Abstract
Background:
Meniscal tears are common in the young, active population. In this group of patients, repair is advised when possible. While inside-out repair remains the standard technique, recent advances in all-inside repair devices have led to a growth in their popularity. Previous reviews on the topic have focused on outdated implants of limited clinical relevance.
Purpose:
To determine the difference in failure rates, functional outcomes, and complications between inside-out and modern all-inside repairs.
Study Design:
Systematic review.
Methods:
A systematic review was registered with PROSPERO and performed following Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines using the MEDLINE, EMBASE, and Cochrane databases. Inclusion criteria were (1) clinical study reporting on all-inside or inside-out repair, (2) evidence levels 1 to 4, and (3) use of modern all-inside implants for all-inside repairs. Exclusion criteria were (1) use of meniscal arrows or screws and (2) concomitant surgical procedures. Study characteristics, subjects, surgical technique, clinical outcomes, and complications were collected and analyzed.
Results:
A total of 481 studies were screened and assessed for eligibility, which identified 27 studies for review. Studies defined clinical failure as persistent mechanical symptoms, effusion, or joint line tenderness, while anatomic failure was incomplete or no healing on MRI or second-look arthroscopy. There were no significant differences in clinical or anatomic failure rates between inside-out and all-inside repairs (clinical failure: 11% vs 10%, respectively, P = .58; anatomic failure: 13% vs 16%, respectively, P = .63). Mean ± SD Lysholm and Tegner scores for inside-out repair were 88.0 ± 3.5 and 5.3 ± 1.2, while the respective scores for all-inside repair were 90.4 ± 3.7 and 6.3 ± 1.3. Complications occurred at a rate of 5.1% for inside-out repairs and 4.6% for all-inside repairs.
Conclusion:
The quality of the evidence comparing inside-out and all-inside meniscal repair remains low, with a majority of the literature being evidence level 4 studies. In this review comparing modern all-inside devices with inside-out repair, no differences were seen in failure rates, functional outcome scores, or complication rates.
Keywords
Meniscal injuries are one of the most common injuries evaluated by orthopaedic surgeons. 37 Meniscal disease has an estimated annual societal cost between $500 million and $5 billion in the United States. 36 Although many meniscal injuries result in a partial meniscectomy (approximately 61 meniscectomies per 100,000 patients annually), meniscal repairs are employed in a select set of tear patterns to attempt to restore the natural function of the meniscus and possibly prevent early degeneration of the affected compartment. In 1948, Fairbank 20 initially described the radiographic changes that occur in patients who have undergone a total meniscectomy, showing the importance of the meniscus in preventing progression of osteoarthritis. Similarly, long-term follow-up studies have shown that an arthroscopic partial meniscectomy will lead to an increased rate of Fairbank changes compared with the anatomically normal contralateral knee. 21 The load seen across the knee joint is proportional to the amount of meniscus that is removed; as such, the goal of a meniscal repair is to maintain a functional meniscus and restore its joint-protective function and possibly prevent osteoarthritis.8,34
Meniscal repairs can be performed via a variety of surgical techniques. Outside-in repair is an option for anterior and middle-segment meniscal tears but is used less commonly than inside-out and all-inside techniques. 16 The gold-standard technique for meniscal repair has been the inside-out technique, in which the surgeon makes an incision either posteromedially or posterolaterally to gain access to the posterior capsule and allow for sutures to be passed through the menisci with long needles in a vertical or horizontal mattress configuration under arthroscopic visualization. 31 Problems specific to this approach include the need for an accessory incision and potential injury to the saphenous vein medially and the peroneal nerve or popliteal neurovascular bundle laterally. 