If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Analysis of Two Different Arthroscopic Broström Repair Constructs for Treatment of Chronic Lateral Ankle Instability in 110 Patients: A Retrospective Cohort Study
Address correspondence to: James M. Cottom, DPM, FACFAS, Florida Orthopedic Foot & Ankle Center Fellowship, 2030 Bee Ridge Road, Suite B, Sarasota, FL 34239.
Chronic lateral ankle instability is a common condition treated by most foot and ankle surgeons. Once conservative treatment has failed, patients often undergo surgical reconstruction, either anatomic or nonanatomic. The present retrospective cohort study compared the clinical outcomes of 2 different arthroscopic Broström procedures. A total of 110 patients (83 females [75.5%] and 27 males [24.5%]) were treated with 1 of the 2 lateral ankle stabilization techniques from October 1, 2014 to December 31, 2015. Of the 110 patients, 75 were included in the arthroscopic lateral ankle stabilization group with an additional suture anchor used proximally and 35 were included in the arthroscopic lateral ankle stabilization group using the knotless design. The age of the cohort was 46.05 ± 17.89 (range 12 to 83) years. The body mass index was 30.03 ± 7.42 (range 18.3 to 52.5) kg/m2. Of the 110 patients, 25 (22.7%) had undergone concomitant procedures during lateral ankle stabilization. Overall, postoperative complications occurred in 14 patients (12.7%). No statistically significant differences were found between the 2 groups regarding the complication rates, use of concomitant procedures, and the presence of diabetes and workers compensation claims. No statistically significant differences were found in the mean age, body mass index, or gender distribution between the 2 groups. The preoperative American Orthopaedic Foot and Ankle Society (AOFAS) Ankle-Hindfoot scores were 50.85 ± 13.56 (range 18 to 76) and 51.26 ± 13.32 (range 18 to 69) in groups 1 and 2, respectively. The postoperative AOFAS Ankle-Hindfoot scores were 88.19 ± 10.72 (range 54 to 100) and 84 ± 15.41 (range 16 to 100) in groups 1 and 2, respectively. No statistically significant difference was found between these 2 groups. The preoperative visual analog scale score was 7.45 ± 1.39 (range 3 to 10) and 6.97 ± 1.25 (range 5 to 10), which had improved to 1.12 ± 1.38 (range 0 to 5) and 1.8 ± 1.98 (range 1 to 9) postoperatively for groups 1 and 2, respectively. The difference in the postoperative visual analog scale score between the 2 groups was statistically significant. The preoperative and postoperative AOFAS scale, Foot Function Index, and Karlsson-Peterson scores showed no statistically significant differences between the 2 groups. From our experience, either procedure is an acceptable treatment option for chronic lateral ankle instability, with the knotless technique showing a trend toward more complications.
Patients with chronic lateral ankle instability often become surgical candidates when their ankle has not responded favorably to nonoperative care. Lateral ankle stabilization procedures are often separated into 2 general groups, anatomic and nonanatomic repairs (
) later modified this technique in 1980. In recent years, numerous investigators have described and advocated for arthroscopic and arthroscopically assisted techniques for lateral ankle stabilization (
). These techniques have the advantage over traditional open techniques of addressing intraarticular pathologic features at the same time as the ligamentous instability.
With more attention given to arthroscopic techniques for lateral ankle stabilization, some investigators have investigated the strength of these different constructs in cadaveric models (
Cottom JM, Baker JS, Richardson P, Maker J. A biomechanical comparison of 3 different arthroscopic lateral ankle stabilization techniques in 36 cadaveric ankles. Presented at the 74th Annual Scientific Conference of the American College of Foot and Ankle Surgeons, Austin, TX, 2016.
). The goal of the present study was to compare 2 different groups of patients who had undergone arthroscopic lateral ankle stabilization procedures using 1 of 2 constructs. The 2 constructs under investigation in the present study have been previously described by Cottom et al (
Cottom JM, Baker JS, Richardson P, Maker J. A biomechanical comparison of 3 different arthroscopic lateral ankle stabilization techniques in 36 cadaveric ankles. Presented at the 74th Annual Scientific Conference of the American College of Foot and Ankle Surgeons, Austin, TX, 2016.
