- Email: [email protected]
Inadvertent Intra-Arterial Drug Injections in the Upper Extremity: Systematic Review
SCIENTIFIC ARTICLE
Inadvertent Intra-Arterial Drug Injections in the Upper Extremity: Systematic Review Chris Devulapalli, MD,* Kevin D. Han, MD,† Ricardo J. Bello, MD, MPH,‡ Dawn M. LaPorte, MD,† C. Tate Hepper, MD,† Ryan D. Katz, MD† Purpose To review the literature pertaining to inadvertent intra-arterial drug injection in the upper extremity, explore the various treatment options and their outcomes, and identify risk factors for limb amputation following intra-arterial injection. Methods A systematic review of Medline, EMBASE, and Cochrane databases (inception to March 2013) was completed for patients presenting with intra-arterial drug injection of the upper extremity. Details on intervention and outcome were extracted and subjected to pooled analysis with amputation as the primary outcome. Results A total of 25 articles (209 patients) were included for review. Mean patient age was 31 8 years (male, 71%; female, 29%). Prescription opioids (33%) were the most commonly injected substance, and the brachial artery (39%) was the most common site. The overall weighted mean amputation incidence was 29%. Anticoagulants were the most common treatment used (77%). From pooled analysis, conditions requiring antibiotic use were significantly associated with a higher incidence of amputation; whereas use of steroids was associated with a lower incidence of amputation. Patients presenting 14 hours or more after injection and those injecting crushed pills rather than pure substances had significantly higher incidences of amputation. Conclusions Intra-arterial drug injection of the upper extremity carries an amputation incidence of nearly 30%. Conditions requiring adjunctive antibiotic use and delay in receiving care were both significantly associated with higher incidences of amputation. No single treatment protocol to date has established superiority. (J Hand Surg Am. 2015;40(11):2262e2268. Copyright Ó 2015 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Therapeutic IV. Key words IVDU, hand ischemia, intra-arterial injections, drug abuse.
I
injection (IADI) of the upper extremity is a rare but potentially devastating phenomenon that can lead to tissue loss or necessitate an amputation. In the past, IADIs were most NADVERTENT INTRA-ARTERIAL DRUG
From the *Department of Surgery, MedStar Georgetown University Hospital, Washington, DC; the †Curtis National Hand Center, MedStar Union Memorial Hospital, Baltimore, MD; and the ‡Universidad Central de Venezuela, Caracas, Venezuela. Received for publication April 29, 2015; accepted in revised form August 3, 2015. No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Ryan D. Katz, MD, care of Anne Mattson, The Curtis National Hand Center, MedStar Union Memorial Hospital, 3333 N. Calvert St., 200, Baltimore, MD, 21218; e-mail: [email protected]. 0363-5023/15/4011-0022$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2015.08.002
2262
r
Ó 2015 ASSH
r
Published by Elsevier, Inc. All rights reserved.
often the result of iatrogenic complications during anesthesia procedures.1e10 A steady rise in recreational intravenous drug abuse in the last few decades has made drug users the patient population now primarily affected by this problem.10e15 The 2011 World Drug Report estimated 14 million intravenous drug users worldwide, most of whom reside in the United States.16 The clinical sequelae from IADI can include skin necrosis, infections, pseudoaneurysm, rhabdomyolysis, compartment syndrome, and distal ischemia resulting in limb loss. This presents a treatment challenge to hand surgeons who have very little guidance and even less consensus from the literature. Prior attempts at treating this problem have included medical and invasive therapies.17 Anticoagulants,18e22
IADI OF UPPER EXTREMITY
arterial vasodilators,19,20,23,24 and antiplatelet agents19,20 have all been tried with varying success. Angiography has been used for diagnosis and treatment with limited success.25 Regardless of the type of treatment provided, the amputation incidence IADI ranges from 20% to 100%.26 To gain insight into the diagnosis and management of IADI of upper extremities, we performed a systematic review of the literature. METHODS Literature search We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The Medline (PubMed), EMBASE, and Cochrane databases were searched from their inception to March 2013. Search criteria were formulated to identify all reports of IADIs in the upper extremities by using these search terms: intra-arterial injection, upper extremity, intravenous drug use, upper extremity intra-arterial injection, and upper extremity intravenous drug use. Study selection Two independent reviewers (C.D. and K.D.H) identified any potentially relevant articles from the list of titles. Studies were included that met the following criteria: primary study written in English, intravenous drug abusers of any age, and patients presenting with IADI to the upper extremity. Any studies reporting iatrogenic arterial drug injections, lower extremity arterial injections, intravenous drug injections, commentaries, reviews, and animal studies were excluded. Citations in the included articles were reviewed to identify any additional articles meeting inclusion requirements. Data extraction Once all relevant articles were selected, manuscript review and data extraction were performed to obtain the following information: patient demographic data, presentation characteristics, substance injected, site injected, diagnostic modalities, medical therapies, surgical interventions, and follow-up time. The main outcome assessed was amputation, defined as any surgical or autoamputation at the digital level or proximal. Statistical analysis A summary estimate of the overall mean amputation incidence from all studies (excluding single-patient case reports), weighted based on sample size, was calculated with 95% confidence intervals (CIs) and a forest plot was built (Appendix A; available on the Journal’s Web site at www.jhandsurg.org). To determine the effect of various treatment modalities on amputation incidence, J Hand Surg Am.
r
2263
data on intervention and amputation outcome were pooled, producing a study cohort from published case series. All studies describing pseudoaneurysm and studies missing information at the individual level were excluded. From this pooled study cohort, a univariate analysis was done to determine association between each treatment modality and amputation incidence. A multivariable regression model was fitted to estimate the independent effect for each treatment modality after adjusting for concurrent treatments. Given the scarcity of data points, only the 2 strongest confounders were adjusted for in each model. These confounders were identified by the change-in-estimates method following Mantel-Haenszel analysis.27 Using a similar pooled analysis, incidences of amputation were compared between injections of crushed pills and pure substances (ie, heroin, cocaine) and a critical time point (designated as the median delay in presentation). Significance was tested using chi-square test. Statistical significance was defined as P < .05. RESULTS Study selection The initial search identified 1,888 articles, which were screened to determine their eligibility. An additional 20 studies were identified through manual cross-referencing of citation lists. After applying the eligibility criteria, 25 studies containing 209 patients with IADI were included in the review (Fig. 1). The median year of publication was 1991, and publications ranged from years 1972 to 2011. Twenty-two articles dealt with patients with distal ischemia following IADI,11,17e19,21e26,28e39 2 articles reported on pseudoaneurysm patients,40,41 and 1 study reported on both.20 All 25 studies were retrospective case series or reports. Study and patient characteristics Detailed characteristics of included studies are listed in Appendix B (available on the Journal’s Web site at www.jhandsurg.org). Mean patient age was 31 8 years; 71% of the patients were male, and 29% were female. The median delay in presentation to the hospital from time of injection was 14 hours. The most commonly injected site was the brachial artery (40%), followed by the radial artery (24%). The most commonly injected substances were nonheroin prescription opioids (33%). Most patients presented with severe pain (78%) and/or cyanosis and mottling (46%). Motor impairment on presentation was reported in 37%, a sensory deficit in 34%, and temperature impairment in 26% of patients. Upon presentation to the hospital, 17% of patients already had some extent of gangrene of Vol. 40, November 2015
2264
IdenƟficaƟon
IADI OF UPPER EXTREMITY
Records idenƟfied through Medline (Pubmed), EMBASE and Cochrane databases searching: intra-arterial injecƟon AND (upper extremity OR intravenous drug use) OR upper extremity intra-arterial injecƟons OR upper extremity intravenous drug use (n = 1,888)
AddiƟonal records idenƟfied through manual cross-referencing (n=20)
Eligibility
Screening
Records aŌer duplicates removed (n = 1,828)
Records screened (n =1,828)
Full-text arƟcles assessed for eligibility (n=293)
Records excluded for nonEnglish language, imperƟnent topic, commentaries, or review format (n =1,535) Full-text arƟcles excluded due to lower extremity injecƟons or iatrogenic drug injecƟons (n =268)
Included
Studies included in qualitaƟve and quanƟtaƟve synthesis (n = 25)
FIGURE 1: Flow diagram depicts the search strategy for inclusion of articles in the systematic review; reported in accordance with the Preferred Reporting Systems for Systematic Reviews and Meta-Analysis statement (PRISMA).
their extremity. The overall mean amputation incidence, weighted for each studies sample size, was 29% (Appendix A; available on the Journal’s Web site at www.jhandsurg.org). Pseudoaneurysm Thirty-five patients in 3 studies20,40,41 presented with a pseudoaneurysm following IADI. The most common presenting symptoms were pulsatile mass (51%) and bleeding (29%). Duplex ultrasonography was used to confirm diagnosis in 29% of patients, and arteriography was used in 1 patient. All 35 patients received broadspectrum antibiotics. Culture positivity incidence was 63%, with Staphylococcus aureus being the most common isolate. Heparin and low-molecular-weight (LMW) dextran were used after surgery for thrombosis prophylaxis in 8 patients (23%).40 All patients underwent surgical intervention, with 80% receiving ligation of the J Hand Surg Am.