17 This technique can also be technically demanding and requires more assistance in the operating room, which can lead to increased operative time. The all-inside technique obviates the need for an accessory incision but adds increased cost and the risk to the neurovascular bundle in posterior lateral tears. In 1991, Morgan 38 published a new “all-inside” technique using a posterior portal, a 70° arthroscope, and suture hook device designed to repair posterior horn meniscal tears. Albrecht-Olsen et al 4 were the next to introduce an absorbable polylactic acid tack implant for all-inside meniscal repair. The new implant allowed for more rapid meniscal repair using the standard anterior arthroscopy portals and a 30° arthroscope. 4 Additional implants based on the polylactic acid tack were fashioned as meniscal arrows, screws, and staples. Because of reports of implant-induced chondral injury and synovitis as well as concern regarding the strength of the implant compared with traditional sutures, the trend in recent years has been the development of suture-based devices. 33
Previously, Grant et al 23 performed a systematic review to compare inside-out and all-inside techniques for repair of isolated meniscal tears. The studies included by these authors primarily used the meniscal arrow, screw, or staple. Given that these implants have fallen out of clinical use, the results of the review are now of limited utility in guiding the choice between all-inside and inside-out fixation. In addition, since the publication of their review, the number of therapeutic studies on meniscal repairs has nearly doubled, justifying the need for an updated systematic review on the topic. 23
The purpose of this study was to provide an updated comparison of failure rates, functional outcome scores, and complication rates between inside-out and modern all-inside meniscal repairs for the treatment of isolated meniscal tears. Our hypothesis was that no differences would exist between these 2 surgical techniques in any of these outcome measures.
Methods
A systematic review was conducted according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines using a PRISMA checklist. 35 The review was registered with the PROSPERO International prospective register of systematic reviews (CRD42015019568). 43
Two reviewers (Y.A.F. and B.J.E.) independently conducted the search in November 2015 using the following databases: MEDLINE, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, EMBASE, and Scopus. The search terms used included the following: all-inside meniscal repair, all-inside meniscus repair, inside-out meniscal repair, inside-out meniscus repair, bucket handle tear, bucket handle meniscal tear, bucket handle meniscus tear, bucket handle meniscal repair, and bucket handle meniscus repair. Since the use of a hyphen (ie, inside-out vs inside-out) in the search terms produced a different number of results, all terms were searched with and without the hyphen. Articles were initially screened to only include (1) English-language studies and (2) publication date of August 2010 to November 2015. The date of August 2010 allowed for all literature published since the search of Grant et al 23 to be included in the current results. Article titles, abstracts, and full texts were further screened to limit the studies to include (1) therapeutic studies, (2) evidence levels 1 to 4 (2011 update by the Oxford Centre for Evidence-Based Medicine 40 ) clinical studies, (3) all-inside suture-based procedure or inside-out procedure for meniscal repair, (4) vertical unstable meniscal tears, and (5) patients aged 18 years and older. Medical conference abstracts were excluded. Additional studies included in the systematic review were those previously identified by Grant et al 23 using the modern all-inside suture-based device or inside-out procedure for meniscal repair. We performed an independent search of the literature before August 2010 to validate the search results of the prior systematic review. To ensure that all studies were located, references within included studies and all review or technical studies were cross-referenced for inclusion if missed by the initial search. If duplicate subject populations were discovered, the population with the longer follow-up or greater number of subjects was included. Level 5 evidence, reviews, letters to the editor, basic science, biomechanical studies, outside-in repair techniques, imaging, surgical technique, and classification studies were excluded. The search process is shown in the flow diagram (Figure 1).
Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) flowchart of study.
