JOURNAL OF ENDODONTICS Printed in U.S.A.

Copyright © 2003 by The American Association of Endodontists VOL. 29, NO. 11, NOVEMBER 2003

Roentgenographic and Direct Observation of

I. B. Bender,* D.D.S., and Samuel Seltzer,** D.D.S., Philadelphia

Artificial lesions simulating pathological conditions diolucency. The purpose of this study was to find out under what
were made in mandibles from human cadavers. specific conditions bone destruction in vitro was not detectable by
Roentgenograms of the lesions were made and roentgenographic examination, and to correlate the known location
compared with the specimens of bone. Lesions in and extent of artificially produced bone lesions with the roentgen-
ograms of those lesions. The general procedure was to create
cortical bone can be detected roentgenographi-
lesions of various sizes, gradations and anatomic locations in
cally only if there is perforation of the bone cortex, mandibles. The visible evidence of bone destruction was then
erosion from the inner surface of the bone cortex, compared with the appearance of these lesions in the usual dental
or extensive erosion or destruction from the outer roentgenograms.
surface. Lesions in cancellous bone cannot be de-
tected roentgenographically. Extensive disease of
bone may be present even when there is no evi- MATERIALS AND METHOD
dence of it on roentgenograms.
Human mandibles obtained at necropsy were studied. Both wet
and dry specimens were used. Color photographs and roentgeno-
grams of the specimens were taken prior to experimentation. The
mandibles were cut into block sections. Experimental lesions of
It is not uncommon to find clinical signs of bone disease in spite various types were then made with diamond stones, endodontic
of negative roentgenographic findings. For example, in many in- reamers, files and dental burs. Roentgenograms were made of the
stances of acute alveolar abscess, even though pus is draining experimental lesions in the bone sections. All roentgenograms
through the root canal, no changes can be detected by roentgeno- were taken at the same target film distance and angulation (except
graphic examinations. In many of the acute total pulpitides the for the deliberate variation of angulation described in the first
initial roentgenograms fail to indicate pathological changes, yet experiment). All films were exposed for two seconds at 65 kvp, 10
roentgenograms taken several days later may show changes in the ma. and were developed together for three minutes.
apical tissues.1 Periodontal lesions involving the bifurcation are
not always seen on roentgenographic examination. Roentgeno-
grams taken after extraction of teeth often fail to show distinct EXPERIMENTAL EVIDENCE
areas of rarefaction, in spite of the extensive cavitation. The
trabeculae over the sockets appear normal and if the lamina dura Effects of Depth of Lesions in Cortical Bone
could somehow be obliterated there would be no way of determin-
ing roentgenographically that a tooth had been removed recently. Holes were drilled into a section of the mandible with round
Instrumentation of the root canal beyond the apex with reamen and burs of various sizes (no. 2 to no. 8). The buccal plates were drilled
files usually does not produce roentgenographic changes in the (1) to a depth of 1 mm., (2) halfway into the buccal cortex, (3) all
trabecular pattern of bone even though the instruments have dis- the way through the buccal cortex (perforation) and (4) deep
placed some of the cancellous bone. In acute mastoid, march enough to perforate both buccal and lingual bone plates. Roent-
fracture, and osteomyelitis, roentgenograms also often appear neg- genograms of these artificial lesions were taken at various angu-
ative. lations.
Thus it seems that extensive bone destruction may occur under
some conditions without being detectable through changes in ra-
Effect of Depth
From Albert Einstein Medical Center, Northern Division, department of dentistry.
* Associate professor in oral medicine.
** Associate professor in histopathology, School of Dentistry, University of
Lesions 1 mm. deep did not appear on the roentgenogram,
Pennsylvania. regardless of the size of bur used. As the depth of lesions increased
† Bender IB, and Seltzer S. Roentgenographic and direct observation of experi- there was greater radiolucency and the shadows became more
mental lesions in bone I. J Am Dent Assoc 62:152-60, 1961. Copyright (c) 1961
American Dental Association. All rights reserved. Reprinted by permission of ADA pronounced, irrespective of bur size. Thus loss of a superficial
Publishing, a Division of ADA Business Enterprises, Inc. layer of the outer surface of bone is not ordinarily detectable. As

702
Vol. 29, No. 11, November 2003 Classic Article 703

FIG 1. A: Four holes of different depths were drilled in the mandible with a no. 8 round bur. Roentgenograms show only two or three holes,
depending on angulation. B: Roentgenogram showing three holes. The fourth is not perceptible. The radiolucency increases with depth of cut.
The shadow is darkest in area where cortex has been perforated. C: Note change in shape and size of holes and gradations in shadow as
angulation is changed. The hole made by cutting halfway through the cortex can hardly be seen. D: Note further change in shadow and shape
of hole as the angle is increased. Also, distal root is completely denuded of bone in B. A change in angulation produces an impression of bone
regeneration.

