AJSM PreView, published on March 10, 2015 as doi:10.1177/0363546515572579

Intact Vitality and Increased Accumulation of Nonmineralized Bone Matrix in Biopsy Specimens of Juvenile Dissecans

A Histological Analysis

Matthias Krause,*yz MD, Daniel Lehmann,§ Michael Amling,z MD, Tim Rolvien,z Karl-Heinz Frosch,y MD, Klaus Pu¨ schel,|| MD, Klaus Bohndorf,{ MD, and Norbert M. Meenen,y§ MD Investigation performed at the Department of Pediatric Sports Medicine, Altona Childrens Hospital and the Department of Osteology and Biomechanics, University Medical Center Hamburg-Eppendorf, Hamburg, Germany

Background: Although commonly proposed to be the starting point of juvenile osteochondritis dissecans (JOCD), avascular os- teonecrosis (AVN) has been an inconsistent finding in histological studies. Analysis of early-stage lesions is required to elucidate the origins of OCD and justify proper treatment. Purpose: To analyze histological sections of JOCD lesions with special emphasis on bone vitality. Study Design: Cross-sectional study; Level of evidence, 3. Methods: Of 64 patients with 74 JOCD lesions (20 females, mean age, 11.4 years; 44 males, mean age, 12.7 years), 34 required surgery because of lesion instability or failed nonoperative treatment. From 9 patients, 11 histological specimens were obtained. Lesions were classified according to the International Cartilage Repair Society (ICRS). Two additional histological control sections were harvested from children without JOCD manifestation. Undecalcified histological sections were histomorphometrically ana- lyzed. To analyze the skeletal health of the patients, biochemical analyses with special emphasis on bone metabolism were performed. Results: Histologically, no osteonecrosis was visible in any of the cases. Osteocyte distribution was similar among OCD lesions and controls. ICRS OCD I lesions (n = 6) showed no intralesional separation. In ICRS OCD II and III lesions (n = 5), there was a subchondral fracture concomitant with histological characteristics of active repair mechanism (increased bone formation: oste- oid volume P = .008, number P = .046; resorption: number P = .005; and tissue fibrosis compared with con- trols). Instead, in ICRS OCD I lesions, subchondral osteoid volume (P = .010) and osteoblast number (P = .046) were significantly increased compared with controls; however, no active repair mechanisms (no increased or fibrous tissue) were detected, suggesting a focal lack of mineralization. Fifty-seven of 64 patients (89.1%) showed a vitamin D deficiency. The median vitamin D serum level of the patients with ICRS OCD I lesions was 13.6 mg/L. Conclusion: In the present study, osteonecrosis was not found in histological specimens of JOCD. As a secondary finding, focal accumulations of nonmineralized bone matrix indicating a lack of mineralization in ICRS OCD I lesions were revealed. This finding correlated with a low level of vitamin D in the affected children. Keywords: osteochondritis dissecans; bone vitality; histology; vitamin D deficiency

Osteochondritis dissecans (OCD) of the knee is an encoun- considered to constitute an acquired, focal, potentially tered cause of pain and is a functional knee disorder among reversible lesion of the subchondral bone. The lesion may children, adolescents, and young adults. Currently, OCD is result in osseous collapse and fragmentation. There is pos- sible involvement of cartilage through tears and delamina- tion. Eventually, lesion instability and dissection/ sequestration of the lesion may be the final result.17 Based The American Journal of Sports Medicine, Vol. XX, No. X DOI: 10.1177/0363546515572579 on physeal maturity, OCD is distinguished in a juvenile Ó 2015 The Author(s) (JOCD) and an adult form, as JOCD lesions have been

1 2 Krause et al The American Journal of Sports Medicine associated with a better prognosis than adult ones.29,56 To exclude patients with a suspected ossification anomaly, However, healing rates up to 50% for nonoperative treat- the JOCD lesion had to present a perilesional high signal ment in stable JOCD lesions are still poor and unpredict- and be located in the anterior and middle third of the fem- able. Hence, a consensus on a general treatment oral condyle.25 Consequently, 27 patients had to be algorithm for JOCD is still missing.12,17,31,57 As for every excluded because an ossification anomaly was finally diag- other disease, the optimal treatment is based on the knowl- nosed. The location of the JOCD was defined in the coronal edge of its origin. However, despite more than 125 years of view (medial or lateral condyles) using the Cahill and mostly clinical research and discussion, a reasonable etio- Berg11 classification and in the sagittal view (anterior, logical concept is still debated.17 Many different theories middle, and posterior third), as previously described.31 To have been discussed, including genetics, ossification abnor- analyze the skeletal health of the patients, biochemical malities, mechanical overloading due to knee malalignment, analyses with a special emphasis on bone metabolism, repetitive or acute trauma, insufficiency fracture, chondro- including serum levels of 25(OH)D, , phosphate, necrosis, or avascular osteonecrosis (AVN).# Although com- and , were performed. Out of the monly proposed, AVN has been an inconsistent finding in remaining 97 patients, another 33 patients had to be histological studies. In a recent meta-analysis regarding excluded due to missing biochemical analyses. None of the histological evaluation of OCD lesions, 7 of 10 studies the patients had other metabolic bone disorders such as found necrotic subchondral bone stock; however, they failed primary hyperparathyroidism, hypocalcaemia, renal insuf- to elucidate whether AVN was the primary lesion or a sec- ficiency, glucocorticoid therapy, or bone cancer. Hence, ondary phenomenon.51 Given the disparity of histological a total of 64 patients with 74 JOCD lesions (20 females, findings, the primary goal of the present study was to ana- mean age, 11.4 6 2.3 years; 44 males, mean age, 12.7 6 lyze histological sections of arthroscopically stable early 1.9 years) were included in the study. Out of 64 patients, stage JOCD lesions with special emphasis on bone vitality. 9 had (14.1%) bilateral JOCD lesions. One of the patients We hypothesized that osteonecrosis is not the starting point had 2 JOCD lesions in the same knee (medial and lateral of JOCD and may therefore not be the predominant finding femur condyle). The local ethics committee approved the in our analysis of early stage JOCD lesions. Furthermore, study protocol of retrospective clinical data analysis and given the frequent appearance of JOCD in highly active biopsy specimen retrieval from patients with OCD mani- patients, repetitive trauma has been considered to be festation (PV3795). a risk factor for JOCD. It is assumed that the subchondral bone structure may be affected, resulting in a stress frac- 10,28 ture. Hence, our secondary goal was to histologically Biochemical Analysis analyze subchondral bone architecture, turnover, and bone metabolism in the affected children. Blood samples available for the cohort, with special empha- sis on 25(OH)D levels, were analyzed using a radioimmuno- assay (DiaSorin) with an interassay coefficient of variation MATERIALS AND METHODS between 8.2% and 11%. The sensitivity of this method was 1.5 mg/L. The 25(OH)D levels were categorized as fol- 20 Study Group lows : sufficient (30 mg/L), insufficient (20-29 mg/L), defi- cient (10-19 mg/L), and severely deficient (\10 mg/L). Between 2009 and 2014, a total of 124 patients with an ini- tial diagnosis of JOCD knee lesions were admitted to our clinic. At initial presentation, patient demographics and Surgical Treatment medical history, including age at surgery, sex, ethnicity, sports activity (none, 1-23 per week, 33 per week), inter- Surgical treatment was indicated in case of JOCD lesion val between diagnosis and surgical treatment, and mag- instability or failure of nonoperative treatment. Lesion netic resonance imaging (MRI), were obtained. MRI instability was defined as the detection of a cartilage images with T1-, PD-, and T2-weighted turbo spin echo breach on MRI3 and/or secondary signs of instability (rim (TSE) sequences in axial, coronal, and sagittal planes of of fluid signal intensity, multiple breaks in the subchon- the knee, which were performed by colleagues from outpa- dral plate, and a second outer rim of low T2-weighted sig- tient clinics, were requested at the first visit in our clinic. nal intensity).27 Out of 64 patients, 34 required surgical intervention because of failed nonoperative treatment. Failure of nonoperative treatment was defined as persis- #References 8, 10, 18, 24, 36, 38, 41, 53, 60. tent pain and functional impairment of the injured knee

