Arthritis, periarthritis, and bursitis...
ARTHRITIS, PERIARTHRITIS, AND BURSITIS OF THE SHOULDER JOINT
IN papers which will appear in future on the class of cases which I have made the chief subject of this book, the main criticism of my point of view may be that I have belittled "rheumatism" in the shoulder joint, and that the cases which I have called ruptures of the tendon, are merely the results of rheumatic joint disease, called "arthritis" nowadays. The findings which Dr. Akerson has published, and which can of course be duplicated wherever a similar class of material is found, may be used as demonstrating arthritis. There will be endless discussion as to whether disease in the tendon is the primary condition, or whether trauma, either chronic or acute, is the cause of the defects. I may say that years ago, when I first began opening these bursas, I ignored many ruptures of the supraspinatus because I thought they were merely "rheumatic joints." As I grew more familiar with the appearance of the ruptures seen at operation at varying periods of time from weeks to months after the accident, I began to realize that such old lesions as depicted in Plate VII, which are found at autopsy, were merely the results of arthritis in consequence of a rupture, or a sequence of trauma-arthritis-trauma-arthritis, and so on. Thus my present belief is that trauma and arthritis play alternating parts, but that trauma is the chief cause of the damage.
The change in my views as time has gone on, may be illustrated by a description of an examination which I recently made at autopsy through the kindness of Dr. Wolbach, Pathologist at the Peter Bent Brigham Hospital. The subject was a man of 83, who entered the hospital moribund and could give no history. He was afflicted with an extreme grade of general arthritis, so that there was distortion of most of the joints in his body. I was permitted to examine the shoulder joints. The findings on both shoulders were essentially the same. I will describe the condition of one. On making the usual incision into the bursa, a little fluid escaped and the bursa itself was found to be extremely large, extending beyond the normal limits in every direction, so that the whole head of the bone, tuberosities and all, well down on the neck externally, formed the base of this large sac. There were no synovial folds between the three different portions, and the finger could be easily passed beneath the acromion as far as the glenoid edge, and far under the coracoid process as well. The posterior limits extended well beyond the insertions of the infraspinatus and teres minor. The attachments of the supraspinatus and part of the infraspinatus had parted from the tuberosity, and an eroded joint surface was exposed in the gap thus made. Much of the greater tuberosity was rounded off. The amount of eroded cartilage corresponded to the portions which could be exposed through the gap. The tendons of the subscapularis and the teres minor were normal in appearance, but rather thin and feeble. The edges of the gap were falciform. The sulcus and eminence had disappeared, that is, were rounded off and covered by a thin, cartilagelike layer of fibrous tissue. The left shoulder joint showed a similar appearance, except that the degree of evulsion of the tendons was not quite as great. The appearance in both shoulders was very much like that in Fig. 8, Plate VII.
Years ago I would have simply said this is a case of arthritis, and therefore of no interest so far as the pathology of the bursa is concerned, but I feel quite the other way at present. My interpretation of the condition would now be that the patient suffered from some toxic disturbance which rendered his joint tissues and the adjacent tendons sub-normal, and even partially necrotic, so that slight traumata which would not affect a normal tendon, easily tore their fibers. Perhaps at first the change was merely degeneration of the structure of the tendon. Presently some slight extra effort, as in putting on a coat, may have torn a few of the fibers, creating some local soreness and disability, and starting up a mild inflammatory process in the adj acent membrane. The patient would have refrained from using the arm for a time, and still further atrophy and degeneration of the tissues would occur. Then another minor accident may have pulled out a few more fibers and more lameness have been experienced. Presently enough fibers may have been torn to allow the fluid of the joint to pass through into the bursa. Continued use of the arm would have produced a mild synovitis with an increase of fluid, and the fluid itself would have lubricated the tissues sufficiently so that friction under the acromion would have been allayed. However, each time the patient raised his arm the fluid would have been forced more into the bursa, so that eventually the bursa would dilate in all directions. As this process went on, more fibers of the tendon would be torn out and the weaker it became, the more o'f a vicious circle would, be established, until eventually it arrived at the condition in which I found it at my examination. Meantime, the portion of the tuberosity to which the tendons were formerly attached would have become absorbed and rounded off as explained on page 92.
