Introduction

Osteotomy around the knee is a well-established and commonly employed surgical option for active patients with unicompartmental knee osteoarthritis [6, 28]. In recent years, a surgical technique has evolved with the introduction of minimally invasive biplanar opening wedge osteotomy with locking plate fixation, which enables early bone union and functional recovery [8, 18, 19, 29, 33]. Based on this progress, satisfactory mid-term outcomes have been reported for opening-wedge high tibial osteotomy (OWHTO) [11, 21]; however, limitations and shortcomings associated with correction in the proximal tibia alone have also been reported [3, 5, 22, 27, 34]. When correction of severe varus deformity is performed with HTO alone, a large wedge opening may result in non-physiologic lateral inclination of the joint line (Fig. 1). This situation can lead to several potential problems such as increased shear force at the cartilage [14, 20], femoral subluxation in full weight-bearing, overloading of the knee joint capsule, and difficulty in converting to total knee arthroplasty (TKA) [13, 15, 25, 26]. Considering the aforementioned shortcomings of HTO, double level osteotomy (DLO) has been introduced with the intention of restoring physiologic joint alignment and orientation for knees following correction of severe varus deformity (Fig. 2) [19, 32]. There have been several studies reporting the outcomes of DLO [2,3,4,5, 16, 27, 38]. In the studies published from the 1960s to the 1980s, the reported outcomes were rather discordant including both positive and negative results [2, 3, 5, 16, 32, 38]; however, those studies from the earlier days included both osteoarthritic and rheumatoid knees, and detailed consideration for knee alignment was not given in the surgical planning and procedure. Consequently, the clinical advantage of DLO over isolated HTO was not readily accepted in those days. Following 2000, the surgical concept of DLO has evolved for restoration of physiologic joint alignment and orientation. Babis et al. and Saragaglia et al. reported a satisfactory outcome of DLO with this concept [3, 27]. To date, however, there still is a paucity of evidence in literatures supporting the efficacy of DLO for restoration of normal anatomy and resultant symptomatic improvement. In addition, surgical technique has been further refined in the last decade with the introduction of the minimally invasive biplanar osteotomy technique and rigid fixation using a locking compression plate (LCP). Although these refinements in surgical technique are expected to improve the surgical outcome, results of the current DLO procedure have not been reported in literatures.

Fig. 1
figure 1

Preoperative planning of double level osteotomy using digital planning software. a Severe varus deformity of the right knee with combined deformity in both the femur and tibia (mLDFA: 90.6°, mMPTA: 81°, mTFA: − 16° varus). b Osteotomy simulation started with isolated OWHTO, and was considered for cases with an anticipated mMPTA of more than 95° or a wedge size of more than 15 mm. Simulation for OWHTO in this case shows that the wedge size was 17 mm and mMPTA was 97° when HTO alone was used to correct the varus deformity

Full size image
Fig. 2
figure 2

In preoperative planning, mLDFA was set at 85°. a In the preoperative surgical simulation, the required wedge size was 12 mm with an mMPTA of 91.7°. b At the postoperative evaluation, measured mTFA, mLDFA, and MPTA were 1.2°, 85°, and 90.9°, respectively. OWHTO opening wedge high tibial osteotomy, mLDFA mechanical lateral distal femur angle, mMPTA mechanical medial proximal tibia angle, mTFA mechanical tibio-femoral angle

Full size image

The purpose of this study, therefore, was to examine the early radiological and clinical outcomes following minimally invasive DLO performed with a modern concept and procedure. It was hypothesized that DLO, which is currently performed in our practice, would accomplish restoration of normal joint alignment and orientation with resultant clinical improvement.

Materials and methods

Twenty consecutive patients who underwent DLO in our institute during the study period from January 2014 to September 2016 were included in the study. The patient population comprised 5 males and 15 females with a mean age of 62.5 ± 6.8 years (range 45–76 years). All patients could be followed-up for at least 1 year after surgery, and the mean follow-up period was 20.7 ± 6.8 months (range 12–36 months).

Radiological and clinical evaluations were performed before surgery and at the follow-ups. In the radiological assessment, the following parameters were measured: mechanical tibiofemoral angle (mTFA), mechanical lateral distal femoral angle (mLDFA), mechanical medial proximal tibial angle (mMPTA), and joint-line convergence angle (JLCA) [9, 30]. Radiological measurement was performed on long-leg weight-bearing radiographs using digital planning software (mediCAD®, Hectec, Germany) [31]. In addition, validated patient-registered outcome measures, Knee Injury and Osteoarthritis Outcome Score (KOOS), and the score of the International Knee Documentation Committee (IKDC) Subjective Knee Evaluation Form were used for quantitative evaluation of clinical outcome.

