J Orthop Sci (2002) 7:226–231

Estimation of the rate of osteogenesis by tetracycline double labelingin intertrochanteric fractures of the femur in different age groups andits correlation with radiological grading of osteoporosis

Sandeep Singh1, Anil Kumar Jain

1, Anil Arora1, Jugesh Khanna

2, Satish Kumar Bhargava3,

and Sudhir Kumar1

1 Department of Orthopaedics, University College of Medical Sciences and Guru Teg Bahadur Hospital, Delhi-110 095, India2 Department of Anatomy, University College of Medical Sciences and Guru Teg Bahadur Hospital, Delhi, India3 Department of Radiology, University College of Medical Sciences and Guru Teg Bahadur Hospital, Delhi, India

turnover. Bone turnover continues to occur throughoutlife. The terms bone formation rate and resorption ratesignify the anabolic and catabolic activities, respec-tively. The balance between the two activities deter-mines the amount of bone present at any given periodof time. The amount of bone present determines thestrength of bone directly, so that low bone mass is aprimary determinant of fracture.1 A decrease in bonemass and mechanical failure of the skeleton that resultsin the occurrence of fracture with minimal trauma isdefined as osteoporosis. The public health impact ofosteoporosis stems from its association with fractures ofthe hip, spine, and forearm.9

Histological heterogeneity in bone has been reportedin patients with osteoporosis.16 Some patients have in-creased resorption alone, some have increased resorp-tion and formation, and some have decreased boneformation. The distinction of low, high, or normalturnover is of interest for the selection of the best thera-peutic agent. Ten years ago, bone histomorphometrywas used as a routine technique for the diagnosis ofosteoporosis and other metabolic bone diseases, butbone mass is now measured by noninvasive methodssuch as dual energy X-ray absorptiometry (DEXA); thelatter, however, do not provide information on bonequality or microarchitecture.2 Bone histomorphometryconsists of measuring parameters reflecting bone struc-ture and turnover. It remains the only method that al-lows the study of bone at the tissue or the cell level toenable measurements at intermediary levels of organi-zation of bone. Any currently available tetracycline an-tibiotic administered in vivo by any acceptable routewill be deposited in newly formed and mineralizingbone, cartilage, and teeth. All the mineral ever depos-ited in lamellar bone matrix is laid down in the initial 4days of mineralization in the zone of demarcation.5 Thehistological measurement of bone formation with a tet-racycline tissue marker is a direct measurement of bonetissue metabolism and also provides dynamic param-

Abstract Our aim was to estimate the rate of osteogenesis byusing tetracycline double labeling and to correlate it with theradiological grading of osteoporosis in patients of intertro-chanteric fracture of the femur in different age groups. Thirty-two cases of intertrochanteric fracture of the femur of eithersex, without any associated osteopenic conditions, who wereadmitted for surgery constituted the clinical material. Thecases were divided into three age groups: group I, �45 years;group II, 45–60 years; and group III, �60 years. Osteoporosiswas graded in all groups on plain X-rays by Singh’s index (SI)and metacarpal index (MI). The patients were labeled with asingle large dose (1.5 g) of tetracycline orally on day 1 and day8. An iliac crest bone biopsy specimen was obtained from eachof the patients during surgical intervention. Thin undecalcifiedsections were made by the hand grinding technique. Thesections were studied under UV light for presence offluorescence, and rate of osteogenesis (mineral appositionalrate, MAR) was calculated for each of the groups. Only 7patients were found to have radiological osteoporosis (bySingh’s index). The average MI and MAR were found to be46.5% and 1.48µm/day, respectively. The single-dose labelingschedule has shown good fluorescent bands, comparable tothe conventional dose schedule. No correlation was foundbetween Singh’s index and metacarpal index in group I andIII, whereas in group II it was significant. There is nosignificant correlation between the rate of new bone forma-tion and the radiological grading of osteoporosis.

Key words Osteoporosis · Intertrochanteric fracture femur ·Tetracycline double labeling

Introduction

Bone is a highly vascular and dynamic tissue and is in aconstant process of resorption at some areas and forma-tion at other sites, a process termed bone remodeling or

Offprint requests to: A.K. JainReceived: May 28, 2001 / Accepted: November 26, 2001

227S. Singh et al.: Bone dynamics

eters.6 The paucity of literature on the rate of boneformation in the Indian population, on correlation ofosteoporosis with intertrochanteric fracture of the fe-mur in different age groups, and on correlation of oste-oporosis with the rate of new bone formation was themotivating factor for the present study.