) and include arthroscopic lateral ankle stabilization using an additional suture anchor proximally and arthroscopic lateral ankle stabilization with a knotless design. Our hypothesis for this investigation was that a measurable, clinical difference would be present between these 2 groups.
Patients and Methods
Medical Record Review
From the senior author's (J.M.C.) surgical records, patients were identified who had undergone an arthroscopic Broström procedure from October 1, 2014 to December 31, 2015 (15 months). This period represents the period the senior author (J.M.C.) was using both of the described constructs (the technique is described in the next section). Once the patients were identified, their medical records were analyzed and data abstracted, including age, gender, side, body mass index (BMI), presence of diabetes mellitus, and whether the treatment was a workers compensation claim. The operative reports were abstracted for the construct used in the procedure and for any concomitant procedures performed at the same surgery. The preoperative pain scores were abstracted from the records of the patients' preoperative encounter. Additionally, the American Orthopaedic Foot and Ankle Society (AOFAS) Ankle-Hindfoot scale and Foot Function Index (FFI) scores were collected preoperatively (
). These scoring systems are used routinely in the senior author's (J.M.C.) practice to measure pain and functional outcomes for surgical patients and are collected at regular intervals postoperatively. The AOFAS and FFI scores were abstracted from the records of the patients' most recent follow-up appointment. In addition, the Karlsson-Peterson ankle scoring system, which measures function and pain levels of the ankle joint (
), was completed postoperatively at the patients' most recent clinical follow-up visit. Finally, data regarding the patients' return to weightbearing and their interval to starting formal physical therapy was abstracted from the medical records.
Surgical Technique
Each patient received a popliteal nerve block by the anesthesia department preoperatively and was then brought into the operating room and placed in the supine position. General anesthesia was then administered. A well-padded thigh tourniquet was placed on the patient, and a thigh holder was positioned to have the foot elevated a few inches off the operating table (Fig. 1). The distal fibula, peroneal tendons, and intermediate dorsal cutaneous nerves were then outlined with a surgical marker (Fig. 2). A noninvasive ankle distractor was then applied and manual traction used to distract the ankle.
Fig. 1The patient was placed in the operating room in the supine position with the operative extremity in a thigh holder.
Fig. 2The landmarks were drawn before starting the procedure and included the distal fibula, peroneal tendons, and the course of the intermediate dorsal cutaneous nerve.
Standard anteromedial and anterolateral portals were placed, and ankle arthroscopy was performed. Extensive arthroscopic debridement was performed using a 4-mm camera and shaver. Any intraarticular pathologic features were addressed during arthroscopy and before proceeding to lateral ankle stabilization. Extraarticular pathologic features, such as peroneal tendon abnormalities, were typically addressed after the lateral ankle stabilization procedure. Attention was then directed toward the anterolateral gutter, where extensive debridement was performed to remove any synovitis that might result in impingement. Debridement of the distal fibula to bone was undertaken using an ablator to facilitate capsular and ligamentous adhesions (Fig. 3). Lateral stabilization of the ankle joint was then obtained with 1 of the 2 constructs.
Fig. 3(A) Extensive, chronic synovitis was noted within the ankle joint and removed using an arthroscopic shaver and arthroscopic ablator. (B) The distal fibula was debrided down to bone to help facilitate capsular and ligamentous adhesion to bone.
