OsteoArthritis and Cartilage (2006) 14, 1119e1125

ª 2006 Published by Elsevier Ltd on behalf of OsteoArthritis Research Society International.doi:10.1016/j.joca.2006.05.003

InternationalCartilageRepairSociety

Results after microfracture of full-thickness chondral defects in differentcompartments in the kneeP. C. Kreuz M.D.y*, M. R. Steinwachs Ph.D.y, C. Erggelet Ph.D.y, S. J. Krausey,G. Konrad M.D.y, M. Uhl M.D.z and N. Sudkamp Ph.D.yyDepartment of Orthopaedic and Trauma Surgery, Albert Ludwig University Freiburg, Hugstetterstrasse 55,79106 Freiburg, GermanyzDepartment of Radiology, Albert Ludwig University Freiburg, Hugstetterstrasse 55, 79106 Freiburg,Germany

Summary

Objective: To determine if the clinical results after microfracture of full-thickness cartilage lesions deteriorate over a period of 36 months.

Methods: Between 1999 and 2002 85 patients (mean age 39.5 years) with full-thickness cartilage lesions underwent the microfractureprocedure and were evaluated preoperatively and 6, 18 and 36 months after surgery. Exclusion criteria were meniscal pathologies, axial mal-positioning and ligament instabilities. Baseline clinical scores were compared with follow-up data by paired Wilcoxon-tests for the modifiedCincinnati knee and the International Cartilage Repair Society (ICRS)-score. The effects of the lesion localization and Magnetic resonanceimaging (MRI) parameters were evaluated using the Pearson correlation and independent samples tests.

Results: Both scores revealed significant improvement 18 months after microfracture (P< 0.0001). Within the second 18 months after surgerythere was a significant deterioration in the ICRS-score (P< 0.0001). The best results could be observed in chondral lesions of the femoralcondyles. Defects in other areas of the knee deteriorated between 18 and 36 months after microfracture. MRI 36 months after surgeryrevealed best defect filling in lesions on the femoral condyles with significant difference in the other areas (P< 0.02). The Pearson coefficientof correlation between defect filling and ICRS-score was 0.84 and significant at the 0.01 level.

Conclusions: Microfracture is a minimal invasive method with good short-term results in the treatment of small cartilage defects. A deteriora-tion of the results starts 18 months after surgery and is most evident in the ICRS-score. The best prognostic factors have young patients withdefects on the femoral condyles.ª 2006 Published by Elsevier Ltd on behalf of OsteoArthritis Research Society International.

Key words: Osteoarthritis, Marrow stimulation technique, Microfracture, Cartilage lesion.

Introduction

Microfracture is one of many methods available to treat ar-ticular cartilage lesions. The technique has been elaboratedby Steadman et al. and applied chiefly in young athletesand in young patients generally1,2. There are some inherentadvantages. The microfracture technique is minimally inva-sive because it is arthroscopic through standard portals inmost cases. In comparison to an abrasion chondroplastythe subchondral bone plate is not completely destructed

1Parts of the study were presented at the nineteenth and twenty-first annual congress of the International Society for OrthopaedicsTraumatology and Sports Medicine, Munich, Germany, June 2004and May 2006 at the first common annual congress for orthopaedicand trauma surgery, Berlin, Germany, October 2005, and at thesixth ICRS-symposium (International Cartilage Repair Society) inSan Diego, USA, January 2006.

No support in form of grants, equipment or other items has occurredfor this project.

*Address correspondence and reprint requests to: PeterCornelius Kreuz, M.D., Department of Orthopaedic Surgery,Albert Ludwig University of Freiburg, Hugstetterstrasse 55, 79106Freiburg, Germany. Tel: 49-761-270-2401; Fax: 49-761-270-2619;E-mail: peter.kreuz@uniklinik-freiburg.de

Received 21 June 2005; revision accepted 9 May 2006.

1

but partially preserved between the microfracture holesimproving load-bearing characteristics following healing3.In contrast to Pridie drilling no heat necrosis or polishingis introduced into the subchondral bone and marrow withmicrofracture4,5. The equipment is standardized and thecosts are minimal since expensive cell cultures are not nec-essary. Unlike osteochondral, perichondral, periosteal orchondral autograft procedures the problem of a harvestsite morbidity is excluded6.

