Primary total knee arthroplasty is a safe procedure with reproducible results for most patients. However, challenges exist for patients with marked preoperative deformity, ligamentous instability, bone loss, and other structural problems affecting the knee. These problems occasionally require increased varus and valgus constraint or even a rotating-hinge construct to achieve adequate stability (Fig. 1). We define complex primary total knee arthroplasty as any primary total knee arthroplasty that requires increased implant constraint at the time of the index surgical procedure to achieve a stable knee. Although this definition does not include deformity that may be corrected with primary implants, it allows us to easily identify the at-risk patient population by implant design utilized at the time of the surgical procedure. This definition encompasses a broad spectrum of patient diagnoses including posttraumatic arthritis, congenital deformity, and Charcot arthropathy.
Currently, to our knowledge, there has been limited information available on the long-term outcomes in patients who require a constrained primary total knee arthroplasty1-3. Additionally, small studies that have evaluated this patient population tend to evaluate only one implant type or have limited intermediate-term follow-up4-15. Long-term outcomes in patients requiring a complex primary total knee arthroplasty are largely unknown. Additionally, we are not aware of a study that has compared the various levels of constraints and their effect on implant survivorship.
To our knowledge, this study is the largest comparative study on complex primary total knee arthroplasty. The null hypothesis for this study was that there would not be any statistical difference between routine and complex primary total knee arthroplasty populations with respect to implant survival, revision, or reoperation rates.
Materials and Methods
Institutional review board approval was obtained per institutional guidelines. The Mayo Clinic Total Joint Registry was evaluated from 1979 to 2013 to identify all primary total knee arthroplasties. After excluding patients who had undergone patellofemoral and unicompartmental knee arthroplasties, as well as patients with unrecorded implants, we identified 28,667 patients who had undergone primary total knee arthroplasties (posterior stabilized or cruciate retaining). Implant stability was assessed independently by 1 of the 10 surgeons performing the primary surgical procedure. In each case, the surgeon attempted to utilize the least amount of constraint to achieve a stable knee. If a cruciate retaining or posterior stabilized design did not achieve adequate stability, a varus and valgus constrained implant or a rotating-hinge implant was used. Implants included varus and valgus constrained as well as hinged implants from 1 of 3 implant manufacturers: DePuy, Zimmer, and Stryker. Selection criteria for our study included primary total knee arthroplasty, a minimum follow-up of 2 years, and a revision total knee implant used at the time of the primary surgical procedure. Utilizing these criteria, 673 patients (2.3%) were identified: 427 patients had a varus and valgus constrained implant and 246 patients had a rotating-hinge implant. Survival analysis was performed, following patients until their revision and reoperation or even to last contact. The median follow-up for the population was 10.1 years (range, 0 to 34.2 years). Demographic and clinical data are routinely updated from inpatient and outpatient electronic medical records to ensure accurate and up-to-date information on patient outcomes in the Joint Registry.
Demographic data were recorded for each implant type, including patient age at the time of the index surgical procedure, sex, body mass index (BMI), and preoperative diagnosis. The standard patient follow-up included a recheck with clinical and radiographic evaluation at time intervals of 3 months, 1 year, 2 years, 5 years, and every 5 years thereafter. Complications and reoperations were recorded on a continuous basis throughout the duration of the study. An all-cause complication rate and the 5 most common complications leading to reoperation or component revision were recorded for each implant. These include wound complication (wound dehiscence requiring closure or debridement), osteolysis and polyethylene wear (requiring revision of polyethylene or all implants), infection (including superficial and deep), stiffness (range of motion <90° requiring a manipulation), and fracture (any periprosthetic fracture requiring reoperation).
Baseline characteristics and patient demographic characteristics were compared among the 3 groups using a chi-square test for discrete variables and an analysis of variance (ANOVA) for continuous variables. In the assessment of the revision and reoperation outcomes, patients were followed to the event of interest, revision or reoperation, or otherwise until death or the last patient contact. The Kaplan-Meier survival was used to estimate the rates of survival free of implant revision and reoperation for each implant type, with estimates reported at 5, 10, 15, and 20 years when available. Cox proportional hazards regression was used to assess the association between implant type and the risk of revision and reoperation. The results of the Cox models are reported as a hazard ratio and a 95% confidence interval (95% CI). The most common mechanisms for revision were examined. Multivariate Cox regression models included age, sex, BMI, diagnosis, and mechanism. The alpha level was set at 0.05 for significance.
