Pe d i a t r i c I m a g i n g • O r i g i n a l R e s e a r c h Bedoya et al. MRI of Proximal Femur in Children

Downloaded from www.ajronline.org by University of Utah on 09/27/14 from IP address 155.97.178.73. Copyright ARRS. For personal use only; all rights reserved

Pediatric Imaging Original Research

Dynamic Gadolinium-Enhanced MRI of the Proximal Femur: Preliminary Experience in Healthy Children Maria A. Bedoya1 Camilo Jaimes1,2 D’mitry Khrichenko1 Jorge Delgado1,3 Bernard J. Dardzinski1,4 Diego Jaramillo1,4 Bedoya MA, Jaimes C, Khrichenko D, Delgado J, Dardzinski BJ, Jaramillo D

Keywords: femur, growth plate, MRI, perfusion imaging, periosteum DOI:10.2214/AJR.13.12341 Received December 4, 2013; accepted after revision January 19, 2014. 1

Department of Radiology, The Children’s Hospital of Philadelphia, 34th St and Civic Center Blvd, Philadelphia, PA 19104. Address correspondence to D. Jaramillo ([email protected]). 2

Present address: Department of Radiology, Massachusetts General Hospital, Harvard Medical School, Boston, MA.

3

Present address: Medical School, CES University, Medellin, Colombia.

4 Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA.

WEB This is a web exclusive article. AJR 2014; 203:W440–W446 0361–803X/14/2034–W440 © American Roentgen Ray Society

W440

OBJECTIVE. The purpose of this study is to use dynamic contrast-enhanced MRI to evaluate the perfusion characteristics of the proximal femur in the growing skeleton. MATERIALS AND METHODS. We evaluated 159 subjects (mean age, 5.67 years) who underwent a well-controlled protocol of contrast-enhanced MRI of the abdomen and hips. Perfusion and permeability parameters (enhancement ratio peak, AUC, time to peak, and rate of extraction) for six regions of the proximal femur were calculated. RESULTS. A decrease with age was found for all contrast kinetics parameters in all regions (p < 0.001). Perfusion parameters differed between the regions (p < 0.001). The highest perfusion and permeability parameters were found in the metaphyseal spongiosa, metaphyseal marrow, and periosteum. The metaphyseal spongiosa had a highly vascular pattern of enhancement and showed the highest enhancement ratio peak, AUC, and rate of extraction and the lowest time to peak. The metaphyseal marrow showed a vascular pattern of enhancement with a lower peak compared with the metaphyseal spongiosa. The periosteum showed prompt nonvascular contrast enhancement that reached a plateau that remained elevated. CONCLUSION. The highest enhancement was seen in areas involved with growth: the metaphyseal spongiosa, which is related to endochondral ossification, and the periosteal cambium, which is related to membranous ossification. The enhancement characteristics are radically different: in the spongiosa; enhancement is brisk and declines, with a vascular pattern, whereas contrast uptake increases with time in the periosteum. Recognition of normal enhancement patterns of the proximal femur is important for distinguishing normal development from pathologic processes.

D

ynamic contrast-enhanced MRI (DCE-MRI) studies are useful to evaluate differences in tissue perfusion. In the growing bones of piglets, DCE-MRI studies have shown that there is greater enhancement of the hematopoietic marrow than the fatty marrow, that epiphyseal and physeal cartilage are almost nonenhancing outside of the vascular canals, and that there is intense enhancement in the metaphyseal spongiosa [1]. The data in children are more limited because most studies have been obtained in the hips of children with epiphyseal abnormalities, primarily osteonecrosis [2–4]. Human studies also show a layer deep in relation to the fibrous periosteum that enhances during childhood [5, 6]. This is a normal component of the periosteum and should not be confused with subperiosteal disease. At our institution, we have performed dynamic contrast-enhanced MR urography in children of all ages for various indications. The

examinations include a functional study of renal perfusion and function that is done by analyzing the enhancement of the kidneys at regular intervals. The imaged anatomy includes the pelvis and hips. In this population of children, mostly without skeletal abnormality, the proximal femurs have been imaged after a well-controlled protocol of contrast enhancement. This provided a unique opportunity to evaluate perfusion of the structures of the growing skeleton. The purpose of this study is to use DCEMRI to evaluate the perfusion characteristics of the various structures of the epiphysis, physis, and metaphysis of the growing skeleton and the changes that occur with age. Materials and Methods Participants This study was approved by the institutional review board of our hospital and was performed in compliance with HIPAA. We conducted a retrospective search to identify all DCE-MRI of the pelvis and

