Safety of, and biological and functional response to, a novel metallic implant for the management of focal full-thickness cartilage defects: Preliminary assessment in an animal model out to 1 year.

Focal full-thickness cartilage lesions of the human medial femoral condyle (MFC) can cause pain and functional impairment. Affected middle-aged patients respond unpredictably to existing treatments and knee arthroplasty may be required, prompting risk of revision. This study assesses the safety of, and biological and functional response to, a metallic resurfacing implant which may delay or obviate the need for traditional arthroplasty. The anatomic contour of the surgically exposed MFC of six adult goats was digitally mapped and an 11 mm diameter full-thickness osteochondral defect was created. An anchor-based Co-Cr resurfacing implant, matching the mapped articular contour, was implanted. Each goat's contralateral unoperated femorotibial joint was used as a control. Postoperative outcome was assessed by lameness examination, radiography, arthroscopy, synoviocentesis, necropsy, and histology up to 26 (n = 3) or 52 (n = 3) weeks. By postoperative week (POW) 4, goats demonstrated normal range of motion, no joint effusion, and only mild lameness in the operated limb. By POW 26 the animals were sound with only occasional very mild lameness. Arthroscopy at POW 14 revealed moderate synovial inflammation and a chondral membrane extending centrally across the implant surface. Radiographs at POWs 14 to 52 implied implant stability in the operated joints, as well as subchondral bone remodeling and mild exostosis formation in the operated and contralateral unoperated joints of some goats. By POW 26, histology revealed new trabecular bone abutting the implant. At POWs 26 and 52 MFC cartilage was metachromatic and intact in the operated and unoperated femorotibial joints. Proximal tibiae of some operated and unoperated limbs demonstrated limited subchondral bone remodeling and foci of articular cartilage fibrillation and thinning. The chondral membrane crossing the prosthesis possessed a metachromatic matrix containing singular and clustered chondrocytes. Our data imply the safety, biocompatibility, and functionality of the implant. Focal articular damage was documented in the operated joints at POWs 26 and 52, but lesions were much reduced over those previously reported in untreated defects. Expanded animal or preclinical human studies are justified.     

Critical‐size defect induces unicompartmental osteoarthritis in a stable ovine knee

Animal models simulating osteoarthritis are frequently associated with irreversible changes in biomechanics. Although these models successfully induce osteoarthritis, results of experimental repair procedures are impaired by biomechanical problems. The aim of this study was to define the critical size of a chondral lesion to induce unicompartmental osteoarthritis in a stable joint. Sixteen sheep were randomly divided into four treatment groups. A cartilage defect (7‐ or 14‐mm diameter) was created in the weight‐bearing zone of the medial femoral condyle. The sheep were mobilized for 6 or 12 weeks. Osteoarthritis was determined by gross assessment, India ink staining, histology (Mankin score), and analysis of COMP in the serum. In the 6‐week group, only minor osteoarthritis was registered for either defect size. After 12 weeks, the 14‐mm defect induced minor osteoarthritis at the femoral condyle and caused significant degenerative changes at the tibial articular cartilage and the meniscus. The 7‐mm defect created focal unicompartmental osteoarthritis at the medial femoral condyle and minor degenerative changes at the corresponding tibia. A 7‐mm full‐thickness chondral defect with a weight‐bearing regimen of 12 weeks induced local osteoarthritis at the medial compartment in an otherwise stable joint as aimed. © 2011 Orthopaedic Research Society Published by Wiley Periodicals, Inc. J Orthop Res 30:214–220, 2012     

Effects of exercise level on biomarkers in a rat knee model of osteoarthritis

The aim of this study was to elucidate whether the levels of serum biomarkers reflect the progression of osteoarthritis (OA) induced by different levels of exercise. Thirty‐five Wistar rats subjected to anterior cruciate ligament transaction (ACLT) were divided into three groups: Control, moderate running, and intense running. Twelve rats (moderate running without ACLT) were allocated as a naive group. Running was performed on a motorized treadmill, at a speed of 18 m/min for 30 min/day (moderate and naive) or 60 min/day (intense) for 3 days per week. After 2 or 4 weeks, OA histopathology in the knees was evaluated using the Osteoarthritis Research Society International (OARSI) score, and the serum levels of cleaved collagen type II (C2C) and procollagen II C‐propeptide (CPII) were analyzed. The OARSI score deteriorated in the intense running group after 2 weeks and the serum C2C/CPII ratio suggested the development of OA. At 4 weeks, the C2C/CPII ratio suggested there would be deterioration in the OARSI score but the score did not differ significantly between the moderate and intense running groups. C2C/CPII ratio had 13–25% correlation with the OARSI histological score. Thus, in rat experimental OA, the OARSI score could be partially predicted by the C2C/CPII ratio as a serum biomarker of OA. © 2013 Orthopaedic Research Society Published by Wiley Periodicals, Inc. J Orthop Res 31:1026–1031, 2013     

