the rib hump inidiopathic scoliosis...64 the journal of bone and joint surgery the rib hump...

64 THE JOURNAL OF BONE AND JOINT SURGERY

THE RIB HUMP IN IDIOPATHIC SCOLIOSIS

MEASUREMENT, ANALYSIS AND RESPONSE TO TREATMENT

T. THULBOURNE, BRIDGE OF EARN, SCOTLAND, and R. GILLESPIE, TORONTO, CANADA

Front the Hospitalfor Sick Children, Toronto, Ontario

This paper describes a simple method for the recording of rib deformity in idiopathic scoliosis. The

relationships have been recorded between the measured rib hump and rib depression deformities and I) the

rotation of the vertebral bodies (as measured by the method of Nash and Moe on the standing radiograph);

2) the degree of lateral curvature (as measured by the method of Cobb on the standing radiograph); and

3) the rib-vertebra angles and their differences (as described by Mehta). No clear linear relationships were

found. Many examples of irregular relationship were recorded, for example, marked spinal rotation with

minimal rib hump. The response of the rib deformities to treatment by Milwaukee brace in fifty-two patients

is described ; the hump is little changed but the depression on the opposite side may be considerably reduced.

Harrington instrumentation may have a similar effect.

In idiopathic scoliosis the most important compo-

nents of the deformity are the thoracic rib hump, the

lateral shift of the chest relative to the pelvis and the

asymmetry of the shoulders. The last two are easily

recorded photographically and radiographically and on

the whole are correctable by either bracing or operation.

The rib hump is the feature of the deformity most

resented by the patient, least understood by the surgeon

and most resistant to treatment.

Great ingenuity has been demonstrated in the de-

velopment of instruments to record the deformities of

scoliosis. These mechanical devices reached what was

probably a peak of complexity in 1913 when Layer-

miccoca described his pantoscoliosogr�fo and plasmo-

scoliosometro. With the general use of radiography,

however, increasing emphasis was placed on the radio-

graphic appearances. Although the rib deformity has

aroused interest because of its variable response to

treatment, few attempts have been made in recent years

to measure its extent accurately. The authors are aware

of only one modern instrument designed to chart this

feature, that described by Rippstein in 1967.

This paper presents a simple method of measuring

and recording the rib deformity and analyses its relation-

ship to lateral curvature, to rotation of the spine and to

the rib-vertebra angles. Finally, the response of the

deformity to treatment by the standard Milwaukee brace

and by standard Harrington instrumentation is described.

THE MEASURING DEVICE

The instrument consists of a series of movable strips

which can be locked in position by a lever on the frame

(Fig. I ). The central strip is marked and the frame

carries a spirit level. In clinical use the patient places her

hands together and flexes hips and spine until the hands

are between the extended knees at the level ofthe patellae.

The instrument is placed across the back, centred on the

spinous process of the apical vertebra and perpendicular

to the spine at this level. After ensuring from the spirit

level that the instrument is horizontal and that each

movable element is in contact with skin, the locking lever

is depressed and the resulting contour is transferred to

graph paper. By aligning the vertical elements of the

instrument with the lines on the graph paper, the profile

of the back is accurately reproduced as a chart. Serial

recordings afford a good visual impression of the pro-

gress of both rib hump and valley during treatment

(Fig. 2).

MATERIAL AND METHODS

Measurement of hump and relationship to radiographic

featuresUsing this device we measured the rib hump deformity infifty-two patients with untreated scoliosis. In order to producea numerical representation of the rib deformity, each chart

was reviewed and three figures, which can be read directly,were obtained (Fig. 3): H-the height of the hump; W-the

distance ofthe peak ofthe hump from the midline; and D-thedepression of the ribs at the same distance. In addition thegradients of the hump (H/W) and of the depression (D/W)were calculated. We believe that these gradients give a moreaccurate representation of the deformity than the heightsalone, because they eliminate the influence of the width ofthe patient’s back.

Measurement of the lateral curvature was carried out onan antero-posterior radiograph of the erect spine by themethod ofCobb (1948). On the same radiograph the rotationof the apical vertebra was measured by the method of Nashand Moe (1969). Recognising that this method becomes

T. Thulbourne, F.R.C.S., Orthopaedic Department, Bridge of Earn Hospital, Bridge of Earn, Perthshire, Scotland.

Dr Robert Gillespie, F.R.C.S.(Ed.), F.R.C.S.(C), Suite 1028, 123 Edward Street, Toronto, Ontario, Canada. (Address for reprints).

:�.�

� ;�

1�.J;TI

�1

Fi-lu. I

Figure I-The measuring device in use.