The intervention leads to a spontaneous repair response,which is based upon therapeutically induced bleeding fromthe opened subchondral bone spaces and subsequentblood-clot formation7. Knutsen has compared in a recentstudy the histological results after microfracture and autolo-gous chondrocyte implantation (ACI). The predominant re-pair tissue after microfracture was fibrocartilage. For thesmall number of biopsies and the short-term follow-up of 2years there was no significant difference in the histologicalresults after ACI8,9. Furthermore animal studies have shownthat the tissue formed after microfracture is fibrous in na-ture10 and not durable11,12. Therefore a long-term cure can-not be expected from the performance of this marrowstimulation technique.

Regardless of these poor biomechanical prerequisitesrecent studies with long-term follow-up over 10 years

119

1120 P. C. Kreuz et al.: Microfracture in full-thickness chondral defects

have shown improvement in function without significantdeterioration over the time. However these studies areonly functional outcome case series of patients with chon-dral lesions, based on special self-administered question-naires sent to the patients after surgical intervention4.There is only one prospective study with a 2-year follow-up and a serial objective observation by clinical and histo-logical evaluation9.

The present prospective study is based on scores witha continuous objective patient evaluation including controlMRI’s over a period of 36 months.

Patients and methods

Between 1999 and 2002 a total of 85 patients with full-thickness chondral defects of the knee underwent themicrofracture procedure. Thereof 70 patients (33 males,37 females) with an average age of 39.5 years (range19e55 years) met the inclusion criteria (below). Dependingon the localization of the defects (femoral condyles, troch-lea, tibia or retropatellar) the patients were assigned tofour different groups. Thirty-two patients had defects onthe femoral condyles, 16 on the trochlea, 11 on the tibiaand 11 on the patella. Regarding the different parametersthe four groups were relatively homogeneous: the averageage ranged between 38.5 and 41.6 years, the body massindex between 25.2 and 26.4 kg/m2 and the defect sizebetween 2.0 and 2.39 cm2. The details of the four studygroups appear in Table I.

INCLUSION CRITERIA

The criteria for inclusion were full-thickness chondraldefects grade III A,B according to the International CartilageRepair Society (ICRS)-classification system13 (InternationalCartilage Repair Society); that includes lesions extendingdown to the calcified layer (IIIB) but not through the subchon-dral bone plate (IIIC). The intact subchondral bone plate wasdetected preoperatively with an Magnetic resonance imag-ing (MRI), showing no subchondral edema or damage tothe bone and intraoperatively with a probe by arthroscopicpalpation. The lesions were sized between 1 and 4 cm2

and had to be localized in only one of the three knee com-partments (medial femoro-tibial joint, lateral femoro-tibialjoint or patello-femoral joint). Furthermore the study popula-tion was limited to patients between 18 and 55 years of age.

EXCLUSION CRITERIA

Excluded were patients with a trauma within the last 4weeks, varus or valgus deformities with a malalignmentover 5(, limits in knee extension or flexion less than130(, patella malalignment with medial or lateral shift ofmore than 0.5 cm, instabilities of the collateral or cruciateligaments, meniscal pathologies, morphine treatment formore than 3 months, intra-articular corticosteroid injections

within the previous month or knee arthroscopy within theprevious 6 months.

STUDY DESIGN

The prospective study with a 36-months follow-up wasconducted to determine the efficacy of microfracture in thepostoperative follow-up. Furthermore we wished to figureout if primary good results 6 or 12 months after surgeryremain on the same level or deteriorate after a period of36 months.

STUDY COURSE

Each patient underwent the same standardized evalua-tion progress pre-, intra- and postoperatively using the mod-ified Cincinnati-score (1¼ excellent, 2¼ good, 3¼ fair,4¼ poor)14 and the ICRS-scoring system (1¼ normal,2¼ nearly normal, 3¼ abnormal, 4¼ severely abnormal)13.For this purpose the patients had to answer standardizedquestionnaires and were examined by an orthopaedic sur-geon preoperatively and 6, 18 and 36 months aftersurgery. The average follow-up was 36 months (range34e38 months). Varus or valgus deformities were excludedby X-ray of the whole leg. The correct patella position wasdetected by a routine medio-lateral, patello-femoral and ax-ial joints radiographs. Furthermore an MRI was performedpreoperatively and 18 and 36 months after the surgicalintervention. According to the classification system ofHenderson et al. the following four criteria were evaluated:defect filling (1¼ complete, 2� 50% of the defect,3< 50%, 4¼ full-thickness defect), cartilage signal(1¼ normal¼ identical to the adjacent articular cartilage,2¼ nearly normal¼ slight areas of hyperintensity, 3¼abnormal¼ larger areas of hyperintensity, 4¼ absent), sub-chondral edema and effusion (both graded as 1¼ absent,2¼mild, 3¼moderate, 4¼ severe)15. Additionally, an over-all MRI-score was given, corresponding to the worst score inthe four categories. The amount of fill reflects the growth ofrepair tissue, the signal might be an indicator of maturity ofthe graft e especially in comparison to the signal of the ad-jacent normal cartilage, the subchondral edema might reflectthe attachment to the underlying bone, cartilage tears andthe process of graft remodeling and the effusion is an indica-tor for chronic irritation or infection of the joint15.

