J Clin Pathol 1983;36:1323-1331

Review article

Histomorphometry of bonePA REVELL

From the Department ofMorbid Anatomy, The London Hospital Medical College, Whitechapel,London El 2AD

SUMMARY This review of the histomorphometry of bone outlines methods of biopsy and proces-sing of specimens in the laboratory, the basic principles of morphometry, and the measurementsmade in order to obtain estimates of the proportional volumes and surfaces occupied by differentcomponents of bone. Variability such as that between methods, observers and laboratories is dis-cussed and a brief outline of automatic and semiautomatic methods of image analysis also given.

Histopathology is a subject normally considered inpurely descriptive terms but a quantitative approachis sometimes of value. This is true of the pathologyof bone. Bone morphometry has a role to play in thestudy of particular metabolic disturbances and theirtreatment. Group means and standard deviationsare best used as a method of comparison under thesecircumstances. Morphometry may also be helpful inthe evaluation of an individual biopsy. Care isneeded over the reproducibility of techniques andthe methods may be time consuming, but theseaspects will almost certainly change with thedevelopment of relatively low cost semiautomaticcomputer-linked image analysis systems.

Bone biopsy

Several factors determine the site selected for bonebiopsy, namely:(a) ease of clinical availability,(b) an area where there is active bone turnover and(c) adequate amounts of trabecular bone.The iliac crest is usually chosen.' 2 Modern techni-ques of biopsy involve obtaining a cortex to cortexcore of bone from the iliac crest with a wide boretrephine.3 Ideally the biopsy should be obtainedfrom a standard site.4 An alternative method is totake a vertical core downwards from the iliac crest.A comparison by Visser et a15 was unable to demon-strate a systematic difference between the two tech-niques with respect to the measurement of volumedensities. Wedges of iliac crest obtained at necropsyshould include the area normally examined in biopsyspecimens.The complications of iliac crest biopsy have

Accepted for publication 29 June 1983

recently been reported by Duncan et a16 who pre-sented data for nearly 15 000 biopsies, three fifths ofwhich were transiliac, the remainder obtained by asuperior approach through the iliac crest. Morbiditywas low with both methods, the most common prob-lem being haematoma in patients with primaryhaematological diseases or those receiving heparinduring haemodialysis.Other complications included neuropathy affect-

ing the lateral cutaneous nerve of the thigh, woundinfection, pain, fracture and osteomyelitis. Theoverall incidence of complications in this largenumber of biopsies was 0-5%. Duncan and col-leagues6 found that the large majority of patientsexperience little pain. A similar finding was obtainedby Johnson et al.7The experience of the person performing the

biopsy plays a part in determining the suitability ofthe specimen for histological examination. Whenbone biopsy is performed as an occasional proce-dure, the specimen is often fragmented and inade-quate.

Laboratory processing of the biopsy

The biopsy may be fixed in formalin, 70% ethanolor methanol. The latter two fixatives give better pre-servation of tetracycline fluorescence.8'-0 It is essen-tial to prepare undecalcified plastic embedded sec-tions and our own preference is for methyl metha-crylate sectioned at 6-7 um on a Jung K microtome.Thicker sections (15 ,um) may be used for ultravioletmicroscopy of tetracycline fluorescence.Changes in the amount of total bone and osteoid

together with the activity of the cells at trabecularsurfaces are the important features in metabolicbone disease. All the necessary details are detect-

1323

copyright. on M

ay 18, 2022 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.36.12.1323 on 1 D

ecember 1983. D

ownloaded from

1324able in haematoxylin-eosin stained sections but it ispreferable to use several other techniques to high-light particular features. The differentiation ofosteoid from mineralised bone is easily achievedusing von Kossa counterstained with van Gieson,eosin, safranin or almost any other similar stain. TheGoldner trichrome method gives good contrast be-tween mineralised and unmineralised bone, as doother trichrome methods. Solochrome cyanin is usedin some centres. A close correlation has been shownbetween results for measurements of osteoid volumewhen solochrome and von Kossa methods are com-pared." 12 Comparability was less good betweensolochrome and trichrome, the latter tending tounderestimate the amount of osteoid."' 13 Featuresat trabecular surfaces are reasonably well seen inhaematoxylin-eosin and Goldner trichromemethods. Toluidine blue or thionin staining givegood definition of cellular details. Osteoclasts areeasily visualised by these methods, but some work-ers are now using cold formalin fixation, equal partsmethyl/glycol methacrylate embedding medium andacid phosphatase staining for the demonstration ofthese cells.'4The width of the osteoid seams depends on:(a) the osteoblastic apposition rate, that is the rate

of production of osteoid by osteoblasts,(b) the rate of mineralisation of the osteoid so pro-