In the first group, the drill guide for the first anchor was placed through the anterolateral portal directly midline in the coronal plane and approximately 1 cm superior to the distal aspect of the fibula. A guide hole was drilled and the anchor then inserted and seated in place using a mallet. The placement of the anchor can be confirmed using the arthroscope (Fig. 4). The anchor system used was a 3.0-mm bioabsorbable anchor (BioComposite SutureTak®; Arthrex, Naples, FL). The drill guide was removed, with the sutures now visualized exiting the anterolateral portal. A microsuture lasso was then used to capture the anterior talofibular ligament, ankle capsule, and inferior extensor retinaculum (Fig. 5). The microsuture lasso was placed percutaneously and angled toward the anterolateral portal, with the first pass placed approximately 1.5 to 2 cm inferiorly and anteriorly to the distal fibula. The nitinol wire was then advanced and used to capture 1 strand from the suture anchor, which was then pulled back through the skin to exit at site 1. Approximately 1 cm anteriorly and superiorly to site 1, the lasso was advanced percutaneously in the same manner as described above, exiting the anterolateral portal. Strand 2 from the suture anchor was then pulled back through the skin to exit site 2. A second bone anchor was inserted using the same technique, with placement in the fibula at the level of the lateral talar dome (Fig. 6). The strands exited the anterolateral portal, and a microsuture lasso was again used to capture the individual strand approximately 1 cm anteriorly and superiorly to the previous strand for site 3 and site 4. Four individual strands should now be exiting the skin (Fig. 7). An accessory portal was then made between sites 2 and 3 using a blade to incise only the skin. A hemostat was used for blunt dissection until the inferior extensor retinaculum was probed (Fig. 8). A probe was then used to gather the strands subcutaneously, with all strands exiting the accessory portal (Fig. 9). The extremity was released from distraction. With the assistant holding the foot in a dorsiflexed and everted position, the strands for each individual bone anchor were tied to the appropriate tension (Fig. 10). The anterior talofibular ligament, ankle capsule, and inferior extensor retinaculum were then advanced and secured to the anterior fibula. The strands should not be cut at this point. A separate 1- to 2-cm incision was then made approximately 3 cm proximally to the distal fibula laterally in the midline of the bone. The fibula was visualized, and, using the 2.9-mm bioabsorbable anchor system (PushLock®; Arthrex), a hole was drilled into the fibula (Fig. 11). A curved hemostat was then directed subcutaneously through the incision, as close to fibula as possible, exiting the accessory portal, where the knots were tied. The hemostat was then used to capture the strands and advanced proximally, exiting at the fibular incision (Fig. 12). The strands were then secured into the fibula using the 2.9-mm bioabsorbable anchor, with additional tension applied to the sutures (Fig. 13). This technique creates a double-row construct using 3 suture anchors.
Fig. 4Placement of the first anchor in the double-row, 3-anchor technique can be confirmed arthroscopically.
Fig. 5(A) The entry point for the second pass with the curved suture passer. Note that this is approximately 1 cm anteriorly and superiorly to the first suture. (B) The suture passer being brought through the anterolateral portal.
Fig. 7With the bone anchors in place in the anterior distal fibula, the sutures have been passed through anterior talofibular ligament, inferior extensor retinaculum, and ankle capsule and can be seen exiting the skin. Note that each strand is spaced approximately 1 cm apart from each other.
Fig. 8An accessory incision was made between suture strands 2 and 3, and a hemostat was used to bluntly dissect down to the inferior extensor retinaculum.
Fig. 10With the surgical assistant holding the foot in a dorsiflexed and everted position, the sutures are tied onto themselves. This tightens the ankle capsule, anterior talofibular ligament, and inferior extensor retinaculum and reapproximates the foot to the anterior distal fibula.
Fig. 11Through a second accessory incision placed approximately 3 cm proximally to the tip of the fibula, a hole was drilled in the lateral fibula for the third anchor.
Fig. 12After the knots have been tied, the 4 suture strands were brought into the proximal accessory incision using a hemostat. Note the 18-gauge spinal needle placed in the drill hole for the additional suture anchor.
Fig. 13The 4 suture strands were then gathered and anchored into the lateral fibula using an additional suture anchor, creating a double-row, 3-anchor construct.
In the second group, a knotless construct was created using 4 suture anchors. The construct incorporates a smooth, 1.5-mm, low-profile polyethylene suture material, Labral Tape (Arthrex). Using a technique similar to that for the first group, the first anchor was placed 1 cm dorsally to the tip of the fibula, with care taken to ensure the anchor did not violate the medial and lateral cortices of the fibula. A second anchor was inserted in the same fashion as for the first group, although it was placed 1 cm proximally to the first anchor. Next, the 2 tape strands, corresponding to each anchor, were passed through the anterior talofibular ligament and inferior extensor retinaculum. This was accomplished with the same suture lasso used in the first group. A separate 1- to 2-cm incision was made approximately 3 cm proximally to the distal fibula. The fibula was visualized, and, using the 2.9-mm suture anchor system, 2 holes were drilled into the fibula, with care taken to not violate the anterior and posterior cortices. At this point, an 18-gauge spinal needle was placed into each drill hole to maintain a visualized position. The foot was dorsiflexed and everted by the assistant. A strand from each anchor in the anterior fibula was then inserted into 1 of the lateral drill holes with a suture anchor (Fig. 14). The last 2 strands were crossed and fixated into the fibula using the exact same method, thus, creating a crossed suture anchor construct. An example of this construct is shown in Fig. 15, demonstrated in a cadaveric specimen.