CURVATURE (degrees) under 21 21-30 31-40 41-50 over 50

NUMBER OF PATIENTS 8 15 13 1 1 5

DEFORMITY Range 0-19 6-38 9-24 10-33 7-43(H . D millimetres) Mean 10 15 16 18 28

HUMP Range 0-022 003-022 007-025 002-036 010-044GRADIENT Mean 009 015 019 019 025

DEPRESSION Range 0-022 0-033 008-037 011-035 0-02-030GRADIENT Mean 016 016 022 024 020

ROTATION (degrees) under 10 10 15 20 over 20

NUMBER OF PATIENTS 13 8 12 13 6

DEFORMITY Range 0-25 7-18 9-31 7-33 10-43(H #{149}D millimetres) Mean 13 13 17 21 19

HUMP Range 0-022 003-022 007-036 007-033 003-044GRADIENT Mean 014 013 018 016 021

DEPRESSION Range 0-030 007-037 003-&33 002-034 013-035GRADIENT Mean 0-13 021 019 0-22 021

THE RIB HUMP IN IDIOPATHIC SCOLIOSIS 65

VOL. 58-B, No. I, FEBRUARY 1976

8

FIG. 2

Figure 2-Serial tracings of a rib cage deformity during treatment.

w

FIG. 3

w

Method of measuring the rib hump chart for analysis.H�hump height. W=distance of hump from midline.

D=rib depression.

TABLE I

LATERAL CURVATURE AND RIB DEFORMITY

TABLE II

VERTEBRAL ROTATION AT APEX OF TI-IORACIC CURVE AND RIB CAGE DEFORMITY

inaccurate beyond 20 degrees, as shown by Mehta (1973), werecorded the six cases with rotation of this severity as “over20 degrees”. The rib-vertebra angles at the apical vertebraewere measured by the method of Mehta (1972) and the R-Vangle differences (concave minus convex) were obtained. Anattempt was then made to determine if the rib cage deformityis directly related to any of these readings.

H/W

0.5

0.4

0.3

0.2

0.1

H+D (mm)

45

40#{149}

35

30

25

20

15’

l3�

5.

0.1

0.2

0.3 890.4

<21 � >50 ,w <21 � >50

LATERAL THORACIC CURVE

(DEGREES)

LATERAL THORACIC CURVE

(DEGREES)

FIG. 4

A histograph showing the relationship between rib deformityand lateral curvature.

H/W

<10 10 15 20 >20 <10 10 15 20 >20

ROTATION (DEGREES) ROTATION (DEGREES)

FIG. 5

A histograph showing the relationship between rib deformityand vertebral rotation.

The effect of bracing and of instrumentation on rib cage

deformityFor some time the authors had thought that Milwaukeebracing could reduce the rib hump. The introduction of amethod of serial recording of the deformity afforded anopportunity to test this opinion in thirty-seven patients overa period of one to two years. More recently we have beenable to observe the results of Harrington instrumentation andwe present the preliminary findings.

66 T. THULBOURNE AND R. GILLESPIE

THE JOURNAL OF BONE AND JOINT SURGERY

RESULTS

Rib deformity and lateral curvature-In order to assess

the influence of lateral curvature alone, without initially

considering the combined effect ofcurvature and rotation,

we allocated fifty-two patients to five groups according

to the severity of curve. Table I shows the range and

means, in each group, of the sums of the hump height

and depression (H+D) and of the two gradients (H/W

+ D/W).The pattern can be appreciated more readily in

histograph form (Fig. 4). As the severity of the lateral

curvature increased there was, as one would expect, a

general upward trend in both the total deformity and in

the hump gradient. This tendency was not so clearly

H/W

0-4

0-3

Q#{149}2

0�1

0

0�1

Q#{149}2

03

Q.4#{149}

D/WFIG. 6

A histograph in which each column represents a single patient.Charting the rib deformity in order of increasing rib humpgradient demonstrates no comparable progression of rib

depression gradient.

� shown in the gradient of the depression. Of more signifi-

cance, however, was the wide overlap in the different

groups. In patients with least curvature (under 21 de-

grees) there were humps of greater severity than the

means of all groups but the most curved. Conversely, in

those most curved (over 50 degrees) there were rib humps

even less than the mean of the group with the mildest

curves.

Rib deformity and vertebral rotation-When the fifty-two

patients were regrouped according to the degree of rota-

tion of the apical vertebra (Table II) and graphically

illustrated in the same manner as for lateral curvature

(Fig. 5), a similar result was obtained. The expected

trend ofincreasing deformity with increasing rotation was

present. Again there was extensive overlap ; some patients

with little rotation had marked rib humps and vice versa.