r
pseudoaneurysm alone, and 20% receiving ligation of pseudoaneurysm and reconstruction of the involved artery either primarily (17%) or with a reverse saphenous vein graft. No patients required amputations of the upper limb. The only complications reported on follow-up were neuromuscular deficits in 2 patients (6%). Distal ischemia Twenty-three studies reported on patients presenting with distal ischemia following IADI (n ¼ 174 patients).11,17e26,28e39 The pooled amputation incidence was 37% (n ¼ 63 of 174). Of these 63 amputations, 76% were digital amputations, 10% were midhand amputations, and 14% were forearm or more proximal amputations. The mean number of digits amputated was 1.58. Thirty-four fasciotomies (20%) were performed for compartment syndrome, 20 (12%) of which were hand fasciotomies, and 14 Vol. 40, November 2015
2265
IADI OF UPPER EXTREMITY
(8%) required both hand and forearm fasciotomies. Additional surgical interventions such as debridements, skin grafts, and open embolectomy were not included in our review because of vast inconsistencies in reporting of these outcomes. The following treatment modalities were employed: angiography (%), anticoagulation (77.0%), antiplatelet therapy (53%), vasodilator medications (25%), steroids (39%), antibiotics (29%), and local anesthetic field blocks (17%). Of the 174 patients, 21% received monotherapy, 76% received combination therapy of 2 or more treatments, and 3% received no treatment. Treatment time and dosage varied greatly among studies. The anticoagulants used most commonly were unfractionated heparin followed by LMW heparin. Low-molecular-weight dextran and aspirin were the most commonly used antiplatelet agents. Vasodilator medications used included reserpine, papaverine, tolazonline, iloprost, nifedipine, lidocaine, prostacyclins, and praxilene. Culture positivity rates were 50% for distal ischemia patients with methicillin-resistant S. aureus being the most common isolate. Anesthesia field blocks performed were brachial plexus anesthesia (60%), stellate ganglion blocks (17%), axillary blocks (17%), and cervical sympathectomy (7%). Sixteen patients (9%) were treated with chemical thrombolytic therapy with a resultant amputation incidence of 69% (11 of 16 patients), although this modality was not included in the regression model because only a small subset received the treatment. Multivariable regression analysis A logistic regression model determined the effect of each treatment modality on amputation (Appendix C; available on the Journal’s Web site at www.jhandsurg.org). A total of 150 patients with distal ischemia following IADI from 22 studies were included in the model.11,17e19,21e26,28e39 Patients requiring antibiotics were associated with increased incidences of amputation on univariate analysis (unadjusted odds ratio [OR], 3.09; 95% CI, 1.34e7.15; P ¼ .006). This association remained significant after adjustment for 2 strongest confounders (ie, steroids and anticoagulants) (adjusted OR, 2.64; CI, 1.05e6.69; P ¼ .040). Steroids were significantly associated with lower amputation incidence on univariate analysis (unadjusted OR, 0.47; CI, 0.23e0.98; P ¼ .038), but the association lost statistical significance after adjusting for confounders (adjusted OR, 0.53; CI, 0.22e1.24; P ¼ .142). There was no significant association between the remaining treatment modalities (angiography, anticoagulants, antiplatelets, vasodilators, and anesthesia blocks) and amputation J Hand Surg Am.
r
TABLE 1. Frequency of Substances Injected in Distal Ischemia Cases and Amputation Outcome Total (174 cases)
Cases Receiving Amputation
Percentage Receiving Amputation
Opiates
43
16
37
Barbiturates
32
10
31
Amphetamines
31
8
26
Benzodiazepines
28
14
50
Unknown
15
9
60
Multiple substances
8
3
38
Other
7
2
29
Heroin
6
0
0
Cocaine
4
1
25
Substance
incidence. Angiography was associated with higher amputation incidence approaching statistical significance on univariate analysis (unadjusted OR, 2.47; CI, 0.92e6.65; P ¼ .064) and multivariable analysis (adjusted OR, 2.70; CI, 0.87e8.36; P ¼ .084). Time of presentation Eighteen studies provided data on delay in presentation (n ¼ 102 patients).11,18,19,22e24,26,28,29,31e39 A critical threshold of 14 hours, the median time of delay, was selected for chi-square analysis. Of 50 patients presenting within 14 hours, 2 (4%) required amputation; whereas 24 (46%) of the 52 patients presenting 14 hours or later after IADI required amputation. This accounted for a 20 times higher odds of amputation for presentation after 14 hours (OR, 20.57; CI, 3.61e117.25; P < .001). Substance injected and effect on amputation Frequency of the various substances injected and their subsequent amputation rates for the 174 patients with distal ischemia following IADI are shown in Table 1. Apart from unknown cases, injection of benzodiazepines resulted in the highest amputation incidence of 50%. Chi-square analysis was performed to determine the effect of injecting crushed substances (ie, originally packaged in a pill or pure substance form) on amputation incidence. Unknown substances were omitted from this analysis. One amputation was required among 13 patients with intra-arterial pure substance injection; whereas 53 (36%) of the 146 intra-arterial crushed pill injections required amputation, with amputation being more than 6 times more likely with crushed substances (OR, 6.84; CI, 0.83e56.03; P ¼ .038). Vol. 40, November 2015
2266
IADI OF UPPER EXTREMITY
DISCUSSION The overall weighted mean amputation incidence was 29% for patients presenting with IADI. One reason for the persistently high morbidity incidence following IADI is the poorly understood pathophysiology of the injury. Several theories have been proposed but none are conclusive. A vasoconstrictive theory has been derived from a rabbit ear model and angiographic findings demonstrating potent vasoconstriction and vasospasm after IADIs.24,42,43 Arterial thrombosis has been repeatedly shown to be a key pathogenic step leading to ischemia in most IADIs.1,21,31,43 A crystal theory has also been put forth that describes the precipitation of crystals following IADI of various substances at higher pH levels than arterial blood. These crystalline irritants cause marked perfusion abnormalities and ischemia.1,3,32,44 Others have postulated that injected compounds cause direct endothelial trauma with resultant inflammation and venous outflow disruptions.7,22,43,45e48 Treatment modalities have been directed toward addressing these proposed causes of tissue injury. Because it has been generally accepted that thrombosis is a common end point following IADI, Sen et al8 recommended the initiation of anticoagulation in all cases (dosage guideline similar to treatment of that for pulmonary emboli) early in treatment barring any serious contraindications. Gaspar et al17 and Treiman et al22 have each published series (19 and 45 patients, respectively), describing their experiences treating IADI with a standardized protocol of dexamethasone, heparin, and LMW dextran. These 2 series had incidences of amputation of 26% and 24%, respectively. Stueber23 used a single-agent vasodilator (reserpine) in 14 patients with 1 requiring amputation. Even though anticoagulation was the most common treatment used (77% of the reviewed population), our analysis showed that anticoagulation conferred no benefit in lowering amputation incidence. Our review found only corticosteroids to be associated with a lower amputation incidence. Although steroids were part of many treatment protocols, the use of corticosteroids as sole therapy for intra-arterial injury is not known. All patients presenting with pseudoaneurysms were treated with antibiotics (likely because of the high culture positivity incidence), whereas antibiotic use in distal ischemia cases was more selective. Treiman et al22 gave antibiotics only to patients showing obvious signs of infection. Patients who did not receive antibiotics in Treiman et al’s study showed no signs of cellulitis, wet gangrene, or infection, nor did they have worse outcomes.22 Rautio and Keski-Nisula25 performed angiography in 7 patients with distal ischemia after IADI, J Hand Surg Am.
