- Email: [email protected]
Randomized Controlled Trial Comparing the Radiologic Outcomes of Conventional and Minimally Invasive Techniques for Total Knee Arthroplasty
The Journal of Arthroplasty Vol. 22 No. 6 2007
Randomized Controlled Trial Comparing the Radiologic Outcomes of Conventional and Minimally Invasive Techniques for Total Knee Arthroplasty Pak Lin Chin, MBBS (Singapore), MRCSEd, MMed (Ortho) (Singapore), Leon Siang Shen Foo, MBBS (Singapore), MRCSEd, MMed (Ortho) (Singapore), Kuang Ying Yang, MBBS (Singapore), FRCS (Edin), Seng Jin Yeo, MBBS (London), FRCS (Edin), FAMS, and Ngai Nung Lo, MBBS (Singapore), FRCS (Edin), MMed (Surgery) (Singapore), FAMS Abstract: The objective of this study was to compare the radiologic outcomes of total knee arthroplasty using the conventional technique with those using minimally invasive surgery (MIS) techniques. Ninety patients were randomized to undergo conventional (control), MIS mini-incision mid-vastus (mini), or MIS sidecutting (side cut) techniques for their total knee arthroplasty. Long-leg radiographs were assessed postoperatively. The mean overall limb varus alignments were 1.038 (SD, 2.588) for the control group, 0.878 (SD, 2.968) for the mini group, and 0.378 (SD 4.278) for the side cut group. The mean overall limb alignments within F3.08 varus/ valgus were 83.3%, 83.3%, and 56.7%, respectively. Femoral implant placement ( P = .028) and overall limb alignment ( P = .024) in the side cut group were significantly poorer as compared with those in the control group. The side cut group also had more outliers in the coronal plane. Results were comparable between the mini and control groups. The side cut technique appears to affect the accuracy of implant placement. Key words: MIS-TKA, Randomized controlled trial, Radiographic alignment. n 2007 Elsevier Inc. All rights reserved.
and controversial orthopedic topic is that of minimally invasive surgery (MIS) for TKA. Minimally invasive surgery is based on the concept of causing as minimal surgical trauma as possible to reduce postoperative pain and facilitate the body’s healing, resulting in earlier functional recovery. This concept is not new to knee arthroplasty. Minimally invasive surgery has already been proposed for unicondylar knee arthroplasty. It is only logical for MIS to be applied to TKA as technology progresses and surgical experience develops [2]. With the extent of surgical trauma aside, the success of a TKA is dependent on several other factors, chief among which is correct component alignment. Malalignment can lead to abnormal patellar tracking, increased polyethylene wear, early loosening, and poor functional outcome. There are concerns that obtaining correct compo-
The existing research evidence suggests that total knee arthroplasty (TKA) is a safe and cost-effective treatment for alleviating pain and restoring physical function in patients who do not respond to nonsurgical therapies [1]. As surgical experience with TKA increases, new operative techniques to expedite functional recovery and improve clinical outcomes are constantly being proposed. A current From the Department of Orthopedic Surgery, Singapore General Hospital, Singapore City, Republic of Singapore. Submitted December 9, 2005; accepted October 3, 2006. No benefits or funds were received in support of this study. Reprint requests: Pak Lin Chin, MBBS (Singapore), MRCSEd, MMed (Ortho) (Singapore), Department of Orthopedic Surgery, Singapore General Hospital, Block 6, Level 7, Outram Road, Singapore City 169608, Republic of Singapore. n 2007 Elsevier Inc. All rights reserved. 0883-5403/07/1906-0004$32.00/0 doi:10.1016/j.arth.2006.10.009
800
Radiographic Alignment ! Chin et al 801
nent alignment may be difficult with the currently proposed minimally invasive techniques for TKA because of poor surgical access. This may translate eventually to poorer functional outcomes and increased failure rates. Hence, the aim of this study was to compare the radiologic outcomes of the conventional technique with those of minimally invasive techniques for TKA.
Materials and Methods From March to October 2004, 90 consecutive patients with osteoarthritis of their knees requiring TKA were enrolled into this single-blind, prospective, randomized, and controlled trial. Patients with genu valgus, previous knee surgery (including high tibial osteotomy and previous knee arthroplasty), an active infection, or malignancy and those deemed medically unfit for surgery were excluded. The study protocol was approved by our hospital’s ethics committee. Each patient was counseled and provided with patient information sheets regarding the conduct of the trial. After we obtained informed consent from the patients, we allocated them to undergo the standard medial parapatellar technique (control; n = 30), the minimally invasive mini-incision midvastus technique (mini; n = 30), or the minimally invasive quadriceps-sparing technique (side cut; n = 30) for their TKA. An independent statistician designed the randomization table. Concealment of allocation was by sealed envelopes, the contents of which were only made known to the operating surgeons at the start of each surgery. The patients, research assistants, physiotherapists, and statisticians were all blinded to the allocations. All patients were seen before surgery, and their demographic characteristics, including age, sex, height, weight, and body mass index, were recorded. A senior anesthetist also assessed them preoperatively to determine their fitness for surgery. All surgeries were performed by the 2 senior surgeons (NNL and SJY), each of whom had performed at least 200 TKAs in the past year. This study was preceded by the surgeons familiarizing themselves thoroughly with the various minimally invasive TKA techniques. This process included cadaveric workshops on 50 cases with the side cut technique to help minimize performance bias. For all techniques used, the manufacturers’ recommended techniques were followed. The standard medial parapatellar technique (control) involved an anterior incision using a capsular
approach that separates the interval between the rectus femoris and the vastus medialis. It allowed for the patella to be everted for better joint visualization and implant placement. The suprapatellar pouch was also dissected free for prosthesis sizing. This was followed by intramedullary femoral and extramedullary tibial instrumentations. Standard cutting blocks were then used for the bone cuts. Two minimally invasive techniques were used for comparison: the mini technique and the side cut technique. The minimally invasive mini technique was carried out as described by Laskin [3,4]. The suprapatellar pouch was minimally disrupted. Zimmer NexGen MIS Mini (Zimmer Inc, Warsaw, Ind) and Johnson & Johnson DePuy Pathway (DePuy Orthopedics Inc, Warsaw, Ind) instrumentation sets were used in this mini study arm. These mini instrumentation sets were essentially miniaturized versions of conventional cutting jigs. Their miniaturization however allowed for less extensive soft tissue dissections. The other minimally invasive technique used was the side cut technique, with which a capsular approach as described by Gesell and Tria [5] was used. Significantly, the extensor mechanism and the suprapatellar pouch were not violated. A
Fig. 1. Angles measured for assessment: angle A, mechanical axis of the lower limb vs the femoral component alignment; angle B, mechanical axis of the lower limb vs the tibial component alignment; angle C, angle bisecting the mechanical axis of the femur and the mechanical axis of the tibia; angle F, angle of femoral component flexion; angle T, posterior slope of the tibial component.