AJR:203, October 2014

Downloaded from www.ajronline.org by University of Utah on 09/27/14 from IP address 155.97.178.73. Copyright ARRS. For personal use only; all rights reserved

MRI of Proximal Femur in Children abdomen performed at our institution during a 2.5year period. This search yielded 165 children who had undergone MR urography. Inclusion criteria were availability of DCE-MRI acquired for at least 10 minutes without interruption after the administration of gadolinium, inclusion of the hip in the FOV, and adequate visualization of the aorta to use as the arterial input. Exclusion criteria were presence of hip abnormality on MRI, medical history of hip abnormality, incomplete depiction of the hip, and technically limited studies. A hip was considered to be completely depicted if the entire acetabulum, proximal femoral epiphysis, and metaphysis at the level of the greater trochanter were seen. For subjects who underwent more than one examination (n = 3), we included only the initial examination to avoid bias. Likewise, when both hips were imaged, only one hip was analyzed. One child with bilateral Legg-CalvéPerthes disease and five with technically limited studies (incomplete coverage, n = 2; motion, n = 2; low signal-to-noise ratio, n = 1) were excluded. Images were analyzed from 159 subjects (76 girls and 83 boys) with mean [± SD] age, 5.67 ± 5.03 years; range, 0.09–15.91 years. Children were grouped according to age distribution as follows: 0–2 years, n = 56; > 2–5 years, n = 32; > 5–10 years, n = 34; and > 10 years, n = 37.

MRI Parameters All examinations were performed using a 1.5-T MR scanner (Avanto, Siemens Healthcare). Either a six- or nine-channel body matrix coil was used in combination with a 24-channel spine coil. Unenhanced and gadopentetate dimeglumine–enhanced images were acquired using a 3D T1-weighted gradient-echo sequence with fat saturation: volumetric interpolated breath-hold examination (VIBE) (Siemens Healthcare). Each 3D volume was acquired in 15 seconds with TR/TE, 3.63/1.23; flip angle, 30°; and matrix, 208 × 256. Fifty-five 3D dynamic image sets were acquired with increasing 2- to 42-second pauses in between, with an average dynamic imaging time of 17 minutes. Eight to twelve unenhanced slabs were acquired. IV gadopentetate dimeglumine (Magnevist, Bayer Schering Pharma) at a dose of 0.1 mmol/kg (minimum, 2 mL and maximum, 20 mL) was administered with a power injector at a speed setting no greater than 0.25 mL/s. As part of the clinical protocol at our institution, all children received furosemide IV 15 minutes before the administration of contrast material (dose, 1 mg/kg; maximum, 20 mg).

Image Segmentation and Analysis We used a two-compartment model to analyze the kinetics of gadolinium enhancement. This model has been used in skeletally immature subjects [4]. The model assumes that gadopentetate dimeglu-