PLATE VII. ASPECTS OF BASE OF BURSA IN OLD LESIONS OF THE SUPRASPINATUS
These drawings (reduced nearly one-half) were made by Dr. Akerson from specimens which he had removed at autopsy. They represent old, chronic lesions, the results of accidents or disease suffered many years before death. The superior aspect of each specimen is shown; i.e., one is looking down on the base of the bursa and sees the defects in it caused by destruction of the tendon of the supraspinatus. The lower margin is anterior in all cases.
The plate, as a whole, gives a very fair idea of the general aspect of specimens obtained in this way. At first sight one would not recognize that they are all instances of the same lesion. Unless one examined them in detail one would say they were instances of chronic arthritis. Without the knowledge obtained from operation on living cases, the writer himself would not have been able to interpret these specimens as the results of ruptures, but with increasing experience he feels quite confident that each lesion should be so interpreted. Owing to the large number of Dr. Akerson's specimens from which we could choose, we have been able to use this plate to illustrate two quite different and antagonistic points. By arranging them in sequence, the first six figures show a steadily increasing degree in the extent of the defect, although each pair came from a single cadaver. Second, by choosing pairs from the same patient the series also illustrates the frequency of the finding of a bilateral condition, which is a strong argument for considering that the original process may be one of disease or perhaps of overuse. The casual pathologist on finding such conditions- would hardly suspect injury as the basis of the lesion, yet this conclusion must be reached after earnest study, even if the rupture may have occurred in degenerated tissue.
No. 1 illustrates the condition of the base of the bursa in a case where the rupture is partial and has made only a small opening between the joint and the bursal cavity, although a chronic bursitis has been set up.
In No. 2 the destruction of the tendon beneath the base of the bursa was more extensive, so that the degree of bursitis is greater.
In No. 8 we see the condition I have called " Straps," where many fibers of the tendon have torn away irregularly. The joint surface is seen in the openings between the straps.
In No. 4 this condition is more advanced so that one sees the eminences on the bare tuberosity from which some of the fibers have been evulsed.
No. 5. The cartilaginous surface is exposed through the more or less triangular rent, of which the bare tuberosity and joint cartilage form the base.
No. 6. The condition is even more pronounced. The upper half of the gap shows cartilage, the border of which is irregular. In the lower half of the gap is seen the bare tuberosities with their eminences of new formed bone.
No. 7 and No. 8 show the extremes of the series. The base of the bursa, with chronically inflamed bands, is present on the left, while the right shows the effect of a severe injury which tore away the supra- and infraspinatus and later, erosion of the cartilage over the area of the articular surface, which was exposed in the gap, had occurred.
Nos. 5 and 6 are typical "complete" ruptures, while the first four and also No. 7 were "incomplete," although small communications existed between joint and bursa in each case. No. 8 represents such a case as that described on p. 109.
It is quite possible also, that instead of this process occurring by gradual changes, there may have been at any time a fall downstairs or other injury which would have torn the fibers of the tendon to a considerable extent, and from that time on, he may never have had proper use of his arm. Perhaps he had two separate injuries—each to one arm—or perhaps, on the other hand, the process may have been so slow and insidious and accompanied by so much greater disturbance in other joints, that no history could have been obtained of a shoulder injury, even if the patient had been able to give one.
Since this patient presented an extreme arthritic condition in most of his other joints, one might say that the condition in the shoulders was simply a part of the whole process. Undoubtedly, this is to a certain extent true, but my contention is that without traumata, even though trivial, the partition formed by the musculotendinous cuff, between the bursa and the j oint, would have remained intact. This contention is amply justified by the fact that in many of my operated cases such lesions were unilateral. In a few early cases, I have found a good, thick, normal appearing tendon with a transverse rupture and so little separation, that I could readily and successfully suture the tear.
This was an extreme case of chronic generalized arthritis, and for the sake of argument I think it is fair to postulate that most of the operated cases had some degree of "arthritic" degeneration in their musculo-tendinous cuff's before their accidents. Yet the fact remains that I sutured their ruptured tendons and obtained good function in some cases. Even in this old man, there may have been a time, years ago, when prompt suture of his tendons, combined with attention to his foci of absorption, might have saved him the use of his arms.
Although cases of arthritis which involve one shoulder joint and do not involve other joints, must be very rare, they do occasionally occur. With the exception of one case of a probable gonorrhoea! infection, and several cases of tuberculosis, the following case is the only one which I have confirmed by an exploratory operation.