duced by the osteoblasts.'5Clearly, an increase in the amount of osteoid rela-tive to total bone tissue (hyperosteoidosis) is notnecessarily due to osteomalacia and it is importantto decide whether there is a calcification defectunder these circumstances. Staining methods whichdemonstrate the mineralisation front include solo-chrome cyanin, Sudan black, cobalt salts andtoluidine blue at pH 2.8.2'5-18 The mineralisationfront appears as a granular purple line at the junc-tion of osteoid and mineralised bone in the toluidineblue method. None of these methods is particularlyreliable and the mineralisation front is best demon-strated by incorporation of a tetracycline label intothe bone before biopsy.The fluorescence of tetracyclines in bone was

described by Milch et al'9 and adapted as a meansof labelling the mineralisation front by variousworkers.20-22

Tetracyclines are bound at sites of activecalcification, as shown by the close anatomical rela-tionship between tetracycline and 45Ca deposition inbone.2223 Further information and references relat-ing to use of tetracycline labelling of bone may beobtained from the literature.8 1024-29A single label gives information about any defect

in calcification, while double-labelling enablesmeasurement of bone mineralisation rate. The time

Revell

required between administration of label and biopsybefore reproducible results are obtained has beenfound to be 48 to 72 h.2Y Tetracycline given for twoor three days followed by an interval of three daysbefore biopsy provides a suitable regimen for singlelabelling, while double labelling is obtained by theuse of two three day courses of tetracycline separ-ated by 10 days.

Basic principles of the morphometry of bone

It is not proposed to give a detailed account of thetheory of morphometry, which is available fromother sources.3032 Basically, the methods involvethe application of probability theory to geometry bythe use of estimates, rather than exact measure-ments. Repeated counting is used in order to makethe estimates as accurate as possible. Although themeasurements are made on two dimensional images(allowing for section thickness), the informationderived may be interpreted on a three dimensionalbasis. Results are usually expressed in percentagesthough some workers use ratios.Bone quantification may be performed using one

of the following:(a) inexpensive simple eye-piece graticules,(b) semiautomatic instruments in which a digitising

tablet is linked to a desk-top microcomputer,(c) fully automatic computer-linked image analysis

equipment.The following account will deal mainly with pointcounting and so-called linear intercept methodsusing eye-piece graticules.

Point counting is performed by the superimposi-tion of a series of points on the microscope field withan eye-piece graticule and enables the estimation ofareas. The principle of point counting is extremelysimple. The number of points or hits occurring on aparticular feature, for example, bone trabeculae iscounted and expressed as a percentage of the totalnumber of possible hits (Fig. 1). The adjacent field iscounted, and so on, until a sufficient number ofmeasurements has been made to obtain an accurateestimate of the true amount of trabecular bone as apercentage of total tissue. Thus if n, n2, n3 etc. arethe numbers of points falling on trabeculae and N1,N2,N3 etc. are the total possible numbers of pointsin each field, the area A, expressed as a percentage,can be calculated as:

nl + n2 + n3............ nxN, +N2 +N3.NXlOO

where x is the number of fields necessary to obtain areproducible result.

copyright. on M

ay 18, 2022 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.36.12.1323 on 1 D

ecember 1983. D

ownloaded from

Histomorphometry of bone

Fig. 1 Diagram to show the principle ofpoint countingusing the Zeiss Integration Plate II eye-piece graticule. Iftheislands are considered to be bone trabeculae, then in thesmall central square there are 8 out of25 points falling onbone-that is 32 % oftotal tissue is bone. Alternatively, 36points fall on bone in the large square which contains 100points-that is, 36% oftotal tissue is bone. (Withpermission of Carl Zeiss (Oberkochen) Ltd.)

--- Nominal value

10 20 30 40 50 60 70 80 90 100Test positions

I , I I

0 500 1000 1500 2000 2500Total No ot points

Check of measured values (test object)

Fig. 2 Diagram to show the effect ofincreasing the numberoffields examined and points counted on the accuracy oftheestimate obtained for a volume measurement. There isconsiderable variation in the results obtained when smallnumbers offields are measured, but the values graduallysettle to a "nominal" value, shown as a horizontal line.Obviously the number offields could be increased further,in theory until a constant value were obtained. (Withpermission of Carl Zeiss (Oberkochen) Ltd.)

1325

The simplest way to overcome the problem ofhow much to count is to calculate the mean valueafter a given number of fields, count more fields andrecalculate the mean value, continuing until themean value settles to a more or less constant level.This value is described as the "nominal value" and isbest understood by reference to Fig. 2.