Fig. 14The smooth, low-profile polyethylene suture material, Labral Tape (Arthrex, Naples, FL) has been gathered and passed through the accessory incision, after being passed through the ankle capsule, anterior talofibular ligament, and inferior extensor retinaculum. The sutures are shown in a crossed technique, which will ultimately be anchored into the lateral fibula more proximally.
Fig. 15The knotless, double-row, 4-anchor construct is depicted in a cadaveric model. Note that 1 strand from each of the anchors in the anterior distal fibula is crossed and brought into a corresponding anchor in the lateral fibula. This creates a knotless, crossed suture anchor fixation.
The mean age, BMI, proportion of female gender, and the prevalence of diabetes and workers compensation cases were compared using Student's t test, χ2 test, or Fisher's exact test, as appropriate, between the 2 groups. The mean preoperative and postoperative AOFAS, FFI, visual analog scale (VAS), and Karlsson-Peterson scores were calculated for each group and compared using a 2-way unpaired t test. The follow-up period was defined as the interval from the date of the index procedure to the last available clinic records. A minimum of 6 months of follow-up time was required for inclusion in the study. The interval to weightbearing with and without crutches, interval to weightbearing with regular shoes, and interval to participation in formal physical therapy were analyzed using Student's t test. The association of complications and concomitant procedures was analyzed using the χ2 test. All statistical analyses were performed using the R Statistical Package (R Foundation, Vienna, Austria).
Results
A total of 128 consecutive arthroscopic lateral ankle stabilization procedures were performed using 1 of the 2 techniques from October 1, 2014 to December 31, 2015 by the senior author (J.M.C.) (mean average of approximately 8.5 procedures per month). Of the 128 patients, 7 were excluded from the analysis because they had undergone concomitant procedures with the lateral ankle stabilization that changed the postoperative weightbearing protocol. Also, 11 patients did not have the minimum 6-month follow-up data available. The follow-up period for the remaining 110 patients was 13.2 ± 12.5 (range 6 to 21) months. Of the 110 patients, 83 were female (75.5%) and 27 were male (24.5%; Table 1). The mean age of the cohort was 46.05 ± 17.89 (range 12 to 83) years. The BMI was 30.03 ± 7.42 (range 18.3 to 52.5) kg/m2. Of the 110 patients, 25 (22.7%) had undergone concomitant procedures with the lateral ankle stabilization (Table 2). Postoperative complications occurred in 14 patients (12.7%) in the cohort (Table 3). No statistically significant differences were found between the 2 groups regarding the complication rates, use of concomitant procedures, or presence of diabetes and workers compensation claims (Table 4).
Table 1Patient characteristics
Characteristic
Additional Suture Anchor (n = 75)
Knotless (n = 35)
p Value
Age (y)
44.7 ± 18.57
48.9 ± 16.52
.260
Gender (n)
.007
Female
51 (68)
32 (91.43)
Male
24 (32)
3 (8.57)
BMI (kg/m2)
30 ± 7.64
30.2 ± 7.04
.907
Abbreviation: BMI, body mass index. Categorical data expressed as count (%) and continuous data as mean ± standard deviation.
The 2 groups were compared as detailed in the “Statistical Analysis” section. Of the 110 patients, 75 were in group 1 and 35 were in group 2. No statistically significant differences were found in the mean age, BMI, or gender distribution between the 2 groups (Table 1). The preoperative AOFAS scores were 50.85 ± 13.56 (range 18 to 76) and 51.26 ± 13.32 (range 18 to 69) in groups 1 and 2, respectively. No statistically significant difference was found between the 2 groups. The postoperative AOFAS scores were 88.19 ± 10.72 (range 54 to 100) and 84 ± 15.41 (range 16 to 100) in groups 1 and 2, respectively. Again, no statistically significant difference was found between the 2 groups. The preoperative VAS score was 7.45 ± 1.39 (range 3 to 10) and 6.97 ± 1.25 (range 5 to 10), which had improved to 1.12 ± 1.38 (range 0 to 5) and 1.8 ± 1.98 (range 1 to 9) postoperatively for groups 1 and 2, respectively. The difference in the postoperative VAS score between groups 1 and 2 was statistically significant (Table 5). The preoperative and postoperative AOFAS, FFI, and Karlsson-Peterson scores showed no statistically significant differences between the 2 groups (Table 5). The return to weightbearing data and the interval to starting physical therapy are listed in Table 6.
