Comparison of primary arthrodesis versus open reduction with internal fixation for Lisfranc injuries: Systematic review and meta-analysisPF Han1, ZL Zhang1, CL Chen1, YC Han2, XC Wei1, PC Li1
1 Department of Orthopaedics, The Second Hospital of Shanxi Medical University, Shanxi Key Laboratory of Bone and Soft Tissue Injury Repair, Taiyuan, Shanxi, P.R. China
2 Department of Upper School, Subsidiary High School of Taiyuan Normal University, Taiyuan, Shanxi, P.R. China
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/jpgm.JPGM_414_18
Source of Support: None, Conflict of Interest: None
Keywords: Arthrodesis, Lisfranc injury, meta-analysis, open reduction with internal fixation
The Lisfranc joint, also known as the tarsometatarsal joint, connects the forefoot and midfoot and consists of three cuneiform bones, one cuboid bone, and five metatarsal bone. This joint was originally named by the surgeon Jacques Lisfranc (1790–1847) in 1815. Fractures of the Lisfranc joint are uncommon, accounting for only ~0.2% of all fractures in the body. Notably, the incidence of fractures of the Lisfranc joint is two to four times higher in males than in the females though, and the reason may be the more frequent participation in high-impact exercise among males. Fracture of the Lisfranc joint can occur at any age but is most frequently seen in young people in their 30s. Early diagnosis and treatment of the fracture are essential to restore and preserve function of the foot. Without proper treatment, the tarsometatarsal joint may cause arthritis or fracture nonunion may occur due to poor resetting and disorder, often leading to collapse of the second metatarsal arch structure. The latter condition causes severe pain, osteoarthritis of the middle foot, and progressive loss of function, all of which negatively affect the patient's quality of life. At present, the most common treatments for fractures involving the Lisfranc joint are primary arthrodesis and open reduction with internal fixation (ORIF).,, However, the optimal choice of surgical treatment remains controversial. Therefore, the present meta-analysis was performed to systematically compare the outcomes achieved with the two methods for the purpose of providing theoretical guidance for clinical practice.
Online databases, including Embase, PubMed, PubMed Central, CINAHL, PQDT, and Cochrane Library, were searched for studies comparing arthrodesis to ORIF for the treatment of Lisfranc injuries. A manual search of journals and references was performed, and we attempted to also find relevant gray literature, such as unpublished academic papers and chapters in monographs. We searched for all pertinent papers without any restriction on language, acquiring translation if necessary. The keywords used in the searches were as follows: arthrodesis, open reduction, internal fixation, and Lisfranc, according to the following strategy: “Lisfranc AND arthrodesis AND open reduction OR internal fixation.”
Inclusion and exclusion criteria
For inclusion in the meta-analysis, the studies had to meet the following inclusion criteria: (1) study population of adults with Lisfranc injuries; (2) case-controlled trial (CCT), prospective study, retrospective study, or cohort study design; (3) surgical treatment performed for all patients; (4) comparison of the results of primary arthrodesis and ORIF; and (5) outcome measurements included anatomic reduction rate, American Orthopedic Foot and Ankle Society (AOFAS) score, return to duty rate, complications, and visual analog scale (VAS) score for pain.
The exclusion criteria for studies were as follows: (1) an article type of letter, comment, editorial, review, proceedings, personal communication, or case report; (2) failure or injury type or treatment to meet the inclusion criteria; (3) a score on the modified Jadad scale ≤4 points; and (4) inaccurate measurements or incomplete data.
Two reviewers (PF H and ZL Z) independently analyzed the retrieved literature to determine whether the studies met the criteria for inclusion in the meta-analysis. Disagreement was resolved by discussion or submitted to a third reviewer when necessary. Basic information, including the name of the first author, the year of publication, the sample sizes for primary arthrodesis and ORIF for Lisfranc injuries, etc., was recorded. The risk-of-bias assessment tool outlined in the Cochrane Handbook was used to measure the methodological quality of CCTs. Six domains were evaluated: random trials, allocation concealment, double-blind, incomplete outcome data, and selective reporting risk. The modified Jadad scale was used to assess the quality of cohort studies, with a maximum score of 7 points and a score >5 points indicating a high-quality study.
The data of the included studies were entered into RevMan 5.3 software by two reviewers independently. Dichotomous outcomes are expressed as odds ratios (ORs); and standard mean differences (SMDs) are given for continuous variables with the corresponding 95% confidence intervals (95% CIs). Chi-square test was used to detect heterogeneity among the data. If no significant heterogeneity was detected (I2 < 50%), a fixed effects model was chosen, and if significant heterogeneity was detected (I2 > 50%), a random effects model was used. Some studies were individually removed for sensitivity analysis, and funnel plots were generated to assess bias. Differences were considered statistically significant when P ≤ 0.05.
