Abstract
We describe a modified technique for the salvage of a total knee arthroplasty after disruption of the extensor mechanism. Between January and December 1992, seven patients had reconstruction of the extensor mechanism with use of a medial or an extended medial gastrocnemius flap. Six of the seven patients were followed for a mean of thirty-three months (range, twenty-six to forty-one months) and were evaluated both preoperatively and postoperatively with regard to the knee and functional scores of The Knee Society as well as the range of motion, extensor lag, walking status, and patellar height. The seventh patient was lost to follow-up six months postoperatively and was excluded from the analysis of the results.Preoperatively, the knee and functional scores were 16 ± 12.3 points and 12 ± 12.1 points (mean and standard deviation), respectively; the mean range of motion was 70 ± 44.0 degrees; and the mean extensor lag was 53 ± 33.4 degrees. Postoperatively, the mean knee and functional scores improved to 82 ± 12.4 points and 51 ± 23.0 points, respectively; the mean range of motion improved to 100 ± 21.8 degrees; and the mean extensor lag decreased to 24 ± 18.8 degrees. After the procedure, all patients who previously had been dependent on a walker were able to walk about the community with or without a cane, and those who had been dependent on a wheelchair were able to walk with the assistance of a walker.Patellar height was measured according to the method of Insall and Salvati for the four patients who had a patella. Preoperatively, the patellar heights were grossly abnormal; postoperatively, they more closely approached accepted normal values for three of the four patients.Reconstruction of a complicated rupture of the extensor mechanism with use of a medial gastrocnemius transposition flap after total knee arthroplasty is a reliable option for treatment.
Rupture of the patellar ligament is an infrequent but serious complication after total knee arthroplasty. The reported prevalence of this complication is 0.17 per cent (fourteen of 8288 knees)22. The options for treatment have included observation21,22, use of a double-upright drop-lock brace21, direct repair with sutures1,21,26, fixation to bone with staples21,26, augmentation with use of the semitendinosus or gracilis tendon3,4,10,12,22, reconstruction with use of an allograft5, and turndown of the quadriceps tendon22. The results obtained with these techniques have varied, with serious complications occurring frequently. Cadambi and Engh reported very good results with use of the semitendinosus tendon for reconstruction; however, that technique requires adequate patellar bone stock and is not indicated for patients who have deficient skin or complete loss of the extensor mechanism. Occasionally, patients have multiple problems, such as an exposed prosthesis or an infection together with loss of the extensor mechanism, or scarred and contracted skin or soft tissue around the proximal aspect of the tibia as the result of multiple operative procedures, a failed repair of the extensor mechanism, or a history of a previous patellectomy. We are not aware of a reliable technique that addresses these problems in patients who have had a total knee arthroplasty.
The purpose of the current report is to describe a modified technique for reconstruction of the extensor mechanism in patients who have had a total knee arthroplasty. Use of a medial gastrocnemius transposition flap or an extended medial gastrocnemius flap (a flap that includes a portion of the Achilles tendon) provides improved soft-tissue coverage and allows for the return of the function of the extensor mechanism. This technique was adapted from previously described methods16,17 for reconstruction of the extensor mechanism after excision of the proximal aspect of the tibia and prosthetic replacement.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
†Sections of Orthopaedic Surgery (J. W. J., C. M. DuB., S. R. S., and H. A. F.) and Plastic Surgery (L. J. G.), Department of Surgery, The University of Chicago, 4646 North Marine Drive, Chicago, Illinois 60640. Please address requests for reprints to Dr. Finn.
Seven patients (seven knees) who had disruption of the extensor mechanism after total knee arthroplasty had reconstruction of the extensor mechanism with use of a medial or an extended medial gastrocnemius flap between January and December 1992. There were six women and one man; the patients had a mean age of sixty-eight years (range, fifty-six to eighty-one years) at the time of the reconstruction. One patient was lost to follow-up six months postoperatively and therefore was excluded from the analysis of the results. The mean duration of follow-up for the remaining six patients was thirty-three months (range, twenty-six to forty-one months).
