The surgical approach selected in total knee arthroplasty has significant implications for the surgeon and patient. The selection of an approach must consider patient characteristics including existing deformity, prior surgical incisions, and level of exposure required to correctly place components and protect the soft tissue. It should provide the surgeon a level of flexibility for intraoperative modification in the setting of unexpected complexity or complication. In this chapter, the technical considerations of each approach will be described, with attention to the advantages, disadvantages, and associated outcomes.

Surgical approach to the knee joint starts with an understanding of the anatomy of the anterior knee soft tissues (Fig. 39-1). The classic approach to the knee, in absence of wound or prior surgical scars, is centered on the anterior knee. The incision is placed over the patella at midline, extending in both proximal and distal direction to a total of approximately 10 cm in length, and terminating 1 cm medial to the tibial tubercle. An alternative is to make a medial parapatellar skin incision, with the goal of achieving less tension across the incision in flexion. No differences have been noted in clinical studies comparing these approaches.¹,²,³ In this region where soft-tissue coverage is limited, it is critical for the surgeon to preserve skin vascularity for effective wound healing. The operating surgeon must be familiar with the vascular anatomy of the knee and minimize superficial dissection as to not devascularize the superficial skin and soft tissue. Moreover, a thick subcutaneous flap must be maintained above the extensor mechanism to preserve blood supply and prevent necrosis.

The neurovascular network of the anterior knee skin and soft tissue originates from the saphenous and descending geniculate arteries on the medial aspect of the knee. Therefore through the classic midline incision, the lateral skin flap experiences a greater level of hypoxia during the surgery.¹,⁴,⁵ The anastomoses that feed the anterior skin exist superficial to the deep fascia of the knee, and thus care should be taken to avoid superficial dissection that may put the overlying skin at risk for necrosis (Fig. 39-2). Likewise, the infrapatellar branch of the saphenous nerve is usually sacrificed with midline incision leading to lateral knee numbness. Though this is associated with no morbidity, the potential for lateral knee numbness after surgery should be disclosed with patients in the preoperative period.

An understanding of the medial-based neurovascular anatomy about the knee is important for all cases but critical for the creation of skin incisions in the patient with previous incisions. When selecting an area to place the incision, the surgeon must stay as lateral as possible to avoid devascularizing a large section of anterior soft tissue. It is also preferable to incorporate previous longitudinal skin incision and incise at right angles to previous horizonal incisions to better preserve skin perfusion and eliminate avascular skin bridges that lead to necrosis. Subcutaneous tissue flaps should be developed directly ventral to the extensor mechanism. This preserves the blood supply for the superficial flap carried through the subcutaneous tissue. A far lateral incision can be selected, and medial exposure can still be obtained because of the ability to lengthen the incision and develop large flexible flaps that provide that necessary exposure. Surgeons should be aware of large scars or soft-tissues defects that may be threatened by an anterior- or medial-based knee incision. Important characteristics include the age of the incision or scar, the circumstances, and history of difficulty with wound healing. In the setting of complex history with any of the above criteria, a plastic surgery consult should be considered preoperatively.

After skin incision and soft-tissue flap preparation, the extensor mechanism is clearly visualized (Fig. 39-3). All subsequent exposures described differ in their management of the extensor mechanism; however, the goal of any approach is to achieve adequate visualization for bone resection, component insertion, and soft-tissue balancing. The major anatomic landmarks and the different intervals that can be utilized for approach are described in Fig. 39-3.

Initially described by von Langenbeck in 1879, the medial parapatellar approach (i.e., median parapatellar or paramedian approach) offers an easily reproducible and extensile approach to the knee.⁶ More recently popularized by Insall, the approach provides excellent exposure and can be modified when additional exposure or soft-tissue release is required (as discussed in the following difficult exposures section). The classic version of this approach splits the quadriceps tendon in the medial one-third and releases both the rectus femoris tendon and vastus medialis fibers from the patella. More recent modifications of this approach have medialized the quadriceps incision, leaving the majority of the quadriceps tendon intact and only removing the vastus medialis from its patellar insertion. Insall’s modification describes dissection of the
extensor mechanism in a straight line over the anterior patella surface, continuing distal along the medial border of the patellar tendon.⁶ From there, the quadriceps expansion is peeled from the anterior patellar surface with sharp dissection until the medial border of the patella is visualized. At this point, the synovium is then divided, the fat pad is split in the midline, and the patella is everted laterally. Closure can be performed in either flexion or extension, as randomized prospective trial has shown no effect on outcome based on knee position during closure.⁷

The authors of this chapter prefer the following step-by-step approach, utilizing different modifications as needed for a repeatable and extensile exposure to the knee (Fig. 39-4).