TREATMENT

The microfracture was performed by two experiencedsurgeons as described by Steadman et al.1,2, who recom-mended to debride first the exposed bone of all remainingunstable cartilage using a curette and a full-radius resectorto create a perpendicular edge of healthy well-attached via-ble cartilage around the defect. After preparation of the bed,the very small micro-holes were generated with an angularawl and distributed across the entire articular cartilage

Table IPreoperative patient information: the P-value between the different defect size, defect grade, body mass index and average age of the four

groups was always >0.05 (ManneWhitney-U-test). This makes the four groups comparable

Group Localizationof defect

Number ofpatients

Averageage (years)

Male:female Body massindex (kg/m2)

Defectsize (cm2)

Outerbridgeclassification

1 Femoral condyles 32 38.7 (19e55) 17:15 25.2 (19e31) 2.02 (1e3) 3.75 (3e4)2 Trochlea 16 41.6 (26e55) 8:8 26.4 (22e32) 2.31 (1e4) 3.94 (3e4)3 Tibia 11 39.7 (22e55) 3:8 25.6 (21e29) 2.39 (1.5e4) 3.64 (3e4)4 Retropatellar 11 38.5 (23e55) 5:6 25.2 (21e29) 2.0 (1e3) 3.73 (3e4)

1121Osteoarthritis and Cartilage Vol. 14, No. 11

lesion site, at a distance of 3e4 mm apart and down toa depth of 4 mm, thus yielding about 3e4 holes/cm2. Thusthe biomechanics of the subchondral bone plate was notdisturbed drastically. Finally the irrigation pump pressurewas reduced to control efficient bleeding from all generatedchannels. One drain without suction was placed intra-artic-ularly to reduce the risk of an arthrofibrosis in case of diffusebleeding. The drain did not affect the postoperative passiverange of motion since it was already removed on day 1 or 2after microfracture depending on the amount of secretion.Postoperative mobilization was started on the first day aftersurgery with a continuous passive motion machine for6e8 h/day16. For lesions on the weight-bearing surfaceson the condyles the initial range of motion was 30e60(and increased after 1 week as tolerated by 10e20( untilfull range of motion was obtained. For lesions of the pa-tello-femoral joint the flexion was limited with a brace for 2weeks to 40(, for another 2 weeks to 60( and until the sixthweek after surgery to 90(. The brace prevents excessiveshear forces on the maturing marrow clot. Cold therapywas used for all patients for 1 week. The patients werekept in hospital for 2e3 days after the surgical procedure.Low molecular weight heparin s.c. as well as non-steroidalanti-inflammatory medication were used. Crutch-assistedtouchdown weight-bearing ambulation was prescribed for6 weeks after the surgical procedure. Afterwards the pa-tients progressed to full weight-bearing and began a morevigorous program of active motion of the knee. Crutchfree walking was permitted after 3 months and return tosports was permitted after 4e6 months depending on theclinical examination.

STATISTICAL ANALYSIS

Descriptive statistics during follow-up (arithmetic mean,standard deviation, range) were calculated using standardformulas. Baseline clinical scores were compared with fol-low-up data by paired Wilcoxon-tests and Friedman-tests forthe modified Cincinnati knee score and the ICRS-score.

The effects of patient age, body mass index, location, gradeand size of lesion were assessed using Pearson correlationwith post hoc independent samples ManneWhitney-U-test.

Furthermore MRI-characteristics and clinical scores were com-pared using the Spearman coefficient of correlation17. All statis-tics were performed with SPSS and reviewed by anindependent statistician.

Results

All surgical procedures were performed without complica-tions. Postoperatively, there were two patients with a tempo-rary tingling around the arthroscopic portals, whichdisappeared completely after 3 months. Seven patientshad a persisting effusion over 4 weeks that had to betreated with knee elevation, anti-inflammatory medicationand cool dressings. There were no patients with postopera-tive infection or limited range of motion.