Two authors (Y.A.F. and B.J.E.) extracted the data from the articles. Study and subject demographic parameters analyzed included year of publication, financial conflict of interest, years of subject enrollment, level of evidence, number of subjects, sex, age, number of meniscal tears (stable and unstable knees), size of tear, location of tear, chronicity of tear, duration of follow-up, surgical technique, and type of device. Clinical outcomes were reported as rate of clinical and anatomic failure, type of complications, and knee-related outcome scores, including International Knee Documentation Committee (IKDC) subjective and objective, Lysholm, Tegner, Hospital for Special Surgery (HSS), Japan Orthopaedic Association (JOA), and Marx activity knee scores. Clinical failure was defined as persistent mechanical symptoms, effusion, or joint line tenderness while anatomic failure was defined as incomplete or no healing on magnetic resonance imaging (MRI) arthrogram or second-look arthroscopy. Study methodological quality was evaluated using the Modified Coleman Methodology Score, and level of evidence (1-4) was assigned to each study in accordance with the system provided in the 2011 update by the Oxford Centre for Evidence-Based Medicine.18,40
Data Synthesis
The pooled data for clinical failures were used to calculate overall and treatment-specific failure rates. Failure rates were compared using contingency analysis and the Fisher exact test. Frequency weighted means were calculated for functional outcome scores. Descriptive statistics were provided for the reported complications.
Results
A total of 481 articles were identified on the initial literature search. Of these, 331 were excluded based on date of publication and language. The remaining 150 article titles, abstracts, or full texts were reviewed for eligibility based on the inclusion and exclusion criteria. Altogether, 17 articles from the current literature search and 10 articles from the prior systematic review were combined for a total inclusion of 27 articles. ‡ Of the 27 studies published between 1988 and 2015, there were 2 randomized or prospective studies, while remaining studies included level 3 (4 studies) and level 4 (21 studies) evidence (Table 1). Thirteen studies reported outcomes of inside-out repairs with 676 tears, and 14 studies reported outcomes of all-inside repairs with 548 tears. The mean duration of follow-up was 57 months; all-inside and inside-out repair was 39 and 70 months, respectively. All reported methods of all-inside were suture-based systems with the Fast-Fix (Smith & Nephew) device used in 12 studies, § the Meniscal Viper Repair System (Arthrex) used in 1 study, 24 and a suture hook device (Linvatec) used in 1 study. 2 Because of the popularity of the Fast-Fix device, subset analysis was performed comparing inside-out and all-inside Fast-Fix repairs. Results of the subset analysis were not different from the primary analysis. Measurements of the average tear size were not reported, but 3 inside-out3,4,19 and 7 all-inside10,11,15,24,26,42,44 studies reported the average number of knots or devices per tear. On average, inside-out tears used 3 knots per meniscal tear, and all-inside tears used 2 devices per meniscal tear. Although it is not a validated measure, using average number of knots or devices per meniscal tear as a proxy of tear size possibly suggests a larger tear size for those treated with inside-out repairs.
Patient Demographics and Study Characteristics a
HSS, Hospital for Special Surgery; IKDC, International Knee Documentation Committee; JOA, Japan Orthopaedic Association; —, not available.
A meaningful total or average could not be calculated because of a lack of data from other studies.
Clinical and Anatomic Failures
Clinical outcomes were reported for all included studies, with clinical failure commonly defined as persistent mechanical symptoms, effusion, or joint line tenderness. The overall clinical failure rate was 11% (range, 0%-37% in individual studies) (Table 2). ‖ The clinical failure rates of inside-out and all-inside techniques were statistically equivalent (11% vs 10%, respectively; P = .58).
Clinical and Anatomic Failures a
Data in italics represent the group of patients included in the total for the denominator in the calculation of the failure rate because not all studies provided the number of failures for isolated tears only. —, not available.
Mean frequency weighted failure rate.
Anatomic failure was reported in only 9 studies (Table 2). ¶ Patients were evaluated for anatomic failure by MRI arthrogram or second-look arthroscopy whereby incomplete or lack of healing of the repaired meniscus was considered a failure. Among the 8 studies, the overall anatomic failure rate was 14%, with the individual technique failure rates for inside-out and all-inside being 13% and 16%, respectively (P = .63).