Since measurements are made on a basically twodimensional object, the value obtained is an area, orarea fraction (Aa), of the whole area of tissuecounted. The Delesse principle, described by aFrench geologist in the middle of the last century,33states that area is an unbiased estimator of volumeso that it is usual to express values as volumes eventhough areas have been measured.The linear intercept method is used for measure-

ments of surface area or simple length measure-ments. An eye-piece graticule enables the superim-position of a series of parallel lines upon the micro-scope field to be examined. The method is easilyunderstood by reference to Fig. 3, which illustratesthe measurement of osteoid surface using five lines.The number of intercepts, or more accuratelyintersections, that osteoid and total trabecular sur-face make with these lines is counted. Further meas-urements are performed on successive adjacentfields until an accurate estimate of the percentage ofsurface occupied by a particular feature, in this caseosteoid, is obtained. Repetition until a nominalvalue is achieved applies in the same way as for

7tn4LR

\~~CFig. 3 Diagram to show the use ofthe linear interceptmethod for measuring surface lengths. The broad blackborders represent areas ofosteoid. There are S places wherethe lines intersect with osteoid-covered surface and 20 placeswhere lines intersect trabecular surface-that is, osteoidsurface is 25% oftrabecular surface. This diagram is asimplified representation of the superimposition of lines inthe Integration Plate II, shown in Fig 1.

13 -

12 -

S 111

10-

9-

*. s~~~~~~~~~~~~ copyright.

on May 18, 2022 by guest. P

rotected byhttp://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.36.12.1323 on 1 Decem

ber 1983. Dow

nloaded from

1326

point counting. Measurements of osteoid, resorp-tion, osteoblastic surfaces may be made by thismethod.The mineralisation front, as seen by ultraviolet

light microscopy after tetracycline labelling of thebiopsy, is also measured in this way. If double label-ling of the biopsy with tetracycline is available, thenit is possible to measure the distance between thetwo labels using a calibrated micrometer eye-piece.This measurement is performed at four equidistantpoints along each surface showing two fluorescentlines, as illustrated in Fig. 4, and the measurementrepeated a sufficient number of times, saytwenty.9242528 The appositional rate is calculated bydividing the mean distance between the labels (d),by the time interval between administration of labels(t). In order to obtain a true appositional rate, it isnecessary to apply a correction factor since the twolines of tetracycline label are actually sectioned inrandom planes varying between perpendicular andhorizontal. The mathematical method for the deri-vation of this correction factor is available in papersby Frost28 and Teitelbaum and Nichols.25 Frost28 hasrecommended a correction factor of 0 74-that is,the true appositional rate may be calculated as 0 74times the value obtained from actual measurement.

Measurements on bone biopsies

There is considerable variation in the literature withrespect to the terms used for the various valuesobtained in bone morphometry. The volume oftrabecular bone as a proportion of total tissue (boneand bone marrow) is termed the trabecular bonevolume, absolute volume of trabecular bone, frac-tional bone volume, fractional trabecular bone vol-ume, volume density of bone or relative trabecularvolume. Surface density is strictly a measure of sur-face area related to volume and expressed in mm2/cm3334-36 but "surface density" and the abbrevia-tion S, are often used more loosely. It is vital there-fore to have a clear idea of what is being expressedwhen examining particular results. A list of para-meters and how they are measured is shown in Table1.The osteoid index provides a useful guide to the

thickness of osteoid seams, and is derivedmathematically from the osteoid volume and osteoidsurface, using the following formula:

Osteoid volumeOsteoid index = x 100.

Osteoid surfaceIt has been shown to correlate well with the actualmeasurement of osteoid seam width."5 A rapid andfairly reliable method of assessing osteoid seams is

Revell

Fig. 4 Diagram to illustrate the measurement ofappositional rate in bone with a double tetracycline label.The distance (d) between two fluorescent lines ofincorporated tetracycline is measured with a calibratedeye-piece graticule at four equidistant points. Theappositional rate is obtained with knowledge ofthe timeinterval (t) between administration ofthe labels-that is,appositional rate = dlt. (See text for references to correctionfactor required for measurement ofthe appositional rate.)

the use of polarised light microscopy, when thenumber of birefringent lamellae (bright lines) maybe counted. Up to four such lamellae are present innormal bone3"38 so that a greater number than this isan indicator of hyperosteoidosis.