大鼠膝关节不稳定骨性关节炎软骨与软骨下骨组织微观结构的变化

[目的]观察大鼠膝关节经前交叉韧带切断术(anterior cruciate ligament transection,ACLT)造成关节不稳定情况下,其关节软骨与软骨下骨骨组织微结构的变化.[方法] 20只3个月龄雄性SD大鼠按体重随机分为对照组与实验组,每组10只动物.实验组切断右膝关节前交叉韧带,对照组只暴露前交叉韧带而不切断,术后12周处死所有动物并取材.取右肢股骨远端膝关节于70％酒精固定后脱钙6周,石蜡包埋切片,经HE染色及Masson染色,普通光学显微镜观察摄片,采用图像分析系统做定量分析;取右肢胫骨膝关节于70％酒精固定后不脱钙做硬组织包埋切片,行Van Kossa染色后,定量分析软骨下骨组织微观结构变化.[结果]术后12周后大体观察发现,实验组关节面较对照组明显失去光泽,呈颗粒感,部分软骨面破溃,关节面周围可见骨赘形成;实验组关节软骨HE及Masson染色着色不均,软骨细胞排列紊乱,且实验组Mankin评分结果显著高于对照组,差异具有统计学意义(P＜0.05);实验组软骨下骨组织微结构较对照组改变明显,差异具有统计学意义(P＜0.05).[结论]ACLT术导致膝关节不稳定,在此情况下对大鼠膝关节软骨形态及软骨下骨组织微结构造成显著破坏,若不能及时采取干预措施,将加速膝关节功能退变进程.     

Improved repair of chondral and osteochondral defects in the ovine trochlea compared with the medial condyle

Associations between topographic location and articular cartilage repair in preclinical animal models are unknown. Based on clinical investigations, we hypothesized that lesions in the ovine femoral condyle repair better than in the trochlea. Full‐thickness chondral and osteochondral defects were simultaneously established in the weightbearing area of the medial femoral condyle and the lateral trochlear facet in sheep, with chondral defects subjected to subchondral drilling. After 6 months in vivo, cartilage repair and osteoarthritis development was evaluated by macroscopic, histological, immunohistochemical, and biochemical analyses. Macroscopic and histological articular cartilage repair and type‐II collagen immunoreactivity were better in the femoral trochlea, regardless of the defect type. Location‐independently, osteochondral defects induced more osteoarthritic degeneration of the adjacent cartilage than drilled chondral lesions. DNA and proteoglycan contents of chondral defects were higher in the condyle, reflecting physiological topographical differences. The results indicate that topographic location dictates the structural patterns and biochemical composition of the repair tissue in sheep. These findings suggest that repair of cartilage defects at different anatomical sites of the ovine stifle joint needs to be assessed independently and that the sheep trochlea exhibits cartilage repair patterns reflective of the human medial femoral condyle. © 2013 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 31:1772–1779, 2013     

Bone marrow stimulation of the medial femoral condyle produces inferior cartilage and bone repair compared to the trochlea in a rabbit surgical model

