The Journal of Arthroplasty 30 (2015) 1233–1236
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
The Valgus Stress Radiograph Does Not Determine the Full Extent of Correction of Deformity Prior to Medial Unicompartmental Knee Arthroplasty Tyler M. Kreitz, MD, Mitchell G. Maltenfort, PhD, Jess H. Lonner, MD Rothman Institute, Thomas Jefferson University, Philadelphia, PA
a r t i c l e
i n f o
Article history: Received 15 October 2014 Accepted 10 February 2015 Keywords: unicompartmental knee arthroplasty stress radiograph coronal alignment computer navigation
a b s t r a c t Routine preoperative stress radiographs have been advocated, in part, to determine “full correctability” of deformities before proceeding with unicompartmental knee arthroplasty (UKA) despite limited data supporting their utility. Fifty consecutive patients undergoing medial UKA with robotic navigation were studied. In 20° of flexion, significantly greater correctability was achieved after removal of osteophytes by an additional 1.8°, with a mean corrected alignment of 2.5° varus. Seventy-four percent of knees were not correctable to neutral alignment or more. In conclusion, preoperative stress radiographs have overstated value in patients undergoing medial UKA since the full extent of correctability of varus deformity cannot be determined until after removal of osteophytes and since most deformities are not fully correctable to neutral in UKA. © 2015 Elsevier Inc. All rights reserved.
Unicompartmental knee arthroplasty (UKA) is proven effective for appropriately selected patients with single compartment knee arthritis. The success of UKA is dependent on several factors, including patient selection, surgical precision, and implant design [1–13]. When these elements are optimized, so too are postoperative knee kinematics and functionality, patient satisfaction and implant durability after UKA [14]. The valgus stress view radiograph has been popularized as a means of determining correctability of the deformity, ensuring maintenance of the lateral joint space, and indirectly assessing the integrity of the anterior cruciate and medial collateral ligaments [1,2,15,16]. It is often discussed as an important criteria for determination of a patient’s candidacy for medial UKA [1,2,12,15,16]. Proponents of the valgus stress xray often state that it must demonstrate “full correctability” of the deformity, without narrowing of the lateral joint space, for a patient to be considered a candidate for medial UKA [1,2,15,16]. The valgus stress radiograph is taken with the patient lying supine, with the knee in 20° of flexion to relax the posteromedial capsule and a valgus stress applied. To avoid a parallax effect, the xray beam is directed 10° caudad [2,15,16].
One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2015.02.008. Reprint requests: Jess H. Lonner, MD, Rothman Institute, 825 Old Lancaster Ave., 2nd Floor, Bryn Mawr, PA 19010. http://dx.doi.org/10.1016/j.arth.2015.02.008 0883-5403/© 2015 Elsevier Inc. All rights reserved.
Although often advocated as a standard tool in the preoperative evaluation for medial UKA, there is a paucity of literature evaluating the utility of preoperative valgus stress radiographs [16–18]. Our hypothesis is that the preoperative valgus stress radiograph is of overstated value for 2 reasons: (1) maximal correctability is not possible until after intraoperative removal of osteophytes; and (2) full correction of the deformity is not routinely achievable. Other potential uses of the valgus stress radiograph, such as indirectly assessing the integrity of the ACL by identifying the presence of a fixed varus deformity, determining integrity of the medial collateral ligament and integrity of the lateral compartment articular cartilage were not studied. Methods This is a prospective series of 50 consecutive patients who underwent robotically assisted medial UKA with intraoperative navigation. Institutional review board approval was obtained for this study and the United States Health Insurance Portability and Privacy Act (HIPPA) regulations were followed throughout. All procedures were performed by the senior surgeon. The mean patient age was 60, ranging from 49 to 85 years old. A medial peripatellar capsular incision was routinely used. Computer navigation methods were utilized to make measurements and plan surgery in conjunction with robotic bone preparation during UKA. The anchoring system is comprised of 2 bicortical pins in both the distal femoral and proximal tibial metaphyses, to which trackers and reflective spherical transmitters are attached. An infrared beam emitted from the hardware module and camera is reflected from the spherical
1234
T.M. Kreitz et al. / The Journal of Arthroplasty 30 (2015) 1233–1236
trackers on the arrays, and anatomic surface landmarks of the joint and mechanical and rotational axes of the limb are mapped and registered. The accuracy of measurements with intraoperative navigation has previously been validated to within 1° [19–21]. In each patient the resting (unstressed) varus mechanical alignment of the limb was determined. The medial osteophytes were assessed and subjectively categorized as being either small or large by the senior surgeon. Prior to removing the osteophytes a maximal valgus stress was applied to the knee to determine “correctability” of the deformity, both in full extension and 20° of flexion, and the corrected mechanical alignment recorded for each position in all patients. The osteophytes were then removed, maximal valgus stress was re-applied and the same measurements were made in both full extension and 20° of flexion. Mechanical alignment was analyzed using positive values for varus, negative values for valgus, and zero for neutral mechanical alignment. A regression analysis was performed, controlling for multiple measurements in each patient – osteophyte size (small or large); corrected (stressed) alignment both before and after osteophyte removal; and differences in stressed (corrected) mechanical alignment at full extension and 20° of flexion.
