Mechanical alignment of tibial stems in revision total knee arthroplasty

The Journal of Arthroplasty Vol. 18 No. 7 Suppl. 1 2003

Mechanical Alignment of Tibial Stems In Revision Total Knee Arthroplasty Brian S. Parsley, MD,* Nobuhiko Sugano, MD, PhD,† Roberto Bertolusso, MS,‡ and Michael A. Conditt, PhD‡

Abstract: This multicenter, retrospective study evaluates the radiographic results of achieving optimal tibial alignment in revision total knee arthroplasty (TKA) using a single modular CoCr cemented or cementless stemmed implant design. Stem size and length also were evaluated. The hundred ninety-nine Revision TKAs were performed between January 1993 and January 1996 by 13 experienced revision knee surgeons. The cases were subdivided into 5 comparative groups: (1) cemented stems, (2) 140-mm length canal-filling stems, (3) 140-mm length non– canal-filling stems, (4) 95-mm length canal-filling stems, and (5) 95-mm length non– canal-filling stems. The anteroposterior (AP) tibial alignment angle was measured. The canal-filling ratio (CFR) was determined by dividing the stem diameter by the endosteal diameter at the stem tip. Overall, the ability to achieve tibial alignment in the AP plane was more predictable when canal-filling (CFR ⱖ 0.85) cementless stems were used. This was further enhanced when long canal-filling cementless stems were selected. The least-predictable results and the highest probability of varus malalignment were achieved with cemented stems. Key words: revision TKA, mechanical alignment, stem fixation, tibia. © 2003 Elsevier Inc. All rights reserved.

Revision total knee arthroplasty (TKA) poses a variety of technically challenging problems. The tibial component must address the concerns of fixation, alignment, coverage, and bony deficiencies in suboptimal bone and soft-tissue conditions. Modular tibial components have been developed to address these varied concerns and problems. The ability to achieve longitudinal alignment in revision knee

arthroplasty is critical to the success of the revision procedure [1–3]. Failure to properly align the knee has been clearly demonstrated to be a major cause of loosening [1,2,4,5]. The redistribution of bone stresses in the proximal tibia through proper tibial component alignment is an important first step in the success of the revision procedure. Whereas the addition of modular stems of varied lengths, diameters, and cemented or press-fit options provides the surgeon with multiple alternatives to address the deficiencies presented, the selection of cemented or cementless stem fixation in revision situations remains a subject of continued debate. In the current study, we review the differences between the use of cemented versus cementless stems in achieving longitudinal alignment when a single contemporary revision knee implant design was used. We further evaluated the benefits of stem canal fill and stem length on the ability to achieve optimal mechanical alignment of the tibial tray.

From the *Department of Orthopedic Surgery, Baylor College of Medicine, Houston, TX, †Department of Orthopedic Surgery, Osaka University, Suita, Japan, and ‡Institute of Orthopedic Research and Education, Houston, TX. Benefits of funds were received in partial or total support of the research material described in this article from DePuy, Warsaw, Indiana, and Johnson & Johnson. Reprint requests: Brian S. Parsley, MD, 6550 Fannin, Suite 2625, Houston, TX 77030. © 2003 Elsevier Inc. All rights reserved. 0883-5403/03/1807-1218$30.00/0 doi:10.1054/S0883-5403(03)00302-4