The influence of the location of cartilage lesions on cartilage repair outcome is incompletely understood. This study compared cartilage and bone repair in medial femoral condylar (MFC) versus femoral trochlear (TR) defects 3 months after bone marrow stimulation in mature rabbits. Intact femurs from adult rabbits served as controls. Results from quantitative histomorphometry and histological scoring showed that bone marrow stimulation produced inferior soft tissue repair in MFC versus TR defects, as indicated by significantly lower % Fill (p = 0.03), a significant increase in collagen type I immunostaining (p < 0.00001) and lower O'Driscoll scores (p < 0.05). 3D micro‐CT analysis showed that repaired TR defects regained normal un‐operated values of bone volume fraction, trabecular thickness, and trabecular number, whereas in MFC defects the repaired bone architecture appeared immature and less dense compared to intact un‐operated MFC controls (p < 0.0001). Severe medial meniscal damage was found in 28% of operated animals and was strongly correlated with (i) low cartilage defect fill, (ii) incomplete bone repair in MFC, and (iii) with a more posterior defect placement in the weight‐bearing region. We conclude that the location of cartilage lesions influences cartilage repair, with better outcome in TR versus MFC defects in rabbits. Meniscal degeneration is associated with cartilage damage. © 2013 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 31:1757–1764, 2013     

Chondroprotective effect of high‐dose zoledronic acid: An experimental study in a rabbit model of osteoarthritis

To address the need to impact the subchondral bone‐articular cartilage interaction for the treatment of degenerative osteoarthritis (OA), bisphosphonates may be used as a means to inhibit the subchondral bone resorption. The purpose of the present study is to evaluate the chondroprotective effect of zoledronic acid (ZOL) in a model of OA. Eighteen adult male rabbits underwent an anterior cruciate ligament transection and were separated into two groups: ZOL group (n = 10) received 0.6 mg/kg intravenous injection of ZOL on day 1, 15, and 29 and placebo group (n = 8) received saline. The animals were euthanized at 8 weeks. Macroscopically, the ZOL group had significantly milder ulcerations, cartilage softening and fibrillation compared to the placebo group. Microscopically, morphology of the articular cartilage was better in the ZOL treated group compared with the placebo group, without complete disorganization in any section of the ZOL group. Furthermore, the chondrocytes in the ZOL treated group were mainly cloning, indicating cartilage repairing and regeneration process, while in the placebo group hypocellularity predominated. Additionally, subchondral necrosis was evident in some specimens of the placebo group. Zoledronic acid, in a high‐dose regimen, proved to be chondroprotective in a well‐established animal model of OA. © 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 32:1646–1651, 2014.     

Glenoid cartilage mechanical properties decrease after rotator cuff tears in a rat model

Rotator cuff repairs are commonly performed to reduce pain and restore function. Tears are also treated successfully without surgical intervention; however, the effect that a torn tendon has on the glenohumeral cartilage remains unknown. Clinically, a correlation between massive rotator cuff tears and glenohumeral arthritis has often been observed. This may be due to a disruption in the balance of forces at the shoulder, resulting in migration of the humeral head and subsequently, abnormal loading of the glenoid. Our lab previously demonstrated changes in ambulation and intact tendon mechanical properties following supraspinatus and infraspinatus rotator cuff tendon tears in a rat model. Therefore, the purpose of this study was to investigate the effects of supraspinatus and infraspinatus rotator cuff tears on the glenoid cartilage. Nine rats underwent unilateral detachment of the supraspinatus and infraspinatus tendons and were sacrificed after 4 weeks. Cartilage thickness significantly decreased in the antero‐inferior region of injured shoulders. In addition, equilibrium elastic modulus significantly decreased in the center, antero‐superior, antero‐inferior, and superior regions. These results suggest that altered loading after rotator cuff injury may lead to damage to the joint with significant pain and dysfunction. Clinically, understanding the mechanical processes involved with joint damage will allow physicians to better advise patients. © 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 30:1435–1439, 2012     