There was a significant difference (p < 0.001 for all) among the implant types for age, sex, diagnosis, and BMI (Table I). The rotating-hinge group had the youngest mean age at 52.3 years followed by the varus and valgus constrained group at 64.6 years. The mean age for the unconstrained implant group was 68.2 years. A relatively higher percentage of female patients received constrained implants (63.4% for the rotating-hinge implant group and 67.2% for the varus and valgus constrained implant group) compared with male patients (36.6% for the rotating-hinge implant group and 32.8% for the varus and valgus constrained implant group). Degenerative joint disease was the most common diagnosis for all groups except for the rotating-hinge group. The varus and valgus constrained and rotating-hinge groups had relatively higher numbers of posttraumatic arthritis and congenital deformity diagnoses compared with the routine total knee arthroplasty group.
The median follow-up among the 28,667 knees was 10.1 years, with an interquartile range from 2.6 years to 11.9 years. The median follow-up in the 3 implant types was 5.7 years for the rotating hinge design, 10.1 years for the routine (unconstrained) total knee arthroplasty, and 5.3 years for the constrained design. During the follow-up period, 9,676 deaths resulted in censoring of total knee arthroplasties in the analyses; by implant type, there were 69 deaths in patients with rotating-hinge implants, 9,517 deaths in patients with routine implants, and 90 deaths in patients with constrained implants. The overall survival free of reoperation for the patients with varus and valgus constrained implants was 77.0% at 10 years and 59.6% at 20 years. The survival rates free of reoperation for the rotating-hinge group were 49.2% at 5 years and 16.9% at 10 years. These rates were significantly lower than in the routine total knee arthroplasty group, which had survival rates free of reoperation of 87.9% at 10 years and 76.7% at 20 years (Table II; Figs. 2-A through 2-D). The multivariate analysis yielded significant hazard ratios for age, sex, and BMI (p < 0.0001 for all). Compared with the reference routine total knee arthroplasty group, the age, sex, and BMI-adjusted hazard ratio for all-cause reoperation was significantly higher (p < 0.001) in both the varus and valgus constrained group (1.74 [95% CI, 1.36 to 2.23]) and the rotating-hinge group (2.07 [95% CI, 1.58 to 2.70]). The most common complications that led to reoperation were wound complications, infection, and stiffness. The hazard ratios for reoperations for wound complications were significantly elevated at 2.88 (95% CI, 1.92 to 4.32) for the varus and valgus constrained group (p < 0.001) and 2.12 (95% CI, 1.22 to 3.70) for the rotating-hinge group (p = 0.008).
The overall implant revision-free survival rates for the varus and valgus constrained implant were 90.0% at 10 years and 72.8% at 20 years (Table III). For the rotating-hinge implant at 10 and 20 years, the revision-free survival rates were 74.6% at 10 years and 40.3% at 20 years. The adjusted hazard ratio for all-cause revision was significantly increased for the varus and valgus constrained group at 1.65 (95% CI, 1.15 to 2.38; p = 0.007) but not for the rotating-hinge group at 1.48 (95% CI, 0.99 to 2.21; p = 0.054) compared with routine total knee arthroplasty. However, it should be noted that the latter nearly reached significance (p = 0.054). Taken together, the rate of component revision for any reason at 10 years was >2 times higher in the complex total knee arthroplasties (11.8%) compared with routine total knee arthroplasties (3.2%). At 20 years after the index surgical procedure, the component revision rate was >3 times higher in the complex total knee arthroplasties (29.4%) compared with routine total knee arthroplasties (7.7%).
The most common reasons for component revision were wear and osteolysis, infection, and fracture. The survival curves for revision by complication and implant type are shown in Figures 3-A through 3-F. At 10 years postoperatively, the survival free of revision for wear and osteolysis was 95.9% for the routine total knee arthroplasty group, which was comparable and not significantly different from the survival rates of 93.7% for the varus and valgus constrained group (p = 0.070) and 91.4% for the rotating-hinge group (p = 0.147) (Table III). The survival curve for wear and osteolysis demonstrates a trend toward increased survival in the routine total knee arthroplasty group; however, this did not reach significance (p > 0.2) (Fig. 3-C). Component revision for infection was more common in the varus and valgus constrained group, with an adjusted hazard ratio of 2.48 (p = 0.045), and the rotating-hinge group, with a hazard ratio of 6.13 (p < 0.001), compared with the routine total knee arthroplasty group. Finally, we did observe a higher rate of fractures requiring component revision in the rotating-hinge group compared with the routine total knee arthroplasty group, with a hazard ratio of 6.32 (p < 0.001) (Table III).