There was a significant improvement in all groupsbetween the preoperative scores and the scores at final fol-low-up (P< 0.0001). The Pearson coefficient of correlationbetween both scores (overall) was 0.81. However theCincinnati-score improved more as the ICRS-score(P< 0.0001) (Table II). The Cincinnati-score increased ingroup 12.34� 0.87 points, compared to 1.06� 1.29 pointsin group 2 and 1.45� 0.98 points in groups 3 and 4. Re-garding the ICRS-score the patients of group 1 improved

Table IIICRS- and Cincinnati-score preoperative and after 36 months in thefour groups. Patients of group 1 improved significantly more ( P<0.02) as patients in group 2, 3 or 4. Thirty-six months after micro-fracture the Pearson coefficient of correlation between both scores

(overall) was 0.81

Group AverageICRS-scorepreoperative

AverageICRS-score

after36 months

AverageCincinnati-

scorepreoperative

AverageCincinnati-score after36 months

1 3.53� 0.51 2.13� 0.75 4.0� 0.0 1.66� 0.872 3.88� 0.34 3.06� 1.0 4.0� 0.0 2.94� 1.293 3.91� 0.3 3.0� 0.77 4.0� 0.0 2.55� 1.044 3.64� 0.5 2.91� 0.83 4.0� 0.0 2.55� 0.93

Overall 3.69� 0.47 2.6� 0.92 4.0� 0.0 2.23� 113

Table III(a, b) Course of the ICRS- and Cincinnati-score between preoperative, 6, 18 and 36 months after microfracture. Listed is always the score-delta D (with its P-value) between two intervals. Significant deterioration (in bold type) occured in both scores in all groups between 18 and 36months postoperative except the patients of group 1. However regarding the whole study period of 36 months all groups improved significantly

in both scores

Group Improvement between preoperative and6 months postoperative

Improvement between 6 and18 months postoperative

ICRS-scoreD (P-value) Cincinnati-scoreD (P-value) ICRS-scoreD (P-value) Cincinnati-scoreD (P-value)

1 0.91 (<0.0001) 1.94 (<0.0001) 0.96 (<0.0001) 0.21 (¼0.008)2 0.5 (¼0.005) 1.5 (<0.0001) 1.0 (¼0.001) 0.25 (¼0.046)3 0.9 (¼0.002) 1.82 (¼0.002) 0.73 (¼0.005) 0.09 (¼0.317)4 0.91 (¼0.015) 1.73 (¼0.002) 0.64 (¼0.008) 0.09 (¼0.56)

Overall 079 (<0.0001) 1.79 (<0.0001) 0.86 (<0.0001) 0.18 (¼0.001)

Group Deterioration between 18 and36 months postoperative

Improvement between preoperative and36 months postoperative

ICRS-scoreD (P-value) Cincinnati-scoreD (P-value) ICRS-scoreD (P-value) Cincinnati-scoreD (P-value)

1 �0.47 (¼0.06) 0.19 (¼0.217) 1.4 (<0.0001) 2.34 (<0.0001)2 �0.68 ([0.013) �0.69 ([0.022) 0.82 (¼0.006) 1.06 (¼0.01)3 �0.73 ([0.005) �0.46 ([0.013) 0.91 (¼0.008) 1.45 (¼0.008)4 �0.82 ([0.007) �0.37 ([0.046) 0.73 (¼0.046) 1.45 (¼0.005)

Overall �0.56 (<0.0001) �0.2 (¼0.059) 1.09 (<0.0001) 1.77 (<0.0001)

1122 P. C. Kreuz et al.: Microfracture in full-thickness chondral defects

most as well. Patients of group 1 improved significantlymore (P< 0.02) compared to patients of the other threegroups (Table III) .

With respect to the follow-up intervals the patients im-proved from preoperative to 6 months and from 6 to 18months after microfracture (P< 0.05) (Fig. 1). However ingroups 3 and 4 there was no more significant improvementin the Cincinnati-score in the second interval. In the third in-terval between 18 and 36 months after microfracture thecourse of both scores changed (Fig. 2). There was signifi-cant deterioration of the ICRS- and the Cincinnati-score ingroups 2, 3 and 4 (P< 0.05). Only the patients of group 1revealed neither a significant improvement nor a significantdeterioration in both scores (P> 0.06) (Table III). Howeverthe older patients over 40 years of group 1 (n¼ 16) deteri-orated significantly in the ICRS-score (P¼ 0.005). In thisgroup younger patients (less than 40 years old; n¼ 16)had a significantly better clinical outcome than did older pa-tients (P¼ 0.008) (Figs. 3 and 4).

MRI 36 months after surgery revealed best defect filling ingroup 1 (Fig. 3) with significant difference when comparedto the other groups (P< 0.02) (Table IV). The Pearson

Fig. 1. Diagnostic arthroscopy 12 months after microfracture. Thedefect is filled with repair tissue, completely integrated and in level

with the surrounding cartilage.