NORMAL VALUESDefinitive normal values for the measurementsmade in bone morphometry are impossible to give,since they vary from laboratory to laboratory. It isessential that each centre obtains a number of nor-mal bone samples at necropsy from the iliac crest ofpreviously ambulant cases dying suddenly with car-diovascular disease and having no other knownpathology. Measurement of these specimens enablesa normal range to be obtained. Examples of typicalnormal values are given in Table 2. Several workershave demonstrated a gradual decrease in thetrabecular bone volume of normal bone withincreasing age2 39`41The mineralisation front should normally occupy

more than 80% of the osteoid surface, but thisdecreases slightly in older subjects.2 A mineralisa-tion front of less than 20% represents a definitecalcification defect. The mineralisation rate (apposi-tional rate) as measured by double-labelling withtetracycline is normally about 1 ,um/day or slightlyless. Active osteoid formation by osteoblasts usuallycomprises around 5% of the total trabecular surface

... .........

copyright. on M

ay 18, 2022 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.36.12.1323 on 1 D

ecember 1983. D

ownloaded from

Table 1

Trabecular bone volume

Osteoid volume

Osteoid surface

(Active) osteoblastic surface

Resorption surface

Osteoclastic resorption surface

Mineralisation front

Osteoid index

Appositional rate(see text, correction factor)

Osteoclastic index

Volume of trabeculaeVolume of trabeculae and marrow

Volume of osteoidVolume of osteoid and mineralised bone

Length of surface occupied by osteoidTotal length of trabecular surface

Length of trabeculae occupied by active osteoblastsTotal length of trabecular surface

Length of trabecular surface occupied by resorption lacunaeTotal length of trabecular surface

Length of surface occupied by osteoclastsLength of resorption lacunae

Line length of mineralisation (Tetracycline fluorescence)Total length of trabecular surface

Osteoid volumeOsteoid surface

- Distance between labelsTime

x 100

x 100

x 100

x 100

x too

x 100

x 100

x 100

gm/day

= Estimate of numbers of osteoclasts as either osteoclasts per high power field,osteoclasts per sq mm or sq cm of tissue or osteoclasts per mm trabecularsurface

Table 2 Normal values for iliac crest bone expressed aspercentages (from Melsen et all')

Age (yr) Male Female

Trabecular bone volume10-19 245-±1-7 31-8±2-720-29 24-6±2-3 26-8+1-730-39 22-9±1-1 26.5±2-440-49 17-4±1-9 22 5±3 550-59 20-4±1+2 20-1±2+460-69 13-7±1-2 15-5±1-870-79 16-2±2-4 17-0±2-0> 80 15-4±2-8 17-7±3-2

Osteoid volume10-19 2-6±0-6 2-1±0-720-29 2 5±0 3 2-4±0-430-39 1-8±0-3 1.7±0-440-49 2-2±0.9 1-3±0-550-59 3-0±0-5 1.9±0.460-69 1-6±0-3 2-0+0-670-79 1-8±0-5 2-10t-5> 80 0-9±0-2 1-3±0-2

Osteoid surface10-19 17-8±2-8 14-7±1-920-29 18-8±2-5 16-7± 1-630-39 14-9t2-2 14-1±2-640-49 17-2±5-5 14-1±1-750-59 20-3±1-2 19-3±3-160-69 16-0±2-4 19-8±3-470-79 18-3±2-5 18-2±3-4> 80 12-0±3-2 12-7t0-8

Resorption surface10-19 3-9±0-4 3-2±0-320-29 4-0±0 5 3 9±0-330-39 3-7±0-3 4-0±0-440-49 4-1±0+8 3-4±0-250-59 4-0±0-3 5.1±0-660-69 4-1±0+6 3-6±0-670-79 50±0-6 4-3±0-5> 80 6-0±0-1 3-7±0-6

and active resorption less than 1 % of total trabecu-lar surface.2

VARIABILITYImportant aspects of variability in bonequantification include differences between (a) sitesin the same bone (b) bones in the same patient (c)observers (d) laboratories, as well as (e) variationsin methods, such as staining techniques andmagnifications used in microscopy.There have been numerous studies in which dif-

ferent sites in the iliac crest were compared. Mini-mal differences have been demonstrated betweenbiopsies at adjacent sites in the same iliac crest2441 42although differences do occur when the bone ismore than 2 cm posterior or inferior to the standardsite of biopsy.4 Comparison of results obtained fromthe left and right iliac crests have been performed byseveral workers45 11374344 and no systematic differ-ences have been demonstrated.The influence of staining methods on bone his-

tomorphometry has already been mentioned, com-parability being good between solochrome and vonKossa for measurements of osteoid volume, and lessgood between solochrome and trichrome methods.Melsen el al4 found higher values for osteoid volumeand osteoid surface using Masson trichrome com-pared with toluidine blue.There are several questions which arise with

respect to the use of tetracycline labelling of bone.The timing of administration of label and biopsy is

Histomorphometry of bone 1327

copyright. on M

ay 18, 2022 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.36.12.1323 on 1 D

ecember 1983. D

ownloaded from

sv

IMean tI SEM

.L I vv

I I

I 24

-23

-22 cEX

-21 c-

-20 '<-19 <

-18

-17

Fig. 5 Surface density (Sv) and volumedensity (Vv) of iliac crest cancellous boneat different magnifications, showing thesystematic increase in the estimate ofsurface density with change inmagnification and no difference in volumedensity. (From Olah,36 with thepermission ofArmour-Montagu,Levallois, France.)

