Introduction

Percutaneous forefoot surgery (PFS) has been developed by Stephen Isham, Mariano de Prado and Pau Golano. They have produced valuable knowledge and built an anatomic and technical foundation for PFS [1, 2]. It is theorized that 50% of forefoot cases are surgically repaired in the USA by means of PFS and minimum incision surgery (MIS) [3]. However, PFS has not been given a great deal of attention in the surgical literature [4].

PFS is performed through 1–3 mm incisions, using a mini-blade for soft tissue and power rotary bur for osseous procedures under image intensification. Theoretical advantages are a potential faster recovery with immediate weight bearing, reduced surgical time, a less painful postoperative period and less stress to the patient. Besides, PFS is performed as an outpatient procedure. The main disadvantages are the requirement for specific equipment and lengthy learning curve.

The purpose of the present study is to evaluate the clinical and radiographic results of PFS for hallux valgus (HV) deformity.

Materials and methods

This is a single forefoot non-specialized center, with a single-surgeon prospective study. The surgeon had a previous surgical experience in PFS technique of 20 cases. All the patients were treated in an outpatient surgery unit.

Between April 2006 and December 2013, 126 patients were included in the study, 108 of them completed the minimum two-year follow-up period (57.3 months, range 24–112), 24 patients with a one-step bilateral surgery. Therefore, a total of 132 feet were studied, with previous informed consent.

The indication was a painful HV with or without metatarsalgia, with less than 60° of HV angle (HVA) and less than 20° of 1–2 intermetatarsal angle (IMA). All patients had adequate ROM of the first metatarsophalangeal (MTP) joint. All patients reported having pain and difficulty wearing shoes, with no improvement from conservative treatments. Patients were excluded if they had rheumatoid arthritis or other inflammatory diseases, previously failed HV surgery or isolated interphalangeal HV.

A clinical and radiological weight-bearing examination, by an independent review, was performed preoperatively and postoperatively at two months, one year and at final follow-up (December 2015). It included visual analog scale (VAS), American Orthopaedic Foot and Ankle Society hallux score (AOFAS) [5, 6], HVA, IMA, distal metatarsal articular angle (DMAA), the metatarsal index (M1 > M2, M1 = M2, M1 < M2) and joint congruency of MTF joint. Besides, VAS was collected one and three weeks postoperative, to assess pain level at immediate postoperative period. The subjective satisfaction with the outcome (very satisfied, satisfied, dissatisfied, or disappointed) was also recorded. Anteroposterior weight-bearing radiographic angles were measured by the Couglin et al. method [7]. Joint congruency of the first MTP was assessed using the criteria defined by Pigott [8]. All the measurements were taken digitally (Ykonos, Sescam©).

Surgical technique

All of the procedures were performed under spinal anesthesia with bupivacaine (0.25%) ankle block for postoperative pain control. At home, patients were recommended to use acetaminophen and/or dipyrone and/or nonsteroidal anti-inflammatory agents according to demand.

The PFS procedures were based on the description of Isham and De Prado [1, 2]. Exostosectomy, lateral metatarsophalangeal arthrolysis and percutaneous Akin osteotomy of the first phalanx were done in all cases [9]. Patients with moderate to severe lateral metatarsalgia underwent distal metatarsal mini-invasive osteotomy (DMMO), without fixation, of 2nd, 3rd and 4th metatarsals. Patients with IMA ≥ 15° underwent proximal closing wedge osteotomy of the first metatarsal with minimally invasive surgery (MIS), and fixed with a superolateral compression staple. Treatment of lesser toes was determined on a case-by-case basis, with a combination of procedures on the soft tissues (flexor and extensor tendon tenotomies) and bones (phalangeal osteotomies) (Fig. 1).

Fig. 1
figure 1

Left: preoperative radiograph, HVA 41°, IMA 17°, DMAA 25°. Right: 6° year postoperative, HVA 5°, IMA 8°, DMAA 27°

Full size image

In every case, a specific dressing is fashioned at the end of the procedure to maintain the correction. Monitoring of this dressing was done at first and third weeks and removed at third week.

