Patients
From July 2013 and October 2015, the medical records of 53 patients who had undergone arthroscopic surgery for symptomatic DLM by an experienced surgeon (J.W) were retrospectively reviewed. Of those, 32 DLM patients in whom we performed an arthroscopic repair were identified. Inclusion criteria were as follows: no former ipsilateral meniscus surgery, no commitment surgery such as anterior cruciate ligament reconstruction, and a meniscal tear size > 10 mm. The exclusion criteria included patients did not underwent arthroscopic repair of torn of DLM and age older than 60 years. Of those patients, 16 were augmented with platelet-rich plasma (PRP), and 16 were performed without PRP augmentation. All 32 patients had radiographic evidence of meniscal pathology seen on magnetic resonance imaging (MRI) (Fig. 1). Despite our efforts, 3 of the selected patients were lost to follow-up due to migration (2 patients in the PRP group and 1 in the non-PRP group). We evaluated the remaining 29 patients, 14 in the PRP group and 15 in the non-PRP group. No second-look imaging or second-look arthroscopy was performed in this study.
PRP preparation
To obtain the PRP, 37 ml of the patient’s blood was collected into a 50-ml injector containing 4 ml 3.8% sodium citrate as anticoagulant. Then, 2 centrifugations were performed: the first at 2000 rpm for 10 min to separate erythrocytes, and the second also at 2000 rpm for 10 min to concentrate platelets, which provided 4 ml of PRP. The preparation method used allowed the number of platelets per milliliter to increase by a mean of 6.4 ± 1.6 times (range, 3.5–8.7) with respect to baseline blood values. Injected PRP in this study also contained leukocytes (leukocyte-rich PRP) 6.1 ± 1.5 times (range, 3.1–8.4) times with respect to the normal blood value.
Operative procedure
The meniscus repair was performed using the inside-out technique. Repair was only performed in the red-red zone or red-white zone of the posterior horn of the medial or lateral meniscus, which is reported to have good healing potential [21]. An arthroscopic examination was performed via anteromedial and anterolateral portals. A hook probe (Smith & Nephew, Andover, MA) was used to confirm a lateral meniscal tear (Fig. 2). Once the tear pattern was confirmed, an arthroscopic punch (Smith & Nephew, Andover, MA) and shaver (Smith & Nephew, Andover, MA) were used to perform saucerization of the discoid meniscus. Meniscal instability was again confirmed, and preparation for repair was undertaken (Fig. 3). The torn margin of the meniscus and adjacent synovium were abraded with a rasp and shaver to improve the vascular supply to the lesion. A medium Graves speculum blade (MedGyn, Addison, IL) was then positioned as a meniscal retractor to aid in retrieving sutures and protecting the neurovascular structures behind the knee by an incision parallel and just posterior to the lateral collateral ligament. Once the retractor was in place, two double-armed needles with 2–0 polyester braided suture (Ethicon, Somerville, NJ) were placed vertically every 4 to 6 mm and were tied to appose the meniscus body to the remaining meniscus rim and attachment using a zone-specific cannula (Linvatec, Largo, FL) (Fig. 4).
After the repair, excess saline was suctioned out. Then, PRP (4 ml) and 500 IU thrombin (1 ml) were injected simultaneously on the repaired site using two injectors (one 5-ml for PRP and the other 1-ml for thrombin) and a cannula needle 2.5 mm in diameter under arthroscopic vision (Fig. 5). The meniscal sutures previously placed were loosened so that the PRP can have the best contact area with the lesion. After the PRP gel clot was formed on the lesion (Fig. 6), the knee was taken to 90° of flexion and the sutures were fastened down and then tied. Finally, the arthroscope was pulled out, and the portals were then sutured. No drainage was used after the surgery.
Rehabilitation protocol
In this study, all patients followed the same rehabilitation protocol. For the first 2 weeks, patients were allowed to bear partial body weight up to 10 kg, with range of motion from 0° to 60°. In the third and fourth weeks, partial weight-bearing of 50% of body weight was permitted, with range of motion from 0° to 90°, After the fourth week, unassisted walking with full weight-bearing was permitted and full joint movement exercises were started. No squatting or deep flexion activities greater than 120° were permitted for 3 months, and running, jumping, and cutting were restricted for 6 months.
Outcome assessment
Patients were evaluated at baseline (the day before surgery) and then 12 and 24 months after the last injection; evaluation included the Lysholm score [22], and Ikeuchi grade [2], Visual analogue score (VAS) for pain [23] and failure rate. Failure was defined by patients developing symptoms of joint line pain and/or locking or swelling or requiring repeat arthroscopy. Outcome data were collected in person or by telephone by an orthopaedic surgeon who was blinded to the treatment received by the patient.
Statistical analyses
All data management and statistical analysis were performed with SPSS version 22.0 software (SPSS Inc., Chicago, IL, USA). Student’s t-tests were used for continuous variables (age, symptom duration, duration of follow-up, Lysholm score and VAS for pain) and Fisher exact test was used to analyze the categorical variables (gender, type of DLM, repaired meniscal zone, Ikeuchi grade and failure rate). In our study, the failure rate was demonstrated by intent-to-treat (ITT) and per-protocol (PP) analyses. In the ITT approach, all patients were included in the analysis in the group to which they were allocated regardless of loss to follow-up. In the PP approach, patients who completed the entire procedure were included in the analysis. The level of significance was set at P < 0.05.
We conducted univariate linear regression analysis to identify variables associated with pain relief and function improvement. Then we performed multivariate linear regression analysis including variables with a P value < 0.05 defined by univariate analysis.
For this study, the sample size calculation for patients was done according to the previous study by Pujol and colleagues [24]. Our hypothesis was that there would be a 6% relative difference in failure rate between the PRP and non-PRP group, which meant that a sample size of 349 patients in each group was needed to get a power of 80% for a significance level of 5%.