Case No. 7. An unmarried woman of 27, who was born in New Brunswick, but who had done housework in the United States for some years, entered the Massachusetts General Hospital in December, 1909. She stated that she had always been well except for scarlet fever as a child, and pneumonia five years before entrance. Ever since the pneumonia she had frequently had sharp pain in her shoulder at night, but not during the day.
In January, 1909, the pain was especially severe one night and the next morning she found she was unable to raise her right arm above her head, and a "jumping pain" started in the arm and shoulder. At the suggestion of her doctor she carried the arm in a sling for three months, but without relief. In September, the arm was immobilized in plaster, but this did not relieve her pain.
The patient was thin and nervous. There was atrophy of the muscles about the whole shoulder, especially of the deltoid. There appeared to be no motion at all in the scapulo-humeral joint. X-ray showed that the true joint was involved and the cartilaginous area greatly damaged. A diagnosis of tuberculosis was made.
At operation the bursa was found to be adherent throughout, except in the portion beneath the acromion. The tendons of the short rotators were not thickened but were atrophied. There was no oedema or tuberculous granulation tissue seen at any time during the operation. The tendon of the subscapularis was divided and the true joint opened. It was found that only a central island of normal articular surface was left on the head of the humerus. There was no pus or fluid in the joint. The inferior surfaces of the tendons of the short rotators were adherent to the bone beneath them over a circular area about one-half inch wide, which took the place of the normal periphery of the articular surface. With a finger in the joint the amount of remaining articular surface was demonstrated to be nearly exactly that shown in the X-ray. After these manipulations, external rotation and abduction were found to be about two-thirds normal. Abduction was limited by contraction of the teres muscles and it was thought best not to risk rupture. A piece of the synovial membrane just at its reflection to the subscapularis tendon was excised for section. Gross observation showed no evidence of tuberculosis nor did the microscopic report by Dr. Whitney. The tenotomy of the subscapularis tendon was not repaired. Muscle loosely closed with catgut and skin with s. w. g. The wound healed by first intention.
A year later the condition of the shoulder was found to be the same as before operation. There had been no improvement at all. I was unable to trace the patient afterward.
My opinion at the end of the operation was that the patient had caries sicca, but the pathologic report did not confirm this and merely showed chronic inflammation. The fact that the symptoms began soon after a pneumonia suggests that there may have been a pneumococcus infection. Since the wound healed by first intention it is very unlikely that the infection was due to any of the ordinary pyogenic organisms. The striking differences of the symptoms in this patient from those of the ordinary case of "frozen shoulder" were absolute immobility of the joint, and a more obtuse angle formed by the axis of the shaft of the humerus and that of the scapula.
What is arteritis in any joint? Inflammation from any cause, whether primarily infection from pathogenic bacteria or from mechanical derangements due to changes which are congenital or traumatic.
Most of the phenomena of idiopathic arthritis are best illustrated in the hinge jqints, the cartilaginous surfaces of which remain in contact at some point and in which the lateral movement is prevented by tense ligaments. Static weight-bearing joints also are readily deformed by the results of inflammation, so that permanent damage with distortion results.
Pathologically, acute arthritis is characterized by swelling of the membrane, increase of joint fluid, oedema and congestion of the ligaments and periarticular structures, and spasm of the neighboring muscles. Chronic arthritis results in areas of absorption of the joint cartilages, lipping of the edges of the bone adjacent to the edge of the joint cartilage and shrinking or thickening of the ligaments, and often redundant fibrous changes in the synovial membrane, which may encroach on the cartilage.
You will find it difficult to demonstrate these changes in an un-traumatized shoulder joint, and the reason is that the shoulder differs from all the other joints in its mechanics, especially in two respects. Its surfaces are not held in contact by ligaments and its ligaments are never tense, except in the extremes of motion. It is not a static weight-bearing joint, and only at intervals is it even a power-bearing joint. Therefore, it is to be expected that the phenomena of arthritis will not be pronounced, even when the joint has been infected or traumatized. Since it does not have to bear weight, it can be relieved of duty on the slightest onset of soreness. The hip, the knee, the ankle, when inflamed, must bear a burden even during the momentary pause as the patient limps. Pressure is brought to bear on the affected cartilage, and the sway of the body tenses the tight ligaments and induces fibrous reparative changes.