人工全膝关节置换术中联合股骨外髁滑移截骨术矫正股骨外弓畸形的疗效分析

目的探讨人工全膝关节置换术(total knee arthroplasty,TKA)中采用股骨外髁滑移截骨术(lateral condyle sliding osteotomy,LCSO)矫正股骨外弓畸形的疗效。方法回顾分析2018年7月—2020年7月TKA中采用LCSO治疗的17例伴股骨外弓畸形的骨关节炎患者临床资料。男3例,女14例;年龄58~68岁,平均63.2岁。股骨外弓畸形病因:股骨发育畸形12例,股骨骨折畸形愈合5例。膝关节骨关节炎KellgrenLawrence分级:Ⅲ级4例,Ⅳ级13例。术前生理外翻角为9.5°~12.5°,平均10.94°。病程3~25年,平均15.1年。术前及末次随访时测量股骨远端机械外侧角(mechanical lateral distal femur angle,mLDFA)、髋-膝-踝角(hip-knee-ankle angle,HKA)、机械轴偏向(mechanical axis deviation,MAD),评估关节外畸形在关节内矫正及下肢机械力线恢复情况;采用膝关节学会评分系统(KSS)膝评分和功能评分、疼痛视觉模拟评分(VAS)、膝关节活动度(range of motion,ROM)评估疗效;行膝内、外翻应力试验,X线片复查截骨片愈合情况,评估关节稳定性及LCSO的安全性。结果术后患者切口均Ⅰ期愈合,无切口感染、下肢深静脉血栓形成等术后早期并发症发生。17例患者均获随访,随访时间12~36个月,平均23.9个月。截骨片均达骨性愈合,愈合时间2~5个月,平均3.1个月。术后膝内、外翻应力试验均为阴性,未发生外侧副韧带松弛、断裂,膝关节不稳,假体松动、翻修、感染等情况。末次随访时mLDFA、HKA、MAD及膝关节ROM、VAS评分、KSS膝评分和功能评分均较术前显著改善,差异有统计学意义(P<0.05)。结论在伴有股骨外弓畸形TKA中应用LCSO疗效确切且安全,关节外畸形在关节内矫正,一次手术可同时恢复下肢机械力线和关节平衡。     

关节镜下微骨折术联合自体骨软骨移植治疗膝关节股骨髁大面积软骨损伤

目的探讨关节镜下微骨折术联合自体骨软骨移植(osteochondral autologous transplantation,OAT)治疗膝关节股骨髁大面积(4~6 cm^2)软骨损伤的疗效。方法2016年3月-2017年6月,采用关节镜下微骨折术联合OAT治疗22例膝关节股骨髁大面积软骨损伤患者。其中男16例,女6例;年龄22~60岁,平均38.6岁。致伤原因:交通事故伤8例,运动损伤14例。病程1~6个月,平均3.4个月。股骨内侧髁损伤15例,外侧髁损伤7例;软骨损伤面积4~6 cm^2,平均4.98 cm^2。软骨损伤国际软骨修复协会(ICRS)分级:Ⅲ级9例,Ⅳ级13例。伴半月板损伤18例。术前疼痛视觉模拟评分(VAS)为(6.36±1.25)分,Lysholm评分为(36.00±7.77)分。结果术后切口均Ⅰ期愈合。患者均获随访,随访时间2~3年,平均2.3年。术后2年时VAS评分为(1.27±0.94)分,Lysholm评分为(77.82±6.21)分,均较术前明显改善(t=16.595,P=0.000;t=21.895,P=0.000)。术后2年,MRI显示软骨缺损区修复良好。结论关节镜下微骨折术联合OAT治疗膝关节股骨髁大面积软骨损伤早期疗效较好。     

Topographical investigation of changes in depth‐wise proteoglycan distribution in rabbit femoral articular cartilage at 4 weeks after transection of the anterior cruciate ligament

In this study, we explore topographical changes in proteoglycan distribution from femoral condylar cartilage in early osteoarthritis, acquired from both the lateral and medial condyles of anterior cruciate ligament transected (ACLT) and contralateral (CNTRL) rabbit knee joints, at 4 weeks post operation. Four sites across the cartilage surface in a parasagittal plane were defined across tissue sections taken from femoral condyles, and proteoglycan (PG) content was quantified using digital densitometry. The greatest depth‐wise change in PG content due to an ACLT (compared to the CNTRL group) was observed anteriorly (site C) from the most weight‐bearing location within the lateral compartment. In the medial compartment, the greatest change was observed in the most weight‐bearing location (site B). The depth‐wise changes in PG content were observed up to 48% and 28% depth from the tissue surface at these aforementioned sites, respectively (p < 0.05). The smallest depth‐wise change in PG content was observed posteriorly (site A) from the most weight‐bearing location within both femoral condyles (up to 20% and up to 5% depth from the tissue surface at lateral and medial compartments, respectively). This study gives further insight into how early cartilage deterioration progresses across the parasagittal plane of the femoral condyle. © 2015 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 33:1278–1286, 2015.     