To assess the influence of vertebral rotation from

another aspect, consider the effect to be expected if a

TElL RIB HUMP lN IDIOPATHIC SCOLIOSIS 67


normal spine were rotated. The upward tilt on one side,

which produces in effect a rib hump, would be associated

with comparable depression on the other side. When,

however, each patient was charted in order of degree of

rib hump gradient (Fig. 6) it could be seen that the length

of the column below the axis, which represents the rib

depression gradient, was distributed in a random manner.

Thus the rib deformity cannot be a direct result of

vertebral rotation alone.

FiG. 7

Histographs showing the combined effect of lateral curvatureand vertebral rotation of 5 degrees or less.

FIG. 8

Histographs showing the combined effect of lateral curvature

and vertebral rotation of 20 degrees.

Rib deformity and combined lateral angulation and verte-

bral rotation-In attempting to determine if the deformity

is related to a combination of angulation and rotation,

we studied two groups of patients, with rotation of 5

degrees or less (Fig. 7) and of 20 degrees (Fig. 8).

The vertical axes represent, as before, the total rib

deformity (H+ D), and the hump and depression gradients

(H/W and D/W). In these two figures, each column

represents an individual patient, and the patients are

arranged in order of increasing lateral curvature. Within

groups of patients having the same degree of vertebral

rotation, the rib deformity was not significantly influenced

by the severity of the scoliosis. A similar analysis of the

remaining groups (10 and 15 degrees of rotation) revealed

a similar lack of correlation.

To illustrate this lack of correlation clinically, two

patients are presented. Figure 9 shows a girl aged fifteen

with a mild double curve. The thoracic curve, rather

unusual in being convex to the left, is the one producing

most of the clinical deformity and measures only 16

degrees from T.2 to T.9, with rotation of barely 5 degrees

at T.6. The lower curve, from T.8 to L.3, measures 22

degrees with no significant rotation. Figure 10 shows a

girl aged sixteen with severe scoliosis. The major curve,

from T.6 to T.l2, measures 50 degrees, with over 20

degrees of rotation at T.9. The upper curve, from T.2

to T.6, is 42 degrees, with 1 5 degrees of rotation at T.4.

As can be seen, the girl with the more deformed rib cage

is the one with the less marked spinal changes.

Rib deformity and rib-vertebra angles-These angles give

an indication of deformation in a coronal plane of the

ribs relative to the apical vertebra. The rib hump and

depression represent deviation of the ribs in a sagittal

plane, both as a result of vertebral rotation and of any

angulation relative to the vertebrae. Because we thought

there might be an interrelationship between these two

features, we regrouped the patients according to R-V

angle differences (concave side minus convex side). How-

ever, a familiar pattern emerged (Table 111, Fig. 1 1). A

slight upward trend in the severity of the deformity was

demonstrated, but again there was a wide overlap.

When one considers, independently, the effect of the

R-V angle of the convex side on the rib hump gradient

(Table 1V, Fig. 12), and of the R-V angle of the concave

side on the rib depression gradient (Table V, Fig. 13),

one sees a tendency for the rib hump to be more severe

if the ribs on that side are more depressed from the

horizontal, and for less marked rib depression as the ribs

on the concave side become more nearly horizontal. There

is, however, no evidence of a direct relationship.

The effect of bracing and instrumentation

The Milwaukee brace-The chart of one girl illustrates

the effect of Milwaukee bracing (Fig. 14). The tracings

made over a period of two years, from thirteen to fifteen,

demonstrated a common pattern of events. The positive

rib hump on the convex side remained fairly constant;

the rib depression, however, gradually became less notice-

able until finally it was actually reversed. Thus the total

deformity was much reduced. This change occurred

without change of the lateral curvature and with no

significant difference in vertebral rotation. The same

train of events, leading to an almost unchanged hump

but elevation of the depressed side, has been observed in

twenty-two of the thirty-seven cases.

Harrington insfrumentation-Although the nulnber of

cases is small, the pattern appears to be the same as that

with Milwaukee brace treatment. The tracings before and

after operation in three cases are given in Figure 15.

DISCUSSION AND CONCLUSIONS

In recent years there has been more emphasis on the

treatment of the rib hump deformity in scoliosis by such

means as resection of part of the prominent ribs or

osteotomy of the transverse process (Goldstein 1966,

:� 1FIG. 10

A severe curve of 50 degrees and over 20 degrees of rotation but with little rib cage deformity.