*Address correspondence to Matthias Krause, MD, Department of Trauma and Reconstructive Surgery, Asklepios Clinic St. Georg, Lohmu¨ hlenstrasse 5, 20099 Hamburg, Germany (email: [email protected]). yDepartment of Trauma and Reconstructive Surgery, Asklepios Clinic St Georg, Hamburg, Germany. zDepartment of Osteology and Biomechanics, University Medical Center Hamburg-Eppendorf, Hamburg, Germany. §Department of Pediatric Sports Medicine, Altona Childrens Hospital, University Medical Center Hamburg-Eppendorf, Hamburg, Germany. ||Department of Legal Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany. {High Field MR Center, Department of Biochemical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria. The authors declared that they have no conflicts of interest in the authorship and publication of this contribution. Vol. XX, No. X, XXXX Histological Analysis of Juvenile OCD 3

TABLE 1 Baseline Data and Vitamin D Levels

Complete Study Group (N = 64) Males (n = 44) Females (n = 20)

Age, y, mean 6 SD 12.27 6 2.13 12.66 6 1.92 11.4 6 2.35 Vitamin D level, n (%)a \10 mg/L (severely deficient) 7 (10.9) 3 (6.8) 4 (20.0) 10-19 mg/L (deficient) 35 (54.7) 25 (56.9) 10 (50.0) 20-29 mg/L (insufficient) 15 (23.5) 10 (22.7) 5 (25.0) 30 mg/L (sufficient) 7 (10.9) 6 (13.6) 1 (5.0)

aClassified according to Gonzalez-Gross et al.20

(n = 17) or lesion size progression as assessed by MRI after BS; %), osteoid volume (OV/BV; %), osteoid surface at least 6 months restriction of repetitive impact sports (OS/BS; %), osteoid thickness (O.Th; mm), bone volume (n = 17). Arthroscopic findings were evaluated according (BV/TV; %), trabecular thickness (Tb.Th; mm), trabecu- to International Cartilage Repair Society (ICRS) OCD lar number (Tb.N; mm21), and trabecular separation classification.6,59 Twenty-nine patients had an ICRS (Tb.Sp; mm), according to the American Society for OCD grades I and II lesion and required drilling, and 1 Bone and Mineral Research standards using the patient required fixation due to instability. Nine OsteoMeasure histomorphometry system (Osteometrics) patients gave informed consent to obtain a histological connected to a Zeiss microscope.15 Two additional, age- specimen while requiring surgery. In 1 male (patient related histological specimens were obtained during No. 2), a follow-up arthroscopic drilling intervention autopsy from 2 females (both 14 years old) without was necessary because of failed bony consolidation 6 JOCD lesions or any bone disorder. Causes of death months after the first arthroscopic intervention. In 1 were postoperative sepsis and pneumonia due to aplastic female patient (patient No. 8), JOCD lesions of the anemia. Written informed consent to participate in the medial and lateral femur of the same knee were diag- study was obtained from each patient or legal guardian. nosed, and biopsy specimens were taken from each lesion The local ethics committee approved the study protocol. inthesamearthroscopicsession. Arthroscopic retroar- ticular drilling was performed with a 1.4 mm–diameter Kirschner wire for marking the lesion under radio- graphic control followed by overdrilling with a 2.9 mm– Statistics diameter cannulated drill. Additionally, an osteochon- Statistical analysis was carried out using SPSS Statistics v21 dral cylindrical tissue sample was gently obtained retro- (IBM Corp). Data distribution is given as mean 6 standard grade, preferably from the center of the lesion, using deviation and median (range) for continuous variables. The a 14-gauge bone-biopsy needle (Ostycut; Angiomed/ Mann-Whitney U test was used to report group differences Bard). A postoperative MRI revealed intralesional dril- concerning different regions of interest. All tests were 2- ling and correct biopsy sampling in all of the cases. sided, and a P value of \.05 was considered significant.