33

34 The Journal of Arthroplasty Vol. 18 No. 7 Suppl. 1 October 2003

Materials and Methods This retrospective study involved 391 cases (370 patients with 21 bilateral revisions) from 13 experienced revision knee surgeons enrolled in a multicenter revision knee study using the DePuy Coordinate Revision Total Knee System (DePuy, Warsaw, IN). All patients had a minimum of 1 year of follow-up. The average age was 69 years, with a range from 36 to 91 years of age. Two patients were withdrawn from the study for inadequate radiographs, leaving 389 cases for review (368 patients; 192 females and 176 males). Three hundred four cases had undergone one prior procedure, whereas 85 cases had undergone 2 or more prior procedures. A modular cobalt-chromium posterior substituting revision total knee implant design with a fixed platform articulation was used in all cases. The tibial trays used were 5 AMK (primary tibial tray), 329 CRK (Morse taper stem-lock revision tray), and 55 Ultra (screw stem-lock revision tray). Eighty-six cases used the cemented taper stem option with a standard taper diameter measuring 50 mm (n ⫽ 14), 70 mm (n ⫽ 58), or 105 mm (n ⫽ 14) in length. The modular press-fit stem options were selected in the 95-mm length for 138 cases and 140-mm length for 160 cases. The diameters ranged from 10 mm to 24 mm in 2-mm increments. Following canal preparation for stem insertion, intramedullary guides with the trial stems attached were used to make the proximal tibial resection. The optimal target alignment goal with this system is perpendicular to the tibial longitudinal alignment in the anteroposterior (AP) plane (90° ⫾ 2°). AP tibial radiographs on 14 ⫻ 17-inch films were reviewed at the 12-month follow-up by a single experienced surgeon. No data were collected on the reproducibility of the measurements of the evaluator. The medial tilt of the tibial tray was measured from the AP radiograph. The canal-fill ratio (CFR) was evaluated in the AP plane for the 95-mm and 140-mm cementless stems. The CFR was determined by dividing the stem diameter at the stem tip by the endosteal diameter at the stem-tip location. Canal-filling stems were defined as stems with a CFR ⱖ0.85. The cases were subdivided into 5 categories: (1) tibias with cemented stems (Cem) using any of the tapered stem extensions of the 50-mm, 70-mm, or 105-mm lengths and the AMK primary trays (n ⫽ 91); (2) tibias with long, canal-filling stems (LCF) (140-mm length, CFR ⱖ 0.85) (n ⫽ 105); (3) tibias with long, non– canal-filling stems (LNCF) (140-mm

Fig. 1. CFR for the noncemented implant types (298 cases). , CFR ⬍ 0.85 (Non Canal Fill): 42%; , CFR ⱖ 0.85 (Canal Fill): 58%.

length, CFR ⬍ 0.85) (n ⫽ 55); (4) tibias with short, canal-filling stems (SCF) (95-mm length, CFR ⱖ 0.85) (n ⫽ 68); and (5) tibias with short, non– canal-filling stems (SNCF) (95-mm length, CFR ⬍ 0.85) (n ⫽ 70). ANOVA, 2-way ANOVA, Levene, Tukey, Sceffe´ , Shapiro, and chi-square tests were used for statistical analysis.

Results The CFR was calculated for the uncemented stem groups. Fifty-eight percent of this subset achieved a CFR ⱖ 0.85 (Fig. 1). A comparative distribution of cases fell into each of the 4 categories. The ability to achieve optimal longitudinal alignment (90° ⫾ 2°) was successful in 76% of the cases overall. A test to reject the concordance to a normal distribution showed no statistical significance (P ⫽ .11). The mean AP tilt of the tibial tray for the overall group postoperatively was 90.04° (STD ⫽ 1.8°) (Fig. 2). Analysis of the groups demonstrated a similar mean AP alignment but a wide difference in range (Fig. 3). The smallest standard deviation (1.35°) and the tightest range (5.7°) of AP alignments were found in the LCF group. The highest standard deviation (2.02°) and range (9.6°) were found in the Cem group. Comparison of the canal-filling stem subgroups (SCF ⫹ LCF) versus the non– canal-filling stem subgroups and cemented stems (SNCF ⫹

Mechanical Alignment of Tibial Stems • Parsley et al.

Fig. 2. AP tibial alignment for the 389 cases. , AP ⬍ 88° (varus): 12%; 88° ⱕ AP ⱕ 92°: 76%; , AP ⬎ 92° (valgus): 12%.

LNCF ⫹ Cem) revealed a significant difference (P ⬍ .001) in achieving the desired perpendicular longitudinal alignment of the tibial tray in the AP plane (STD ⫽ 1.45° vs. STD ⫽ 1.96°). The ability to obtain

Fig. 3. Median AP tibial alignment by implant type (Cem: cemented [n ⫽ 91]; LCF: long, canal-filling [n ⫽ 105]; LNCF: long, non– canal-filling [n ⫽ 55]; SCF: short, canal-filling [n ⫽ 68]; and SNCF: short, non– canal-filling [n ⫽ 70]).

35

Fig. 4. Proportion of outliers—AP alignment outside the range of 88°–92°— by implant type (■, Cem: cemented; 䊐, LCF: long, canal-filling; ,LNCF: long, non– canalfilling; , SCF: short, canal-filling; and , SNCF: short, non– canalfilling).