When the shoulder joint is involved and other joints are not, it may be taken for granted that the trouble is not due to arthritis of constitutional origin.
The shoulder joint does become involved in some of the general infections, as in cases of acute and gonorrheal rheumatism and in some of the exaggerated cases of general chronic arthritis of septic or toxic origin, but it is noticeable that the shoulder usually clears up before the weight-bearing or hinge joints do. It is seldom left permanently damaged. Isolated arthritis in the shoulder joint is most unusual. How common it is in the knee or the- great toe joint, or even in the hip, compared to its incidence in the shoulder! Dealing With shoulders as much as I have for thirty years, I can recall but one case of a permanently damaged shoulder from a local non-tuberculous arthritis.
Consider by itself in relation to this joint each of the individual changes characteristic of arthritis.
Fluid. Did you ever recognize an excess of fluid in the shoulder joint? You have many times in the knee or in the ankle. I have rarely seen it, although I have been on the lookout for it. When it occurs you may recognize it in two ways. First, by the position of the humerus in relation to the scapula. The arm and scapula are fixed at an obtuse angle, readily demonstrated by comparing the axis of the spine of the scapula with that of the humerus. The tense capsule forces the bones into this position. This sign appears only in very severe cases when the fluid is under tension. Careful palpation beneath the posterior edge of the acromion or in the axilla is the other way, for a rounded swelling can be felt and the sensation of fluctuation obtained. A third way should be by the X-ray, for the joint surfaces would be forced apart, if the amount of fluid is great. One may also, with a sudden push upward of the humerus, knock the joint surfaces together. When joint and bursa communicate through a rupture in the capsule, you may find what I call the fluid sign. (See p. 155.) Fluid in the shoulder joint never causes anterior swelling unless the bursa is also involved.
Why is distention of the capsule of this joint so rare? In the first place, because there is room for so much fluid, that long before a large amount forms, the irritation commands the j oint to rest. We do not limp around on a lame shoulder as we do on a knee. In the second place, there are no normal protruding tags of membrane in the shoulder joint, as about the alar ligaments in the knee, to get caught, bruised or twisted and thus to start synovitis.
If congestion of the tissues should start in the shoulder, it has plenty of room, and the spasm initiated in the short rotators would protect it until the inflammation subsides. Thus the shoulder joint, in cases of general arthritis, recovers more readily than do the tense joints held by firm ligaments.
Villous conditions are likewise unlikely, because normally the whole synovial lining is smooth, and no clefts have to be filled as in the knee during normal flexion and extension. Hence we rarely find fibrosis of inflamed villi and "lipoma arborescens" to impinge in the shallow surface of the glenoid.
Serious oedema and thickening of the ligaments are scarcely possible, since there are no thick ligaments and but a thin capsule.
Absorption of the cartilage does take place in some old cases of severe rupture of the supraspinatus, but this is usually a local area where pressure is exerted on the joint surface by the under side of the acromion, as the head of the bone gains a fulcrum when the arm is abducted. This cannot occur unless the supraspinatus tendon has given way, for normally it protects the cartilage from this contact. It is very significant that Akerson found joint erosion localized by and proportionate to the extent of the ruptures in the tendon. I doubt whether cartilage erosion occurs in the shoulder joint unless the supraspinatus has been damaged or when suppurative inflammation, which is rare in this joint, has occurred.
I have not found cases where a "pannus," as described by Nichols in the other joints, has invaded the cartilage from the joint edge and gradually reduced its extent. The loss of cartilage in the gap caused by retraction of the tendon is symmetrical with the exposed areas and is quite superficial.
Deformity of the articular surface does occur in the shoulder, but only after very destructive traumatic lesions or suppurative infection, tuberculosis, fracture, or operation. Most deformities of joints are due to combinations of disease (softened bone and tissue), with weight bearing or power bearing in sufficient degree to cause one side of the joint to yield. The shoulder joint is relatively free from this combination. With the slightest symptom of inflammation, all weight or power bearing is instinctively stopped. As a whole, arthritic deformities in the shoulder are rare, and most of the old museum specimens which show deformities are probably the late results of fracture and dislocation. I have never seen a monarticular arthritis of this joint sufficiently severe to cause deformity. Even in the generalized case detailed above, the heads of the bones were mere rounded knobs, not much misshapen and with no hypertrophic excrescences.