802 The Journal of Arthroplasty Vol. 22 No. 6 September 2007 Table 1. Patients’ Profiles
Age (y) Male/female ratio Body mass index (kg/m2) Total operation time (min) Total blood loss (mL) Length of inpatient hospitalization stay (d)
Control group
Mini group
Side cut group
63.4 (47-80) 3:27 29.44 (22.7-40.0) 92.0 (55-120) 40.3 (30-980) 6.1 (2-10)
67.4 (56-80) 6:24 28.50 (22.1-40.0) 111.7 (65-150) 411.8 (40-1310) 6.0 (4-15)
69.0 (57-80) 6:24 27.53 (18.6-34.2) 111.5 (75-145) 357.8 (30-1040) 6.1 (3-16)
Values in brackets are ranges. All P values, except for total operative time, are not significant ( P N .05) with analysis of variance.
specially designed Zimmer NexGen MIS quadriceps-sparing instrumentation (Zimmer Inc) was used. This instrument was characterized by miniaturization and side-cutting jigs. The knee replacement implants used in this trial included the Zimmer NexGen Complete Knee Solution Legacy posterior-stabilized knee system (Zimmer Inc) and the Johnson & Johnson DePuy PFC Sigma fixed-bearing cruciate-substituting knee system (DePuy Orthopedics Inc) for the control and mini study arms. As for the side cut study arm, only the Zimmer NexGen Complete Knee Solution Legacy posterior-stabilized knee system (Zimmer Inc) was used. The patella was resurfaced in every patient. A Palacos G antibiotic-loaded bone cement (BioMet Orthopedics, Inc, Warsaw, Ind) was used in all the cases. Antibiotic prophylaxis (2 g of intravenous cefazolin) was administered 30 minutes before the commencement of surgery. This was administered provided that there was no contraindication. If a patient was allergic to cefazolin, 1 g of intravenous vancomycin was administered instead. This antibiotic prophylaxis was continued for 24 hours after the operation. All patients had a Redivac drain inserted intraoperatively, which was removed 24 hours after the operation. The total blood loss sustained intraoperatively and that at 24 hours after the surgery were calculated. The total operating time was also recorded. Postoperative pain relief was administered via patient-controlled analgesia using intravenous
morphine for the first 24 hours after the operation. Oral analgesia was commenced 24 hours after the operation. This included 1 g of paracetamol every 6 hours, 550 mg of naproxen every 12 hours, and 20 mg of omeprazole every 12 hours. If a patient was allergic to these medications, 50 mg of tramadol every 8 hours was served instead. All patients underwent a standardized postoperative physiotherapy program. This involved continuous passive movements and active exercises as soon as possible. The total length of inpatient hospital stay was recorded, and any postoperative complication was documented by blinded research assistants. All patients had a standardized series of plain x-rays taken when they were able to weight bear without any flexion deformity; these included long-leg standing anteroposterior and lateral x-rays of the limbs [6-8]. All radiographs were performed at the same radiology center. Radiologic data were collected separately and independently on 3 occasions by 3 research assistants [7]. All 3 research assistants were blinded to patient allocations. The following radiologic data were collected (reference is drawn to the article by Chin et al [6] regarding the measurement of radiologic angles in postoperative longleg standing x-rays of patients with total knee replacements; Fig. 1): 1. Mechanical axis of the lower limb vs the femoral component alignment (angle A) 2. Mechanical axis of the lower limb vs the tibial component alignment (angle B)
Table 2. Angles in the Coronal Axis Control group Angle A (femur implant) Angle B (tibial implant)
86.0 to 91.5 1.5 to 4.0 (1.17 varus) 88.0 to 97.0 7.0 to 2.0 (0.77 valgus)
Mini group 84.0 to 93.0 3.0 to 6.0 (1.28 varus) 87.0 to 98.0 8.0 to 3.0 (1.12 valgus)
Side cut group 85.0 to 92.0 2.0 to 5.0 (1.64 varus) 88.0 to 98.0 8.0 to 2.0 (1.91 valgus)
Values are presented as degrees. All P values are not statistically significant ( P N .05) with analysis of variance.
Radiographic Alignment ! Chin et al 803 Table 3. Good Outcomes for Angles in the Coronal Axis Good outcome
n (%)
P
Angle A (femur implant) Control group
– 27 (90)
Mini group
.448
.519
25 (83.3) Side cut group
.166 23 (76.7)
Angle B (tibial implant) Control group
Mechanical Axis of the Lower Limb on Coronal X-Ray Measurements (Angle C)
– 27 (90)
Mini group
.448
.222
25 (83.3) .053
Side cut group 21 (70.0)
3. Angle bisecting the mechanical axis of the femur and the mechanical axis of the tibia (angle C) 4. Angle of femoral component flexion (angle F) 5. Posterior slope of the tibial component (angle T) Statistical Analysis All data were collected at the end of the trial, and the patient groups were blinded for further studies. A blinded statistician (Stephanie Fook-Chong) who was independent of the surgical and research team analyzed the data using SPSS 13.0 for Windows (SPSS Inc, Chicago, Ill). Pearson’s v 2 analysis, Fisher’s exact test, and analysis of variance between groups were performed as appropriate.
Results
Femoral and Tibial Component Alignments on Coronal X-Ray Measurements (Angles A and B) The results showed that the control and mini study arms had more consistent femoral and tibial Table 4. Angle C (Mechanical Axis of the Femur vs Mechanical Axis of the Tibia) in the Coronal Axis Alignment 6.0 to 6.0 (1.03 varus) 4.0 to 7.0 (0.87 varus) 8.0 to 8.0 (0.37 varus)
All 3 study arms had a mean varus alignment. The control and mini study arms had significantly more patients with good outcome measurements for the mechanical axis of the lower limb as compared with the side cut study arm (Tables 4 and 5). Femoral Component Flexion and Tibial Component Posterior Slope on Sagittal X-Ray Measurements (Angles T and F) All 3 study arms gave consistent tibial and femoral component alignments on the sagittal x-ray measurements (Table 6). Total Operation Time Operations performed with the standard medial parapatellar technique (control) took a shorter total time as compared with the 2 minimally invasive techniques ( P b .001). There was no statistical difference in the total operating time between the 2 minimally invasive study arms. Total Blood Loss
The patients’ profiles are listed in Table 1. The cohort was homogeneously distributed among the 3 study arms.