mine is delivered by a vessel (the aorta, considered the central compartment) and “leaks” into the tissues (peripheral compartment), that the exchange between compartments is linear and bidirectional, and that the gadopentetate dimeglumine is irreversibly eliminated from one or both compartments. The exchange of gadopentetate dimeglumine between compartments leads to changes in signal intensity (SI) over time, which can be used to estimate pharmacokinetic parameters. Perfusion and permeability were calculated for six different ROIs, which served as peripheral compartments. Enhancement data were loaded into software that was written in-house. The software was developed in a Linux environment using Interactive Data Language (IDL). The dynamic sequences were converted into a single 4D array. A vascular ROI was manually drawn in the aorta (Fig. 1). A graph of the average SI over time was generated and the number of unenhanced time points was calculated automatically on the basis of the vascular mask. The δ SI maximum intensity projection (MIP) through time was generated for every slice. To get the δ SI MIP of each slice, we created an image with the highest δ SI (δ SI = SIat determinate time period − SIbase) through time; the highest δ SI of each pixel was selected. All ROIs were manually drawn on the δ SI MIP images by a research assistant with 6 years of experience in processing dynamic gadolinium-enhanced MRI and verified by a physician. Independent ROIs in six different regions in the proximal femur were placed as follows: the unossified epiphysis (referred to as epiphyseal cartilage), the secondary ossification center, the primary physis, the metaphyseal spongiosa, the marrow of the metaphysis, and the periosteum (Fig. 2). The epiphyseal cartilage, secondary ossification center, and metaphyseal marrow were readily identified on the basis of their anatomic location and enhancement pattern on δ SI MIP images. The band of very high SI in the proximal metaphysis was considered to represent the metaphys­eal spongiosa [7], and the layer of cartilage immediately adjacent to the latter on the epiphyseal side was considered to represent the physis [1]. The enhancing periosteum was seen as a thin band of high SI surrounding the bone of the metadiaphysis of the proximal femur [5, 6], just deep in relation to the low SI of the fibrous periosteum. Because the proximal femur was peripheral in the FOV and the structures became too thin, the epiphyseal cartilage could not be drawn in children more than 2 years old or in the physis in children more than 10 years old. The ROIs were exported in a custom binary format on a case-by-case basis for batch processing. ROI volume, enhancement over time curves, and DICOM tags were saved in the binary file. Another custom program was developed in a Linux environment using IDL to analyze the bi-

nary ROI files. For each ROI we calculated the enhancement ratio peak; AUC, defined as the AUC from the time of maximal enhancement of the aorta to 60% enhancement of the aorta; time to peak (TTP) enhancement after contrast administration; and flow from the vessel into the peripheral compartment or rate of extraction (k1 = Ktrans [transfer constant]). To reduce variability in calculation of AUC, we arbitrarily analyzed the curves up to 10 minutes after the injection of gadolinium. Enhancement ratio peak was defined as enhancement ratio peak = (SIpeak − SIbase) / SIbase, where SIpeak is the maximum SI after injection of contrast material and SIbase is the averaged SI from the unenhanced slabs. Enhancement curves were numerically interpolated to a 1-second temporal resolution from the time of contrast injection until 10 minutes after injection for every case (Fig. 3). To generate the enhancement curves, enhancement of each time period was calculated with the following formula: (SIat determinate time period − SIbase) / SIbase. To make individual enhancement curves comparable, all curves were normalized by making the areas under each curve equal to 1. On the basis of these normalized curves, we generated average enhancement curves for all patients and ROI subgroups (Fig. 4).

Statistical Analysis Scatterplots showing age versus enhancement values generally suggested a nonlinear association. Therefore, to examine associations between age and enhancement values, relationships were modeled as power regressions rather than linear fits [i.e., ln(Y) = ln(b0) + (b1 × ln(t)]. For each perfusion measurement (i.e., enhancement ratio peak, AUC, TTP, k1), a linear mixed model analysis was used to examine differences between the different ROIs. If a significant main effect was observed, simple-effect analyses using the post hoc Bonferroni correction was used to determine specific differences. Age group dif-

Fig. 1—8-month-old girl. Change in signal intensity maximum-intensity-projection image from dynamic contrast-enhanced coronal 3D T1-weighted gradientecho fat-saturated sequence shows vascular ROI that was manually drawn in aorta (red).

AJR:203, October 2014 W441

Fig. 2—8-month-old girl with normal right hip and well-developed secondary ossification center. A and B, Change in signal intensity maximumintensity-projection images of dynamic contrastenhanced coronal 3D fat-saturated T1-weighted gradient-recalled echo sequence in midcoronal plane of right femoral head with manually drawn ROIs in B: epiphyseal cartilage (dark blue), secondary ossification center (green), primary physis (yellow), metaphyseal spongiosa (light pink), marrow of metaphysis (dark pink), and periosteum (light blue).

ferences were analyzed using the same statistical tool. The significance level for differences was accepted at p value < 0.05. We used Mann-Whitney U tests to examine the sex differences in the perfusion measurements for each age group. Given 24 measurements (four perfusion measurements in six ROIs) in each age group, to control family-wise error a Bonferroni correction was applied (i.e., p = 0.05/24 = 0.002). All analyses were conducted using SPSS, version 20.0.0 (IBM).