X40 x63 x100 x160 x250Magnitication

x400

l |I I I I l

128 80 50 32 20 13 0*8Resolution ( n)

Student /pothologistr = 0 85S= 0 00001SEE=401 76n= 20y= 086x 820 3

r = 0.95s = 0* 00001SEE = 175 4n=16y= 082x + 3137

Fig. 6 Determination ofthesurface density (Sv) oftrabecularbone. Sixteen values obtained by astudent at the beginning and end ofa study of150 normal cases, 20values ofa student compared witha experienced pathologist on thesame biopsies. (From Delling etal,45 with the permission ofArmour-Montagu, Levallois,France.)

0 700 2100 3500 4900 63001400 2800 4200 5600

important and biopsy should normally be two orthree days after the last label. If the interval betweentwo labels is much longer than three weeks then thepercentage of osteoid bearing two labels will bedecreased.The influence of microscopic resolution on the

results obtained by bone morphometry has beendescribed.3637 Changes in magnification between 25and 400 times did not influence estimations of vol-ume (volume density), but over the same range ofmagnification there was a systematic increase in theestimate of surface density.36 (Fig. 5)

Delling and his colleagues45 have shown that thereis greater variability in results obtained for surfacemeasurements compared with volume measure-ments. This was particularly the case where cellular

details at trabecular surfaces were being evaluted,and many more fields had to be measured to obtainreproducibility under these circumstances. The rela-tively small amount of cellular activity at trabecularsurfaces and the irregular distribution of changes arecontributing factors. The same authors also per-formed comparisons between observers andlaboratories.45 The experience of the observer wasfound to be important, especially with respect tothose surface measurements relating to cellularactivity. The volume and surface densities oftrabecular bone were not affected by observerexperience in the sense that comparison of a studentwith himself and the student with an experiencedpathologist both showed good statistical correlation(see Fig. 6). The shift in the line for the pathologist/

1328

36-mEu 34-EE 32-

(.40x 30-

' 28-

a 26-

u 24-

) 22-

Revell

x25

7000

6300-

5600-

4900

4200-jn

3500(I)

2800*

2100

1400

700

7000

I I v I I I I

aI I I I I I I

0

0 0

0

0 Student lst / 2no Student/pothol

copyright. on M

ay 18, 2022 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.36.12.1323 on 1 D

ecember 1983. D

ownloaded from

Histomorphometry of bone

student comparison suggests that the pathologistwas noting features which were not recognised bythe student.

Differences between centres looking at the samebiopsies are also of interest. When four differentmorphometry groups were asked to assess 10 biop-sies, there was considerable variation with respect tovalues obtained for osteoid surface measurementsthough trabecular bone volume showed much smal-ler variation.45

Semiautomatic and automatic methods in bonequantification

The examination of large numbers of specimens istime-consuming and tedious using the simple eye-piece graticule. Automated and semiautomatedcomputer-linked systems offer an alternativeapproach. Both types of equipment have the advan-tage that they enable a much greater throughput ofmaterial. Disadvantages include the amount of timerequired setting up the equipment before sectionsmay be examined, and the need to stain sections in away which is suitable for image analysis. Automaticimage analysers may be divided into three maincategories, according to whether they work bysource-plane-scanning, specimen-plane-scanning orimage-plane-scanning.32

Image-plane-scanning is the basis on which alltelevision-linked systems operate. The best knownsystem of this type is the Quantimet 720 and the fol-lowing account is based on personal experience of itsuse.

Computer-linked image analysis using the Quan-timet may be considered in three stages. Input is bymeans of a closed circuit television linked to a lightmicroscope. The signals resulting from the scanningof the microscope image by the television cameraare used to produce an image on a television screenand at the same time for computer analysis of par-ticular features. The results of the computer analysisare displayed numerically at the top of the televisionscreen. It is also possible to superimpose the compu-ter analysis display as an image on the televisionscreen.The machine of which the author has experience

has television display, standard function computer, acontrol system to set light sensitivity and shade cor-rection, variable frame and scale settings,densitometer function and interfaces with a desk topmicrocomputer. Area, intercept, perimeter and sizecan be measured and a programmer module enablesthe machine to run automatically through a series ofmeasurements on each field.The Quantimet is used in our own laboratories in