r
concluding it was helpful in finding predictive, ominous signs such as delayed or absent flow but had no effect on amputation outcome. In this review, patients who required antibiotics and those who had angiography had higher incidence of amputation than those who did not (OR, 2.64 and 2.70, respectively). A possible explanation for these findings is that these diagnostic and treatment modalities are reserved for those patients with more severe symptoms on presentation. Such patients may already have passed a critical threshold beyond salvage. From these findings, we recommend the selective use of both antibiotics and angiography: antibiotics in pseudoaneurysm patients and distal ischemia patients with obvious signs of infection, and angiography in patients with unclear diagnosis of intra-arterial injury and patients with clinical deterioration in spite of treatment. No significant association was found between vasodilators, antiplatelet drugs, anesthesia field blocks and amputation incidence. The low prevalence of these treatments, heterogeneity of treatment protocols between studies, small sample sizes, and study designs without adequate control groups all pose limitations to uncovering any true association between individual treatments and amputation outcome. Therefore, we are unable to recommend one specific treatment regimen with proven superiority for limb salvage following IADI. Our review found a strong association with delay in presentation past 14 hours and increased amputation incidence. Presentation within 14 hours resulted in only 2 of 50 amputations (4%), whereas presenting after 14 hours led to 24 of 52 amputations (46%). Similarly, Stueber23 found significantly worse clinical outcomes if treatment was delayed 16 or more hours after injury. These findings illustrate the likely presence of a relative period of reversible damage following IADI. After injection of diazepam into a rabbit ear model, histological changes examined on microscopic sections revealed an immediate vasoconstrictive insult within minutes, and then a temporary period of normal blood flow was observed before arterial thrombi and necrotic changes began 24 to 48 hours after initial insult.48,49 This may explain a period of relative reversibility, during which time treatment is more effective in preventing ischemia and necrosis. We also noted that injection of crushed pills led to a significantly higher incidence of amputation compared with pure drug substances (ie, heroin, cocaine). A plausible explanation is the damaging effect of preservatives, found in packaged pills, to the endothelial lining of arteries. Goldberg et al32 found that injection of pure codeine into femoral arteries of dogs was harmless, but addition of microcrystalline cellulose, a common preservative agent in oral tablets, produced immediate Vol. 40, November 2015
IADI OF UPPER EXTREMITY
ischemic symptoms and gangrene. The most commonly injected substances were prescription opioid tablets and other packaged pills. This review has limitations. Inadvertent intra-arterial drug injection is rare, and we found no high-quality (level IeII evidence) randomized, controlled studies with sufficient sample size for inclusion. Thus, this review’s findings are based on retrospective case series and reports. We are also aware that heterogeneity in patient populations between studies, differences in treatment protocols, and variation in reporting of outcomes create bias at the study and outcome level and ultimately diminish validity of pooled analysis. Despite these limitations, this review of the literature may help clinicians make decisions when treating this potentially devastating injury. The clinical course of patients with IADI, despite any intervention, is often quite predictable and ends with tissue loss. Specifically, these patients do not seem to benefit from the treatment we are most likely to recommend—anticoagulation. Furthermore, barring an angiogram showing a filling defect suggestive of embolism, the role of surgery is questionable for what is likely a “field effect” injury to the vascular endothelium over a distance. The risks of surgery (often with postoperative anticoagulation) are quite real and include bleeding, compartment syndrome, the need for blood products, donor site morbidity if bypass vessel harvested, and failure to alter the clinical course. Of this list, the last risk is worthy of further discussion for, if we surgeons are not able to preserve tissue through our intervention, we ultimately are violating our oath of “primum non nocere.” The best treatment for IADIs should be debated. REFERENCES 1. Brown SS, Lyons SM, Dundee JW. Intra-arterial barbiturates. A study of some factors leading to intravascular thrombosis. Br J Anaesth. 1968;40(1):13e19. 2. Burn JH. Why thiopentone injected into an artery may cause gangrene. Br Med J. 1960;2(5196):414e416. 3. Waters DJ. Intra-arterial thiopentone. A physico-chemical phenomenon. Anaesthesia. 1966;21(3):346e356. 4. Lundy JS. Clinical Anaesthesia. Philadelphia: WB Saunders, 1942;542. 5. Van der Post CWH. Report of a case of mistaken injection of pentothal sodium into aberrant artery. S Afr Med J. 1942;16:182. 6. Macintosh RR, Heyworth PSA. Intra-arterial injection of pentothal: a warning. Lancet. 1943;242:571. 7. Cohen SM. Accidental intra-arterial injection of drugs. Lancet. 1948;2(6524):409e417. 8. Sen S, Chini EN, Brown MJ. Complications after unintentional intraarterial injection of drugs: risks, outcomes, and management strategies. Mayo Clin Proc. 2005;80(6):783e795. 9. Gramenz P, Roberts D, Schrag L. Intra-arterial self-injection of methadone tablets into the femoral artery. J Emerg Med. 2010;39(3): e125ee127.
J Hand Surg Am.
r
2267
10. Righini M, Angellillo-Scherrer A, Gueddi S, Le Gal G, Bounameaux H. Management of severe ischemia of the hand following intra-arterial injection. Thromb Haemost. 2005;94(1):219e221. 11. Hamer R, Phelps D. Inadvertent intra-arterial injection of phentaramine: a complication of drug abuse. Ann Emerg Med. 1981;10(3): 148e150. 12. Bhabra MS, Meshikhes AN, Thomson GJ, Craig P, Parrott NR. Intraarterial temazepam: an important cause of limb ischaemia in intravenous drug abusers. Eur J Vasc Surg. 1994;8(2):240e242. 13. Hawkins LG, Lischer CG, Sweeney M. The main line accidental intra-arterial drug injection. Clin Orthop Relat Res. 1973;94: 268e274. 14. Charney MA, Stern PJ. Digital ischemia in clandestine intravenous drug users. J Hand Surg Am. 1991;16(2):308e310. 15. Bittner C, Zuber M, Eisner L. Acute ischemia of the hand in a drug addict after accidental intra-arterial injection [in German]. Swiss Surg. 2002;8(6):281e284. 16. United Nations Office on Drugs and Crime. World Drug Report 2011. United Nations Publication, Sales No. E.11.XI.10. Available at: http://www.unodc.org/documents/data-and-analysis/WDR2011/ World_Drug_Report_2011_ebook.pdf. Accessed April 28, 2015. 17. Gaspar MR, Hare RR. Gangrene due to intra-arterial injection of drugs by drug addicts. Surgery. 1972;72(4):573e577. 18. Borrero E. Treatment of “trash-hand” following intraarterial injection of drugs in addicts: cse studies. Vasc Endovascular Surg. 1995;29(1): 71e75. 19. Partanen TA, Vikatmaa P, Tukiainen E, Lepäntalo M, Vuola J. Outcome after injections of crushed tablets in intravenous drug abusers in the Helsinki University Central Hospital. Eur J Vasc Endovasc Surg. 2009;37(6):704e711. 20. Benitez PR, Newell MA. Vascular trauma in drug abuse: patterns of injury. Ann Vasc Surg. 1986;1(2):175e181. 21. Lindfors NC, Vilpponen L, Raatikainen T. Complications in the upper extremity following intra-arterial drug abuse. J Hand Surg Eur Vol. 2010;35(6):499e504. 22. Treiman GS, Yellin AE, Weaver FA, Barlow WE, Treiman RL, Gaspaee MR. An effective treatment protocol for intraarterial drug injection. J Vasc Surg. 1990;12(4):456e465. 23. Stueber K. The treatment of intraarterial pentazocine injection injuries with intraarterial reserpine. Ann Plast Surg. 1987;18(1):41e46. 24. Silverman SH, Turner WW Jr. Intraarterial drug abuse: new treatment options. J Vasc Surg. 1991;14(1):111e116. 25. Rautio R, Keski-Nisula L. Inadvertent intra-arterial drug injections: the role of angiographic and clinical findings. Acta Radiol. 2006;47(6):554e558. 26. Ferrante AMR, Cina G. Upper limb inschemia following intraarterial herion injection: a case report. Vasc Endovascular Surg. 1990;24(7): 519e523. 27. Greenland S. Modeling and variable selection in epidemiologic analysis. Am J Public Health. 1989;79(3):340e349. 28. Petrie PW, Lamb DW. Severe hand problems in drug addicts following self administered injections. Hand (N Y). 1973;5(2):130e134. 29. Cooper JC, Griffiths AB, Jones RB, Raffery AT. Accidental intraarterial injection in drug addicts. Eur J Vasc Surg. 1992;6(4): 430e433. 30. Begg EJ, McGrath MA, Wade DN. Inadvertent intra-arterial injection. A problem of drug abuse. Med J Aust. 1980;2(10):561e563. 31. Stover MC, Perrone J. Images in clinical medicine. Vascular occlusion after intraarterial cocaine injection. N Engl J Med. 2006;355(19): 2021. 32. Goldberg I, Bahar A, Yosipovitch Z. Gangrene of the upper extremity following intra-arterial injection of drugs. A case report and review of the literature. Clin Orthop Relat Res. 1984;188:223e229. 33. Seak CK, Kooi XJ, Seak CJ. Acute hand ischemia after intra-arterial injection of meprobamate powder. J Emerg Med. 2012;43(3):468e471. 34. Gouny P, Gaitz JP, Vayssairat M. Acute hand ischemia secondary to intraarterial buprenorphine injection: treatment with iloprost and dextran-40—a case report. Angiology. 1999;50(7):605e606.
Vol. 40, November 2015
2268
IADI OF UPPER EXTREMITY
35. Corser G, Masey S, Jacob G, Kernoff P, Browne D. Ischaemia following selfadministered intra-arterial injection of methylphenidate and diamorphine. A case report of treatment with intra-arterial urokinase and review. Anaesthesia. 1985;40(1):51e54. 36. Adir Y, Halpern P, Nachum Z, Bitterman H. Hyperbaric oxygen therapy for ischaemia of the hand due to intra-arterial injection of methadone and flunitrazepam. Eur J Vasc Surg. 1991;5(6): 677e679. 37. Funk L, Grover D, de Silva H. Compartment syndrome of the hand following intra-arterial injection of heroin. J Hand Surg Br. 1999;24(3):366e367. 38. Lloyd WK, Porter JM, Lindell TD, Rösch J, Dotter CT. Accidental intraarterial injection in drug abuse. Am J Roentgenol Radium Ther Nucl Med. 1973;117(4):892e895. 39. Blackwell SJ, Huang TT, Lewis SR. Intra-arterial drug abuse. Tex Med. 1978;74(3):64e68. 40. Behera A, Menakuru SR, Jindal R. Vascular complications of drug abuse: an Indian experience. ANZ J Surg. 2003;73(12):1004e1007. 41. Siu WT, Yau KK, Cheung HY, et al. Management of brachial artery pseudoaneurysms secondary to drug abuse. Ann Vasc Surg. 2005;19(5):657e661.