more and more bone is removed from a given region from the outer structure was removed until there was a complete hollow within
cortex inward the shadow deepens and the area of rarefaction the bone. All the bone marrow was removed up to the junction of
becomes more pronounced, hence more readily observed. the innermost surface of the cortex and the cancellous portion of
The roentgenograms were similar, whether taken from the buc- the bone. Roentgenographic examination still failed to show
cal or lingual side. From the roentgenographic evidence alone it changes in the trabecular pattern of the bone or in its radiolucency.
would not have been possible to determine on which side the When a bur eroded the innermost surface of the bone cortex a clear
lesions had been made. and distinct radiolucent shadow appeared.
This evidence indicates that destruction of the cancellous por-
tion of the bone does not produce changes discernible in roent-
Effect of Angulation
genograms taken under the conditions described. Changes become
manifest only if there is encroachment on the innermost surface of
As the roentgenographic angulation was increased, the rounded
the cortical bone or if there is frank perforation (Fig. 2 and 3).
shadows became elongated and their densities decreased. These
effects of angulation were especially pronounced for perforations.
The elongation of shadows increased with the depth of the cut as
well as with the degree of angulation, a fact to be taken into Artificial Lesions in the Cortex and its Junction with
Cancellous Bone
consideration in evaluating roentgenograms. It was not so much
the diameter of the bur as it was the depth of the cut that produced
A block section of the mandible was prepared in the manner
a recognizable lesion (Fig. 1).
previously described. The cancellous bone up to the junction with
The angulation effect provides an explanation of the roentgen-
the cortex was removed and a roentgenogram was taken. Then the
ographic images of the nasopalatine and the mental foramens. At
bone was split in half mesiodistally, to yield separate lingual and
some angulations, particularly in roentgenograms of the nasopal-
buccal plate sections, each consisting of cortex and trabeculae at
atine area, the image produced is more cylindrical than circular. In
the junction. In the lingual section, the trabeculae at the junction
an occlusal film, made with the x-ray tube positioned on the
foramen, a more distinct circular shadow is obtained. were scraped with a bone chisel. In the buccal section, the tissue at
the junction was left undisturbed. Roentgenograms of the section
with junction trabeculae reduced, showed loss of trabecular struc-
Artificial Lesions in Cancellous Bone ture or pattern. The bone had a foamy appearance (Fig. 3,C).
Roentgenograms of the section with junction trabeculae intact
The cancellous bone of the mandible was drilled to various showed an intact trabecular pattern (Fig. 3,D).
depths and roentgenograms were made. No evidence of drilling A part of the separate buccal plate was sliced sagitally from the
could be seen in the roentgenograms. More and more cancellous outer surface with a diamond stone to remove the cortical bone
704 Bender and Seltzer Journal of Endodontics

FIG 2. A: After cancellous bone has been removed. There is no apparent evidence of any disturbance. B and C: Specimen of bone showing
amount of cancellous structure that was removed. Notched area is part of lingual canal. D: Side view shows absence of cancellous bone and
small amount of trabeculae present at the innermost surface of the bone cortex. Trabeculae are absent in notched area that forms lingual aspect
of mandibular canal. Roots of molar tooth articulate with the buccal cortex. Distal root appears to be within cancellous bone.

(Fig. 3,G,H,I). The other part was not cut so that comparative
effects could be observed on the same specimen. The cortex was
removed until the junction area of the cortical and cancellous bone
was exposed. The specimen thus produced was about 2 mm. thick.
Bur marks were then drilled in (1) the outer surface of the buccal
plate, (2) the inner surface of the junction area where the outer
buccal plate had not been removed and (3) the inner surface of the
junction area where the cortex had been removed.
Roentgenograms revealed definite and distinct areas of rarefac-
tion where the bur had been drilled into the inner or outer surface
of the cortex. The bur mark was barely discernible in the junction
area although the bone was almost perforated. The trabecular
pattern was intact on both sides of the bone, but there was a
pronounced difference in the densities (Fig. 3,E and F).
From these experiments it may be inferred that areas of rarefac-
tion manifest themselves only if there is erosion of the cortex from
the inner or outer surface or if there is frank perforation. The
trabeculae indicated on the roentgenograms are those which are
present at the junction of cortex and cancellous bone. If the
trabeculae in the interior of the cancellous bone or in the marrow
are destroyed, the difference of the trabecular pattern on the roent-
genogram is not disturbed.