optimal alignment also was significantly better when the canal-filling stem groups (LCF ⫹ SCF) were selected as compared with the cemented stem group (Cem) (P ⬍ .001) (STD ⫽ 1.45° vs. STD ⫽ 2.02°). Significant differences were noted between the cemented stem group and each of the uncemented stem groups (P ⬍ .001). Evaluation of the alignment outliers (AP ⬍ 88° and AP ⬎ 92°) for each group revealed a significant difference (P ⬍ .01) in the proportions of outliers between the cemented and uncemented stems. When evaluating the varus outlier cases (AP ⬍ 88°), the cemented stems had a significantly higher probability (29.7%) of malaligning the stem in the varus direction than the uncemented stems (6.7%). The long, canal-filling stem group had the least chance of varus malalignment (3.8%). Valgus alignment was least likely to occur in the cemented stem group (5.5%). The short, canal-filling stem group had the highest percentage of valgus outliers (20.6%). The highest percentage of outliers (both varus and valgus) was seen in the cemented stem group (35.2%), whereas the lowest percentage was seen in the long, canalfilling stem group (16.2%) (Fig. 4). A 2-way ANOVA analysis failed to demonstrate any significant difference in AP alignment variability as a result of interaction between surgeons and implant selection (P group:surgeon ⫽ .59).

36 The Journal of Arthroplasty Vol. 18 No. 7 Suppl. 1 October 2003

Discussion The ability to achieve AP tibial alignment was more predictable and reliable when the canal-filling cementless stems were used in this multicenter study with the same implant system design. Evaluation of the subgroups of stems revealed that AP alignment could be consistently achieved in LCF stems. This is consistent with Bertin et al.’s report on stem revision arthroplasty in which he was able to achieve more consistent alignment with longer, canal-filling stems as well [1]. The ability to achieve appropriate longitudinal alignment improves the overall success of revision arthroplasty. Gustilo et al. reported on 56 consecutive revision TKAs and noted 5 cases of progressive radiolucencies. Three were revised and 2 are asymptomatic at this time. He noted that the majority of the failures were in varus alignment at the time of failure. Our study also demonstrates that the long, canal-filling cementless stems were more predictable in achieving AP alignment within a tighter range of variance. This was particularly evident when compared with cemented tibial stems of varying lengths. This group was less predictable in achieving AP alignment and had the highest probability of varus malalignment [2]. The mechanical benefit of the intramedullary stem extension has been supported by multiple authors [1,3,4,6,7]. Mechanical testing by Yoshii et al. on uncemented tibial prostheses reported a statistically significant decrease in subsidence (permanent sinking displacement of the tray relative to the bone surface on the loaded side) on the long-stem group (150-mm) when compared with the stemless or short-stemmed (75-mm) groups, and in mean lift-off (permanent lifting displacement of the tray relative to the bone surface on the unloaded side) on the long-stem group compared with the stemless group. The authors concluded that the tibial tray with a long stem could achieve better initial fixation of the implant to bone when compared with shortstem or no-stem groups [3]. Cameron and Jung [4] also reported on the benefits of diaphyseal stem

extensions in TKA. Stem extension decreased the occurrence of radiolucencies when 50-mm and 100-mm stem extensions were used [8]. Albrektsson et al., using Roentgen stereophotogrammetric analysis (RSA), found that the addition of metal backing and a 110-mm stem extension significantly reduced both migration and tilt over 2-year follow-up in primary total knees [6]. In conclusion, the ability to achieve tibial alignment in the AP plane was more predictable and reliable when canal-filling cementless stems were used. This was further enhanced when long, canalfilling cementless stems were selected. The least predictable results with the highest number of varus outliers occurred with cemented stems.

References 1. Bertin KC, Freeman MAR, Samuelson KM, Ratcliffe SS, Todd RC: Stemmed revision arthroplasty for aseptic loosening of total knee replacement. J Bone Joint Surg 67:242, 1985 2. Gustilo T, Comadoll JL, Gustilo RB: Long-term results of 56 revision total knee replacements. Orthopedics 19:99, 1996 3. Yoshii I, Whiteside LA, Milliano MT, White SE: The effect of central stem and stem length on micromovement of the tibial tray. J Arthroplasty 7:433, 1992 4. Cameron HU, Jung YB: Noncemented stem tibial component in total knee replacement: the 2- to 6-year results. Can J Surg 36:555, 1993 5. Friedman RJ, Hirst P, Poss R, Kelley K, Sledge CB: Results of revision total knee arthroplasty performed for aseptic loosening. Clin Orthop Rel Res 255:235, 1990 6. Albrektsson BFJ, Ryd L, Carlsson LV, Freeman MAR, Herberts P, Regner L, Selvik G: The effect of a stem on the tibial component of knee arthroplasty. J Bone Joint Surg 72:252, 1990 7. Murray PB, Rand JA, Hanssen AD: Cemented longstem revision total knee arthroplasty. Clin Orthop Rel Res 309:116, 1994 8. Cameron HU: Clinical and radiologic effects of diaphyseal stem extension in noncemented total knee replacement. Can J Surg 38:45, 1995