Immediate full weight bearing was allowed with a rigid, flat-soled postoperative shoe for 5 weeks, and no deep venous thrombosis prophylaxis was used. Only cases with compression staple osteosynthesis received one antibiotic prophylaxis dose with cefazolin.

Statistical analysis

Continuous data were described as means and standard deviations. Categorical data were described as absolute and relative frequencies. The Wilcoxon signed-rank test was used to compare values before and after surgery and during follow-up. A 0.05 level of significance was used throughout. Statistical analyses were performed with the aid of PASW 18.0 statistical analysis software (SPSS Inc, Chicago, IL, USA).

Results

The demographics of age, sex and operated side are shown in Table 1. The associated pathologies to HV and surgical procedures performed are listed in Table 2. It is essential to highlight a greater number of cases with metatarsalgia treated with DMMO in the group. Also in the 27.3% of patients have been performed a proximal closing wedge osteotomy of the first metatarsal with MIS. Both facts affect the clinical and radiological results.

Table 1 Demographics of PFS patients
Full size table
Table 2 Associated pathologies to hallux valgus and surgical procedures
Full size table

Mean average preoperative AOFAS hallux score was 50.6 ± 11.2 points, and at follow-up, it improved to 85.9 ± 13.8 points (Table 3). The vast majority of patients had a preoperative pain level moderate or severe (96.2%) (VAS 4–9). One week after the procedure, patients with none or mild pain (VAS 0–3) were 59%. At the third week, that percentage increased to 74.6%. Final follow-up, 76.5% had none or mild pain. Mean preoperative pain score (6.3 ± 1.5 points) decreased progressively at first (3.2 ± 2 points) and third week (2.4 ± 1.8 points) (p < 0.001), but without changes between third week and second month (2.4 ± 2.2 points) (p = 0.792). A further improvement in pain occurred between the second month and the first year (1.8 ± 2.5 points) (p = 0.007). No changes in the level of pain until the final follow-up (1.9 ± 2.4 points) (p = 0.588).

Table 3 AOFAS hallux scores
Full size table

The mean average HVA improved from 34.3° ± 9.3° before the operation to 16° ± 8.2° at second month, a correction of 53.3%. During the first year, there was a loss of postoperative correction of 6.5°, implying a loss of 18.9%, but stabilizing angulation after the first year. Preoperative IMA was 13.1° ± 3.5°, decreasing 2.4° (18.3%), at final follow-up. The DMAA suffered a slight, but progressive worsening with a mean of 18.5° ± 7.3° preoperative and 23.6° ± 9.7° at final follow-up (Table 4).

Table 4 Radiologic evaluation
Full size table

The number of cases with preoperative metatarsal index M1 < M2 was 47.7% and M1 > M2 was 23.5%. At final follow-up, they were 43.5% and 29%, respectively. Preoperative joint congruency of the first MTP was 47% cases, improving at second month to 95% (p < 0.001), but decreasing at fist year (80.3%, p < 0.001). Without changes between fist-year and final follow-ups (73.3%, p = 0.150).

At final follow-up, 76.5% of the subjects were satisfied or very satisfied. Only 6.8% of cases were disappointed. Recurrence of medial 1st MT head pain happened in 22 cases (16.7%). Complications registered are shown in Table 5.

Table 5 Complications
Full size table

Discussion

The VAS is validated for various orthopedic outcomes and has shown to be reliable [10–12]. In our study, PFS has shown a low immediate postoperative pain level, with mean VAS score of 3.2 ± 2 points at first week. Pain level continued improving until third week (2.4 ± 1.8 points), lower than Nikolau et al. (3.84 ±  1.3 points) [13] with Mitchell’s osteotomy, and at first year (1.8 ± 2.5 points) (p = 0.007), without worsening at final follow-up. Our final VAS score was lower than that reported by Deveci et al. [14] with Scarf osteotomy (2.4 ± 1). This shows us a rapid recovery of pain. The final follow-up AOFAS hallux score in our study was 85.9 ± 1.8 points, previous reports with different procedures presented similar outcomes [14–18], although in our study we have treated also patients with metatarsalgia, increasing pain level and delaying recovery.

