The Costovertebral Joints. Anatomical-Clinical Observations in Arthritis By H. NATHAN, H. WEINBERG, G. C . ROBINAND I. AVIAD Costovertebral arthritis was found to be a frequent condition in a study of 346 skeletons and 100 roentgenograms. It followed a definite pattern of distribution, which seems to be related to the anatomical characteristics of the joints and the ribs involved at the different thoracic levels. No differences were found between the races and sexes. In a few patients in which the diagnosis of costovertebral arthritis was established, local treatment was tried with good results. In un studio de 346 skeletos e 100 roentgenogrammas, il esseva trovate que arthritis costo-vertebral es un frequente condition. 1110 sequeva un definite configuration in su distribution, lo que pare esser relationate con le characteristicas anatomic del articulationes e del costas afficite a1 varie nivellos thoracic. Nulle differentias esseva trovate inter le racias e le sexos. In un micre numero de patientes in qui le diagnose de arthritis costo-vertebral esseva establite, tractamento local esseva essayate con bon resultatos. I N THE COURSE of a study of the changes in the vertebral column due to observations on arthritic reactions of the articular surfaces of the costovertebral joints were recorded. Contrary to previous these reactions were found to be fairly frequent. Moreover the distribution of the affected joints appeared to follow a characteristic pattern. The costovertebral joints have attracted little attention, and few references can be traced in the English literature of the past twenty five year^.^^^ Our findings may perhaps help in the comprehension of the etiology of some syndromes otherwise ill-understood, where pain or other disabilities are related to the vertebral column or the thoracic cage; and may in this way contribute to more adequate treatment of these cases. ANATOMYOF THE COSTOVERTEBRAL JOINTS Each costal arch articulates with the vertebral column by two synovial joints, one between the head of the rib and the vertebral bodies, and thc other between the tubercle of the rib and the transverse process. The costovertebral joints, between rib head and vertebral bodies, (the ones discussed in this work) differ in their structure at various thoracic levels (fig. 1). The second to tenth costovertebral joints are characterized Gy a rib head of wedge shape. Each head articulates with the hemi-facets of two adjacent vertebrae and with the intervertebral disc. Besides the joint capsule, a strong intra-articular ligament binds the head of the rib to the intervertebral disc and divides the joint into two separate synovial cavities. The first, eleventh and twelfth costovertebral joints differ from the others, normally, in that each rib articulates only with its corresponding vertebral body by a single full facet. The intervertebral disc plays no part in the formation of the joint and there is no intra-articular ligament. It can thus be under228 ARTHRITISAND RHEUMATISM, VQL.7, No,3 (JUNE), 1964 229 THE COSTOVERTEBRAL JOINTS \ \ / Fig. 1.-Diagrammatic representation of the two types of costovertebral joint. The tenth rib (ClO) in the figure articulates with the adjacent bodies of T9 and T10 by two hemi-facets, and also with the intervertebral disc. The intra-articular ligament divides the joint into two separate synovial cavities. Joints of this structure are found between T2 and T10. The eleventh rib (C11) articulates only with the single full facet of its corresponding vertebra. Such joints are also found in T1 and T12. stood that the first thoracic body has a full facet for the first rib head, and a hemi-facet for the second rib, while the tenth thoracic vertebra usually possesses one hemi-facet only, articulating with the lower part of the tenth rib head. (fig. 1). Variations do occur, however, where the tenth thoracic vertebra articulates with the tenth rib by a full facet, similar to those of the eleventh and twelfth vertebrae. MATERIALSAND METHODS Vertebral columns of 346 white and Negro subjects from the Todd Collection in the Department of Anatomy of the Western Reserve University, were studied in this present work. Their distribution according to age, sex and race is shown in table I. Table 1.