Fig. 2. Diagnostic arthroscopy 3 years after microfracture. Therepair tissue is soft and degenerated with large fissures and loose

cartilage fibers.

coefficient of correlation between defect filling and ICRS-score was 0.84 and significant at the 0.01 level. The clinicalscores 36 months after microfracture correlated best withthe overall MRI-score and the MRI parameter ‘‘defect filling’’(Table V).

Discussion

Recent studies have revealed promising results aftermicrofracture. In this context Steadman et al. described im-proved function in 95% of their study population witha mean follow-up of 11.3 years1. There was significant im-provement in patients’ ability to do activities of daily living,strenuous work and sports from preoperative scores toscores obtained during follow-up after surgery. These re-sults were astonishing since microfracture (as well as Pridiedrilling and abrasion arthroplasty) belongs to the classicalmarrow stimulation techniques that involve surgical accessto the bone marrow spaces underlying regions of damagedarticular cartilage and thus promote resurfacing with

Fig. 3. Sagittal T1-weighted MRI (repetition time, 567 ms; echotime, 15 ms) of the right knee of a 32-year-old woman 3 years aftermicrofracture; the defect is completely filled and the signal of the re-pair tissue is identical to the adjacent articular cartilage. The former

lesion is marked with two arrows.

Fig. 4. Sagittal T1-weighted MRI (repetition time, 572 ms; echotime, 15 ms) of the right knee of a 48-year-old man 3 years aftermicrofracture; on the medial femoral condyle is a full-thickness car-tilage defect with no more repair tissue and a subchondral edema.

1123Osteoarthritis and Cartilage Vol. 14, No. 11

Table IVMRI 36 months after surgery: the best defect filling could be detected in group 1 with significant difference when compared to the other groups( P< 0.02). The Pearson coefficient of correlation between defect filling and subchondral edema was 0.89 and significant at the 0.01 level

Group Defect filling Subchondral edema Cartilage signal Effusion Overall MRI-score

1 1.69� 0.74 1.50� 0.62 1.63� 0.61 1.31� 0.47 1.78� 0.712 2.56� 0.96 2.31� 0.96 2.31� 0.94 1.88� 0.62 2.56� 0.963 2.64� 0.92 2.55� 0.93 2.55� 0.93 2.36� 1.03 2.64� 0.924 2.55� 0.93 2.45� 1.04 2.36� 0.92 2.18� 0.98 2.64� 1.03

Overall 2.17� 0.95 2.0� 0.93 2.04� 0.88 1.74� 0.81 2.23� 0.94

predominantly fibrocartilaginous repair tissue of inferiorquality. In this context a fibrous type of cartilage tissuehas been found in rabbits3,10,18 and dogs19 that underwentabrasion chondroplasty20. The histological findings wereidentical in laboratory experiments with rabbits that under-went Pridie drilling21 or canines treated with microfracture22.

There are only few studies that examine the durability offibrous cartilage. In rabbits subjected to Pridie drilling, repairtissue endured for a longer period than did in animalstreated by abrasion chondroplasty3. This may be due tothe preserved subchondral bone plate between the drillholes with its biomechanical characteristics23. Shapiro ex-amined the durability of fibrous tissue in 122 New Zealandwhite rabbits over a period of 48 weeks11. He found pro-gressive differentiation of mesenchymal stem cells in full-thickness drilled defects of articular cartilage in fibroblasts,osteoblasts, articular chondroblasts and chondrocyteswithin the first 12 weeks. However between 24 and 48weeks after surgical intervention progressive failure of ap-parently well healed cartilage occurred. Reason for theworse durability was the lack of physical and chemicalbonding between the macromolecular components of therepair tissue and the residual adjacent cartilage allowing mi-cromotion and macromotion between them. Furthermore hefound that the repair cartilage over the re-established prom-inent tidemark was only half as thick as the surroundingoriginal cartilage. Since bone is more than 1000 times stifferthan cartilage the thin and soft repair tissue is predisposedto degenerate over the hard and hypertrophic subchondralbone plate11,24,25 (Fig. 5). These results could be confirmedby a recent sheep study with microfracture of created carti-lage lesions. The initially well formed repair tissue degener-ated over a stiff and hypertrophic subchondral bone plateafter a period of 12 months12. Similar results could be de-tected in a recent prospective cohort study with microfrac-ture of 48 isolated cartilage defects of the femur: MRIevaluation revealed osseous overgrowth in 25% of the pa-tients and persistent gaps between the native and repair tis-sue in 92% of the microfracture repairs26. In the controlMRI’s of our study population we detected an osseous over-growth in 19 of all 70 patients.