Results SI versus time curves were successfully generated on all studies included in the analyses (Fig. 3). The metaphyseal spongiosa, metaphyseal marrow, and periosteum enhanced avidly but had different patterns of enhancement. At all ages, the metaphyseal spongiosa had a highly vascular pattern of enhancement, with a brisk increase in SI after contrast administration and a high peak followed by a prompt decrease in SI. The peak decreased with age but the enhancement pattern persisted (Fig. 4). The metaphyseal marrow showed a simi-

Fig. 3—Graph shows enhancement versus time curves of different ROIs in 8-month-old girl: epiphyseal cartilage (dark blue), secondary ossification center (green), physis (yellow), metaphyseal spongiosa (light pink), metaphyseal marrow (dark pink), and periosteum (light blue). SI = signal intensity.

W442

A

B

lar curve, with early enhancement and washout, and overall with lower peak than that of the metaphyseal spongiosa. In patients older than 10 years the vascular pattern of enhancement was less discernible. The periosteum also showed prompt enhancement after gadolinium administration; however, the SI reached a plateau and remained elevated for the duration of the examination. In patients older than 11 years, the periosteum continued the ascending enhancement curve without reaching a plateau (Fig. 4). The secondary ossification center, physis, and epiphyseal cartilage showed modest enhancement. The secondary ossification center showed slow uptake and washout in patients less than 0.8 years (9.6 months); after that age, the SI reached a plateau simultaneously with the physis enhancement and showed a very slight washout. The epiphyseal cartilage had minimal enhancement, which remained stable for the duration of the examination. The metaphyseal marrow was successfully sampled in all 159 subjects. The metaphyseal spongiosa could be sampled in 153 sub-

jects (96%); the mean age of subjects with visible metaphyseal spongiosa was 5.29 years (median, 3.95 years; range, 0.09–15.6 years) and that of subjects without a visible metaphyseal spongiosa was 15.57 years (median, 15.6 years; range, 15.2–15.9 years). The enhancing periosteum could be sampled in 152 subjects (95.6%); the mean age of the subjects with visible periosteum was 5.9 years (median, 4.3 years; range: 0.09–15.91 years) and that of subjects without visible periosteal enhancement was 0.36 years (median, 0.48 years; range, 0.12–0.62 years). A retrospective analysis of the anatomic images in the subjects without a well-seen periosteal enhancement revealed faint enhancement in four subjects and hips placed in a position inconvenient for segmentation in three studies. The secondary ossification center ROI was sampled in 136 subjects (85.5%); the mean age for subjects with a secondary ossification center ROI was 6.59 years (median, 5.4 years; range, 0.26–15.9 years) and that of subjects without a secondary ossification center (imaged before the appearance of

Enhancement (SIat determinate time period – SIbase) / SIbase

Downloaded from www.ajronline.org by University of Utah on 09/27/14 from IP address 155.97.178.73. Copyright ARRS. For personal use only; all rights reserved

Bedoya et al.

5

4

3

2

1

0

0

5

10 Time (min)

15

20

AJR:203, October 2014

MRI of Proximal Femur in Children 0.003

Normalized Enhancement

Normalized Enhancement

Downloaded from www.ajronline.org by University of Utah on 09/27/14 from IP address 155.97.178.73. Copyright ARRS. For personal use only; all rights reserved

0.003

0.002

0.001

0

0

2

4

6

8

10

0.002

0.001

0

0

2

Time (min)

4

6

8

10

Time (min)

A

B

Fig. 4—Normalized enhancement curves. A and B, Graphs show average of normalized enhancement curves (AUC = 1) of metaphyseal spongiosa (A) and periosteum (B) from time of contrast injection (0 minutes) until 10 minutes after injection by age group: 0–2 years (red), > 2–5 years (blue), > 5–10 years (green) and > 10 years (yellow).