Study inclusion and characteristics
The literature screening process is outlined in [Figure 1]. From the described searches, a total of 107 relevant, nonduplicated articles were retrieved. Upon reading the titles and abstracts, we excluded 49 noncontrolled studies and 41 articles that were not related to the purpose of this study. The full text of the remaining 17 articles was read, and the inclusion and exclusion criteria were applied. Finally, seven studies met all criteria for inclusion in this meta-analysis. The basic characteristics of these studies are summarized in [Table 1] and [Table 2]. The characteristics (e.g., age and gender) and conditions of the patients in the included studies were compared, and no differences were found (all P > 0.05).
Anatomic reduction rate
Anatomic reduction refers to the restoration of the normal anatomic positioning of the fractured segment through reduction, with perfect alignment (exact contact between the surfaces of the two fractured ends) and apposition (the relationship between the two fractured segments on the vertical axis). The anatomic reduction rate refers to the proportion of cases in which anatomic reduction was achieved among all patients. Four trials involving 127 patients compared the anatomic reduction rate between patients treated with primary arthrodesis or ORIF. A fixed effects model was employed in this meta-analysis, because the heterogeneity between the studies was not significant (I2 =0%). The meta-analysis showed that the anatomic reduction rates [95% CI (0.38, 4.17), P = 0.71] were similar between the treatment groups [Figure 2].
In evaluation of the AOFAS ankle-hindfoot score, pain, walking ability, gait, leg range, ankle stability, and joint line were evaluated separately. The maximum total score was 100, and a score greater than 75 was considered satisfactory. Three trials including 75 patients compared the AOFAS score between primary arthrodesis and ORIF groups. A fixed effects model was employed in the meta-analysis, because the heterogeneities between the studies were not significant (I2 =46%). The results showed that the AOFAS ankle-hindfoot score was significantly lower in patients treated with ORIF than in those treated with arthrodesis [95% CI (1.60, 8.81), P = 0.005; [Figure 3].
Return to duty rate
Return to duty was defined as a patient's return to their full work-related duty and passing his or her service-specific physical fitness test. Four trials including 216 patients compared the rates of postoperative return to duty between primary arthrodesis and ORIF groups. A fixed effects model was employed in the meta-analysis, because the heterogeneities between the studies were not significant (I2 =7%). The results showed that the return to duty rate in the ORIF group was significantly less than that in the arthrodesis group [95% CI (1.06, 4.99), P = 0.04; [Figure 4].
Seven studies including 1,189 patients compared the rates of complications between the arthrodesis and ORIF groups. Specifically, the need for removal of hardware, the need for revision surgery, and the total rate of complications were compared. A random effects model was employed in the meta-analysis, because the heterogeneities between the studies and subgroups were significant (I2 >50%). The results showed that the incidence of hardware removal was less in the primary arthrodesis group than in the ORIF group [95% CI (0.03, 0.14), P < 0.001], whereas the incidence of revision surgery [95% CI (0.15, 7.33), P = 0.96] and total complication rate [95% CI (0.30, 2.27), P = 0.71] did not differ significantly between the treatment groups [Figure 5].
It is worth mentioning that the I2 value for this outcome was >50 (this is the only outcome that shows significant heterogeneity). We performed a sensitivity analysis that removed each study individually and found that the I2 value remained >50 with the removal of each study. This means that no specific trial was the main source of heterogeneity. To better solve the problem of heterogeneity, we modified the meta-analysis model and performed subgroup analysis. After the above steps for homogenization, the results can be considered reliable.
The VAS is commonly used to assess the degree of postoperative pain in patients. The VAS is represented by a line 10 cm in length, on which one end is labeled 0 and indicates no pain and the other end is labeled 10 and indicates severe pain. Patients rank their pain according to the 11-point scale, and the sensitivity and reliability of the method are well-established. Three trials including 79 patients compared postoperative VAS scores reported 15–30 months after surgery between primary arthrodesis and ORIF groups. A fixed effects model was employed in the meta-analysis, because the heterogeneities between the studies were not significant (I2=0%). The results showed that the postoperative VAS score in the arthrodesis group was significantly less than that in the ORIF group [95% CI (−1.69, −0.15), P = 0.02; [Figure 6].
The seven studies included in this meta-analysis were evaluated with strict quality assessment. Two studies were randomized control trials (RCTs), and five studies were CCTs; thus, the possibility of a bias was low.
Direct sources of clinical and methodological heterogeneity among the studies include variations in the evaluation standards applied by the different trials. It is well known that the consistency of the meta-analysis calculation is based on the included data. In principle, with a greater degree of overlap of the confidence intervals between studies, the possibility of statistical homogeneity between studies is increased, and conversely, with smaller overlap, more heterogeneity exists between studies. The heterogeneity is usually evaluated by the Q or I2 test, which can detect the presence and degree of heterogeneity. In this study, the I2 test indicated some heterogeneity in the included studies (only one I2 value was >50).