Trabecular Patterns in Roentgenograms of Intact and FIG 3. The split halves of bone specimen in Figure 2 are shown
Dissected Junction Areas above. A: The lingual portion in which the trabeculae were scraped.
Foamy appearance can be seen in the roentgenogram, C; as a result
of reducing trabecular structure. Mandibular canal can also be seen.
A mandible obtained from a 48 hour autopsy specimen was
B: Buccal portion in which the trabeculae were left intact. The
sectioned and graded amounts of cortex were removed from a roentgenogram, D, was taken prior to drilling holes in trabecular
portion of both lingual and buccal sides. Graded amounts of the structure. E: Before removal of buccal plate up to the junction area.
junction area also were removed so that a 3 mm. and a 2 mm. Note that trabecular pattern can still be discerned. F: Showing
thickness of cancellous bone remained. distinct areas of radiolucency as a result of drilling the cancellous
In roentgenograms, the trabecular pattern and trabeculae could and cortical bone to the junction area from cancellous and cortical
be seen when the junction was not encroached upon. As more and sides. G, H and I: Specimens of bone illustrating side views, the
more of the junction area was removed, the bone became more sagittal cut to the junction area and the bur mark drilled to junction
radiolucent and the trabeculae and the trabecular pattern became area. Note denuded areas of bone on root surface.
Vol. 29, No. 11, November 2003 Classic Article 705

FIG 4. Roentgenographic changes occurred as cortical bone was FIG. 5. Numerous nutrient foramens are present in anterior portion of
removed mesial to first molar. When junction area of bone was mandible. On lingual surface foramens run in horizontal direction,
removed there was loss of trabecular pattern with hardly any visu- whereas on labial aspect they run in vertical direction. Reamers
alization of cancellous structure. A: Before cortex was removed. B: perforate cortex and enter cancellous portion of bone.
After removal of buccal plate to junction area. C: After removal of
lingual and buccal plate. Some junction area is present. D: Further
removal of junction area. E and F: Complete removal of junction then dissected out by removing the buccal or lingual plate of bone
area. Views from lingual and buccal sides respectively. G: Specimen with diamond stones. The nutrient canals were observed as small
before experiment. H and I: After experiment. tubes that traverse the cortex vertically on the buccal side. They
also traverse the cortex in a general horizontal direction (penetrat-
more difficult to discern (Fig. 4). The inference is that the trabec- ing the bone at a 45 degree angle) on the lingual side in the region
ular pattern originates at the junction of the cortex and cancellous of the genio tubercles as shown in Figure 5.
bone. The mental foramen was most radiolucent when the central
Changes in trabecular pattern were the same whether bone was beam was parallel with the orifice. At 180 degree angulation the
removed from the junction area on the outer (cortical) side or foramen appeared darker and more distinct; at 90 degrees it ap-
trabecular extensions were reduced in depth from the inner (can- peared less dark and more diffuse. The roentgenographic image of
cellous) side at the junction area. the foramen opening varied from circular to funnel shaped.
When the bone plate over dentigerous regions is removed com- The mandibular canal appeared more distinct and rarefied when
pletely, changes in pattern do not readily occur because the lamina the cortex was removed. Examination of the specimens in some
dura has junction areas of cancellous bone that affect the ultimate instances showed the canal notched throughout the body of the
trabecular appearance. This explanation is in accord with the mandible in the molar and bicuspid region on the lingual side. In
observations of Goldman, Millsap and Brenman.2 other instances the canal was surrounded by a distinct layer of bone
Cancellous bone is comparatively radiolucent and it is less plate suspended in the cancellous structure.
dense than cortex or alveolar bone proper. The alveolar bone The roots, in several instances, appeared to be denuded of
proper is dense because of its relatively smaller content of fibrillar buccal bone near the apical third (Fig. 3,F). This condition could
matter and large content of cementing substance. The cementing not be detected by means of roentgenograms. In many instances,
substance is particularly dense because of its greater content of the bone was extremely thin in the apical areas on the buccal side.
calcium salt per unit volume.3 Superimposition of bone or a mon- Such denuded areas were not observed on the lingual side.
tage can increase the bone quantity to such an extent that the effect The teeth were encased completely in lamina dura which joined
of greater density is produced in the roentgenograms. This effect is with the outer cortex. The alveolus and the lamina dura are cortical
observed, for instance, in roentgenograms of the external and bone; the lamina dura therefore has a junction area in contact with
internal oblique ridge region. trabeculae. The presence of lamina dura accounts for the greater
amount of trabeculae found in dentigerous regions, especially
between the teeth and towards the alveolar ridge (Fig. 3,D).
Other Observations Where the alveolar process was thick in the molar region and
there was a layer of cancellous bone between the lamina dura and
The nutrient canals could be seen roentgenographically only the cortical plate, trabeculae could be seen across the roots of the
when they were present in the bone cortex. For direct examination, teeth on the alveolar process. The bone appeared normal when the
the foramens were located with endodontic reamers or files and teeth were extracted. Where the cortex was thin, or thick but with
706 Bender and Seltzer Journal of Endodontics