Materials and methods

This study was conducted in the Department ofOrthopaedics, in conjunction with the Department ofAnatomy and the Department of Radiodiagnosis, Uni-versity College of Medical Sciences and Guru TegBahadur Hospital, Delhi. Thirty-two patients of variousages, both males and females, with intertrochantericfracture femur who were admitted to the orthopaedicswards for surgical procedures and were otherwise medi-cally fit constituted the clinical material for the study.

The patients were divided into three groups by age:group I, �45 years; group II, 45–60 years; and group III,�60 years. Informed consent was obtained from eachpatient. X-rays of the pelvis including both hip joints byanteroposterior (AP) view in 20° internal rotation weretaken to grade osteoporosis according to Singh’s index(SI)14 (Fig. 1). Osteoporosis was also graded for eachpatient by determining the combined cortical thicknessof the second metacarpal (metacarpal index, MI) of thedominant hand for which an X-ray hand — AP view wastaken (Fig. 2).

MI � Combined width of layers of cortex on each side

Total width of shaft� 100

The foregoing values were calculated at the midpoint ofthe shaft of the second metacarpal and expressed inpercentages.

Patients selected for the study were given a singlelarge dose (1.5g) of tetracycline orally under direct su-pervision on day 1, followed by an interval of 6 days.The same dosage was repeated on day 8 while preparingthe patient for surgery. The patients refrained from tak-ing food, especially dairy products, 1 h before and 2 hafter intake of the tetracycline capsules. Care was takenthat there was no delay in the treatment of the patientsby labeling them with tetracycline.

A 5-mm-diameter modified Patiala core biopsy nee-dle was used to obtain the iliac crest bone biopsy speci-mens from the patients. These samples were taken atthe time of surgical intervention, which was at least 48 hafter the last dose of tetracycline. The samples thusobtained were kept in 70% ethanol at 4°–8°C in a do-mestic refrigerator until they were further processed.

Thin undecalcified sections were made from the sam-ples by hand grinding them on sandpaper. The sectionswere ground until they became almost transparent, sothat the print was visible through them when held overa newspaper. Sections were given repeated thoroughwashings with normal saline solution. Sections were airdried and transferred to a grease-free slide, fixed withDPX mounting medium, and then covered with a cover-slip. They were then left to dry at room temperatureafter mounting. The sections were studied under ultra-

Fig. 1. X-ray of hip joint in AP view in slight internal rotationto determine Singh’s index

Fig. 2. Right-hand AP view to determine metacarpal index

228 S. Singh et al.: Bone dynamics

violet light for presence of fluorescence using a Zeissultraviolet microscope with an H.B.O. 50 W mercuryvapor bulb, 3650 A° excitation filter, and a 3970 A°barrier filter.

The rate of osteogenesis was estimated by the fluores-cent bands present at the mineralization front. Aneyepiece micrometer calibrated with a standard mi-crometer was used to measure the distance between thetwo bands. The readings were taken from the startingpoint of one band to the starting point of the otherband. The mineral appositional rate (MAR), i.e., rate ofbone formation at the basic multicellular unit (BMU)level, was calculated as follows:

MAR �Distance between the starting points of the 2 bands

Time interval between the 2 doses (�7 days)

The units used were micrometers per day. The readingfor each case was taken at least at three different loci,and the average (mode) of these was taken as the finalreading. All measurements utilized a 16� objective toobtain the maximum corrected value.

At 16� objective power, 100 ocular divisions corre-sponded with 62 divisions of the stage micrometer:

1 ocular division � 0.62

mm � 6.2 � 10�3

mm100

� 6.2 µm

Results

Thirty-seven consecutive patients of intertrochantericfracture of femur of various ages and either sex wereinitially chosen for the study. A total of 64 slides wereprepared from core biopsy specimens of these patients.Tetracycline fluorescence could be visualized in 58slides, and in 34 slides well-defined double bands wereappreciated (Fig. 3). Five patients were eliminated fromthe study as double bands were not obtained in theslides prepared from their core biopsy material (includ-ing 1 patient in whose slide double bands were not welldefined). Thus, the final number of subjects in our studywas 32. To calculate the mineral oppositional rate

(MAR), i.e., rate of new bone formation at theBMU level, the interband distance was measured. Themean MAR for each of the three groups was asshown in Table 1. The distribution of subjects in variousgroups of Singh’s index is shown in Table 2, and thedistribution of subjects for the metacarpal index isshown in Table 3.