The primary diagnoses for the six patients included a rupture of the patellar ligament during a revision total knee arthroplasty, an acute rupture of the patellar ligament two weeks after a primary total knee arthroplasty, a chronic rupture of the patellar ligament (duration, twenty-two months), an intraoperative takedown of the patellar ligament because of an ankylosed knee, a chronic rupture of the quadriceps tendon (duration, nine months), and an infected knee with an exposed prosthesis and necrosis of the entire extensor mechanism (Table I). All patients had contributing pre-existing comorbid conditions. Two patients (Cases 1 and 5) had rheumatoid arthritis and had had a previous total knee arthroplasty.
All patients had reconstruction of the extensor mechanism with either a medial gastrocnemius flap (Cases 1 through 4) or an extended medial gastrocnemius flap (Cases 5 and 6). Use of a medial gastrocnemius flap generally is indicated when the patella is present and provides vascularized tissue over the tibial tubercle for attachment of the patellar ligament. An extended medial gastrocnemius flap includes a portion of the Achilles tendon and is used when the patella is absent. The tendinous portion of the flap bridges the defect from the tibial tubercle to the quadriceps tendon.
All operative procedures were performed by the senior one of us (H. A. F.). The patients were evaluated by two independent observers (C. M. DuB. and S. R. S. ). The evaluation included a comparison of the preoperative and postoperative knee and functional scores of The Knee Society6,12, with a total possible score of 100 points for each. The preoperative and postoperative ranges of motion and extensor lags were measured with a hand-held goniometer. An extensor lag was defined as the difference between the greatest passive extension and the greatest active extension of the knee. Walking status also was assessed preoperatively and postoperatively. Patellar tracking was assessed both clinically and radiographically.
Patellar height was measured with use of the method described by Insall and Salvati11. Only four patients (Cases 2 through 5) could be evaluated with this method. Of the two remaining patients, one (Case 1) had had a patellectomy before the reconstruction and the other (Case 6) had a patellectomy as part of the reconstruction. Measurements were made on the preoperative lateral radiographs of the knee. The method of Insall and Salvati is used to determine the ratio between the length of the patellar ligament (on its posterior surface) and the greatest diagonal length of the patella11. A ratio of 1.02 ± 0.13 (mean and standard deviation) for normal knees was used as the standard for comparison11.
Five of the six patients had a revision to a constrained total knee arthroplasty; this procedure was performed concurrently with the reconstruction of the extensor mechanism. Four of the patients (Cases 1, 3, 5, and 6) had the revision because of a combination of severe ligamentous and osseous deficiencies and loosening or malpositioning of a component. Another patient (Case 4) had the revision as a result of soft-tissue contractures, myositis, and ankylosis associated with the treatment of a supracondylar femoral fracture after a total knee arthroplasty. The remaining patient (Case 2) had a primary constrained total knee replacement because of severe multiple-plane instability. This patient was morbidly obese; two weeks after the procedure, she sustained a traumatic disruption of the patellar ligament during a fall, which necessitated reconstruction of the extensor mechanism.
Operative Technique
The incision from the previous total knee arthroplasty is extended down the leg, just medial to the tibia (Fig. 1). The superficial fascia is incised, and a plane is developed between the medial gastrocnemius and soleus muscles, where the plantaris tendon usually is visualized. The medial gastrocnemius is divided at its distal insertion into the Achilles tendon (Fig. 2). Proximally, the median raphe is identified between the bellies of the medial and lateral gastrocnemius muscles. With blunt and sharp dissection, the medial gastrocnemius is divided up to the level of the tibial condyles to provide an adequate arc of rotation (Fig. 3). Throughout the procedure, the medial sural artery must be preserved to provide an adequate blood supply to the medial gastrocnemius (Fig. 2). The muscle is transposed anteriorly at the level of the knee joint to cover the tibial tubercle and is sutured to the fascia of the anterior compartment. The patellar ligament and the anterior aspect of the joint capsule are sutured to the medial border of the medial gastrocnemius flap (Figs. 4 and 5). A split-thickness skin graft may be used to cover part of the medial gastrocnemius flap if approximation of the skin is not possible.
