REVIEW

The periosteumPart 1: Anatomy, histology and molecular biology

Goran Augustin *, Anko Antabak, Slavko Davila

Injury, Int. J. Care Injured (2007) 38, 1115—1130

www.elsevier.com/locate/injury

Clinical Hospital Center Zagreb, Kispaticeva 12, 10000 Zagreb, Croatia

Accepted 21 May 2007

KEYWORDSPeriosteum;Fibrous layer;Cambium layer;Sharpey’s fibres;Periosteal circulation;Bone formation;Bone resorption;Perichondrialossification groove

Summary The periosteum is a thin layer of connective tissue that covers the outersurface of a bone in all places except at joints (which are protected by articularcartilage). As opposed to bone itself, it has nociceptive nerve endings, making it verysensitive to manipulation. It also provides nourishment in the form of blood supply tothe bone. The periosteum is connected to the bone by strong collagenous fibres calledSharpey’s fibres, which extend to the outer circumferential and interstitial lamellaeof bone. The periosteum consists of an outer ‘‘fibrous layer’’ and inner ‘‘cambiumlayer’’. The fibrous layer contains fibroblasts while the cambium layer containsprogenitor cells which develop into osteoblasts that are responsible for increasingbone width. After a bone fracture the progenitor cells develop into osteoblasts andchondroblasts which are essential to the healing process. This review discusses theanatomy, histology and molecular biology of the periosteum in detail.# 2007 Elsevier Ltd. All rights reserved.

Contents

Historical aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1116Anatomical considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1116Microscopic features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1116Periosteal circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1120

Intrinsic periosteal system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1121Periosteocortical (cortical capillary) anastomoses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1122Musculoperiosteal system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1122Nutritive periosteal system (fascioperiosteal system) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1122

Periosteal bone formation during growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1123Periosteal bone formation in adulthood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1124Periosteal bone resorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1125The perichondrial ossification groove . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1126

* Corresponding author. Tel.: +385 915252372.E-mail addresses: augustin.goran@gmail.com (G. Augustin), anko.antabak@zg.t-com.hr (A. Antabak), slavko.davila@yahoo.com

(S. Davila).

0020–1383/$ — see front matter # 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.injury.2007.05.017

1116 G. Augustin et al.

Extrinsic mechanical effects of the periosteum on the growth plate . . . . . . . . . . . . . . . . . . . . . . . . . . . 1126Conflict of interest statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1127References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1127

Historical aspects

Since the time of Duhamel and John Hunter it hasbeen the belief of anatomists and surgeons that theperiosteum is osteogenic. In 1757 Duhamel andMonceau reflected the periosteum from the boneand fitted around it a silver ring, over which theperiosteum was sewed. After a period of severalmonths the ring was completely covered with boneand from this observation they concluded that theperiosteum secreted bone.35 In the mid 1800s,Dupuytren proposed that the cartilage of fracturecallus originated from periosteum and bone mar-row.36 In 1867 Ollier proved that the deep cellular orosteogenic layer of a free periosteal graft is able toproduce bone. This view was not disputed until in1912 when Sir W. Macewan published his work TheGrowth of Bone in which he described many experi-ments which seemed to demonstrate that the peri-osteum cannot be considered osteogenic, and that itmust be viewed merely as a limiting membrane ofmuch the same nature as the sheath of a muscle orthe capsule of one of the viscera. This observation ofa periosteum as merely a limiting membrane wasconfirmed by the Gallie and Robertson in 1914.44

Then Lacroix in 1945 demonstrated the osteogeniccapability of mature periosteum.65

Figure 1 Cortex (K), periosteum (P) and muscle (M).Collagen fibres (Sharpey’s fibres, blue arrows) penetratefrom periosteum to bone matrix.

Anatomical considerations

The periosteum is specialised fibrous tissue in a formof fibro-vascular membrane. This well vascularisedfibrous sheath, covers the external surface of mostbones and is absent from articular surfaces, tendoninsertions, or sesamoid bone surfaces.60 The peri-osteum and bones are bound together by collagenfibres called Sharpey’s fibres that penetrate intobone. The direction of collagen fibres is determinedby tension forces (Fig. 1). These fibres penetrateentire cortex at the sites exposed to the high tensionforces and the results are tight junctions of tendonsand bones.136 In the region of the diaphyses of longbones the periosteum is thicker (2—3 mm) and easilyseparated from the underlying bone. It is stronglyfused with bones in the metaphyseal and epiphysealregion where it is thinner.

The main feature of children’s bone is to grow,wrappedwith elastic, firm periosteum. This explainswhy childrens’ fractures have some specific biome-

chanical features: bone fractures without the dis-ruption of the periosteum (subperiosteal fractures)or intact periosteum on the concave side of thefracture (greenstick fracture).59 With growth, theperiosteum becomes thinner and loses elasticity andfirmness.91 It is especially compliant on tensileforces and tearing which results in the disruptionof the periosteum in the level of bone fracture inadults. The periosteum is highly vascularised andinnervated and contains large amounts of lymphaticvessels.53 It contains different types of nerves: sen-sory and vasomotor nerves. These vasomotor nervesregulate vessel tone by regulation of precapillarysphincters and capillary blood flow. Pain fibres withnociceptors are highly expressed which explains theintense pain that follow periosteal injuries.74

Microscopic features

Generally, periosteum is composed of an outerfibrous and inner cellular layers and does notsupply epithelial cells, though periosteum hasthe potential to produce collagen.25 The structureof the periosteum in terms of ultrastructure andfunctional organisation was not definitively under-stood until recently. The original division into twoanatomical layers was made by Tonna in 1965, andonly in 1986 Tang and Chai clearly delineatedosteogenic cells of the cambium from fibroblasts(fibrous layer).124,128

The periosteum 1117

Figure 2 The periosteum of sheep tibia. (a) Magnification 250� and (b) magnification 25�. Photomicrograph of normalperiosteum attaching to bone. Periosteum consists of two clearly divided layers: osteogenic, cambium (K) and fibrous (F)layer. Periosteal surface (P) adjacent to the cortex.

Microscopically (Fig. 2), the periosteumconsists ofan outer, fibrous, firm layer (collagen and reticularfibres) and an inner, proliferative layer (cambium)which lies adjacent to bone and contains osteoblastand osteoprogenitor cells (Fig. 3). Cambium is cap-able of: (a) forming normal lamellar bone appositionon cortical bone that grows in width and (b) formingprimary, woven bone after a fracture.54,103,124,129

The outer fibrous layer provides elasticity and flex-ibility, whereas the inner cambium is the osteogeniclayer and contains three or four cell layers, includingosteoblasts and preosteoblastic cells.24,27,119

The first division of the periosteum into threelayers was made by Squier et al.119 in 1990 with the

Figure 3 Periosteal covering of the human femoral midshperiosteal surface comprising the cambium layer stained withRef. [3].

analysis of periosteal morphology of the dog withlight and electron microscopy.

Zone I consists mainly of osteoblasts arranged inthe layer adjacent to the bone surface in a form ofsimple epithelium and a supraosteoblast layer ofsmaller, compact cells.6 Adjacent to primary (imma-ture) bone, during intense synthesis of extracellularmatrix, osteoblasts are cuboidal, arranged as stra-tified epithelium, with basophilic cytoplasm withhigh levels of alkaline phosphatase (Fig. 4).38 Withthe decrease of activity, osteoblasts elongateand basophilic characteristics of the cytoplasmdecrease. The layer over the osteoblasts consistsof small, spindle cells with scarce endoplasmic reti-

aft. Note the abundance of cells (arrowheads) near theMasson trichrome. Magnification 400�, bar = 25 mm. From

1118 G. Augustin et al.

Figure 4 Periosteum of the sheep tibia. Zone I: basophilic osteoprogenitor cells (red arrows) of germinative layer intransition to Zone II (blue line). Zone II: transparent zone with capillaries (yellow arrows) consists of extracellular matrixand fibroblasts. Magnification 25�, bar = 15 mm. Hemalaun-eosin. Imunohistochemical staining with CD 31 and CD 34 (vonWillenbrand factor).

culum that are similar to fibroblasts. These areosteogenic progenitor cells which differentiate intoosteoblasts. Fibrous tissue consists mainly of col-lagen and small amount of elastic fibres.125 Fibro-blasts are scarce and blood vessels are almostcompletely lacking.114 This is the thinnest part ofthe periosteum (also called germinative layer).