the ossification center) was 0.26 years (median, 0.20 years; range, 0.09–0.8 years). All subjects above 8 months of age had a visible secondary ossification center. The physis was seen in 121 subjects (76%); the mean ages of subjects with and without physis were 3.32 years (median, 2.72 years; range, 0.09–9.94 years) and 13.18 years (median, 13.25 years; range, 8.11–15.9 years), respectively. The epiphyseal cartilage was seen in 56 subjects (35%); the mean ages of subjects with and without epiphyseal cartilage was 0.73 years (median, 0.59 years; range, 0.09–1.99 years) and 8.37 years (median, 8.05 years; range, 2.06–15.91 years), respectively. There were no significant sex differences in the perfusion parameters measured. Although the secondary ossification center enhancement ratio peak and the periosteum enhancement ratio peak in the age group 0–2 years showed sex differences at the uncorrected p level (with uncorrected p values around 0.03), no measurement showed significant sex differences at the family-wise corrected p value.

2 years, a subtle increase in enhancement ratio was observed. R2 values in a power regression model were as follows: metaphyseal spongiosa (0.662), metaphyseal marrow (0.647), periosteum (0.399), secondary ossification center (0.508), physis (0.168), and epiphyseal cartilage (0.305) (Fig. 5). There were significant differences in the enhancement ratio peak among the different regions (p < 0.001). In the overall population, the metaphyseal spongiosa had the highest enhancement ratio peak, followed by the periosteum, metaphyseal marrow, physis, secondary ossification center, and epiphyseal cartilage. Linear mixed-model analyses with post hoc Bonferroni correction confirmed that, for the whole population, enhancement ratio peak in the metaphyseal spongiosa, metaphyseal marrow, and periosteum were significantly different (p < 0.001). In children older than 5 years, the enhancement ratio peak of the metaphyseal spongiosa and the periosteum could no longer be differentiated (p > 0.999).

Enhancement Ratio Peak For all ROIs, a significant association between age and enhancement ratio peak values was observed (p < 0.001), with enhancement ratio peak values decreasing in relation to age. In the metaphyseal spongiosa, metaphyseal marrow, secondary ossification center, and periosteum there was a marked decrease in the enhancement ratio during the first 2 years of life, followed by a slower rate of decrease after this age. The physis showed a similar curve during infancy; however, after reaching a nadir at around

AUC There were significant differences in the AUC among the different bone regions (p < 0.001). In the overall population, the metaphyseal spongiosa had the highest AUC, followed by the periosteum, metaphyseal marrow, physis, secondary ossification center, and epiphyseal cartilage. Linear mixed-model analyses with post hoc Bonferroni correction revealed that for the whole population, only the differences between the metaphyseal spongiosa and all other regions showed statistical significance (p < 0.001). Just as for the enhancement ratio

peak curves, there was an age-related decrease in AUC in all regions, which was particularly obvious during the first 2 years of life. This decline, however, was not as notable as the decline for the enhancement ratio peak (R2 < 0.213). After this period of rapid decrease, all the structures showed minimal changes with age, except the physis, which showed a minor increase. Time to Peak There were significant differences in the TTP among the different regions (p < 0.001). In the overall population, the first structure to reach peak enhancement was the metaphyseal spongiosa, followed by the metaphyseal marrow; secondary ossification center; physis; epiphyseal cartilage; and, lastly, the periosteum. Linear mixed-model analyses with post hoc Bonferroni correction revealed that, for the whole population, the metaphyseal spongiosa showed a statistical difference (p = 0.007) compared with all the other ROIs. The metaphyseal spongiosa presented a significant association between age and TTP (p < 0.001), showing a slight increase of TTP with age (R2 = 0.258). All other comparisons were not significant. Microvascular Permeability: Rate of Extraction There were significant differences in the rate of extraction among the different regions (p < 0.001). The rate of extraction was higher in the metaphyseal spongiosa, followed by the metaphyseal marrow and periosteum. The rate of extraction was low and similar for the physis and secondary ossification center and lowest in the epiphyseal

AJR:203, October 2014 W443

8.0

8.0

7.0

7.0 Enhancement Ratio Peak

Enhancement Ratio Peak

connective tissue surrounded by a dense microcirculation [8]. The periosteal cambium layer is involved with fracture healing and bone remodeling and contains cells that