The technique can be altered to include reconstruction of the entire extensor mechanism for patients in whom a repair of a rupture of the quadriceps tendon has failed, such as those who have had a patellectomy or a patellar fracture with disruption of the extensor mechanism, those who have had loss of the extensor mechanism secondary to infection, and those who have had takedown of a fusion. In these situations, the incision is extended distally and the medial one-third to one-half of the Achilles tendon is obtained in continuity with the medial gastrocnemius muscle (Fig. 6). The entire flap then is transposed anteriorly as already described, and the deep fascia of the gastrocnemius is approximated to the proximal aspect of the tibia with sutures or wires placed through drill-holes (Fig. 7). With the knee in full extension, the Achilles-tendon portion of the flap is rotated superiorly and is approximated to the remaining portion of the quadriceps tendon under as much tension as possible. A split-thickness skin graft can be used for closure when necessary.
Postoperatively, the patients were immobilized in an above-the-knee cast with the knee in extension. Gait-training and isometric exercises for the quadriceps were begun immediately. At six weeks postoperatively, the cast was removed and active extension was verified by means of straight-leg raising. Gentle, passive range-of-motion exercises then were started. Eight weeks postoperatively, the patients progressed to active range-of-motion and strengthening exercises. Intensive strengthening exercises were started twelve weeks postoperatively. The patients were placed in a hinged, full-length brace for eight to ten weeks after the removal of the cast, with progressive flexion dialed into the brace.
The patients were evaluated preoperatively and postoperatively with regard to the knee and functional scores of The Knee Society6,12, range of motion, extensor lag, walking status, and patellar height (Table II).
Knee score: Preoperatively, the knee score was a mean (and standard deviation) of 16 ± 12.3 points (range, 0 to 37 points). Postoperatively, the score increased to a mean of 82 ± 12.4 points (range, 68 to 100 points). All patients had little or no pain as well as an improved knee score.
Functional score: The mean preoperative functional score was 12 ± 12.1 points (range, 0 to 30 points). Postoperatively, the score improved to a mean of 51 ± 23.0 points (range, 20 to 90 points).
Range of motion: The mean preoperative range of motion was 70 ± 44.0 degrees (range, 0 to 120 degrees). Postoperatively, the mean range of motion improved to 100 ± 21.8 degrees (range, 65 to 130 degrees). Four of the six patients had improvement in the range of motion, whereas two had a small decrease (Table II). Of the latter two patients, one (Case 1) had had a preoperative range of motion from 0 to 120 degrees, and a 20-degree flexion contracture had developed postoperatively. The other patient (Case 6) had had a preoperative range of motion from 0 to 90 degrees and lost 10 degrees of flexion postoperatively.
Extensor lag: The mean preoperative extensor lag was 53 ± 33.4 degrees (range, 5 to 90 degrees). (The preoperative data for one patient, Case 4, were not considered to be applicable because he had a severely ankylosed knee.) Postoperatively, the mean value improved to 24 ± 18.8 degrees (range, 0 to 50 degrees). The mean postoperative extensor lag for the four patients (Cases 1 through 4) who had a simple medial gastrocnemius flap was 14 ± 13.9 degrees (range, 0 to 30 degrees).
Walking status: Preoperatively, all patients were dependent on either a walker (Cases 1 through 4) or a wheelchair (Cases 5 and 6). After the procedure, all of the patients who previously had been able to walk in the home but had depended on a walker were able to walk about the community either without the use of assistive devices (Cases 2 and 4) or with the assistance of a cane only (Cases 1 and 3). Both of the patients who had been dependent on a wheelchair preoperatively were able to walk with the assistance of a walker postoperatively.
Patellar height: Patellar height could be measured for four of the six patients. Three of these patients (Cases 2, 3, and 5) had an improvement toward the reported normal mean for patellar height when measured with the method of Insall and Salvati11. The preoperative ratios for these patients had been 1.58, 1.92, and 0.70, respectively; postoperatively, the values improved to 1.03, 1.33, and 1.02. The fourth patient (Case 4) had had a preoperative ratio of 1.02 and had a postoperative ratio of 1.97.
Complications: There were two complications. One patient (Case 4), who had had an ankylosed knee preoperatively, needed manipulation eight weeks postoperatively because of a limited range of motion (0 to 65 degrees). After manipulation, the range of motion improved to 82 degrees, but a 30-degree extensor lag developed with a range of 80 to 30 degrees. At thirty-one months postoperatively, the range of motion was 0 to 65 degrees, with a 30-degree extensor lag. Another patient (Case 6), who had had an infected knee with an exposed prosthesis preoperatively, had a small region of necrosis at the margin of the wound postoperatively. The patient was managed with débridement and a split-thickness skin graft.