Zone II is a relatively transparent zone withcapillaries and amorphous extracellular matrix mak-ing the most voluminous part (Figs. 5 and 6). Thefibroblasts constitute most of the cellular compo-nent and collagen fibres are abundant and bothstructures occupy one quarter of this layer. Fibro-blasts are arranged in thin bunches, thinner than inother layers of the periosteum.119

Blood vessels are numerous in this layer, mostlycapillaries (Figs. 4 and 6). Together with a densecapillary network this layer contains an abundanceof endothelial pericytes.32 Pericytes are poly-morphic cells of mesenchymal origin, which containmultiple, branching cytoplasmic processes that par-

Figure 5 Periosteal division into three zones. Zone II (transpMagnification 25�, bar = 15 mm. Hemalaun-eosin. Imunohistofactor).

tially surround capillaries. Pericytes are found in themicrovasculature of connective tissue, nervous tis-sue, muscle tissue and the lungs.115 These cells havethe ability to contract and hencemay regulate bloodflow in the microvasculature.23 Pericytes may alsofunction as resting stem cells and differentiate intosmooth muscle cells.81 They may also play a regu-latory role in controlling capillary proliferation dur-ing wound healing,31 and support capillaries inmaintaining structural rigidity of the micro-vesselwall.29 Pericytes are cells in physical contact withcapillary endothelial cells, with the ability to differ-entiate into numerous cell types, including osteo-blasts.14,101,133 These cells may serve as asupplementary source of osteoprogenitor cells32

and may be more important in periosteal boneformation due to their greater abundance in peri-osteum20 than in endosteal bone surface apposi-tion.14 Cultured pericytes mineralise in vitro andsynthesise the osteoblast marker, alkaline phospha-tase, as well as bone matrix proteins, including

arent zone) consists of extracellular matrix and fibrolasts.chemical staining with CD 31 and CD 34 (von Willenbrand

The periosteum 1119

Figure 6 Zone II of the sheep tibial periosteum. Zone IIwith capillaries (green arrows). Capillary diameter is5.55—6.49 mm. Magnification 250�, bar = 15 mm. Hema-laun-eosin. Imunohistochemical staining with CD 31 andCD 34 (von Willenbrand factor).

osteocalcin,20 osteonectin, osteopontin, and bonesialoprotein. These cells form an osteogenic tissuethat mimics bone-derived tissue, both spatially andtemporally,38 and responds to osteogenic stimuli,such as BMP and parathyroid hormone.101 Sympa-thetic nervous fibres in this layer are much denserthan in the bone.74 Extracellular matrix and fibro-blasts are less susceptible to histological stainingand this layer of the periosteum is less salient and isbrighter, and together with zone I is called cambium(from Latin, meaning to exchange). The only proteinthat is present in the higher amount in periosteumthan in the bone is periostin.55,72 Predominantly it islocated in the preosteoblasts which secrete perios-tin in the extracellular matrix. The original term forthis protein is OSF-2, and the highest concentrationis found in the disrupted periosteum. The synthesisof periostin is increased four-fold during the first 3days after the fracture.72 The concentrationdecreases with the progression of differentiationof osteogenic progenitor cells and the activity ofosteoblasts. Still the synthesis and the role of theperiostin are unclear. It seems that it is responsiblefor the interaction of cells and extracellular matrix,as a mediator, during mechanical changes in theperiosteum. Also its role is probably in osteoblastdifferentiation.55,90

Zone III consists of numerous fibroblasts withcollagen fibres in scarce extracellular matrix(Fig. 5). The blood vessels are scarce, mostly capil-laries. This zone is easily perceivable because of thehigh amount of collagen fibres and their suscept-ibility for histological staining. The most important

characteristics of collagen are firmness, inextensi-bility and insolubility. Collagen fibres make a net-work of thin fibres in ambiguous directions.103,119

This layer is called ‘fibrous layer’ of periosteum.Zone I is thin in contrast to zones II and III which areseveral fold thicker. These significant quantitativedifferences in the periosteal structure in all threezones are constant despite of the region and thelocation on the bone and indicate persistent peri-osteal microanatomy.4,72,119 Today, it is clear thatthe morphology of the periosteum depends not onlyupon the species but also upon the age. Periostealfibroblast number and fibrous layer thicknessdecrease with age,127 although atrophy of thefibrous layer is less than that of the cambium layer.91

Vessel density throughout the periosteum alsodeclines with age but retains the capacity toincrease when activated by mechanical loading orfracture repair.38 These age-induced changes mayhelp explain why periosteal cells from older subjectsfail to form mineralised nodules in culture,85 andwhy periosteal bone formation rate40 and respon-siveness to hormones and cytokines95 decline withage. During aging the size and the number of thecells decrease while the size and the thickness of thecollagen fibres increase.28 Cellular density of thecambium layer is three-fold higher than the fibrouslayer but the ratio is constant and does not changewith aging. Absolute and relative values of totalperiosteum thickness and the thickness of each layerare decreased.4,72,91 The main feature of the mor-phological changes of the cambium layer duringaging is dramatic decrease127 and elongation38 ofosteoblasts. This reduction in osteoblast numbermay contribute to the apparent atrophy and thin-ning of the cambium layer that occurs with age.91

Periosteal fibroblast number and fibrous layer thick-ness also decrease with age,127 although atrophy ofthe fibrous layer is less than that of the cambiumlayer.38,91 This biologically impaired and reducedperiosteum has small reparatory potential with aslower response rate to stimulation with cytokinesand hormones (longer fracture healing time). Peri-osteal expansion occurs throughout life. The rate ofexpansion is high during puberty,17 slower during theadult years106,117 and in women, accelerated againafter the menopause.1 Independently of otherchanges, expansion of the periosteal surfaceincreases the strength of long bones and decreasesthe risk of fracture.89

Site-specific differences in periosteal anatomy oractivity clearly exist throughout the skeleton. It iswell know that the calvarial periosteum is uniquelyregulated compared to the axial skeleton, and thatcellular periosteum is scarce at the femoral neck.98

The existence of periosteum at the femoral neck is

1120 G. Augustin et al.

Figure 7 Location of intrinsic periosteal system between (G) germinative, cambium and (F) fibrous layer (redarrowheads).

commonly debated. Early observational92,96 andhistological10 studies suggest that human femoralneck lacks a periosteum. The absence of callusformation following femoral neck fractures in adultssupports these observations.37,41,60,92,122 Despitethese studies there are some opposing observationsclaiming that the femoral neck periosteal coveringexists.8,34,98,112 Periosteal cellularity at the femoralneck is significantly lower than in the diaphysealregion even in young adults. Twenty percent of thefemoral neck surface has cellular periosteum whichsuggests that anabolic osteogenic therapies may beeffective in strengthening this clinically relevantsite. Periosteal cells have greater sensitivity tomechanical61 and pharmacological82 stimuli com-pared to marrow cells and even limited cellularperiosteum may be sufficient for enhancing perios-teal apposition. These cells probably do serveto expand the periosteal diameter, as the femoralneck experiences age-associated radial expan-sion.11,106,117 It may be, however, that the limitedquantity of cells limits the rate of expansion, result-ing in less than optimal bone geometry and there-fore elevated fracture risk. Alternatively, these datamay present supporting evidence that the femoralneck exhibits an alternative means of periostealapposition. Previous studies have documented thatboth periosteal calcification and calcified fibrocar-tilage undergo osteonal remodeling.134,140 Althoughthis study did not document any calcified fibrocar-tilage, the abundant periosteal mineralised tissuedid contain individual osteons, clearly separatedfrom the periosteal bone surface, in some regions.Such a mechanism could be an alternative explana-tion for femoral neck periosteal expansion with age.Thus, rather than circumferential lamellae beinglaid down on the periosteal surface and subse-quently remodeled into osteons, as occurs in dia-physeal bone, mineral accumulates separate fromthe periosteal surface with subsequent osteonalremodeling necessary for incorporation into theexisting bone. The highly irregular surface of the

femoral neck, as compared to the relatively smoothperiosteal surface of diaphyseal bone, certainlysupports this hypothesis although further study isnecessary.2

There are few studies that specifically addressthe site-specific differences,4,83,113 yet clear differ-ences in periosteal bone formation rates existamong skeletal sites.

