Acute arterial thrombosis associated with total knee arthroplasty

Transcription

1 Acute arterial thrombosis associated with total knee arthroplasty Keith D. Calligaro, MD, Dominic A. DeLaurentis, MD, Robert E. Booth, MD, Richard H. Rothman, MD, Ronald P. Savarese, MD, and Matthew J. Dougherty, MD, Philadelphia, Pa. Purpose: Acute arterial thrombosis associated with total knee arthroplasty (TKA) is a rare but limb-threatening complication. The purpose of this report was to determine the incidence and optimal management of these complications by reviewing our extensive orthopedic experience and the English-language literature. Methods: Between April 1989 and March 1994 seven (0.17%) patients had development of acute limb-threatening ischemia after 4097 TKAs that were performed at our hospital. Management of these complications included (1) emergency arteriography to define inflow and outflow arteries, (2) use of autologous vein from the contralateral leg when arterial bypasses were necessary (because TKAs are associated with a high incidence of deep vein thrombosis), and (3) early, aggressive revascularization that often required difficult distal bypasses to achieve limb salvage. Management of our cases are compared with treatment of 13 patients described in the literature. Results: Ten patients treated at other hospitals by arterial thrombectomy alone (six cases), sympathectomy alone (two cases), fasciotomy alone (one case) or delayed arterial bypass resulted in seven major amputations and one death. All seven of our patients and three patients treated elsewhere underwent emergency femorodistal bypasses (six tibial, three below-knee popliteal, one pedal). All 10 patients had limb salvage after long-term follow-up (average 18 months; range 1 to 58). Conclusion: Thrombectomy alone for acute arterial thrombosis associated with TKA generally is unsuccessful and associated with unacceptably high amputation rates. Dismal results without emergency bypass is due to underlying chronic occlusive atherosclerotic disease found in these patients and intimal plaque disruption that can occur with knee manipulation or tourniquet compression. Acute arterial occlusion after TKA is best managed by emergency arteriography and a femoroinfrageniculate bypass. (J VASC SURG 1994;20: ) In a previous report by DeLaurentis et al., 1 a prospective study of patients who underwent total knee arthroplasty (TKA) documented an unexpectedly low incidence (2%) of chronic arterial insufficiency in this elderly population and identified risk factors for acute arterial ischemic complications. 1 Associated risk factors for development of these complications included history of lower extremity From the Sections of Vascular Surgery and Orthopedic Surgery (Drs. Booth and Rothman), Pennsylvania Hospital/Thomas Jefferson Medical College, Philadelphia. Supported by a grant from the John F. Connelly Foundation. Presented at the Eighth Annual Meeting of the Eastern Vascular Society, Montrdal, Qudbec, Canada, May 12-15, Reprint requests: Keith D. Calligaro, MD, 700 Spruce St., Suite 101, Philadelphia, PA Copyright 1994 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter /94/$ /6/59522 arterial insufficiency, absent pedal pulses, and calcification of the superficial femoral or popliteal artery.1 The purpose of this report was to determine the', optimal treatment of patients who had developmem: of acute arterial thrombosis associated with TKA. We reviewed the English-language literature and the extensive orthopedic experience at our hospital, which represents the largest reported experience of acute arterial thrombosis associated with TKA at any one medical center. PATIENTS AND METHODS Between April 1989 and March 1994, 4097' TKAs were performed by the Section of Orthopedic Surgery at Pennsylvania Hospital in Philadelphia. Approximately 8% (332) of these patients were referred for revision knee surgery. Ninety percent (3687) of the operations were performed with use of 927

2 928 Calligaro et al. December 1994 a tourniquet. Seven (0.17%) patients had development acute arterial thrombosis associated with TKA. Thirteen patients reported in the English-language literature (Medline access through Bibliographic Retrieval Services through February 1994) treated at other hospitals were also reviewed. 27 Based on our unique experience we developed a protocol to treat all patients at our institution who had development of these thrombotic complications. First, emergency arteriography was obtained after the knee surgery and before revascularization to identify patent inflow and outflow arteries. The only exception to this rule is when palpable pedal pulses were documented before operation by a reliable source. Second, we favored use of the contrallateral saphenous vein for distal bypasses. TKAs are associated with a high incidence of ipsilateral deep vein thrombosis, and excision of the saphenous vein from the same leg will potentially remove an important conduit for venous drainage. 