J Hand Surg Am.
r
42. Burn JH, Hobbs R. Mechanism of arterial spasm following intraarterial injection of thiopentone. Lancet. 1959;1(7083):1112e1115. 43. Kinmonth JB, Shepherd RC. Accidental injection of thiopentone into arteries: studies of pathology and treatment. Br Med J. 1959;2(5157): 914e918. 44. Jones NC. Inadvertent intra-arterial injection of drugs: Why does it still occur. J Intensive Care. 1995;51:66e68. 45. Engler HS, Freeman RA, Kanavage CB, et al. Production of gangrenous extremities by intra-arterial injections. Am Surg. 1964;30: 602e607. 46. Dodd TJ, Scott RN, Woodburn KR, et al. Limb ischaemia after intraarterial injection of Temazepam gel: histology of nine cases. J Clin Pathol. 1994;47(6):512e514. 47. Maxwell TM, Olcott C, Blaisdell FW. Vascular complications of drug abuse. Arch Surg. 1972;105(6):875e882. 48. Knill RL, Evans D. Pathogenesis of gangrene following intra-arterial injection of drugs: a new hypothesis. Can Anaesth Soc J. 1975;22(6): 637e646. 49. Crawford CR, Terranova WA. The role of intraarterial vasodilators in the treatment of inadvertant intraarterial injection injuries. Ann Plast Surg. 1990;25:279e282.
Vol. 40, November 2015
IADI OF UPPER EXTREMITY
2268.e1
APPENDIX A. Forest Plot Shows Amputation Rates of All Included Studies With Sample Size of 2 or More*
*The mean amputation incidence, weighted based on sample size of each individual study, is reported.
J Hand Surg Am.
r
Vol. 40, November 2015
2268.e2
APPENDIX B.
Included Study and Patient Characteristics, Treatment Modalities, and Amputation Outcome Delay in Presentation (h)
Amphetamine
DI
1
37
Opiate
DI
168
CR
29
Opiate þ amphetamine
DI
12
1
CR
20
Heroin
DI
1
Adir et al, 199136
1
CR
37
Opiate þ benzo
DI
Gouny et al, 199934
1
CR
27
Opiate
Stover & Perrone, 200631
1
CR
54
Seak et al, 201233
1
CR
Lloyd et al, 197338
2
Funk et al, 199937 Blackwell et al, 197839
Study Type
Hamer & Phelps, 198111
1
CR
26
Goldberg et al, 198432
1
CR
Corser et al, 198535
1
Ferrante & Cina, 199026
Author, y
Mean Age (y)
Substance Injected (n of Patients)
Angiography (n of Patients)
Pharmacotherapy (n of Patients Who Received) ASA, LMW dextran, tolazoline, hydrocortisone
Yes (1)
Anesthetic Blocks (n of Patients) Axillary
Abx
Amputation Rate (%) 0
100
J Hand Surg Am. r
Vol. 40, November 2015
Heparin, phentolamine, Axillary Abx
0
Heparin, lidocaine
0
288
ASA, LMW dextran, dypyridamole, papaverine, HBO
0
DI
8
LMW heparin, LMW dextran, iloprost, nifedipine
0
Cocaine
DI
96
23
Other
DI
16
CR
34.5
Barbiturate (1), opiate (1)
DI
110
2
CR
26
Heroin (2)
DI
4
3
CR
20
Benzo (1), opiate (1), other (1)
DI
14
Yes (1)
Yes (1)
Yes (2)
Heparin
100
Heparin, LMW dextran, nifedipine, dexamethasone, HBO
100
SGB (1) Axillary (1) Heparin (1), LMW dextran (2), papaverine/lidocaine (1), reserpine (1), Abx (2)
100
None
0
Axillary (1) Heparin (1), LMW dextran (1), papaverine (1), steroid (1), Abx (1)
0
(Continued)
IADI OF UPPER EXTREMITY
DI vs PSA
Number of Cases
APPENDIX B.
Included Study and Patient Characteristics, Treatment Modalities, and Amputation Outcome (Continued) Delay in Presentation (h)
Angiography (n of Patients)
Heroin (1), amphetamine (1), opiate (1)
DI
12
Yes (3)
30.3
benzo þ opiate (1), other (1)
DI
2
CR
22
Barbiturate (2), opiate (1), cocaine (1)
DI
5
CR
25
Barbiturate (3), amphetamine (2)
Borrero, 199518
5
CR
35.6
Opiate (5)
Siu et al, 200541
7
CS
48.1
Rautio & KeskiNisula, 200625
7
Retro
Behera et al, 200340
8
Stueber, 198723
Study Type
Mean Age (y)
Silverman & Turnerl, 1991
3
CR
24.7
Cooper et al, 199229
3
CR
Petrie & Lamb, 197328
4
Begg et al, 198030
Author, y
Substance Injected (n of Patients)
Pharmacotherapy (n of Patients Who Received)
Anesthetic Blocks (n of Patients)
Heparin, tolazoline
Amputation Rate (%) 0
J Hand Surg Am.
67
50
Abx (4)
50
DI
NR
Heparin (4), ASA (1), SGB (1), axillary (1), cervical (1) papaverine (4), nicotynl tartrate (1), steroid (1), Abx (5)
20
DI
18
Heparin, LMW Dextran (3), Abx (5)
60
Heroin þ benzo (5), heroin (2)
PSA
96
Yes (1)
Abx (7)
27
Benzo (4), opiate (1), opiate þ benzo (1), other (1)
DI
240
Yes (7)
LMW heparin (5), ASA (5), iloprost (3), nifedipine (1)
57
Retro
33
Opiate (8)
PSA
240
Heparin, LMW dextran, Abx (8)
0
14
CS
33
Opiate (14)
DI
21
Reserpine (11), dextran (2), none (1)
7
Partanen et al, 200919
18
Retro
26
Benzo(16),opiate(7), amphetamine (1)
DI
63
Gaspar & Hare, 197217
19
Retro
36
Barbiturate (18), opiate (1), other (1)
DI
12
r
Heparin (2), SGB (1), cervical (1) Prostacyclin (1), Praxilene (1), Abx (3)
Vol. 40, November 2015
Yes (3)
0
Daltaperin (11), lovenox (3), acetylsalycilate (3), clopidogrel (1), iloprost (2)
44
Heparin, dextran 40, dexamethasone (19)
26
2268.e3
(Continued)
IADI OF UPPER EXTREMITY
DI vs PSA
Number of Cases
2268.e4
APPENDIX B.
Included Study and Patient Characteristics, Treatment Modalities, and Amputation Outcome (Continued) DI vs PSA
Lindfors et al, 201021
24
Retro
31
Benzo (12), opiate (6), amphetamine (1), other (5)
Benitez & Newell, 198620
32
Retro
37
Opiate (15), DI (12), barbiturate (2), PSA heroin (1), multi (2) (20)
Treiman et al, 199022
45
Retro
32
Ritalin (26), barbiturate (8), opiate (7), heroin (2), other (5), multi (8)
DI
DI
Delay in Presentation (h) 324
Angiography (n of Patients) Yes (3)
NR
26.3
Yes (2)
Pharmacotherapy (n of Patients Who Received) LMW heparin (22), iloprost (2), Abx (19)
Anesthetic Blocks (n of Patients)
Amputation Rate (%)
Brachial plexus (18)
54
J Hand Surg Am.
Heparin, dextran, vasodilator (unspecified) (12), Abx (32)
25
Heparin, dextran 40, dexamethasone (45)
24
Abx, antibiotics; ASA, aspirin; benzo, benzodiazepine; cervical, cervical sympathectomy; CR, case report; CS, case series; DI, distal ischemia; HBO, hyperbaric oxygen; multi, multiple agents; PSA, pseudoaneurysm; Retro, retrospective review; NR, not reported; SGB, stellate ganglion block.
r
Vol. 40, November 2015
IADI OF UPPER EXTREMITY
Study Type
Author, y
Mean Age (y)
Substance Injected (n of Patients)
Number of Cases
2268.e5
IADI OF UPPER EXTREMITY
APPENDIX C. Univariate and Multivariable Logistic Regression Analysis for Treatment Modality and Amputation Outcome Univariate Analysis Treatment
Unadjusted OR*
Multivariable Analysis
95% CI
P Value
Adjusted OR*
95% CI
P Value
Angiography
2.47
0.92e6.65
.064
2.70
0.87e8.36
.084
Anticoagulants
1.14
0.52e2.52
.740
1.89
0.73e4.89
.190
Antiplatelets
0.68
0.34e1.36
.271
1.14
0.44e2.96
.787
Vasodilators
0.97
0.45e2.08
.942
0.43
0.16e1.15
.093
Steroids
0.47
0.23e0.98
.038
0.53
0.22e1.24
.142
Antibiotics
3.09
1.34e7.15
.006
2.64
1.05e6.69
.040
Blocks
2.09
0.63e6.92
.217
2.26
0.57e8.98
.248
*OR represents the odds of receiving amputation compared with all other patients not receiving each specified treatment.