Outcome Scores
Several outcome scores were reported among the studies, which included the following: Lysholm (15 studies # ), Tegner (13 studies**), IKDC (5 studies11,22,26,44,49), HSS knee (2 studies2,47), JOA knee (1 study 24 ), and Marx activity (1 study 11 ) scores. Lysholm and Tegner scores were the only scores reported in more than one study for both inside-out and all-inside techniques. As a result, frequency-weighted means were calculated only for Lysholm and Tegner scores. The overall mean ± SD Lysholm score after meniscal repair was 89.0 ± 3.8. The mean Lysholm scores for inside-out and all-inside repairs were 88.0 ± 3.5 and 90.4 ± 3.7, respectively (Table 3). The overall mean Tegner score after meniscal repair was 5.7 ± 1.3. Mean Tegner scores for inside-out and all-inside repair techniques were 5.3 ± 1.2 and 6.3 ± 1.3, respectively (Table 4).
Lysholm Scores a
—, not available.
Frequency-weighted score, mean ± SD.
Tegner Scores a
—, not available.
Frequency-weighted score, mean ± SD.
Complications
Complication rates among the included studies were not consistently reported, with only 11 of 27 studies citing the number and type of complications (Table 5). †† The studies reporting complications were split, with 7 inside-out and 4 all-inside studies. Since not reporting complications does not necessarily mean they did not occur, only studies explicitly citing the absence of complications were assigned a 0% complication rate. The study using the Arthrex Viper meniscal repair system cited occasional slight chondral injury caused by the rigid shaft of the device when used in a narrow joint gap, but this study failed to provide an incidence for the complication. 24 Overall, the number of complications was low, with a total of 22 complications for an overall incidence of 5.0%. The complication rate for inside-out repair was 5.1%, with the listed complications being superficial infection, deep infection, peroneal and saphenous nerve palsies, thrombophlebitis, and deep venous thrombosis.4,14,19,25,30,45,47 All-inside repairs had a complication rate of 4.6%, with the listed complications being deep infection, deep venous thrombosis, pulmonary embolism, chondral damage, and intraoperative failure of the device.2,15,32,33 Given the low percentage of studies reporting complications, no statistical comparison was performed between the 2 repair techniques.
Complications a
DVT, deep venous thrombosis; PE, pulmonary embolism; —, not available.
Mean frequency-weighted complication rate.
Discussion
The principal findings of this systematic review are that inside-out and modern all-inside meniscal repairs achieve similar functional outcomes, with equivalent anatomic and clinical failure rates and similar complication profiles, thus confirming our initial hypothesis and the findings of previous reviews on the topic. Given the current trend toward increased rates of meniscal repair over debridement, understanding the current literature on repair techniques will allow for improved evidence-based decisions. 1
Our results corroborate previous biomechanical studies showing equivalent fixation properties of modern all-inside implants and traditional mattress stitches.9,12 The biomechanical strength of the Fast-Fix device has been shown to be comparable with vertical mattress sutures while maintaining the same load to failure when placed in a horizontal fashion.9,12 In addition, Borden et al 12 demonstrated biomechanical comparison of the Fast-Fix device as superior to the previously used meniscal arrow. Similarly, the results of the biomechanical analysis correlated with the lower clinical failure rates when the Fast-Fix device was compared with the meniscal arrow in a randomized clinical trial. 32
Review of the literature demonstrates 3 systematic reviews in the past 5 years have been published comparing inside-out and all-inside meniscal repairs.7,23,39 The systematic reviews by Nepple et al 39 and Ayeni et al 7 only included all-inside studies with the meniscal arrow. Grant et al 23 was the only systematic review to compare inside-out and all-inside repairs, but the all-inside repairs included the outdated and modern all-inside meniscal repair devices. The results of the current review provide a similar conclusion as Grant et al, 23 with no significant difference in the clinical results between inside-out and all-inside meniscal repair. Interestingly, clinical failure rates were lower in our study compared with the previous review, with inside-out and all-inside failure rates of 11% and 10% compared with the previously reported rates of 17% and 19%. 23 Whether this represents a true improvement in outcomes over time is unclear, but this could be explained in light of improved implant design and progress along our collective learning curve. In contrast, Lysholm scores remain similar between Grant et al 23 and the current review. Grant et al 23 reported Lysholm scores of 87.8 for inside-out and 90.2 for all-inside, while our analysis demonstrated scores of 88.0 for inside-out and 90.4 for all-inside. Finally, Tegner scores were also higher for all-inside repair in our review. Grant et al 23 reported the Tegner scores for inside-out and all-inside techniques to be 5.6 and 5.5, in contrast with 5.3 for inside-out and 6.3 for all-inside repairs in the present study. Unfortunately, due to the retrospective nature of the included studies, the baseline characteristics of the included patients are unknown, and we therefore cannot determine whether these differences are real or related to differences in patient populations.