8 Also, because patients undergoing TKA have development of postoperative leg edema, lymphatic drainage may be interrupted if the ipsilateral saphenous vein is dissected. Third and most importantly, emergency aggressive revascularization procedures were performed, even when distal tibial or pedal bypasses proved necessary. Because our patients were given heparin for arterial surgery, drains placed for TICA were left in place to prevent hematoma formation. Before the revascularization procedure, the lower extremity was prepared, and sterile adhesive drapes were used to isolate the drains on the lateral aspect of the leg. Drains were reconnected to suction after operation. Additionally, patients who undergo TKA at our institution are given low-dose warfarin (Coumadin) the night before or the night of knee surgery (exact dose and timing has varied during this review), which further emphasizes the need for continued drainage. Three patients at other hospitals were treated with the same aggressive protocol as our seven patients (Table I). Nine of 10 patients underwent preoperative arteriography, and all 10 underwent distal arterial bypasses (six tibial, three below-knee popliteal, one pedal). Two required four-compartment leg fasciotomy. Bypasses were performed with autologous vein in nine of 10 cases. Our patients underwent revascularization by use of the contralateral saphenous vein whenever possible (five of seven cases). In situ vein grafts were used in two patients because the contralateral saphenous vein had been excised for coronary artery bypass in one patient and simultaneous bilateral knee replacements were performed in the other patient. In three of our patients, signs and symptoms of acute arterial ischemia did not become severe enough to warrant revascularization until 4 or 5 days after TKA. These three patients were found to have a slightly cooler foot with minimal rest pain or paresthesias but with normal sensation and motor function and weak but present pedal Doppler signals. In all three patients the signs and symptoms slowly worsened, which prompted revascularization. Only one of the three patients underwent preoperative noninvasive vascular laboratory studies, which revealed an ankleflorachial index of She also underwent preoperative arteriography because of the high likelihood that she would require revascularization at some time after the knee surgery. Ten patients at other hospitals were treated with a less aggressive approach, which included thrombectomy in six patients, sympathectomy in two patients, and fasciotomy and delayed bypass in one patient each (Table II). RESULTS Seven of 10 patients treated elsewhere who had development of acute arterial thrombosis associated with TKAs and did not undergo arterial bypasses required major amputations (three above-knee, one below-knee, one transmetatarsal, two unspecified). One other patient died (cause not reported) after undergoing arterial thrombectomy 7 days after the TKA was performed. All 10 patients treated with our aggressive revascularizafion protocol survived, with a 100% limb salvage rate. Of the seven patients we treated, six patients are alive and fully ambulatory with patent bypasses. One patient died of pneumonia 3 months after the TKA and bypass. DISCUSSION We believe that thrombectomy alone is generally inadequate to treat acute arterial thrombosis associated with TKA because of underlying chronic atherosclerotic disease in these patients. In our previous report, acute ischemia occurred only in patients with underlying atherosclerotic occlusive disease. 1 Chronic lower extremity arterial occlusion and intimal plaque disruption caused by knee manipulation or tourniquet compression most likely accounts for thc high failure rate of thrombectomy alone. 1-3 Because routine surgical inspection of the superficial femoral and popliteal arteries was not performed, it was difficult to determine the exact cause of arterial thrombosis in these patients. Nonvisualization of the distal superficial femoral artery and popliteal artery on arteriography can be caused by a thigh tourniquet