two different ways for the measurement of bone his-

1329

tology. Osteoid volume and trabecular bone volumeare measured using the densitometer function. Sig-nals from the television camera are passed to adetection module in which grey level thresholds areset in 64 steps from black to white. It is importantthat the image being analysed can be resolved intoclearly distinguishable grey levels. Our preference isto use von Kossa stained sections counterstainedwith van Giesen and treated in such a way that col-oration of the bone marrow features is deliberatelyleached out. The image is resolved into black, greyand white areas corresponding to mineralised bone,osteoid and bone marrow. Careful selection of thegrey level settings enables accurate detection at thecorrect boundary point for the three features whichare to be detected and measured. The areas overwhich there are particular grey levels are measuredand values for "white, grey and black" obtained.Raw data for the areas of osteoid, mineralised boneand total tissue are then routed to the desk top com-puter (output terminal). The process is repeated foradjacent fields and all the raw data accumulated inthe output computer. Osteoid volume and trabecu-lar bone volume are calculated automatically by aprogramme in this computer, in just the same way aswould be performed in conventional point counting.Although it is theoretically possible to obtain

perimeter measurements using the same basicmethod, we have preferred to use the Quantimet asa semiautomatic image analyser. It is possible to out-line trabecular surface, osteoid and resorption sur-faces and lengths occupied by active osteoblasts andosteoclasts with a light pen (Fig. 7). The line lengthsof each outlined feature are routed to the outputcomputer for each field examined and the measure-ments calibrated so that it is possible to obtain anabsolute value for the lengths measured. The outputcomputer programme accumulates data for eachmeasurement and after a previously specifiednumber of fields has been examined, calculatesosteoid surface, active osteoblastic surface, resorp-tion surface, osteoclastic resorption surface and totallength of trabecular surface. The numbers of osteo-clasts seen is accumulated in the output computer aseach field is examined, so that it is possible to obtainan osteoclast index, expressed as osteoclasts/mmtrabecular surface.

Semiautomatic systems for quantitative analysisof histology consist of a "digitising tablet", which isan electronic drawing board, linked to a microcom-puter. The microscopic field may be projected ontothis drawing board or a side-arm drawing tube used.Commercially available semiquantitative equipmentdesigned for this purpose includes the Leitz ASM,Reichert-Jung MOP and Videoplan. The digitisingtablet enables the outlining of features seen in the

copyright. on M

ay 18, 2022 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.36.12.1323 on 1 D

ecember 1983. D

ownloaded from

Revell

Fig. 7 Photograph of the Quantimet 720television screen, showing the use of a light peni(white line) to outline the trabecular surface. Theline is occupying 1941 picture points.

microscope field and measurements of line length,perimeter, intercept or area, are then rapidlyobtained. These measurements can be expressed inabsolute terms bv previous calibration. Area andsurface measurements for each field may be summedand the various bone morphometry parameters cal-culated by use of a suitable programme in the mic-rocomputer. Semiautomatic methods are subjective,in that they rely on the drawing of features by theobserver. The use of digitising tablets for bonequantification is described elsewhere.46-48

COMPARISON OF BONE QUANTITATIONMETHODSA good correlation has been found between theactual volume of bone, measured by water dis-placement, and the point counting technique.3849 Adifference of 1 5% was found between the use of theZeiss eye-piece graticule and the Quantimet methodby Giroux, Courpron and Meunier." No differencein accuracy has been found between point-countingand a semiautomatic method using a digitiser tab-let.47 Smaller numbers of fields needed to be meas-

ured using the semiautomatic method to achieve thesame coefficients of variance with respect to bothnon-cellular and cellular (bone formation andresorption) parameters.

References

Matrajt H, Bordier P, Martin J, Hioco D. Technique pour l'inclu-sion des biopsies osseuses nondecalcifies. J Microscopie1967;6:499-504.

2 Rasmussen H, Bordier P. The physiological and cellular basis ofmetabolic bone disease. Baltimore: Williams and Wilkins.1974.

Bordier P. Matrajt H, Miravet B, Hioco D. Mesure histologiquede la masse et de la resorption des travees osseuses. PatholBiol (Paris) 1964;12:1238-43.

4 Melsen F. Melsen B, Mosekilde L. An evaluation of the quantita-tive parameters applied in bone histology. Acta Pathol Mic-robiol Scand 1978;86:63-9.

' Visser WJ, Niermans HJ, Roelofs JMM, Raymakers JA.Duursma SA. Comparative morphometry of bone biopsiesobtained by two different methods from the right and left iliaccrest. In: Meunier PJ, ed. Bone histomorphometry. 2nd Inter-national Workshop. Societe de la Nouvelle Imprimerie Four-nie; Toulouse, France, 1976:79-87.

6 Duncan H, Dao SD, Parfitt AM. Complications of bone biopsy.Metab Bone Dis et Rel Res 1980;2:suppl 483-6.

Johnson KA, Kelly PH, Jowsey J. Percutaneous biopsy of theiliac crest. Clin Orthop Rel Res 1977;123:34-6.

Frost HM. Tetracycline-based histological analysis of boneremodelling. Calcif Tissue Res 1969;3:211-37.