Histological Analysis RESULTS The cylindrical tissue samples were subjected to an unde- calcified infiltration process and embedded in methyl Characterization of the Complete Study Group methacrylate for 2-dimensional histomorphometry.4,30,33 The sections were cut to a thickness of 4 mm with Patient characteristics are reported in Table 1. Except for 5 a heavy-duty microtom (Carl Zeiss) and stained with mod- patients (7.1%), all participated in repetitive sports activi- ified von Kossa stain, toluidine blue, and Goldner tri- ties. Fifty-one patients (72.9%) participated in organized chrome. The undecalcified histological specimens competitive sport activities at high levels (33 per week) obtained were divided into 2 regions of interest (ROIs): before the onset of symptoms. Soccer was the most common a ‘‘juxta-articular,’’ directly subchondral ROI and a ‘‘periph- sport exercised, followed by athlete trampoline jumping. eral’’ ROI, distant to the articular surface and juxta-articular None of the patients reported an incidental memorable ROI.TheseROIswereevaluatedwithrespecttoviability, injury leading to functional decline and pain in the knee. subchondral bone architecture, and bone turnover. Parame- Thirty-five of 64 patients (54.7%) showed a deficient vita- ters of static histomorphometry were quantified, including min D level. Seven patients (10.9%) had severe vitamin osteocyte number (N.Ot/BPm; mm21), number of empty oste- D deficiency. Overall, 89.1% of patients had a vitamin D ocyte lacunae (N.Empty.Lc/BPm; mm21), osteoblast number level below the recommended 30 mg/L.5,45 There was no (N.Ob/BPm; mm21), osteoblast surface (ObS/BS; %), osteo- significant difference between males and females (P = clast number (N.Oc/BPm; mm21), osteoclast surface (OcS/ .352). 4 Krause et al The American Journal of Sports Medicine

TABLE 2 Demographic Data and Histological Findings of Biopsy Specimensa

25(OH)D ICRS Time to Intralesional AVN, Level, OCD Lesion Arthroscopy, fracture Peripheral/ Major Histological Patient Sex Age, y mg/L Grade Location mo (depth) Juxta-articular Finding

1 M 14 13.6 II Lateral, mid third 1 Yes (1 mm) No/no Increased bone turnover with high bone resorption in the juxta-articular ROI; bone marrow fibrosis; high vascularization 2.1 M 13 13.1 I Medial, mid third 7 No No/no Regularly shaped trabeculae; increased bone turnover (the juxta-articular . peripheral) with detached ; regular bone marrow 2.2 M 14 37.2 I Medial, mid third 13 No No/no Regularly shaped trabeculae; increased bone turnover (the juxta-articular . peripheral) with detached osteoclasts; regular bone marrow 3 M 13 13.2 I Medial, mid third 6 No No/no Regularly shaped trabeculae; increased bone turnover (the juxta-articular . peripheral) with detached osteoclasts; regular bone marrow 4 M 14 12.5 II Lateral, mid third 12 No No/no Increased bone turnover with high bone resorption in the juxta-articular ROI; bone marrow fibrosis; high vascularization 5 M 15 14.9 II Medial, mid third 7 Yes (5 mm) No/no Increased bone turnover with high bone resorption close to intralesional cleft; bone marrow fibrosis; fibrocartilage within cleft; high vascularization 6 M 12 24.3 II Medial, mid third 1 Yes (3 mm) No/no Increased bone turnover with high bone resorption; bone marrow fibrosis; fibrocartilage and premature callus formation within cleft; high vascularization 7 F 13 13.6 I Lateral, mid third 6 No No/no Regularly shaped trabeculae; increased bone turnover (the juxta-articular . peripheral) with detached osteoclasts; regular bone marrow 8.1 F 11 23.8 I Lateral, mid third 10 No No/no Regularly shaped trabeculae; increased bone turnover (the juxta-articular . peripheral) with detached osteoclasts; regular bone marrow 8.2 F 11 27.8 I Medial, mid third 10 No No/no Regularly shaped trabeculae; increased bone turnover (the juxta-articular . peripheral) with detached osteoclasts; regular bone marrow 9 F 12 15.3 III Medial, mid third 48 Yes (6 mm) No/no Increased bone turnover with high bone resorption close to intralesional cleft; bone marrow fibrosis; fibrocartilage within cleft; high vascularization Control 1 F 14 — — — No No/no Regularly shaped trabeculae; regular bone turnover; regular bone marrow; regular mineralization Control 2 F 14 — — — No No/no Regularly shaped trabeculae; regular bone turnover; regular bone marrow; regular mineralization

aAVN, avascular osteonecrosis; F, female; ICRS, International Cartilage Repair Society; M, male; OCD, osteochondritis dissecans; ROI, region of interest.

Characterization of the Patient Group arthroscopically normal without suspicious color change, With Histological Analysis fissures, or softening in the center of the lesion (ICRS OCD I) (Figure 1). The median vitamin D serum level Eleven histological specimens were obtained from 6 males was 13.6 mg/L (range, 13.1-37.2 mg/L) in the patients (mean age, 13.7 6 1.0 years) and 3 females (mean age, 12.0 with ICRS OCD I lesions (n = 4). In another 4 lesions in 6 1.0 years) with open epiphyses (Table 2). The median 4 patients, the cartilage was fissured either partially or time between diagnosis and arthroscopy was 7 months. completely at the edge of the lesion but was stable while In 6 of 11 JOCD lesions, the cartilage surface appeared arthroscopic probing (ICRS OCD II) (Figure 1). In patient Vol. XX, No. X, XXXX Histological Analysis of Juvenile OCD 5

Figure 1. Magnetic resonance imaging (MRI) and arthroscopic findings of juvenile osteochondritis dissecans (JOCD) knee lesions. ICRS OCD I (patient No. 3): Coronal fat-saturated proton density–weighted and sagittal T1-weighted turbo spin echo sequences show a primarily stable subchondral bone lesion. Surgical treatment through retrograde drilling was indicated after 6 months of unsuccessful nonoperative treatment with persistent symptoms. Arthroscopy demonstrated no cartilage lesion. ICRS OCD II grade (patient No. 6): T2-weighted MRI images show a potentially unstable JOCD lesion at initial presentation due to signal change, thin- ning of cartilage, and a high signal line between the lesion and the normal epiphysis. Arthroscopy demonstrated a protruding defect partially surrounded by fissures. However, while probing the defect was stable, indicating drilling without primary fixation.