Standard Mini Side cut
component alignments with less outliers as compared with the side cut study arm (Table 2). The proportion of patients with good outcomes for femoral and tibial component alignments was also higher for the control and mini study arms as compared with the side cut study arm (Table 3). Descriptively, the groups ranked as follows: control group = mini group N side cut group.
Table 5. Good Outcomes of Angle C (Mechanical Axis of the Femur vs Mechanical Axis of the Tibia) Good outcome
P
n (%)
P
Control group
– 1.000 .597
Total blood loss was less in the minimally invasive technique study arms as compared with the control study arm ( P = .015). Three of the patients in the control study arm (10%) required postoperative blood transfusions. None of the patients in either minimally invasive technique group needed any blood transfusion. Between the 2 minimally invasive technique groups, the side cut
.127
Values are presented as degrees. Statistical analysis was conducted with analysis of variance.
– 25 (83.3)
Mini group
1.000 25 (83.3) .024
Side cut group 17 (56.7)
.0240
804 The Journal of Arthroplasty Vol. 22 No. 6 September 2007 Table 6. Angles in the Sagittal Axis
Angle F (femoral flexion) Angle T (tibial slope)
Control group
Mini group
Side cut group
5.0 to 10.0 (3.17 flexion) 2.0 to 10.0 (4.95 posterior slope)
0 to 5.0 (2.75 flexion) 0 to 12.0 (4.17 posterior slope)
0 to 10.0 (3.05 flexion) 0 to 7.0 (3.55 posterior slope)
Values are presented as degrees. All P values are not significant ( P N .05) with analysis of variance.
study arm had less blood loss as compared with the mini study arm, although this difference was not statistically significant ( P = 1.000). Length of Inpatient Hospitalization Stay There was no statistical difference in the length of inpatient hospitalization stay among the 3 study arms. This finding may be unique to our local medical insurance coverage terms, for which there is no significant financial impetus to keep inpatient hospitalization stays as short as possible. Postoperative Complications Two patients who underwent the standard medial parapatellar technique (control) knee arthroplasty developed proximal deep vein thrombosis, which was treated expectantly with antithrombotic therapy. One patient who had a side cut operation developed superficial wound infection, which was successfully treated with 2 weeks of intravenous antibiotics. One patient in each minimally invasive study arm developed postoperative knee stiffness that required manipulation under general anesthesia 3 months after the operation. There was no mortality case seen in any of the 3 study arms.
Discussion Minimally invasive surgery has become a popular topic among orthopedic surgeons and among patients. Increasing numbers of patients are asking for an MIS after reading about its proposed advantages of less postoperative pain, quicker functional recovery, and better cosmesis (shorter scars). Although the concept behind MIS appears to be attractive, there is first a need to ascertain its efficacy with well-designed and well-administered clinical trials before proclaiming it as the next standard of care. There is also a need for consensus among orthopedic surgeons as to what exactly constitutes an MIS. Minimally invasive surgery does not necessarily equate with shorter wound incisions; it is best appreciated as a surgical approach that results in less
intraoperative surgical trauma. This in turn reduces postoperative pain, facilitates early rehabilitation, and enhances clinical and functional recovery. Two of the more commonly described minimally invasive techniques for TKA are the mini technique and the side cut technique. Both techniques cite minimal disruption of the quadriceps mechanism as their advantage, with each technique having its own proponents. The minimally invasive mini technique has been described as an efficacious alternative to the standard medial parapatellar quadriceps-splitting approach that still allows for satisfactory surgical exposure. On the other hand, the side cut technique has been advocated as a useful technique to completely avoid violating the quadriceps mechanism and the suprapatellar pouch. However, there are concerns about the adequacy of surgical access and the accuracy of bone cuts with the customized side-cutting jigs used with the side cut technique. This in turn may affect component alignment, resulting in abnormal wear and eventual clinical failure. Tria [9] reported on favorable early results from his series of 120 minimally invasive TKAs. His results indicated that minimally invasive TKAs were associated with less blood loss, less postoperative pain, better early motion, and a shorter hospital stay as compared with the conventional techniques, with no compromise of accuracy. Laskin et al [10] performed 32 minimally invasive TKAs using the mini approach. They reported that the average visual analog pain score and the total amount of pain medication among the patients with minimally invasive TKAs were lower as compared with those among the patients who had a standard medial parapatellar technique surgery. In addition, they found that improvements in the Knee Society score 6 months after the operation were statistically higher as compared with those in the control group. They attributed this to limited disruption of the extensor mechanism and the suprapatellar pouch, allowing for more rapid restoration of quadriceps muscle control. They also noted that the radiographic alignment and position of implant components were satisfactory in all the patients from both groups.
