Human Fetal Bone Development: Histomorphometric Evaluation of the Proximal Femoral Metaphysis

Transcription

1 Bone Vol. 30, No. 6 June 2002: Human Fetal Bone Development: Histomorphometric Evaluation of the Proximal Femoral Metaphysis B. L. SALLE, 1 F. RAUCH, 2 R. TRAVERS, 2 R. BOUVIER, 3 and F. H. GLORIEUX 2 1 Department of Neonatology, Hôpital Edouard Herriot, Lyon, France 2 Genetics Unit, Shriners Hospital and McGill University, Montréal, Qc, Canada 3 Department of Pathology, Hôpital Edouard Herriot, Lyon, France Quantitative data on metaphyseal bone histology during early human development are scarce. In the present study the proximal femoral metaphysis of 35 fetuses and newborns (gestational age weeks) was analyzed by histomorphometry. Averaged over the entire metaphyseal area, the relative amount of bone and cartilage was higher in the third compared to the second trimester. Osteoid thickness increased with gestational age, whereas indices of bone resorption decreased. The relative amount of cartilage decreased with increasing distance from the growth plate, whereas the relative amount of bone increased. This was due to trabecular thickening, which occurred at an estimated rate of 3 m/day in areas close to the growth plate. Despite this rapid rate of net bone gain, osteoid indices were relatively low, indicating that mineralization occurred very rapidly after bone deposition. These observations suggest that modeling, not remodeling, is the predominant mechanism responsible for the development of femoral metaphyseal cancellous bone in utero. (Bone 30: ; 2002) 2002 by Elsevier Science Inc. All rights reserved. Key Words: Bone development; Femur; Fetus; Histomorphometry; In utero; Metaphysis. Introduction Skeletal development can be regarded schematically as a succession of two major phases. 4,9 The first is skeletal patterning, the formation of skeletal elements as soft tissue templates. The basic shape of all bones is preformed during this process, which in humans is complete by the end of embryonic development. In the second phase of skeletal development, the soft tissue template undergoes gradual ossification and growth. This leads to a dramatic increase in size, but relatively little changes in shape. This phase lasts from the early stages of fetal life until adulthood. In the development of long bones, the soft tissue model consists of cartilaginous tissue. Ossification first starts in the midshaft region, where a bony collar is formed. 9 This is followed by vascular invasion of the cartilaginous center and replacement of cartilage tissue by mineralized trabecular bone. This primary center of ossification rapidly expands in each direction toward the two ends. Soon afterwards the trabeculae in the center of the Address for correspondence and reprints: Dr. Frank Rauch, Genetics Unit, Shriners Hospital for Children, 1529 Cedar Avenue, Montreal, Qc H3G 1A6, Canada. frauch@shriners.mcgill.ca shaft region are resorbed to form the medullary cavity. The three basic parts of a growing long bone are thus in place: the epiphysis, which still consists of cartilage; the metaphysis with trabeculae surrounded by cortical bone; and the diaphysis, where only cortical bone is present. Tubular bones grow in length by a continuation of the endochondral ossification process outlined earlier. Epiphyseal cartilage cells situated close to the metaphysis proliferate continuously, secrete cartilaginous matrix, and finally hypertrophy. 14 Columns of hypertrophied cartilage cells are separated by longitudinal septa. At the metaphysis/growth cartilage junction (MGCJ), the cartilage is invaded by blood vessels, which pave the way for bone cell precursors. 11 Eighty percent of the longitudinal septa of the growth cartilage are rapidly resorbed in the metaphyseal zone immediately behind the invading blood vessels. 5 The remaining longitudinal septa serve as scaffolds, on which osteoblasts deposit bone matrix. Thus, primary trabeculae are formed, which consist of a mixture of cartilage and bone tissue. With increasing distance from the growth cartilage these primary trabeculae thicken and undergo rapid turnover in a process commonly referred to as remodeling. 1 At the border between metaphysis and diaphysis, all metaphyseal trabeculae are removed as long as the bone grows in length. The diaphysis is therefore devoid of trabeculae. Thus, metaphyseal trabeculae of long bones are transient structures during bone development, which undergo a life cycle of creation at the MGCJ and destruction at the diaphyseal side of the metaphysis. 9 Detailed qualitative descriptions of metaphyseal bone histology during early development have been available for decades. 