No. 9, open reduction and internal fixation of the osteo- comparable with controls and ICRS OCD II and III lesions. chondral lesion with resorbable PLLA pins was necessary The number of empty osteocyte lacunae even tended to be (ICRS OCD III). The median vitamin D serum level in lower than in controls (Figure 3). Trabeculae showed a nor- patients with ICRS OCD II and III lesions was 15.1 mg/L mal thickness of 120 mm35 and were regularly shaped but (range, 12.5-24.3 mg/L). Of these 9, only 1 patient, who thinner compared with controls (Tb.Th P = .008) (see had a vitamin D deficiency at his initial visit and was pre- Appendix Figure A1, available online at http://ajsm.sage- scribed vitamin D supplementation, had a sufficient vita- pub.com/supplemental). In addition, trabeculae were cov- min D level at his follow-up visit 6 months later. The ered with significantly more nonmineralized bone matrix most common location of the JOCD lesion was the classic (OV/BV, OS/BS each P = .015, O.Th P = .039) (Figure 4) site at the mid- to posterolateral aspect of the medial fem- and (N.Ob/BPm P = .023) compared with con- oral condyle (Table 2). trols. However, osteoclasts were only sporadically found (Figure 3). Controls Juxta-articular ROI. In the juxta-articular ROI, there was also no osteonecrosis visible in any of the cases (Fig- In the 2 control patients, the overlying cartilage looked nor- ures 2G, 2H, and 2J). Osteocytes were evenly distributed, mal from surface to bottom with respect to viability and and the number of empty osteocyte lacunae was even lower potential degeneration (Figure 2A). Bone marrow consisted than in controls (Figure 3). As an additional finding, we of regular fat cells and regular vascularization. Trabecular found an unexpectedly high content of osteoid accumula- bone stock was viable, as lacunae were filled with osteocytes tion (OV/BV . 2%) (Figure 4) without signs of active repair throughout the complete histological section (Figure 2, B mechanisms (increased resorption or tissue fibrosis), indi- and C). Only sporadic accumulation of nonmineralized cating focal defects of mineralization (Figure 2J).45,59 Oste- bone matrix (osteoid) was detected in the peripheral and oid content increased significantly compared with the the juxta-articular ROI (Figure 2D). At the cartilage-bone peripheral ROI (OV/BV P = .042) (Figures 2H and 2I). Con- junction, remnants of partial ossification in the center of comitantly, osteoblast numbers and size were significantly viable single trabeculae were visible (Figure 2E). increased compared with controls. The overlying hyaline cartilage revealed a regular appearance of the chondro- ICRS OCD I Lesions cytes with respect to size and number.

In all ICRS OCD I lesions (n = 6), no subchondral bony dis- ICRS OCD II and III Lesions ruption, cleft, or separation was found (Figure 2F). Peripheral ROI. In the peripheral ROI, viable trabecu- In 3 of 4 ICRS OCD II lesions and in the 1 ICRS OCD III lae were visible (Figure 2G). Osteocyte number was lesion, a complete or partial disruption of an osseous 6 Krause et al The American Journal of Sports Medicine

Figure 2. (A-E) Undecalcified histological sections of a control and (F-J, patient No. 3; toluidine blue stain) an ICRS OCD I lesion. (B, G) In the peripheral region of interest (ROI), there was viable bone stock with osteocytes (white arrows) and regular bone mar- row (hash sign) in controls and ICRS OCD I lesions (toluidine blue stain). (C, H) In the juxta-articular ROI in controls and ICRS OCD I lesions, trabeculae were viable (white arrows), with remnants of partial ossification (red arrow), regular bone marrow (hash sign), and sporadically occurring osteoclasts (black arrow; toluidine blue stain). ICRS OCD I lesions showed a significantly increased accumulation of osteoid compared with controls (dotted line, toluidine blue stain). (D, I) At lower magnification: osteoid = red ver- sus mineralized bone matrix = black (modified von Kossa stain). (E, J) In both groups, remnants of partial ossification (red dotted line, red arrows) were seen. In addition increased bone formation (white dotted lines) but no increased bone resorption (black arrow) was visible at the viable (white arrows) cartilage-bone junction (toluidine blue stain). fragment was detected (n = 5). Mean disruption depth was lesions (N.Oc/BPm P \ .001, Oc.S/BS P = .001), these 3.75 6 2.22 mm (Table 2). were accompanied by a great number of bone resorbing Peripheral ROI. Similar to controls and ICRS OCD I osteoclasts (Figure 3F). The additional finding of a high lesions, the peripheral ROI consisted of viable trabeculae vascularization, lymphocyte infiltration, and tissue fibrosis with numerous osteocytes and regular bone marrow (Fig- suggested an active repair mechanism.16,59 Although, in 1 ure 3; Figure 4, B and C). There was no evidence of ICRS OCD II lesion, no obvious cleft was detected, but increased cell death indicated by a low number of empty there were comparable signs of increased bone turnover osteocyte lacunae (Figure 3). The cleft between the periph- in the course of an active repair mechanism including eral ROI and the juxta-articular ROI was filled with fibro- increased bone resorption, bone marrow fibrosis, and ele- cartilaginous tissue or fibrous tissue containing numerous vated osteoid indices. We assume that the biopsy specimen blood vessels (Figure 4D). In 1 case, premature callus for- was taken very close to an intralesional cleft. mation could be observed, representing the healing poten- tial of the lesion (Figure 4E). Juxta-articular ROI. In the juxta-articular ROI, again, DISCUSSION no signs of AVN, such as a high number of empty osteo- cytes lacunae or low number of osteocytes, could be Despite several attempts to elucidate the origin and patho- observed. Compared with controls, irregularly formed tra- physiology of OCD, currently no universally accepted con- beculae were covered with more osteoblasts producing cept exists. Different theories are still equivalently osteoid (N.Ob/BPm P = .046; Ob.S/BS P = .032) (Figure proposed to be causative of OCD. Although OCD may 3). However, in contrast to controls (N.Oc/BPm P = .005, most likely have a multifactorial origin, 3 of the most com- Oc.S/BS P = .032) and, most importantly, ICRS OCD I monly discussed causes include genetics, repetitive Vol. XX, No. X, XXXX Histological Analysis of Juvenile OCD 7