no intervening cancellous bone, the bone specimen had a foamy large lesions may go undetected with the usual roentgenographic
appearance and trabeculae could not be seen in the region where procedures. Early metastatic carcinoma in bones often cannot be
the teeth were extracted. detected by means of a roentgenographic examination, even
though the patient has bone pain. As the disease progresses it
destroys more of the marrow spaces, invades the cortex and pro-
DISCUSSION duces lesions which show up on roentgenograms. Shackman and
Harrison6 have demonstrated that a patient may have extensive
The assumption that results obtained with cadaver material are metastases without demonstrable roentgenographic abnormality.
valid for bone under clinical conditions receives support from the However, with tomography large areas in cancellous bone may be
experimental work of Ardran.4 He removed about one half in or of detected by tomograms made in the correct plane.
the inner cancellous structure of vertebrae obtained from human Although similar experiments have not been made on the max-
cadavers and found that the artificial lesions could be seen in illa, similar findings could reasonably be expected.
roentgenograms taken laterally. When water was put into the bone
cavities, the lesions could not be seen from any angle. In such
experiments water is added to produce an environment more SUMMARY
closely approximating the radiolucency of the tissue fluid associ-
ated with bone in vivo. Ardran’s work seems to indicate that Mandibles from human cadavers were dissected and artificial
roentgenographic examination of bone in vivo would be even less lesions were made to simulate pathological conditions. A compar-
discriminating than it would be for specimens obtained from ca- ison was made of roentgenogram with the actual sections of bone,
davers. as observed visually.
There is much histologic evidence of apical or periodontal It is evident from these experiments that, by the methods ordi-
disease without roentgenographic manifestation. A review of Bur- narily employed for taking roentgenograms, lesions in cortical
ket’s5 protocols of human necropsy material shows that in many bone can be detected roentgenographically only if there is perfo-
instances roentgenographic examination yielded negative results ration on the bone cortex, erosion from the inner surface of the
when cancellous bone was diseased, and sometimes even when bone cortex, or extensive erosion or destruction from the outer
cortex was involved—probably because only superficial cortex surface. Lesions in cancellous bone cannot be detected roentgen-
was removed. ographically.
Goldman, Millsap and Brenman2 observed that the removal of The apparent cancellous destruction that is manifest on roent-
the buccal and lingual alveolar plates had no effect on the trabec- genograms is really an erosion of the innermost surface of the bone
ular pattern around the teeth. This observation is not contrary to our cortex at the junction are between cortex and cancellous bone.
finding that trabecular patterns change if the innermost surface of No defect can be visualized beyond the junction area as it
the cortex is removed. They did not attach any significance to the encroaches on the marrow spaces.
junction area; they were describing the region around the teeth in Extensive disease of bone, therefore may be present even when
a dry specimen, whereas we were observing the body of the there is no evidence of it on roentgenograms.
mandible in a wet specimen. In and around the teeth there is more
cortex, crestal bone or lamina dura, and therefore more junction
surface to exhibit trabeculae. Where junction surface is present the
trabecular pattern does not change.
References
In view of the experimental evidence presented, early stages of
bone disease cannot be detected by means of roentgenograms. Also 1. Seltzer S. The role of endodontics in complete mouth reconstruction.
the size of a rarefied area on the roentgenogram is not correlated J.A.D.A. 1955;51:320.
2. Goldman HM, Millsap JS, Brenman HS. Origin of registration of the
with the amount of tissue destruction. A small area of rarefaction architectural pattern the lamina dura and the alveolar crest in the dental
can be indicative of as much or more bone destruction as a large radiograph. Oral Surg., Oral Med. & Oral Path. 1957;10:741.
3. Sicher H. Some principles of bone pathologies. J. Oral Surg. 1949;7:
rarefied area on the roentgenogram. 104.
The afore-mentioned experiments emphasize that routine roent- 4. Ardran GM. Bone destruction not demonstrated by radiography. Brit. J.
genograms may not detect the presence of secondary neoplasms or Radiol. 1951;24:107.
5. Burket LW. Human necropsy protocols. Unpublished data.
inflammation causing bone destruction. Whereas this point has 6. Shackman R, Harrison CV. Occult metastases. Brit. J. Sur. 1948;35:
been recognized for minute lesions, it has not been realized that 385.