3 Volume 20, Number 6 Calligaro et al. 929 Table I. Acute arterial thrombosis associated with TKA patients treated by urgent or emergency arterial bypass Tourniquet Time of X-RAY Distal Press Time treatment Outcome Author Age~sex Ca2 + * Pulse (ram Hg) (rain) Redo (POD )t Arteriogram. Treatment (follow-up) Rush et al.s Starsiak 6 Present series 58/M - None - 90 No 0 SFA occlusion FEM-BKP RVG; Limb salvage fasciotomy (8 months) 56/M Yes Pedal No 0 AKP occlusion FEM-PT RVG Limb salvage (7 months) 64/M Yes Pedal No 0 SFA occlusion FEM-BKP Dacron Limb salvage 63/F None Pedal Yes 0 - AKP-BKP RVG Limb salvage (40 months) 87/F - None - - No 5 AKP occlusion AKP-AT RVG Limb salvage (58 months) 74/M None None No 0 SFA occlusion SFA-PT RVG Limb salvage (17 months) 65/M Yes None No 0 SFA occlusion SFA-PT RVG; Limb salvage fasciotomy (24 months) 75/M - None (Used) - NO 0 SFA occlusion SFA-PT ISVG Limb salvage (3 months) 70/M Yes None No 4 POP occlusion SFA-peroneal trans- Limb salvage located vein graft (3 months) 81/F None None No 4 SFA occlusion SFA-DP ISVG Limb salvage (1 month) X-RAY Ca2 +, Calcified arteries on plain radiograms of the knee; POD, postoperative day; SFA, superficial femoral artery; FEM, femoral (specific site not mentioned); BKP, below-knee popliteal artery; RVG, reversed vein graft;akp, above-knee popliteal artery; PT, posterior; AT, anterior; ISVG, in situ vein graft; DP, dorsalis pedis artery. *Knee films are routinely obtained before operation before all patients undergo TKA. tzero represents same day as TKA. Information was not available in all reports. Table II. Acute arterial thrombosis associated with TKA patients not treated by urgent or emergency arterial bypass Tourniquet Time of X-RAY Distal Press Time treatment Outcome Author Age~sex Ca2 + * Pulse (ram Hg) (min) Redo (POD )t Arteriogram Treatment (fbllow-up) Fortune 2 55/M - Pedal - - No 9 No intimal Thrombectomy damage Hagan and Kaufman s 80/M Yes Pedal No 1 SFA occlusion Thrombectomy McCauley et al.4 60/M Yes Popliteal (Used) - No 0 SFA defect, AKP Thrombectomy occlusion Zahrani and Cuschieri 7 62/M.... No 7 SFA occlusion Sympathectomy Rush et al. s Thrombectomy Thrombectomy Thrombectomy Fasciotomy Sympathectomy Delayed bypass Limb salvage (36 mos) Limb salvage AKA AKA Amputation Amputation Death AKA TMA BKA X-RAY Ca2+, Calcified arteries on plain radiograms of the knee; POD, postoperative day; SEA, superficial femoral artery; AKP, above-knee popliteal artery; AKA, above-knee amputation; TMA, transmetatarsal amputation; BKA, below-knee amputation. *Knee films are routinely obtained before operation before all patients undergo TKA. tzero represents same day as TKA. Information was not available in all reports. that caused an intimal flap, thrombus in the popliteal artery that propagated proximally, or underlying chronic arterial occlusion. A tourniquet was applied during all TK&s (11 of 11) associated with acute arterial thrombosis (this information was not available in the other nine cases). We recommend that the orthopedic surgeon not use a tourniquet, if possible, for patients with a preop-

4 930 Calligaro et al. December 1994 erative anlde/brachial index less than 0.50 or poor quality pulse volume recordings at the ankle or transmetatarsal level. However, there have been several patients at our institution with an ankle/brachial index between 0.30 to 0.49 who have tolerated a tourniquet without adverse effects. We generally do not favor performing a leg bypass before TKA in patients at high risk because use of a tourniquet in one patient who underwent a TKA before this series resulted in graft occlusion. When we were very concerned about the risk of arterial thrombosis in a patient with severe chronic arterial disease, we obtained an arteriogram before the TKA to diminish the duration of leg ischemia after the knee surgery. The patient could then immediately be brought to surgery for emergency revascularization if thrombosis occurred. We followed this course of action in two patients, and both required post TKA femorotibial bypasses. Arterial thrombectomy alone after TKA may occasionally be successful, as reported in two of six patients treated at other hospitals. 2,3 Two additional cases have been reported when thrombectomy was successful after a thigh tourniquet was applied for podiatric surgery and resulted in disruption of an atheromatous plaque in the superficial femoral artery. 9,1 Therefore it is reasonable to attempt passing an embolectomy catheter if a preoperative arteriogram suggests that this treatment may be feasible. However, the surgeon should be prepared to proceed immediately with an additional revascularization procedure if a completion arteriogram does not document a technically satisfactory result of the thrombectomy and adequate pedal Doppler signals or palpable pulses do not return. If this is not done, simple thrombectomy will be generally unsuccessful and is associated with an unacceptably high amputation rate. An aggressive attitude regarding limb salvage is mandatory in treating these complications. We believe that in all patients undergoing TKA, a careful peripheral pulse examination should be done before operation, and, if pulses are abnormal, noninvasive vascular studies should be documented on the medical record. Nurses in the recovery room routinely check the peripheral pulses and status of Doppler signals in the foot. If there is any question of ischemia, a vascular surgeon is immediately notified. A bypass to a tibial or pedal artery, preferably with contralateral saphenous vein, will be required