These results confirm and may explain the outcome ofour study: good results with high score levels after 6 and18 months deteriorated after the study period of 36 months.

Table VPearson coefficient of correlation between MRI-characteristics andthe clinical scores 36 months after microfracture. The defect fillingand the overall MRI-score correlated best with the clinical symp-

toms. All correlations were significant at the 0.01 level

Score Defectfilling

Subchondraledema

Cartilagesignal

Effusion OverallMRI-score

ICRS 0.84 0.76 0.76 0.71 0.85Cincinnati 0.87 0.77 0.75 0.69 0.85

However also the population of Steadman’s studyincluded patients without improvement: 17% consideredthe pain in their knees unchanged as compared with the pre-operative status and 4% considered the pain worse postop-eratively; 53.5% of all patients reported about mild and14.1% about moderate pain. Swelling decreased from preop-erative scores up to year 3 but did not change from years 3through 7. Seven percent of all knees had to undergo a sec-ond surgery for persistent swelling and discomfort1.

This heterogenity in outcome between worse and excel-lent results over more than 10 years could be explainedby some factors influencing the final result: until now thestructure of the tissue formed after microfracture is still un-clear. Recent histological studies in an equine model haveshown the microfracture repair cartilage to be not only fi-brous but a mixture of fibrocartilage (48%) and hyalinecartilage (20%)27. Knutsen found in a randomized trial com-paring microfracture and ACI more fibrous as hyaline carti-lage in the histological evaluation of biopsy specimenstaken during a second-look arthroscopy after microfracture.Fifty percent of the biopsies in the ACI group showed somehyaline tissue. For the small number of biopsies and theshort-term follow-up of 2 years there was no significant dif-ference in the histological results between both groups9.Peterson found in his ACI-study 80% hyaline cartilage inthe biopsies, analysed by independent pathologists28,29.In the study of Brittberg et al., 11 of 15 patients had ‘‘hya-line-like cartilage’’ after ACI30. Frisbie et al. found 70%type II collagen at 1 year after microfracture, however theaggrecan content was less than ideal27. The described dif-ferences in the structure of the repair tissue may lead to dif-ferent results because of changing biomechanical qualities:the stiffness of hyaline cartilage is 3.07 N , of hyaline-likecartilage 2.77 N and of fibrous cartilage 1.27 N31. Glaserand Putz have shown that the reduced stiffness of

Fig. 5. Intralesional osteophytes as result of osseous growth into thedefect site after microfracture with thinning the overlying repair tissue.

1124 P. C. Kreuz et al.: Microfracture in full-thickness chondral defects

fibrocartilage leads to fissures in the tissue texture and thusto progressive osteoarthritis32.

Besides the quality of the repair tissue some other factorsmay influence the result. An age over 35 years1, atrophicmuscles with insufficient knee stabilization and missingcontinuous passive motion16,18,20,33 turned out to be disad-vantageous. Compared to Steadman’s patients our studypopulation was less sportive and the mean age was higher.Furthermore our patients underwent serial objective obser-vation instead of a subjective evaluation by questionnaires.Finally the ICRS-score is more critical as the Cincinnati-score since the lowest grade within a group determinesthe group grade.

These factors may explain the deterioration of results inour study group compared to Steadman’s study group.

For the disappointing results after microfracturing patello-femoral chondral lesions, we tend to perform alternativetreatment methods in this compartment. In chondral lesionsover 2 cm2 we use for the patella different techniques of ACIand for the trochlea ACI as well or autologous osteochon-dral transplantation33e36. However all resurfacing tech-niques may have success only if the biomechanicalproperties of the joint with tibiofemoral or patella malalign-ment and ligament instabilities are considered and treatedas well33,34.

The worse results of patients over 40 years compared toyounger patients may be explained by the reduced regener-ation capacity of the repair tissue. As stem cells age, theirtelomere length, mitochondrial function and their mitoticand synthetic activity decline, and their responsiveness toanabolic mechanical stimuli and growth factors de-creases37,38. For these reasons chondrocyte senescencecontributes to the age-related increase in the prevalenceof osteoarthritis and decrease in the efficacy of cartilagerepair.

Therefore we have limited the indication for the microfrac-ture procedure in older patients to chondral lesions smallerthan 2 cm2. In larger lesions microfracture might have onlythe potential to relieve pain for some years or to delay theimplantation of an endoprosthesis.