6,9 In contrast, quantitative histomorphometric data are scarce. One study has investigated the structural aspects of trabecular morphogenesis at the costochondral junction in children from 3 to 36 weeks of age. 5 Another study has analyzed iliac crest bone in a small group of premature and newborn babies. 2 However, metaphyseal bone at these locations is different from that in long bone metaphysis, as the ribs and ilium do not have a diaphysis and therefore trabeculae persist instead of being resorbed. 17 We previously have reported histomorphometric results in femoral metaphyses from two human fetuses at 19 weeks of gestation and five newborns at term. 12 In the present study we extend these observations to a group of 35 fetuses and newborns covering the gestational age range from 16 to 41 weeks. Apart from establishing morphometric reference data for the proximal femoral metaphysis, these data provide the opportunity to gain insight into long bone growth in humans by Elsevier Science Inc /02/$22.00 All rights reserved. PII S (02)00724-X

2 824 B. L. Salle et al. Bone Vol. 30, No. 6 Histomorphometry of the fetal femoral metaphysis June 2002: All histomorphometric analyses were made in one frontal section, spanning the whole cancellous compartment of the metaphysis. The measured area was mm 2 (mean SD). Because bone structure and turnover vary as a function of the distance from the MGCJ, measurements were performed separately in eight 0.75-mm-wide consecutive bands starting at 0.75 mm below the growth cartilage (Figure 1). Measurements were restricted to cancellous bone. Histomorphometric analyses were performed as described previously. 13 The following primary measures were obtained in cancellous bone: tissue area, bone area, bone perimeter, cartilage area, osteoid area, osteoid perimeter, and osteoclast perimeter. Osteoblast perimeter was not determined, because osteoblasts could not be identified with sufficient precision in the cell-rich areas close to the MGCJ. All parameters were derived from these primary measures, as described. 13 Nomenclature and abbreviations follow the recommendations of the American Society for Bone and Mineral Research. 18 Statistical Analysis Figure 1. Schematic representation of the proximal femoral metaphysis in the human fetus. The fields of histomorphometric analysis are indicated. Each field covers an area of 0.56 mm 2. Fields carrying the same number were grouped together as bands that run parallel to the MGCJ. Materials and Methods Femur Samples Samples from 35 fetuses and newborns were studied. Gestational age varied from 16 to 41 weeks. All fetuses of a gestational age up to 22 weeks had been stillborn. Babies born after more than 22 weeks of gestation had died within 1 h after birth. Malformed fetuses or newborns were not included into this study. After consent had been obtained from the parents as required under French laws, femurs were removed in toto at autopsy and collected by one of us (R.B.) in the Department of Pathology at the Hôpital Edouard Herriot (Lyon, France). Sample Processing Specimens were cleaned of soft tissues, their lengths measured with a caliper, cut transversely at midshaft, and fixed in buffered formaldehyde (ph 7.1). A frontal cut of each femoral fragment was made with a rotating diamond saw (Isomet Low Speed Saw, Buehler, Ltd.) to expose epiphyseal cartilage, metaphyseal area, and part of the diaphysis. After dehydration and clearing, specimens were embedded in methylmetacrylate. After polymerization, the blocks were trimmed to orient the plane of section to have a complete frontal view of the specimen. Sections of 6 m were cut and stained with Goldner trichrome, as described previously. 13 Histomorphometry The specimens were separated into three gestational age groups: samples from the second trimester (gestational age weeks); the first half of the third trimester (28 33 weeks); and the second half of the third trimester (34 41 weeks). Differences between gestational age groups were tested for significance using one-way analysis of variance (ANOVA). For post hoc analyses of differences between individual groups, Bonferroni s adjustment was used. Differences between metaphyseal bands were tested by ANOVA for repeated measures. Differences between two groups were evaluated by paired or unpaired t-tests, as appropriate. p 0.05 was considered statistically significant. These calculations were performed using SPSS 6.0 for Windows software (SPSS, Inc., Chicago, IL). Results Femoral shaft length increased from 34 mm at 16 weeks to 95.5 mm at term. Table 1 gives histomorphometric results measured in the three gestational age groups. All measurement fields of a specimen were pooled for this analysis. The relative amount of bone (BV/TV) was about one third higher in the last trimester as compared with the second trimester. As shown in Figure 2, bone volume appeared to increase steadily from the early part of the