Because of ligament and tendon muscle attach-ments and fibrocartilage on some areas of the peri-osteal surface, periosteal cells are exposed todifferent physical environments in contrast to morefrequently studied endosteal cells, which arebathed in hematopoietic marrow. Compared toendosteal osteoblasts, periosteal osteoblasts exhi-bit greater mechanosensitivity to strain,61 a lowerthreshold of responsiveness to parathyroid hor-mone,82 higher levels of expression of proteins suchas periostin55,90,123 and more oestrogen a recep-tors.21 These differences in threshold sensitivityto physical, hormonal, and mechanical stimulimay underlie the differences in periosteal and endo-steal surface responses to therapy.39 Periosteum hascholinergic sympathetic innervation (Fig. 7). Adultperiosteum contains VIP-immunoreactive fibresassociated with periosteum, as well as catechola-minergic fibres associated with blood vessels.52,53

VIPergic and cholinergic properties are present inthe same fibres.121 Tracing studies indicate thatperiosteal VIP-immunoreactive fibres of the ribsand sternum originate from thoracic sympatheticganglia.53

Periosteal circulation

The arterial supply of the long bones consists of thenutritional arteries and of numerous vessels enter-ing the bone from the periosteum.5,51 The periostalcirculation is an important part of bone vascularisa-tion. The blood supply of the periosteum is derivedfrom four vascular systems.116

The periosteum 1121

Intrinsic periosteal system

The intrinsic periosteal system is located betweenthe fibrous layer and cambium, mostly in zone II(Fig. 7).116 These are terminal branches of nutritiveperiosteal system. These branches form a net of (a)longitudinal blood and lymphatic vessels where thevessels run parallel to the long axis of the bone and(b) circular vessels where the vessels encircle thebone. These vessels interconnect with (c) shortbranches with no predominant direction.58,105,116

Capillaries are the smallest vessels of the bloodcirculatory system and form a complex interlinkingnetwork. The capillary wall is composed of endothe-lial cells, a basement membrane, and occasionalscattered contractile cells called pericytes. A capil-lary consist of one, two or three epithelial cells. Thecapillaries form a dense network of narrow, shorttubesmeasuring from 3 to 4 mm in diameter (i.e. halfthe diameter of red blood cells) up to 30—40 mm(these large blood spaces are usually known as sinu-soids). On average, capillaries have a diameter of 6—8 mm and are approximately 750 mm to 1 mm long.

Figure 8 Vascular supply of cortical bone. Periostocortical ansupply.

Their average volume is 40 mm3 and blood flow 0.1—0.5 m/s.43 Oxygen rich blood flows from arteriolesinto the capillary bed and deoxygenated blood istransported from capillaries to venules. Pressuredifference forces the blood from the capillary bedto venules. Blood from arterioles travels to terminalarterioles, also calledmetaarterioles.Metaarterioleshave a discontinuous layer of smooth muscle cells (incontrast to arterioles). Capillary density in tissue isdirectly proportional to metabolic activity of thetissue. Capillary density is the highest in the brain,kidneys, liver, heart andmuscles and low in bones, fatand fibrous tissue. There are no exact data aboutcapillary density in the periosteum. Periosteal veinshave a thinner vessel wall with a higher quantity ofcollagen fibres than arteries often leading to luminalcollapse during microscopic examination. The layerscannot be strictly differentiated. Periosteal veinscontain lesser amount of elastin than periostealarteries and these fibres are scattered with no pre-dominant direction. Lymphatic vessels have thinnerwalls than veins and lack distinct layers. The lumen isirregular and its wall consists of endothelial

astomosis connects periosteal and nutritional artery blood

1122 G. Augustin et al.

cells surrounded by fibrous tissue. Only larger lym-phatic vessels have a muscular layer that containssmooth muscle cells in both longitudinal and circulardirection.

Periosteocortical (cortical capillary)anastomoses

Periosteal arterioles run longitudinally without adecrease in the diameter and give branches thatare directed to bone. Normally, these branches areperpendicular to this main periosteal vessel. In theouter third of the cortex, in the nearest centralcanal of external osteons they anastomose with themedullary system (Fig. 8). The number and thediameter of periosteocortical anastomoses increaseprogressively from the diaphysis to the metaphy-sis.131 In some cases branches of periosteal arteriesand arterioles pass through the whole cortex andsupply sinusoids and other vessels of the medullarysystem.75,102 This system represents a direct con-nection of periosteal blood supply with the nutri-tional arteries. These periosteocortical arterieshave concomitant veins, a system characteristicfor all mammals that is responsible for survival ofouter cortex when nutritive or medullary bloodsupply is diminished or blocked.

Musculoperiosteal system

Musculoperiosteal anastomoses with surroundingmuscle have a significant role in periosteal callusformation.84,141 Their role is evenmore significant inconditions of insufficient intrinsic (nutritive) peri-osteal circulation. The epimysium is well nourishedand fused with the fibrous layer of the periosteum inaway that pulling themuscle from the bone resultedin stripping of the periosteum. The blood supply ofthe epimysium is derived from two sources: themainmuscular branch (Fig. 9a) and branches of segmen-

Figure 9 Musculoperiosteal anastomoses: (a) main muscepimysium.

tal arteries (Fig. 9b). Musculoperiosteal anasto-moses can also be divided (according to the sizeof the vessels) into musculonutritive arteries withconcomitant veins and less valuable anastomoses ata capillary level.137

Nutritive periosteal system(fascioperiosteal system)

The periosteum is vascularised by several segmen-tal arteries. Distribution of these segmentalarteries differs from bone to bone because of dif-ferent insertion of tendons and fascia.75 The nutri-tive arterial system is accompanied by a venoussystem. Every artery is accompanied by two veins.99

As an example, periosteal circulation of human tibiais presented in detail.80 Nutritive periosteal circu-lation of the human tibia is divided into four regions(Fig. 10). These regions are connected at thecapillary level. Seventy to 80% of cortical bloodflow is delivered by periosteal circulation and90—100% of venous blood is drained by periostealcirculation depending on the anatomic and boneregion.22,27,116,130,139

The anterolateral sector of the proximal fifth ofthe tibial periosteum is nourished by a recurrentbranch of the anterior tibial artery (ATA). Anasto-moseswere found proximallywith the lateral inferiorgeniculate artery with branches of the medial infer-ior geniculate artery on the tibial tuberosity andunder the distal part of the patellar ligament. Inthe proximal fifth the latter artery supplies the ante-romedial side of the tibia and the medial part of thedorsal side. The lateral part of the dorsal side of theupper fifth of the tibia is nourished by the recurrentposterior tibial artery coming from the ATA. At thelateral condyle the supply is supported by the lateralinferior genicular artery from the popliteal artery.The lateral surface of the proximal diaphysis is nour-ished by periosteal branches from the ATA, mainly

ular branch and (b) branches of segmental artery for

The periosteum 1123

Figure 10 Periosteal arteries of the tibia. The schemedemonstrates sectors being supplied by one or multiplearteries which nourish the periosteum and the outer partof the cortex. ATA: anterior tibial artery; ARTA: anteriorrecurrent tibial artery; FA: fibular artery; ILGA: inferiorlateral genicular artery; IMGA: inferor medial genicularartery, PTA: posterior tibial artery; PRTA: posterior recur-rent tibial artery.

running in a transverse or slightly ascending butseldom in a descending direction. There are 5—1280

or 2—899 of these branches. Both authors found thatthe arterial vessels of the periosteum are accompa-nied by two veins. The periosteal branches of theproximal diaphysis partly extend to the medial sur-face where they merge with periosteal branches ofthe posterior tibial artery (PTA). In addition, the PTAgives support to the nutritient artery for supplyingthe posterior surface. At the level of proximal dia-physis there are vertical and also circular segmentalanastomoses of semicircular branches from the ATAand PTA. The nutritient tibial artery often arises fromthe PTA, seldom from the ATA.