J Hand Surg Am.
r
Vol. 40, November 2015
Inadvertent Intra-Arterial Drug Injections in the Upper Extremity: Systematic Review Chris Devulapalli, MD,* Kevin D. Han, MD,† Ricardo J. Bello, MD, MPH,‡ Dawn M. LaPorte, MD,† C. Tate Hepper, MD,† Ryan D. Katz, MD† Purpose To review the literature pertaining to inadvertent intra-arterial drug injection in the upper extremity, explore the various treatment options and their outcomes, and identify risk factors for limb amputation following intra-arterial injection. Methods A systematic review of Medline, EMBASE, and Cochrane databases (inception to March 2013) was completed for patients presenting with intra-arterial drug injection of the upper extremity. Details on intervention and outcome were extracted and subjected to pooled analysis with amputation as the primary outcome. Results A total of 25 articles (209 patients) were included for review. Mean patient age was 31 8 years (male, 71%; female, 29%). Prescription opioids (33%) were the most commonly injected substance, and the brachial artery (39%) was the most common site. The overall weighted mean amputation incidence was 29%. Anticoagulants were the most common treatment used (77%). From pooled analysis, conditions requiring antibiotic use were significantly associated with a higher incidence of amputation; whereas use of steroids was associated with a lower incidence of amputation. Patients presenting 14 hours or more after injection and those injecting crushed pills rather than pure substances had significantly higher incidences of amputation. Conclusions Intra-arterial drug injection of the upper extremity carries an amputation incidence of nearly 30%. Conditions requiring adjunctive antibiotic use and delay in receiving care were both significantly associated with higher incidences of amputation. No single treatment protocol to date has established superiority. (J Hand Surg Am. 2015;40(11):2262e2268. Copyright Ó 2015 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Therapeutic IV. Key words IVDU, hand ischemia, intra-arterial injections, drug abuse.
I
injection (IADI) of the upper extremity is a rare but potentially devastating phenomenon that can lead to tissue loss or necessitate an amputation. In the past, IADIs were most NADVERTENT INTRA-ARTERIAL DRUG
From the *Department of Surgery, MedStar Georgetown University Hospital, Washington, DC; the †Curtis National Hand Center, MedStar Union Memorial Hospital, Baltimore, MD; and the ‡Universidad Central de Venezuela, Caracas, Venezuela. Received for publication April 29, 2015; accepted in revised form August 3, 2015. No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Ryan D. Katz, MD, care of Anne Mattson, The Curtis National Hand Center, MedStar Union Memorial Hospital, 3333 N. Calvert St., 200, Baltimore, MD, 21218; e-mail: [email protected]. 0363-5023/15/4011-0022$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2015.08.002
2262
r
Ó 2015 ASSH
r
Published by Elsevier, Inc. All rights reserved.
often the result of iatrogenic complications during anesthesia procedures.1e10 A steady rise in recreational intravenous drug abuse in the last few decades has made drug users the patient population now primarily affected by this problem.10e15 The 2011 World Drug Report estimated 14 million intravenous drug users worldwide, most of whom reside in the United States.16 The clinical sequelae from IADI can include skin necrosis, infections, pseudoaneurysm, rhabdomyolysis, compartment syndrome, and distal ischemia resulting in limb loss. This presents a treatment challenge to hand surgeons who have very little guidance and even less consensus from the literature. Prior attempts at treating this problem have included medical and invasive therapies.17 Anticoagulants,18e22
IADI OF UPPER EXTREMITY
arterial vasodilators,19,20,23,24 and antiplatelet agents19,20 have all been tried with varying success. Angiography has been used for diagnosis and treatment with limited success.25 Regardless of the type of treatment provided, the amputation incidence IADI ranges from 20% to 100%.26 To gain insight into the diagnosis and management of IADI of upper extremities, we performed a systematic review of the literature. METHODS Literature search We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The Medline (PubMed), EMBASE, and Cochrane databases were searched from their inception to March 2013. Search criteria were formulated to identify all reports of IADIs in the upper extremities by using these search terms: intra-arterial injection, upper extremity, intravenous drug use, upper extremity intra-arterial injection, and upper extremity intravenous drug use. Study selection Two independent reviewers (C.D. and K.D.H) identified any potentially relevant articles from the list of titles. Studies were included that met the following criteria: primary study written in English, intravenous drug abusers of any age, and patients presenting with IADI to the upper extremity. Any studies reporting iatrogenic arterial drug injections, lower extremity arterial injections, intravenous drug injections, commentaries, reviews, and animal studies were excluded. Citations in the included articles were reviewed to identify any additional articles meeting inclusion requirements. Data extraction Once all relevant articles were selected, manuscript review and data extraction were performed to obtain the following information: patient demographic data, presentation characteristics, substance injected, site injected, diagnostic modalities, medical therapies, surgical interventions, and follow-up time. The main outcome assessed was amputation, defined as any surgical or autoamputation at the digital level or proximal. Statistical analysis A summary estimate of the overall mean amputation incidence from all studies (excluding single-patient case reports), weighted based on sample size, was calculated with 95% confidence intervals (CIs) and a forest plot was built (Appendix A; available on the Journal’s Web site at www.jhandsurg.org). To determine the effect of various treatment modalities on amputation incidence, J Hand Surg Am.
r
2263
data on intervention and amputation outcome were pooled, producing a study cohort from published case series. All studies describing pseudoaneurysm and studies missing information at the individual level were excluded. From this pooled study cohort, a univariate analysis was done to determine association between each treatment modality and amputation incidence. A multivariable regression model was fitted to estimate the independent effect for each treatment modality after adjusting for concurrent treatments. Given the scarcity of data points, only the 2 strongest confounders were adjusted for in each model. These confounders were identified by the change-in-estimates method following Mantel-Haenszel analysis.27 Using a similar pooled analysis, incidences of amputation were compared between injections of crushed pills and pure substances (ie, heroin, cocaine) and a critical time point (designated as the median delay in presentation). Significance was tested using chi-square test. Statistical significance was defined as P < .05. RESULTS Study selection The initial search identified 1,888 articles, which were screened to determine their eligibility. An additional 20 studies were identified through manual cross-referencing of citation lists. After applying the eligibility criteria, 25 studies containing 209 patients with IADI were included in the review (Fig. 1). The median year of publication was 1991, and publications ranged from years 1972 to 2011. Twenty-two articles dealt with patients with distal ischemia following IADI,11,17e19,21e26,28e39 2 articles reported on pseudoaneurysm patients,40,41 and 1 study reported on both.20 All 25 studies were retrospective case series or reports. Study and patient characteristics Detailed characteristics of included studies are listed in Appendix B (available on the Journal’s Web site at www.jhandsurg.org). Mean patient age was 31 8 years; 71% of the patients were male, and 29% were female. The median delay in presentation to the hospital from time of injection was 14 hours. The most commonly injected site was the brachial artery (40%), followed by the radial artery (24%). The most commonly injected substances were nonheroin prescription opioids (33%). Most patients presented with severe pain (78%) and/or cyanosis and mottling (46%). Motor impairment on presentation was reported in 37%, a sensory deficit in 34%, and temperature impairment in 26% of patients. Upon presentation to the hospital, 17% of patients already had some extent of gangrene of Vol. 40, November 2015
2264
IdenƟficaƟon
IADI OF UPPER EXTREMITY
Records idenƟfied through Medline (Pubmed), EMBASE and Cochrane databases searching: intra-arterial injecƟon AND (upper extremity OR intravenous drug use) OR upper extremity intra-arterial injecƟons OR upper extremity intravenous drug use (n = 1,888)
AddiƟonal records idenƟfied through manual cross-referencing (n=20)
Eligibility
Screening
Records aŌer duplicates removed (n = 1,828)
Records screened (n =1,828)
Full-text arƟcles assessed for eligibility (n=293)
Records excluded for nonEnglish language, imperƟnent topic, commentaries, or review format (n =1,535) Full-text arƟcles excluded due to lower extremity injecƟons or iatrogenic drug injecƟons (n =268)
Included
Studies included in qualitaƟve and quanƟtaƟve synthesis (n = 25)
FIGURE 1: Flow diagram depicts the search strategy for inclusion of articles in the systematic review; reported in accordance with the Preferred Reporting Systems for Systematic Reviews and Meta-Analysis statement (PRISMA).
their extremity. The overall mean amputation incidence, weighted for each studies sample size, was 29% (Appendix A; available on the Journal’s Web site at www.jhandsurg.org). Pseudoaneurysm Thirty-five patients in 3 studies20,40,41 presented with a pseudoaneurysm following IADI. The most common presenting symptoms were pulsatile mass (51%) and bleeding (29%). Duplex ultrasonography was used to confirm diagnosis in 29% of patients, and arteriography was used in 1 patient. All 35 patients received broadspectrum antibiotics. Culture positivity incidence was 63%, with Staphylococcus aureus being the most common isolate. Heparin and low-molecular-weight (LMW) dextran were used after surgery for thrombosis prophylaxis in 8 patients (23%).40 All patients underwent surgical intervention, with 80% receiving ligation of the J Hand Surg Am.
