone was as effective as the combination of saline flushout and aspiration with a liposuction cannula. Most (86%) of these patients healed without any soft tissue loss.
On the basis of these limited findings, we recommend a complete removal of the extravasated substance by stab incisions and flushing with 500 mL of normal saline. This measure should be done as soon as possible, but certainly within 24 hours of extravasation injury (2,3). Drains should be inserted and should remain in place for 24 hours (2).
There are contradictory reports on the efficacy of treating extravasations with topically applied drugs (2,18). Specifically, neither clinical nor animal-based trials have shown a positive effect of topically applied cortisone (2,10,18). Other recommendations include subcutaneous infiltration with hyaluronidase, which may accelerate diffusion (10,19), and antidotes such as dimethyl sulfoxide (20) and nitrates (13,21). The intradermal application of sodium thiosulfate or sodium bicarbonate has also been recommended (10,16). However, sodium bicarbonate can also be toxic to tissues (Tables 2 and 3). After extravasation of vasopressors such as epinephrine, norepinephrine, vasopressin, and dopamine, early infiltration (within 10 minutes) of phentolamine seems to be effective (22,23).
Even when extravasation is recognized, underestimation of the risk for subsequent tissue damage is common (see Cases 1 and 2). This often results in inadequate management (6,18,24). In extravasation injuries by vesicants, a wait-and-see policy is not appropriate. Obviously, there will be no controlled human studies on the relation between the amount of vesicant infused and the degree of tissue damage. All data available are from case reports and retrospective case series. In the face of the devastating consequences of tissue destruction, the flushout technique should be strongly considered. If the vesicant is not eliminated within the first 24 hours, advice from an experienced surgeon should be sought. In addition, to evaluate the extent of deep tissue damage, magnetic resonance imaging (T1- and T2-weighted images) is advised before surgery. Magnetic resonance imaging findings of severity of damage have been shown to be in good agreement with intraoperative findings (25).
We conclude that extravasation of a vesicant, whether a typical antineoplastic drug or hyperosmolar solutions, is a medical emergency that requires rapid action to minimize the extent of the damage. A wait-and-see policy is not justified. Flushing and drainage of the affected area, which can be performed at the bedside, are essential. Any approach that allows extravasated material to remain—whether diluted, cooled, or treated with an antidote—is less than ideal (3).
1. Yosowitz P, Ekland DA, Shaw RC, Parsons RW. Peripheral intravenous infiltration necrosis. Ann Surg 1975;182:553–6.
2. Upton J, Mulliken JB, Murray JE. Major intravenous extravasation injuries. Am J Surg 1979;137:497–506.
3. Gault DT. Extravasation injuries. Br J Plast Surg 1993;46:91–6.
4. Boyle DM, Engelking C. Vesicant extravasation: myths and realities. Oncol Nurs Forum 1995;22:57–67.
5. Beason R. Antineoplastic vesicant extravasation. J Intraven Nurs 1990;13:111–4.
6. Hastings-Tolsma M, Yucha CB. IV infiltration: no clear signs, no clear treatment? RN 1994;57:34–8 ; quiz 9.
7. Heckler FR. Current thoughts on extravasation injuries. Clin Plast Surg 1989;16:557–63.
8. Findlay JY. White veins after propofol. Anaesthesia 1994;49:838.
9. Raszka WV Jr, Kueser TK, Smith FR, Bass JW. The use of hyaluronidase in the treatment of intravenous extravasation injuries. J Perinatol 1990;10:146–9.
10. Kassner E. Evaluation and treatment of chemotherapy extravasation injuries. J Pediatr Oncol Nurs 2000;17:135–48.
11. Brown AS, Hoelzer DJ, Piercy SA. Skin necrosis from extravasation of intravenous fluids in children. Plast Reconstr Surg 1979;64:145–50.
12. Casanova D, Bardot J, Magalon G. Emergency treatment of accidental infusion leakage in the newborn: report of 14 cases. Br J Plast Surg 2001;54:396–9.
13. Brothers TE, Von Moll LK, Niederhuber JE, et al. Experience with subcutaneous infusion ports in three hundred patients. Surg Gynecol Obstet 1988;166:295–301.
14. Wallenborn J, Kuhnert I. Do position control methods for central venous catheters prevent complications? Hydromediastinum caused by an initially correctly placed tri-lumen subclavian catheter by using intra-atrial ECG recording: a case report. Anaesthesiol Reanim 2002;27:131–7.
15. Walker C, Jackson D, Dolan S. The potential for extravasation using a new five lumen catheter. Anaesthesia 1997;52:716–7.
16. Gaze NR. Tissue necrosis caused by commonly used intravenous infusions. Lancet 1978;2:417–9.
17. Edwards JJ, Samuels D, Fu ES. Forearm compartment syndrome from intravenous mannitol extravasation during general anesthesia. Anesth Analg 2003;96:245–6.
18. Dorr RT. Antidotes to vesicant chemotherapy extravasations. Blood Rev 1990;4:41–60.
19. O'Reilly C, McKay FM, Duffty P, Lloyd DJ. Glyceryl trinitrate in skin necrosis caused by extravasation of parenteral nutrition. Lancet 1988;2:565–6.
20. Alberts DS, Dorr RT. Case report: topical DMSO for mitomycin-C-induced skin ulceration. Oncol Nurs Forum 1991;18:693–5.
21. Mayo DJ. Fibrin sheath formation and chemotherapy extravasation: a case report. Support Care Cancer 1998;6:51–6.
22. Chang KA, Jawan B, Luk HN, et al. Bullous eruptions caused by extravasation of mannitol: a case report. Acta Anaesthesiol Sin 2001;39:195–8.
23. Bey D, El-Chaar GM, Bierman F, Valderrama E. The use of phentolamine in the prevention of dopamine-induced tissue extravasation. J Crit Care 1998;13:13–20.
24. Riley RH, Westhoff GP. Extravasation of propofol. Anaesth Intensive Care 1993;21:720–1.
25. Yama N, Tsuchida Y, Nuka S, et al. Usefulness of magnetic resonance imaging for surgical management of extravasation of an antitumor agent: a case report. Jpn J Clin Oncol 2001;31:122–4.
26. Loth TS. Minimal surgical debridement for the treatment of chemotherapeutic agent-induced skin extravasations. Cancer Treat Rep 1986;70:401–4.