The midvastus approach has been well described by Engh and Parks.¹² It utilizes an interval in the midsubstance of the vastus medialis at the superomedial border of the patella. By using this interval, it avoids disruption of the quadriceps tendon and vastus medialis insertion thereby maintains an intact extensor mechanism.

A standard anterior midline incision is created in flexion, and dissection is performed to fascia. The medial aspect of the patella and vastus medialis insertion should be adequately exposed with creation of the medial soft-tissue flap. The prepatellar bursa is reflexed medially off the anterior surface of the patella, while the dissection in the distal aspect of the incision is limited to skin and fat only. With the knee in a flexed position, the superomedial corner of the patella is identified and blunt dissection is performed full thickness through the vastus medialis parallel to muscle fibers approximately 4 cm proximal to the superomedial patella. The arthrotomy is then performed in this region, proximal to the patella and deep to the vastus split, and is carried through the retinaculum along the medial aspect of patella to 1 cm medial to the distal aspect of the tibial tubercle (Fig. 39-5). Care must be taken to preserve a cuff of soft tissue on the patella for closure. A subperiosteal release is then performed to the midcoronal plane while capsule, synovium, and bursa are reflected to the medial border of the tubercle. Necessary soft-tissue releases from the proximal tibia may now be performed. Note that, in order to fully evert the patella from this approach, the capsular folds of the suprapatellar pouch must also be released. With lateral patellofemoral
ligament release and fat pad excision, the joint capsule is reflected until the lateral plateau is clearly visualized, completing the approach. Closure of the arthrotomy and approach should be performed at 60° of flexion, starting at the intersection of the capsular and muscle-splitting region of the incision. The muscle split itself does not need to be sutured.¹³

Proponents of this approach contend the midvastus approach avoids disruption of the quadriceps tendon while providing excellent exposure. However, this approach should not be performed in obese patients, hypertrophic arthridities, knees that lack greater than 80° of flexion, and in the setting of previous high tibial osteotomies.¹² In addition, the surgeon must be aware of the maximal safe distance for proximal sharp dissection, 4.5 cm, which was demonstrated by Cooper et al.¹⁴ In comparing the midvastus approach to the median parapatellar approach, the avoidance of the extensor mechanism demonstrates some early improvement in pain control and return of function. An early study from White et al showed postoperative improvement in pain at both 8 day and 6 week time points, with an improved time to achieve straight leg raise. However, no differences were seen at 6-month follow-up between the midvastus and median.¹⁵ From an intraoperative standpoint, fewer lateral releases were observed with the midvastus approach. Early pain control and return of quadriceps function were confirmed in a study by Dalury and Jiranek,¹⁶ but refuted by Keating et al¹⁷ who found no significant differences at early or late time point. Interestingly, Parentis et al showed the midvastus split to alter EMG findings suggestive of an underlying component of nerve injury to the vastus medialis. They also noted a decrease in the number of lateral releases and overall blood loss in the midvastus approach compared to medial parapatellar approach.¹⁸ Despite these findings, no corresponding clinical deficits or outcome differences were noted in the study. A more recent meta-analysis performed by Liu et al combined 32 randomized controlled trials with a total of 2451 total knee arthroplasty cases to evaluate both the midvastus approach and subvastus approach to the median parapatellar. The midvastus approach demonstrated improved visual analog scale (VAS) pain scores and knee motion in the 1 to 2 week postoperative period but an overall
increase in operative time. No significant differences were noted in long-term clinical outcome.¹⁹ Overall, there is evidence of the theoretical benefit of sparing the extensor mechanism to decrease the early pain and aid in early recovery of knee motion and quadriceps function. However, this improvement does not persist at later time points and amounts to no significant clinical improvement at long-term follow-up.

The subvastus (or Southern approach) to the knee joint was first described in 1929 but has been more recently described and popularized by Hoffman et al.²⁰ Proponents of this approach state that it is a more anatomic approach, respecting the boundaries of the normal knee anatomy. In comparison to the medial parapatellar approach, the extensor mechanism is left undisrupted with preservation of the medial patellar blood supply. However, the subvastus approach has important contraindications to recognize for use and cannot be extended in cases of difficult exposure. These contraindications include revision arthroplasty, prior knee arthrotomy, previous high tibial osteotomy, and obese patients.²⁰,²¹ In these cases, proper exposure will be difficult to obtain because the arthrotomy cannot be extended, and there are limited options to further mobilize the extensor mechanism.