Radiographic Alignment ! Chin et al 805
Similar results were reported by Haas et al [11] with their series of 40 minimally invasive TKAs performed via the mini technique. However, there is, to our knowledge, no published article describing a 3-arm prospective randomized controlled trial comparing the radiologic outcomes of the standard medial parapatellar technique with those of the minimally invasive techniques of the mini and side cut approaches. Furthermore, most of the earlier studies on minimally invasive TKA were cohort studies and not prospective randomized controlled studies [12-18]. Radiologic measurements in these earlier studies were also often based on short knee films and not long-leg standing x-rays, which give more accurate radiologic measurements. Computed tomographic scan assessment can be very enticing; however, radiographs, which can be done with the patient standing, give a better reflection of the situation. The MIS techniques needed a longer operative time because the minimal surgical exposure needed more time for careful placement of instrument guides and implant placement. Moreover, with the small incision, intuitive optimization of the moving surgical window was required, in which the knee is ranged more frequently, hence accounting for longer surgical duration. This is similar to previous authors’ experience [2-4]. The most clinically relevant radiologic measurement is the overall limb alignment against the mechanical axis of the lower limb, angle C. A varus/valgus malalignment greater than 38 has been shown to lead to early failure [6]. The subanalysis of the placement of the femoral component (angle A) and the tibial component (angle B) would help identify the deficiencies in the particular minimally invasive technique, be it in the femoral or the tibial component alignment. Our results showed that the femoral and tibial component positions have comparable error risks (Table 3). This may be a result of the decreased exposure and difficulty in placing the cutting guides close to the bone in the ideal position. This was most evident in the side cut group. The sagittal-axis assessments of good outcomes are arbitrarily set and, to our knowledge, have not been investigated thoroughly, along with their bearing on survivorship. We feel that the outcome will not be compromised as long as the femoral and tibial components are not placed in extension. Our results suggest that the mini technique appears to give radiologic results comparable with those of the standard medial parapatellar technique. The side cut technique, although generally giving a satisfactory overall standing femorotibial lower limb alignment, does show significant vari-
ance in the accuracy of the implant component alignment. This we postulate is related to the less ideal surgical access provided for by the side cut technique. There is poor line of sight, making the operation more technically difficult. We also had frequent difficulties in ensuring accurate bone cuts with the customized sidecutting jigs used with the side cut technique. This was especially true for bone cuts to the lateral femoral condyle and tibial plateau. This was because the side-cutting jigs were more prone to displacement and loosening during sawing of the bone. It is important to keep the jigs well secured. We found it useful to keep the intramedullary guide rod within the femur to maintain alignment of the femoral side-cutting jig. The use of multiple pins, particularly threaded pins, to secure the jigs in place is also advisable. Another consideration is the need for the saw blade to transverse greater distances when using the side cut technique. This increased bone-cut transversing distance magnifies any error that may have resulted from incorrect jig placement or loosening of the jigs from the jig-bone interfaces. Minimally invasive surgery for TKA is still in its infancy. There is a steep learning curve, and much research work remains to be done. Based on our experience and the difficulties encountered in the cited cases, we recommend that efforts be made to design more user-friendly, more secure, and more accurate instrumentations. This is especially true for side-cutting jigs. Further miniaturization of instrumentations and particularly that of the prosthesis would be most welcome. Currently, incision length is largely restricted by the size of the prosthesis.
Conclusions The side cut technique but not the mini technique appears to have a significant negative impact on the accuracy of TKA. The errors may be caused by intrinsic problems with the side cut technique, which include restricted surgical access with limited line of sight, the need for the saw blade to transverse a longer bone-cut distance, magnifying any cutting jig-bone alignment error, and inadequately rigid cutting jig-bone fixation, causing loss of cutting jig alignment. Total knee arthroplasty is a very successful surgery. It is imperative that outliers or surgical errors be minimized. Having good functional outcome using new techniques at the expense of poor consistency in implant placement is clinically unacceptable. The ultimate aim is to align the knee along the mechanical axis. Studies
806 The Journal of Arthroplasty Vol. 22 No. 6 September 2007 have shown that a malalignment greater than F38 will lead to early failure [19]. Additional caution needs to be exercised when using minimally invasive side cut techniques to avoid these pitfalls.
Acknowledgments We thank Ms Stephanie Fook-Chong of the Singapore General Hospital Clinical Trials Center for her invaluable assistance with the statistical analysis.
References 1. NIH Consensus Panel. NIH consensus statement on total knee replacement December 8-10, 2003. J Bone Joint Surg Am 2004;86-A:1328. 2. Tria Jr AJ, Coon TM. Minimal incision total knee arthroplasty: early experience. Clin Orthop Relat Res 2003;185. 3. Laskin RS. Minimally invasive total knee replacement using a mini–mid vastus incision technique and results. Surg Technol Int 2004;13:231. 4. Laskin RS. Minimally invasive total knee arthroplasty: the results justify its use. Clin Orthop Relat Res 2005;440:54. 5. Gesell MW, Tria Jr AJ. MIS unicondylar knee arthroplasty: surgical approach and early results. Clin Orthop Relat Res 2004;53. 6. Chin PL, Yang KY, Yeo SJ, et al. Randomized control trial comparing radiographic total knee arthroplasty implant placement using computer navigation against conventional technique. J Arthroplasty 2005;20:618. 7. Perlick L, Bathis H, Perlick C, et al. Revision total knee arthroplasty: a comparison of postoperative leg alignment after computer-assisted implantation versus the conventional technique. Knee Surg Sports Traumatol Arthrosc 2005;13:167 [Epub 2004 Apr].
8. Sparmann M, et al. Positioning of total knee arthroplasty with and without navigation support. A prospective, randomised study. J Bone Joint Surg Br 2003;85:830. 9. Tria Jr AJ. Advancements in minimally invasive total knee arthroplasty. Orthopedics 2003;26(8 Suppl); s859. 10. Laskin RS, Beksac B, Phongjunakorn A, et al. Minimally invasive total knee replacement through a mini-midvastus incision: an outcome study. Clin Orthop 2004;74. 11. Haas SB, Cook S, Beksac B. Minimally invasive total knee replacement through a mini midvastus approach: a comparative study. Clin Orthop 2004;68. 12. Dalury DF, Jiranek WA. A comparison of the midvastus and paramedian approaches for total knee arthroplasty. J Arthroplasty 1999;14:33. 13. White Jr RE, Allman JK, Trauger JA, et al. Clinical comparison of the midvastus and medial parapatellar surgical approaches. Clin Orthop 1999;117. 14. Keating EM, Faris PM, Meding JB, et al. Comparison of the midvastus muscle–splitting approach with the median parapatellar approach in total knee arthroplasty. J Arthroplasty 1999;14:29. 15. Weinhardt C, Barisic M, Bergmann EG, et al. Early results of subvastus versus medial parapatellar approach in primary total knee arthroplasty. Arch Orthop Trauma Surg 2004;124:401 [Epub 2004 May 07]. 16. Matsueda M, Gustilo RB. Subvastus and medial parapatellar approaches in total knee arthroplasty. Clin Orthop 2000;161. 17. Cila E, Guzel V, Ozalay M, et al. Subvastus versus medial parapatellar approach in total knee arthroplasty. Arch Orthop Trauma Surg 2002;122:65. 18. Faure BT, Benjamin JB, Lindsey B, et al. Comparison of the subvastus and paramedian surgical approaches in bilateral knee arthroplasty. J Arthroplasty 1993; 8:511. 19. Ritter MA, et al. Postoperative alignment of total knee replacement. Its effect on survival. Clin Orthop 1994;153.