6.0 5.0 4.0 3.0 2.0 1.0 0.0

6.0 5.0 4.0 3.0 2.0 1.0

0

5

10 Age (y)

15

0.0

20

0

8.0

8.0

7.0

7.0

6.0 5.0 4.0 3.0 2.0 1.0 0.0

4.0 3.0 2.0 1.0

0

5

10 Age (y)

15

0.0

20

0

Rate of Extraction

Rate of Extraction

0.040 0.030 0.020 0.010

5

10 Age (y)

15

20

A

15

20

Fig. 5—Enhancement ratio peak and age. A–D, Scatterplot and power regression graphs show enhancement ratio peak versus age in years of metaphyseal spongiosa (A), metaphyseal marrow (B), periosteum (C), and secondary ossification center (D).

0.050

0

10 Age (y)

D

0.050

0

5

C

0.050

0.010

20

5.0

0.060

0.020

15

6.0

0.060

0.030

10 Age (y)

B

0.060

0.040

5

A

Enhancement Ratio Peak

Discussion Our study shows that gadopentetate dimeglumine enhancement varies significantly among the various structures of the growing skeleton. Enhancement is greatest in the periosteum and metaphyseal spongiosa, followed by the marrow of the metaphyseal bone, the epiphyseal ossification center, the epiphyseal cartilage, and the physis. After contrast material administration, SI in all the regions has an early peak followed by a washout period. The only exception is the periosteum, where SI increases steadily with time. Peaks of enhancement ratio decrease with age in all the bone regions but decrease faster in the epiphyseal than in the metaphyseal structures, with the periosteum having the slowest rate of decrease. Permeability is greatest in the metaphyseal spongiosa; it peaks at 2 years and decreases thereafter.

The periosteum has two layers: an outer fibrous layer and an inner structure called the cambium, which lies adjacent to the bone and contains cells embedded in a loose

Enhancement Ratio Peak

cartilage. Linear mixed-model analyses with post hoc Bonferroni correction revealed that for the whole population only the metaphyseal spongiosa showed a statistical difference (p < 0.001) compared with all other ROIs. As with the enhancement parameters, we observed an age-related decrease in k1 (p < 0.001) that was particularly obvious during the first 2 years of life (R2 < 0.606) in all regions, except for the physis. In infants, permeability in the secondary ossification center was high, decreased abruptly during the first year of life, and remained stable thereafter (R2 = 0.362). Permeability in the metaphyseal marrow, metaphyseal spongiosa, and periosteum also was higher during the first 2 years of life and decreased with age (R2 < 0.606) (Fig. 6).

Rate of Extraction

Downloaded from www.ajronline.org by University of Utah on 09/27/14 from IP address 155.97.178.73. Copyright ARRS. For personal use only; all rights reserved

Bedoya et al.

0

0.040 0.030 0.020 0.010

0

5

10 Age (y)

15

20

0

0

5

10 Age (y)

15

20

B

C

Fig. 6—Rate of extraction and age. A–C, Scatterplot and power regression graphs show rate of extraction versus age in years of metaphyseal spongiosa (A), metaphyseal marrow (B), and periosteum (C).

W444

AJR:203, October 2014

Downloaded from www.ajronline.org by University of Utah on 09/27/14 from IP address 155.97.178.73. Copyright ARRS. For personal use only; all rights reserved

MRI of Proximal Femur in Children can undergo osteochondrogenic differentiation during the healing process [8]. The periosteal cambium is responsible for membranous growth and is most active during the growing years. Membranous growth is steady throughout the growing years without a direct hormonal influence [9, 10]. The relatively slow decline in enhancement through the years may reflect that periosteal activity persists beyond puberty and adolescence. The poor visualization of the cambium in the youngest subjects likely reflects limitations of spatial resolution and not a true structural change [6]. The metaphyseal spongiosa is the area where endochondral ossification occurs. This process requires the junction of the columns of cartilage and the vessels of the spongiosa. Abundant vascularity is necessary for the cartilage to be transformed into bone [5, 11, 12]. The amount of endochondral transformation is related to the rate of growth and therefore is subject to hormonal influence. Although it would be expected that vascularity would also follow the rate of growth and thus increase during the growth spurt, our study and prior animal studies show that vascularity is greater during infancy and decreases steadily thereafter [1, 13]. The epiphyseal ossification center of the proximal femur appears normally at around 6 months on MRI, and the epiphyseal marrow could only be sampled after that age [14]. The greater rate of decline in enhancement of epiphyseal than metaphyseal marrow has been documented in prior studies [13]. Throughout childhood, the greater vascularity of metaphyseal marrow is responsible for the greater incidence of hematogenous pathology in the metaphysis than in the epiphysis and for the greater incidence of ischemic pathology in the epiphyseal marrow [13, 15]. The enhancement ratio peak in the physeal and epiphyseal cartilage decreases rapidly after infancy and is minimal in later childhood and adolescence. This correlates well with what is seen histologically: The physis becomes avascular after 18 months, and the epiphyseal vascular canals become sparser after the first 2 years of life [16]. Highly vascular structures typically have a greater enhancement ratio peak, higher peak intensity, shorter TTP, larger area under the time-intensity curve, and greater permeability. Prior studies have not contained pharmacokinetic analyses and have been limited to an evaluation of enhancement ratios [7, 13, 17, 18]. This type of analysis is

particularly useful when looking at the difference between the enhancement of the metaphyseal spongiosa and the periosteal cambium. These two very vascular structures have markedly high enhancement ratio peaks, but the TTP is low in the spongiosa (which has a near-arterial configuration) and high in the periosteal cambium, in which the contrast material accumulates with time. Contrast material washes out quickly in the metaphyseal spongiosa but remains in the periosteal cambium. Presumably in the periosteal cambium, the contrast material leaks into the loose connective tissue and remains there, whereas in the metaphyseal spongiosa it is washed away by the abundant metaphyseal vessels [1, 7]. The possible physiologic repercussions are intriguing given that the metaphyseal spongiosa is highly related to endochondral ossification (longitudinal growth), whereas the periosteal cambium is related to membranous ossification (appositional growth) [8, 19]. The study is limited because the MRI examinations were not tailored to evaluate the skeletal structures, and therefore spatial resolution was sometimes suboptimal. All patients were evaluated for a specific abnormality. Although there was no known musculoskeletal abnormality in the subjects, they were not free of disease. In conclusion, the structures of the growing skeleton have different patterns of enhancement. The highest enhancement is seen in areas involved with growth: the metaphyseal spongiosa, related to endochondral ossification, and the periosteal cambium, related to membranous ossification. Their enhancement characteristics are markedly different: In the spongiosa, enhancement is brisk and declines, with a vascular pattern, whereas contrast uptake increases with time in the cambium. Enhancement in all areas decreases with age. The differences in perfusion help in understanding the distribution of certain diseases in the growing skeleton. The rich vascularity of the metaphyseal spongiosa and the periosteal cambium predisposes these areas to blood-borne diseases, such as bacterial invasion or seeding of malignant cells [20, 21]. This explains the greater incidence of hematogenous osteomyelitis and metastasis to bone in the metaphysis. Subperiosteal abscesses and subperiosteal tumoral elevation are attributable to the rich vascularization of the periosteal cambium and the loose attachment of the periosteum to the bone cortex along the di-

aphysis [5]. Thus, recognition of normal enhancement patterns of the proximal femur is important for distinguishing normal development from pathologic processes and for better understanding of the physiology of the immature skeleton. References 1. Jaramillo D, Villegas-Medina OL, Doty DK, et al. Age-related vascular changes in the epiphysis, physis, and metaphysis: normal findings on gadolinium-enhanced MRI of piglets. AJR 2004; 182:353–360 2. Jaramillo D, Villegas-Medina OL, Doty DK, et al. Gadolinium-enhanced MR imaging demonstrates abduction-caused hip ischemia and its reversal in piglets. Pediatr Radiol 1995; 25:578– 587 3. Kim EY, Kwack KS, Cho JH, Lee DH, Yoon SH. Usefulness of dynamic contrast-enhanced MRI in differentiating between septic arthritis and transient synovitis in the hip joint. AJR 2012; 198:428–433 4. Workie DW, Dardzinski BJ, Graham TB, Laor T, Bommer WA, O’Brien KJ. Quantification of dynamic contrast-enhanced MR imaging of the knee in children with juvenile rheumatoid arthritis based on pharmacokinetic modeling. Magn Reson Imaging 2004; 22:1201–1210 5. Laor T, Jaramillo D. MR imaging insights into skeletal maturation: what is normal? Radiology 2009; 250:28–38 6. Laor T, Chun GF, Dardzinski BJ, Bean JA, Witte DP. Posterior distal femoral and proximal tibial metaphyseal stripes at MR imaging in children and young adults. Radiology 2002; 224:669–674 7. Menezes NM, Olear EA, Li X, et al. Gadolinium-enhanced MR images of the growing piglet skeleton: ionic versus nonionic contrast agent. Radiology 2006; 239:406–414 8. Frey SP, Jansen H, Doht S, Filgueira L, Zellweger R. Immunohistochemical and molecular characterization of the human periosteum. ScientificWorldJournal 2013; 2013:341078 9. Gosman JH, Hubbell ZR, Shaw CN, Ryan TM. Development of cortical bone geometry in the human femoral and tibial diaphysis. Anat Rec (Hoboken) 2013; 296:774–787 10. Franz-Odendaal TA. Induction and patterning of intramembranous bone. Front Biosci Landmark Ed 2011; 16:2734–2746 11. Ecklund K, Jaramillo D. Imaging of growth disturbance in children. Radiol Clin North Am 2001; 39:823–841 12. Mackie EJ, Ahmed YA, Tatarczuch L, Chen KS, Mirams M. Endochondral ossification: how cartilage is converted into bone in the developing skeleton. Int J Biochem Cell Biol 2008;

AJR:203, October 2014 W445

Downloaded from www.ajronline.org by University of Utah on 09/27/14 from IP address 155.97.178.73. Copyright ARRS. For personal use only; all rights reserved

Bedoya et al. 40:46–62 13. Dwek JR, Shapiro F, Laor T, Barnewolt CE, Jaramillo D. Normal gadolinium-enhanced MR images of the developing appendicular skeleton. Part 2. Epiphyseal and metaphyseal marrow. AJR 1997; 169:191–196 14. Jaramillo D, Laor T, Hoffer FA, et al. Epiphyseal marrow in infancy: MR imaging. Radiology 1991; 180:809–812 15. Jaramillo D, Galen TA, Winalski CS, et al. Legg-Calvé-Perthes disease: MR imaging evaluation during manual positioning of the hip—

W446

comparison with conventional arthrography. Radiology 1999; 212:519–525 16. Rivas R, Shapiro F. Structural stages in the development of the long bones and epiphyses: a study in the New Zealand white rabbit. J Bone Joint Surg Am 2002; 84-A:85–100 17. Barnewolt CE, Shapiro F, Jaramillo D. Normal gadolinium-enhanced MR images of the developing appendicular skeleton. Part 1. Cartilaginous epiphysis and physis. AJR 1997; 169:183– 189 18. Menezes NM, Connolly SA, Shapiro F, et al.

Early ischemia in growing piglet skeleton: MR diffusion and perfusion imaging. Radiology 2007; 242:129–136 19. Trueta J, Morgan JD. The vascular contribution to osteogenesis. I. Studies by the injection method. J Bone Joint Surg Br 1960; 42-B:97–109 20. Jaramillo D. Infection: musculoskeletal. Pediatr Radiol 2011; 41(suppl 1):S127–S134 21. Bussard KM, Gay CV, Mastro AM. The bone microenvironment in metastasis; what is special about bone? Cancer Metastasis Rev 2008; 27:41–55

AJR:203, October 2014

Dynamic gadolinium-enhanced MRI of the proximal femur: preliminary experience in healthy children.

The purpose of this study is to use dynamic contrast-enhanced MRI to evaluate the perfusion characteristics of the proximal femur in the growing skele...
974KB Sizes 0 Downloads 5 Views