A regression equation was calculated to look for cor-relation between the following:

i. Singh’s index and metacarpal index (SI vs. MI)ii. Singh’s index and mineral appositional rate (SI vs.

MAR)iii. Metacarpal index and mineral appositional rate (MI

vs. MAR)

No significant correlation was found between Singh’sindex, metacarpal index, and the mineral appositionalrate in group I and III; however, Singh’s index signifi-cantly correlated with the metacarpal index in group II.A significant correlation could also be established forthe subject population as a whole (n � 32) (Table 4).

Discussion

Patients who have osteoporosis are at increased risk offracture because of reduced bone quantity as well as

Fig. 3. Discrete single and double fluorescent tetracyclinelabels characterize active bone remodeling (�16)

Table 1. Patient data: age, SI, MI, and MAR

AverageAge Average MI

Group (years) Average SI (%) Average MAR (µm/day)

I (n � 11) 33.36 4.73 49.62 (SD � 5.56) 1.45 (SD � 0.35)II (n � 8) 54.75 3.75 48.10 (SD � 10.21) 1.49 (SD � 0.58)III (n � 13) 70.08 4.00 42.80 (SD � 8.97) 1.50 (SD � 0.29)

Total (n � 32) 53.63 4.16 46.47 (SD � 8.63) 1.48 (SD � 0.38)

SI, Singh’s index; MI, metacarpal index; MAR, mineral appositional rate; SD, standard deviation

229S. Singh et al.: Bone dynamics

bone quality. As with other clinically silent conditions(e.g., hypertension, hypercholestrolemia), the only wayto identify people who have osteoporosis is with appro-priate testing. Previously, routine X-rays were used todiagnose osteoporosis, however it is now well estab-lished that a loss of at least 30% of bone mineral isrequired before an unequivocal Roentgen logic diagno-sis of osteoporosis can be made.8 Now, newertechniques called bone densitometry are available tomeasure bone quantity that are more accurate and pre-cise, but these methods are very expensive and are in aphase of evaluation in developing countries. We haveused conventional X-rays to grade osteoporosis by

Singh’s index and metacarpal index for the followingreasons:

1. The method is technically simple (only X-rays ofpelvis with bilateral hip—AP view in 20° internalrotation and X-ray right hand AP view are required).

2. It is inexpensive.3. It is easily accessible and quite safe.4. It is universally available and can be undertaken

even at district level.

Although Singh’s index correlated well with metacarpalindex in our study, inter-observer variations were no-ticed at times. Considerable subjectivity and intero-

Table 2. Distribution of patients according to Singh’s index

Grading of osteoporosis by Singh’s index

Subgroup(age in

Group years) I II III IV V VI

Group I �30 — — — — 3 (27.3%) 1 (9.1%)31–44 — 1 (9.1%) — 3 (27.3%) 2 (18.1%) 1 (9.1%)

Group II 45–60 1 (12.5%) — 2 (25%) 2 (25%) 3 (37.5%) —Group III 61–74 — — 2 (15.4%) 2 (15.4%) 4 (30.8%) —

�75 — 1 (7.6) — 4 (30.8%) — —

Total 1 (3.1%) 2 (6.2%) 4 (12.4%) 11 (34.5%) 12 (37.6%) 2 (6.2%)

Table 3. Distribution of patients according to metacarpal index (MI)

Metacarpal index (%)

Group Age (years) 21–30 31–40 41–50 51–60 61–70

Group I �30 — — 3 1 —31–44 — 1 6 — —

Group II 45–60 1 — 2 4 1Group III 61–74 — 4 3 — 1

�75 — 2 3 — —

Table 4. Correlation between SI, MI, and MAR

Group Variables Regression equation P value Significance

I (n � 11) SI vs. MI — 0.943 Not significantSI vs. MAR — 0.522 Not significantMI vs. MAR — 0.501 Not significant