Randomized Controlled Trial Comparing the Radiologic Outcomes of Conventional and Minimally Invasive Techniques for Total Knee Arthroplasty Pak Lin Chin, MBBS (Singapore), MRCSEd, MMed (Ortho) (Singapore), Leon Siang Shen Foo, MBBS (Singapore), MRCSEd, MMed (Ortho) (Singapore), Kuang Ying Yang, MBBS (Singapore), FRCS (Edin), Seng Jin Yeo, MBBS (London), FRCS (Edin), FAMS, and Ngai Nung Lo, MBBS (Singapore), FRCS (Edin), MMed (Surgery) (Singapore), FAMS Abstract: The objective of this study was to compare the radiologic outcomes of total knee arthroplasty using the conventional technique with those using minimally invasive surgery (MIS) techniques. Ninety patients were randomized to undergo conventional (control), MIS mini-incision mid-vastus (mini), or MIS sidecutting (side cut) techniques for their total knee arthroplasty. Long-leg radiographs were assessed postoperatively. The mean overall limb varus alignments were 1.038 (SD, 2.588) for the control group, 0.878 (SD, 2.968) for the mini group, and 0.378 (SD 4.278) for the side cut group. The mean overall limb alignments within F3.08 varus/ valgus were 83.3%, 83.3%, and 56.7%, respectively. Femoral implant placement ( P = .028) and overall limb alignment ( P = .024) in the side cut group were significantly poorer as compared with those in the control group. The side cut group also had more outliers in the coronal plane. Results were comparable between the mini and control groups. The side cut technique appears to affect the accuracy of implant placement. Key words: MIS-TKA, Randomized controlled trial, Radiographic alignment. n 2007 Elsevier Inc. All rights reserved.
and controversial orthopedic topic is that of minimally invasive surgery (MIS) for TKA. Minimally invasive surgery is based on the concept of causing as minimal surgical trauma as possible to reduce postoperative pain and facilitate the body’s healing, resulting in earlier functional recovery. This concept is not new to knee arthroplasty. Minimally invasive surgery has already been proposed for unicondylar knee arthroplasty. It is only logical for MIS to be applied to TKA as technology progresses and surgical experience develops [2]. With the extent of surgical trauma aside, the success of a TKA is dependent on several other factors, chief among which is correct component alignment. Malalignment can lead to abnormal patellar tracking, increased polyethylene wear, early loosening, and poor functional outcome. There are concerns that obtaining correct compo-
The existing research evidence suggests that total knee arthroplasty (TKA) is a safe and cost-effective treatment for alleviating pain and restoring physical function in patients who do not respond to nonsurgical therapies [1]. As surgical experience with TKA increases, new operative techniques to expedite functional recovery and improve clinical outcomes are constantly being proposed. A current From the Department of Orthopedic Surgery, Singapore General Hospital, Singapore City, Republic of Singapore. Submitted December 9, 2005; accepted October 3, 2006. No benefits or funds were received in support of this study. Reprint requests: Pak Lin Chin, MBBS (Singapore), MRCSEd, MMed (Ortho) (Singapore), Department of Orthopedic Surgery, Singapore General Hospital, Block 6, Level 7, Outram Road, Singapore City 169608, Republic of Singapore. n 2007 Elsevier Inc. All rights reserved. 0883-5403/07/1906-0004$32.00/0 doi:10.1016/j.arth.2006.10.009
800
Radiographic Alignment ! Chin et al 801
nent alignment may be difficult with the currently proposed minimally invasive techniques for TKA because of poor surgical access. This may translate eventually to poorer functional outcomes and increased failure rates. Hence, the aim of this study was to compare the radiologic outcomes of the conventional technique with those of minimally invasive techniques for TKA.
Materials and Methods From March to October 2004, 90 consecutive patients with osteoarthritis of their knees requiring TKA were enrolled into this single-blind, prospective, randomized, and controlled trial. Patients with genu valgus, previous knee surgery (including high tibial osteotomy and previous knee arthroplasty), an active infection, or malignancy and those deemed medically unfit for surgery were excluded. The study protocol was approved by our hospital’s ethics committee. Each patient was counseled and provided with patient information sheets regarding the conduct of the trial. After we obtained informed consent from the patients, we allocated them to undergo the standard medial parapatellar technique (control; n = 30), the minimally invasive mini-incision midvastus technique (mini; n = 30), or the minimally invasive quadriceps-sparing technique (side cut; n = 30) for their TKA. An independent statistician designed the randomization table. Concealment of allocation was by sealed envelopes, the contents of which were only made known to the operating surgeons at the start of each surgery. The patients, research assistants, physiotherapists, and statisticians were all blinded to the allocations. All patients were seen before surgery, and their demographic characteristics, including age, sex, height, weight, and body mass index, were recorded. A senior anesthetist also assessed them preoperatively to determine their fitness for surgery. All surgeries were performed by the 2 senior surgeons (NNL and SJY), each of whom had performed at least 200 TKAs in the past year. This study was preceded by the surgeons familiarizing themselves thoroughly with the various minimally invasive TKA techniques. This process included cadaveric workshops on 50 cases with the side cut technique to help minimize performance bias. For all techniques used, the manufacturers’ recommended techniques were followed. The standard medial parapatellar technique (control) involved an anterior incision using a capsular
approach that separates the interval between the rectus femoris and the vastus medialis. It allowed for the patella to be everted for better joint visualization and implant placement. The suprapatellar pouch was also dissected free for prosthesis sizing. This was followed by intramedullary femoral and extramedullary tibial instrumentations. Standard cutting blocks were then used for the bone cuts. Two minimally invasive techniques were used for comparison: the mini technique and the side cut technique. The minimally invasive mini technique was carried out as described by Laskin [3,4]. The suprapatellar pouch was minimally disrupted. Zimmer NexGen MIS Mini (Zimmer Inc, Warsaw, Ind) and Johnson & Johnson DePuy Pathway (DePuy Orthopedics Inc, Warsaw, Ind) instrumentation sets were used in this mini study arm. These mini instrumentation sets were essentially miniaturized versions of conventional cutting jigs. Their miniaturization however allowed for less extensive soft tissue dissections. The other minimally invasive technique used was the side cut technique, with which a capsular approach as described by Gesell and Tria [5] was used. Significantly, the extensor mechanism and the suprapatellar pouch were not violated. A
Fig. 1. Angles measured for assessment: angle A, mechanical axis of the lower limb vs the femoral component alignment; angle B, mechanical axis of the lower limb vs the tibial component alignment; angle C, angle bisecting the mechanical axis of the femur and the mechanical axis of the tibia; angle F, angle of femoral component flexion; angle T, posterior slope of the tibial component.