Table 5Results of outcome measures
Outcome Measure
Additional Suture Anchor (n = 75)
Knotless (n = 35)
p Value
AOFAS score (range 0 to 100)
Preoperatively
50.85 ± 13.56
51.25 ± 13.18
.883
Postoperatively
88.16 ± 10.73
84 ± 15.41
.101
VAS score (range 1 to 10)
Preoperatively
7.45 ± 1.39
6.97 ± 1.25
.083
Postoperatively
1.12 ± 1.38
1.8 ± 1.98
.039
FFI
Preoperatively
66.17 ± 15.31
64.03 ± 17.8
.517
Postoperatively
19.57 ± 14.80
26.14 ± 20.35
.058
Postoperative KP score
85.26 ± 11.61
82.17 ± 17.79
.277
Abbreviations: AOFAS, American Orthopaedic Foot and Ankle Society (Ankle-Hindfoot scale); FFI, Foot Function Index; KP, Karlsson-Peterson scale; VAS, visual analog scale.
Numerous arthroscopic and arthroscopically assisted lateral ankle stabilization procedures have been described in recent years, with the proponents of these techniques citing advantages such as smaller incisions, earlier weightbearing and functional rehabilitation, and the ability to address intraarticular pathologic features at the same time as addressing the instability (
Cottom JM, Baker JS, Richardson P. The “all-inside” arthroscopic Broström procedure with an additional suture anchor augmentation: a prospective study of 45 consecutive patients. Presented at the 74th Annual Scientific Conference of the American College of Foot and Ankle Surgeons, Austin, TX, 2016.
With more attention being given to these arthroscopic techniques, the strength of the particular construct has been questioned. The strength of various constructs has been tested in cadaveric laboratory studies (
Cottom JM, Baker JS, Richardson P, Maker J. A biomechanical comparison of 3 different arthroscopic lateral ankle stabilization techniques in 36 cadaveric ankles. Presented at the 74th Annual Scientific Conference of the American College of Foot and Ankle Surgeons, Austin, TX, 2016.
Cottom JM, Baker JS, Richardson P, Maker J. A biomechanical comparison of 3 different arthroscopic lateral ankle stabilization techniques in 36 cadaveric ankles. Presented at the 74th Annual Scientific Conference of the American College of Foot and Ankle Surgeons, Austin, TX, 2016.
), the biomechanical strength of 3 different constructs was tested in a cadaveric model. The 2 constructs investigated in the present study were 2 of the 3 tested. The construct used in group 1 in the present study showed an ultimate load to failure of 246.82 ± 82.37 N. The construct used in group 2 had an ultimate load to failure of 206.62 ± 55.62 N (
Cottom JM, Baker JS, Richardson P, Maker J. A biomechanical comparison of 3 different arthroscopic lateral ankle stabilization techniques in 36 cadaveric ankles. Presented at the 74th Annual Scientific Conference of the American College of Foot and Ankle Surgeons, Austin, TX, 2016.
). In the biomechanical study, this difference was not found to be statistically significant. However, the difference in the strength of both of these constructs was statistically significant in terms of the load to failure compared with the arthroscopic Broström procedure without modification (
Cottom JM, Baker JS, Richardson P, Maker J. A biomechanical comparison of 3 different arthroscopic lateral ankle stabilization techniques in 36 cadaveric ankles. Presented at the 74th Annual Scientific Conference of the American College of Foot and Ankle Surgeons, Austin, TX, 2016.