-Distribution of 346 Vertebral Columns Examined According to Age, Sex and Race - Number of Vertebral Columns Examined White Age Group Male Female Male 0-19 20-29 30-39 40-49 50-59 60-69 70-79 80 and over 3 21 17 20 20 21 20 15 2 9 9 9 14 12 14 7 7 TOTALS 137 76 (YeSrS) _____ Negroes Female Total 12 10 8 10 8 7 5 7 15 11 10 6 8 5 4 19 57 47 47 50 49 67 66 346 46 31 230 NATHAN, WEINBERG, ROBIN AND AVIAD 2A 20 (See figure legend, facing page.) In this study, the articular facets of the costovertebral joints in each thoracic body of each of the vertebral columns were examined. Distinct lipping of the border, or erosion of the articular surfaces were considered to be indicative of arthritis (figs. 2 and 3 ) . The articular surfaces of the transverse processes and those of the corresponding rib heads were not examined, and no observations on them have been included in this study. Secondly a random series of 100 roentgenograms of the thoracic spine, taken for various indications were studied. Lastly a number of patients complaining of symptoms which could theoretically be related to costovertebral arthritis were examined and radiologicd examination, particularly of these joints was undertaken. 231 THE COSTOVERTEBRAL JOINTS 2c Fig. 2.-a) Articulated 11th and 12th thoracic vertebrae. The costovertebral articular facet of the 11th vertebra is smooth and its borders are clean (no arthritis) ; whereas the same facet of the 12th vertebra (arrow) is irregular, with distinct lipping of its borders (arthritis). b) Isolated 11th thoracic vertebra. Shows marked lipping in the posterior-inferior border of the costovertebral articular facet (arrow). Osteophytes are also present in the superior border of vertebral body. c) Isolated 12th thoracic vertebra. The surface of the costovertebral facet (arrow) is irregular; distinct lipping is present in the postero-inferior part of the border.. OBSERVATIONS AND DESCRIPTION In 168 of the 346 specimens examined (48 per cent) costovertebral arthritis was noted in one or more joints. The lipping was generally of a small degree (fig. 2 ) . In some cases, however, spurs appeared fairly large (fig. 3); and in a few of them the head of the rib was even found to be totally fused to the vertebra. Sometimes lipping of the costo-vertebral joints and osteophytes of the vertebral margins were present simultaneously in the same vertebrae (fig. 2 ( b ) ); but no correlation between both processes seemed to exist. The relationship between the presence of costo-vertebral arthritis and age is shown in table 2. As can be seen arthritis made its appearance in the third decade. In the following ten-year age group the frequency increased rapidly to 57 per cent of the spines examined. Only minor, non-significant fluctuations in frequency were noted in the succeeding decades. NO significant differences in age frequency were noted between Negro and white vertebral columns or between the sexes. The distribution of the affected joints in the various thoracic vertebral bodies follows a characteristic pattern (fig.4). The first, eleventh and twelfth bodies (i.e. those presenting full facets) appear to be the most affected. Among the remaining thoracic bodies, those usually having hemi-facet joints, a rising frequency of arthritis was found from the second to the 68th vertebrae and thereafter the frequency fell to the tenth thoracic body. 232 NATHAN, WEINBERG, ROBIN AND AVIAD 3A 35 Fig. 3.-a) Six thoracic vertebra. Shows a large spur (arrow) projecting downward from the superior margin of the inferior costovertebral articular hemi-facet. The superior hemi-facet is not affected. b ) Eighth thoracic vertebra; inferior view. Spurs (arrows) are seen on both sides, coming from the superior margins of the inferior costovertebral articular hemi-facets. The right one is bigger. No osteophytes are seen coming from the margin of the vertebral bodies. The curves of frequency were similar in all groups of skeletons, whether classified by race and sex. The small differences noted do not appear to be significant. The distribution of costovertebral arthritis appeared to show no marked difference when left and right sides were compared (fig. 5 ) but there was a noticeable difference when the superior and inferior hemi-facets of each vertebral body were compared (fig. 6 ) . It can be seen that the inferior hemi-facet of each vertebral body (i.e. those facets corresponding to the upper 233 THE COSTOVERTEBRAL JOINTS Table 2.