Results The mean unstressed resting mechanical axis in full extension was 8.2° varus alignment (range 1° to 16°). Before osteophyte removal, the mean corrected mechanical alignment was 5.1° ± 2.5° varus in full extension and 3.4° ± 2.7° varus in 20° of flexion. After osteophyte removal, the mean corrected mechanical alignment was 3.2° ± 2.4° varus in full extension and 1.6° ± 2.5° varus in 20° of flexion (range 6° varus to 4° valgus) (Table 1). A total of 13 patients (26%) were correctable to either neutral or valgus alignment. Thirty-seven patients (74%) had a postoperative varus stress alignment and were not correctable to neutral. Varus deformities were more correctable when medial capsule and ligaments were stressed in 20° of flexion than in full extension. Prior to and after osteophyte removal, the alignment was 1.7° ± 0.8° and 1.6° ± 0.1° more correctable in 20° of flexion compared to full extension, respectively. With the knees flexed 20°, 32% of knees with mechanical varus of 10° or less were correctable to neutral or beyond. Alternatively, 7.7% of those with mechanical varus of 10° or more were correctable to neutral or beyond. Significantly greater correctability was achieved after removal of osteophytes by an additional 1.8° ± 0.1° (P = 2.00E-16) in 20° of flexion. Additionally, after osteophyte removal, correctability in 20° of flexion was significantly greater than in full extension by an additional 1.6°± 0.1° (P = 2.00E-16) (Table 2). Regression analysis showed no significant change in stressed alignment whether the osteophytes removed were small or large. There was a mean of 0.2° ± 0.52° less correctability when small osteophytes were removed compared to large osteophytes (P = 0.653).
Discussion UKA has a specific set of indications for patients with single compartment osteoarthritis of the knee [3–5]. “Full correctability” of deformity on stress radiographs is often cited as a prerequisite for a successful outcome after medial UKA [1,15,16,18,22]. However, clinical studies suggest that slight undercorrection (i.e. leaving the knee in slight mechanical varus) may improve midterm and long-term outcomes and durability [6–10]. Additionally, even in the hands of those who have been strong advocates of using valgus stress radiographs to ensure that deformities are “fully correctable” to determine candidacy for UKA, 25% of medial UKAs were left in varus, without compromising clinical or radiographic outcomes [23]. Further, in addition to the absence of clinical data supporting the benefit of valgus stress radiographs on clinical outcomes, these studies are utilized in less than 20% of patients being considered for UKA [24]. Finally, there is discordance regarding the optimal role of stressed radiographs and how “full correctability” should best be defined. Our hypothesis was that preoperative valgus stress radiographs cannot fully determine the extent of correctability of the deformity. While proponents suggest that preoperative valgus stress radiographs are important to determine that deformities are “fully correctable,” our study shows that determination of maximal correctability cannot be achieved until after removal of osteophytes and that “full correctability” to a neutral mechanical axis is not achievable in the majority of patients undergoing medial UKA. In the present study, in 20° of flexion, application of maximal valgus stress after osteophyte removal produced an additional 1.8° of correction of the deformity. This statistically significant difference underscores that full determination of correctability is not achievable until after osteophyte removal. Second, in this consecutive nonselected series “full correction” to neutral alignment or beyond was only possible in 26% of patients undergoing medial UKA, even after osteophyte removal. Only 8% of patients with preoperative mechanical varus of more than 10° and 32% with preoperative varus of 10° or less were correctable to neutral or beyond. Further, given the success of medial UKA when leaving the deformity in slight mechanical varus, recommending “full correctability” is both impractical and unsubstantiated [6–10,23,25]. Other potential roles of the stress radiograph, such as ruling out a fixed deformity which would suggest chronic ACL deficiency, determining integrity of the medial collateral ligament and integrity of the lateral compartment articular cartilage were not studied. Our study supports the conclusions by Waldstein et al [17] that stress radiographs may be unnecessary, although the basis for those conclusions differs in our 2 studies. They concluded that they provided no additional information in the assessment of the quality of cartilage in the lateral compartment or correctability of deformity in 84 patients (91 knees) undergoing total knee arthroplasty for non-inflammatory varus arthritis [17]. They found that preoperative alignment of 10° or less mechanical varus was correctable to 0.1° varus (95% CI, 0.6° valgus to 0.5° varus), with 93% correctable to between 3° varus and 3° valgus
Table 2 Impact of Variables on Correctability. Table 1 Resting and Corrected (Stressed) Mechanical Alignment.
Variable Mean ± SD
Unstressed (resting) alignment Stressed alignment before osteophyte removal: full extension Stressed alignment before osteophyte removal: 20° of knee flexion Stressed alignment after osteophyte removal: full extension Stressed alignment after osteophyte removal: 20° of knee flexion
8.2° ± 2.9° 5.1° ± 2. 5° 3.4° ± 2.7° 3.2° ± 2.4° 1.6° ± 2.5°
Shows the mean mechanical axis alignment and SD of the limb in full extension without applying a stress (resting alignment), and corrected (stressed) alignment in full extension and 20° of flexion, both before and after osteophyte removal for 50 patients undergoing medial UKA.
Mean ± SD
P
Difference in stressed alignment after removal of small 0.2° ± 0.5° 0.653 vs large osteophytes Difference in stressed alignment in 20° flexion before vs −1.8° ± 0.1° 2.00E-16 after osteophyte removal After osteophyte removal, difference in stressed −1.6°± 0.1° 2.00E-16 alignment in 20° flexion vs full extension Shows impact of multiple variables on changes in alignment with applied valgus stress: size of osteophytes; stressed alignment in 20° flexion before and after osteophyte removal; and stressed alignment in 20° flexion compared to full extension after osteophyte removal. Positive values represent an increase in stress alignment and negative values a decrease. P b 0.05 was considered significant.