With these biomechanical findings, the goal of the present study was to evaluate these 2 constructs in the clinical setting to determine whether any significant differences would exist between the 2 in terms of the clinical outcomes. The findings from the present study have shown that any clinical difference between the use of 1 of the 2 constructs versus the other is minimal. No significant differences were found between the 2 groups in terms of the AOFAS, FFI, or Karlsson-Peterson scores either preoperatively or postoperatively. The difference in the preoperative VAS scores between the 2 groups was not statistically significant; however, the difference in the postoperative VAS score was statistically significant. The postoperative VAS score for group 1 was 1.12 ± 1.38 (range 0 to 5), but the postoperative VAS score for group 2 was 1.8 ± 1.98 (range 1 to 9). It is unclear whether this difference would be clinically significant, in particular, because the functional outcomes scores were not significantly different between the 2 groups.
The complication rate of the 2 groups was also compared. The complication rate was greater in group 2 (17.1%) than in group 1 (10.7%). We believe that group 2 had a greater percentage of complications owing to the technique and materials used. The suture tape used in the construct could have a negative effect on the outcomes because its foot print is larger than the more traditional arthroscopic Broström technique with fiber wire. This, theoretically, could lead to a greater incidence of soft tissue impingement, nerve entrapment, and other common complications. Also, it is likely that more complications were reported owing to the high number of patients who were not available for follow-up examinations. Most of these patients were in group 2, which could have changed the outcomes for several different measures, including the incidence of complications.
) described their technique for the arthroscopic Broström procedure and their experience with this technique. They used a similar technique to that used in group 1 of the present study. However, they limited the use of the additional suture anchor to those patients who require additional strength in the repair after tying of the suture knots. In the present study, this additional suture anchor was used routinely in the senior author's (J.M.C.) practice.
) described a technique for arthroscopic lateral ankle stabilization. They described a technique that uses a single accessory portal, in addition to the standard arthroscopy portals. They reported that at a mean follow-up period of 22.3 months, the patients had improved from a mean AOFAS score of 67 to 97. They had treated 16 patients with this technique, with no complications.
The present study had several limitations. First, we did not perform a power analysis to determine the number of patients needed for the present study. Second, the construct used in each patient was not blinded. Although this theoretically would not have had an effect on the subjective nature of the questionnaires or on the complication rate, it might have had an effect on the objective portion of the AOFAS questionnaire. Third, the follow-up time was a minimum of 6 months; thus, only the short-term results of the procedures could be evaluated. However, because the goal of the present study was to compare 2 different constructs, we believed this study period would be adequate to assess any differences between the 2 groups. The abbreviated follow-up time also allowed us to include the data from more patients in the statistical analysis, which should have served to increase the power of the study.
In conclusion, the present study found no statistically significant differences between the 2 constructs evaluated in terms of clinical outcome. A trend was found toward fewer complications using the 3-anchor construct described in group 1. In the senior author's (J.M.C.) practice, this trend has led to using the construct used in group 1 exclusively for arthroscopic lateral ankle stabilization procedures and abandoning the knotless technique. As arthroscopic lateral ankle stabilization becomes more widely used, further research will be needed to evaluate different constructs of fixation, different postoperative rehabilitation protocols, the indications, the contraindications, and numerous other factors to optimize patient outcomes.
Cottom JM, Baker JS, Richardson P, Maker J. A biomechanical comparison of 3 different arthroscopic lateral ankle stabilization techniques in 36 cadaveric ankles. Presented at the 74th Annual Scientific Conference of the American College of Foot and Ankle Surgeons, Austin, TX, 2016.
Cottom JM, Baker JS, Richardson P. The “all-inside” arthroscopic Broström procedure with an additional suture anchor augmentation: a prospective study of 45 consecutive patients. Presented at the 74th Annual Scientific Conference of the American College of Foot and Ankle Surgeons, Austin, TX, 2016.
30362-9&id=gr1.jpg){kind=link}
30362-9&id=gr2.jpg){kind=link}
30362-9&id=gr3.jpg){kind=link}
30362-9&id=gr4.jpg){kind=link}
30362-9&id=gr5.jpg){kind=link}
30362-9&id=gr6.jpg){kind=link}
30362-9&id=gr7.jpg){kind=link}
30362-9&id=gr8.jpg){kind=link}
30362-9&id=gr9.jpg){kind=link}
30362-9&id=gr10.jpg){kind=link}
30362-9&id=gr11.jpg){kind=link}
30362-9&id=gr12.jpg){kind=link}
30362-9&id=gr13.jpg){kind=link}
30362-9&id=gr14.jpg){kind=link}
30362-9&id=gr15.jpg){kind=link}