Comparison of the subvastus and paramedian surgical approaches in bilateral knee arthroplasty

The Journal of Arthroplasty Vol. 8 No. 5 1993

Comparison of the Subvastus and P a r a m e d i a n Surgical A p p r o a c h e s in Bilateral Knee Arthroplasty B r u c e T. F a u r 6 , M D , * J a m e s B. B e n j a m i n , MD,-]- B e t s y L i n d s e y , PT,:I: R o b e r t G. Volz, MD,-]- a n d Del S c h u t t e , M D w

Abstract: A prospective randomized study was performed on 20 patients undergoing one-stage bilateral knee arthroplasty. One knee was exposed using a standard median parapatellar arthrotomy and the other knee with a subvastus arthrotomy. All patients underwent quantitative strength testing before surgery and at I week, l month, and 3 months after surgery. "l'he knees were also evaluated for range of motion, and patients, who were blinded as to the approach used, completed questionnaires at each evaluation period as to their preference, if any, regarding knee pain and level of function. There was no difference in the range of motion between knees exposed with the paramedian or snbvastus arthrotomy at any time period. The subvastus knees demonstrated significantly greater strength at the l - w e e k and l - m o n t h intervals, but there was no strength difference at the 3-month interval. There were more lateral releases performed in the paramedium knees, and three minor complications were related to the subvastus approach. Patients who expressed a preference chose the subvastus knee 4: l over the paranaedian knee. The subvastus approach offers a reasonable alternative to the paramedian arthrotomy and preserves greater quadriceps strength in the early postoperative period. Key w o r d s : parapatellar arthrotomy, subvastus arthrotomy, range of motion, strength.

Total knee arthroplasty (TKA) has been procedurally refined over the last decade and a half to yield consistently good results with few" complications. The standard median parapatellar arthrotomy approach, initially described by Von Langenbeck, ~7 is routinely employed as it provides excellent joint exposure. Postoperative patellar complications, including avascular necrosis of,:the patella, patellar malalignment, subluxation, and dislocation have, however, been associated with this approach. ~-3.~ Recently, interest has been focused on the subvastus,

or quadriceps-sparing arthrotomy because of reports of decreased wound complications, a shorter hospital stay, decreased analgesia requirements, and an earlier return to function when compared to the standard median parapatellar arthrotomy. 7-8.~2.~6 The theoretical advantages of the subvastus approach are based on the preservation of an intact quadriceps mechanism, Rather than severing the insertion of tile vastus medialis from the quadriceps tendon and medial patellar retinaculum, the subvastus approach preserves the extensor tendon mechanism by elevating the inferior border of the vastus medialis from the medial intermuscular septum and adductor aponeurosis. 9"~~ Because of the sizable number of patients uridergoing one-stage bilateral TKA at the University of Arizona, it was felt that such a patient population would be ideal to study the relative merits of these two ap-

Front the *Milwaukee Medical Clinic SG Milwaukee, Wisconsht, tSection of Orthopedic Sttrgery and :fRehabilitation Services, University of Arizona llealth Sciettces Center, Tttcsott, Arizona, and w Surgery, Medical University of South Carolina, Charh'ston, South Carolina.

Reprint requests: James B. Benjamin, MD, Section of Orthopedic Surgery, Universityof Arizona Heahh Sciences Center, 1501 North CampbellAvenue,Tucson, AZ 85724.

511

5.12 The Journal of Arthroplasty Vol. 8 No. 5 October 1993

proaches. 4-6.~3 Using a prospective randomized study of one-stage bilateral TKAs, with one knee undergoing a subvastus arthrotomy and the other undergoing the standard median parapatellar approach, we assessed the relative benefits of these two surgical exposures on the patients' postoperative recovery.