Use of varus and valgus constrained or rotating-hinge implants in the primary total knee arthroplasty setting is occasionally necessary when bone loss, deformity, and/or collateral ligament laxity are severe. However, the benefits of the implant constraint can negatively affect implant survival and complication rates, which may lead to reoperation or revision. To our knowledge, there have been relatively few studies that have examined the outcomes of complex primary total knee arthroplasty, and we were unable to identify any study that compared the varus and valgus constrained and hinged implants in the setting of complex primary total knee arthroplasty16. We found that approximately 2% of patients in our study underwent a complex primary total knee arthroplasty. Demographic analysis revealed that these patients were significantly younger and were more likely to have distorted anatomy from underlying abnormalities such as trauma, previous infection, or even a previous surgical procedure. It is likely that our percentage of patients undergoing complex primary total knee arthroplasty is slightly higher than that in the general population because we are a tertiary referral center.
Overall, we found significantly higher reoperation and revision rates in patients undergoing complex primary total knee arthroplasties compared with those undergoing routine total knee arthroplasties. Most commonly, the reoperations were for wound complications, infection, and stiffness. There were no differences observed in the revision rates for wear and osteolysis between groups. However, the patients undergoing complex primary total knee arthroplasties did have a higher rate of component revision for infection and the rotating-hinge group had a higher revision rate for fracture.
The varus and valgus constrained implant group was noted to have good implant survivorship, with 90% survival at 10 years and 73% survival at 20 years. However, when a rotating-hinge implant was necessary to achieve stability, survival free from revision was significantly lower at 10 years (75%) and 20 years (40%). It is likely that the implant survivorship in the rotating-hinge group is lower than that in the varus and valgus constrained group because of the nature of the underlying abnormalities present in this cohort of patients. The use of hinged implants was reserved for marked collateral ligament deficiency or for patients in whom the flexion-extension mismatch was not correctable with lesser constraint.
One might assume that revision implants used in the primary setting would not perform as well as standard implants do secondary to the increased constraint and underlying abnormalities. However, we now counsel patients that we expect 73% implant survivorship at 20 years if they require a varus and valgus constrained implant and the results will be worse (around 40% survival) if a hinged implant is required. Taken together, the relative risk of all-cause revision at 10 years was >2 times higher in patients receiving a constrained implant compared with patients receiving an unconstrained implant at the time of the index surgical procedure. At 20 years postoperatively, this was >3 times higher. This is important to note for preoperative counseling when the arthroplasty surgeon suspects that constraint will be necessary, but the degree of constraint required will be an intraoperative decision.
This study does not allow us to directly compare implant designs. Patients who required a hinged implant required it because they could not achieve adequate stability with a varus and valgus constrained implant. As such, these data should not discourage use of constrained implants when necessary in the primary total knee arthroplasty setting. However, it may serve as a reference when discussing complications and revision rates with patients requiring complex arthroplasty. In general, the lowest amount of implant constraint should be used at the primary surgical procedure, realizing that approximately 2% of patients may require a constrained primary implant to achieve a stable knee.
The retrospective nature of the study involving a large patient population has its inherent limitations. Joint registry data rely on hard end points and can be limited when patients are lost to follow-up or undergo a revision surgical procedure at other institutions. Our definition of complex primary total knee arthroplasty included only patients who required a revision total knee arthroplasty implant design and therefore excluded patients who had substantial deformity that was correctable with a primary implant. Finally, because of the large patient population studied, we did not extensively evaluate clinical and radiographic data to support these results.
In summary, this study provides valuable information on the long-term outcomes of patients who underwent complex primary total knee arthroplasty. We compared 3 different implant designs that are most commonly utilized for patients with marked preoperative deformity or ligamentous instability. The relative risk of all-cause revision at 10 years was >2 times higher in patients receiving a constrained primary implant. At 20 years, this was >3 times higher. This information should be utilized to preoperatively counsel these patients on implant longevity and complication rates when the surgeon suspects that constraint will be required to achieve a stable knee at the time of the index surgical procedure. The implant failure and complication rates are significantly higher in this patient population compared with patients undergoing routine primary total knee arthroplasty.