The distal diaphysis is exclusively supplied bybranches of the ATA, which form a capillary networkwith circular and vertical anastomoses. The lateralsurface is nourished through periosteal brancheswhich merge on the medial surface with periostealperforators. The latter originate from the ATA andsupply the posterior surface before reaching themedial surface. The total number of existing peri-osteal perforators is two to five.48,80

In two of three cases, the periosteum of thelateral surface around the fibular notch at the cau-dal fifth of the tibia is nourished by perforators ofthe fibular artery (FA) which is branching into anascending and descending branch. The other partof the lateral side is supplied by the periostealbranches from the ATA. In one-third, the wholelateral area is nourished by branches of the ATA.In cases when perforators are not developed the ATAgives off a strong branch which copies the course ofthe first mentioned artery. The variations of theperiosteal perforators are well documented byHyrtl.56 The caudal fifth of the posterior surface ismainly supplied by a transversally running periostealbranch of the FA which splits up into multiple smallvessels. These capillaries reach the medial surfaceand anastomose with periosteal branches comingfrom the lateral surface. Additionally there arebranches of the PTA for the supply of the caudalarea of the periosteum of the posterior surface. Thelateral surface is chiefly nourished by branches ofthe ATA, whereas the posterior surface is supplied bybranches arising from both ATA and PTA and minorparts by the FA and the inferior medial and lateralgenicular arteries. Thus, the lateral, as well as theposterior surface, are supplied by direct branches ofthe major arteries of the lower leg. In contrast themedial surface is nourished only by vessels comingfrom the lateral and posterior surface, respectively.

From this anatomical consideration it is obviousthat the anterior tibial artery is of great importancefor the arterial supply of the tibial periosteum withan autonomous region at the distal diaphysis. Thismedial aspect of the three-quarters of the tibialperiosteum is nourished only by small capillarybranches of the anterior tibial artery. This is of asignificant clinical importance because this area hasa high incidence of pseudoarthrosis.16 The perios-teal circulation represents a significant part of tibialvascularisation and periosteal disruption impairsand diminishes the cortical blood supply.64,138 Inshort: an osteocorticotomy should neither be madeat the distal diaphysis nor in the upper part of theproximal diaphysis, because of disruption of thenutritient artery.

Periosteal bone formation duringgrowth

The growth plate components go through a sequen-tial process of cell proliferation, extracellularmatrix synthesis, cellular hypertrophy, matrixmineralisation, localised vascular invasion andapoptosis. These highly coordinated activities leadto longitudinal bone growth and bone formation

1124 G. Augustin et al.

Figure 11 Apposition of bone around periosteal vessels presented in four phases.

at the physeal—metaphyseal region by the mechan-ism of enchondral ossification. The growth cartilagereplenishes itself through the germinal zone and iscontinually replaced by bone at the physeal—meta-physeal junction. The length of the entire boneincreases; the physes at either end are displacedprogressively further away from the centre of thebone, and the physis itself maintains the sameheight throughout the growth period. At the sametime, there is radial growth of the diaphysis andparts of the metaphysis caused by direct appositionof cortical bone by osteoblasts from the inner cam-bial layer of the periosteum (intramembranous boneformation) (Fig. 3). Apposition of bone around andbetween periosteal vessels results in formation ofperiosteal ridges, which, in subsequent phases unitearound periosteal vessels thus producing Haversiancanal, osteons (Fig. 11).

There are 16 stages and with several additionalsubstages of long bone and epiphyseal developmentthat represent the timing and coordination of thegrowth process.42 Periosteal bone apposition is acardinal feature of skeletal development. Long bonesgrow wider as they grow taller, and it is commonlyrecognised that there is wide individual variation inthis process (‘‘big-boned’’ versus ‘‘small-boned’’). Infact, after adjustment for height orweight, there is awide range in bone size, indicating that periostealapposition is affected by a distinct set of determi-nants.76 In humans, some of the most obvious aregender (males > females) and race (blacks > whi-whites > asians).73,86,87 Geographical differences inbone size are also marked, even within racial bound-aries.30 Disorders of bone size expansion, such aschildhood illness at critical periods of development,have been proposed to contribute to the variation inadult bone strength and fracture likelihood.17 Animalstudies support a positive effect of androgens anda negative effect of oestrogens on periostealbone formation rates.132 At puberty in males, the

periosteum expands due to androgen action withlittle change in the endocortical (medullary dia-meter), so that cortical width increases. At pubertyin females, the periosteal expansion ceases. Endo-cortical (medullary) diameter decreases as the endo-cortical bone formation occurs. This endocorticalcontraction contributes 25% of the total corticalthickness.45 Males and females have the same cor-tical thickness but the bone diameter is greater inmales, conferring greater breaking strength. Thus,reduced cortical thickness may be the result ofexcessive radial expansion of the endocortical sur-face relative to the periosteal surface before andduring puberty. This may be due to either increasedresorption and/or reduced bone formation. A role forinsulin-like growth factor 1 (IGF-1) in the regulationof periosteal apposition has long been postulated,especially in concert with sex steroids during pub-erty.15 Many other factors are probably involved aswell. For instance, mechanical force applied in vivoinduces the expression of a variety of genes in theperiosteum78 anda rapid transformation of quiescentperiosteal surfaces to those on which bone formationoccurs.94 In fact, it has been suggested that themechanical loading environment is a primary mod-ulator of periosteal apposition.135 Also, genetic ana-lyses have implicated a variety of chromosomalregions (and genes) in the control of bone size inhumans and mice.62,63 In the light of their effects onbone formation in other skeletal compartments,other lifestyle andenvironmental factors (e.g.,nutri-tion, alcohol and tobacco use)108,135 may modulateperiosteal bone formation, but their effects have notbeen well examined.

Periosteal bone formation in adulthood

Animal studies, from rodents to primates, documentthe persistence of periosteal bone formation

The periosteum 1125

throughout life, albeit at a slower rate than duringgrowth, and there is the strong suggestion that bonesize may continue to increase during adulthood. Atpresent, most evaluations of change in bone size inhumans are small and cross-sectional and are subjectto limited power and cohort effects,69,77,106,107 butsome longitudinal studies support the increase inbone size with age.12,13,69 Mechanical events haveusually been assumed to underlie the observationthat bone size can increase inadults.70Oneattractivemodel posits that gradual endosteal bone loss withaging leads to cortical thinning and thus more bend-ing stress on theouter surface of bone, in turn leadingto the stimulation of periosteal bone apposition as abiomechanical compensation.13,69 On the otherhand, periosteal expansion also seems to occur inearly adulthood, at a timewhen endosteal resorptionhas not begun, suggesting that events at the perios-teum do not only reflect mechanical influences.69

Moreover, less loaded bones (metacarpal, skull) alsoexperience periosteal expansion in adults. Althoughprobably important, the relative role of mechanicalforces in the determination of periosteal responsesare unknown, as during growth, other factors mayalso influencebone sizeduringaging (nutrition, endo-crine factors, lifestyle variation, etc.).

The periosteal effects of selective oestrogenreceptor modulators or nongenotropic oestrogensare unclear. Might other factors known to adverselyaffect osteoblast viability or bone formation (e.g.,glucocorticoids, alcohol, renal dysfunction, vitaminD deficiency, etc.) contribute to a failure of perios-teal expansion and increased fracture propensity?Conversely, stimulators of periosteal bone forma-tion should offer new opportunities to improve bonestrength. For instance, parathyroid hormone ther-apy (and even mild hyperparathyroidism) mayincrease bone size and strength through complexeffects on bone forming elements on the periostealsurface.93 If the postulated sex difference in bonesize is a result of androgen action, as some animalstudies suggest,88 it lends support to the potentialuse of androgenic compounds, acting through aneffect on bone size, in the prevention of age-relatedfracture. The emergence of the periosteum as atarget for pharmacotherapeutics, for instance withparathyroid hormone or androgenic agents, pro-mises to alter approaches to fracture risk reduction.