Figure 3. Histomorphometric analysis of bone cells with respect to the peripheral and juxta-articular region of interest in controls, ICRS OCD grade I lesions and ICRS OCD grade II and III lesions. Bars represent mean 6 standard deviation of osteocyte number (N.Ot/BPm) and number of empty osteocyte lacunae (N.Empty.Lc/BPm), osteoblast number (N.Ob/BPm), osteoblast surface (Ob.S/BS), osteoclast number (N.Oc/BPm), and osteoclast surface (OcS/BS). Osteoblasts, toluidine blue stain; osteoclasts, Gold- ner trichrome stain; osteocytes, scanning electron microscopy imaging (SEM). *P \ .05. trauma, and ischemia.18 In a recent review analyzing his- cases, and viable bone stock in 1 case. The authors con- tological studies, 7 of 10 publications reported necrosis of cluded that AVN could be the initial incident or secondary the subchondral bone.51 Unfortunately, these studies var- to fracture of the subchondral bone. Subsequently, necrotic ied in the sources of the specimens (detached vs partially bone would be remodeled, and new bone would be gener- detached vs stable OCD lesions), histological methods ated, explaining the simultaneous occurrence of necrotic used, and patients included (JOCD vs adult OCD vs mixed and viable bone in the same biopsy specimen. Our present cohort).51 It is commonly believed that the majority of results of stable juvenile OCD biopsy specimens, reflecting adult OCD lesions represent persistent, uncured juvenile the early stage of OCD, do not support this consideration of OCD lesions.9,17,29 Therefore, and in contrast to previous Uozumi et al.54 The number of empty osteocyte lacunae in histological studies, our histological analysis of subchon- ICRS OCD I lesions, defining osteonecrosis, was even dral bone vitality included juvenile OCD biopsy specimens lower than in skeletally healthy controls. Also, even in only. Our results demonstrate that the subchondral bone more advanced stages of JOCD (ICRS OCD II and III stock of juvenile OCD lesions is viable and shows no signs lesions), there were evenly distributed osteocytes connected of AVN, neither in the peripheral ROI nor in the juxta- by canaliculi, representing viable bone (Figure 3). Our articular ROI in any of the cases. Despite inconsistent results are in accordance with previous findings by Yone- results, reduced blood supply is still under debate to cause tani et al,59 who also demonstrated viable subchondral tra- focal AVN at the lateral aspect of the medial femoral con- beculae in stable JOCD lesions (ICRS OCD I and II). On the dyle, the predilection site of an OCD lesion.51 In an analy- contrary, in 7 histological studies previously reporting AVN, sis of 5 arthroscopically unstable adult OCD lesions with osteonecrosis was either found in loose bodies13,21,39,44 or subchondral bone stock, Uozumi et al54 found necrotic tra- partially detached JOCD lesions.38,54,58 Hence, these lesions beculae in 2 cases, necrotic next to viable trabeculae in 2 primarily represented arthroscopically unstable JOCD 8 Krause et al The American Journal of Sports Medicine

Figure 4. (A) Static histomorphometry and (B-F, patient No. 6) undecalcified histological section of an unstable ICRS OCD III lesion. (A) Osteoid indices with respect to the peripheral and juxta-articular regions of interest in controls, ICRS OCD grade I lesions, and ICRS OCD grade II and III lesions. Bars represent mean 6 SD of osteoid volume (OV/BV), osteoid surface (OS/ BS), and osteoid thickness (O.Th). (B) Overview of ICRS grade III lesion: 1, viable peripheral bone stock; 2, intralesional cleft; 3, dissected fragment adjacent to metaplastic cartilage; 4, regular articular cartilage (Goldner trichrome stain). (C) The viable host bone (white arrows) showed regular bone marrow proximal to the fragment site. Closer to the intralesional cleft, there was increased bone turnover (white dotted line, formation; red dotted line, resorption) in addition to tissue fibrosis (red stars; tolu- idine blue stain). (D) The intralesional cleft was filled with fibrous or fibrocartilage tissue (toluidine blue stain). (E) In another case, the intralesional cleft was filled with immature callus (red dotted lines; toluidine blue stain). (F) The fragmented OCD lesion was viable (white arrows). Increased bone turnover (dotted lines, formation; black arrows, resorption) was present (toluidine blue stain; inset = 4003 magnification). *P \ .05.

lesions, which were naturally either completely or partially the present study in ICRS OCD grade II and III lesions with separated from their subchondral blood supply. In addition, an intralesional cleft. These data are also in line with previ- previous studies demonstrated that the blood supply of the ous findings of histological sections with intralesional sepa- distal femur, especially the lateral aspect of the medial fem- ration or the margins of loose bodies.13,21,38,39,44,54,58,59 Most oral condyle, 1 of the common sites for OCD, was not interestingly, in 1 case with an intralesional separation in impaired.47,48 Early findings by Chiroff and Cooke14 and the present study, immature callus formation could be Koch et al28 even showed viable subchondral bone stock in observed between the 2 fragments. This supported the con- unstable or dissected JOCD lesions. Based on our present cept of an active fracture repair mechanism in ICRS II and results, we agree with the assumption of Edmonds and III lesions and emphasized the healing and reunion poten- Polousky17 that osteonecrosis associated with OCD rather tial of subchondral fractures identified on MRI. seems secondary to lesion detachment than to be the initial Regarding our secondary goal, surprisingly, in all 6 his- event. In addition, the histological evaluation of OCD tological sections of ICRS OCD I lesions, no circumscribed lesions by Chiroff and Cooke14 and Koch et al28 showed healing processes were observed. Instead, regular fat cells high bone formation, resorption, osteoid accumulation, and were seen without fibrocartilage, phagocytes, increased tissue fibrosis, representing classic characteristics of cancel- vascularization, or increased bone resorption.59 In fact, lous fracture repair.15,17,33 These findings were confirmed in osteoclasts were only sporadically seen. Postoperative Vol. XX, No. X, XXXX Histological Analysis of Juvenile OCD 9