9Parfitt AM, Villanueva AR, Crouch MM, Mathews CHE. Dun-can H. Classification of osteoid seams by combined use of cellmorphology and tetracycline labelling. Evidence for intermit-tency of mineralisation. In: Meunier PJ. ed. Bone mor-

phometry. Societ& de la Nouvelle Imprimerie Fournie;Toulouse, France, 1976:299-310.

'Melsen F, Mosekilde L. Interpretation of single labels after invivo double labelling. Metab Bone Dis et Rel Res 1980;2:suppl171-8.

"Giroux JM, Courpron P, Meunier P. Histomorphometrie del'osteopenie physiologique senile. Monographie du laboratoirede Researches sur I'Histodynamique osseuse. Lyon, 1975.

2 Meunier P, Edouard C. Quantification of osteoid tissue intrabecular bone. Methodology and results in normal iliac bone.In: Jaworski ZFG, ed. Proceedings of the 1st Workshop on

Bone Morphometry. Ottawa: University of Ottawa Press.1976:191-6.

Meunier P, Edouard C. Courpron P, Toussaint F. Morphometricanalysis of osteoid in iliac trabecular bone. Methodology.

1330

copyright. on M

ay 18, 2022 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.36.12.1323 on 1 D

ecember 1983. D

ownloaded from

Histomorphometry of bone

Dynamical significance of the osteoid parameters in vitamin Dand problems related to uremic disease. Norman AW et al,eds. Berlin: Gruyter, 1975:149-55.

14 Evans RA, Dunstan CR, Hills EE. Extent of resorbing surfacesbased on histochemical identification of osteoclasts. In: JeeWSS, Parfitt AM, eds. Bone histomorphometry. 3rd Interna-tional Workshop. Metab Bone Dis Rel Res 1980;2:suppl 29-34.

5 Meunier P, Edouard C, Richard D, Laurent J. Histomor-phometry of osteoid tissue. The hyperosteoidoses. In: MeunierPJ, ed. Bone histomorphometry. 2nd International Workshop.Societe de la Nouvelle Imprimerie Fournie; Toulouse, France,1976:249-62.

16 Irving JT. A histological stain for newly calcified tissues. Nature1958;181:704-5.

'' Irving JT. The sudanophil material at sites of calcification. ArchOral Biol 1963;8:735-45.

Matrajt J, Hioco D. Solochrome cyanine R as an indicator dye ofbone morphology. Stain Technol 1972;41:97-100.

91 Milch RA, Hall DP, Tobie JE. Fluorescence of tetracycline anti-biotics in bone. J Bone Jt Surg [Am] 1958;40:897-910.

20 Baud CA, Dupont DH. Histologie intrastructurale sur labifluorescence du tissu, osseux tonite par les t6tracyclines. C RSeances Acad Sci 1962;254:3129-30.

21 Frost HM. Tetracycline labelling of bone and the zone of demar-cation. Can J Biochem 1962;40:485-9.

22 Harris WH, Jackson RH, Jowsey J. The in vivo distribution oftetracycline in canine bone. J Bone Jt Surg [Am]1962;44: 1308-20.

23 Urist MR, Ibsen KH. Chemical reactivity of mineralised tissuewith oxytetracycline. Arch Pathol 1963;76:484-96.

24 Frost HM, Meunier P. Histomorphometry of trabecular bone. II.

An empirical test for the theoretical correction for apposi-tional rate measurements. In: Meunier PJ, ed. Bone histomor-phometry. 2nd International Workshop. Societe de la NouvelleImprimerie Fournie; Toulouse, France, 1976:371-81.

25 Teitelbaum SL, Nichols SH. Tetracycline-based morphometricanalysis of trabecular bone kinetics. In: Meunier PJ, ed. Bonehistomorphometry. Societe de la Nouvelle Imprimerie Four-nie; Toulouse, France, 1976:311-9.

26 Flora L. Idiosyncrasies of the measurement bone dynamics withfluorescent labels. In: Meunier PJ, ed. Bone histomor-phometry. 2nd International Workshop. Societe de la NouvelleImprimerie Fourni6; Toulouse, France, 1976:321-6.

27 Bordier PJ, Marie P, Miravet L, Ryckewaert A, Rasmussen H.Morphological and morphometrical characteristics of themineralisation front. A vitamin D regulated sequence of thebone remodelling. In: Meunier PJ, ed. Bone histomor-phometry. Societe de la Nouvelle Imprimerie Fournie;Toulouse, France, 1976:335-54.

28 Frost HM. Histomorphometry of trabecular bone. I. Theoreticalcorrection of appositional rate measurements. In: Meunier PJ,ed. Bone histomorphometry. 2nd International Workshop.Societe de la Nouvelle Imprimerie Fournie; Toulouse, France,1976:361-70.