T.M. Kreitz et al. / The Journal of Arthroplasty 30 (2015) 1233–1236
on stress radiographs. In knees with greater than 10° varus, valgus force corrected deformities to a mean of 5.5° varus (95% CI, 4.3° to 6.8° varus); most (91%) undercorrected between 3.8° varus and 10.3° varus. The authors concluded that since most small deformities (those most commonly considered for UKA) were corrected to between 3° varus and 3° valgus, preoperative stress radiographs provided no additional information in the preoperative evaluation of patients who may be considered for UKA and were therefore unnecessary [17]. Their indirect determination of corrected mechanical alignment has inherent error both due to the method of measure and the error of using short films with a knee that is flexed [22] and may explain the greater extent of correction than we observed in our study. So while Waldstein et al felt that stress radiographs were unnecessary since most patients were correctable to “neutral”, ours questioned the role of stress radiography since we found that full correctability is not possible until osteophytes are removed and most knees could not be corrected to neutral. Our results and conclusions differed from those from a study by Tashiro et al which reported that preoperative full-length valgus stress radiographs were useful for evaluating correctability of varus deformities prior to UKA in 37 consecutive patients [18]. In their study, an arthrometer was used to impart a 150-N valgus force to the studied knees and mechanical axis deviation was measured using full length radiographs. Patients had a mean unstressed varus alignment of 6.5° ± 2.8° (range: 2.3° to 11.8° varus). With a valgus stress, the alignment was corrected to 0.2° ± 2.6° mechanical varus. It is not clear how many were corrected to 0° or beyond, but none were corrected beyond 2° valgus. It appears that the authors of that study achieved greater correction of deformity with instrumented and calibrated stress application than we achieved. It is not known how much manual force was applied in each case by the senior surgeon conducting our study, whether it was consistent throughout cases, and how it compared to the 150N valgus force applied in Tashiro’s study. Nonetheless, while our manual method may create some inherent variability in applied force and thereby confound accurate measurements, we feel that the method used in our study is more representative of the uncalibrated technique used by most surgeons performing preoperative valgus stress radiographs in clinical practice. While we are using navigated measures of alignment rather than preoperative stress radiographs to determine correctability, the navigated method of alignment measurement used in this study has advantages compared to the use of stress views with short plate radiographs. A study by Lonner et al found that as little as 10° of knee flexion has a statistically significant impact on the measured tibiofemoral alignment after TKA using short radiographs [22]. That study showed that the tibiofemoral valgus alignment is reduced by 1° when the knee is flexed 10° and it can be reduced further with rotation of the limb. Therefore the apparent correction of alignment that occurs with application of a valgus stress with the knee flexed 20° is overestimated when using short films, particularlly if the limb is even slightly rotated during the radiograph. This may explain why the studies by Waldstein et al [17] and Gibson et al [16] found greater apparent correctability than we did in the current study. Our study limits the error by using the mechanical, rather than the tibiofemoral, axis. However, we could not control for subtle errors in rotation that may have occurred during valgus stressing. Additionally, Yaffe et al found that there is a mean 5° discrepancy in measured mechanical alignment between navigated measurements and full-length films, with variations that can be as much as 12° [26]. Therefore, the short films commonly used when performing valgus stress xrays, as well as the arbitrary determination that the knee has been flexed 15°–20° and not rotated may render determination of tibiofemoral alignment on valgus stress radiographs relatively inaccurate. Further study can be considered to compare correctability by navigated and radiographic measures; however, our opinion based on the referenced studies is that navigation, as applied in the current study, is a reasonable surrogate for preoperative stress radiographs.
1235
Limitations of this study include the subjective sizing of osteophytes. Osteophytes were categorized as either small or large without specific measurement criteria. Nonetheless, the size of the osteophytes did not appear to make a difference in the extent of correctability of the deformities. Preoperative and corrected alignment of the mechanical axis was measured using a validated navigated system, allowing for consistency between measurements. However, we are accepting that our use of navigation is accurate to within 1°, based on published studies that have validated this method [19–21], and it is our assumption that our use of navigation is more accurate than short films. If this assumes a maximum 1° error, then our difference in correctability may be reduced to 0.8°, rather than 1.8°. The valgus stress applied for each patient was neither instrumented nor measured and therefore inconsistencies of the forces applied may have resulted, causing potential errors in measures of alignment. However, while this variability in applied force may have some inherent error, this technique is representative of what is commonly used in clinical practice. Finally, our study used a more narrow determination of “neutral” alignment compared to other studies. We considered “neutral” to be 0°, since in UKA the goal is typically to leave the knee in slight residual varus and not overcorrect beyond 0°. Nonetheless, even if we broadened the range of “neutral” alignment, still the majority of our patients were not correctable beyond 3° of mechanical varus. The clinical utility of the valgus stress radiograph in the preoperative assessment of patients undergoing UKA has not been proven in the literature. Our data demonstrate that the full extent of correctability cannot be determined until after removal of osteophytes in patients undergoing medial UKA with varus deformity, and that full correctability to neutral or valgus resting alignment is not achievable in the majority of patients undergoing medial UKA. In light of the finding that most varus deformities are only partially correctable even after osteophyte resection, surgeons should not recommend “full correctability” as a criterion for UKA if that means correction to a neutral mechanical axis. Additionally, using intraoperative navigated measures as a proxy for preoperative stress radiographs, our data do not support the routine use of preoperative valgus stress radiographs for determining the extent of correction of the deformity in patients undergoing medial UKA, although there may be other potential roles of the valgus stress radiograph that were not investigated in this study. References 1. Argenson JN, Parratte S. Medial unicompartmental knee arthroplasty: Fixed BEARING TECHNIQUES. Chapter 10 , In: Berend KR, Cushner FD, editors. Partial Knee Arthroplasty: Techniques for Optimal Outcomes. Philadelphia, PA: Elsevier Saunders; 2012. 2. Hurst JM, Berend KR. Mobile-bearing unicondylar knee arthroplasty: the Oxford experience. Clin Sports Med 2014;33(1):105. 3. Borus T, Thornhill T. Unicompartmental knee arthroplasty. J Am Acad Orthop Surg 2007;15:9. 4. Newman MT, Lonner JH, Ries M. Partial knee arthroplasty – patellofemoral, unicompartmental, and bicompartmental. Ch 9 , In: Glassman AH, Lachiewicz PF, Tanzer M, editors. Orthopaedic Knowledge Update: Hip and Knee Reconstruction 4. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2011. 5. Jamali AA, Scott RD, Rubash HE, et al. Unicompartmental knee arthroplasty: Past, present, and future. 2009;38:564. 6. Berger RA, Meneghini RM, Jacobs JJ, et al. Results of unicompartmental knee arthroplasty at a minimum of ten years of follow-up. J Bone Joint Surg Am 2005;87:999. 7. Pennington DW, Swienckowski JJ, Lutes WB, et al. Unicompartmental knee arthroplasty in patients sixty years of age or younger. J Bone Joint Surg Am 2003; 85:1968. 8. Argenson JA, Chevrol-Benkeddache Y, Aubaniac JM. Modern unicompartmental knee arthroplasty with cement: A three to ten-year follow-up study. J Bone Joint Surg Am 2002;84(12):2235. 9. Hernigou P, Deschamps G. Alignment influences wear in the knee after medial unicompartmental arthroplasty. Clin Orthop 2004;423:161. 10. Cartier P, Sanouiller JL, Grelsamer RP. Unicompartmental knee arthroplasty surgery: 10-year minimum follow-up period. J Arthroplasty 1996;11:782. 11. Emerson RH, Higgins LL. Unicompartmental knee arthroplasty with the Oxford prosthesis in patients with medial compartment arthritis. J Bone Joint Surg Am 2008;90:118. 12. Argenson JA, Parratte S. The unicompartmental knee: design and technical considerations in minimizing wear. Clin Orthop Relat Res 2006;452:137. 13. Meek RMD, Masri BA, Duncan CP. Minimally invasive unicompartmental knee replacement: rationale and correct indications. Orthop Clin N Am 2004;35:191.
1236
T.M. Kreitz et al. / The Journal of Arthroplasty 30 (2015) 1233–1236
14. Patil S, Colwell Jr CW, Ezzet KA, et al. Can normal knee kinematics be restored with unicompartmental knee replacement? J Bone Joint Surg Am 2005;87(2):332. 15. Oxford Partial Knee. Manual of the Surgical Technique. Biomet UK Ltd Waterton Industrial Estate Bridgend, South Wales CF31 3XA, United Kingdom; 2015. http:// www.biomet.nl/resource/5980/Oxford-Knee-Optec.pdf. 16. Gibson PH, Goodfellow JW. Stress radiography in degenerative arthritis of the knee. JBJS 1986;68-B:608. 17. Waldstein W, Bou Monsef J, Buckup J, et al. The value of valgus stress radiographs in the workup for medial unicompartmental arthritis. Clin Orthop Relat Res 2013; 471(12):3998. 18. Tashiro Y, Matsuda S, Okazaki K, et al. The coronal alignment after medial unicompartmental knee arthroplasty can be predicted: usefulness of full-length valgus stress radiography for evaluating correctability. Knee Surg Sports Traumatol Arthrosc 2014;22:3142. 19. Hauser R. Computer-aided 3D-navigation systems: a plea for an error model. HNO 2000;2:71.
20. Oberst M, Bertsch C, Lahm A, et al. Regression and correlation analysis of preoperative versus intraoperative assessment of axes during navigated total knee arthroplasty. Comput Aided Surg 2006;11:87. 21. Picard F, Gregori A, Leitner F. Computer Assisted Total Knee Arthroplasty: Validation of the Image Free Concept. Berlin, Germany: Pro Business; 2007. 22. Lonner JH, Laird M, Stuchin SA. Effect of rotation and knee flexion on radiographic alignment in total knee arthroplasties. CORR 1996;331:102. 23. Gulati A, Pandit H, Jenkins C, et al. The effect of leg alignment on outcome of unicompartmental knee replacement. J Bone Joint Surg (Br) 2009;91B:469. 24. Schindler OS, Scott WN, Scuderi GR. The practice of unicompartmental knee arthroplasty in the United Kingdom. J Orthop Surg 2010;18:312. 25. Chatellard R, Sauleau V, Colmar M, et al. Medial unicompartmental knee arthroplasty: Does tibial component position influence clinical outcomes and arthroplasty survival? Orthop Traumatol Surg Res 2013;99S:S219. 26. Yaffe MA, Koo SS, Stulberg SD. Radiographic and navigation measurements of TKA limb alignment do not correlate. Clin Orthop 2008;466:2736.
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
The Valgus Stress Radiograph Does Not Determine the Full Extent of Correction of Deformity Prior to Medial Unicompartmental Knee Arthroplasty Tyler M. Kreitz, MD, Mitchell G. Maltenfort, PhD, Jess H. Lonner, MD Rothman Institute, Thomas Jefferson University, Philadelphia, PA
a r t i c l e
i n f o
Article history: Received 15 October 2014 Accepted 10 February 2015 Keywords: unicompartmental knee arthroplasty stress radiograph coronal alignment computer navigation
a b s t r a c t Routine preoperative stress radiographs have been advocated, in part, to determine “full correctability” of deformities before proceeding with unicompartmental knee arthroplasty (UKA) despite limited data supporting their utility. Fifty consecutive patients undergoing medial UKA with robotic navigation were studied. In 20° of flexion, significantly greater correctability was achieved after removal of osteophytes by an additional 1.8°, with a mean corrected alignment of 2.5° varus. Seventy-four percent of knees were not correctable to neutral alignment or more. In conclusion, preoperative stress radiographs have overstated value in patients undergoing medial UKA since the full extent of correctability of varus deformity cannot be determined until after removal of osteophytes and since most deformities are not fully correctable to neutral in UKA. © 2015 Elsevier Inc. All rights reserved.
Unicompartmental knee arthroplasty (UKA) is proven effective for appropriately selected patients with single compartment knee arthritis. The success of UKA is dependent on several factors, including patient selection, surgical precision, and implant design [1–13]. When these elements are optimized, so too are postoperative knee kinematics and functionality, patient satisfaction and implant durability after UKA [14]. The valgus stress view radiograph has been popularized as a means of determining correctability of the deformity, ensuring maintenance of the lateral joint space, and indirectly assessing the integrity of the anterior cruciate and medial collateral ligaments [1,2,15,16]. It is often discussed as an important criteria for determination of a patient’s candidacy for medial UKA [1,2,12,15,16]. Proponents of the valgus stress xray often state that it must demonstrate “full correctability” of the deformity, without narrowing of the lateral joint space, for a patient to be considered a candidate for medial UKA [1,2,15,16]. The valgus stress radiograph is taken with the patient lying supine, with the knee in 20° of flexion to relax the posteromedial capsule and a valgus stress applied. To avoid a parallax effect, the xray beam is directed 10° caudad [2,15,16].
One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2015.02.008. Reprint requests: Jess H. Lonner, MD, Rothman Institute, 825 Old Lancaster Ave., 2nd Floor, Bryn Mawr, PA 19010. http://dx.doi.org/10.1016/j.arth.2015.02.008 0883-5403/© 2015 Elsevier Inc. All rights reserved.
Although often advocated as a standard tool in the preoperative evaluation for medial UKA, there is a paucity of literature evaluating the utility of preoperative valgus stress radiographs [16–18]. Our hypothesis is that the preoperative valgus stress radiograph is of overstated value for 2 reasons: (1) maximal correctability is not possible until after intraoperative removal of osteophytes; and (2) full correction of the deformity is not routinely achievable. Other potential uses of the valgus stress radiograph, such as indirectly assessing the integrity of the ACL by identifying the presence of a fixed varus deformity, determining integrity of the medial collateral ligament and integrity of the lateral compartment articular cartilage were not studied. Methods This is a prospective series of 50 consecutive patients who underwent robotically assisted medial UKA with intraoperative navigation. Institutional review board approval was obtained for this study and the United States Health Insurance Portability and Privacy Act (HIPPA) regulations were followed throughout. All procedures were performed by the senior surgeon. The mean patient age was 60, ranging from 49 to 85 years old. A medial peripatellar capsular incision was routinely used. Computer navigation methods were utilized to make measurements and plan surgery in conjunction with robotic bone preparation during UKA. The anchoring system is comprised of 2 bicortical pins in both the distal femoral and proximal tibial metaphyses, to which trackers and reflective spherical transmitters are attached. An infrared beam emitted from the hardware module and camera is reflected from the spherical
1234
T.M. Kreitz et al. / The Journal of Arthroplasty 30 (2015) 1233–1236
trackers on the arrays, and anatomic surface landmarks of the joint and mechanical and rotational axes of the limb are mapped and registered. The accuracy of measurements with intraoperative navigation has previously been validated to within 1° [19–21]. In each patient the resting (unstressed) varus mechanical alignment of the limb was determined. The medial osteophytes were assessed and subjectively categorized as being either small or large by the senior surgeon. Prior to removing the osteophytes a maximal valgus stress was applied to the knee to determine “correctability” of the deformity, both in full extension and 20° of flexion, and the corrected mechanical alignment recorded for each position in all patients. The osteophytes were then removed, maximal valgus stress was re-applied and the same measurements were made in both full extension and 20° of flexion. Mechanical alignment was analyzed using positive values for varus, negative values for valgus, and zero for neutral mechanical alignment. A regression analysis was performed, controlling for multiple measurements in each patient – osteophyte size (small or large); corrected (stressed) alignment both before and after osteophyte removal; and differences in stressed (corrected) mechanical alignment at full extension and 20° of flexion.
Results The mean unstressed resting mechanical axis in full extension was 8.2° varus alignment (range 1° to 16°). Before osteophyte removal, the mean corrected mechanical alignment was 5.1° ± 2.5° varus in full extension and 3.4° ± 2.7° varus in 20° of flexion. After osteophyte removal, the mean corrected mechanical alignment was 3.2° ± 2.4° varus in full extension and 1.6° ± 2.5° varus in 20° of flexion (range 6° varus to 4° valgus) (Table 1). A total of 13 patients (26%) were correctable to either neutral or valgus alignment. Thirty-seven patients (74%) had a postoperative varus stress alignment and were not correctable to neutral. Varus deformities were more correctable when medial capsule and ligaments were stressed in 20° of flexion than in full extension. Prior to and after osteophyte removal, the alignment was 1.7° ± 0.8° and 1.6° ± 0.1° more correctable in 20° of flexion compared to full extension, respectively. With the knees flexed 20°, 32% of knees with mechanical varus of 10° or less were correctable to neutral or beyond. Alternatively, 7.7% of those with mechanical varus of 10° or more were correctable to neutral or beyond. Significantly greater correctability was achieved after removal of osteophytes by an additional 1.8° ± 0.1° (P = 2.00E-16) in 20° of flexion. Additionally, after osteophyte removal, correctability in 20° of flexion was significantly greater than in full extension by an additional 1.6°± 0.1° (P = 2.00E-16) (Table 2). Regression analysis showed no significant change in stressed alignment whether the osteophytes removed were small or large. There was a mean of 0.2° ± 0.52° less correctability when small osteophytes were removed compared to large osteophytes (P = 0.653).
Discussion UKA has a specific set of indications for patients with single compartment osteoarthritis of the knee [3–5]. “Full correctability” of deformity on stress radiographs is often cited as a prerequisite for a successful outcome after medial UKA [1,15,16,18,22]. However, clinical studies suggest that slight undercorrection (i.e. leaving the knee in slight mechanical varus) may improve midterm and long-term outcomes and durability [6–10]. Additionally, even in the hands of those who have been strong advocates of using valgus stress radiographs to ensure that deformities are “fully correctable” to determine candidacy for UKA, 25% of medial UKAs were left in varus, without compromising clinical or radiographic outcomes [23]. Further, in addition to the absence of clinical data supporting the benefit of valgus stress radiographs on clinical outcomes, these studies are utilized in less than 20% of patients being considered for UKA [24]. Finally, there is discordance regarding the optimal role of stressed radiographs and how “full correctability” should best be defined. Our hypothesis was that preoperative valgus stress radiographs cannot fully determine the extent of correctability of the deformity. While proponents suggest that preoperative valgus stress radiographs are important to determine that deformities are “fully correctable,” our study shows that determination of maximal correctability cannot be achieved until after removal of osteophytes and that “full correctability” to a neutral mechanical axis is not achievable in the majority of patients undergoing medial UKA. In the present study, in 20° of flexion, application of maximal valgus stress after osteophyte removal produced an additional 1.8° of correction of the deformity. This statistically significant difference underscores that full determination of correctability is not achievable until after osteophyte removal. Second, in this consecutive nonselected series “full correction” to neutral alignment or beyond was only possible in 26% of patients undergoing medial UKA, even after osteophyte removal. Only 8% of patients with preoperative mechanical varus of more than 10° and 32% with preoperative varus of 10° or less were correctable to neutral or beyond. Further, given the success of medial UKA when leaving the deformity in slight mechanical varus, recommending “full correctability” is both impractical and unsubstantiated [6–10,23,25]. Other potential roles of the stress radiograph, such as ruling out a fixed deformity which would suggest chronic ACL deficiency, determining integrity of the medial collateral ligament and integrity of the lateral compartment articular cartilage were not studied. Our study supports the conclusions by Waldstein et al [17] that stress radiographs may be unnecessary, although the basis for those conclusions differs in our 2 studies. They concluded that they provided no additional information in the assessment of the quality of cartilage in the lateral compartment or correctability of deformity in 84 patients (91 knees) undergoing total knee arthroplasty for non-inflammatory varus arthritis [17]. They found that preoperative alignment of 10° or less mechanical varus was correctable to 0.1° varus (95% CI, 0.6° valgus to 0.5° varus), with 93% correctable to between 3° varus and 3° valgus
Table 2 Impact of Variables on Correctability. Table 1 Resting and Corrected (Stressed) Mechanical Alignment.
Variable Mean ± SD
Unstressed (resting) alignment Stressed alignment before osteophyte removal: full extension Stressed alignment before osteophyte removal: 20° of knee flexion Stressed alignment after osteophyte removal: full extension Stressed alignment after osteophyte removal: 20° of knee flexion
8.2° ± 2.9° 5.1° ± 2. 5° 3.4° ± 2.7° 3.2° ± 2.4° 1.6° ± 2.5°
Shows the mean mechanical axis alignment and SD of the limb in full extension without applying a stress (resting alignment), and corrected (stressed) alignment in full extension and 20° of flexion, both before and after osteophyte removal for 50 patients undergoing medial UKA.
Mean ± SD
P
Difference in stressed alignment after removal of small 0.2° ± 0.5° 0.653 vs large osteophytes Difference in stressed alignment in 20° flexion before vs −1.8° ± 0.1° 2.00E-16 after osteophyte removal After osteophyte removal, difference in stressed −1.6°± 0.1° 2.00E-16 alignment in 20° flexion vs full extension Shows impact of multiple variables on changes in alignment with applied valgus stress: size of osteophytes; stressed alignment in 20° flexion before and after osteophyte removal; and stressed alignment in 20° flexion compared to full extension after osteophyte removal. Positive values represent an increase in stress alignment and negative values a decrease. P b 0.05 was considered significant.