in most cases. A dedicated effort by the vascular surgeon should lead to salvage of a functional lower extremity in patients who have development of acute arterial thrombosis associated with TKA. REFERENCES 1. DeLaurentis DA, Levistsky KA, Booth RE, et al. Arterial and ischemic aspects of total knee arthroplasty. Am J Surg 1992; 164: Fortune WP. Complications of total and partial arthroplasty in the knee. In: Epps CH, ed. Complications in orthopaedic surgery. Philadelphia: JB Lippincott, 1978; Hagan PF, Kaufman EE. Vascular complication of knee arthroplasty under tourniquet: a case report. Clin Orthop 1990;257: McAuley CE, Steed DL, Webster MW. Arterial complications of total knee replacement. Arch Surg 1984;119: Rush JH, Vidovich JD, Johnson MA. Arterial complications of total knee replacement: The Australian experience. J Bone Joint Surg Br 1987;69: Starsiak CR. Arterial occlusion after total knee arthroplasty. J Am Osteopath Assoc 1980;79: Zahrani HA, Cuschieri RJ. Vascular complications after total knee replacement. J Cardiovasc Surg 1989;30: Prevention of venous thrombosis and pulmonary embolism: consensus conference. JAMA 1986;256: Giaunestras NJ, Cranley ]], Lentz M. Occlusion of a tibial artery after a foot operation under tourniquet: a case report. J Bone Joint Surg Am 1977;59A: Williams TA, Baerg RH, Veal WS. Acute arterial occlusion secondary to the use of a pneumatic thigh tourniquet. J Am Podiatr Med Assoc 1986;76: Submitted June 2, 1994; accepted July 28, DISCUSSION Dr. Richard K. Spence (Camden, N.J.). If most of these complications occurred within hours after surgery, is there a relation to use of collected blood? I believe that blood that came out of the knee after the tourniquet was released was returned to the patient by autotransfusion. That blood is very high in activated dotting factors. Is it possiblc that this, coupled with an arterial injury, may be the cause of some of these complications? When did you see these complications? Was it after the patients underwent autotransfusion? Dr. Keith D. Calligaro. Most of our patients and those reported in literature were found to have acute

5 Volume 20, Number 6 Calligaro et al. 931 ischemia in the recovery room or on the operating table as soon as the knee operation was finished. There were a couple of other patients who did not undergo surgery until 3 or 4 days later. It was more than likely accounted for either by the use of the tourniquet, which may have led to a plaque disruption, or, maybe even more likely, some collateral vessels may have been disrupted during the manipulation. Dr. Syed A. Razvi (Brighton, Mass.). Our total knee replacements that are done at our institution are done with epidural anesthetic, which is continued after operation for pain management, and we had a patient who complained of loss of sensation and motor function, which was interpreted by the orthopedic service as a result of use of epidural anesthetic, and the patient was not seen by the vascular service until the limb was nonviable. I suggest that if a patient is going to receive epidural anesthetic and continue to receive it for postoperative pain management, they should undergo very dose arterial monitoring. Dr. Calligaro. I give a lot of credit to our orthopedic surgeons, who are really attuned to this problem. They consult with us frequently on any patient they are concerned about. Dr. Victor J. Celani (Allentown, Pa.). I can recall one patient, an 85-year-old woman, who was undergoing operation on her infected knee for the second time; ischemia developed. The patient would have needed a tibial bypass, but the family refused. You may be right that most of the time you may need a bypass, but I don't think you ought not do a thrombectomy first and see what the results are. We've had some success with that without doing a bypass. So I don't disagree with performing arteriography and saying that most people would require a tibial bypass, but I think you have to try a thrombectomy first and see what your result is, because some of those will succeed. Dr. Calligaro. I can't disagree with that. If you are positive that you can get a very good result, you can stop if a completion angiogram shows a good result. On the other hand, the mistake that has been made in the past is that, when thrombectomy has been attempted and some trickling of blood results, people assume they can stop. Two of these 10 patients were treated by thrombectomy with success. Dr. Robert E. Madden (Valhalla, N.Y.). I would like to take advantage of the extensive experience of the Pennsylvania Hospital to help solve a problem that I recently had regarding an elderly patient who already had a femorotibial in situ bypass and then required knee replacement surgery? How would you approach that? Dr. Calligaro. That is the one situation where we would emphasize as strongly as possible to the orthopedic surgeon not to use a tourniquet. There have been almost 4000 total knee replacements over the past 5 years. Ten percent were used without a tourniquet. Dr. Jack L. Cronenwett (Lebanon, N.H.). I agree with your interpretation. Most of these patients require tibial bypasses, and none of them have palpable pulses if they are the ones who have very diffuse disease to begin with, but its at a subclinical level. The problem that we face is that our orthopedic surgeons are really tuned in to this, and they frequently send patients before operation who do not have palpable pulses and ask us whether they can safely perform this type of orthopedic surgery. Have you had a chance to look at any of ankle/brachial indexes in a large group of patients with this? Based on that, do you have a threshold where you think they should take any special precautions during operation? Do you ever recommend preoperative bypass in a patient who does not have more traditional indications for it? Dr. Cailigaro. That is a very important question. If ~L ankle/brachial index is around 0.4 or 0.5 or above and the: patients do not have diabetes, we will almost always clear that patient for a knee replacement. We will spend extra time with the patient and their family, explaining that they may require a additional emergency arterial bypass. I think that's important. But if the ankle/brachial index is less than 0.4 or 0.3 or the pulse volume recordings are essentially absent at the ankle or foot, we perform arteriography before knee replacement, as we have done on two patien~ during the past 5 years. Both patients ended up requiring an emergency femorotibial bypass. In answer to your question, if it's 0.4 or above we will clear them for surgery, if it's below 0.4 we do not generally recommend an arterial bypass first because we are afraid that it will occlude if the surgeons are going to use a tourniquet. Dr. Michael P. Rade (West Seneca, N.Y.). Did any of the patients whom you treated with the bypass end up with an infected knee prosthesis? How long did the patients receive antibiotics after operation? Dr. Calligaro. We have left that to the orthopedic surgeons. From our standpoint we routinely give prophylactic antibiotic within a half hour of the surgery and another dose. If the antibiotics are given longer than that:, that's the orthopedic surgeon's decision. One other technical adjunct is that the orthopedic surgeons always have a drain that's placed, which we think is important. We remove the collecting device, put an adhesive drape over the suction catheter, put that over the lateral aspect of the knee, prepare both legs, and then put another sterile adhesive drape over that. After operation we then connect the suction catheter so the patients don't get a hematoma; we hope that flaat will decrease complications and infection. Dr. Dominic A. DeLaurentis (Philadelphia, Pa.). Because we will be asked to examine these patients, it is nice to have a good format as what you do. When we did the prospective study we had a discussion regarding prospective versus retrospective. In the prospective study, we studied more than 200 patients. We found a 7% incidence of chronic lower extremity ischemia. When we studied 1100 charts in a retrospective manner, we found an incidence of only about 0.9%, which again demonstrated how poorly a retrospective study is in picking up pulses in the ischemic status of legs. But the overall rate of chronic, lower extremity ischemia was about 2%. Even though these

6 932 Calligaro et al. December 1994 patients are in the proper age range, they don't have a lot of symptomatic atherosclerotic occlusive disease. Its relatively rare. The important point is that when these complications occur, they almost invariably occurred in the patients who had lower extremity ischemia. Chronic lower extremity ischemia is not a common disease. But when you identify a patient who has it, you have to treat it. I tell the orthopedic surgeons that if they want to use a tourniquet, they are going to be faced with this. If the leg is really ischemic, get the arteriogram before operation and, if the patient has rest pain, do the vascular surgery first. Dr. Brooke Roberts (Philadelphia, Pa.). In as much as you obtain these arteriograms promptly, you said many of the patients are recognized, even in the operating room. Have you had a chance to determine whether the tourniquet is causing the obstruction or the manipulation at the knee joint? Obviously if the arteriogram is delayed, the thrombosis may occur. Would you give some information about the actual location of the original trouble? Dr. Calligaro. We did not directly explore those sites to know for sure, but there were a couple of instances where we knew it was due to the tourniquet, based on what we found in those cases. We knew that acute lower extremity ischemia was due to the knee manipulation. It's hard to know in all cases.