29 Treharne RW, Brighton C(. The use and possible misuse of tet-racycline as a vital stain. Clin Orthop Rel Res 1979;140:240-6.

30 DeHoff RT, Rhines FN. Quantitative microscopy. New York:McGraw-Hill, 1968.

31 Williams MA. Quantitative methods in biology. Amsterdam, NewYork, Oxford: North Holland Publishing Company, 1977.

32 Aherne WA, Dunnill MS. Morphometry. London: EdwardArnold, 1982.

33 Delesse MA. Proc6dd mechanique pour determiner la composi-

1331

tion des roches. Annales des Mines 1848;13:379-88.3 Merz WA, Schenk RK. Quantitative structural analysis of human

cancellous bone. Acta Anat 1970;75:54-66.35 Merz WA, Schenk RK. A quantitative histological study on bone

formation in human cancellous bone. Acta Anat 1970;76: 1-15.

36 Olah AJ. Influence of microscopic resolution on the estimation ofstructural parameters in cancellous bone. In: Meunier PJ, ed.Bone histomorphometry. 2nd International Workshop. Societede la Nouvelle Imprimerie Fournie: Toulouse, France,1976:55-61.

37 Woods CG, Morgan DB, Peterson CR, Gossman HH. Measure-ment of osteoid in bone biopsy. J Pathol Bacteriol1968;95:441-7.

38 Ellis HA. Peart KM. Quantitative observations on mineralisedand non-mineralised bone in the iliac crest. J Clin Pathol1972;25:277-86.

3 Courpron P, Meunier P, Bressot C, Giroux JM. Amount of bonein iliac crest biopsy. Significance of the trabecular bone vol-ume. Its values in normal and in pathological conditions. In:Meunier PJ, ed. Bone histomorphometry. 2nd InternationalWorkshop. Societ6 de la Nouvelle Imprimerie Fourni6;Toulouse, France, 1976:29-53.

40 Beck JS, Nordin BEC. Histological assessment of osteoporosisby iliac crest biopsy. J Pathol Bacteriol 1960;80:391-7.

41 Melsen F, Melsen B, Mosekilde L, Bergmann S. Histomor-phometric analysis of normal bone from the iliac crest. ActaPathol Microbiol Scand 1978;86:70-81.

42 Garner A, Ball J. Quantitative observations on mineralised andunmineralised bone in chronic renal azotaemia and intestinalmalabsorption syndrome. J Pathol Bacteriol 1966;91:545-61.

43 Ritz E, Krempien B, Mehls 0, Malluche J. Skeletal abnor-malities in chronic renal insufficiency before and duringmaintenance hemodialysis. Kidney Int 1973;4:116-27.

4 Visser WJ, Roelofs JMM. Peters JPJ, Lentferink MHF, DuursmaSA. Sampling variation in bone histomorphometry. In: JeeWSS, Parfit AM, eds. Bone histomorphometry. 3rd Interna-tional Workshop. Metab Bone Dis Rel Res 1980;2: suppl 429-34.

45 Delling G, Luehmann H, Baron R, Mathews CHE, Olah A.Investigation of intra and inter-reader reproducibility. In: JeeWSS, Parfitt AM, eds. Bone histomorphometry. 3rd Interna-tional Workshop. Metab Bone Dis Rel Resl980;2: suppl 419-27.

40 Birkenhager-Frenkel DH, Clermonts ECGM, Richter H. His-tomorphometry by means of an x-y tabloid. Design of a com-puter programme; Disposition of Equipment. In: Jee WSS,Parfitt AM, eds. Bone histomorphometry. 3rd InternationalWorkshop. Metab Bone Dis Rel Resl980;2: suppl 453-7.

47 Malluche HH, Sherman D, Manaka R, Massey SG. Comparisonbetween different histomorphometric methods. In: Jee WSS,Parfitt AM, eds. Bone histomorphometry. 3rd InternationalWorkshop. Metab Bone Dis Rel Res 1980;2: suppl 449-5 1.

48 Manaka RC, Malluche HH. A program package for quantitativeanalysis of histologic structure and remodelling dynamics ofbone. Comput Programs Biomed 1981;13:191-201.

49 Schwartz MP, Reeker RR. Direct and histomorphometricdeterminations of surface density and volume. In: Jee WSS,Parfitt AM, eds. Bone histomorphometry. 3rd InternationalWorkshop. Metab Bone Dis Rel Res 1980;2: suppl 279-80.

Requests for reprints to: Dr PA Revell, Department ofMorbid Anatomy, The London Hospital Medical College,Whitechapel, London El 2AD.

copyright. on M

ay 18, 2022 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.36.12.1323 on 1 D

ecember 1983. D

ownloaded from