Rupture of the patellar ligament is an infrequent but serious complication after total knee arthroplasty. However, when it does occur, the result can be disastrous21. Rupture of the patellar ligament after total knee arthroplasty has received little attention; we were able to find a total of only sixty-nine instances that have been reported1,5,27. The etiology is multifactorial and can include traumatic rupture3,22, atraumatic rupture in patients receiving high doses of steroids or in those who have diabetes or collagen vascular disease22, devascularization resulting from extensive lateral retinacular release or multiple operations14, use of a non-rotating-hinge prosthesis19, malalignment of a component resulting in poor patellar tracking8, loss of fixation after deliberate detachment for exposure of an ankylosed knee22 or for a distal realignment procedure21, and inadvertent or traumatic detachment during exposure of the knee21. The options for treatment include observation21,22, bracing21, fixation with sutures or staples1,21,26, autologous-tissue augmentation with use of the semitendinosus or gracilis tendon3,4,10,13,22, turndown of the quadriceps tendon22, and reconstruction with use of an allograft5. Non-operative treatment often relegates the patient to a wheelchair or to the home, or both. Previous options for operative treatment, including direct repair, augmentation with an autologous graft3, and reconstruction with an allograft5, have been associated with a high rate of complications and have not addressed the associated problems of contracted, devascularized skin flaps or deficient patellar bone stock.
The medial gastrocnemius flap has been shown to be suitable for providing soft-tissue coverage of the proximal aspect of the tibia, the knee, and the distal aspect of the femur15. Exposed knee prostheses also can be covered successfully with the gastrocnemius flap2,9,23. Feldman et al. dissected twenty cadaveric specimens and confirmed that the medial and lateral heads of the gastrocnemius muscle each receive one primary neurovascular bundle, which proceeds longitudinally for the entire length of the muscle and receives no additional arterial inflow. This anatomy makes elevation and rotation of the muscle belly on its single, proximal pedicle a simple and reliable procedure17.
Until the 1980's, malignant tumors of the proximal aspect of the tibia were treated primarily with above-the-knee amputation. One of the major obstacles to preservation of the extremity was the inability to reconstruct a functional extensor mechanism after resection of the proximal aspect of the tibia and insertion of a prosthesis. The emphasis on limb salvage in the last decade necessitated a solution for reconstruction of the extensor mechanism. Malawer and Price17 developed a technique for reconstruction of the extensor mechanism in which the patellar ligament is attached to the tibial component and the component, together with the patellar ligament, is covered with a medial gastrocnemius transposition flap. The patellar ligament is sutured to the fascia of the gastrocnemius flap. Malawer and Price17 reported good results with this technique. Our technique is an adaptation of such previously accepted techniques16,17 for reconstruction of the extensor mechanism after excision of the proximal aspect of the tibia and prosthetic replacement.
To our knowledge, we are the first to report on the use of a medial gastrocnemius transposition flap for the treatment of disruption of the extensor mechanism after total knee arthroplasty. The technique offers several advantages. The flap provides a vascularized muscle bed, which allows for improved healing of the reconstructed extensor mechanism. There is excellent soft-tissue coverage of the proximal aspect of the tibia and of the components of the total knee arthroplasty. This technique does not rely on the bone quality of the patella or the proximal aspect of the tibia; rather, it relies on soft-tissue healing, and therefore late failure is unlikely. The extensile approach and the rotation of the medial gastrocnemius flap can be performed easily; in addition, the technique can be altered to include reconstruction of the entire extensor mechanism with a portion of the Achilles tendon. Finally, there is no loss of function after the transfer because the soleus and the remaining lateral head of the gastrocnemius function synergistically to provide plantar flexion of the ankle17,18,20,25.