Lipping of the articular margins is the most characteristic appearance in cases of hypertrophic arthritis in all joints. It occurs exactly at the border of the articular cartilage. In most hinge joints the lateral ligaments are inserted just at this point, and hypertrophic spurs or rims may go up into these ligaments, or even into the attachments of tendons. In the shoulder joint, since there are no lateral ligaments, there is very little tendency for lipping, so far as the articular surface of the humerus is concerned. In the glenoid, there may be some lipping due to bone formation in the fibrous rim of the shallow j oint surface. The finding by X-ray of such instances of lipping of the glenoid is usually accidental, and generally accompanied by no symptoms. I doubt if it is significant at all, so far as arthritis is concerned. Examination of many desiccated shoulder blades shows that the glenoid surface is quite variable, and such instances as there are of unusually misshapen conditions of the edge, are probably due to relaxation of the joint from conditions such as paralysis of the adjoining muscles, recurrent dislocation, or looseness of the capsule due to rupture of some of the short rotators. These conditions would permit the head of the bone to ride over unduly on the edge of the glenoid and cause bone proliferation there. On the humeral head in occasional desiccated bones, one sees a very slightly raised rim on the edge of the articular cartilage. This appears to be the only evidence of arthritis that one could determine by X-ray. Such appearances are frequently found in elderly people who have never had a symptom. I wish to call especial attention to Diagram C,- which shows why hypertrophic appearances on the tip of the tuberosity, or external to it, are to be attributed to lesions of the bursa and not to arthritis of the joint. They are not in the joint.
Synostoses. Bony ankylosis of the shoulder joint rarely occurs except in children, following tuberculosis and osteomyelitis. Both of these conditions are very rare in adults. In many of the other joints the severe grades oi arthritis end in fixation, but the shoulder joint is of such a structure that bony ankylosis can only be produced with difficulty even when it is the intention of the surgeon to procure it. The amount of surface for bony contact is so small in proportion to the degree of leverage exerted by the arm, that it is a real achievement to get a synostosis. The tendency of the joint, even in severe infections, is to slowly form fibrous adhesions which permit only a small arc of motion.
We have now reviewed the individual changes characteristic of arthritis in other joints, and made it clear that in the shoulder these pathologic conditions either do not exist or have a tendency to remain at the minimum. We cannot say that there is no such thing as non-traumatic arthritis in the shoulder joint, because inflammation of the joint can and does occur, but we can see that owing to the structure of the joint it does not readily occur and, if it does, the lesions are at a minimum compared to those which occur in the other joints. In fact, we might say with much truth that there is no such condition as a chronic arthritis per se of the shoulder, for, owing to the structure of the j oint, it would always be a periarthritis; that is, inflammation can only show itself in the tissue which is the important one in the joint; namely, the fused musculo-tendinous cuff which forms the major part of this articulation. I contend that the defects in this cuff so frequently found at autopsy were originally largely traumatic, although unhealthy tendon may have suffered the trauma.
Periarthritis. The structures besides the musculo-tendinous cuff which might be involved in this condition are the prolongation of the synovial lining of the joint down into the bicipital groove, the bursas under the infraspinatus and subscapularis, the subacromial bursa, and the glenoid half of the capsule. It would be a very severe periarthritis which would affect to any considerable extent the structures external to these. There is occasionally some secondary atrophy of the muscles further from the joint (e.g., the deltoid, the pectorals, and the teres major), due to prolonged spasm or inaction. The reader may refer to Chapter I, p. 14, which discusses the outer and inner set of muscles in the shoulder. It is the inner set which is involved regularly, but the outer set which operates the scapula is as a rule not atrophied.
In periarthritis, what is the exact condition which prevents the patient from performing active motion, or the surgeon from performing passive motion? Such a case might have all the following changes:
1. Adhesions between the roof and floor of the subacromial bursa involving (a) the subacromial portion; (fo) the subdeltoid portion; (c) the subcoracoid portion.