Mean of final HVA correction was 12° in our study. A similar correction was shown in several reports with Scarf [19–21], Chevron [17, 22], Kramer [18] and Reverdin–Isham osteotomy [23, 24] (12°–16°), with similar follow-up. Our correction was lower compared to those reported by Lee et al. [25], Deveci et al. [14], Jones et al. [26] and Bonnel et al. [27] (22°–25°, with Scarf and Chevron osteotomies), but with lower mean follow-up (7–26 months). Lee et al. [16], with Chevron proximal and distal osteotomy, reported 24° of HVA correction with similar follow-up. Mean IMA correction in several reports was between 6° and 9° (Scarf, proximal and distal Chevron and Kramer osteotomies) [14, 16, 18, 21, 25, 26], which is better than 3° of our study. But, Bauer et al. [23, 24], with Reverdin–Isham osteotomy, obtained a correction equal to ours. The proximal closing wedge osteotomy of the first metatarsal performance (27.3% of patients) has affected the clinical and radiological results. But we have considered it a necessary associated procedure to decrease the pathologic IMA. We have to take into account that intraobserver and interobserver reliability rates are high for measuring the HVA and IMA (<5°, 95% confidence interval) [28].

76.5% of patients were “satisfied” or “very satisfied,” and this percentage is significantly lower than those reported with Scarf, proximal or distal Chevron and Reverdin–Isham osteotomies (85–95%) [16, 19, 20, 23, 29]. Recurrence of medial 1st MT head pain happend in 16.7% of patients, of which, only 5.3% request revision surgery. A total of 10.6% of patients needed revision surgery. Rates of HV recurrence vary in the literature from 2.7 to 20% [30–34], and therefore, recurrence is common in HV surgery, with a multifactorial cause including patient-related and surgical factors [35]. In PFS, recurrence of medial 1st MT head pain may be in relation to insufficient resection with peripheral bony prominence. Moreover, an inadequate lateral MTP release can cause lack of HVA correction. Both mistakes are more frequent at the beginning of the learning curve.

There was no case of HV deformity or deep venous thrombosis. The most numerous complications were metatarsalgia after DMMO (11.4%) (without relationship with HV deformity), with a lower rate than that reported by García-Fernández et al. (40%) [36] and similar to Henry et al. (14%) [37] reports. Although Bauer [38] considers DMMO an easily reproducible procedure with a short learning curve, we consider it a technically demanding procedure with a prolonged forefoot edema and high complications rate (22.9%, in our study). Therefore, we do not recommend its systematic realization.

PFS could be an option for surgical HV deformity treatment, but without improving the patient satisfaction rate results published with conventional procedures. The main advantage is a low postoperative pain level, but with an insufficient HVA correction. Despite the worse radiographic results in our study, compared with previous reports using different procedures, we present similar clinical outcomes. Which is in line with what was published by Thordarson et al. [39].

The major limitation of the present study is the difficulty to standardize treatment in HV surgery, with several additional procedures, that have some influence on outcomes (metatarsalgia, proximal closing wedge osteotomy of the first metatarsal and lesser toes deformities). Also we have not studied how this might impact outcomes. Nevertheless, this study presents our results with this surgical strategy. The surgeon who participated in the study only had twenty cases of surgical experience previous to the study. As a consequence, the results may apply to surgeons with little experience. The study’s strengths are the prospective study, the cohort number and the mean final follow-up.

Conclusion

PFS, in our study, does not improve the radiological and patient satisfaction rate results compared with conventional procedures. The main advantage is a low postoperative pain level, but with an insufficient HVA correction. The technique requires a learning curve before being able to produce reliably acceptable results.