802 The Journal of Arthroplasty Vol. 22 No. 6 September 2007 Table 1. Patients’ Profiles
Age (y) Male/female ratio Body mass index (kg/m2) Total operation time (min) Total blood loss (mL) Length of inpatient hospitalization stay (d)
Control group
Mini group
Side cut group
63.4 (47-80) 3:27 29.44 (22.7-40.0) 92.0 (55-120) 40.3 (30-980) 6.1 (2-10)
67.4 (56-80) 6:24 28.50 (22.1-40.0) 111.7 (65-150) 411.8 (40-1310) 6.0 (4-15)
69.0 (57-80) 6:24 27.53 (18.6-34.2) 111.5 (75-145) 357.8 (30-1040) 6.1 (3-16)
Values in brackets are ranges. All P values, except for total operative time, are not significant ( P N .05) with analysis of variance.
specially designed Zimmer NexGen MIS quadriceps-sparing instrumentation (Zimmer Inc) was used. This instrument was characterized by miniaturization and side-cutting jigs. The knee replacement implants used in this trial included the Zimmer NexGen Complete Knee Solution Legacy posterior-stabilized knee system (Zimmer Inc) and the Johnson & Johnson DePuy PFC Sigma fixed-bearing cruciate-substituting knee system (DePuy Orthopedics Inc) for the control and mini study arms. As for the side cut study arm, only the Zimmer NexGen Complete Knee Solution Legacy posterior-stabilized knee system (Zimmer Inc) was used. The patella was resurfaced in every patient. A Palacos G antibiotic-loaded bone cement (BioMet Orthopedics, Inc, Warsaw, Ind) was used in all the cases. Antibiotic prophylaxis (2 g of intravenous cefazolin) was administered 30 minutes before the commencement of surgery. This was administered provided that there was no contraindication. If a patient was allergic to cefazolin, 1 g of intravenous vancomycin was administered instead. This antibiotic prophylaxis was continued for 24 hours after the operation. All patients had a Redivac drain inserted intraoperatively, which was removed 24 hours after the operation. The total blood loss sustained intraoperatively and that at 24 hours after the surgery were calculated. The total operating time was also recorded. Postoperative pain relief was administered via patient-controlled analgesia using intravenous
morphine for the first 24 hours after the operation. Oral analgesia was commenced 24 hours after the operation. This included 1 g of paracetamol every 6 hours, 550 mg of naproxen every 12 hours, and 20 mg of omeprazole every 12 hours. If a patient was allergic to these medications, 50 mg of tramadol every 8 hours was served instead. All patients underwent a standardized postoperative physiotherapy program. This involved continuous passive movements and active exercises as soon as possible. The total length of inpatient hospital stay was recorded, and any postoperative complication was documented by blinded research assistants. All patients had a standardized series of plain x-rays taken when they were able to weight bear without any flexion deformity; these included long-leg standing anteroposterior and lateral x-rays of the limbs [6-8]. All radiographs were performed at the same radiology center. Radiologic data were collected separately and independently on 3 occasions by 3 research assistants [7]. All 3 research assistants were blinded to patient allocations. The following radiologic data were collected (reference is drawn to the article by Chin et al [6] regarding the measurement of radiologic angles in postoperative longleg standing x-rays of patients with total knee replacements; Fig. 1): 1. Mechanical axis of the lower limb vs the femoral component alignment (angle A) 2. Mechanical axis of the lower limb vs the tibial component alignment (angle B)
Table 2. Angles in the Coronal Axis Control group Angle A (femur implant) Angle B (tibial implant)
86.0 to 91.5 1.5 to 4.0 (1.17 varus) 88.0 to 97.0 7.0 to 2.0 (0.77 valgus)
Mini group 84.0 to 93.0 3.0 to 6.0 (1.28 varus) 87.0 to 98.0 8.0 to 3.0 (1.12 valgus)
Side cut group 85.0 to 92.0 2.0 to 5.0 (1.64 varus) 88.0 to 98.0 8.0 to 2.0 (1.91 valgus)
Values are presented as degrees. All P values are not statistically significant ( P N .05) with analysis of variance.
Radiographic Alignment ! Chin et al 803 Table 3. Good Outcomes for Angles in the Coronal Axis Good outcome
n (%)
P
Angle A (femur implant) Control group
– 27 (90)
Mini group
.448
.519
25 (83.3) Side cut group
.166 23 (76.7)
Angle B (tibial implant) Control group
Mechanical Axis of the Lower Limb on Coronal X-Ray Measurements (Angle C)
– 27 (90)
Mini group
.448
.222
25 (83.3) .053
Side cut group 21 (70.0)
3. Angle bisecting the mechanical axis of the femur and the mechanical axis of the tibia (angle C) 4. Angle of femoral component flexion (angle F) 5. Posterior slope of the tibial component (angle T) Statistical Analysis All data were collected at the end of the trial, and the patient groups were blinded for further studies. A blinded statistician (Stephanie Fook-Chong) who was independent of the surgical and research team analyzed the data using SPSS 13.0 for Windows (SPSS Inc, Chicago, Ill). Pearson’s v 2 analysis, Fisher’s exact test, and analysis of variance between groups were performed as appropriate.
Results
Femoral and Tibial Component Alignments on Coronal X-Ray Measurements (Angles A and B) The results showed that the control and mini study arms had more consistent femoral and tibial Table 4. Angle C (Mechanical Axis of the Femur vs Mechanical Axis of the Tibia) in the Coronal Axis Alignment 6.0 to 6.0 (1.03 varus) 4.0 to 7.0 (0.87 varus) 8.0 to 8.0 (0.37 varus)
All 3 study arms had a mean varus alignment. The control and mini study arms had significantly more patients with good outcome measurements for the mechanical axis of the lower limb as compared with the side cut study arm (Tables 4 and 5). Femoral Component Flexion and Tibial Component Posterior Slope on Sagittal X-Ray Measurements (Angles T and F) All 3 study arms gave consistent tibial and femoral component alignments on the sagittal x-ray measurements (Table 6). Total Operation Time Operations performed with the standard medial parapatellar technique (control) took a shorter total time as compared with the 2 minimally invasive techniques ( P b .001). There was no statistical difference in the total operating time between the 2 minimally invasive study arms. Total Blood Loss
The patients’ profiles are listed in Table 1. The cohort was homogeneously distributed among the 3 study arms.