MRI follow-up assessments revealed intralesional drilling the medial femoral condyle, the contact of the medial and correct biopsy sampling. However, we observed an intercondylar tubercle against the medial condyle while unexpectedly higher accumulation of osteoid (osteoidosis) the tibia is internally rotated at knee extension has increasing from the peripheral to the juxta-articular ROI been suggested to constitute 1 potential stress raiser.19 compared with controls (Figure 3A). Osteoidosis, repre- At the lateral condyle, modified biomechanics (eg, due to senting a mineralization defect (definition: osteoid volume leg axis deformities, gait abnormalities, meniscal injury, [OV/BV] . 2%45), can be observed in cases of malignancies or discoid lateral meniscus), which were present in 1 such as bone forming tumors, osteoid osteoma, or onco- patient (patient No. 8) with 2 JOCD lesions of the medial genic hypophosphatemic ; systemic diseases and lateral femoral condyle of the same knee, have been like renal osteodystrophy, primary hyperparathyroidism, associated with JOCD occurrence.40 Hence, JOCD is fibrous dysplasia, or Paget disease; in the course of frac- a multifactorial disease of an adolescent knee. In light ture healing, chemical bone demineralization, and meta- of the low serum vitamin D levels of our patients, we bolic diseases like malabsorption, fluorosis, malnutrition, hypothesize that our histological findings indicating or vitamin D deficiency.** There was no evidence for any a focal lack of mineralization in ICRS OCD I lesions could, of these diseases or conditions in our patients. However, at least in part, be explained by vitamin D deficiency. we found low serum vitamin D levels in all but 1 patient Future interventional prospective studies are needed to and concluded that the focal lack of mineralization in elucidate a potential association. ICRS OCD grade I lesions could, at least in part, be This study had several limitations. First, the most explained by vitamin D deficiency. This conclusion may important limitation is the limited number of histological be supported by a case study linking vitamin D deficiency sections of JOCD and controls. However, to the best of with OCD manifestation in a patient with rickets and our knowledge, this is the first study to analyze subchon- bilateral OCD of the knee.46 A vitamin D level .30 mg/L dral bone stock of early stage, stable ICRS OCD grade I is considered a prerequisite for skeletal health and proper and II lesions compared with ICRS OCD III lesions, and fracture healing.42 In the present study, the median vita- nonaffected controls. A higher number of controls would min D serum level in the patients who were biopsied was have been desirable, but the number was limited to autop- 14.9 mg/L. In our cohort of 64 patients with a JOCD lesion, sies due to ethical reasons. Unfortunately, the vitamin D 57 (89.1%) showed a vitamin D level below recommenda- level of the controls could not be evaluated. Second, only tions of 30 mg/L.20,45 Seven patients (10.9%) even showed 1 biopsy specimen of each JOCD lesion was obtained. a severe vitamin D deficiency to a nondetectable vitamin One biopsy specimen might not represent the complete his- D level. Yet, vitamin D deficiency is not a unique feature tological spectrum of the entire JOCD lesion.51 Third, the in patients requiring orthopaedic treatment and is a com- gold standard to assess mineralization defects is the histo- mon finding among European adolescents. The numbers logical analysis of iliac crest biopsy specimens, which was range between 81% and 87%.20,22,55 Hence, vitamin D defi- not part of the present study. However, laboratory meas- ciency is not the sole explanation for OCD occurrence. urements revealed vitamin D deficiency in all but 1 patient However, given our present findings of a potential sub- with biopsy sampling. Fourth, in cases of osteoid accumu- chondral fracture repair in ICRS OCD II and III lesions, lation, the method of choice to differentiate osteoidosis it is interesting that it could be shown that callus forma- from either lack of mineralization or increased bone forma- tions in a fracture site consumed high rates of vitamin D tion is labeling before taking a biopsy specimen. Unfortu- metabolites, thus highlighting its importance for proper nately, because of limited approval from the ethics mineralization in bone union.26 In addition, osteoidosis committee, labeling was not part of the study protocol. impairs the biomechanical competence of bone, and repet- However, increased osteoid accumulation compared with itive trauma may lead to fracture.7,23 In this context, vita- controls, only sporadically seen osteoclasts, and vitamin min D deficiency has been identified as an important D deficiency as a laboratory finding seen in our ICRS predisposing risk factor for insufficiency fractures.5,37,49 OCD I patients suggested a lack of mineralization as a rea- In a randomized, double-blind, placebo-controlled study son for osteoidosis. including 5201 young military recruits, sufficient vitamin D treatment reduced the incidence of insufficiency frac- 37 tures by 20%. Given that many patients with JOCD par- CONCLUSION ticipate in high-level sports, repetitive trauma has been accepted to pose 1 of the most likely predisposing factors In the present study, we were able to confirm our hypothesis 1 for JOCD occurrence. Some authors even considered and demonstrated that the subchondral bone of ICRS OCD 10,29 JOCD to constitute an insufficiency fracture. In the I, II, and III lesions was viable. We conclude that ischemia present study group, 92.9% of the patients regularly per- leading to avascular osteonecrosis is not a predominant eti- formed repetitive impact sports like soccer or trampoline ological feature of OCD. As a secondary finding, histological jumping. Of the patients, 72.9% of the patients practiced analysis revealed osteoidosis in early stage JOCD lesions. sports at high levels before first clinical symptoms. At We speculate that this lack of mineralization makes this bone segment more prone to mechanical stress, which **References 2, 4, 16, 32, 34, 43, 45, 50, 52. may result in subchondral insufficiency fractures. 10 Krause et al The American Journal of Sports Medicine