second trimester to a gestational age of about 30 weeks. This increase in the amount of bone was due to an increase in trabecular thickness (Tb.Th), whereas the number of trabeculae (Tb.N) remained about stable (Table 1). The relative amount of cartilage (Cg.V/TV) also increased between the second and third trimester. The cartilage-to-bone ratio (Cg.V/BV) did not vary with gestational age. As for parameters of bone formation, the relative amount of bone surface covered with osteoid (OS/BS) varied widely and was similar between gestational age groups. In contrast, osteoid thickness (O.Th) increased significantly. Relative osteoid volume (OV/BV) did not change with gestational age. All indices of bone resorption decreased with gestational age. Metaphyseal cancellous bone structure and turnover of growing individuals depend on the distance from the growth cartilage. 5,15,22 Therefore, histomorphometric parameters were analyzed separately in eight adjacent bands starting at 0.75 mm below the growth cartilage (Figure 1). Table 2 shows histomorphometric parameters of the three gestational age groups in band 1. Compared with the results for the entire area of analysis (Table 1), there was less bone but more cartilage in band 1. A larger amount of bone surface was covered with osteoid in the third than in the second trimester. Indices of bone resorption were higher in band 1 than in the pooled measurement area. The changes in structural parameters with increasing distance from the MGCJ are shown in Figure 3. Because the pattern of these changes was similar in all three gestational age groups, the specimens of all gestational ages were analyzed together. Bone volume increased markedly from the first to the last band, because trabecular thickening more than offset the decrease in

3 Bone Vol. 30, No. 6 June 2002: B. L. Salle et al. Histomorphometry of the fetal femoral metaphysis 825 Table 1. Description of study material and histomorphometric results in three gestational age groups (all metaphyseal regions are pooled) weeks weeks weeks p n Gestational age (weeks) bc ac ab Structural parameters BV/TV (%) bc a a Tb.Th ( m) bc a 93 9 a Tb.N (/mm) Cg.V/TV (%) Cg.V/BV (%) Bone formation OS/BS (%) O.Th ( m) c a OV/BV (%) Bone resorption Oc.S/BS (%) c a 0.03 See Results section for abbreviations of bone parameters. Values expressed as mean SD. p values represent results of analysis of variance (ANOVA). Results of post hoc unpaired t-test using Bonferroni s adjustment are indicated as follows: a significant difference (p 0.05) vs. data from the week group; b significant difference (p 0.05) vs. data from the week group; and c significant difference (p 0.05) vs. data from the week group. trabecular number. The amount of cartilage decreased with increasing distance from the growth cartilage. All of these structural changes were more rapid close to the growth cartilage (band 1 to 4) than in metaphyseal regions closer to the diaphysis (bands 5 8). Figure 4 shows the changes in formation and resorption indices with increasing distance from the growth cartilage. Osteoid surface exhibited a biphasic pattern. First, there was an increase between bands 1 and 2, which was of borderline significance (p 0.08; paired t-test). Further away from the growth cartilage, osteoid surface decreased continuously to 20% of its peak value, but the thickness of osteoid seams remained unchanged. Accordingly, osteoid volume decreased between bands 1 and 8. Parameters of bone resorption decreased continuously with increasing distance from the MGCJ. Discussion In this study we quantitatively analyzed cancellous bone in the proximal femoral metaphysis of fetuses, as well as in preterm and full-term babies. Because none of the babies had survived for more than 1 h after birth, our results reflect mostly intrauterine bone development. Although none of the fetuses and newborns in this study had obvious signs of a skeletal disorder, it must be noted that our data may have been influenced by the conditions that finally led to death. This is a caveat that applies to most of Table 2. Histomorphometric results in band 1 (closest to the growth cartilage) according to gestational age weeks weeks weeks p Structural parameters BV/TV (%) c a Tb.Th ( m) bc a 58 7 a Tb.N (/mm) Cg.V/TV (%) Cg.V/BV (%) Bone formation OS/BS (%) c a O.Th ( m) c a 0.02 OV/BV (%) c a 0.01 Bone resorption Oc.S/BS (%) See Results section for abbreviations of bone parameters. Values expressed as mean SD. p values represent results of analysis of variance (ANOVA). Results of post hoc unpaired t-test using Bonferroni s adjustment are as follows: a significant difference (p 0.05) vs. data from the week group; b significant difference (p 0.05) vs. data from the week group; and c significant difference (p 0.05) vs. data from the week group.

4 826 B. L. Salle et al. Bone Vol. 30, No. 6 Histomorphometry of the fetal femoral metaphysis June 2002: Figure 2. Individual results for bone volume as a function of gestational age. All metaphyseal regions are pooled. what we know about the histology of human intrauterine development. With the limitations inherent to such studies, the data presented in Table 1 might be regarded as histomorphometric reference values for the proximal femoral metaphysis during early development. These results might be useful for studies that aim at a quantitative characterization of lethal bone dysplasias. 21 In addition, the present study offers the opportunity to shed some light on the process of metaphyseal development in a human long bone. Femur length increased from 34 mm at a gestational age of 16 weeks to 95.5 mm at 41 weeks, corresponding to a growth rate of 0.35 mm per day. This is consistent with previous reports on the development of the fetal femur. 6,10 During in utero development about 45% to the longitudinal growth of the femur is contributed by the proximal growth cartilage. 6,9 Thus, the MGCJ and the remainder of the proximal femoral metaphysis advance at a speed of approximately 0.16 mm/day. We performed our measurements in adjacent bands that were parallel to the growth cartilage and had a width of 0.75 mm. Therefore, the mean age of the bone (i.e., the time that had passed since release from the MGCJ) in adjacent bands differed by 0.75/0.16 days 4.7 days. These considerations allow estimating the dynamics of trabecular development from our findings. Trabeculae were 41 m thicker in band 4 than in band 1 (Table 2). As the distance between these two bands corresponds to an age difference of about 14 days, trabecular thickness must have increased at a rate of approximately 3 m/day. This estimated rate of the increase in trabecular thickness is roughly 240 times faster than the rate of trabecular thickening occurring in the iliac bone of children between 1.5 and 11 years of age. 13,19 Our estimates may carry some error, but they nevertheless show that net bone formation activity must be at least two orders of magnitude higher in the fetal bone metaphysis than in iliac cancellous bone of children. This difference in speed of bone apposition points to a fundamental difference between bone metabolism in fetal metaphyseal and in secondary cancellous bone. The prevailing metabolic activity in secondary cancellous bone is remodeling, which is defined as the cyclical succession of bone resorption and formation on the same bone surface. 16 This process is slow (several months for one cycle), and due to its cyclical nature is an inefficient mechanism to achieve rapid changes in bone mass. In fact, the net positive balance of a remodeling cycle in the secondary bone of children is only about 10% of the amount of bone formed. 19,20 Remodeling adds or removes only a few microns of bone to trabecular surfaces per year. 16,19 In light of these considerations, it appears implausible that the rate of trabecular thickening between band 1 and band 4 3 m per day could be achieved by remodeling. Changes of this speed are typical for bone modeling. 7 In modeling, osteoblasts are active without the need for prior resorption, and they remain at a high activity level for a long period of time. 16 For these reasons modeling allows far more rapid changes in bone mass than remodeling. 7,8,16 Our study provides only indirect evidence that trabecular thickening at the fetal femoral metaphysis is due to modeling. However, this conclusion is supported by observations made in the proximal tibial metaphysis of growing rats. Figure 3. Structural parameters. Variation with distance from the growth cartilage (mean SEM). The significance of the difference vs. results in band 1 indicated by asterisks (*p 0.05, **p 0.01, ***p 0.001, by paired t-test).

5 Bone Vol. 30, No. 6 June 2002: B. L. Salle et al. Histomorphometry of the fetal femoral metaphysis 827 Figure 4. Parameters of bone formation and resorption. Variation with distance from the growth cartilage (mean SEM). The significance of the difference vs. results in band 1 indicated by asterisks (*p 0.05, **p 0.01, ***p 0.001, by paired t-test). Multiple fluorochrome labeling revealed that modeling and not remodeling is the predominant mechanism of bone metabolism at that location. 3 This suggests that concepts about the cellular mechanisms of bone turnover, which have been derived from studies in mature cancellous bone, are not necessarily applicable to the metaphysis of developing long bones. The relatively low results for osteoid indices in the femoral metaphysis still seem to be at variance with the rapid increase in trabecular thickness. This could be partly explained by a very short mineralization lag time. Indeed, matrix mineralization appears to occur very rapidly, as shown in two other fetuses where dynamic parameters could be measured. 12 In addition, the most rapid increase in trabecular thickness occurring close to the growth cartilage is due to woven bone formation, whereby no osteoid seams arise. 