Funnel plots are primarily used to observe whether the results of a systematic review or meta-analysis are biased, due to publication bias or any other bias. (1) If the included study has no bias, the points on the funnel plot are symmetrically scattered around the estimated true values of the independent study effect points, showing an inverted symmetric funnel shape; (2) if any bias exists, an asymmetrical funnel plot will be observed. With more obvious asymmetry, the degree of bias is greater, which may be associated with the incomplete collection of relevant literature, an insufficient sample size, and differences in the experience of clinical physicians. Our corresponding funnel plot showed that no bias existed [Figure 7]. Sensitivity analysis also showed good overall results [Figure 8] and [Figure 9].
Injuries of the Lisfranc joint that require surgical repair are relatively rare. However, with the increasing application of computed tomography, magnetic resonance imaging, and other imaging methods in recent years, the incidence of this injury has increased, suggesting the previously reported incidence may have been underestimated., Such injuries are easily missed or misdiagnosed due to the complexity of the Lisfranc joint anatomy, difficulty distinguishing these injuries from common forefoot sprains, and overshadowing by other serious combined injuries. Research has shown that ~40% of patients with Lisfranc injuries receive no treatment in the first week after injury, and these injuries are missed in 20% of patients with multiple fractures. Lack of treatment or improper treatment often leads to permanent disabilities such as chronic foot pain and deformity, corresponding to an extremely high disability rate in these patients. In recent years, considerable progress has been made in understanding the pathogenesis of Lisfranc injuries and diagnosing these injuries., However, controversy persists regarding the optimal treatment of such injuries, and thus, a relatively uniform treatment approach has yet to be established. At present, most cases are treated with either ORIF or joint arthrodesis., Previous studies demonstrated the effectiveness of partial arthrodesis as a salvage operation, allowing preservation of most functions of the affected foot., However, primary arthrodesis is recommended only for cases with severe damage, in which the joint cannot be anatomically reconstructed or the patient has severe pain due to traumatic arthritis. Some studies have also reported that joint arthrodesis surgery can be used as not only a salvage operation, with Lisfranc fusion able to reduce pain and increase joint stability with an even better long-term therapeutic effect than ORIF. The present meta-analysis was performed to better understand the relative efficacy and safety of primary arthrodesis and ORIF for treating fractures of the Lisfranc joint.
We selected five outcomes (i.e., the anatomic reduction rate, AOFAS score, return to duty rate, complications, and VAS scores) in an effort to fully compare the efficacy and safety between the two procedures. The results of the meta-analysis revealed that primary arthrodesis was associated with a lower incidence of hardware removal [95% CI (0.03, 0.14), P < 0.001], a lower postoperative VAS score [95% CI (-1.69, -0.15), P = 0.02], a higher return to duty rate [95% CI (1.06, 4.99), P = 0.04], and a better AOFAS score [95% CI (1.60, 8.81), P = 0.005] compared with ORIF. Based on the above factors, we conclude that primary arthrodesis is better than ORIF for the treatment of Lisfranc injuries. We also expect that with further advancements in relevant techniques and conditions across various fields, the deficiencies of arthrodesis for the treatment of Lisfranc injuries will be resolved, and this approach will show even greater superiority.
This systematic review has its limitations. First, heterogeneity among the data extracted from the different studies may also create bias in the meta-analysis results and increase the proportion of error. The included studies were published over 15 years (2002–2017), which could introduce bias. Furthermore, although I2 values <50% are considered to indicate low heterogeneity, ignoring this heterogeneity could have potentially increased bias in our results.
Second, although two studies were RCTs, in the other studies, patients were informed of the therapeutic options and chose the treatment, which reduces the methodological quality of the results and could impact the reliability of meta-analysis results. Also, because some studies contained insufficient information for pooled analyses, data from only three to seven articles were included in each comparison, and the heterogeneity between groups is increased with the use of fewer studies. Finally, Lisfranc injuries are complicated, and analyses of treatments for specific types of injuries (traumatic vs nontraumatic, bony vs ligamentous vs both) would be valuable.
Nevertheless, we identified and screened relevant articles carefully using multiple strategies, with two researchers applying strict exclusion and inclusion criteria independently. Although our findings need to be further verified by more RCTs with higher quality and larger sample sizes, our study provides the most detailed and up-to-date analysis on the surgical treatment of Lisfranc injuries.
In conclusion, the present meta-analysis provides evidence that primary arthrodesis offers a better therapeutic effect than ORIF for the treatment of Lisfranc injuries. Although it has been considered a salvage operation previously, our results showed that primary arthrodesis resulted in better functional recovery and less pain than ORIF. Additional well-designed RCTs with large sample sizes are needed to confirm these findings.
In meta-analysis, special attention should be paid to identifying and handling heterogeneity of data. The results of our analysis must be considered keeping in mind the unavoidable limitations in this study.
Financial support and sponsorship
This study was supported by National Project of International Science and technology Cooperation program of China (2015DFA33050), National Natural Science Foundation of China (81601949).
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9]
[Table 1], [Table 2]