Microfracture is a minimal invasive and cheap methodwith good short-term results especially in young activepatients with small cartilage defects. A deterioration of theresults starts 18 months after surgery and is most evidentin the ICRS-score. The best prognostic factors have youngpatients with defects on the femoral condyles.

Acknowledgments

No support in form of grants, equipment or other items hasoccurred for this project. The authors have received nothingof value.This manuscript does not contain information about medicaldevices.

References

1. Steadman JR, Briggs KK, Rodrigo JJ, Kocher MS,Gill TJ, Rodkey WG. Outcomes of microfracture fortraumatic chondral defects of the knee: average11-year follow-up. Arthroscopy 2003;19:477e84.

2. Steadman JR, Miller BS, Karas SG, Schlegel TF,Briggs KK, Hawkins RJ. The microfracture techniquein the treatment of full-thickness chondral lesions ofthe knee in National Football League players.J Knee Surg 2003;16:83e6.

3. Menche D, Frenkel S, Blair B, Watnik N, Toolan B,Yahhoubian R, et al. A comparison of abrasion arthro-plasty and subchondral drilling in the treatment of full-thickness cartilage lesions in the rabbit. Arthroscopy1996;12:280e6.

4. Pridie KH. A method of resurfacing osteoarthritic kneejoints. J Bone Joint Surg Br 1959;41:618e9.

5. Tippet JW. Articular cartilage drilling and osteotomyin osteoarthritis of the knee. In: McGinty JB,Caspari RB, Jackson RW, Poehling GG, Eds. Opera-tive Arthroscopy. 2nd edn. Philadelphia, New York:Raven Press 1996:411e26.

6. Sledge SL. Microfracture techniques in the treatment ofosteochondral injuries. Clin Sports Med 2001;20:365e77.

7. Hunziker EB. Articular cartilage repair: basic scienceand clinical progress. A review of the current statusand prospects. Osteoarthritis Cartilage 2002;10:432e63.

8. Knutsen G, Engebretsen L, Ludvigsen TC, Drogset JO,Grontvedt T, Solheim E, et al. Autologous chondrocyteimplantation versus microfracture e a prospectiverandomised Norwegian multicenter-trial. ICRS-Symposium June 2002, Toronto, Canada Orthopae-dics Today Intern 2002;5:25.

9. Knutsen G, Engebretsen L, Ludvigsen TC, Drogset JO,Grontvedt T, Solheim E, et al. Autologous chondrocyteimplantation compared with microfracture in the knee.A randomized trial. J Bone Joint Surg Am 2004;86(3):455e64.

10. Furukawa T, Eyre DR, Koide S, Glimcher MJ. Biochem-ical studies on repair cartilage resurfacing experimen-tal defects in the rabbit knee. J Bone Joint Surg 1980;62A:79e89.

11. Shapiro F, Koide S, Glimcher MJ. Cell origin anddifferentiation in the repair of full-thickness defects ofarticular cartilage. J Bone Joint Surg 1993;75A:532e53.

12. Nehrer S, Dorotka R, Bindreiter U, Losert U, WindbergerU, Macfelda K, et al. Microfracture in the treatment ofchondral defects in a sheep model. Abstract book ofthe 5th symposium of the International Cartilage RepairSociety in Gent, Belgium, 2004; p. 42.

13. Brittberg M, Aglietti P, Gambardella R, Hangody L,Hauselmann HJ, Jakob RP, et al. The ICRS clinicalcartilage injury evaluation system e 2000. 3rd ICRSmeeting in Goteborg, Sweden, 2000.

14. Noyes FR, McGinniss GH, Grood ES. The variablefunctional disability of the anterior cruciate ligament-deficient knee. Orthop Clin North Am 1985;16:47e67.

15. Henderson IJP, Tuy B, Connell D, Oakes B,Hettwer WH. Prospective clinical study of autologouschondrocyte implantation and correlation with MRI atthree and 12 months. J Bone Joint Surg Br 2003;85B:1060e6.

16. Rodrigo JJ, Steadman JR, Silliman JF, Fulstone HA.Improvement in full-thickness chondral defect healingin the human knee after debridement and microfrac-ture using continuous passive motion. Am J KneeSurg 1994;7:109e16.

17. Hosmer DW, Lemeshow S. Applied Logistic Regres-sion. New York: John Wiley 1989. pp. 106e43.

18. Kim HWK, Moran ME, Salter RB. The potential forregeneration of articular cartilage in defects createdby chondral shaving and subchondral abrasion e anexperimental investigation in rabbits. J Bone JointSurg 1991;73A:1301e15.