Materials and Methods Twenty patients with symmetric disabling arthritis undergoing one-stage bilateral knee arthroplasty at the University of Arizona Health Sciences Center (Tucson, AZ) between June 1990 and March 1991 were prospectively evaluated. All patients underwent strength testing before surgery and at 1 week, 1 month, and 3 months after surgery. Quadriceps strength was quantitated by measuring peak torques obtained during concentric isokinetic extension of each knee on a LIDO (Loredan, Davis, CA) device. Measurements were made at both 60~ and 120~ testing intervals. Five to seven trials Were performed at each speed and the average torque was recorded. Range of motion for each knee was also recorded at each testing interv.al. Strength testing and range of motion measurements were performed by a single observer (B. L.) who was blinded as to the surgical approach used. Patients, who also were blinded as to the surgical approach and had consented prior to surgery, completed questionnaires at each evaluation period as to their preference, if any, regarding knee pain and level of function. The results of the questionnaires were unknown to the surgeon until the completion of the study. All knees were surgically approached through a midline incision. By random selection, decided by the senior surgeon at the time of surgery, one knee was exposed by a standard median parapatellar arthrotomy and the other utilizing a subvastus approach as described by Hoffman.7.s Surgery was performed under tourniquet control, which was released after wound closure, and either tricompartmental or unicompartmental arthroplasty was performed based on Lhe pathology encountered at the time of.surgery. All unicompartmental prostheses were cemented. Tricompartmental arthroplasties were performed in a hybrid fashion with a cemented metal-backed tibial component and an all-polyethylene patellar component. The femoral component was a press:-fit, porous-coated design. In cases of severe osteopenia or loss of bone stock, the femoral component was also cemented. In all patients, the procedure performed on each knee was identical excluding the surgical approach and lateral

release when performed. Intraoperative patellar tracking was assessed in each knee after the definitive prostheses were inserted. The patella was reduced in the trochlear groove and t h e knee was passively flexed from 0 ~ to 90 ~ with no digital pressure on the patella or extensor mechanism. If patellar tilt or lateral subluxation resulted in failure of the medial facet from contacting the femur during this arc, a lateral release was performed. All knees were drained, postoperative drainage was collected using a cell saver, and the red blood cells salvaged during the first 4 hours were returned to the patient. ~~The drains were then connected to standard closed-suction wound drain reservoirs and subsequent drainage was recorded until the drains were discontinued at 48 hours. All patients received prophylactic antibiotics intravenously for 48 hours. Bedside physical therapy was initiated on tile first postoperative day, with knee motion and ambulation permitted on day 2. Continuous passive motion was utilized only if knee range of motion failed to show at least 60 ~ flexion, an extension lag of 20 ~ or a motion arc of less than 50 ~ by the third postoperative day. Statistical analysis of the data was performed using a MANOVA approach, with t-tests using a between/within analysis of variant design.

Results The study group consisted of 11 men and 9 women with an average age of 70 years (range, 55-81 years). Nineteen patients were diagnosed with osteoarthritis and one with rheumatoid arthritis. Twelve patients presented before surgery with genu varum deformities averaging 7.5 ~ (range, 5~ ~varus), the remaining eight patients presented with genu valgum deformities averaging 15~ (range, 50-25 ~ valgus). Fourteen patients underwent bilateral TKA and six underwent bilateral unicompartmental arthroplasty. Tourniquet times for the entire group averaged 74 minutes for knees exposed with the subvastus approach and 71 minutes for knees exposed with a parapatellar arthrotomy. Tourniquet times for tricompartmental arthroplasties averaged 79 and 73 minutes, while unicompartmental arthroplasties averaged 55 and 66 minutes for the subvestus and parapatellar approaches, respectively. Despite the apparent time benefit of the subvastus approach in the unicompartmental subgroup, there were no significant differences in either group, with respect to tourniquet time. Intraoperative assessment of patellar tracking was performed by visualization of patellofemoral congruency and tracking when the knee was placed through a passive arc of 0~ ~with no digital

S u b v a s t u s and P a r a m e d i a n A p p r o a c h e s in TKA

9

Faur~ et al.

513

Table 1. Range of Motion (degrees) Preoperative Flexion: Subvastus vs Parapatellar All Total Knee arthroplasty Unilateral knee arthroplasty Extension: Subvastus vs Parapatellar All Total knee arthroplasty Unilateral knee arthroplasty

1 Week

1 Month

3 Months

l 12/111 108/105 125/129

87/87 85/85 94/92

97/97 94/95 103/102

107/107 103/105 118/115

- 7/- 5 - 7/- 6 -6/-2

-10/-11 -lO/-11 -10/-12

-8/-6 -7/-6 -9/-6

-4/-4 -4/-4

pressure on the quadriceps mechanism. This assessment led to lateral retinacular release in five knees in the median parapatellar group and two knees in the subvastus group. All knees requiring lateral release were tricompartmental arthroplasties. W o u n d drainage following postoperative blood salvage averaged 375 mL in the subvastus group and 411 mL in the parapatellar group. Postoperative drainage in the tricompartmental subgroup averaged 428 mL and 473 mL, while unicompartmental arthroplasties averaged 151 mL and 142 mL for the stlbvastus and parapatellar approaches, respectively. Ahhough postoperative blood loss did differ significantly between the tricompartmental and unicompartmental subgroups, there were no significant differences related to the type of surgical exposure. Preoperative range of motion averaged 6 ~ ~ This arc of motion was essentially regained over a 3m o n t h peri()d. Postoperative range of motion failed to show any difference between knees with respect to surgical approach. Assessment of flexion and extension at 1 week, 1 month, and 3 months after surgery showed little variation, regardless of approach, differing by no more than 3 ~ of flexion or 3 ~ of exten-

-2/-5

sion at any time interval (Table 1). These differences were not significant. Quadriceps strength as measured with tile LIDO did demonstrate a difference between knees at the 91-week and 1-month testing intervals. At 1 week, the knee exposed with the subvastus approach showed 35% greater strength at the 60~ testing and 38% greater strength at the 120~ testing. At 1 month, the strength differences were 16% and 12%, respectively, again favoring the knee exposed with the subvastus approach. At both time intervals these strength differences were significant ( P < .05). By 3 months, there were no significant strength differences between the knee s exposed with the subvastus or parapatellar arthrotomy (Figs. 1, 2). These findings were consistent for both the tricompartmental and unicompartmental subgroups. Strength measurements obtained at the 3-month interval did not differ significantly from measurements obtained before surgery (P < .05), which would su,~gest that the patients had returned to their baseline strength. The patient questionnaires demonstrated that after surgery nine patients preferred the knee exposed with the subvastus approach in regard to the level

Sub-Vastus [ ]

vs Parapatellar E J

..Q

Fig. 1. Quadriceps strength measured as peak torques at 60~ at each test interval.

E

(D

Q_

ALL TKA UKA Pre-Op

ALL TKA UKA 1 Week

ALL TKA UKA 1 Month

ALL TKA UKA 3 Months

514

The Journal of Arthroplasty Vol. 8 No. 5 O c t o b e r 1993

80--~

Fig. 2. Quadriceps strength measured as peak torques at 120~ at each test interval.

Sub-Vastus

[]

vs Parapatellar

[]

70-

--

60

o

5040-

x O

n

30-

2010_

ALL TKA UKA Pre-Op

of comfort and perceived strength, two patients preferred the knee approached with the parapatellar arthrotomy, and nine patients expressed no preference. Of the nine patients who preferred the subvastus knee, three identified that knee to be the more disabling or painful knee before surgery. Of the two patients who preferred the paramedian knee, one considered that knee to be the more disabling knee before surgery. Three knees developed postoperative hematomas. Two subvastus knees developed medial-thigh hematomas that were transiently symptomatic. The presence of the hematomas did not affect overall knee performance or range of motion. One knee approached with a paramedian arthrotomy developed an intracapsular hematoma that required aspiration. One patient failed to achieve satisfactory range of motion and required bilateral manipulation on tile 9 ninth postoperative day, as he had achieved only 60 ~ of flexion in both knees. Symptomatic patellar subluxation developed in a snbvastus knee 2 months after surgery. Before surgery the patient had genu valgum deformities, and the knee exposed with the paramedian arthrotomy underwent a lateral release at the time of surgery to improve patellar tracking. Patellar tracking in the knee exposed with the subvastus approach appeared to be satisfactory at the time of surgery and was not released. The patient has improved significantly with quadriceps strengthening exercises and declines further surgical intervention. No infections or episodes of deep vein thrombosis were noted in any patient.

Discussion In this study, the subvastus arthrotomy appeared to offer an early advantage to the knee arthroplasty patient in postoperative rehabilitation as demon-

ALL TKA UKA 1 Week

ALL TKA UKA 1 Month

ALL TKA ~KA 3 Months

strated by significantly improved quadriceps strength at 1 week and, to a lesser extent, 1 month after surgery. This finding correlates with previous reports of earlier attainment of unassisted straight leg raises 7"8" 12.16 and unassisted crutch walking. 7 However, this early increased quantitative quadriceps strength does not appear to provide any long-term advantage, as knees exposed with the parapatellar approach attained equivalent levels of quadriceps strength by 3 months after surgery. As with other reports, 7"~2 our study showed no advantage in postoperative range of motion provided by the subvastus approach at any measured time interval, nor significant differences in operative time, blood loss, or eventual patient satisfaction. Despite the limited advantages-provided by the subvastus arthrotomy in the early postoperative phase, there may be several anatomic advantages helpful for the long-term survival and function of the arthroplasty, particularly in TKA. The patella receives its blood supply from a parapatellar extraosseous ring of vessels fed by the supreme (descending) geniculate, superomedial and inferomedial geniculate, superolateral and inferolateral geniculate, and anterior tibial recurrent arteries. ~4The median parapatellar arthrotomy severs the contribution of tile supreme, superomedial, and inferomedial geniculate vessels to the perivascular ring. The additional requirement in some knees for lateral retinacular release has been shown to significantly decrease intraosseous patellar blood flow. ~' Although not observed frequently, it has been suggested that the combination of lateral release and parapatellar arthrotomy may precipitate avascular necrosis of the patella, possibly increasing the incidence of patcllar fractures and component loosening.~-3 The subvastus arthrotomy predominantly preserves the contribution of the medial patellar blood supply, and theoretically

Subvastus and Paramedian Approaches in TKA

may decrease the likelihood of patellar avascular necrosis, although this complication is rarely seen. Two postoperative medial-thigh hematomas were encountered in knees utilizing the subvastus approach. The vascular a n a t o m y around the knee has been well described in the literature. ~'~]'~4 Proximal elevation of the vastus medialis from the medial intermuscular septum is limited to a distance of approximately 10 cm by the presence of articular branches of the supreme (descending) geniculate artery and its accompanying veins. Further proximal dissection risks damaging either the intermuscular septal branch, coursing towards its parapatellar anastomosis with tile superomedial and superolateral geniculates, or the articular branch on the medial intermuscular septum. Injury to these vessels represent the probable source of hematomas in our series (Fig. "3). Care must be taken during both the blunt elevation of the vastus medialis and in the stretching of the vastus medialis w h e n everting the patella and flexing the knee to avoid injury to these branches. Other technical considerations in utilizing the subvastus approach are the difficulty in initially visualizing the lateral compartment and the danger of increased stresses on the patellar tendon insertion on the tibial tubercle. Contracted lateral soft tissues, relatively short femoral stature, and obesity contribute to these obstacles. The risk of avulsion of the patellar tendon insertion can be lessened significantly by gradually bringing the knee into flexion to allow for the viscoelastic nature of the quadriceps to adjust, and, if needed, through release of the infrapatellar fat pad and anterior horn of the lateral meniscus. A

d""

L

;l

s

..

"~-

Faur6 et al.

515

small relaxing incision of the distal aspect of the vastus medialis tendon on the capsular retinaculum can also be utilized. These steps can be sequentially performed as needed for exposure in difficult cases but are not routinely required. Even after patellar eversion and maximal knee flexion, the lateral compartment can be difficult to visualize. Completion of the femoral flexion gap cuts usually relaxes the soft tissues sufficiently to allow unencumbered lateral visualization. It is extremely important to observe the patellar tendon during the initial dislocation of the patella, as avulsion of the tendon from the tibial tubercle is a devastating complication. Despite symmetric deformities and disease processes in the 20 paired knees studied, 5 lateral releases were required in the median parapatellar knees, in contrast to 2 releases in tile contralateral subvastus knees. It has been suggested that the maintenance of the quadriceps mechanism in an intact state, as offered by the subvastus approach, permits more accurate assessment of patellar tracking and more inherent stability of the patella. 7 Certainly, less intraoperative patellar tilt was evident in the assessment of patellar tracking in the subvastus knees in this study, leading to the decreased incidence of lateral retinacular releases. In one patient, however, this proved costly, as subsequent patellar subluxation occurred in the unreleased subvastus knee at 2 months. After surgery, a merchant's view revealed lateral patellar tilt, not present in the parapatellar approach knee, which u n d e r w e n t lateral release. The presence of preoperative patellar tracking anomalies on clinical exam or merchant's view shSuld be taken

I:

Fig. 3. Anteroposterior view of a knee specimen demonstrating the location of the supreme descending geniculate artery and vein (arrow) and their relationship to the vastus medialis and joint line.

9

516

The Journal of Arthroplasty Vol. 8 No. 5 October 1993

into consideration despite apparent acceptable intraoperative patellar tracking following a subvastus arthrotomy, since the intact vastus medialis can obscure visualization of the patellofemoral articulation and allow incomplete assessment of this area. This prospective randomized study, comparing the subvastus and parapatellar arthrotomies in patients undergoing bilateral knee arthroplasty, d e m o n strated improved quadriceps strength up to 1 m o n t h after surgery. This functional advantage dissipated by 3 months. There was no advantage to either approach in terms of postoperative range of motion. Although more patients expressed a preference for the knee with the subvastus approach, a majority expressed no preference between the two approaches. Previous sttldies have s h o w n reduced analgesic requirements, earlier straight leg raising, and short hospital stays in patients undergoing unilateral TKA. 7"12"16 In o u r experience, the subvastus arthrotomy's early strength advantage and potential anat o m i c and vascular benefits m a k e it a viable alternative to the traditional parapatellar approach. W e found that this a p p r o a c h m a y be used safely with a wide variety of deformities and body habjtus.

4.

5.

6.

7.

8.

9.

10.

l 1.

12.

Acknowledarnents 13. The authors Would like to t h a n k Karen Sue Burkett for careful preparation of the manuscript and Donald Witzke, PhD, for statistical analysis.

14. 15.

References I. Brick GW, Scott RD: The patellar femoral component of total knee arthroplasty. Clin Orthop 231 : 163, 1988 2. Cameron tt, Fedorkow DM: The patella in total knee arthroplasty. Clin.Orthop 165:197, 1982 3. Claytpn ML, Thirnpathi R: Patellar complications after

16.

17.

total condylar arthroplasty. Clin Onhop 170:132, 1982 Dodd CAF, Hungerford DS, Krackow A: Total knee arthroplas W fixation: comparison of early reports of paired cemented versus uncemented porous coated anatomic prostheses. Clin Orthop 260:66, 1990 "Dorr LD, Ochsuer JL, Gronlem J, Perry J: Functional comparison of posterior cruciate-retained versus cmciated sacrificed total knee arthroplasty. C]in Orthop 236:36, 1988 Enis JE, Gardner R, Robledo MA et al: Comparison of patellar resurfacing versus non-resurfacing in bilateral total knee arthroplasty. Clin Orthop 260:38, 1990 Hoffman AA, Brodie MR, Murdock LE: A new approach to the knee for total knee arthroplasty, ttarrington Society Transactions, Harrington Arthritis Research Center, Phoenix, 1989 Hoffman AA, Plaster RL, Murdock LE: Subvastus (southern) approach for primary total knee arthroplasty. Clin Orthop 269:70, 1991 Insall J: A midline approach to the knee. J Bone Joint Surg 53A:1584, 1971 Insall J: Surgical Approaches to the knee. p. 41. In Insall J (ed): Surgery of the knee. Churchill Livingstone, New York, 1984 Kayler DE, Lyttle D: Surgical interruption of patellar blood supply by total knee arthroplasty. Clin Orthop 229:221, 1988 Maric Z, Ou DM, Karpman RR: The standard versus the sub-vastus (southern) approach for total knee arthroplasty: a randomized prospective study. Orthop Trans 15:43, 1991 Morrey B, Adams R, llstrup D, Bryan R: Complications and mortality associated with bilateral or unilateral total knee arthroplasty. J Bone Joint Surg 69A: 484, 1987 Scapinelli R: Blood supply of the human patella. J Bone Joint Surg 49B:463, 1967 Slagis SV, Benjamin JB, Volz RG, Giordano GF: Postoperative blood salvage in total hip and knee arthroplasty: a randomized controlled trial J Bone Joint Surg 73B:591, 1991 Smigielski M, Gustke K: Southern approach for total knee replacement. Florida Orthopedic Society, Tampa, 1989 Von Langenbeck B: zur resection des kniegeinke. Verh Dtsch En Geseuch F Chir VII:23, 1879