r
pseudoaneurysm alone, and 20% receiving ligation of pseudoaneurysm and reconstruction of the involved artery either primarily (17%) or with a reverse saphenous vein graft. No patients required amputations of the upper limb. The only complications reported on follow-up were neuromuscular deficits in 2 patients (6%). Distal ischemia Twenty-three studies reported on patients presenting with distal ischemia following IADI (n ¼ 174 patients).11,17e26,28e39 The pooled amputation incidence was 37% (n ¼ 63 of 174). Of these 63 amputations, 76% were digital amputations, 10% were midhand amputations, and 14% were forearm or more proximal amputations. The mean number of digits amputated was 1.58. Thirty-four fasciotomies (20%) were performed for compartment syndrome, 20 (12%) of which were hand fasciotomies, and 14 Vol. 40, November 2015
2265
IADI OF UPPER EXTREMITY
(8%) required both hand and forearm fasciotomies. Additional surgical interventions such as debridements, skin grafts, and open embolectomy were not included in our review because of vast inconsistencies in reporting of these outcomes. The following treatment modalities were employed: angiography (%), anticoagulation (77.0%), antiplatelet therapy (53%), vasodilator medications (25%), steroids (39%), antibiotics (29%), and local anesthetic field blocks (17%). Of the 174 patients, 21% received monotherapy, 76% received combination therapy of 2 or more treatments, and 3% received no treatment. Treatment time and dosage varied greatly among studies. The anticoagulants used most commonly were unfractionated heparin followed by LMW heparin. Low-molecular-weight dextran and aspirin were the most commonly used antiplatelet agents. Vasodilator medications used included reserpine, papaverine, tolazonline, iloprost, nifedipine, lidocaine, prostacyclins, and praxilene. Culture positivity rates were 50% for distal ischemia patients with methicillin-resistant S. aureus being the most common isolate. Anesthesia field blocks performed were brachial plexus anesthesia (60%), stellate ganglion blocks (17%), axillary blocks (17%), and cervical sympathectomy (7%). Sixteen patients (9%) were treated with chemical thrombolytic therapy with a resultant amputation incidence of 69% (11 of 16 patients), although this modality was not included in the regression model because only a small subset received the treatment. Multivariable regression analysis A logistic regression model determined the effect of each treatment modality on amputation (Appendix C; available on the Journal’s Web site at www.jhandsurg.org). A total of 150 patients with distal ischemia following IADI from 22 studies were included in the model.11,17e19,21e26,28e39 Patients requiring antibiotics were associated with increased incidences of amputation on univariate analysis (unadjusted odds ratio [OR], 3.09; 95% CI, 1.34e7.15; P ¼ .006). This association remained significant after adjustment for 2 strongest confounders (ie, steroids and anticoagulants) (adjusted OR, 2.64; CI, 1.05e6.69; P ¼ .040). Steroids were significantly associated with lower amputation incidence on univariate analysis (unadjusted OR, 0.47; CI, 0.23e0.98; P ¼ .038), but the association lost statistical significance after adjusting for confounders (adjusted OR, 0.53; CI, 0.22e1.24; P ¼ .142). There was no significant association between the remaining treatment modalities (angiography, anticoagulants, antiplatelets, vasodilators, and anesthesia blocks) and amputation J Hand Surg Am.
r
TABLE 1. Frequency of Substances Injected in Distal Ischemia Cases and Amputation Outcome Total (174 cases)
Cases Receiving Amputation
Percentage Receiving Amputation
Opiates
43
16
37
Barbiturates
32
10
31
Amphetamines
31
8
26
Benzodiazepines
28
14
50
Unknown
15
9
60
Multiple substances
8
3
38
Other
7
2
29
Heroin
6
0
0
Cocaine
4
1
25
Substance
incidence. Angiography was associated with higher amputation incidence approaching statistical significance on univariate analysis (unadjusted OR, 2.47; CI, 0.92e6.65; P ¼ .064) and multivariable analysis (adjusted OR, 2.70; CI, 0.87e8.36; P ¼ .084). Time of presentation Eighteen studies provided data on delay in presentation (n ¼ 102 patients).11,18,19,22e24,26,28,29,31e39 A critical threshold of 14 hours, the median time of delay, was selected for chi-square analysis. Of 50 patients presenting within 14 hours, 2 (4%) required amputation; whereas 24 (46%) of the 52 patients presenting 14 hours or later after IADI required amputation. This accounted for a 20 times higher odds of amputation for presentation after 14 hours (OR, 20.57; CI, 3.61e117.25; P < .001). Substance injected and effect on amputation Frequency of the various substances injected and their subsequent amputation rates for the 174 patients with distal ischemia following IADI are shown in Table 1. Apart from unknown cases, injection of benzodiazepines resulted in the highest amputation incidence of 50%. Chi-square analysis was performed to determine the effect of injecting crushed substances (ie, originally packaged in a pill or pure substance form) on amputation incidence. Unknown substances were omitted from this analysis. One amputation was required among 13 patients with intra-arterial pure substance injection; whereas 53 (36%) of the 146 intra-arterial crushed pill injections required amputation, with amputation being more than 6 times more likely with crushed substances (OR, 6.84; CI, 0.83e56.03; P ¼ .038). Vol. 40, November 2015
2266
IADI OF UPPER EXTREMITY
DISCUSSION The overall weighted mean amputation incidence was 29% for patients presenting with IADI. One reason for the persistently high morbidity incidence following IADI is the poorly understood pathophysiology of the injury. Several theories have been proposed but none are conclusive. A vasoconstrictive theory has been derived from a rabbit ear model and angiographic findings demonstrating potent vasoconstriction and vasospasm after IADIs.24,42,43 Arterial thrombosis has been repeatedly shown to be a key pathogenic step leading to ischemia in most IADIs.1,21,31,43 A crystal theory has also been put forth that describes the precipitation of crystals following IADI of various substances at higher pH levels than arterial blood. These crystalline irritants cause marked perfusion abnormalities and ischemia.1,3,32,44 Others have postulated that injected compounds cause direct endothelial trauma with resultant inflammation and venous outflow disruptions.7,22,43,45e48 Treatment modalities have been directed toward addressing these proposed causes of tissue injury. Because it has been generally accepted that thrombosis is a common end point following IADI, Sen et al8 recommended the initiation of anticoagulation in all cases (dosage guideline similar to treatment of that for pulmonary emboli) early in treatment barring any serious contraindications. Gaspar et al17 and Treiman et al22 have each published series (19 and 45 patients, respectively), describing their experiences treating IADI with a standardized protocol of dexamethasone, heparin, and LMW dextran. These 2 series had incidences of amputation of 26% and 24%, respectively. Stueber23 used a single-agent vasodilator (reserpine) in 14 patients with 1 requiring amputation. Even though anticoagulation was the most common treatment used (77% of the reviewed population), our analysis showed that anticoagulation conferred no benefit in lowering amputation incidence. Our review found only corticosteroids to be associated with a lower amputation incidence. Although steroids were part of many treatment protocols, the use of corticosteroids as sole therapy for intra-arterial injury is not known. All patients presenting with pseudoaneurysms were treated with antibiotics (likely because of the high culture positivity incidence), whereas antibiotic use in distal ischemia cases was more selective. Treiman et al22 gave antibiotics only to patients showing obvious signs of infection. Patients who did not receive antibiotics in Treiman et al’s study showed no signs of cellulitis, wet gangrene, or infection, nor did they have worse outcomes.22 Rautio and Keski-Nisula25 performed angiography in 7 patients with distal ischemia after IADI, J Hand Surg Am.