In most endosteal indices of bone adaptationendosteal adaptation of both the loaded and controltibiae is identical. Moreover, endosteal adaptationdid not increase with strain rate. These results ofabsence of large endosteal adaptive responses, inthe presence of large periosteal adaptive responses,are consistent in the literature.46,71,79,120 That theadaptive response was largely confined to the peri-

osteal surface has significant implications for resis-tance to bending. For a given amount of bone, bonelocalised on the surface furthest away from theneutral axis of bending can most effectively resistbending by efficiently elevating the cross-sectionalmoment of inertia, and that surface represented theperiosteal surface.67

Periosteal bone resorption

Despite much recent attention to the potentialimportance of periosteal bone formation, therehas been very little consideration of the occurrenceor importance of osteoclastic resorption on the sur-face of bone. Periosteal resorption is somehow aheretical concept. It is frequently assumed thatthere is an inexorable expansion of the periosteumthrough isolated new bone formation, or modeling,and that resorption is rare on the periosteal surface.However, it is unequivocal that periosteal resorptionoccurs in some situations. Parfitt93 has pointed outthe drift in bone surfaces that accompanies growth,including the dramatic resorption that must occuron the medial ileal surfaces during pelvic enlarge-ment. Analogous events occur in other flat bones(mandible, skull, scapulae). Similarly, longitudinalgrowth of appendicular bones is accompanied byrapid periosteal resorption of the metaphysis(‘‘waisting’’) to create the more slender diaphysis.Essentially, the periosteal radius (and size) of thebone shrinks during that process100 and strength ismaintained by simultaneous endocortical boneapposition to form a thickened cortex. While thereis simply very little information concerning thepresence or absence of resorption on most adultperiosteal surfaces, Epker and Frost39 actuallydescribed periosteal resorption (and remodeling)in adults on the surface of ribs almost 40 yearsago and Balena et al.9 examined periosteal remo-deling on the surface of the ileum in women. Infurther studies, the extent of eroded periostealsurface equalled that on the endocortical surface(although there were fewer osteoclasts present onthe periosteal surface and in general the remodelingrate was considered much slower than on the endo-steal surfaces). It was estimated that the boneformation by the periosteum occurred on previouslyeroded surfaces–—in other words, bone formationoccurred only as part of remodeling and did notresult frommodeling. Virtually no other informationexists concerning the nature of periosteal remodel-ing events or their impact on bone health. Never-theless, there are clear illustrations of thisphenomenon. For instance, the alveolar ridge ofthe mandible can be rapidly lost after tooth loss,

1126 G. Augustin et al.

which reduces the mechanical forces on it.7

One example of how the disease can affect theperiosteum is that hyperparathyroidism has beenclassically associated with ‘‘subperiosteal’’ boneresorption. In severe forms, a reduction in miner-alised bone size (classically of the phalanges) can beobserved radiologically. Whether some or all of thisosteoclastic activity originates on the periostealsurface or occurs as a result of exuberant Haversianremodeling (tunnelling) within the subperiostealcortex is unclear. However, the result is a reductionof the effective circumference of bone and arguablyits resistance to fracture. To what extent theselosses of periosteal bone contribute to the increasedfracture risk of advanced hyperparathyroidism isunexplored. In summary, the circumference andto some extent the biomechanical strength of boneshould be considered a function of the balancebetween periosteal bone formation and resorption.However, the rate of periosteal remodelling and thefactors that influence it at critical skeletal sites(vertebrae, proximal femur) are unknown.89

The perichondrial ossification groove

The perichondrial ossification groove of Ranvierthat contains the circumferential bony ring ofLacroix, sometimes referred to as the ‘‘bone bark,’’surrounds the periphery of the growth plate as adifferentiated cell and tissue structure with fibresarranged in three directions: vertically, circumfer-entially, and obliquely. Its components function tocontribute to latitudinal growth of the growthplate by appositional addition of chondrocytes,to contain mechanically and support the physesby its outer fibrous sheath, inner osteogenic layer,and bony ring and to elongate cortical intramem-branous bone formation by osteoprogenitorcells.18,19,50,57,66,110,111 The groove of Ranvier sur-rounds the growth plate and is the specific struc-tural and functional region where the cartilage ofthe endochondral sequence meets the two-layeredperiosteum of the intramembranous sequence. Theouter layer of the periosteum is continuous from thediaphysis toward both bone ends enclosing themetaphysis, the growth plate, and inserting intothe epiphyseal cartilage beyond the physis. Theinner layer of the periosteum with osteogenic cellsalso covers the growth plate and forms an accumu-lation of cells at the depth of the groove adjacent tobut separate from the cells of the germinal andproliferating layers of the physis. The groove regionconsists of an outer layer formed by fibroblasts andcollagen fibres, which is a continuation of the outerfibrous layer of the periosteum; undifferentiated

loosely packed cells, which are cartilage precur-sors; and a group of densely packed cells thatmature into osteoblasts. In those bones or partsof bones in which the diameter of the metaphysis isthe same as that of the adjacent diaphysis (i.e.,there is no cut-back zone), the inner, cambial layerof the periosteum is continuous into the groove, asis the cortex.Where the cutback zone is prominent,the inner cambial layer and the bone ring of theossification groove are discontinuous with the innercambial layer of the periosteum and the diaphysealcortex. The outer fibrous layer is always continuousand serves as a fibroelastic sheath connecting theepiphyseal cartilage at one end of a bone to theepiphyseal cartilage at the other end and enclosingboth physes. The increase in the transverse dia-meter of the physis is achieved by interstitialgrowth in the resting cell layer50,68,104,110 and appo-sitional growth from the region of loosely packedcells (perichondrium) of the groove.66,110,118,126,127

Extrinsic mechanical effects of theperiosteum on the growth plate

The periosteum has an essential role in the forma-tion of cortical bone by its inner osteogenic (cam-bial) layer. The metaphyseal cortical bone is formedby the coalescence of peripheral endochondral tra-becular bone from the physis with intramembranousbone from the inner osteogenic layer of the perios-teum.26

The outer fibrous layer of the periosteum coversnot only diaphyses and metaphyses but also sur-rounds and mechanically supports the epiphysealregions of the growing bone, particularly in Ran-vier’s groove, and eventually attaches into the epi-physeal cartilage beyond the physis. The periostealsleeve has a strong fibroelastic mechanical effect onthe physis. Circumferential cutting of the perios-teum reduces the force by 80% needed to produceepiphysiolysis in rats whereas its partial section inthe proximal medial tibia causes valgus deforma-tion.111 Haasbeek et al.47 showed in two clinicalcases that angular deformations occur when theperiosteum is thickened adjacent to the physis.Dimitriou et al.33 compared the effects of surgicallyinduced longitudinal and transverse sectioning ofthe periosteum, and observed that only the latterincreased longitudinal growth of the long bones.These experimental observations support themechanical theory that a reduction of tension onthe periphyseal region has a beneficial effect ongrowth whereas increased tension slows growth.After removal of the periosteum of the diaphysisin rats, no notable differences were observed in the

The periosteum 1127

heights of the resting, proliferating, and hyper-trophic cell layers or in cell proliferation and therate of longitudinal growth, in comparison withcontrol groups.49 Growth stimulation in childrenafter femoral shaft fractures is considered to becaused by increased periosteal and periphyseal vas-cularity affecting the entire bone.109

Conflict of interest statement

All the authors declare that there are no financialand personal relationships with other people, ororganisations, and that there are no conflicts ofinterest of any kind.

References

1. Alhlborg H, Johnell O, Turner C, et al. Bone loss and bonesize after menopause. N Engl J Med 2003;349:327—34.

2. Allen MR, Burr DB. Human femoral neck has less cellularperiosteum, and more mineralized periosteum, thanfemoral diaphyseal bone. Bone 2005;36:311—6.