16 The increase in osteoid surface between bands 1 and 2 therefore may reflect the shift from woven bone to lamellar bone formation. In band 1 the amount of bone was already about 80% of the average for all fields analyzed, reflecting the extremely rapid rate of bone formation immediately adjacent to the MGCJ. Bone volume clearly differed between the gestational age groups in band 1. Thus, the difference between gestational ages in bone volume results mostly from events taking place within a distance of approximately 1 mm to the growth cartilage, corresponding to 6 7 days of development. Bone metabolism clearly slowed down toward the diaphysis, as indicated by a decrease in formation and resorption parameters. This is also suggested by the observation that differences in trabecular thickness and number were less between bands 5 and 8. Trabecular number appeared to stabilize in the bands furthest away from the MGCJ, but our analysis did not include the border between metaphysis and diaphysis, where the remaining trabeculae should disappear. We therefore cannot address the question of whether the ultimate removal of metaphyseal trabeculae is due to a re-increase in resorptive activity or a further decrease in formation activity. In summary, we have presented quantitative reference material for cancellous bone histomorphometry of the proximal femoral metaphysis in the human fetus. These data show that metaphyseal trabeculae are undergoing very rapid changes during their life cycles. Our findings support the hypothesis that modeling, not remodeling, is the predominant mechanism of cancellous bone development in at least a part of the metaphysis. Acknowledgments: The authors thank Guy Charette for processing the bone samples, and Mark Lepik for illustrations. This work was supported by the Shriners of North America. References 1. Baron R. E. Anatomy and ultrastructure of bone. In: Favus M. J., ed. Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism. 3rd Ed. Philadelphia: Lippincott-Raven; 3 10; Beyers N., Esser M., Alheit B., Roodt M., Wiggs B., and Hough S. F. Static bone histomorphometry in preterm and term babies. Bone 15:1 4; Erben R. G. Trabecular and endocortical bone surfaces in the rat: Modeling or remodeling? Anat Rec 246:39 46; Erlebacher A., Filvaroff E. H., Gitelman S. E., and Derynck R. Toward a molecular understanding of skeletal development. Cell 80: ; Fazzalari N. L., Moore A. J., Byers S., and Byard R. W. Quantitative analysis of trabecular morphogenesis in the human costochondral junction during the postnatal period in normal subjects. Anat Rec 248:1 12; Felts W. J. The prenatal development of the human femur. Am J Anat 94:1 44; Frost H. M. Skeletal structural adaptations to mechanical usage (SATMU): 1. Redefining Wolff s law: the bone modeling problem. Anat Rec 226: ; Frost H. M. Skeletal structural adaptations to mechanical usage (SATMU): 2. Redefining Wolff s law: the remodeling problem. Anat Rec 226: ; Gardner E. Osteogenesis in the human embryo and fetus. In: Bourne G. H.. 2nd Ed.. The Biochemistry and Physiology of Bone, Vol. 3. London: Academic; ; 1971.

6 828 B. L. Salle et al. Bone Vol. 30, No. 6 Histomorphometry of the fetal femoral metaphysis June 2002: Gardner E., and Gray D. J. The prenatal development of the human femur. Am J Anat 129: ; Gerber H. P., Vu T. H., Ryan A. M., Kowalski J., Werb Z., and Ferrara N. VEGF couples hypertrophic cartilage remodeling, ossification and angiogenesis during endochondral bone formation. Nat Med 5: ; Glorieux F. H., Salle B. L., Travers R., and Audra P. H. Dynamic histomorphometric evaluation of human fetal bone formation. Bone 12: ; Glorieux F. H., Travers R., Taylor A., Bowen J. R., Rauch F., Norman M., and Parfitt A. M. Normative data for iliac bone histomorphometry in growing children. Bone 26: ; Howell D. S., and Dean D. D. The biology, chemistry, and biochemistry of the mammalian growth plate. In: Coe F. L., & Favus M. J., eds. Disorders of Bone and Mineral Metabolism. New York: Raven; ; Kimmel D. B., and Jee W. S. A quantitative histologic analysis of the growing long bone metaphysis. Calcif Tissue Int 32: ; Parfitt A. M. Bone-forming cells in clinical conditions. In: Hall B. K., ed. The Osteoblast and Osteocyte, Vol. 1. Caldwell, NJ: Telford; ; Parfitt A. M. The two faces of growth: Benefits and risks to bone integrity. Osteopor Int 4: ; Parfitt A. M., Drezner M. K., Glorieux F. H., Kanis J. A., Malluche H., Meunier P. J., Ott S. M., and Recker R. R. Bone histomorphometry: standardization of nomenclature, symbols, and units. Report of the ASBMR Histomorphometry Nomenclature Committee. J Bone Miner Res 2: ; Parfitt A. M., Travers R., Rauch F., and Glorieux F. H. Structural and cellular changes during bone growth in healthy children. Bone 27: ; Rauch F., Travers R., Parfitt A. M., and Glorieux F. H. Static and dynamic bone histomorphometry in children with osteogenesis imperfecta. Bone 26: ; Spranger J., and Maroteaux P. The lethal osteochondrodysplasias. Adv Hum Genet 19:1 103; Turner R. T. Cancellous bone turnover in growing rats: Time-dependent changes in association between calcein label and osteoblasts. J Bone Miner Res 9: ; Date Received: November 26, 2001 Date Revised: January 31, 2002 Date Accepted: February 5, 2002