2. Adhesions between the bicipital groove and the tendon of the long head of the biceps.
3. Necrotic changes and inflammatory stiffening in the musculotendinous cuff.
4. Chronic inflammation in the synovial membrane of the joint and of its capsule.
5. Adhesions in the extensions (bursa subscapularis and bursa infraspinati) of the joint underlying the infraspinatus and subscapularis.
Different cases may have varying proportions of pathology in these structures, extreme grades in any one of which would suffice to limit the motion of the joint; e.g., firm adhesions at any portion of the subacromial bursa might be the major factor, or joint motion might not be complete owing to adhesions between the intact biceps tendon and the sheath in the groove, even if no other parts of the periarticular structures were involved. Likewise, the bursa might be free; the sheath of the biceps might be free, but restriction of motion might be maintained by the inflammation in the musculotendinous cuff.
I am inclined to believe that all these factors are present to some degree in the worst cases, but I am very sure that in many cases the adhesions in the bursa are most significant and important from the point of view of treatment. However, I believe that the adhesions in the bursa are very often primarily due to the peculiar form of inflammation characteristic of the tendons. This subject of the tendinous involvement is more fully discussed in Chapter VII.
When I first began my work on "Stiff and Painful Shoulders," the usual diagnosis assigned to such cases was "periarthritis." Following my articles drawing attention to the anatomic characteristics of the bursa and the importance of its recognition, the term "subdeltoid bursitis" replaced that of "periarthritis," and since my second paper, the adjective "subacromial" has largely replaced that of "subdeltoid." The present status is, that the bursitis has been accentuated rather than the tendinitis. I now feel that in most cases the bursa, like the peritoneum, is only secondarily involved, and that the commonest causes are: (1) in traumatic cases, a rupture of the fibers of the supraspinatus tendon; and (2) in spontaneous cases a necrosis in this tendon, and even in these cases possibly an initial severe trauma to the tendon. My present view might be expressed as follows: The starting point of most lesions of the shoulder centers in the tendon of the supraspinatus. Thence it involves the bursa and the adjoining tendons of the other short rotators, but the inflammation of the bursa gives the most pronounced and often the only painful symptoms.
Thus, from arthritis, we have passed through periarthritis to bursitis.
Bursitis. Formerly I described bursitis as consisting of three types, the acute, the chronic adherent and the chronic non-adherent. These are still useful clinical divisions. I now feel that an attempt to describe bursitis as an entity would be equivalent to attempting to describe peritonitis as an entity. I believe that the bursa, like the peritoneum, has a very great capacity for the formation of protective adhesions which later may disappear. It is not a structure where disease starts, so much as a structure which limits disease in the adjacent structures by temporary adhesions, causing fixation of the parts. Later its physiology is such that adhesions disappear and two shining, adjacent, frictionless surfaces are again formed to allow motion between the parts. After inflammation the single sac may be replaced by a number of smaller spaces. Lesions in the supraspinatus tendon are the common cause of bursitis, just as lesions in the appendix are the common cause of peritonitis.
Since the success of treatment depends in most conditions on our knowledge of the exact pathology, a surgeon treating shoulder lesions should try in each individual case to picture in his mind the relative proportions of the elements of the joint which may be involved. Personally, I believe that the sheath of the biceps tendon is less apt to be involved than are the other structures. I have never proved its involvement in a single case. I think that the substance of the tendon of the supraspinatus is the most often involved. In many cases there is a secondary bursitis, but in some there is none. I am quite sure that patients may have small ruptures, small calcified deposits, and small areas of necrosis in the tendon, without any bursitis being caused.
I am willing to admit that I cannot readily define just what signs give me the impression in individual cases that any particular structure is especially involved. It is a matter of touch which can only be learned with experience, and deceptive even after much experience. The difficulty of estimating the proportions in which each structure is pathologic is made still greater by the degree of spasm present. In acute cases of bursitis, such as those caused by perforation of a calcified deposit from the tendon into the bursa, spasm alone may hold the whole joint rigid, although on the previous day the full arc of motion could be performed. A month later the joint may be wholly free again, or perhaps even six months later the spasm may have become at a minimum, and yet the joint will be just as rigid from adhesions and stiffening, which are wholly mechanical and prevent relaxation even under ether. All degrees between these extremes exist, because stiffening from spasm fades into an adherent condition by imperceptible degrees.