This study had several limitations. The primary limitation to internal validity was the low quality of the studies available for inclusion (evidence levels 3-4). While the patient and meniscal tear characteristics were similar between groups, it is possible that there was inherent selection bias in determining which patients were candidates for inside-out or all-inside repair. Indeed, many surgeons believe that all-inside is warranted for small tears requiring only 1 or 2 implants, while they prefer inside-out repair for larger tears. We support the notion of favoritism toward inside-out repairs for larger tears since inside-out repairs were done with an average of 3 knots as opposed to 2 devices for all-inside repairs. Another limitation is the risk of instrument bias, since the only functional outcomes scores available were the Tegner and Lysholm. The Lysholm score is well known to have a ceiling effect, and given that the scores for both groups were in the 90s, it is possible that this instrument lacked the sensitivity to detect true clinical differences between groups. 13
With respect to external validity, the primary limitation is that a narrow type of meniscal tear was included in most studies. These were exclusively vertical tears in the well-vascularized portion of the meniscus near the capsular attachment. The results of this study cannot be extrapolated to horizontal cleavage tear, radial tears, root tears, or tears with more questionable vascularity. In addition, studies included only patients 18 years or older, so the applicability of our findings in the pediatric population (where meniscal repair is strongly indicated) is unknown.
Conclusion
This systematic review highlights the shortcomings of the available evidence comparing all-inside and inside-out meniscal repair, with overall low levels of evidence and a limited number of outcomes instruments available for comparison. Within the relatively narrow spectrum of medium-sized vertical meniscal tears, modern all-inside and inside-out repairs have equivalent failure rates, functional outcomes, and complication rates. Future prospective studies are warranted and should include more robust outcome instruments. In addition, information about the direct and indirect costs of all-inside and inside-out repairs needs to be studied for a formal cost analysis. If inside-out and all-inside repair techniques can be shown to have equivalent outcomes in high-level studies, the additional cost at our institution of $213 per all-inside device compared with the inside-out knots will be called into question and amount to large cost savings for our health care system.
Footnotes
One or more of the authors has declared the following potential conflict of interest or source of funding: B.R.B. receives research support from Arthrex, CONMED Linvatec, DJ Orthopaedics, Ossur, Smith & Nephew, and Tornier. A.B.Y. receives research funding from Arthrex and NuTech.
An online CME course associated with this article is available for 1 AMA PRA Category 1 Credit™ at 
. In accordance with the standards of the Accreditation Council for Continuing Medical Education (ACCME), it is the policy of The American Orthopaedic Society for Sports Medicine that authors, editors, and planners disclose to the learners all financial relationships during the past 12 months with any commercial interest (A ‘commercial interest’ is any entity producing, marketing, re-selling, or distributing health care goods or services consumed by, or used on, patients). Any and all disclosures are provided in the online journal CME area which is provided to all participants before they actually take the CME activity. In accordance with AOSSM policy, authors, editors, and planners’ participation in this educational activity will be predicated upon timely submission and review of AOSSM disclosure. Noncompliance will result in an author/editor or planner to be stricken from participating in this CME activity.