3. Allen MR, Hock JM, Burr DB. Periosteum: biology, regulation,and response to osteoporosis therapies. Bone 2004;35:1003—12.

4. Anderson CD, Danylchuk K. Bone-remodeling rates of thebeagle: a comparison between different sites on the samerib. Am J Vet Res 1978;39:1763—5.

5. Anseroff NJ. Die arterien der langen Knochen des Menschen.Z Anat EntwGesch 1934;103:793—812.

6. Aubin J, Triffit J. Mesenchymal stem cells and osteoblastdifferentiation. In: Bilezikian J, Raisz LG, Rodan GA, edi-tors. Principles of bone biology. San Diego: Academic Press;2002. p. 59—81.

7. August M, Kaban LB. The aging maxillofacial skeleton. In:Rosen CJ, Glowacki J, Bilezikian JP, editors. The agingskeleton. San Diego: Academic Press; 1999. p. 359—71.

8. Bagi CM, Wilkie D, Georgelos K, et al. Morphological andstructural characteristics of the proximal femur in humanand rat. Bone 1997;21:261—7.

9. BalenaR,ShihM,ParfittAM.Bone resorptionand formationonthe periosteal envelope of the ilium: a histomorphometricstudy in healthy women. J Bone Miner Res 1992;7:1475—82.

10. Banks H. Healing of the femoral neck fracture. In: Proceed-ings of the conference on aseptic necrosis of the femoralhead; 1964. p. 465—82.

11. Beck TJ, Ruff CB, Bissessur K. Age-related changes in femalefemoral neck geometry: implications for bone strength.Calcif Tissue Int 1993;53(Suppl. 1):S41—6.

12. Beck TJ, Oreskovic TL, Stone KL, et al. Structural adaptionto changing skeletal load in the progression toward hipfragility: the study of osteoporotic fractures. J Bone MinerRes 2001;16:1108—19.

13. Beck TJ, Stone KL, Oreskovic TL, et al. Effects of current anddiscontinued estrogen replacement therapy on hip struc-tural geometry: the study of osteoporotic fractures. J BoneMiner Res 2001;16:2103—10.

14. Bianco P, Riminucci M, Gronthos S, Robey PG. Bone marrowstromal stem cells: nature, biology, and potential applica-tions. Stem Cells 2001;19:180—92.

15. Bikle D, Majumdar S, Laib A, et al. The skeletal structure ofinsulin-like growth factor I-deficient mice. J Bone Miner Res2001;16:2320—9.

16. Boer FC, Patka P, Bakker FC, Haarman HJ. Current conceptsof fracture healing, delayed unions und nonunions. OsteoTrauma Care 2002;10:1—7.

17. Bradney M, Karlsson MK, Duan Y, et al. Heterogeneity in thegrowth of the axial and appendicular skeleton in boys:implications for the pathogenesis of bone fragility inmen. J Bone Miner Res 2000;15:1871—8.

18. Brighton CT. Structure and function of the growth plate. ClinOrthop 1978;136:22—32.

19. Brighton CT, Magnuson PB, Iannotti JP. Growth and devel-opment of bone: the growth plate and its abnormalities. In:Instructional course lectures. Rosemont: American Academyof Orthopedic Surgeons; 1992. p. 105.

20. Brighton CT, Lorich DG, Kupcha R, et al. The pericyte as apossible osteoblast progenitor cell. Clin Orthop 1992;275:287—99.

21. Bord S, Horner A, Beavan S, Compston J. Estrogen receptorsalpha and beta are differentially expressed in developinghuman bone. J Clin Endocrinol Metab 2001;86:2309—14.

22. Brookes M. The blood supply of bone. London: Butter-worth’s; 1984. p. 88.

23. Bruns RR, Palade GE. Studies on blood capillaries. I. Generalorganization of blood capillaries in muscle. J Cell Biol 1968;37:244—76.

24. Buckwalter JA, Glimcher MJ, Cooper RR, Recker R. Bonebiology. I. Structure, blood supply, cells, matrix and miner-alization. In: Pritchard DJ, editor. Instructional courselectures, vol. 45. Rosemont, IL: American Academy ofOrthopaedic Surgeons; 1996. p. 371—86.

25. Burstein FD, Canalis RF, Canalis EM, et al. Scanning electronmicroscopy and gel electrophoresis of vascularized perios-teal autografts. Plast Reconstr Surg 1989;83:500—10.

26. Cadet ER, Gafni RI, McCarthy EF, et al. Mechanisms respon-sible for longitudinal growth of the cortex: coalescence oftrabecular bone into cortical bone. J Bone Joint Surg A 2003;85:1739—48.

27. Chanavaz M. Anatomy and histophysiology of the perios-teum: quantification of the periosteal blood supply to theadjacent bone with 85Sr and gamma spectrometry. J OralImplantol 1995;21:214—9.

28. Chong DA, Evans CA, Heeley JD. Morphology and maturationof the periosteum of the rat mandible. Arch Oral Biol1982;27:777—85.

29. Cogan DG, Toussaint D, Kuwabara T. Retinal vascular pat-terns. IV. Diabetic retinopathy. J Diabet Retinopathy ArchOpthalmol 1961;66:366—78.

30. Crabtree N, Lunt M, Holt G, et al. Hip geometry, bonemineral distribution, and bone strength in European menand women: the EPOS study. Bone 2000;27:151—9.

31. Crocker DJ, Murad TM, Geer JC. Role of the pericyte inwound healing. An ultrastructural study. Exp Mol Pathol1970;13:51—65.

32. Diaz-Flores L, Gutierrez R, Lopez-Alonso A, et al. Pericytesas a supplementary source of osteoblasts in periostealosteogenesis. Clin Orthop 1992;280—6.

33. Dimitriou CG, Kapetanos GA, Symeonides PP. The effect ofpartial periosteal division on growth of long bones: anexperimental study in rabbits. Clin Orthop 1988;236:265—9.

34. Dixon T, Benjamin J, Lund P, et al. Femoral neck buttressing:a radiographic and histologic analysis. Skelet Radiol 2000;29:587—92.

35. Duhamen D, Monceau H. Lettre sur la formation des os dansles animaux, et du bois dans les arbres. Rec period Obs MedChir Pharm 1757;7:153—60.

1128 G. Augustin et al.

36. Dupuytren C. On the injuries and diseases of bones: beingselections from the collected edition of the clinical lecturesof baron Dupuytren. London: Sydenham Society; 1847.

37. Einhorn T. Bone strength: the bottom line. Calcif Tissue Int1992;51:333—9.

38. Ellender G, Feik SA, Carach BJ. Periosteal structure anddevelopment in a rat caudal vertebra. J Anat 1988;158:173—87.

39. Epker BN, Frost HM. A histological study of remodeling at theperiosteal, haversian canal, cortical endosteal, and trabe-cular endosteal surfaces in human rib. Anat Rec 1965;152:129—35.

40. Epker BN, Frost HM. Periosteal appositional bone growthfrom age two to age seventy in man. A tetracycline evalua-tion. Anat Rec 1966;154:573—7.

41. Ferretti JL, Frost HM, Gasser JA, et al. Perspectives onosteoporosis research: its focus and some insights from anew paradigm. Calcif Tissue Int 1995;57:399—404.

42. Forriol F, Shapiro F. Bone development: interaction of mole-cular components and biophysical forces. Clin Orthop RelatRes 2005;432:14—33.

43. Freitas RA. Nanomedicine, Vol. I: Basic capabilities. George-town: Landes Bioscience; 1999. p. 5.

44. Gallie WE, Robertson DE. The periosteum. Can Med Assoc J1914;41:33—6.

45. Garn SM. The earlier gain and later loss of cortical bone.Springfield: C.C. Thomas; 1970.

46. Gross TS, Srinivasan S, Liu CC, Clemens TL, Bain SD. Non-invasive loading of the murine tibia: an in vivo model for thestudy of mechanotransduction. J Bone Miner Res 2002;17:493—501.