-Frequency of Occurrence of Costouertebral Arthritis in the Various Age Groups Age Group (years) 0-19 20-29 30-39 40-49 50-59 60-69 70-79 80 and over No. of Spines Examined No. of Spines Showing Costo-vertebral Arthritis Percentage of Affected Spines 19 57 47 47 50 49 46 31 0 1 87 21 33 36 32 18 0 2 57 44 - 66 73 70 58 part of each costovertebral joint) were more frequently affected than the superior hemi-facets. These differences were more obvious at the peak of the curve of frequencies, that is in the sixth to eighth vertebral bodies. The radiological study of a random group of thoracic spine roentgenograms showed a surprisingly high percentage of costovertebral arthritis, especially in the first costovertebral joints. Out of the 100 cases, obvious arthritic changes were present in 17. It is of interest to note that until this retrospective study was performed, no notice had been taken by the examining radiologists of these appearances, and in no case had the presence of arthritic changes been reported. From the small number of patients studied clinically, details are given of three. case 1 Mrs. T. M. aged 55 complained of pain radiating from the interscapular region to the right shoulder for several years. The pain was sometimes aggravated by coughing. On examination there was focal tenderness just to the right side of the vertebra prominens, some limitation of movements of the cervical spine and of the right shoulder, especially of internal rotation. There were no abnormal neurological signs. She had previously been treated by physiotherapy, short wave diathermy, and cervical traction with no effect. X-ray (fig. 7 ) showed cervical spondylosis, bilateral cervical ribs and severe osteoarthritic changes in the first thoracic costovertebral joint on the right side. She was treated by two injections of Hydrocortone into the affected costovertebral joint with complete relief. Case 2 A maIe aged 60 years complained for three years of increasingly severe radiating pain in the left upper limb, with interscapular pain, especially after effort. Apart from local tenderness to the left of the mid-line of the first thoracic spine and some slight limitation of movement of the left shoulder no abnormalities were noted on examination. Cardiac causes of pain were excluded after intensive cardiac investigation. The X-ray presence of costovertebral arthritis, particularly in the first costovertebral joint, (fig. 8) prompted a trial with intra-articular Hydrocortone, and after two injections he returned to his previous occupation as a carpet repairer, free of pain. Case 3 A 33-year old woman suffered from interscapular pain radiating along the right upper limb for two years, which had not responded to previous physiotherapy. Examination again revealed local tenderness near the spinous process of the first thoracic vertebra, limitation of internal rotation of the shoulder on the affected side, but no neurological abnormalities. Vascular studies, including capillaroscopy were negative. X-ray showed obvious NATHAN, WEINBERG, ROBIN AND AVIAD 22 18 a) White moles b ) White femoles c) N e q r o moles d ) Neqro fernole8 14 10 6 2 34 30 26 22 18 14 10 6 2 TI T2 T3 T4T5 T6 T7 T8T9 TIOTll l 1 2 22 18 e ) All groups Full f a c e t 1%1 Hemifocet 14 10 6 2 TI Tz T3 T4 T5 T6 T7 Ts T 9 T I O T I I TIZ Fig. 4.-Histograms showing the characteristic pattern of distribution of costovertebral arthritis in the different levels of the thoracic spine, according to sex and race: the higher frequencies of arthritis in the full facet joints (Tl, T11 and T12) as compared to those of the hemi-facet joints is evident; as is the regular curve of frequencies in the hemi-facet joints from T2 to T10. No significant differences are shown between Negro and whites or between males and females. arthritic changes in the first costovertebral joint on the right side (fig. 9 ) . The patient received treatment with intra-articular Hydrocortone, and her pain, which was especially noted on the ulnar border of the arm w a s relieved.' _______~ ~ ~~ 'A follow-up of the present cases will be presented in a separate paper together with additional clinical studies. THE COSTOVERTEBRAL JOINTS 235 % 12 Right 10 a Left Fig. 5.