Indications for a double level osteotomy and planning

In our unit, DLO is indicated when there is a combined deformity in the femur and the tibia in the active patient population and the patient wishes to retain high levels of activity or continue sitting on his or her heels (Japanese-style sitting). As a preoperative functional requirement, the level of pain should still allow the patient to have an active life style.

Radiological assessment and surgical planning were carried out on long-leg weight-bearing radiographs using digital planning software (mediCAD®, Hectec, Germany). In preoperative surgical planning, deformity correction by osteotomy was simulated to achieve restoration of physiologic joint alignment and orientation. Normal values for each of the radiological parameters reported in literatures are as follows: − 1° to 3° for mTFA, 85°–90° for mLDFA, 85°–90° for mMPTA, and 0°–2° for JLCA [23, 24]. Regarding the intended parameter values during the preoperative surgical simulation and planning of the osteotomy, the mTFA was set to 0.5°–1° (a slight valgus alignment), while mLDFA and mMPTA were set to 85° and 90°, respectively. Preoperatively, the surgical simulation was started with isolated OWHTO. If the predicted mMPTA in the surgical simulation was 95° or greater (or the wedge size was 15 mm or greater) for deformity correction with OWHTO alone, DLO was considered as a surgical option (Fig. 2) and surgical planning was conducted accordingly [20].

Surgical procedure

All surgeries were performed by the first author (H.N.) under general anaesthesia. Arthroscopy was performed prior to the osteotomy procedure. During arthroscopy, osteophytes in the notch were removed when present to improve extension. DLO was started with closing-wedge distal femoral osteotomy (CWDFO). A 4–5 cm longitudinal incision was made at the lateral side of the femur just above the femoral epicondyle according to the biplanar technique, which was developed by a working group from the Netherlands organized by R. van Heerwaarden [7, 10, 12]. Fixation of the osteotomy was accomplished following the minimally invasive plate osteosynthesis (MIPO) technique which was established by the same group [37] (Figs. 3, 4). The TomoFix medial distal femur (MDF) anatomical plate (DePuy Synthes, Solothurn, Switzerland) originally designed for the contralateral femur was bent according to the individual’s anatomy and positioned under the vastus lateralis muscle for osteotomy fixation. A drain was placed, and the layers were closed. At this stage, the alignment following CWDFO was checked under fluoroscopy for confirmation of correspondence with preoperative planning (Fig. 5). Afterwards, the osteotomy procedure was completed with subsequent OWHTO. The OWHTO procedure performed in this study followed the technique described by Staubli et al. and Lobenhoffer and Agneskirscner [19, 33]. After the osteotomy was made, the medial wedge was gradually opened while the limb alignment was monitored on fluoroscopy by checking the position of the alignment rod at the knee. Once the optimal location of the alignment rod at the knee corresponding to the preoperative planning was attained (with %MA of approximately 52–55%), the osteotomy site was fixed with that position. With the intent of improving the mechanical properties of the osteotomy site and enhancing bony union, a wedge-shaped bone substitute (ß-TCP: Osferion 60; Olympus Terumo Biomaterials, Tokyo, Japan) was inserted into the osteotomy gap in all of the cases included in this study [17, 35]. For fixation, a TomoFixTM medial high tibial (MHT) plate (DePuy Synthes, Solothurn, Switzerland) was placed anteromedially and the osteotomy site was fixed using lag screws (Figs. 6, 7). Finally, autogenous bony pieces and chips obtained from the femoral closing wedge and osteophytes were grafted into the tibial opening wedge to fill the remaining gap in the tibial opening wedge [1].

Fig. 3
figure 3

Surgical sequences for DFO in the left knee. a Small skin incision for DFO. b Four Kirschner wires inserted according to the planning. c Proximal vertical cut. d Ascending cut. DFO distal femoral osteotomy

Full size image
Fig. 4
figure 4

Intraoperative fluoroscopic sequences for DFO. a After biplane cut of DFO. b The osteotomy site closed by application of valgus stress. DFO distal femoral osteotomy

Full size image
Fig. 5
figure 5

Preoperative assessment and surgical planning using digital planning software (a and b) and the intraoperative fluoroscopic image (c, white arrows indicate the mechanical axis). a The radiological parameters measured by digital planning software. b Surgical planning for DFO in the preoperative surgical simulation. c Intraoperative fluoroscopic image after DFO showing the mechanical axis (indicated by the metal alignment rod) passing through the same position as in the preoperative planning. DFO distal femoral osteotomy

Full size image
Fig. 6
figure 6

Preoperative surgical planning using digital planning software and the intraoperative fluoroscopic image. a The preoperative planning using the digital planning software. b Intraoperative fluoroscopic image shows the mechanical axis passing through the same position as in the preoperative planning

Full size image
Fig. 7
figure 7

A 67-year-old female undergoing double level osteotomy for both knees with severe varus knee deformity. a Preoperative standing whole leg radiograph. b Postoperative standing whole leg radiograph

Full size image

Postoperative rehabilitation

The operated knee was not immobilized after surgery, and range of motion exercise was started as tolerated on the following day. Weight-bearing was not allowed for 3 weeks after surgery. Partial weight-bearing using crutches with 20 kg was started at 3 weeks with progression to full weight-bearing at 4 weeks.

Postoperative follow-up

Postoperative follow-up evaluations were periodically performed every 3 months within a year and then every 6 months afterward. Comprehensive radiological and clinical evaluations were conducted at 6 months and 1 year, and the results were compared to the preoperative status. To examine whether the surgical results could be attained as planned, radiological parameter values at the preoperative planning and the 1-year postoperative evaluation were compared for each case, and the difference was regarded an indicator for surgical accuracy. The study design was approved by the Institutional Review Board of Hyogo College of Medicine [No. 2218], and informed consent was obtained from all study subjects.

Statistical analysis

Differences between the pre- and postoperative clinical and radiological parameters were statistically assessed using the paired Student’s t test with a significance level set at P < 0.05. The statistical analysis was performed using SPSS version 19 software (IBM, Armonk, New York, USA). A post hoc power analysis was conducted for comparison of the pre-and postoperative radiological parameters using a 2-sided test. Consequently, it was shown that the sample size of 20 in this study could achieve an adequate power of 0.99 with an α of 0.05.

Results

As for postoperative complications, a major intraoperative arterial injury was encountered in a 44-year-old female patient. In this case, the popliteal artery was injured during the distal femoral osteotomy. Although major bleeding was stopped by local compression, immediate postoperative examination revealed clinical signs of circulatory deficiency. CT angiographic findings indicated injury to the popliteal artery. Six hours after surgery, the injured site was primarily repaired by the vascular surgeons. Although the operated limb could be salvaged, peroneal nerve palsy and drop foot due to a subsequent compartment syndrome remained as a sequel. During the study period, no other complication related to the surgical procedure was noted, and solid bony union was achieved in all knees before 1 year .

Radiological results (Table 1)

Table 1 Pre- and postoperative radiological parameter values for alignment and orientation of the knee
Full size table

Preoperatively, all knees exhibited varus deformity with the mean mTFA angle of 13.5° ± 3.1° varus (range 10.2°–23.6°). The mean mLDFA was 91.2° ± 1.4° (range 88.8°–95°), and the mean mMPTA was 82.3 ± 2.0° (range 77.2°–86.4°) indicating that sources of the deformity were present in both femoral and tibial sides. The mean JLCA was 5.8 ± 2.3° (range 2.5°–10.8°) showing increased joint convergence compared to the normal value.

Radiological evaluation at 6 months showed restoration of normal knee alignment with the mean TFA corrected from 13.5° varus to 1.1° valgus. Other radiological parameters for joint orientation, such as mLDFA and mMPTA, were corrected to normal ranges with significant postoperative changes. JLCA was also significantly reduced after surgery. When these values were repeatedly measured at 12 months, no changes were noted during the subsequent postoperative period, and thus restoration of normal alignment and orientation confirmed at 6 months was maintained up to 12 months after surgery. In the preoperative surgical simulation, the correction angles for the femur and the tibia were planned to be 6.5° ± 1.3° and 9.4° ± 2.4° on average, while the angles actually attained at surgery were 5.6° ± 1.8° and 8.2° ± 3.4°, respectively, based on the differences between the angle values measured before surgery and at 12 months. Consequently, surgical accuracy values, defined as the differences between the preoperative planning and 12-month postoperative measurements for mTFA, mLDFA, and mMPTA angles, ranged from 1.4° to 2.1° on average (Table 2), and satisfactory surgical accuracy could be accomplished.

Table 2 Surgical accuracy indicated by the difference between the radiological parameter values at preoperative planning and postoperative evaluation
Full size table

Clinical results

As shown in Table 3, clinical outcomes as assessed by the KOOS and IKDC subjective scores significantly improved after surgery. Clinical improvement was already evident as early as 6 months after surgery and maintained at 12 months.

Table 3 Pre- and postoperative clinical scores based on patient-registered outcome measures
Full size table

Discussion

The most important findings of the present study were that DLO performed for severe varus osteoarthritic knees in our practice could achieve restoration of normal knee alignment and orientation with resultant improvement in patient-registered clinical outcomes. No changes in radiological parameters from 6 to 12 months and satisfactory clinical improvement as early as 6 months indicate that satisfactory bony union and early functional recovery could be accomplished by the refined surgical procedure with minimally invasive biplanar osteotomy and rigid internal fixation.

The procedure and outcomes of DLO were initially reported in the 1960s by Benjamin [4]. He reviewed his experiences with DLO for knees with rheumatoid arthritis (RA) and osteoarthritis (OA) and reported favourable results. Subsequent to this paper, several studies examined the efficacy of DLO as a surgical option for knees with joint destruction and severe impairment [2, 16, 32, 38]. The majority of these studies reported rather unfavourable surgical outcomes with unpredictable symptomatic relief and a substantially high incidence of postoperative motion loss. Schüller et al. have stated that DLO should be abandoned in management of RA knees [32], while Iveson et al. reported a higher incidence of complication following DLO compared to single (isolated tibial) osteotomy [16]. However, it should be noted that surgical indication of DLO in these studies included both RA and OA knees and strict consideration for knee joint alignment and orientation was not given. Moreover, extensive surgical exposure was required, and no internal fixation was used to fix the osteotomy site. Therefore, the rather negative reports in those days cannot be extrapolated to DLO with modern surgical concepts and procedures.

In 2002, Babis et al. [3] presented a paper describing indication and results of double level osteotomy for varus osteoarthritic knees with the application of a modern surgical concept and procedure. They precisely analyzed the alignment for each knee using computer-aided radiological analysis, and indication for DLO was determined based on preoperative planning using a computer-based system (OASIS: Osteotomy Analysis and Simulation Software). In their series, DLO was performed for 29 knees with an average preoperative mTFA of -13.9° (varus), which corresponded to the amount of varus deformity in our patient population. When conversion to TKA or severe pain was deemed to be a surgical failure, the cumulative survival rate at an average follow-up period of 100 months was 95%. In 2012, Saragaglia et al. reported satisfactory results of DLO performed for osteoarthritic knees with severe varus [27]. They utilized the principles of computer-assisted surgery to achieve the intended amount of correction. CWDFO followed by OWHTO of the tibia and the use of intraoperative navigation in their procedure enabled accurate and consistent deformity correction. In addition, the use of a rigid plate fixation helped prevent postoperative alignment change caused by malunion. These two studies have shown that DLO can achieve satisfactory mid- to long-term outcome if the appropriate surgical indicate and procedure are followed; however, subsequent to these two papers, there have been no papers reporting radiological and functional outcomes of DLO. In recent years, the surgical technique for DLO has further evolved with the introduction of minimally invasive biplanar osteotomy and rigid internal fixation with a locking compression plate system. The intent of the present study was to report the surgical outcome using these current procedure modifications.

The theoretical advantage of DLO over single-level HTO is correction of severe varus knee deformity with restoration of physiological joint line inclination. Joint line obliquity following a large amount of correction by performing a tibial osteotomy alone may lead to several potential problems [13, 20, 25, 26]. Based on the follow-up results after single level HTO, Terauchi et al. reported that knees with varus inclination of the distal femur bear the risk for postoperative recurrence of varus deformity and recommended TKA or overcorrection for those knees rather than considering the increased surgical invasion associated with DLO [36].

In the surgical planning in our current practice, the intended radiological parameters were adjusted to the normal ranges with an mTFA of 0.5°–1° (a slight valgus), an mMPTA of 90°, and an mLDFA of 85°. To date, no consensus has been reached regarding the optimal alignment and orientation to aim for. Although overcorrection in coronal alignment with lateral deviation of the mechanical axis at the joint level is advocated in some papers [3, 36], our surgical procedure with only mild overcorrection (mean postoperative mTFA: 0.8°) and restoration of physiological joint orientation parameters (mLDFA and mMPTA) could attain satisfactory functional improvement. In addition, a satisfactory bony union without postoperative alignment change and early functional recovery as early as 6 months were noted in this study, which further support the efficacy of the current procedure with minimally invasive biplanar osteotomy with rigid internal fixation.

There are several limitations included in the study. First, the follow-up period (minimum of 1 year) was too short, and the number of included subjects was small. Second, the goal of alignment for DLO in this study was close to neutral; however, the optimal postoperative alignment has yet to be defined. Moreover, all radiological assessments were performed by one observer which may induce bias. Although inter- and intraobserver reliabilities for the radiological assessments were not examined in this study, high intra- and interrater reliabilities with the use of digital planning software has been confirmed in a previous study [31].

Conclusions

The minimally invasive DLO technique is a valuable surgical technique accomplishing restoration of physiologic knee joint alignment and orientation with significant improvement in patient-registered clinical outcomes in the early postoperative evaluation.