47. Haasbeek JF, Rang MC, Blackburn N. Periosteal tether caus-ing angular growth deformity: report of two clinical casesand an experimental model. J Pediatr Orthop 1995;15:677—81.

48. Henle J. Handbuch der Gefaßlehre des Menschen. Braunsch-weig: Vierweg; 1868.

49. Hernandez JA, Serrano S, Marinoso ML, et al. Bone growthand modeling changes induced by periosteal stripping in therat. Clin Orthop 1995;320:211—9.

50. Hert J. Growth of the epiphyseal plate in circumference.Acta Anat 1976;82:420—36.

51. Hert J, Hladokova J. Die Gefaßlehre des Menschen. In: LangJ, Wachmuth W, editors. Praktische anatomie I4, bein undstatik. Berlin: Springer; 1972.

52. Hill EL, Elde R. Distribution of cGRP-, VIP-, SP-, and NPYim-munoreactive nerves in the periosteum of the rat. CellTissue Res 1991;264:469—80.

53. Hohmann EL, Elde RP, Rysavy JA, et al. Inervation of peri-osteum and bone by sympathetic vasoctive intestinal pep-tide-containing nerve fibres. Science 1986;232:868—70.

54. Holtrop ME. The ultrastructure of bone. Ann Clin Lab Sci1975;5:264—71.

55. Horiuchi K, Amizuka N, Takeshita S, et al. Identification andcharacterization of a novel protein, periostin, withrestricted expression o periosteum and periodontal liga-ment and increased expression by transforming growthfactor beta. J Bone Miner Res 1999;14:1239—49.

56. Hyrtl J. 7th ed., Handbuch der topographischen Anatomieund ihrer praktisch medizinisch-chirurgischen Anwendun-gen, vol. 2, 7th ed. Wien: Braumuller; 1882.

57. Iannotti JP. Growth plate physiology and pathology. OrthopClin North Am 1990;21:1—17.

58. Iwaku F. Microvasculature of bone and bone marrow. In:Motta PM, Murakami T, Fujita H, editors. Scanning electronmicroscopy of vascular casts: methods and applications.Amsterdam: Kluwer Academic Publishers; 1992. p. 72—86.

59. Jacobsten FS. Periosteum. Its relation to pediatric frac-tures. J Pediatr Orthop B 1997;6:84—90.

60. Jee WS. Integrated bone tissue physiology: anatomy andphysiology. In: Cowin S, editor. Bone mechanics handbook.Boca Raton7: CRC Press; 2001.

61. Jones DB, Nolte H, Scholubbers JG, et al. Biochemical signaltransduction of mechanical strain in osteoblast-like cells.Biomaterials 1991;12:101—10.

62. Turner R, Skinner L, et al. Mapping quantitative trait locithat influence femoral cross sectional area in mice. J BoneMiner Res 2002;17:1752—60.

63. Koller DL, Liu G, Econs MJ, et al. Genome screen forquantitative trait loci underlying normal variation infemoral structure. J Bone Miner Res 2001;16:985—91.

64. Kowalski MJ, Schemitsch EH, Kregor PJ, et al. Effect ofperiosteal stripping on cortical bone perfusion: a laserDoppler study in sheep. Calcif Tissue Int 1996;59:24—6.

65. Lacroix P. Recherches experimentales sur l’osteogeneseperiostique. Arch Biol 1945;56:185—97.

66. Lacroix P. The organization of bone. NewYork: McGraw-Hill;1951.

67. LaMothe JM, Hamilton NH, Zernicke RF. Strain rate influ-ences periosteal adaptation in mature bone. Med Eng Phys2005;27:277—84.

68. Langenskiold A, Heikel HVA, Nevalainen T, et al. Regenera-tion of the growth plate. Acta Anat 1989;134:113—23.

69. Lazenby RA. Continuing pariosteal apposition. I. Documen-tation, hypotheses, and interpretaton. Am J Phys Anthropol1990;82:461—72.

70. Lazenby RA. Continuing periosteal apposition. II. The sig-nificance of peak bone mass, strain equilibrium, and age-related activity differentials for mechanical compensationin human tubular bones. Am J Phys Anthropol 1990;82:478—84.

71. Lee KC, Maxwell A, Lanyon LE. Validation of a technique forstudying functional adaptation of the mouse ulna inresponse to mechanical loading. Bone 2002;31:407—12.

72. Litvin J, Selim AH, Montgomery M, et al. Expression andfunction of periostin-isoforms in bone. J Cell Biochem2004;92:1044—106.

73. Looker AC, Beck TJ, Orwoll ES. Does body size account forgender differences in femur bone density and geometry? JBone Miner Res 2001;16:1291—9.

74. Mach DB, Rogers SD, Sabino MC, et al. Originis of skeletalpain; sensory and sympathetic innervation of the mousefemur. Neuroscience 2002;13:155—66.

75. Macnab I, Haas WG. The role of periosteal blood supply inthe healing of fractures of tibia. Clin Orth Relat Res1974;105:27—33.

76. Marshall LM, Lang TF, Cauley JA, et al. Femoral bone sizeamong older men and its relation to lean and fat mass. JBone Miner Res 2002;17(Suppl 1):S264.

77. Martin RB, Atkinson PJ. Age and sex-related changes in thestructure and strength of the human femoral shaft. J Bio-mech 1977;10:223—31.

78. Matsumoto T, Nakayama K, Kodama Y, et al. Effect ofmechanical unloading and reloading on periosteal boneformation and gene expression in tail-suspended rapidlygrowing rats. Bone 1998;22:89S—93S.

79. Meade JB, Cowin SC, Klawitter JJ, et al. Bone remodelingdue to continuously applied loads. Calcif Tissue Int 1984;36(Suppl 1):S25—30.

80. Menck J, Bertram C, LierseW. Sectorial angioarchitecture ofthe human tibia. Acta Anat (Basel) 1992;143:67—73.

81. Meyrick B, Reid L. Ultrastructural features of the distendedpulmonary arteries of the normal rat. Anat Rec 1979;193:71—97.

The periosteum 1129

82. Midura RJ, Su X, Morcuende JA, et al. Parathyroid hormonerapidly stimulates hyaluronan synthesis by periosteal osteo-blasts in the tibial diaphysis of the growing rat. J Biol ChemDec 2003;278:51462—8.

83. Miller SC, Bowman BM, Miller MA, et al. Calcium absorptionand osseous organ-, tissue-, and envelope-specific changesfollowing ovariectomy in rats. Bone 1991;12:439—46.

84. Muller ME, Allgower M, Schneider R, Willenegger H. Technikder operativen Frakturenbechandlung. Berlin: Springer;1963. p. 38.

85. Nakahara H, Goldberg VM, Caplan AI. Culture-expandedhuman periosteal-derived cells exhibit osteochondralpotential in vivo. J Orthop Res 1991;9:465—76.

86. Nakamura T, Turner CH, Yoshikawa T, et al. Do variations inhip geometry explain differences in hip fracture riskbetween Japanese and white Americans? J Bone MinerRes 1994;9:1071—6.

87. Nelson DA, Barondess DA, Hendrix SL, Beck TJ. Cross-sec-tional geometry, bone strength, and bone mass in theproximal femur in black and white postmenpausal women.J Bone Miner Res 2000;15:1992—7.

88. Orwoll ES. Androgens: basic biology and clinical implication.Calcif Tissue Int 2001;69:185—8.

89. Orwoll ES. Toward an expanded understanding of the role ofthe periosteum in skeletal health. J Bone Miner Res2003;18:949—54.

90. Oshima A, Tanabe H, Yan T, et al. A novel mechanism for theregulation of osteoblast differentiation: transcription ofperiostin, a member of the fasciclin I family, is regulatedby the bHLH transcription factor, twist. J Cell Biochem2002;86:792—804.

91. O’Driscoll SW, Saris DB, Ito Y, Fitzimmons JS. The chondro-genic potential of periosteum decreases with age. J OrthopRes 2001;19:95—103.