In the current series, the knee and functional scores increased by a mean of approximately 66 and 39 points, respectively. Five of the six patients had little or no pain and had improvement in the knee and functional scores. The range of motion improved in four patients and decreased slightly in two. The extensor lag decreased by a mean of 29 degrees. As mentioned earlier, one patient (Case 4) was not included in the preoperative calculation of the mean extensor lag. This patient had complete ankylosis with the knee in extension, and the patellar ligament was disrupted during operative exposure at the time of a revision total knee replacement. The walking status improved dramatically in all patients. Of the four patients for whom patellar height could be assessed, three had normalization of the patellar height postoperatively and one had patella alta.
The four patients (Cases 1 through 4) who had reconstruction with use of a medial gastrocnemius flap can be considered as a subgroup. Postoperatively, all had dramatic improvement in the knee and functional scores, were able to walk about the community with or without the use of a cane, and had a substantially decreased extensor lag (mean, 14 ± 13.9 degrees). It appears that patients who have reconstruction of a ruptured patellar ligament with use of a medial gastrocnemius flap can have a predictable return of function of the extensor mechanism and can return to community activities.
The two patients (Cases 5 and 6) who had reconstruction with use of an extended medial gastrocnemius flap also had substantial improvement in the knee score, but only one of these patients (Case 6) had improvement in the functional score. In addition, the postoperative extensor lag was large (50 and 40 degrees) in these two patients. Nevertheless, both patients did regain some function of the extensor mechanism, and the reconstruction also provided soft-tissue coverage for the components. The fact that these two patients did not report giving-way of the knee suggests that the reconstruction has an important impact on the function of the extensor mechanism, thus allowing walking with use of a walker.
Instability of the knee has received little attention as an etiological factor in rupture of the extensor mechanism; however, five of our six patients had major instability of the knee that necessitated the use of a constrained prosthesis in addition to the reconstruction of the extensor mechanism. We believe that it is important to address any instability of the knee at the time of the reconstruction of the extensor mechanism in order to place as little stress as possible on the reconstruction postoperatively.
In conclusion, the use of a medial or an extended medial gastrocnemius flap appears to be a reliable option for the reconstruction of a ruptured extensor mechanism after total knee arthroplasty. It may be best suited for the treatment of complicated disruptions of the extensor mechanism in patients who have had a failed repair of the quadriceps or who have a rupture of the patellar ligament and poor soft-tissue coverage, poor healing potential, inadequate patellar bone stock, or an exposed prosthesis with loss of the extensor mechanism.
Abril, J. C.; Alvarez, L.; and Vallejo, J. C.: Patellar tendon avulsion after total knee arthroplasty. A new technique. J. Arthroplasty,10: 275-279, 1995.10275
1995
[PubMed]
Bengtson, S.; Carlsson, A.; Relander, M.; Knutson, K.; and Lidgren, L.: Treatment of the exposed knee prosthesis. Acta Orthop. Scandinavica,58: 662-665, 1987.58662
1987
Cadambi, A., and Engh, G. A.: Use of a semitendinosus tendon autogenous graft for rupture of the patellar ligament after total knee arthroplasty. A report of seven cases. J. Bone and Joint Surg.,74-A: 974-979, Aug. 1992.74-A974
1992
Ecker, M. L.; Lotke, P. A.; and Glazer, R. M.: Late reconstruction of the patellar tendon. >J. Bone and Joint Surg.,61-A: 884-886, Sept. 1979.61-A884
1979
Emerson, R. H., Jr.; Head, W. C.; and Malinin, T. I.: Extensor mechanism reconstruction with an allograft after total knee arthroplasty. Clin. Orthop.,303: 79-85, 1994.30379
1994
[PubMed]
Ewald, F. C.: The Knee Society total knee arthroplasty roentgenographic evaluation and scoring system. Clin. Orthop.,248: 9-12, 1989.2489
1989
[PubMed]
Feldman, J. J.; Cohen, B. E.; and May, J. W., Jr.: The medial gastrocnemius myocutaneous flap. Plast. and Reconstr. Surg.,61: 531-539, 1978.61531
1978
Figgie, H. E., III; Goldberg, V. M.; Figgie, M. P.; Inglis, A. E.; Kelly, M.; and Sobel, M.: The effect of alignment of the implant on fractures of the patella after condylar total knee arthroplasty. J. Bone and Joint Surg.,71-A: 1031-1039, Aug. 1989.71-A1031
1989
Gerwin, M.; Rothaus, K. O.; Windsor, R. E.; Brause, B. D.; and Insall, J. N.: Gastrocnemius muscle flap coverage of exposed or infected knee prostheses. Clin. Orthop.,286: 64-70, 1993.28664
1993
[PubMed]
Gustilo, R. B., and Thompson, R.: Quadriceps and patellar tendon ruptures following total knee arthroplasty. In Total Arthroplasty of the Knee. Proceedings of The Knee Society, 1985-1986, pp. 41-47. Edited by J. A. Rand and L. D. Dorr. Rockville, Maryland, Aspen, 1987.