-Histogram. Distribution of costovertebral arthritis along the left and right sides of the thoracic vertebrae. No significant differences between the two sides are present. DISCUSSION AND CONCLUSIONS The most conspicuous feature of the anatomical part of this study is the definite pattern of distribution of costovertebral arthritis (Sgs. 4-6). The high frequency of arthritis in the first, eleventh and twelfth costovertebral joints, and the regular curve of frequency rising from the second to the sixth, seventh and eighth joints, and falling to the tenth were noted in each group of skeletons studied. An etiological interpretation of costovertebral arthritis can be inferred from this regular pattern of distribution. The higher frequency of arthritis noted particularly in the full facet joints of T1, T11 and T12, in which the intervertebral disc plays no part, suggests that this type of joint is more vulnerable to the mechanical irritation of constant rib motion, than is the type of joint seen in the remaining vertebrae, where the two hemi-facets are separated by the intervertebral disc and the intra-articular ligament. The presence of these soft structures seems therefore to play some role in reducing mechanical trauma in the costovertebral joints as it does in the intervertebral joints themselves. On the other hand the regular curve of distribution of costovertebral arthritis between the second and tenth thoracic vertebrae would appear to parallel the length of the ribs and the breadth of the thorax. The peak of the curve at T&T8 corresponds to the level of the longer ribs, whose movements are more extensive. This again would point to a simple mechanical 236 NATHAN, WEINBERG, ROBIN AND AVIAD O/e 16 14 Inferior 12 Superior 10 8 I a. 0 . e. '. *. '.. h -I, .' .* ...' .* Fig. 6.-Histogram. Distribution of costovertebral arthritis in the superior and inferior hemi-facets of the thoracic vertebrae. The inferior hemi-facets (those corresponding to the upper part of each costo-vertebral joint) appear to be more frequently affected that the superior hemi-facets of the vertebrae. The greatest differences are found in the region of the peak of highest frequencies of costovertebral arthritis (T6 to T8). factor being involved in the production of the arthritis of the costovertebral joints. The higher frequency of arthritis observed in the inferior hemi-facets of the vertebral bodies (that is the upper hemi-facet of the costovertebral joint) compared with the frequency found in the superior hemi-facets of the vertebral bodies would appear to be related to the general obliquity of the ribs, which point medially and upward against the vertebral bodies. Thus compressive forces transmitted through the ribs would tend to affect preferentially the upper facet of the joint, that is the inferior hemi-facet of the vertebral body. The fact that this inferior hemi-facet is usually smaller than the superior, and thus may bear a larger work-load from rib movement may also be of significance in the production of arthritis. Goldthwait3 suggests that abnormal obliquity and twisting of the ribs THE: COSTOVERTEBRAL JOINTS Fig. ‘I.-X-Ray of first costovertebral joints Case 1, 237 showing osteoarthritic changes on right side. may be factors in the production of changes in the costovertebral joints, but the high incidence noted in our anatomical study as well as the particular pattern of distribution of the costovertebral arthritis would suggest that such factors are of minor importance, and that mechanical irritation from constant rib motion is of greater importance in the production of the arthritis. From a clinical viewpoint, the syndrome to be expected, if in fact costovertebral arthritis does give rise to any disability, would be one of local pain, tenderness near the mid-line of the vertebral column, aggravated by respiration and coughing, and on chest compression. The close relationship of intercostal nerves, or in the case of the first costovertebral joint of the lower trunk of the brachal plexus and of the sympathetic chain, could account for pain radiating along the distribution of the peripheral nerves. Such a syndrome, commonIy encountered, has hitherto been ascribed to many factors, including spondylosis, disc prolapse, cervical ribs, scalenus anterior syndrome, etc. Only GoldthwaiP has drawn attention to the possibilities of costovertebral joint involvement as being of importance in pains referred to 238 NATHAN, WEINBERG, ROBIN Fig. 8.