92. Pankovich A. Primary internal fixation of femoral neckfractures. Arch Surg 1975;110:20—6.

93. Parfitt AM. Parathyroid hormone and periosteal bone expan-sion. J Bone Miner Res 2002;17:1741—3.

94. Pead MJ, Skerry TM, Lanyon LE. Direct transformation fromquiescence to bone formation in the adult periosteumfollowing a single brief period of bone loading. J Bone MinerRes 1985;3:647—56.

95. Pfeilschifter J, Diel I, Pilz U, et al. Mitogenic responsivenessof human bone cells in vitro to hormones and growth factorsdecreases with age. J Bone Miner Res 1993;8:707—17.

96. Phemister D. The pathology of ununited fractures of theneck of the femur with special reference to the head. J BoneJoint Surg Am 1939;21:681—93.

98. Power J, Loveridge N, Rushton N, et al. Evidence for boneformation on the external bperiostealQ surface of thefemoral neck: a comparison of intracapsular hip fracturecases and controls. Osteoporos Int 2003;14:141—5.

99. Prescher A. Uber eigentumliche Furchenbildungen auf derFacies lateralis tibiae. Verhandlungsband der AnatomischenGesellschaft, 85. Versammlung Anat Anz Suppl 1991;170:207—8.

100. Rauch F, Neu C, Manz F, Schoenau E. The development ofmetaphyseal cortex-implications for distal radius fracturesduring growth. J Bone Miner Res 2001;16:1547—55.

101. Reilly TM, Seldes R, Luchetti W, Brighton CT. Similarities inthe phenotypic expression of pericytes and bone cells. ClinOrthop 1998;346:95—103.

102. Reinelander FW. Circulation in bone. Biochemistry andphysiology of bone, vol. II. New York: Academic Press;1972 . p. 38—46.

103. Rhodin JAG. Histology. A text and atlas. New York: OxfordUniversity Press; 1974. p. 196.

104. Rigal WM. The use of tritiated thymidine in studies ofchondrogenesis. In: McLean FC, Lacroix P, Buddy AM,editors. Radioisotopes and bone. Oxford: Blackwell Scien-tific Publications; 1962. p. 197—225.

105. Rucker M, Roesken, Vollmar B, Menger DM. A novel approachfor comparative study of periosteum, muscle, subcutis andskin microcirculation by intravital fluorescence microscopy.Microvasc Res 1998;56:30—42.

106. Ruff CB, Hayes WC. Subperiosteal expansion and corticalremodeling of the human femur and tibia with aging.Science 1982;217:945—8.

107. Seeman E. From density to structure: growing up andgrowing old on the surfaces of bone. J Bone Miner Res1997;12:509—21.

108. Seeman E. An exercise in geometry. J Bone Miner Res2002;17:373—80.

109. Shapiro F. Fractures of the femoral shaft in children. Theovergrowth phenomenon. Acta Orthop Scand 1981;52:649—55.

110. Shapiro F, Holtrop ME, Glimcher MJ. Organization and cel-lular biology of the perichondrial ossification groove ofRanvier. J Bone Joint Surg 1977;59A:703—23.

111. Shapiro F. Pediatric orthopedic deformities. Basic science,diagnosis, and treatment. San Diego: Elsevier; 2002.

112. Shea J, Vajda EG, Bloebaum RD. Evidence of a hyperminer-alised calcified fibrocartilage on the human femoral neckand lesser trochanter. J Anat 2001;198:153—62.

113. Sheng M, Baylink D, Beamer W, et al. Histomorphometricstudies show that bone formation and bone mineral apposi-tion rates are greater in C3H/HeJ (high-density) thanC57BL/6J (low-density) mice during growth. Bone1999;25:421—9.

114. Simon TM, Sickle DC, Dennis H. Cambium cell stimulationfrom surgical release of the periosteum. J Ortho Res 2003;21:470—80.

115. Sims DE. The pericyte–—a review. Tissue Cell 1986;18:153—74.

116. Simpson AH. The blood supply of periosteum. J Anat1985;140:697—770.

117. Smith Jr RW, Walker RR. Femoral expansion in aging women:implications for osteoporosis and fractures. Science1964;145:156—7.

118. Solomon L. Diametric growth of the epiphyseal plate. J BoneJoint Surg B 1966;48:170—7.

119. Squier CA, Ghoneim S, Kremenak CR. Ultrastructure of theperiosteum from membrane bone. J Anat 1990;171:233—9.

120. Srinivasan S, Weimer DA, Agans SC, et al. Lowmagnitudemechanical loading becomes osteogenic when rest isinserted between each load cycle. J Bone Miner Res2002;17:1613—20.

121. Stephen E, Asmus SP, Story C. Landis developmentalchanges in the transmitter properties of sympathetic neu-rons that innervate the periosteum. J Neurosci 2000;20:1495—504.

122. Szechinski JW, Grigorian MA, Grainger AJ, et al. Femoralneck and intertrochanteric fractures: radiographic indica-tors of fracture healing. Orthopedics 2002;25:1365—8.

123. Takeshita S, Kikuno R, Tezuka K, Amann E. Osteoblast-specific factor 2: cloning of a putative bone adhesion pro-tein with homology with the insect protein fasciclin I.Biochem J 1993;294:271—8.

124. Tang XM, Chai BF. Ultrastructural investigation of osteogeniccells. Chin Med J (Engl) 1986;99:950—6.

125. Taylor JF. The periosteum and bone growth. In: Hall BK,editor. Bone growth VI. Boca Raton7: CRC Press; 1992.

126. Tonna EA. The cellular complement of the skeletal systemstudied autoradiographically with tritiated thymidine

1130 G. Augustin et al.

(H3TDR) during growth and ageing. J Biophys Biochem Cytol1961;9:813—24.

127. Tonna EA, Cronkite EP. The periosteum. Autoradiographicstudies on cellular proliferation and transformation utilizingtritiated thymidine. Clin Orthop 1963;30:218—33.

128. Tonna EA. Response of the cellular phase of skeleton totrauma. Periodontics 1965;4:105—14.

129. Tonna EA. Electron microscopy of aging skeletal cells. III.Periosteum. Lab Invest 1975;61:609—32.

130. Triffitt PD, Cieslak CA, Gegg PJ. A quantitative study of theroutes of blood flow to the tibial diaphys after an osteotomy.J Orthop Res 1993;11:49—57.

131. Trueta J. Blood supply and the rate of healing of tibialfractures. Clin Orthop Relat Res 1974;105:11—27.

132. Turner RT, Wakley GK, Hannon KS. Differential effects ofandrogens on cortical bone histomorphometry in gonadec-tomized male and female rats. J Orthop Res 1990;8:612—7.

133. Urist MR, DeLange RJ, Finerman GAM. Bone cell differentia-tion and growth factors. Science 1983;220:680—6.

134. Vajda E, Bloebaum R. Age-related hypermineralization in thefemale proximal human femur. Anat Rec 1999;255:202—11.

135. van der Meulen MCH, Ashford MW, Kiratli BJ, et al. Deter-minants of femoral geometry and structure during adoles-cent growth. J Orthop Res 1996;14:22—9.

136. Wheater PR, Gurkitt HG, Daniesls VG. Functional histology.New York: Churchill Livingstone; 1987. p. 149.

137. Whiteside LA, Lesker P. The effects of extraperiosteal andsubperiosteal dissection. I. On blood flow in muscle. J BoneJoint Surg 1978;60:23—6.

138. Whiteside LA, Lesker P. The effects of extraperiosteal andsubperiosteal dissection. II. On fracture healing. J BoneJoint Surg 1978;60:26—9.

139. Wilson JW, Rhinelander FW. Bone supply to developing,mature and healing bone. In: Sumner-Smith, editor. Bonein clinical orthopedics: a study of clinical osteology. Phila-delphia: WB Saunders; 1986. p. 57—63.

140. Zagba-Mongalima G, Goret-Nicaise M, Dhem A. Age changesin human bone: a microradiographic and histological studyof subperiosteal and periosteal calcifications. Gerontology1988;34:264—76.

141. Zucman J. Studies on the vascular connections betweenperiosteum, bone and muscle. Br J Surg 1960;48:324—8.