1125Osteoarthritis and Cartilage Vol. 14, No. 11

19. Altman R, Kates J, Chun L, Dean D, Eyre D. Prelimi-nary observations of chondral abrasion in a caninemodel. Ann Rheum Dis 1992;51:1056e62.

20. Johnson L. Arthroscopic abrasion arthroplasty historicaland pathologic perspective: present status. Arthros-copy 1986;2:54e69.

21. Hice G, Freedman D, Lemont H, Khoury S. Scanningand light microscopic study of irrigated and nonirri-gated joints following burr surgery performed througha small incision. J Foot Surg 1990;29:337e44.

22. Breinan H, Martin S, Spector M. Healing of canine artic-ular cartilage defects treated with microfracture,a type-II collagen matrix, or cultured autologous chon-drocytes. J Orthop Res 2000;18:781e9.

23. Rand JA. Role of arthroscopy in osteoarthritis of theknee. Arthroscopy 1991;7:358e63.

24. Wong M, Hunziker EB. Articular cartilage biology andbiomechanics. In: Erggelet C, Steinwachs M, Eds.Gelenkknorpeldefekte. Darmstadt: Steinkopf Press2001:15e27.

25. Simon W. Scale effects in animal joints. I. Articular car-tilage thickness and compressive stress. ArthritisRheum 1970;13:244e56.

26. Mithoefer K, Williams RJ III, Warren RF, Potter HG,Spock CR, Jones EC, et al. The microfracture tech-nique for the treatment of articular cartilage lesionsin the knee. J Bone Joint Surg Am 2005;87A(9):1911e20.

27. Frisbie DD, Trotter GW, Powers BE. Arthroscopicsubchondral bone plate microfracture technique aug-ments healing of large chondral defects in the radialcarpal bone and medial femoral condyle of horses.Vet Surg 1999;28:242e55.

28. Peterson L. Autologous chondrocyte transplantation elong term results. In: Hendrich C, Noth U, Eulert J,Eds. Cartilage Surgery and Future Perspectives. Hei-delberg, New York: Springer Press 2003:19e27.

29. Peterson L, Minas T, Brittberg M, Nilsson A, Sjogren-Jansson E, Lindahl A. Two-to-9-year outcome afterautologous chondrocyte transplantation of the knee.Clin Orthop 2000;374:212e34.

30. Brittberg M, Lindahl A, Nilsson A, Ohlsson C,Isaksson O, Peterson L. Treatment of deep cartilagedefects in the knee with autologous chondrocyte trans-plantation. N Engl J Med 1994;331:889e95.

31. Peterson L, Brittberg M, Kiviranta I, LundgrenAkerlund E, Lindahl A. Autologous chondrocyte trans-plantation. Biomechanics and long-term durability. AmJ Sports Med 2002;30:2e12.

32. Glaser C, Putz R. Functional anatomy of articularcartilage under compressive loading. Quantitativeaspects of global, local and zonal reactions of the col-lagenous network with respect to the surface integrity.Osteoarthritis Cartilage 2002;10:83e99.

33. Steinwachs MR, Kreuz PC. Combinations of differentcartilage resurfacing techniques. Z Orthop IhreGrenzbeg 2003;141:625e8.

34. Steinwachs MR, Kreuz PC. Clinical results of ACTusing a collagen membrane. In: Hendrich C, Noth U,Eulert C, Eds. Cartilage Surgery and Future Perspec-tives. Berlin, Heidelberg, New York: Springer Press2003:37e47.

35. Kreuz PC, Steinwachs M, Erggelet C, Lahm A, Henle P,Niemeyer P. Mosaicplasty with autogenous talar auto-graft for osteochondral lesions of the talus after failedprimary arthroscopic management. A prospectivestudy with a 4-years follow-up. Am J Sports Med2005;33(12):1e9.

36. Kreuz PC, Steinwachs MR, Edlich M, Kaiser T, Mika J,Lahm A, et al. The anterior approach for the treatmentof posterior osteochondral lesions of the talus: com-parison of different surgical techniques. Arch OrthopTrauma Surg 2006;126:241e6.

37. Maneiro E, Martin MA, de Andres MC, Lopez-Armada MJ, Fernandez-Sueiro JL, del Hoyo P, et al.Mitochondrial respiratory activity is altered in osteoar-thritic human articular chondrocytes. Arthritis Rheum2003;48(3):700e8.

38. Martin JA, Buckwalter JA. The role of chondrocyte se-nescence in the pathogenesis of osteoarthritis and inlimiting cartilage repair. J Bone Joint Surg Am 2003;85-A(Suppl 2):106e10.