-X-Ray of first costovertebral joints Case 2, changes on left side. AND AVIAD showing osteoarthritic the chest or the abdomen. In an attempt at a clinical trial therefore, to see whether costovertebral arthritis does in fact give rise to any clinical disability, treatment was afforded to a select series of patients, by an accepted method against arthritis, i.e. by the intra-articular injection of Hydrocortone. For technical reasons, (ease of injection) only such cases in which the first costovertebral joint was involved were selected for this treatment. The success of the treatment, although encouraging is of course not conclusive. However, the success of mechanical methods of treatment reported by Goldthwait3 tends to support our findings. We feel, therefore, that the possibility of costovertebral arthritis being of clinical importance should be borne in mind. Further studies are of course necessary to attempt to correlate the types of symptoms, the physical and radiological findings and the presence of costo-vertebral arthritis. SUMMARY Costo-vertebTa1 arthritis was found in 168 (48 per cent of 346 skeletons and in 17 out of 100 random roentgenograms of the thoracic spine. Arthritis THE COSTOVERTEBRAL JOINTS 239 Fig. 9.-X-Ray of first costovertebral joints Case 3. Osteoarthritic changes are evident on the right side. was first tound in the 3rd decade and increased rapidly towards maximal incidence during the 4th decade. Thereafter, only minor non-significant fluctuations in frequency were observed. The distribution of the affected joints in the various thoracic levels followed a characteristic pattern: 1 ) The lst, 11th and 12th costovertebral joints (i.e., those formed by full facets) showed much higher frequencies of arthritis than the 2nd to 10th costovertebral joints (i.e,, those formed by 2 hemi-facets with the intervertebral disc between them). These observations suggest that the anatomical structure of the costovertebral joints is of some importance in the production of arthritis. Perhaps the intervertebral disc plays some role in reducing the mechanical trauma of the rib movements, as it does between the vertebral bodies. 2 ) The frequency of arthritis in costovertebral joints 2 through 10 followed a normal curve of distribution, the peak of the curve being about the 6th to 8th vertebral level. This distribution seems to be related to the length of the ribs, and to the extension of their movements. 240 NATHAN, WEINBERG, ROBIN AND AVIAD 3 ) The inferior costovertebral hemi-facets of the vertebrae (that is thc the superior hemi-facets of the costovertebral joints) were found to be more affected by arthritis than the superior hemi-facets. This may be related to the obliquity of the ribs that point medially and upwards against the vertebral bodies. No significant differences were found between Negroes and whites or between males and females. Three cases selected from a group of patients suffering from costovertebral arthritis are presented. They were all suffering for long periods from pain and other disabilities which were ascribed to other causes. They did not improve until the correct diagnosis of costovertebral arthritis was established, and an adequate treatment ( intra-articular injections of Hydrocortone ) was instituted. Heretofore, costovertebral arthritis received little attention. The present observations show that costovertebral arthritis is a frequent condition. And it should be more frequently borne in mind when a differential diagnosis of the many painful syndromes of the thorax or abdomen is done. REFERENCES 1. Abrams, N. R.: In Arthritis and Allied Conditions. Ed. 6, Ed. J. L. Hollander. Philadelphia, Lea 8~ Febiger, 1960, p. 824. 2 . Collins, D. H.: Post-graduate Medicine and Surgery: Rheumatic Diseases. Philadelphia, W. B. Saunders Go., 1952, cited by Abrams.1 3. Goldthwait, J. E.: The rib joints. New England J. Med. 223:56%573, 1940. 4. Jones, M. D.: Limitation of hypertrophic spur formation by the costo-vertebral articulations. Radiology. 75:584-588, 1960. 5. Nathan, H.: Spondylolysis. Its anatomy and mechanism of development, J. Bone & Joint Surg. 41A:303-320, 1959. 6. Nathan, H.: Osteophytes of the vertebral column. An anatomical study of their development according to age, race, and sex with considerations as to their etiology and significance. J. Bone & Joint Surg. 44A:243-268, 1962. Hilel Nathan, hl.D., Lecturer, Department of A n a t m y , The Hebrew University Hadassah Medical School, Jerusalem, Israel. Haim Weinberg, M.D., Chief Physician, Instructor Orthopedic Surgery, The Orthopedic Department, Rothschild Hadassah University Hospital, Jerusalem, Israel. Gordon C. Robin, M.D., Ch.D., F.R.S.P.S., F.R.S.C., Chief Physician, Instructor Orthopedic Surgery, The Orthopedic Department, Rothschild Hadassah University Hospital, Jerusalem. Israel. Ithanur Aoiad, hf.D., Chief Physician, Diagnostic X-Ray Department, Instructor in Roentgenology, Rothchild Hadassah University Hospital, Jerusalem, Israel.