r
concluding it was helpful in finding predictive, ominous signs such as delayed or absent flow but had no effect on amputation outcome. In this review, patients who required antibiotics and those who had angiography had higher incidence of amputation than those who did not (OR, 2.64 and 2.70, respectively). A possible explanation for these findings is that these diagnostic and treatment modalities are reserved for those patients with more severe symptoms on presentation. Such patients may already have passed a critical threshold beyond salvage. From these findings, we recommend the selective use of both antibiotics and angiography: antibiotics in pseudoaneurysm patients and distal ischemia patients with obvious signs of infection, and angiography in patients with unclear diagnosis of intra-arterial injury and patients with clinical deterioration in spite of treatment. No significant association was found between vasodilators, antiplatelet drugs, anesthesia field blocks and amputation incidence. The low prevalence of these treatments, heterogeneity of treatment protocols between studies, small sample sizes, and study designs without adequate control groups all pose limitations to uncovering any true association between individual treatments and amputation outcome. Therefore, we are unable to recommend one specific treatment regimen with proven superiority for limb salvage following IADI. Our review found a strong association with delay in presentation past 14 hours and increased amputation incidence. Presentation within 14 hours resulted in only 2 of 50 amputations (4%), whereas presenting after 14 hours led to 24 of 52 amputations (46%). Similarly, Stueber23 found significantly worse clinical outcomes if treatment was delayed 16 or more hours after injury. These findings illustrate the likely presence of a relative period of reversible damage following IADI. After injection of diazepam into a rabbit ear model, histological changes examined on microscopic sections revealed an immediate vasoconstrictive insult within minutes, and then a temporary period of normal blood flow was observed before arterial thrombi and necrotic changes began 24 to 48 hours after initial insult.48,49 This may explain a period of relative reversibility, during which time treatment is more effective in preventing ischemia and necrosis. We also noted that injection of crushed pills led to a significantly higher incidence of amputation compared with pure drug substances (ie, heroin, cocaine). A plausible explanation is the damaging effect of preservatives, found in packaged pills, to the endothelial lining of arteries. Goldberg et al32 found that injection of pure codeine into femoral arteries of dogs was harmless, but addition of microcrystalline cellulose, a common preservative agent in oral tablets, produced immediate Vol. 40, November 2015
IADI OF UPPER EXTREMITY
ischemic symptoms and gangrene. The most commonly injected substances were prescription opioid tablets and other packaged pills. This review has limitations. Inadvertent intra-arterial drug injection is rare, and we found no high-quality (level IeII evidence) randomized, controlled studies with sufficient sample size for inclusion. Thus, this review’s findings are based on retrospective case series and reports. We are also aware that heterogeneity in patient populations between studies, differences in treatment protocols, and variation in reporting of outcomes create bias at the study and outcome level and ultimately diminish validity of pooled analysis. Despite these limitations, this review of the literature may help clinicians make decisions when treating this potentially devastating injury. The clinical course of patients with IADI, despite any intervention, is often quite predictable and ends with tissue loss. Specifically, these patients do not seem to benefit from the treatment we are most likely to recommend—anticoagulation. Furthermore, barring an angiogram showing a filling defect suggestive of embolism, the role of surgery is questionable for what is likely a “field effect” injury to the vascular endothelium over a distance. The risks of surgery (often with postoperative anticoagulation) are quite real and include bleeding, compartment syndrome, the need for blood products, donor site morbidity if bypass vessel harvested, and failure to alter the clinical course. Of this list, the last risk is worthy of further discussion for, if we surgeons are not able to preserve tissue through our intervention, we ultimately are violating our oath of “primum non nocere.” The best treatment for IADIs should be debated. REFERENCES 1. Brown SS, Lyons SM, Dundee JW. Intra-arterial barbiturates. A study of some factors leading to intravascular thrombosis. Br J Anaesth. 1968;40(1):13e19. 2. Burn JH. Why thiopentone injected into an artery may cause gangrene. Br Med J. 1960;2(5196):414e416. 3. Waters DJ. Intra-arterial thiopentone. A physico-chemical phenomenon. Anaesthesia. 1966;21(3):346e356. 4. Lundy JS. Clinical Anaesthesia. Philadelphia: WB Saunders, 1942;542. 5. Van der Post CWH. Report of a case of mistaken injection of pentothal sodium into aberrant artery. S Afr Med J. 1942;16:182. 6. Macintosh RR, Heyworth PSA. Intra-arterial injection of pentothal: a warning. Lancet. 1943;242:571. 7. Cohen SM. Accidental intra-arterial injection of drugs. Lancet. 1948;2(6524):409e417. 8. Sen S, Chini EN, Brown MJ. Complications after unintentional intraarterial injection of drugs: risks, outcomes, and management strategies. Mayo Clin Proc. 2005;80(6):783e795. 9. Gramenz P, Roberts D, Schrag L. Intra-arterial self-injection of methadone tablets into the femoral artery. J Emerg Med. 2010;39(3): e125ee127.
J Hand Surg Am.
r
2267
10. Righini M, Angellillo-Scherrer A, Gueddi S, Le Gal G, Bounameaux H. Management of severe ischemia of the hand following intra-arterial injection. Thromb Haemost. 2005;94(1):219e221. 11. Hamer R, Phelps D. Inadvertent intra-arterial injection of phentaramine: a complication of drug abuse. Ann Emerg Med. 1981;10(3): 148e150. 12. Bhabra MS, Meshikhes AN, Thomson GJ, Craig P, Parrott NR. Intraarterial temazepam: an important cause of limb ischaemia in intravenous drug abusers. Eur J Vasc Surg. 1994;8(2):240e242. 13. Hawkins LG, Lischer CG, Sweeney M. The main line accidental intra-arterial drug injection. Clin Orthop Relat Res. 1973;94: 268e274. 14. Charney MA, Stern PJ. Digital ischemia in clandestine intravenous drug users. J Hand Surg Am. 1991;16(2):308e310. 15. Bittner C, Zuber M, Eisner L. Acute ischemia of the hand in a drug addict after accidental intra-arterial injection [in German]. Swiss Surg. 2002;8(6):281e284. 16. United Nations Office on Drugs and Crime. World Drug Report 2011. United Nations Publication, Sales No. E.11.XI.10. Available at: http://www.unodc.org/documents/data-and-analysis/WDR2011/ World_Drug_Report_2011_ebook.pdf. Accessed April 28, 2015. 17. Gaspar MR, Hare RR. Gangrene due to intra-arterial injection of drugs by drug addicts. Surgery. 1972;72(4):573e577. 18. Borrero E. Treatment of “trash-hand” following intraarterial injection of drugs in addicts: cse studies. Vasc Endovascular Surg. 1995;29(1): 71e75. 19. Partanen TA, Vikatmaa P, Tukiainen E, Lepäntalo M, Vuola J. Outcome after injections of crushed tablets in intravenous drug abusers in the Helsinki University Central Hospital. Eur J Vasc Endovasc Surg. 2009;37(6):704e711. 20. Benitez PR, Newell MA. Vascular trauma in drug abuse: patterns of injury. Ann Vasc Surg. 1986;1(2):175e181. 21. Lindfors NC, Vilpponen L, Raatikainen T. Complications in the upper extremity following intra-arterial drug abuse. J Hand Surg Eur Vol. 2010;35(6):499e504. 22. Treiman GS, Yellin AE, Weaver FA, Barlow WE, Treiman RL, Gaspaee MR. An effective treatment protocol for intraarterial drug injection. J Vasc Surg. 1990;12(4):456e465. 23. Stueber K. The treatment of intraarterial pentazocine injection injuries with intraarterial reserpine. Ann Plast Surg. 1987;18(1):41e46. 24. Silverman SH, Turner WW Jr. Intraarterial drug abuse: new treatment options. J Vasc Surg. 1991;14(1):111e116. 25. Rautio R, Keski-Nisula L. Inadvertent intra-arterial drug injections: the role of angiographic and clinical findings. Acta Radiol. 2006;47(6):554e558. 26. Ferrante AMR, Cina G. Upper limb inschemia following intraarterial herion injection: a case report. Vasc Endovascular Surg. 1990;24(7): 519e523. 27. Greenland S. Modeling and variable selection in epidemiologic analysis. Am J Public Health. 1989;79(3):340e349. 28. Petrie PW, Lamb DW. Severe hand problems in drug addicts following self administered injections. Hand (N Y). 1973;5(2):130e134. 29. Cooper JC, Griffiths AB, Jones RB, Raffery AT. Accidental intraarterial injection in drug addicts. Eur J Vasc Surg. 1992;6(4): 430e433. 30. Begg EJ, McGrath MA, Wade DN. Inadvertent intra-arterial injection. A problem of drug abuse. Med J Aust. 1980;2(10):561e563. 31. Stover MC, Perrone J. Images in clinical medicine. Vascular occlusion after intraarterial cocaine injection. N Engl J Med. 2006;355(19): 2021. 32. Goldberg I, Bahar A, Yosipovitch Z. Gangrene of the upper extremity following intra-arterial injection of drugs. A case report and review of the literature. Clin Orthop Relat Res. 1984;188:223e229. 33. Seak CK, Kooi XJ, Seak CJ. Acute hand ischemia after intra-arterial injection of meprobamate powder. J Emerg Med. 2012;43(3):468e471. 34. Gouny P, Gaitz JP, Vayssairat M. Acute hand ischemia secondary to intraarterial buprenorphine injection: treatment with iloprost and dextran-40—a case report. Angiology. 1999;50(7):605e606.
Vol. 40, November 2015
2268
IADI OF UPPER EXTREMITY
35. Corser G, Masey S, Jacob G, Kernoff P, Browne D. Ischaemia following selfadministered intra-arterial injection of methylphenidate and diamorphine. A case report of treatment with intra-arterial urokinase and review. Anaesthesia. 1985;40(1):51e54. 36. Adir Y, Halpern P, Nachum Z, Bitterman H. Hyperbaric oxygen therapy for ischaemia of the hand due to intra-arterial injection of methadone and flunitrazepam. Eur J Vasc Surg. 1991;5(6): 677e679. 37. Funk L, Grover D, de Silva H. Compartment syndrome of the hand following intra-arterial injection of heroin. J Hand Surg Br. 1999;24(3):366e367. 38. Lloyd WK, Porter JM, Lindell TD, Rösch J, Dotter CT. Accidental intraarterial injection in drug abuse. Am J Roentgenol Radium Ther Nucl Med. 1973;117(4):892e895. 39. Blackwell SJ, Huang TT, Lewis SR. Intra-arterial drug abuse. Tex Med. 1978;74(3):64e68. 40. Behera A, Menakuru SR, Jindal R. Vascular complications of drug abuse: an Indian experience. ANZ J Surg. 2003;73(12):1004e1007. 41. Siu WT, Yau KK, Cheung HY, et al. Management of brachial artery pseudoaneurysms secondary to drug abuse. Ann Vasc Surg. 2005;19(5):657e661.