Insall, J., and Salvati, E.: Patella position in the normal knee joint. Radiology,101: 101-104, 1971.101101
1971
[PubMed]
Insall, J. N.; Dorr, L. D.; Scott, R. D.; and Scott, W. N.: Rationale of The Knee Society clinical rating system. Clin. Orthop.,248: 13-14, 1989.24813
1989
[PubMed]
Kelikian, H.; Riashi, E.; and Gleason, J.: Restoration of quadriceps function in neglected tear of the patellar tendon. Surg., Gynec. and Obstet.,104: 200-204, 1957.104200
1957
Lynch, A. F.; Rorabeck, C. H.; and Bourne, R. B.: Extensor mechanism complications following total knee arthroplasty. J. Arthroplasty,2: 135-140, 1987.2135
1987
[PubMed]
McCraw, J. B.; Dibbell, D. G.; and Carraway, J. H.: Clinical definition of independent myocutaneous vascular territories. Plast. and Reconstr. Surg.,60: 341-352, 1977.60341
1977
Malawer, M. M., and McHale, K. A.: Limb-sparing surgery for high-grade malignant tumors of the proximal tibia. Surgical technique and a method of extensor mechanism reconstruction. Clin. Orthop.,239: 231-248, 1989.239231
1989
[PubMed]
Malawer, M. M., and Price, W. M.: Gastrocnemius transposition flap in conjunction with limb-sparing surgery for primary bone sarcomas around the knee. Plast. and Reconstr. Surg.,73: 741-750, 1984.73741
1984
Mathes, S. J., and Nahai, F.: Muscle flap transposition with function preservation: technical and clinical considerations. Plast. and Reconstr. Surg.,66: 242-249, 1980.66242
1980
Mochizuki, R. M., and Schurman, D. J.: Patellar complications following total knee arthroplasty. J. Bone and Joint Surg.,61-A: 879-883, Sept. 1979.61-A879
1979
Murray, M. P.; Guten, G. N.; Sepic, S. B.; Gardner, G. M.; and Baldwin, J. M.: Function of the triceps surae during gait. Compensatory mechanisms for unilateral loss. J. Bone and Joint Surg.,60-A: 473-476, June 1978.60-A473
1978
Rand, J. A.; Morrey, B. F.; and Bryan, R. S.: Patellar tendon rupture following total knee arthroplasty. Tech. Orthop.,3: 45-48, 1988.345
1988
Rand, J. A.; Morrey, B. F.; and Bryan, R. S.: Patellar tendon rupture after total knee arthroplasty. Clin. Orthop.,244: 233-238, 1989.244233
1989
[PubMed]
Sanders, R., and O'Neill, T.: The gastrocnemius myocutaneous flap used as a cover for the exposed knee prosthesis. J. Bone and Joint Surg.,63-B(3): 383-386, 1981.63-B(3)383
1981
Scuderi, C.: Ruptures of the quadriceps tendon. Study of twenty tendon ruptures. Am. J. Surg.,95: 626-635, 1958.95626
1958
[PubMed]
Simon, S. R.; Mann, R. A.; Hagy, J. L.; and Larsen, L. J.: Role of the posterior calf muscles in normal gait. J. Bone and Joint Surg.,60-A: 465-472, June 1978.60-A465
1978
Siwek, C. W., and Rao, J. P.: Ruptures of the extensor mechanism of the knee joint. J. Bone and Joint Surg.,63-A: 932-937, July 1981.63-A932
1981
Zanotti, R. M.; Freiberg, A. A.; and Matthews, L. S.: Use of patellar allograft to reconstruct a patellar tendon-deficient knee after total joint arthroplasty. J. Arthroplasty,10: 271-274, 1995.10271
1995
[PubMed]