Summarizing, we hypothesize that JOCD of the knee is 19. Fairbanks H. Osteochondritis dissecans. Br J Surg. 1933;21:67-82. a multifactorial burden, and osteonecrosis only constitutes 20. Gonzalez-Gross M, Valtuena J, Breidenassel C, et al. Vitamin D sta- a secondary finding, for example, due to a subchondral tus among adolescents in Europe: the Healthy Lifestyle in Europe by Nutrition in Adolescence study. Br J Nutr. 2012;107(5):755-764. fracture and consecutive detachment. We additionally 21. Green WT, Banks HH. Osteochondritis dissecans in children. J Bone hypothesize that vitamin D deficiency supports the devel- Joint Surg Am. 1953;35(1):26-47. opment of OCD lesions. Vitamin D deficiency, a recognized 22. Hintzpeter B, Scheidt-Nave C, Muller MJ, Schenk L, Mensink GB. risk factor for insufficiency fracture, was a common finding Higher prevalence of vitamin D deficiency is associated with immi- among our group of patients with JOCD. Vitamin D sup- grant background among children and adolescents in Germany. J plementation is proposed as a potential additional thera- Nutr. 2008;138(8):1482-1490. 23. Holick MF. Resurrection of vitamin D deficiency and rickets. J Clin peutic approach to classic nonoperative treatment in Invest. 2006;116(8):2062-2072. patients with stable JOCD and vitamin D deficiency. 24. Jacobi M, Wahl P, Bouaicha S, Jakob RP, Gautier E. Association Future prospective studies are needed to evaluate the effi- between mechanical axis of the leg and osteochondritis dissecans ciency of vitamin D supplementation on JOCD healing. of the knee: radiographic study on 103 knees. Am J Sports Med. 2010;38(7):1425-1428. 25. Jans LB, Jaremko JL, Ditchfield M, Huysse WC, Verstraete KL. MRI REFERENCES differentiates femoral condylar ossification evolution from osteo- dissecans. A new sign. Eur Radiol. 2011;21(6):1170-1179. 1. Aichroth P. Osteochondritis dissecans of the knee. A clinical survey. 26. Jingushi S, Iwaki A, Higuchi O, et al. Serum 1alpha,25-dihydroxyvita- J Bone Joint Surg Br. 1971;53(3):440-447. min D3 accumulates into the fracture callus during rat femoral frac- 2. Amling M, Delling G. [Biology of osteoclasts; their role in bone metas- ture healing. Endocrinology. 1998;139(4):1467-1473. tases]. Orthopade. 1998;27(4):214-223. 27. Kijowski R, Blankenbaker DG, Shinki K, Fine JP, Graf BK, De Smet 3. Bohndorf K. Osteochondritis () dissecans: a review AA. Juvenile versus adult osteochondritis dissecans of the knee: and new MRI classification. Eur Radiol. 1998;8(1):103-112. appropriate MR imaging criteria for instability. Radiology. 2008; 4. Breer S, Krause M, Busse B, et al. Analysis of retrieved hip resurfacing 248(2):571-578. arthroplasties reveals the interrelationship between interface hyperos- 28. Koch S, Kampen WU, Laprell H. Cartilage and bone morphology in teoidosis and demineralization of viable bone trabeculae. J Orthop osteochondritis dissecans. Knee Surg Sports Traumatol Arthrosc. Res. 2012;30(7):1155-1161. 1997;5(1):42-45. 5. Breer S, Krause M, Marshall RP, Oheim R, Amling M, Barvencik F. 29. Kocher MS, Tucker R, Ganley TJ, Flynn JM. Management of osteo- Stress fractures in elderly patients. Int Orthop. 2012;36(12):2581- chondritis dissecans of the knee: current concepts review. Am J 2587. Sports Med. 2006;34(7):1181-1191. 6. Brittberg M, Winalski CS. Evaluation of cartilage injuries and repair. 30. Krause M, Breer S, Hahn M, et al. Cementation and interface analysis J Bone Joint Surg Am. 2003;85(suppl 2):58-69. of early failure cases after hip-resurfacing arthroplasty. Int Orthop. 7. Busse B, Bale HA, Zimmermann EA, et al. Vitamin D deficiency indu- 2012;36(7):1333-1340. ces early signs of aging in human bone, increasing the risk of frac- 31. Krause M, Hapfelmeier A, Moller M, Amling M, Bohndorf K, Meenen ture. Sci Transl Med. 2013;5(193):193ra188. NM. Healing predictors of stable juvenile osteochondritis dissecans 8. Caffey J, Madell SH, Royer C, Morales P. Ossification of the distal knee lesions after 6 and 12 months of nonoperative treatment. Am femoral epiphysis. J Bone Joint Surg Am. 1958;40(3):647-654. J Sports Med. 2013;41(10):2384-2391. 9. Cahill BR. Osteochondritis dissecans of the knee: treatment of juve- 32. Krause M, Keller J, Beil B, et al. Calcium gluconate supplementation nile and adult forms. J Am Acad Orthop Surg. 1995;3(4):237-247. is effective to balance calcium in patients with gastrec- 10. Cahill BR, Ahten SM. The three critical components in the conserva- tomy [published online November 13, 2014]. Osteoporos Int. tive treatment of juvenile osteochondritis dissecans (JOCD). Physi- doi:10.1007/s00198-014-2965-1. cian, parent, and child. Clin Sports Med. 2001;20(2):287-298. 33. Krause M, Museyko O, Breer S, et al. Accuracy of trabecular struc- 11. Cahill BR, Berg BC. 99m-Technetium phosphate compound joint ture by HR-pQCT compared to gold standard muCT in the radius scintigraphy in the management of juvenile osteochondritis disse- and tibia of patients with and long-term bisphospho- cans of the femoral condyles. Am J Sports Med. 1983;11(5):329-335. nate therapy. Osteoporos Int. 2014;25(5):1595-1606. 12. Cahill BR, Phillips MR, Navarro R. The results of conservative man- 34. Krause M, Oheim R, Catala-Lehnen P, et al. Metaphyseal bone forma- agement of juvenile osteochondritis dissecans using joint scintigra- tion induced by a new injectable beta-TCP-based bone substitute: phy. A prospective study. Am J Sports Med. 1989;17(5):601-605. a controlled study in rabbits. JBiomaterAppl. 2014;28(6):859-868. 13. Campbell CJ, Ranawat CS. Osteochondritis dissecans: the question 35. Krause M, Rupprecht M, Mumme M, Puschel K, Amling M, Barvencik of etiology. J Trauma. 1966;6(2):201-221. F. Bone microarchitecture of the talus changes with aging. Clin 14. Chiroff RT, Cooke CP III. Osteochondritis dissecans: a histologic and Orthop Relat Res. 2013;471(11):3663-3671. microradiographic analysis of surgically excised lesions. J Trauma. 36. Laor T, Zbojniewicz AM, Eismann EA, Wall EJ. Juvenile osteochon- 1975;15(8):689-696. dritis dissecans: is it a growth disturbance of the secondary physis 15. Dempster DW, Compston JE, Drezner MK, et al. Standardized of the epiphysis? AJR Am J Roentgenol. 2012;199(5):1121-1128. nomenclature, symbols, and units for bone histomorphometry: 37. Lappe J, Cullen D, Haynatzki G, Recker R, Ahlf R, Thompson K. Cal- a 2012 update of the report of the ASBMR Histomorphometry cium and vitamin D supplementation decreases incidence of stress Nomenclature Committee. J Bone Miner Res. 2013;28(1):2-17. fractures in female navy recruits. JBoneMinerRes. 2008;23(5):741-749. 16. Diamond TH, Clark WA, Kumar SV. Histomorphometric analysis of 38. Linden B, Telhag H. Osteochondritis dissecans. A histologic and fracture healing cascade in acute osteoporotic vertebral body frac- autoradiographic study in man. Acta Orthop Scand. 1977;48(6): tures. Bone. 2007;40(3):775-780. 682-686. 17. Edmonds EW, Polousky J. A review of knowledge in osteochondritis 39. Milgram JW. Radiological and pathological manifestations of osteo- dissecans: 123 years of minimal evolution from Konig to the ROCK chondritis dissecans of the distal femur. A study of 50 cases. Radiol- study group. Clin Orthop Relat Res. 2013;471(4):1118-1126. ogy. 1978;126(2):305-311. 18. Eismann EA, Pettit RJ, Wall EJ, Myer GD. Management strategies for 40. Mizuta H, Nakamura E, Otsuka Y, Kudo S, Takagi K. Osteochondritis osteochondritis dissecans of the knee in the skeletally immature ath- dissecans of the lateral femoral condyle following total resection of lete. J Orthop Sports Phys Ther. 2014;44(9):665-679. the discoid lateral meniscus. Arthroscopy. 2001;17(6):608-612. Vol. XX, No. X, XXXX Histological Analysis of Juvenile OCD 11