�w. .

p 4��’ � �

FIG. 9

A mild curve of 16 degrees and only 5 degrees of rotation but with a marked rib cage deformity.

D/W

0.4

0.3

0.2

0.1

D,,w8-V DIFFERENCE R-V DIFFERENCE

R-V DIFFERENCE (degrees) NEGATIVE 0-5 6-10 � 11-15 16-20 over 20

NUMBER OF PATIENTS 7 13 � 9 9 6 � 8

DEFORMITY Range 7-32 6-28 0-31 7-25 10-29 11-43(H-� D millimetres) Mean , 16 15 15 15 18 22

HUMP Range 003-030 0i0-0�2l 0-036 0-031 013-022 010-044GRADIENT Mean � 013 016 013 015 016 023

DEPRESSION Range 008-027 0-037 � 0-027 012-034 007-028 0-030GRADIENT Mean 021 � 017 018 023 022 018



FIG. II

Histographs showing the relationship of rib deformity to rib-vertebra angle difference.

H<60 60- 66- 71- >75 <7171- 76- 81- >85

65 70 75 75 80 85

CONVEX 8-V ANGLE CONCAVE 8-V ANGLE

(DEGREES) (DEGREES)

FIG. 12 FIG. 13

Figure 12-Histograph showing the re�ationship of theconvex R-V angle to the rib hump gradient. FigureI 3-Histograph showing the relationship between

concave R-V angle and rib depression gradient.

TABLE III

RIB-VERTEBRA ANGLE DIFFERENCE AND RIB CAGE DEFORMITY

TABLE IV

CONVEX R-V ANGLE AND RIB HUMP GRADIENT

CONVEX R-V ANGLE (degrees) � under 60 � 60-65 66-70 71-75 over 75

NUMBEROFPATIENTS 10 10 II 10 11

HUMP Range 0-044 0�02-036 009-025 0-02 1GRADIENT Mean 022 016 016 012

003-030012

TABLE V

CONCAVE R-V ANGLE AND RIB DEPRESSION GRADIENT

CONCAVE R-V ANGLE (degrees) under 71 � 71-75 , 76-80 81-85 over 85

NUMBEROFPATIENTS 13 II 8 8 12

DEPRESSION Range � 022-027 003-037 007-024 0-034GRADIENT Mean 017 020 019 020

0-#{216}�35021

1973), and the use of pressure pads within post-operative

casts (Cotrel and Morel 1964). This interest has increased

the need for a simple method of recording the deformity

rather than reliance on measurements of variations in

the height or slope of the back or on clinical photographs.

With a true outline of the back at its most deformed

level, one is better equipped to draw conclusions on the

effect of treatment. The authors believe that the instru-

ment described fulfils these requirements.

The severity of the rib cage deformity is widely

accepted as being a function of the degree of vertebral

rotation. Less accepted is the view that the rib changes

1-c.

FIG. 14

A series of rib hump graphs in a patient during treatment ina Milwaukee brace. Note the almost unchanged rib hump

but marked improvement in rib depression. FIG. 16

A diagram to show how rib hump and rib depression may beproduced by deformity of the posterior vertebral elements

without rotation of the vertebral body.

JUL

72

FE B74

MAY72

NOV

72

MAY

73T6-T11��5�

T6�T11:27’ \

FIG. 15

The improvement in rib depression but little change in ribhump shown in three patients after Harrington instrumen-

tation.



are always related to the lateral curve ; for example,

Goldstein (1973) stated that “some 75-degree lateral

thoracic curves have a milder and more correctable rib

�- -- ��1�26’

cage deformity than other patients with a 50- or 55-degree

curve”. We have been able to demonstrate, in this study

of fifty-two untreated cases, that the rib cage deformity

bears no direct relationship either to the degree of verte-

bral rotation or to the lateral angulation or to both.

Even the rib-vertebra angles and their differences, which

one might expect to bear some relationship to the de-

formity of those same ribs, give no indication of the

severity of the deformity.

As the rib hump and rib depression are not related

directly to any of the radiographic features normally

measured, there remains the question as to what is their

true causation. It is well known that in scoliosis the

posterior elements show a torsional deformity (Fig. 16),

with deviation of the spinous process towards the con-

cavity of the curve and asymmetry of pedicles, laminae

and transverse processes (Roaf 1960 ; James 1967 ; Enne-

king and Harrington 1969). This deformation occurs

independently of rotation. In our opinion, it is this

deformation which is the major factor contributing to

the altered shape of the rib cage, with vertebral angulation

and rotation contributing relatively little. Unfortunately,

the degree of variation in the shape of the posterior

elements cannot be assessed on a normal radiograph as

the alignment of the transverse processes is not given.