Investigation performed at the Mayo Clinic, Rochester, Minnesota
A commentary by Harry B. Skinner, MD, PhD, is linked to the online version of this article at jbjs.org.
Disclosure: There was no outside source of funding used for this research. On the Disclosure of Potential Conflicts of Interest forms, which are provided with the online version of the article, one or more of the authors checked “yes” to indicate that the author had a relevant financial relationship in the biomedical arena outside the submitted work.
1. Springer BD, Hanssen AD, Sim FH, Lewallen DG. The kinematic rotating hinge prosthesis for complex knee arthroplasty. Clin Orthop Relat Res. 2001 ;392:283–91.
2. Lombardi AV Jr, Berend KR, Leith JR, Mangino GP, Adams JB. Posterior-stabilized constrained total knee arthroplasty for complex primary cases. J Bone Joint Surg Am. 2007 ;89(Suppl 3):90–102.
3. Baker P, Critchley R, Gray A, Jameson S, Gregg P, Port A, Deehan D. Mid-term survival following primary hinged total knee replacement is good irrespective of the indication for surgery. Knee Surg Sports Traumatol Arthrosc. 2014 ;22(3):599–608. Epub 2012 Dec 14.
4. Springer BD, Sim FH, Hanssen AD, Lewallen DG. The modular segmental kinematic rotating hinge for nonneoplastic limb salvage. Clin Orthop Relat Res. 2004 ;421:181–7.
5. Molenaers B, Arnout N, Bellemans J. Complex total knee arthroplasty using resection prostheses at mid-term follow-up. Knee. 2012 ;19(5):550–4. Epub 2011 Oct 5.
6. Sheth NP, Lonner JH. Clinical use of porous tantalum in complex primary total knee arthroplasty. Am J Orthop (Belle Mead NJ). 2009 ;38(10):526–30.
7. Barrack RL, Lyons TR, Ingraham RQ, Johnson JC. The use of a modular rotating hinge component in salvage revision total knee arthroplasty. J Arthroplasty. 2000 ;15(7):858–66.
8. Donaldson WF 3rd, Sculco TP, Insall JN, Ranawat CS. Total Condylar III knee prosthesis. Long-term follow-up study. Clin Orthop Relat Res. 1988 ;226:21–8.
9. Böhm P, Holy T. Is there a future for hinged prostheses in primary total knee arthroplasty? A 20-year survivorship analysis of the Blauth prosthesis. J Bone Joint Surg Br. 1998 ;80(2):302–9.
10. Guenoun B, Latargez L, Freslon M, Defossez G, Salas N, Gayet LE. Complications following rotating hinge Endo-Modell (Link) knee arthroplasty. Orthop Traumatol Surg Res. 2009 ;95(7):529–36. Epub 2009 Oct 17.
11. Hernández-Vaquero D, Sandoval-García MA. Hinged total knee arthroplasty in the presence of ligamentous deficiency. Clin Orthop Relat Res. 2010 ;468(5):1248–53.
12. Lachiewicz PF, Soileau ES. Ten-year survival and clinical results of constrained components in primary total knee arthroplasty. J Arthroplasty. 2006 ;21(6):803–8.
13. Morgan H, Battista V, Leopold SS. Constraint in primary total knee arthroplasty. J Am Acad Orthop Surg. 2005 ;13(8):515–24.
14. Sprenger TR, Doerzbacher JF. Long-term follow-up of the GSB II total knee used in primary total knee arthroplasty. J Arthroplasty. 2002 ;17(2):176–83.
15. Yang JH, Yoon JR, Oh CH, Kim TS. Primary total knee arthroplasty using rotating-hinge prosthesis in severely affected knees. Knee Surg Sports Traumatol Arthrosc. 2012 ;20(3):517–23. Epub 2011 Jul 20.
16. Hartford JM, Goodman SB, Schurman DJ, Knoblick G. Complex primary and revision total knee arthroplasty using the condylar constrained prosthesis: an average 5-year follow-up. J Arthroplasty. 1998 ;13(4):380–7.