A standard anterior incision over the midline is performed with the knee in 90° of flexion. Dissection is performed to the superficial fascia. The superficial fascia is then incised in line with the skin incision and curving slightly medial at the level of the patella. This creates a plane that is then bluntly dissected to separate the fascia layer from the perimuscular fascia of the vastus medialis at the site of insertion. The inferior edge of the vastus is identified, and blunt dissection is continued as the surgeon lifts the muscle off the periosteum and intermuscular septum to a distance of approximately 10 cm proximal to the adductor tubercle. The insertion of the vastus medialis to the medial capsule must then be identified by placing the dissected muscle belly under tension. At this location, a transverse incision of the insertion is performed at the midpatellar level without disrupting the underlying synovium. The remainder of the vastus insertion remains attached to the patella and quadriceps tendon (Fig. 39-6). The extensor mechanism is then retracted in an anterolateral direction, while a curvilinear arthrotomy is performed from the suprapatellar pouch to the tibial tubercle underneath the tensioned muscle belly. The fat pad is incised along the medial edge, and soft-tissue releases are performed around the proximal tibia. With the knee in extension, the patella is then everted and dislocated laterally. The knee is then placed back into flexion slowly, while additional blunt dissection is performed to prevent excessive tension through the vastus medialis and patellar insertion. Once instrumentation is complete, patellar tracking can be assessed. If a lateral release is required, it is performed in an outside-to-inside fashion with the knee in full flexion. This differs from the standard inside-to-outside direction as the lateral aspect of the dorsal portion of the extensor mechanism has limited exposure. Closure of the approach is then initiated with a suture at the apex of the curvilinear fascial incision and completed with interrupted close of both capsule and fascial layers.

The reported advantages of this approach focus on the “anatomic-nature” of the dissection and arthrotomy, with preservation of the extensor mechanism and patellar blood supply. Early studies comparing the subvastus approach with the median parapatellar demonstrated improved patellar tracking and a decreased need for lateral release.²² In addition, multiple more recent analyses have reported a faster return of quadriceps function with decreased pain, improved motion, and outcome scores at early time points.¹⁹,²³,²⁴,²⁵,²⁶,²⁷ However, these same studies fail to demonstrate persistence of these findings or superiority to other approaches at later time points with the exception of one.¹⁹ A most recent systematic review reiterates a faster return of straight leg raise postoperatively but highlights the inconsistency and variability in the findings of improved Knee Society Scores at both early and late time points.²⁸ A table has been provided for comparison (see Table 39-1).²⁴,²⁶,²⁹,³⁰,³¹,³²,³³,³⁴,³⁵,³⁶,³⁷,³⁸,³⁹,⁴⁰,⁴¹ Preservation of the extensor mechanism may possibly decrease pain and enhance function in the immediate postoperative period, but these benefits do not persist at later time points.

First described by Bramlett,⁴² the medial trivector-retaining approach combines the extensile approach of the medial parapatellar and extensor mechanism preservation of the subvastus approach. The underlying principle is that the quadriceps muscle has medial, lateral, and superior vectors that contribute to the extensor mechanism. With an incision through the vastus medialis, the muscle never becomes detached from the quadriceps tendon and therefore a maximal amount of the medial vector is unaltered. In one clinical study, straight leg raise returned 2 days faster in the trivector-retaining approach, but no clinical significance was demonstrated.⁴³ Though quadriceps function may be improved with this approach in the immediate perioperative period, the long-term effect of dividing the muscle fibers has not been extensively studied or validated.

In this approach, a standard anterior midline incision is utilized with dissection performed to expose the extensor mechanism. Approximately 6 cm proximal to the patella, an incision and separation through the vastus medialis is performed approximately 10 to 15 mm medial to the
vastus medialis insertion to quadriceps tendon. This split is carried distal until approximately 1 cm medial to the patella and then is continued on to the tibial tubercle. Eversion of the patella and all subsequent soft-tissue release are performed as needed to achieve visualization of femur and tibia. Proponents claim a similar level of exposure comparable to the medial parapatellar, while preserving vastus medialis tendon contribution to the extensor mechanism. However, as previously highlighted, the long-term effect in terms of healing and innervation of the intermuscular incision has not been well studied. No long-term clinical superiority has been demonstrated with this approach.⁴³

First described by Keblish,⁴⁴ the lateral parapatellar approach was developed as a counter to the medial-based approach in the treatment of valgus knee deformity. Proponents of the approach believe that a medial-based approach does not provide direct access to the tight lateral structures, exacerbates the need for external tibial rotation, and requires additional lateral release for patellar tracking and soft-tissue balancing in the valgus knee. Therefore, a lateral-based approach would theoretically provide better access to the lateral joint pathology while incorporating necessary lateral releases into the approach

and arthrotomy.⁴⁴ It is important to understand that a lateral-based approach decreases medial side visualization, makes patellar eversion more difficult, and requires diligent soft-tissue management to prevent a large lateral defect upon closure.