T.M. Kreitz et al. / The Journal of Arthroplasty 30 (2015) 1233–1236
on stress radiographs. In knees with greater than 10° varus, valgus force corrected deformities to a mean of 5.5° varus (95% CI, 4.3° to 6.8° varus); most (91%) undercorrected between 3.8° varus and 10.3° varus. The authors concluded that since most small deformities (those most commonly considered for UKA) were corrected to between 3° varus and 3° valgus, preoperative stress radiographs provided no additional information in the preoperative evaluation of patients who may be considered for UKA and were therefore unnecessary [17]. Their indirect determination of corrected mechanical alignment has inherent error both due to the method of measure and the error of using short films with a knee that is flexed [22] and may explain the greater extent of correction than we observed in our study. So while Waldstein et al felt that stress radiographs were unnecessary since most patients were correctable to “neutral”, ours questioned the role of stress radiography since we found that full correctability is not possible until osteophytes are removed and most knees could not be corrected to neutral. Our results and conclusions differed from those from a study by Tashiro et al which reported that preoperative full-length valgus stress radiographs were useful for evaluating correctability of varus deformities prior to UKA in 37 consecutive patients [18]. In their study, an arthrometer was used to impart a 150-N valgus force to the studied knees and mechanical axis deviation was measured using full length radiographs. Patients had a mean unstressed varus alignment of 6.5° ± 2.8° (range: 2.3° to 11.8° varus). With a valgus stress, the alignment was corrected to 0.2° ± 2.6° mechanical varus. It is not clear how many were corrected to 0° or beyond, but none were corrected beyond 2° valgus. It appears that the authors of that study achieved greater correction of deformity with instrumented and calibrated stress application than we achieved. It is not known how much manual force was applied in each case by the senior surgeon conducting our study, whether it was consistent throughout cases, and how it compared to the 150N valgus force applied in Tashiro’s study. Nonetheless, while our manual method may create some inherent variability in applied force and thereby confound accurate measurements, we feel that the method used in our study is more representative of the uncalibrated technique used by most surgeons performing preoperative valgus stress radiographs in clinical practice. While we are using navigated measures of alignment rather than preoperative stress radiographs to determine correctability, the navigated method of alignment measurement used in this study has advantages compared to the use of stress views with short plate radiographs. A study by Lonner et al found that as little as 10° of knee flexion has a statistically significant impact on the measured tibiofemoral alignment after TKA using short radiographs [22]. That study showed that the tibiofemoral valgus alignment is reduced by 1° when the knee is flexed 10° and it can be reduced further with rotation of the limb. Therefore the apparent correction of alignment that occurs with application of a valgus stress with the knee flexed 20° is overestimated when using short films, particularlly if the limb is even slightly rotated during the radiograph. This may explain why the studies by Waldstein et al [17] and Gibson et al [16] found greater apparent correctability than we did in the current study. Our study limits the error by using the mechanical, rather than the tibiofemoral, axis. However, we could not control for subtle errors in rotation that may have occurred during valgus stressing. Additionally, Yaffe et al found that there is a mean 5° discrepancy in measured mechanical alignment between navigated measurements and full-length films, with variations that can be as much as 12° [26]. Therefore, the short films commonly used when performing valgus stress xrays, as well as the arbitrary determination that the knee has been flexed 15°–20° and not rotated may render determination of tibiofemoral alignment on valgus stress radiographs relatively inaccurate. Further study can be considered to compare correctability by navigated and radiographic measures; however, our opinion based on the referenced studies is that navigation, as applied in the current study, is a reasonable surrogate for preoperative stress radiographs.
1235
Limitations of this study include the subjective sizing of osteophytes. Osteophytes were categorized as either small or large without specific measurement criteria. Nonetheless, the size of the osteophytes did not appear to make a difference in the extent of correctability of the deformities. Preoperative and corrected alignment of the mechanical axis was measured using a validated navigated system, allowing for consistency between measurements. However, we are accepting that our use of navigation is accurate to within 1°, based on published studies that have validated this method [19–21], and it is our assumption that our use of navigation is more accurate than short films. If this assumes a maximum 1° error, then our difference in correctability may be reduced to 0.8°, rather than 1.8°. The valgus stress applied for each patient was neither instrumented nor measured and therefore inconsistencies of the forces applied may have resulted, causing potential errors in measures of alignment. However, while this variability in applied force may have some inherent error, this technique is representative of what is commonly used in clinical practice. Finally, our study used a more narrow determination of “neutral” alignment compared to other studies. We considered “neutral” to be 0°, since in UKA the goal is typically to leave the knee in slight residual varus and not overcorrect beyond 0°. Nonetheless, even if we broadened the range of “neutral” alignment, still the majority of our patients were not correctable beyond 3° of mechanical varus. The clinical utility of the valgus stress radiograph in the preoperative assessment of patients undergoing UKA has not been proven in the literature. Our data demonstrate that the full extent of correctability cannot be determined until after removal of osteophytes in patients undergoing medial UKA with varus deformity, and that full correctability to neutral or valgus resting alignment is not achievable in the majority of patients undergoing medial UKA. In light of the finding that most varus deformities are only partially correctable even after osteophyte resection, surgeons should not recommend “full correctability” as a criterion for UKA if that means correction to a neutral mechanical axis. Additionally, using intraoperative navigated measures as a proxy for preoperative stress radiographs, our data do not support the routine use of preoperative valgus stress radiographs for determining the extent of correction of the deformity in patients undergoing medial UKA, although there may be other potential roles of the valgus stress radiograph that were not investigated in this study. References 1. Argenson JN, Parratte S. Medial unicompartmental knee arthroplasty: Fixed BEARING TECHNIQUES. Chapter 10 , In: Berend KR, Cushner FD, editors. Partial Knee Arthroplasty: Techniques for Optimal Outcomes. Philadelphia, PA: Elsevier Saunders; 2012. 2. Hurst JM, Berend KR. Mobile-bearing unicondylar knee arthroplasty: the Oxford experience. Clin Sports Med 2014;33(1):105. 3. Borus T, Thornhill T. Unicompartmental knee arthroplasty. J Am Acad Orthop Surg 2007;15:9. 4. Newman MT, Lonner JH, Ries M. Partial knee arthroplasty – patellofemoral, unicompartmental, and bicompartmental. Ch 9 , In: Glassman AH, Lachiewicz PF, Tanzer M, editors. Orthopaedic Knowledge Update: Hip and Knee Reconstruction 4. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2011. 5. Jamali AA, Scott RD, Rubash HE, et al. Unicompartmental knee arthroplasty: Past, present, and future. 2009;38:564. 6. Berger RA, Meneghini RM, Jacobs JJ, et al. Results of unicompartmental knee arthroplasty at a minimum of ten years of follow-up. J Bone Joint Surg Am 2005;87:999. 7. Pennington DW, Swienckowski JJ, Lutes WB, et al. Unicompartmental knee arthroplasty in patients sixty years of age or younger. J Bone Joint Surg Am 2003; 85:1968. 8. Argenson JA, Chevrol-Benkeddache Y, Aubaniac JM. Modern unicompartmental knee arthroplasty with cement: A three to ten-year follow-up study. J Bone Joint Surg Am 2002;84(12):2235. 9. Hernigou P, Deschamps G. Alignment influences wear in the knee after medial unicompartmental arthroplasty. Clin Orthop 2004;423:161. 10. Cartier P, Sanouiller JL, Grelsamer RP. Unicompartmental knee arthroplasty surgery: 10-year minimum follow-up period. J Arthroplasty 1996;11:782. 11. Emerson RH, Higgins LL. Unicompartmental knee arthroplasty with the Oxford prosthesis in patients with medial compartment arthritis. J Bone Joint Surg Am 2008;90:118. 12. Argenson JA, Parratte S. The unicompartmental knee: design and technical considerations in minimizing wear. Clin Orthop Relat Res 2006;452:137. 13. Meek RMD, Masri BA, Duncan CP. Minimally invasive unicompartmental knee replacement: rationale and correct indications. Orthop Clin N Am 2004;35:191.
1236
T.M. Kreitz et al. / The Journal of Arthroplasty 30 (2015) 1233–1236
14. Patil S, Colwell Jr CW, Ezzet KA, et al. Can normal knee kinematics be restored with unicompartmental knee replacement? J Bone Joint Surg Am 2005;87(2):332. 15. Oxford Partial Knee. Manual of the Surgical Technique. Biomet UK Ltd Waterton Industrial Estate Bridgend, South Wales CF31 3XA, United Kingdom; 2015. http:// www.biomet.nl/resource/5980/Oxford-Knee-Optec.pdf. 16. Gibson PH, Goodfellow JW. Stress radiography in degenerative arthritis of the knee. JBJS 1986;68-B:608. 17. Waldstein W, Bou Monsef J, Buckup J, et al. The value of valgus stress radiographs in the workup for medial unicompartmental arthritis. Clin Orthop Relat Res 2013; 471(12):3998. 18. Tashiro Y, Matsuda S, Okazaki K, et al. The coronal alignment after medial unicompartmental knee arthroplasty can be predicted: usefulness of full-length valgus stress radiography for evaluating correctability. Knee Surg Sports Traumatol Arthrosc 2014;22:3142. 19. Hauser R. Computer-aided 3D-navigation systems: a plea for an error model. HNO 2000;2:71.
20. Oberst M, Bertsch C, Lahm A, et al. Regression and correlation analysis of preoperative versus intraoperative assessment of axes during navigated total knee arthroplasty. Comput Aided Surg 2006;11:87. 21. Picard F, Gregori A, Leitner F. Computer Assisted Total Knee Arthroplasty: Validation of the Image Free Concept. Berlin, Germany: Pro Business; 2007. 22. Lonner JH, Laird M, Stuchin SA. Effect of rotation and knee flexion on radiographic alignment in total knee arthroplasties. CORR 1996;331:102. 23. Gulati A, Pandit H, Jenkins C, et al. The effect of leg alignment on outcome of unicompartmental knee replacement. J Bone Joint Surg (Br) 2009;91B:469. 24. Schindler OS, Scott WN, Scuderi GR. The practice of unicompartmental knee arthroplasty in the United Kingdom. J Orthop Surg 2010;18:312. 25. Chatellard R, Sauleau V, Colmar M, et al. Medial unicompartmental knee arthroplasty: Does tibial component position influence clinical outcomes and arthroplasty survival? Orthop Traumatol Surg Res 2013;99S:S219. 26. Yaffe MA, Koo SS, Stulberg SD. Radiographic and navigation measurements of TKA limb alignment do not correlate. Clin Orthop 2008;466:2736.