II (n � 8) SI vs. MI MI � 23.906 6.452 � SI 0.004 SignificantSI vs. MAR — 0.605 Not significantMI vs. MAR — 0.604 Not significant

III (n � 13) SI vs. MI — 0.097 Not significantSI vs. MAR — 0.353 Not significantMI vs. MAR — 0.525 Not significant

Total (n � 32) SI vs. MI MI � 30.817 3.766 � SI 0.004 SignificantSI vs. MAR 0.682 Not significantMI vs. MAR 0.639 Not significant

SI, Singh’s index; MI, metacarpal index; MAR, mineral appositional rate

230 S. Singh et al.: Bone dynamics

bserver variation have been well documented, forSingh’s index.3

Readings were difficult to interpret at times becauseof the inability of the patient to maintain a position ofinternal rotation at the hip, as a consequence of whichfull profile of the femoral neck could not be obtained,and inability to see the trabeculae clearly due to softtissue shadow in a fat patient with bulky thigh andbuttocks.

In our study, 7 of the 32 patients were found to haveosteoporosis according to Singh’s index (i.e., grade IIIor less), one in group I and 3 patients each in group IIand III, i.e., they had trabecular osteoporosis. Seven-teen of the patients had a metacarpal index of less than50%: 2 in group I, 3 in group II, and 12 in group III. Nocorrelation was found between Singh’s index and meta-carpal index for patients in groups I and III, indicating adisproportionate loss of cortical and trabecular bone inthese patients, whereas in group II a significant correla-tion was found indicating a proportional loss of corticalas well as trabecular bone in these patients.

Frost (1969) labeled subjects with demethylchlorte-tracycline (i.e., Declomycin), 10–15mg/kg/day given for4 days orally. An identical second label was given 10days later, and a core biopsy specimen was taken at least4–10 days later.7 Ramser et al. proposed that if osteoma-lacia is suspected or known, the interval between thetwo labels should be approximately 3 weeks and theinterval between the last label and the biopsy should bethe same.12 In 1989, Sherrard and Maloney introducedthe single-dose tetracycline labeling regimen.13 Theyfound that when conventional and single-dose labelswere given to the same patient, label lengths were virtu-ally identical. We used the single-dose regime in ourstudy because of the following advantages: Tetracyclinedoses could be given under direct supervision, and thetotal time period for this schedule is 8 days as comparedto 18 days for the conventional schedule proposed byFrost et al., which ensured that there is no delay in thesurgical intervention planned for the patients; the corebiopsy had to be taken during the surgery to make theprocedure painless.

It seems that improved compliance and shortenedlabel duration make the single dose label a preferablealternative to previous methods. As the smallest in-crement of the micrometer used for taking the mea-surements was 6.2µm and the distance between the twobands too low (average, 10.36µm) in our study,difficulty was encountered while taking the measure-ments. However, the dose-free interval of 6 days (re-sponsible for the short interband interval) was justified,as all the patients had to undergo surgical interventionat the earliest possible time.

In our study, the distance between the two bands wasmeasured from the point of commencement of one

band to the point of commencement of the other band,considering that the start of each tetracycline band sig-nifies the beginning of deposition of the label and isa constant point in relationship to the mineralizationprocess. The width of any fluorescent band is assumedto depend on the growth rate at the time that the labelwas given and on the length of time that the label wasavailable for deposition. Thus, the midpoint of a bandis likely to be quite variable in position in relation tothe point of deposition of tetracycline at the end of thephase of rapid mineralization. Measurements based onthe midpoints of successive bands (Frost 19604; Frost19697; Melsen and Mosekilde 198110; Qui et al. 199011)are likely to be much less accurate than those based onthe points of commencement of the bands.15 Measure-ments based on this argument are also useful in caseswhere the second band is followed by a diffuse fluores-cence and its endpoint or midpoint could not be identi-fied accurately.