Admittedly the extra deviation of the spinous process

relative to the rotated vertebral body can be seen, but

this produces only indirect evidence. Axial tomography

(Gargano, Jacobson and Rosomoff 1974) could demon-

strate the true shape of the apical vertebrae but the

equipment for it is not available for our use at the present

time. Even if it were, we doubt whether the resulting

information would justify the large amount of extra

irradiation (equivalent to that of a myelogram) in a

growing girl who is already undergoing repeated radio-

graphic examinations.

A study of the modification of the rib cage deformity

during treatment has produced an interesting, if unex-

pected, result. Ever since the introduction of the Mu-

waukee brace, opinions have differed widely as to its

effect on the rib hump. We are now able to state that

in most cases the true rib hump on the convex side of a

thoracic curve is little influenced, whereas the contra-

lateral rib depression can be markedly reduced, and in


VOL. 58-B, No. 1, FEBRUARY 1976

the most responsive cases even completely corrected.

Thus for patients where the rib depression contributes a

disproportionate part of the total deformity, one may

maintain a rather more optimistic view as to the ultimate

clinical appearance than can be held when the hump is

the more obvious feature. As this elevation of ribs may

occur without alteration of either curvature or rotation

and is on the side of the chest remote from any pressure

pad, it is difficult to picture what forces produce the

improvement. When treating a child in a Milwaukee

brace we place great emphasis on exercises both in and

out of the brace. One exercise which we regard as most

important is arching of the back in the brace, with the

breath held in deep inspiration. It is interesting to

speculate that this manoeuvre, originally intended to

control associated thoracic lordosis, might be forcing the

ribs on the concave side backwards and producing an

unexpected beneficial effect.

Because few cases are as yet fully documented, we

have been unable to prove conclusively that Harrington

instrumentation always produces the same favourable

effect on the rib cage deformity, but as in brace treatment,

we feel that the clinical appearance may be much im-

proved in patients with relatively severe rib depression.

It should be noted that our patients had a standard form

of instrumentation without osteotomy of the transverse

processes or special after-care aimed at reduction of the

rib hump. Perhaps such additional procedures merit

further consideration when the rib hump provides the

greater element of deformity.

REFERENCES

Cobb, J. R. (1948) Outline for the study of scoliosis. American Academy of Orthopaedic Surgeons. In Instructional Course Lectures,5, 261-275.

Cotrel, Y., and Morel, B. (1964) La technique de l’E.D.F. dans la correction des scolioses. Revue de Chirurgie orthop#{233}dique et r#{233}paratricede I’Appareil Moteur (Paris), 50, 59-75.

Enneking, W. F., and Harrington, P. (1969) Pathological changes in scoliosis. Journal of Bone and Joint Surgery, 51-A, 165-184.

Gargano, F. P., Jacobson, R., and Rosomoff, H. (1974) Transverse axial tomography of the spine. Neuroradiology, 6, 254-258.Goldstein, L. A. (1966) Surgical management of scoliosis. Journal ofBone and Joint Surgery, 48-A, 167-196.Goldstein, L. A. (1973) The surgical treatment of idiopathic scoliosis. Clinical Orthopaedics and Related Research, 93, 13 1-157.

James, J. L P. (1967) Scoliosis. Edinburgh and London : E. & S. Livingstone Ltd.

Laveriniccoca, A. (1913) Sulla scoliosometria-binostereoscopia-pantoscoliosografo-plasmascoliosometro. Atti di oriopedia e Iraurnato-logia, 8, 490-521.

Mehta, M. H. (1972) The rib-vertebra angle in the early diagnosis between resolving and progressive infantile scoliosis. Journal of Boneand Joint Surgery, 54-B, 230-243.

Mehta, M. H. (1973) Radiographic estimation of vertebral rotation in scoliosis. Journal ofBone andJoint Surgery, 55-B, 5 13-520Nash, C. L., Jun., and Moe, J. H. (1969) A study of vertebral rotation. Journal ofBone and Joint Surgery, 51-A, 223-229.

Rippstein, J� (1967) Deux nouveaux instruments pour l’examen clinique des scolioses. Acta orthopaedica Belgica, 33, 595-597.

Roaf, R. (1960) Vertebral growth and its mechanical control. Journal ofBone andJoint Surgery, 42-B, 40-59.

the rib hump inidiopathic scoliosis...64 the journal of bone and joint surgery the rib hump...

Documents