J Hand Surg Am.
r
42. Burn JH, Hobbs R. Mechanism of arterial spasm following intraarterial injection of thiopentone. Lancet. 1959;1(7083):1112e1115. 43. Kinmonth JB, Shepherd RC. Accidental injection of thiopentone into arteries: studies of pathology and treatment. Br Med J. 1959;2(5157): 914e918. 44. Jones NC. Inadvertent intra-arterial injection of drugs: Why does it still occur. J Intensive Care. 1995;51:66e68. 45. Engler HS, Freeman RA, Kanavage CB, et al. Production of gangrenous extremities by intra-arterial injections. Am Surg. 1964;30: 602e607. 46. Dodd TJ, Scott RN, Woodburn KR, et al. Limb ischaemia after intraarterial injection of Temazepam gel: histology of nine cases. J Clin Pathol. 1994;47(6):512e514. 47. Maxwell TM, Olcott C, Blaisdell FW. Vascular complications of drug abuse. Arch Surg. 1972;105(6):875e882. 48. Knill RL, Evans D. Pathogenesis of gangrene following intra-arterial injection of drugs: a new hypothesis. Can Anaesth Soc J. 1975;22(6): 637e646. 49. Crawford CR, Terranova WA. The role of intraarterial vasodilators in the treatment of inadvertant intraarterial injection injuries. Ann Plast Surg. 1990;25:279e282.
Vol. 40, November 2015
IADI OF UPPER EXTREMITY
2268.e1
APPENDIX A. Forest Plot Shows Amputation Rates of All Included Studies With Sample Size of 2 or More*
*The mean amputation incidence, weighted based on sample size of each individual study, is reported.
J Hand Surg Am.
r
Vol. 40, November 2015
2268.e2
APPENDIX B.
Included Study and Patient Characteristics, Treatment Modalities, and Amputation Outcome Delay in Presentation (h)
Amphetamine
DI
1
37
Opiate
DI
168
CR
29
Opiate þ amphetamine
DI
12
1
CR
20
Heroin
DI
1
Adir et al, 199136
1
CR
37
Opiate þ benzo
DI
Gouny et al, 199934
1
CR
27
Opiate
Stover & Perrone, 200631
1
CR
54
Seak et al, 201233
1
CR
Lloyd et al, 197338
2
Funk et al, 199937 Blackwell et al, 197839
Study Type
Hamer & Phelps, 198111
1
CR
26
Goldberg et al, 198432
1
CR
Corser et al, 198535
1
Ferrante & Cina, 199026
Author, y
Mean Age (y)
Substance Injected (n of Patients)
Angiography (n of Patients)
Pharmacotherapy (n of Patients Who Received) ASA, LMW dextran, tolazoline, hydrocortisone
Yes (1)
Anesthetic Blocks (n of Patients) Axillary
Abx
Amputation Rate (%) 0
100
J Hand Surg Am. r
Vol. 40, November 2015
Heparin, phentolamine, Axillary Abx
0
Heparin, lidocaine
0
288
ASA, LMW dextran, dypyridamole, papaverine, HBO
0
DI
8
LMW heparin, LMW dextran, iloprost, nifedipine
0
Cocaine
DI
96
23
Other
DI
16
CR
34.5
Barbiturate (1), opiate (1)
DI
110
2
CR
26
Heroin (2)
DI
4
3
CR
20
Benzo (1), opiate (1), other (1)
DI
14
Yes (1)
Yes (1)
Yes (2)
Heparin
100
Heparin, LMW dextran, nifedipine, dexamethasone, HBO
100
SGB (1) Axillary (1) Heparin (1), LMW dextran (2), papaverine/lidocaine (1), reserpine (1), Abx (2)
100
None
0
Axillary (1) Heparin (1), LMW dextran (1), papaverine (1), steroid (1), Abx (1)
0
(Continued)
IADI OF UPPER EXTREMITY
DI vs PSA
Number of Cases
APPENDIX B.
Included Study and Patient Characteristics, Treatment Modalities, and Amputation Outcome (Continued) Delay in Presentation (h)
Angiography (n of Patients)
Heroin (1), amphetamine (1), opiate (1)
DI
12
Yes (3)
30.3
benzo þ opiate (1), other (1)
DI
2
CR
22
Barbiturate (2), opiate (1), cocaine (1)
DI
5
CR
25
Barbiturate (3), amphetamine (2)
Borrero, 199518
5
CR
35.6
Opiate (5)
Siu et al, 200541
7
CS
48.1
Rautio & KeskiNisula, 200625
7
Retro
Behera et al, 200340
8
Stueber, 198723
Study Type
Mean Age (y)
Silverman & Turnerl, 1991
3
CR
24.7
Cooper et al, 199229
3
CR
Petrie & Lamb, 197328
4
Begg et al, 198030
Author, y
Substance Injected (n of Patients)
Pharmacotherapy (n of Patients Who Received)
Anesthetic Blocks (n of Patients)
Heparin, tolazoline
Amputation Rate (%) 0
J Hand Surg Am.
67
50
Abx (4)
50
DI
NR
Heparin (4), ASA (1), SGB (1), axillary (1), cervical (1) papaverine (4), nicotynl tartrate (1), steroid (1), Abx (5)
20
DI
18
Heparin, LMW Dextran (3), Abx (5)
60
Heroin þ benzo (5), heroin (2)
PSA
96
Yes (1)
Abx (7)
27
Benzo (4), opiate (1), opiate þ benzo (1), other (1)
DI
240
Yes (7)
LMW heparin (5), ASA (5), iloprost (3), nifedipine (1)
57
Retro
33
Opiate (8)
PSA
240
Heparin, LMW dextran, Abx (8)
0
14
CS
33
Opiate (14)
DI
21
Reserpine (11), dextran (2), none (1)
7
Partanen et al, 200919
18
Retro
26
Benzo(16),opiate(7), amphetamine (1)
DI
63
Gaspar & Hare, 197217
19
Retro
36
Barbiturate (18), opiate (1), other (1)
DI
12
r
Heparin (2), SGB (1), cervical (1) Prostacyclin (1), Praxilene (1), Abx (3)
Vol. 40, November 2015
Yes (3)
0
Daltaperin (11), lovenox (3), acetylsalycilate (3), clopidogrel (1), iloprost (2)
44
Heparin, dextran 40, dexamethasone (19)
26
2268.e3
(Continued)
IADI OF UPPER EXTREMITY
DI vs PSA
Number of Cases
2268.e4
APPENDIX B.
Included Study and Patient Characteristics, Treatment Modalities, and Amputation Outcome (Continued) DI vs PSA
Lindfors et al, 201021
24
Retro
31
Benzo (12), opiate (6), amphetamine (1), other (5)
Benitez & Newell, 198620
32
Retro
37
Opiate (15), DI (12), barbiturate (2), PSA heroin (1), multi (2) (20)
Treiman et al, 199022
45
Retro
32
Ritalin (26), barbiturate (8), opiate (7), heroin (2), other (5), multi (8)
DI
DI
Delay in Presentation (h) 324
Angiography (n of Patients) Yes (3)
NR
26.3
Yes (2)
Pharmacotherapy (n of Patients Who Received) LMW heparin (22), iloprost (2), Abx (19)
Anesthetic Blocks (n of Patients)
Amputation Rate (%)
Brachial plexus (18)
54
J Hand Surg Am.
Heparin, dextran, vasodilator (unspecified) (12), Abx (32)
25
Heparin, dextran 40, dexamethasone (45)
24
Abx, antibiotics; ASA, aspirin; benzo, benzodiazepine; cervical, cervical sympathectomy; CR, case report; CS, case series; DI, distal ischemia; HBO, hyperbaric oxygen; multi, multiple agents; PSA, pseudoaneurysm; Retro, retrospective review; NR, not reported; SGB, stellate ganglion block.
r
Vol. 40, November 2015
IADI OF UPPER EXTREMITY
Study Type
Author, y
Mean Age (y)
Substance Injected (n of Patients)
Number of Cases
2268.e5
IADI OF UPPER EXTREMITY
APPENDIX C. Univariate and Multivariable Logistic Regression Analysis for Treatment Modality and Amputation Outcome Univariate Analysis Treatment
Unadjusted OR*
Multivariable Analysis
95% CI
P Value
Adjusted OR*
95% CI
P Value
Angiography
2.47
0.92e6.65
.064
2.70
0.87e8.36
.084
Anticoagulants
1.14
0.52e2.52
.740
1.89
0.73e4.89
.190
Antiplatelets
0.68
0.34e1.36
.271
1.14
0.44e2.96
.787
Vasodilators
0.97
0.45e2.08
.942
0.43
0.16e1.15
.093
Steroids
0.47
0.23e0.98
.038
0.53
0.22e1.24
.142
Antibiotics
3.09
1.34e7.15
.006
2.64
1.05e6.69
.040
Blocks
2.09
0.63e6.92
.217
2.26
0.57e8.98
.248
*OR represents the odds of receiving amputation compared with all other patients not receiving each specified treatment.
J Hand Surg Am.
r
Vol. 40, November 2015