Standard Mini Side cut
component alignments with less outliers as compared with the side cut study arm (Table 2). The proportion of patients with good outcomes for femoral and tibial component alignments was also higher for the control and mini study arms as compared with the side cut study arm (Table 3). Descriptively, the groups ranked as follows: control group = mini group N side cut group.
Table 5. Good Outcomes of Angle C (Mechanical Axis of the Femur vs Mechanical Axis of the Tibia) Good outcome
P
n (%)
P
Control group
– 1.000 .597
Total blood loss was less in the minimally invasive technique study arms as compared with the control study arm ( P = .015). Three of the patients in the control study arm (10%) required postoperative blood transfusions. None of the patients in either minimally invasive technique group needed any blood transfusion. Between the 2 minimally invasive technique groups, the side cut
.127
Values are presented as degrees. Statistical analysis was conducted with analysis of variance.
– 25 (83.3)
Mini group
1.000 25 (83.3) .024
Side cut group 17 (56.7)
.0240
804 The Journal of Arthroplasty Vol. 22 No. 6 September 2007 Table 6. Angles in the Sagittal Axis
Angle F (femoral flexion) Angle T (tibial slope)
Control group
Mini group
Side cut group
5.0 to 10.0 (3.17 flexion) 2.0 to 10.0 (4.95 posterior slope)
0 to 5.0 (2.75 flexion) 0 to 12.0 (4.17 posterior slope)
0 to 10.0 (3.05 flexion) 0 to 7.0 (3.55 posterior slope)
Values are presented as degrees. All P values are not significant ( P N .05) with analysis of variance.
study arm had less blood loss as compared with the mini study arm, although this difference was not statistically significant ( P = 1.000). Length of Inpatient Hospitalization Stay There was no statistical difference in the length of inpatient hospitalization stay among the 3 study arms. This finding may be unique to our local medical insurance coverage terms, for which there is no significant financial impetus to keep inpatient hospitalization stays as short as possible. Postoperative Complications Two patients who underwent the standard medial parapatellar technique (control) knee arthroplasty developed proximal deep vein thrombosis, which was treated expectantly with antithrombotic therapy. One patient who had a side cut operation developed superficial wound infection, which was successfully treated with 2 weeks of intravenous antibiotics. One patient in each minimally invasive study arm developed postoperative knee stiffness that required manipulation under general anesthesia 3 months after the operation. There was no mortality case seen in any of the 3 study arms.
Discussion Minimally invasive surgery has become a popular topic among orthopedic surgeons and among patients. Increasing numbers of patients are asking for an MIS after reading about its proposed advantages of less postoperative pain, quicker functional recovery, and better cosmesis (shorter scars). Although the concept behind MIS appears to be attractive, there is first a need to ascertain its efficacy with well-designed and well-administered clinical trials before proclaiming it as the next standard of care. There is also a need for consensus among orthopedic surgeons as to what exactly constitutes an MIS. Minimally invasive surgery does not necessarily equate with shorter wound incisions; it is best appreciated as a surgical approach that results in less
intraoperative surgical trauma. This in turn reduces postoperative pain, facilitates early rehabilitation, and enhances clinical and functional recovery. Two of the more commonly described minimally invasive techniques for TKA are the mini technique and the side cut technique. Both techniques cite minimal disruption of the quadriceps mechanism as their advantage, with each technique having its own proponents. The minimally invasive mini technique has been described as an efficacious alternative to the standard medial parapatellar quadriceps-splitting approach that still allows for satisfactory surgical exposure. On the other hand, the side cut technique has been advocated as a useful technique to completely avoid violating the quadriceps mechanism and the suprapatellar pouch. However, there are concerns about the adequacy of surgical access and the accuracy of bone cuts with the customized side-cutting jigs used with the side cut technique. This in turn may affect component alignment, resulting in abnormal wear and eventual clinical failure. Tria [9] reported on favorable early results from his series of 120 minimally invasive TKAs. His results indicated that minimally invasive TKAs were associated with less blood loss, less postoperative pain, better early motion, and a shorter hospital stay as compared with the conventional techniques, with no compromise of accuracy. Laskin et al [10] performed 32 minimally invasive TKAs using the mini approach. They reported that the average visual analog pain score and the total amount of pain medication among the patients with minimally invasive TKAs were lower as compared with those among the patients who had a standard medial parapatellar technique surgery. In addition, they found that improvements in the Knee Society score 6 months after the operation were statistically higher as compared with those in the control group. They attributed this to limited disruption of the extensor mechanism and the suprapatellar pouch, allowing for more rapid restoration of quadriceps muscle control. They also noted that the radiographic alignment and position of implant components were satisfactory in all the patients from both groups.