In vivo contrast-enhanced micro MR-imaging of experimental osteoarthritis in the rabbit knee joint at 7.1T1

OBJECTIVE: In this longitudinal MR study the early stages of joint pathology in two surgically-induced rabbit models of osteoarthritis (OA) were monitored by in vivo contrast-enhanced MRI at 7.1T. Qualitative and quantitative MR data were compared with macroscopic and microscopic findings. METHOD: Scanning of mature, male New Zealand White rabbits (N=12) was performed before surgery, and at 2, 4, and 8 weeks after unilateral transection of the anterior cruciate ligament (ACLT), medial meniscectomy (ME), or sham operation. MR-images were simultaneously obtained of both knee joints after intravenous injection of Magnevist. We implemented a 2D T1-weighted (T1w) coronal, fat-saturated gradientecho protocol (68 x 138 microm2, slice thickness 1 mm). Additionally, consecutive 3D gradientecho images were obtained from two sham-operated and two rabbits of the ME group (234 x 273 x 234 microm(3)). ACLT animals were sacrificed at 2 weeks (N=1), and 8 weeks (N=3), ME animals were sacrificed at 4 weeks (N=2), and 8 weeks (N=4), and sham-operated animals were sacrificed at 2 weeks (N=1) and 8 weeks (N=1), respectively. RESULTS: Both OA models reflected important characteristics of the clinical picture of OA. With MR we were able to monitor time dependently the decline of synovial effusion and the formation of osteophytes. Morphologic MR examination showed a moderate to high accuracy for detecting synovial effusion (75%), meniscus (86%) and cruciate ligament (91%) lesions, and osteophytes (88%) as assessed by macroscopic examination. False-negative MR findings for gross macroscopic changes were due to the relative high slice thickness in 2D scans and the fact that the slices only covered the main weightbearing area of the femorotibial joint. Contour abnormalities of articular cartilage were not reliably detected. Quantitative analysis revealed a statistically significant increase of cartilage signal intensity in medial tibial cartilage (48+/-9% ACLT, and 29+/-9% ME in 2D datasets) as compared to contralateral control knees in two-week scans. Signal enhancement persisted or increased at later dates. CONCLUSION: With high-resolution contrast-enhanced MRI at 7.1T the time course of gross pathologic changes in rabbit knees with surgically induced OA can be monitored. Still insufficient spatial resolution and image contrast of the applied 2D protocols limit the sensitivity and prohibit detection of articular cartilage contour abnormalities. However, signal alterations in the cartilage layer indicate alterations of tissue composition at a very early stage of OA development. When used with 3D protocols, contrast-enhanced MRI offers a promising tool for qualitative and quantitative in vivo monitoring of OA in rabbit models.     

急性滑脱性髌股关节撞击症1例

患者,女,18岁,不慎跌倒致右膝关节肿痛、活动受限以及屈膝外翻外旋于2019年6 月23 日入院.行 X 线、MRI检查显示:右股骨外侧髁骨折并膝关节游离体形成,右髌骨内侧支持带撕裂并髌骨半脱位,关节腔积液.查体:右膝关节肿胀,关节活动疼痛伴活动受限,髌骨内侧支持带压痛,肢端感觉、活动可,生理反射存在,病理反射未引出....     

Impact of aging on gene expression in a rat model of ischemic cutaneous wound healing

BACKGROUND: Tissue ischemia and aging are independent features associated with the healing impairment of cutaneous wounds. However, the pathophysiology of these processes as they relate to impaired-healing wounds is poorly understood. MATERIALS AND METHODS: A single full-thickness biopsy wound was made on both ears of young (3-6 month) and aged (>24 month) Fisher rats. One ear was rendered ischemic by transection of the vasculature at the ear base, while the other ear served as an internal nonischemic control. Wounds were harvested from 3 to 7 days and were evaluated histologically for either granulation tissue formation and epithelialization. Total RNA from wounds harvested at postoperative day 7 was probed using a nylon-based cDNA array to assess global genetic expression alterations. RESULTS: Healing in the rat ear model is impaired by both ischemia and advanced age as measured by granulation tissue formation and wound epithelialization. Granulation tissue formation was affected to a greater degree by ischemia than age (-58% versus -21%, respectively) while epithelialization displayed an opposite response (-17% versus -53%, respectively). Global analysis of gene expression suggests that ischemia engenders a marked increase in genes displaying altered expression in aged animals compared to young animals. Importantly, all possible alterations in gene expression are found in samples from aged ischemic wounds, indicating that gene regulation is not simply depressed by advanced age. CONCLUSIONS: Wound epithelialization appears to be affected to a greater degree by advanced age than by ischemia. The results demonstrate the distinctive phenotype presented by the clinically relevant combination of age and ischemia in an in vivo model of cutaneous wound healing.     