41. Mubarak SJ, Carroll NC. Familial osteochondritis dissecans of the 51. Shea KG, Jacobs JC Jr, Carey JL, Anderson AF, Oxford JT. Osteochon- knee. Clin Orthop Relat Res. 1979;140:131-136. dritis dissecans knee histology studies have variable findings and theo- 42. Palermo NE, Holick MF. Vitamin D, bone health, and other health ben- ries of etiology. Clin Orthop Relat Res. 2013;471(4):1127-1136. efits in pediatric patients. J Pediatr Rehabil Med. 2014;7(2):179-192. 52. Simon MJ, Beil FT, Ruther W, et al. High fluoride and low calcium lev- 43. Park EA. Observations on the pathology of rickets with particular ref- els in drinking water is associated with low bone mass, reduced bone erence to the changes at the cartilage-shaft junctions of the growing quality and fragility fractures in sheep. Osteoporos Int. 2014;25(7): : Harvey Lecture, February 16, 1939. Bull N Y Acad Med. 1891-1903. 1939;15(8):495-543. 53. Smillie IS. Treatment of osteochondritis dissecans. J Bone Joint Surg 44. Portigliatti Barbos M, Brach del Prever E, Borroni L, Salvadori L, Bat- Br. 1957;39(2):248-260. tiston B. Osteochondritis dissecans of the femoral condyles. A histo- 54. Uozumi H, Sugita T, Aizawa T, Takahashi A, Ohnuma M, Itoi E. His- logical study with pre-operative fluorescent bone labelling and tologic findings and possible causes of osteochondritis dissecans of microradiography. Ital J Orthop Traumatol. 1985;11(2):207-213. the knee. Am J Sports Med. 2009;37(10):2003-2008. 45. Priemel M, von Domarus C, Klatte TO, et al. Bone mineralization 55. Vierucci F, Del Pistoia M, Fanos M, Erba P, Saggese G. Prevalence of defects and vitamin D deficiency: histomorphometric analysis of iliac hypovitaminosis D and predictors of vitamin D status in Italian healthy crest bone biopsies and circulating 25-hydroxyvitamin D in 675 adolescents. Ital J Pediatr. 2014;40:54. patients. J Bone Miner Res. 2010;25(2):305-312. 56. Wall E, Von Stein D. Juvenile osteochondritis dissecans. Orthop Clin 46. Quan AW, Beall DP, Berry ER, Ly JQ, Sweet CF, Fish JR. A case of North Am. 2003;34(3):341-353. osteochondritis dissecans in rickets. Emerg Radiol. 2005;11(4):219-221. 57. Wall EJ, Vourazeris J, Myer GD, et al. The healing potential of stable 47. Reddy AS, Frederick RW. Evaluation of the intraosseous and extra- juvenile osteochondritis dissecans knee lesions. J Bone Joint Surg osseous blood supply to the distal femoral condyles. Am J Sports Am. 2008;90(12):2655-2664. Med. 1998;26(3):415-419. 58. Yonetani Y, Matsuo T, Nakamura N, et al. Fixation of detached 48. Rogers WM, Gladstone H. Vascular foramina and arterial supply of osteochondritis dissecans lesions with bioabsorbable pins: clinical the distal end of the femur. J Bone Joint Surg Am. 1950;32(4):867- and histologic evaluation. Arthroscopy. 2010;26(6):782-789. 874. 59. Yonetani Y, Nakamura N, Natsuume T, Shiozaki Y, Tanaka Y, Horibe 49. Ruohola JP, Laaksi I, Ylikomi T, et al. Association between serum S. Histological evaluation of juvenile osteochondritis dissecans of the 25(OH)D concentrations and bone stress fractures in Finnish young knee: a case series. Knee Surg Sports Traumatol Arthrosc. men. J Bone Miner Res. 2006;21(9):1483-1488. 2010;18(6):723-730. 50. Seitz S, Priemel M, Zustin J, et al. Paget’s disease of bone: histologic 60. Ytrehus B, Carlson CS, Ekman S. Etiology and pathogenesis of analysis of 754 patients. J Bone Miner Res. 2009;24(1):62-69. osteochondrosis. Vet Pathol. 2007;44(4):429-448.

For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav