Arch Orthop Trauma Surg (2014) 134:843–852 DOI 10.1007/s00402-014-1991-y
Arthroscopy and Sports Medicine
Osteochondral transplantation for the treatment of osteochondral defects at the talus with the Diamond twin system® and graft harvesting from the posterior femoral condyles Wolf Petersen · Pouria Taheri · Benedikt Schliemann · Andrea Achtnich · Cara Winter · Phillip Forkel
Received: 10 November 2013 / Published online: 18 April 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract Background The aim of this study is to analyze clinical results after osteochondral cylinder transplantation for osteochondral defects at the medial or lateral talar dome using the Diamond twin system (Karl Storz). We hypothesize that grafts harvesting from the posterior femoral condyles are associated with less donor site morbidity than reported by previous studies. Methods We have surgically treated 20 patients with an osteochondral defect of the talus by osteochondral transplantation with the Diamond twin system via an osteotomy of the ankle. The osteochondral cylinders were harvested from the posterior aspects of the femoral condyles of the ipsilateral knee. The defects at donor site were filled with a bone substitute of tricalcium phosphate (Synthricer, Karl Storz). The mean age was 25.4 years. After a mean time of 12.6 months, the screws at the medial malleolus were removed and an arthroscopy was performed. The functional outcome was evaluated with the visual analog scale for pain at walking, running, stair climbing, quality of life at the time of implant removal and at a mean follow-up of 25.8 months. Activity was assessed with the Tegner scale. Knee function was evaluated with the Lysholm score. Results In one case, the osteochondral cylinder did not heal and an osteochondral fragment was removed arthroscopically. In all other cases, the osteochondral cylinder W. Petersen (*) · P. Taheri · A. Achtnich · C. Winter · P. Forkel Department of Orthopaedic and Trauma Surgery, Martin Luther Hospital, Grunewald, Caspar Theyss Strasse 27‑31, 14193 Berlin, Germany e-mail: w.petersen@mlk‑berlin.de B. Schliemann Department of Trauma‑, Hand‑ and Reconstructive Surgery, University Hospital Muenster, Muenster, Germany
was stable with surrounding cartilage. The average ICRS Cartilage Repair Assessment was 10.1 points (±1.3). All malleolar osteotomies healed radiologically. In 15 patients, a synovectomy and local debridement of the ankle were performed at second-look arthroscopy. Ankle pain at walking, running and stair climbing as measured by a visual analog scale (10-0) decreased significantly from preoperatively to the first follow-up (mean 12.6 months) and to the second follow-up (mean 25.8 months). The ankle-related quality of life increased significantly from preoperatively to postoperatively. There was no significant change in the Lysholm score. The activity measured with the Tegner activity scale increased significantly from preoperatively to the last follow-up, but only two out of nine patients continued pivoting sports. Conclusions Autologous osteochondral grafting with the Diamond twin system is a reliable treatment option for symptomatic osteochondral defects of the talus. After 1 year, the majority of patients had still some complaints. However, after screw removal and second-look arthroscopy, the pain and ankle-related quality of life further improved. Clinical relevance The donor site morbidity after graft harvesting from the posterior aspects of the femoral condyles is lower than previously reported. Keywords Osteochondral lesion · Tricalcium phosphate · Ceramic bone substitute · Activity · Donor site morbidity · Osteoarthritis · Cartilage repair · Subchondral bone plate
Introduction Osteochondral grafting is a treatment option for stage IV osteochondral defects of the talus. Several techniques for osteochondral grafting have been described in the literature
13
844
[1, 2, 4, 6–8, 10, 11, 13, 15, 18, 22, 25–27]. All techniques have in common that cylindrical osteochondral grafts are harvested from the knee to reconstruct a defect at the talar dome. With the so-called mosaicplasty technique, several small cylinders are used to cover the defect [6–8, 25]. This technique can be used when the defect is predominantly chondral. Otherwise the cylinders are anchored in necrotic bone. With other systems such as with the OATS, SDS or Diamond twin system one or two larger osteochondral cylinders are transferred [4, 10, 14, 22]. With these techniques, good clinical results have been reported for osteochondral transplantation at the talus [1, 2, 4, 6–8, 10–12, 14, 17, 22, 25–27]. For the Diamond twin system, clinical results for the treatment of osteochondral defects of the talus have not been described until know. Most authors use the femoropatellar joint as donor site for the osteochondral grafts [1, 2, 4, 6–8, 10, 11, 13, 15, 18, 22, 25–27]. In our experience, the harvesting of osteochondral grafts from the femoropatellar joint may be associated with donor site morbidity such as anterior knee pain [14]. Several other authors could confirm and described significant donor site morbidity after graft harvesting from the femoropatellar joint [4, 19, 21, 25–27]. An alternative donor site for osteochondral grafts is the posterior aspect of the femoral condyles [14]. Good-toexcellent clinical results with low donor site morbidity have been described for the treatment of osteochondral defects of the knee [14]. A disadvantage of this technique was that the cylindrical grafts were harvested with the patient in prone position. That means, the patient has to be turned around with new skin disinfection and draping to implant the graft into the defect. Because of these disadvantages we modified the original technique. With the modified technique, the osteochondral grafts are harvested via a small horizontal incision in the popliteal fossa with the patient in lateral decubitus position (Fig. 1). To reach the ankle without new skin disinfection the leg has to be rotated in the hip. The aim of this study is to analyze radiological, arthroscopical, and clinical results after osteochondral transplantation for defects at the talus using the Diamond twin system. The grafts were harvested from the posterior femoral condyles of the ipsilateral knee. Our hypothesis was that with this technique, the donor site morbidity after graft harvesting from the posterior aspects of the femoral condyles is lower than previously reported.
Patients and methods We prospectively followed 20 patients who had been treated by autologous osteochondral grafting for symptomatic
13
Arch Orthop Trauma Surg (2014) 134:843–852
osteochondral defects of the talus between 2008 and 2011. The study was approved by the institutional review board. All patients gave written informed consent to be included in a scientific study. Twelve men and eight women were treated. The mean age was 25.4 years (18–44 years). Inclusion criteria for the study were: Age >18 years and a stage III or IV osteochondral defect at the talar dome (ICRS classification). All patients had stage IV osteochondral defect of the talus according to the ICRS classification. The osteochondral defect was located laterally in one and medially in 19 patients. The lesions were diagnosed by plain radiography and MRI. The mean size of the defects of the talar dome was 16.95 (±2.5) × 10.4 (±1.1) mm. All patients had previous unsuccessful arthroscopic surgery with excision of the fragment and microfracture. Mean time to previous surgery was 17 months (range 3–46 months). Surgical technique All surgical procedures were performed under general anesthesia. The defect size was determined on the preoperative MRI. The operation starts with graft harvesting at the knee. Therefore, precise information about the defect size is necessary to determine the number of grafts needed to cover the defect. For osteochondral cylinder transplantation the Diamond twin system® (Karl Storz, Tuttlingen, Germany) was used (Fig. 1). This system consists of various diamond bone cutters. The mutually adjusted sizes of the bone cutters allow a press-fit implantation. The donor cylinder is 0.05 mm larger in diameter than the defect cylinder. Graft harvesting The operation starts with harvesting of the osteochondral donor cylinders from the posterior femoral condyles. To reach the donor site the patient was placed in a lateral position (Fig. 1). After application of a thigh tourniquet the skin was disinfected and draped. A transverse skin incision was performed in the fossa poplitea. The fascia was opened with a scissor and the posterior joint capsule of the lateral femoral condyle was exposed with four Langenbeck retractors after blunt preparation with the index finger. The lateral gastrocnemius muscle must be pushed to the side with a periosteal elevator. Then, the capsule was opened with a scalpel and a Hohmann retractor was placed behind the femoral condyles. The cartilage is cut with an appropriately sized extractor and then the bone is cut with the diamond cutter. To prevent heat damage the cutter has to be cooled with NaCl. The depth of the donor cylinder should be between 15 and 20 mm according to the size of the defect.
Arch Orthop Trauma Surg (2014) 134:843–852
845
Fig. 1 This figure shows the operative technique for graft harvesting from the posterior femoral condyles. a The patient is positioned on the contralateral side and the posterior femoral condyles are exposed via a horizontal incision. b The graft cylinder is cut with a diamond
bone cutter under continuous cooling with NaCl. c The cylinder is extracted with a specific extractor. d The harvest defect is filled with a ceramic bone substitute made of tricalcium phosphate (Synthricer, Karl Storz, Germany)
The depth can be read on the tiller. The donor cylinder is finally extracted with an extractor by a quick rotation. After rinsing, the defect is filled with a cylindrical bone graft substitute optimized for use with the instruments (Syntricer®—biodegradable β-tricalcium phosphate ceramic, Karl Storz, Tuttlingen). The cylindrical bone graft substitute is gently pressed into the donor defect until the cylinder surface is 1 or 2 mm below the surrounding cartilage (Fig. 1). The donor site is rinsed with NaCl and the capsule is closed by resorbable sutures. If a second cylinder is needed the same procedure has to be repeated at the medial side. In 15 cases, two cylinders were used.
protect the posterior tibial tendon from damage of the oscillating saw. The osteotomy plane was marked by two K wires with the use of a fluoroscope and the osteotomy was performed with an oscillating saw and a chisel. At the lateral side an oblique distal fibular osteotomy was performed below the syndesmosis. The defect was prepared by excising the necrotic sequestrum with a bone cutter one size smaller as used for harvesting of the donor cylinders (Fig. 2). The technique for defect preparation is the same as for harvesting the donor cylinder. The depth of the defect hole has to be measured and in case of a mismatch the donor cylinder should be shortened. The donor cylinders were pressed manually into the recipient talar site to fill the defect. In 17 patients, two cylinders were used and in four patients only one cylinder was used. The diameter of the cylinders ranged between 9.45 and 14 mm. In case of two cylinders, they overlap. The osteotomy was rigidly fixed using either titanium screws (medial malleolus) or with an AO 1/3 tubular plate for the fibula.
Osteochondral transplantation at the talus In 19 cases, a medial malleolar osteotomy was performed (Fig. 2). In one case, a lateral fibular osteotomy was performed. The medial malleolus was exposed by two Hohmann retractors after skin incision. The posterior retractor should
13
846
Arch Orthop Trauma Surg (2014) 134:843–852
Fig. 2 a The osteochondral defect is exposed via a malleolar osteotomy. b The defect cylinder is excised with the diamond-coated grinder and c the extractor. Both instruments are one size smaller than that used for graft harvesting. d In this case, the defect is completely filled with two osteochondral donor cylinders
Rehabilitation protocol The patients were mobilized with partial weight-bearing for 6 weeks and with early physiotherapy to both the ankle and the knee. Unprotected weight-bearing was allowed after a radiological control. CPM for the ankle was used for the first 6 weeks. Sensomotoric training started after 8 weeks postoperatively. Jogging was allowed after 6 months and impact activities after screw removal after 1 year. Follow‑up All peri- or postoperative complications were documented. One follow-up examination was performed after a mean of
13
12.6 months and a second follow-up was performed at a minimum of 24 months (mean 25.8 months) after surgery. The first follow-up was the time of implant removal which was performed in all patients. In 15 cases, a secondlook arthroscopy was performed at this time point. The indication for second-look arthroscopy was ankle pain. The arthroscopic findings have been documented (synovitis, adhesions). The quality of the grafts was assessed using the International Cartilage Repair Society (ICRS) cartilage repair assessment which is a semiquantitative outcome measure [23]. This score evaluates the degree of cartilage repair, integration to the border zone and macroscopic appearance. In each category, 4 score points can be assigned. By adding these points the overall score is calculated (12 points: normal, 11–8 points: nearly normal, 7–5
Arch Orthop Trauma Surg (2014) 134:843–852
847
points: abnormal, 4–0 points: severely abnormal). In all cases, a conventional X-ray of the ankle and the knee was performed to examine healing of the malleolar osteotomy and the defect site before. At the 12.6-month follow-up the knee and ankle were examined for range of motion, stability, local tenderness and crepitation. Subjective pain at the ankle was evaluated using a visual analog scale (VAS, 10 always pain, 0 never pain) preoperatively, at the 12.6-month and at the 25.8-month follow-up. Pain was assessed for walking on even ground, walking on uneven ground, running, stair climbing. Ankle-related quality of life was also assessed using a visual analog scale (VAS, 10 = 100 % reduction of life quality, 0 = no influence on life quality) preoperatively, at the 12.6-month and at the 25.8-month follow-up. For quality of life assessment the patients were asked: How much is your quality of life affected by your ankle problems. Knee function was assessed with the Lysholm score [16]. The activity was assessed with the Tegner activity scale [9, 24]. At the 25.8month follow-up, the patients were asked to rate themselves on the German version of the Tegner scale before start of the complaints, immediately before surgery and at the 25.8-month follow-up. They were further asked which sports they performed before start of the complaints, immediately before surgery and at the 25.8-month follow-up. Statistics The statistical analysis was performed at the Medistat Institute in Kiel, Germany. The Shapiro–Wilk test was used to test if the measurements were normal distributed. For nonparametric parameters, nonparametric test was used for statistical analysis of group differences. For global temporal changes the Friedman test was used. The Wilcoxon test (p 5° of flexion or extension in comparison to the contralateral side). Fourteen patients had local tenderness at the medial malleolus. In ten cases, there was local tenderness over the posterior tibial tendon behind the medial malleolus. No patient had any local tenderness at the knee joint. Range of motion of the knee was free in all patients with no crepitation.
13
848
Arch Orthop Trauma Surg (2014) 134:843–852
Fig. 4 Postoperative MRI after 11 months of the patient with a broken bone cylinder (see Fig. 3a)
At second-look arthroscopy, adhesions and synovitis in the anterior ankle compartment were seen in all 15 patients. A partial synovectomy and arthrolysis were performed in all these patients. In one patient we observed a broken graft cylinder. The broken graft was removed arthroscopically in this case. The broken cylinder was also diagnosed on a preoperative MRI (Fig. 4). In all other patients the graft was stable at second-look arthroscopy (Fig. 5). In 5 cases, the graft level was raised in comparison of the surrounding cartilage, in two cases the graft level was below the surrounding cartilage, and in eight cases the surface of the graft cylinder had the same level as the surrounding cartilage. The overall ICRS arthroscopic score for the graft was 10.1 (±1.3). The most common findings were fibrillated surface of the graft cylinders. Two patients had a score of 7 which is considered as abnormal. Both patients were also clinically considered as failures because in both patients there was no significant improvement in pain. Ankle pain at walking, running, jumping and stair climbing assessed with a visual analog scale (VAS, 10 always pain, 0 never pain) decreased significantly from the preoperative status to both follow-ups (Fig. 6a–c). Ankle-related quality of life also improved significantly (Fig. 7a). Activity as rated with the Tegner activity scale is shown in Fig. 7b. Before the beginning of symptoms nine patients performed pivoting sports (soccer, handball, basketball). At the last follow-up, only two patients performed pivoting sports.
13
Fig. 5 Arthroscopic findings at second-look arthroscopy. a Osteochondral cylinder with a smooth surface which is in level with the surrounding cartilage. b In this patient the graft level is raised in comparison to the superficial cartilage. c This cylinder has superficial fibrillations but the graft is in level with the surrounding cartilage
Arch Orthop Trauma Surg (2014) 134:843–852
849
◂ Fig. 6 a The Shapiro–Wilk test showed a nonparametric distribution
of the measurements with the VAS for pain at walking (p
Arthroscopy and Sports Medicine
Osteochondral transplantation for the treatment of osteochondral defects at the talus with the Diamond twin system® and graft harvesting from the posterior femoral condyles Wolf Petersen · Pouria Taheri · Benedikt Schliemann · Andrea Achtnich · Cara Winter · Phillip Forkel
Received: 10 November 2013 / Published online: 18 April 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract Background The aim of this study is to analyze clinical results after osteochondral cylinder transplantation for osteochondral defects at the medial or lateral talar dome using the Diamond twin system (Karl Storz). We hypothesize that grafts harvesting from the posterior femoral condyles are associated with less donor site morbidity than reported by previous studies. Methods We have surgically treated 20 patients with an osteochondral defect of the talus by osteochondral transplantation with the Diamond twin system via an osteotomy of the ankle. The osteochondral cylinders were harvested from the posterior aspects of the femoral condyles of the ipsilateral knee. The defects at donor site were filled with a bone substitute of tricalcium phosphate (Synthricer, Karl Storz). The mean age was 25.4 years. After a mean time of 12.6 months, the screws at the medial malleolus were removed and an arthroscopy was performed. The functional outcome was evaluated with the visual analog scale for pain at walking, running, stair climbing, quality of life at the time of implant removal and at a mean follow-up of 25.8 months. Activity was assessed with the Tegner scale. Knee function was evaluated with the Lysholm score. Results In one case, the osteochondral cylinder did not heal and an osteochondral fragment was removed arthroscopically. In all other cases, the osteochondral cylinder W. Petersen (*) · P. Taheri · A. Achtnich · C. Winter · P. Forkel Department of Orthopaedic and Trauma Surgery, Martin Luther Hospital, Grunewald, Caspar Theyss Strasse 27‑31, 14193 Berlin, Germany e-mail: w.petersen@mlk‑berlin.de B. Schliemann Department of Trauma‑, Hand‑ and Reconstructive Surgery, University Hospital Muenster, Muenster, Germany
was stable with surrounding cartilage. The average ICRS Cartilage Repair Assessment was 10.1 points (±1.3). All malleolar osteotomies healed radiologically. In 15 patients, a synovectomy and local debridement of the ankle were performed at second-look arthroscopy. Ankle pain at walking, running and stair climbing as measured by a visual analog scale (10-0) decreased significantly from preoperatively to the first follow-up (mean 12.6 months) and to the second follow-up (mean 25.8 months). The ankle-related quality of life increased significantly from preoperatively to postoperatively. There was no significant change in the Lysholm score. The activity measured with the Tegner activity scale increased significantly from preoperatively to the last follow-up, but only two out of nine patients continued pivoting sports. Conclusions Autologous osteochondral grafting with the Diamond twin system is a reliable treatment option for symptomatic osteochondral defects of the talus. After 1 year, the majority of patients had still some complaints. However, after screw removal and second-look arthroscopy, the pain and ankle-related quality of life further improved. Clinical relevance The donor site morbidity after graft harvesting from the posterior aspects of the femoral condyles is lower than previously reported. Keywords Osteochondral lesion · Tricalcium phosphate · Ceramic bone substitute · Activity · Donor site morbidity · Osteoarthritis · Cartilage repair · Subchondral bone plate
Introduction Osteochondral grafting is a treatment option for stage IV osteochondral defects of the talus. Several techniques for osteochondral grafting have been described in the literature
13
844
[1, 2, 4, 6–8, 10, 11, 13, 15, 18, 22, 25–27]. All techniques have in common that cylindrical osteochondral grafts are harvested from the knee to reconstruct a defect at the talar dome. With the so-called mosaicplasty technique, several small cylinders are used to cover the defect [6–8, 25]. This technique can be used when the defect is predominantly chondral. Otherwise the cylinders are anchored in necrotic bone. With other systems such as with the OATS, SDS or Diamond twin system one or two larger osteochondral cylinders are transferred [4, 10, 14, 22]. With these techniques, good clinical results have been reported for osteochondral transplantation at the talus [1, 2, 4, 6–8, 10–12, 14, 17, 22, 25–27]. For the Diamond twin system, clinical results for the treatment of osteochondral defects of the talus have not been described until know. Most authors use the femoropatellar joint as donor site for the osteochondral grafts [1, 2, 4, 6–8, 10, 11, 13, 15, 18, 22, 25–27]. In our experience, the harvesting of osteochondral grafts from the femoropatellar joint may be associated with donor site morbidity such as anterior knee pain [14]. Several other authors could confirm and described significant donor site morbidity after graft harvesting from the femoropatellar joint [4, 19, 21, 25–27]. An alternative donor site for osteochondral grafts is the posterior aspect of the femoral condyles [14]. Good-toexcellent clinical results with low donor site morbidity have been described for the treatment of osteochondral defects of the knee [14]. A disadvantage of this technique was that the cylindrical grafts were harvested with the patient in prone position. That means, the patient has to be turned around with new skin disinfection and draping to implant the graft into the defect. Because of these disadvantages we modified the original technique. With the modified technique, the osteochondral grafts are harvested via a small horizontal incision in the popliteal fossa with the patient in lateral decubitus position (Fig. 1). To reach the ankle without new skin disinfection the leg has to be rotated in the hip. The aim of this study is to analyze radiological, arthroscopical, and clinical results after osteochondral transplantation for defects at the talus using the Diamond twin system. The grafts were harvested from the posterior femoral condyles of the ipsilateral knee. Our hypothesis was that with this technique, the donor site morbidity after graft harvesting from the posterior aspects of the femoral condyles is lower than previously reported.
Patients and methods We prospectively followed 20 patients who had been treated by autologous osteochondral grafting for symptomatic
13
Arch Orthop Trauma Surg (2014) 134:843–852
osteochondral defects of the talus between 2008 and 2011. The study was approved by the institutional review board. All patients gave written informed consent to be included in a scientific study. Twelve men and eight women were treated. The mean age was 25.4 years (18–44 years). Inclusion criteria for the study were: Age >18 years and a stage III or IV osteochondral defect at the talar dome (ICRS classification). All patients had stage IV osteochondral defect of the talus according to the ICRS classification. The osteochondral defect was located laterally in one and medially in 19 patients. The lesions were diagnosed by plain radiography and MRI. The mean size of the defects of the talar dome was 16.95 (±2.5) × 10.4 (±1.1) mm. All patients had previous unsuccessful arthroscopic surgery with excision of the fragment and microfracture. Mean time to previous surgery was 17 months (range 3–46 months). Surgical technique All surgical procedures were performed under general anesthesia. The defect size was determined on the preoperative MRI. The operation starts with graft harvesting at the knee. Therefore, precise information about the defect size is necessary to determine the number of grafts needed to cover the defect. For osteochondral cylinder transplantation the Diamond twin system® (Karl Storz, Tuttlingen, Germany) was used (Fig. 1). This system consists of various diamond bone cutters. The mutually adjusted sizes of the bone cutters allow a press-fit implantation. The donor cylinder is 0.05 mm larger in diameter than the defect cylinder. Graft harvesting The operation starts with harvesting of the osteochondral donor cylinders from the posterior femoral condyles. To reach the donor site the patient was placed in a lateral position (Fig. 1). After application of a thigh tourniquet the skin was disinfected and draped. A transverse skin incision was performed in the fossa poplitea. The fascia was opened with a scissor and the posterior joint capsule of the lateral femoral condyle was exposed with four Langenbeck retractors after blunt preparation with the index finger. The lateral gastrocnemius muscle must be pushed to the side with a periosteal elevator. Then, the capsule was opened with a scalpel and a Hohmann retractor was placed behind the femoral condyles. The cartilage is cut with an appropriately sized extractor and then the bone is cut with the diamond cutter. To prevent heat damage the cutter has to be cooled with NaCl. The depth of the donor cylinder should be between 15 and 20 mm according to the size of the defect.
Arch Orthop Trauma Surg (2014) 134:843–852
845
Fig. 1 This figure shows the operative technique for graft harvesting from the posterior femoral condyles. a The patient is positioned on the contralateral side and the posterior femoral condyles are exposed via a horizontal incision. b The graft cylinder is cut with a diamond
bone cutter under continuous cooling with NaCl. c The cylinder is extracted with a specific extractor. d The harvest defect is filled with a ceramic bone substitute made of tricalcium phosphate (Synthricer, Karl Storz, Germany)
The depth can be read on the tiller. The donor cylinder is finally extracted with an extractor by a quick rotation. After rinsing, the defect is filled with a cylindrical bone graft substitute optimized for use with the instruments (Syntricer®—biodegradable β-tricalcium phosphate ceramic, Karl Storz, Tuttlingen). The cylindrical bone graft substitute is gently pressed into the donor defect until the cylinder surface is 1 or 2 mm below the surrounding cartilage (Fig. 1). The donor site is rinsed with NaCl and the capsule is closed by resorbable sutures. If a second cylinder is needed the same procedure has to be repeated at the medial side. In 15 cases, two cylinders were used.
protect the posterior tibial tendon from damage of the oscillating saw. The osteotomy plane was marked by two K wires with the use of a fluoroscope and the osteotomy was performed with an oscillating saw and a chisel. At the lateral side an oblique distal fibular osteotomy was performed below the syndesmosis. The defect was prepared by excising the necrotic sequestrum with a bone cutter one size smaller as used for harvesting of the donor cylinders (Fig. 2). The technique for defect preparation is the same as for harvesting the donor cylinder. The depth of the defect hole has to be measured and in case of a mismatch the donor cylinder should be shortened. The donor cylinders were pressed manually into the recipient talar site to fill the defect. In 17 patients, two cylinders were used and in four patients only one cylinder was used. The diameter of the cylinders ranged between 9.45 and 14 mm. In case of two cylinders, they overlap. The osteotomy was rigidly fixed using either titanium screws (medial malleolus) or with an AO 1/3 tubular plate for the fibula.
Osteochondral transplantation at the talus In 19 cases, a medial malleolar osteotomy was performed (Fig. 2). In one case, a lateral fibular osteotomy was performed. The medial malleolus was exposed by two Hohmann retractors after skin incision. The posterior retractor should
13
846
Arch Orthop Trauma Surg (2014) 134:843–852
Fig. 2 a The osteochondral defect is exposed via a malleolar osteotomy. b The defect cylinder is excised with the diamond-coated grinder and c the extractor. Both instruments are one size smaller than that used for graft harvesting. d In this case, the defect is completely filled with two osteochondral donor cylinders
Rehabilitation protocol The patients were mobilized with partial weight-bearing for 6 weeks and with early physiotherapy to both the ankle and the knee. Unprotected weight-bearing was allowed after a radiological control. CPM for the ankle was used for the first 6 weeks. Sensomotoric training started after 8 weeks postoperatively. Jogging was allowed after 6 months and impact activities after screw removal after 1 year. Follow‑up All peri- or postoperative complications were documented. One follow-up examination was performed after a mean of
13
12.6 months and a second follow-up was performed at a minimum of 24 months (mean 25.8 months) after surgery. The first follow-up was the time of implant removal which was performed in all patients. In 15 cases, a secondlook arthroscopy was performed at this time point. The indication for second-look arthroscopy was ankle pain. The arthroscopic findings have been documented (synovitis, adhesions). The quality of the grafts was assessed using the International Cartilage Repair Society (ICRS) cartilage repair assessment which is a semiquantitative outcome measure [23]. This score evaluates the degree of cartilage repair, integration to the border zone and macroscopic appearance. In each category, 4 score points can be assigned. By adding these points the overall score is calculated (12 points: normal, 11–8 points: nearly normal, 7–5
Arch Orthop Trauma Surg (2014) 134:843–852
847
points: abnormal, 4–0 points: severely abnormal). In all cases, a conventional X-ray of the ankle and the knee was performed to examine healing of the malleolar osteotomy and the defect site before. At the 12.6-month follow-up the knee and ankle were examined for range of motion, stability, local tenderness and crepitation. Subjective pain at the ankle was evaluated using a visual analog scale (VAS, 10 always pain, 0 never pain) preoperatively, at the 12.6-month and at the 25.8-month follow-up. Pain was assessed for walking on even ground, walking on uneven ground, running, stair climbing. Ankle-related quality of life was also assessed using a visual analog scale (VAS, 10 = 100 % reduction of life quality, 0 = no influence on life quality) preoperatively, at the 12.6-month and at the 25.8-month follow-up. For quality of life assessment the patients were asked: How much is your quality of life affected by your ankle problems. Knee function was assessed with the Lysholm score [16]. The activity was assessed with the Tegner activity scale [9, 24]. At the 25.8month follow-up, the patients were asked to rate themselves on the German version of the Tegner scale before start of the complaints, immediately before surgery and at the 25.8-month follow-up. They were further asked which sports they performed before start of the complaints, immediately before surgery and at the 25.8-month follow-up. Statistics The statistical analysis was performed at the Medistat Institute in Kiel, Germany. The Shapiro–Wilk test was used to test if the measurements were normal distributed. For nonparametric parameters, nonparametric test was used for statistical analysis of group differences. For global temporal changes the Friedman test was used. The Wilcoxon test (p 5° of flexion or extension in comparison to the contralateral side). Fourteen patients had local tenderness at the medial malleolus. In ten cases, there was local tenderness over the posterior tibial tendon behind the medial malleolus. No patient had any local tenderness at the knee joint. Range of motion of the knee was free in all patients with no crepitation.
13
848
Arch Orthop Trauma Surg (2014) 134:843–852
Fig. 4 Postoperative MRI after 11 months of the patient with a broken bone cylinder (see Fig. 3a)
At second-look arthroscopy, adhesions and synovitis in the anterior ankle compartment were seen in all 15 patients. A partial synovectomy and arthrolysis were performed in all these patients. In one patient we observed a broken graft cylinder. The broken graft was removed arthroscopically in this case. The broken cylinder was also diagnosed on a preoperative MRI (Fig. 4). In all other patients the graft was stable at second-look arthroscopy (Fig. 5). In 5 cases, the graft level was raised in comparison of the surrounding cartilage, in two cases the graft level was below the surrounding cartilage, and in eight cases the surface of the graft cylinder had the same level as the surrounding cartilage. The overall ICRS arthroscopic score for the graft was 10.1 (±1.3). The most common findings were fibrillated surface of the graft cylinders. Two patients had a score of 7 which is considered as abnormal. Both patients were also clinically considered as failures because in both patients there was no significant improvement in pain. Ankle pain at walking, running, jumping and stair climbing assessed with a visual analog scale (VAS, 10 always pain, 0 never pain) decreased significantly from the preoperative status to both follow-ups (Fig. 6a–c). Ankle-related quality of life also improved significantly (Fig. 7a). Activity as rated with the Tegner activity scale is shown in Fig. 7b. Before the beginning of symptoms nine patients performed pivoting sports (soccer, handball, basketball). At the last follow-up, only two patients performed pivoting sports.
13
Fig. 5 Arthroscopic findings at second-look arthroscopy. a Osteochondral cylinder with a smooth surface which is in level with the surrounding cartilage. b In this patient the graft level is raised in comparison to the superficial cartilage. c This cylinder has superficial fibrillations but the graft is in level with the surrounding cartilage
Arch Orthop Trauma Surg (2014) 134:843–852
849
◂ Fig. 6 a The Shapiro–Wilk test showed a nonparametric distribution
of the measurements with the VAS for pain at walking (p