Spontaneous osteonecrosis of the lateral femoral condyle of the knee: a report of 11 cases

BACKGROUND: Spontaneous osteonecrosis (SON) of the lateral femoral condyle of the knee joint is very rare and there have been only a few articles about this condition. MATERIALS: We reviewed data for 11 patients (7 men, 4 women) with unusual SON of the lateral femoral condyle of the knee. The average age of patients at onset was 61.9 years (range 47-76 years). No patient had underlying disease or had undergone steroid therapy, although one underwent lateral meniscectomy. RESULTS: According to Aglietti's radiographic classification, three patients had stage 1 disease, two had stage 2 disease, three had stage 3 disease, one had stage 4 disease, and two had stage 5 at first examination. The average alignment of affected limbs on standing was 5.9 degrees valgus (range 0 degrees -11 degrees ). Although the osteonecrotic lesion was seen on the lateral side, the mechanical axis passed the medial compartment in three patients. Six patients were treated conservatively and the remaining five required surgery, which consisted of lateral unicompartmental knee arthroplasties. CONCLUSION: The pathology of the necrosis of the lateral femoral condyle was considered to be different from that of the medial femoral condyle regarding clinical features, limb alignment, and radiographic findings.     

Changes of articular cartilage after immobilization in a rat knee contracture model

The objective was to determine the changes of articular cartilage of the knee joint during immobilization in a rat model. The knee joints of adult male rats were immobilized at 150° of flexion using an internal fixator for 3 days, and 1, 2, 4, 8, and 16 weeks. The articular cartilage from the medial midcondylar region of the knee was obtained, divided into three areas (non‐contact area, transitional area, contact area), and in each area, a degree of degeneration was evaluated by gross observation, histomorphometric grading, and measurements of thickness and number of chondrocytes. Elasticity of the articular cartilage was estimated by measuring the sound speed with use of scanning acoustic microscopy. Degeneration of the articular cartilage was mainly observed in the contact and transitional areas. Matrix staining intensity by safranin‐O and number of chondrocytes were decreased in these two areas. The thickness of the articular cartilage in the non‐contact and contact areas was unchanged, but it was increased in the transitional area. Decrease in sound speed was observed in the transitional area of both the femoral and tibial cartilage, indicating the softening of the articular cartilage. The changes of articular cartilage became obvious as early as 1 week after immobilization. These changes may be due to a lack of mechanical stress or a lack of joint fluid circulation during immobilization. Although we do not know the reversibility of these changes of articular cartilage, early mobilization is preferable to avoid these cartilage changes. © 2008 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 27:236–242, 2009     

Delivering rhFGF‐18 via a bilayer collagen membrane to enhance microfracture treatment of chondral defects in a large animal model

Augmented microfracture techniques use growth factors, cells, and/or scaffolds to enhance the healing of microfracture‐treated cartilage defects. This study investigates the effect of delivering recombinant human fibroblastic growth factor 18 (rhFHF18, Sprifermin) via a collagen membrane on the healing of a chondral defect treated with microfracture in an ovine model. Eight millimeter diameter chondral defects were created in the medial femoral condyle of 40 sheep (n = 5/treatment group). Defects were treated with microfracture alone, microfracture + intra‐articular rhFGF‐18 or microfracture + rhFGF‐18 delivered on a membrane. Outcome measures included mechanical testing, weight bearing, International Cartilage Repair Society repair score, modified O'Driscoll score, qualitative histology, and immunohistochemistry for types I and II collagen. In animals treated with 32 μg rhFGF‐18 + membrane and intra‐articularly, there was a statistically significant improvement in weight bearing at 2 and 4 weeks post surgery and in the modified O'Driscoll score compared to controls. In addition, repair tissue stained was more strongly stained for type II collagen than for type I collagen. rhFGF‐18 delivered via a collagen membrane at the point of surgery potentiates the healing of a microfracture treated cartilage defect. © 2015 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 33:1120–1127, 2015.