Radiographic Alignment ! Chin et al 805
Similar results were reported by Haas et al [11] with their series of 40 minimally invasive TKAs performed via the mini technique. However, there is, to our knowledge, no published article describing a 3-arm prospective randomized controlled trial comparing the radiologic outcomes of the standard medial parapatellar technique with those of the minimally invasive techniques of the mini and side cut approaches. Furthermore, most of the earlier studies on minimally invasive TKA were cohort studies and not prospective randomized controlled studies [12-18]. Radiologic measurements in these earlier studies were also often based on short knee films and not long-leg standing x-rays, which give more accurate radiologic measurements. Computed tomographic scan assessment can be very enticing; however, radiographs, which can be done with the patient standing, give a better reflection of the situation. The MIS techniques needed a longer operative time because the minimal surgical exposure needed more time for careful placement of instrument guides and implant placement. Moreover, with the small incision, intuitive optimization of the moving surgical window was required, in which the knee is ranged more frequently, hence accounting for longer surgical duration. This is similar to previous authors’ experience [2-4]. The most clinically relevant radiologic measurement is the overall limb alignment against the mechanical axis of the lower limb, angle C. A varus/valgus malalignment greater than 38 has been shown to lead to early failure [6]. The subanalysis of the placement of the femoral component (angle A) and the tibial component (angle B) would help identify the deficiencies in the particular minimally invasive technique, be it in the femoral or the tibial component alignment. Our results showed that the femoral and tibial component positions have comparable error risks (Table 3). This may be a result of the decreased exposure and difficulty in placing the cutting guides close to the bone in the ideal position. This was most evident in the side cut group. The sagittal-axis assessments of good outcomes are arbitrarily set and, to our knowledge, have not been investigated thoroughly, along with their bearing on survivorship. We feel that the outcome will not be compromised as long as the femoral and tibial components are not placed in extension. Our results suggest that the mini technique appears to give radiologic results comparable with those of the standard medial parapatellar technique. The side cut technique, although generally giving a satisfactory overall standing femorotibial lower limb alignment, does show significant vari-
ance in the accuracy of the implant component alignment. This we postulate is related to the less ideal surgical access provided for by the side cut technique. There is poor line of sight, making the operation more technically difficult. We also had frequent difficulties in ensuring accurate bone cuts with the customized sidecutting jigs used with the side cut technique. This was especially true for bone cuts to the lateral femoral condyle and tibial plateau. This was because the side-cutting jigs were more prone to displacement and loosening during sawing of the bone. It is important to keep the jigs well secured. We found it useful to keep the intramedullary guide rod within the femur to maintain alignment of the femoral side-cutting jig. The use of multiple pins, particularly threaded pins, to secure the jigs in place is also advisable. Another consideration is the need for the saw blade to transverse greater distances when using the side cut technique. This increased bone-cut transversing distance magnifies any error that may have resulted from incorrect jig placement or loosening of the jigs from the jig-bone interfaces. Minimally invasive surgery for TKA is still in its infancy. There is a steep learning curve, and much research work remains to be done. Based on our experience and the difficulties encountered in the cited cases, we recommend that efforts be made to design more user-friendly, more secure, and more accurate instrumentations. This is especially true for side-cutting jigs. Further miniaturization of instrumentations and particularly that of the prosthesis would be most welcome. Currently, incision length is largely restricted by the size of the prosthesis.
Conclusions The side cut technique but not the mini technique appears to have a significant negative impact on the accuracy of TKA. The errors may be caused by intrinsic problems with the side cut technique, which include restricted surgical access with limited line of sight, the need for the saw blade to transverse a longer bone-cut distance, magnifying any cutting jig-bone alignment error, and inadequately rigid cutting jig-bone fixation, causing loss of cutting jig alignment. Total knee arthroplasty is a very successful surgery. It is imperative that outliers or surgical errors be minimized. Having good functional outcome using new techniques at the expense of poor consistency in implant placement is clinically unacceptable. The ultimate aim is to align the knee along the mechanical axis. Studies
806 The Journal of Arthroplasty Vol. 22 No. 6 September 2007 have shown that a malalignment greater than F38 will lead to early failure [19]. Additional caution needs to be exercised when using minimally invasive side cut techniques to avoid these pitfalls.
Acknowledgments We thank Ms Stephanie Fook-Chong of the Singapore General Hospital Clinical Trials Center for her invaluable assistance with the statistical analysis.
References 1. NIH Consensus Panel. NIH consensus statement on total knee replacement December 8-10, 2003. J Bone Joint Surg Am 2004;86-A:1328. 2. Tria Jr AJ, Coon TM. Minimal incision total knee arthroplasty: early experience. Clin Orthop Relat Res 2003;185. 3. Laskin RS. Minimally invasive total knee replacement using a mini–mid vastus incision technique and results. Surg Technol Int 2004;13:231. 4. Laskin RS. Minimally invasive total knee arthroplasty: the results justify its use. Clin Orthop Relat Res 2005;440:54. 5. Gesell MW, Tria Jr AJ. MIS unicondylar knee arthroplasty: surgical approach and early results. Clin Orthop Relat Res 2004;53. 6. Chin PL, Yang KY, Yeo SJ, et al. Randomized control trial comparing radiographic total knee arthroplasty implant placement using computer navigation against conventional technique. J Arthroplasty 2005;20:618. 7. Perlick L, Bathis H, Perlick C, et al. Revision total knee arthroplasty: a comparison of postoperative leg alignment after computer-assisted implantation versus the conventional technique. Knee Surg Sports Traumatol Arthrosc 2005;13:167 [Epub 2004 Apr].
8. Sparmann M, et al. Positioning of total knee arthroplasty with and without navigation support. A prospective, randomised study. J Bone Joint Surg Br 2003;85:830. 9. Tria Jr AJ. Advancements in minimally invasive total knee arthroplasty. Orthopedics 2003;26(8 Suppl); s859. 10. Laskin RS, Beksac B, Phongjunakorn A, et al. Minimally invasive total knee replacement through a mini-midvastus incision: an outcome study. Clin Orthop 2004;74. 11. Haas SB, Cook S, Beksac B. Minimally invasive total knee replacement through a mini midvastus approach: a comparative study. Clin Orthop 2004;68. 12. Dalury DF, Jiranek WA. A comparison of the midvastus and paramedian approaches for total knee arthroplasty. J Arthroplasty 1999;14:33. 13. White Jr RE, Allman JK, Trauger JA, et al. Clinical comparison of the midvastus and medial parapatellar surgical approaches. Clin Orthop 1999;117. 14. Keating EM, Faris PM, Meding JB, et al. Comparison of the midvastus muscle–splitting approach with the median parapatellar approach in total knee arthroplasty. J Arthroplasty 1999;14:29. 15. Weinhardt C, Barisic M, Bergmann EG, et al. Early results of subvastus versus medial parapatellar approach in primary total knee arthroplasty. Arch Orthop Trauma Surg 2004;124:401 [Epub 2004 May 07]. 16. Matsueda M, Gustilo RB. Subvastus and medial parapatellar approaches in total knee arthroplasty. Clin Orthop 2000;161. 17. Cila E, Guzel V, Ozalay M, et al. Subvastus versus medial parapatellar approach in total knee arthroplasty. Arch Orthop Trauma Surg 2002;122:65. 18. Faure BT, Benjamin JB, Lindsey B, et al. Comparison of the subvastus and paramedian surgical approaches in bilateral knee arthroplasty. J Arthroplasty 1993; 8:511. 19. Ritter MA, et al. Postoperative alignment of total knee replacement. Its effect on survival. Clin Orthop 1994;153.