Frost (1969) had shown that the rate of new boneformation decreases with increasing age7; Melsen andMosekilde (1981) showed that the mineral appositionalrate is independent of age.10 In our study, we found thatthere was a marginal increase in MAR with age that isnot statistically significant; i.e., MAR is independentof age. Frost (1966) concluded, from his work, that inosteoporosis there is no characteristic change in boneformation, resorption, or remodeling rate.1 Melsen andMosekilde (1981) found that MAR in patients of oste-oporosis was either normal or reduced as compared tothat of normal subjects.10 Sherrard (1989) noticed thatMAR in a single case of osteoporosis was lower thanthat in the normal.13 In our study, no correlation wasestablished between the severity of osteoporosis (on thebasis of grading by Singh’s index and metacarpal index)and the rate of new bone formation in either of the agegroups and as a whole. In other words, the severity ofosteoporosis does not influence the rate of new boneformation. The rate of new bone formation in a healthysubject and a patient of severe osteoporosis can be thesame (see Table 4).

No other study on correlation of radiological gradingof osteoporosis and rate of osteogenesis is available inthe literature for comparison.

Conclusion

1. The single large dose regimen, as introduced bySherrard, is far superior to the conventional doseregime of tetracycline because the doses can be givenunder direct supervision and the total time period forthe tetracycline labeling is significantly reduced from18 to only 8 days, thus avoiding any delay in opera-tive intervention.

231S. Singh et al.: Bone dynamics

2. Mineral appositional rate does not increase signifi-cantly with age, i.e., MAR is independent of agefactor.

3. There is no correlation between the severity of oste-oporosis (on the basis of grading by Singh’s indexand the metacarpal index) and the rate of new boneformation in either sex and among the various agegroups.

4. Osteoporosis in a patient can be caused by dispro-portionate cortical and medullary bone loss.

5. In our study, only 7 of a total of 32 patients withintertrochanteric fracture of the femur were found tohave definite radiological osteoporosis. This findingimplies that intertrochanteric fractures are equally orrather more common in non-osteoporotic individu-als, at least in this part of the world.

References

1. Bell GH, Dunbar O, Beck JS, et al. Variations in strengthof vertebrae with age and their relation to osteoporosis. CalcifTissue Res 1967;1:75–86.

2. Chanassieux P, Arlot M, Meunier PJ. Clinical use of bone biopsy.In: Marcus M, editor. Osteoporosis. San Diego: Academic Press;1996. p. 1113–21.

3. Edeiken J, Dalinka M, Karasick D. Edeiken’s Roentegen diagno-sis of diseases of bone. 4th ed. Baltimore: Williams & Wilkins;1990. p. 1085–315.

4. Frost HM, Villanueva AR. Observations on osteoid seams. HenryFord Hosp Med Bull 1960;8:212.

5. Frost HM. Tetracycline labeling of bone and the zone of demarca-tion of osteoid seams. Can J Biochem Physiol 1962;40:485–9.

6. Frost HM. Bone dynamics in metabolic bone disease. J BoneJoint Surg Am 1966;48A(6):1192–203.

7. Frost HM. Tetracycline based histological analysis of boneremodeling. Calcif Tissue Res 1969;3:211–37.

8. Greenspan A. Orthopaedic radiology: a practical approach. Phila-delphia: Gower; 1998. p. 19.1–19.7.

9. Kanis JA. Bone density measurements and osteoporosis. J InternMed 1997;241:173–5.

10. Melsen F, Mosekilde L. The role of bone biopsy in the diagnosisof metabolic bone disease. Orthop Clin North Am 1981;12:571–602.

11. Qui Ming-Cai, Wang Jue-Ying, LiWen-Sheng, et al. Dynamicstudy of iliac trabeculae of normal Chinese. Chin Med J 1990;103:363–8.

12. Ramser JR, Villanueva AR, Frost HM. Cortical bone dynamicsin osteomalacia measured by tetracycline bone-labelling. ClinOrthop 1966;49:89–102.

13. Sherrard DJ, Maloney NA. Single-dose tetracycline labellingfor bone histomorphometry. Am J Clin Pathol 1988;91(6):682–7.

14. Singh M, Nagrath AR, Maini PS. Changes in trabecular patternof the upper end of femur as an index of osteoporosis. J BoneJoint Surg 1970;52(A):457–67.

15. Tam CS, Reed R, Campbell JE, et al. Bone growth kinetics. I:Modifications of the tetracycline labelling technique. J Pathol1974;113:27–38.

16. Whyte MP, Bergfeld MA, Murphy WA, et al. Postmenopausalosteoporosis: a heterogenous disorder as assessed by histomor-phometric analysis of iliac crest bone from untreated patients. AmJ Med 1982;72:193–201.