I may summarize by saying that owing to the peculiar mechanics of the shoulder, the avascular and inert supraspinatus tendon is the most vulnerable part of the joint, and that inflammation in it is apt to be painless until the adjacent subacromial bursa is involved, which, being abundantly supplied with vessels and nerves, produces the symptoms of which the individual complains.
The explanation of the relative weakness of the supraspinatus is in my opinion to be given on evolutionary grounds by the change from the horizontal to the vertical animal. This led to obliging the supraspinatus to lift the arm instead of swinging the foreleg. As suggested on page 7 the bursa is a relatively newly developed structure. However, we must not lose sight of the fact that the tendons of the other short rotators are often involved in all these conditions. Furthermore, although the capsule of the joint is only tense in extreme positions, the anatomic position in which we usually carry our arms as we walk or sit, does put some strain on the upper portion represented by the insertion of the supraspinatus. Lockhart has pointed out that in the cadaver this portion of the capsule will hold the head of the humerus up, even if all the remaining capsule is removed. He uses this fact as an argument against the accepted theory that air pressure helps to maintain the joint in position. While I do not agree that this is true, I do believe that in the upright position there is always some tension on the tendon of the supraspinatus, for this is almost an extreme position.
DEFINITIONS AND CLASSIFICATION OF THE VARIOUS FORMS - OF SUBACROMIAL BURSITIS
An acute subacromial bursitis is one in which there is sharply localized tenderness and protective spasm of sudden onset. Acute bursitis rarely arises from any cause except one of the following:
(1) A direct bruise on top of the anterior part of the shoulder when the arm is in dorsal flexion. This type is not proved by actual observation on the operating table. It is merely supported by the clinical fact that such bruises are followed by acute local tenderness over the bursa, accompanied by scapulo-humeral spasm. Such cases promptly recover and do not have a chronic phase. They are rare.
(2) Prolonged (an hour or two) hyperabduction. The best instance is bursitis following hyperabduction of the arm in operations, as for cancer of the breast. Instances of this form are not common, and there is no absolute proof that the strain of this position does not primarily affect some other part of the joint. Such cases may become subacute but not chronic (lasting over six months).
(3) Inflammation extending from about a calcified deposit which has either burst into the bursa or has approached the bursal surface closely enough to inflame its lining. This is by far the most common cause, and is abundantly proved by the findings at operations. These cases may be acute only, or last for months, but they rarely give symptoms for over two years, even in the worst cases. Many recover within a few weeks after perforation.
(4) A rupture of the supraspinatus tendon by trauma of sufficient degree to make a direct opening between the joint and the bursa through the gap in the tendon. This must lead to bleeding into the bursa and more or less distention with joint fluid. I have never yet opened a bursa immediately after such a trauma, but more or less acute bursitis must necessarily occur in such a case, for the floor of the bursa is involved. The acute phase is brief. Complete ruptures always produce a permanent chronic bursitis because the bursa is continually irritated by friction on the irregular base.
(5) Minor ruptures may occasionally produce an acute bursitis, but they may be almost symptomless and usually are subacute in character, and sometimes become truly chronic. In my opinion the great majority of sore shoulders come in this class as subacute cases.
Any other conditions which suddenly produce an acute bursitis must be very rare. I have seen only one suppurating acute case from pathogenic bacteria. Even this case was doubtful and was probably to be explained by contamination of the culture.
Subacute or chronic subacromial bursitis may be divided into two classes:
(1) Those in which there is no restriction of motion, but in which there is a painful point causing a wince or jog in the motion of elevation.
(2) Those in which there is limitation of motion from adhesions, or contractures, although spasm may still be present.
The two classes are not perfectly distinct, but serve in a general way in the study of the subject.
The cases of the non-adherent type are caused by (a) inflamed villi, folds or bands. These cases do not often have an acute phase, (b) Irregularities of the base due to calcified deposits in the tendons, about which there is only a little chronic inflammation. These cases may have already passed through an acute phase or may at any time exhibit one. (c) Defects in the base of the bursa due to a partial or complete rupture of the supraspinatus tendon. The acute phase is brief and occurs only immediately after an accident.
The adherent type of subacute and chronic cases seldom follows types one and two of the acute cases. It is a common sequence of the third type; is very infrequent in the fourth type; not uncommon in the fifth type, especially in those cases where the trauma is trivial and which will be discussed later under "Tendinitis."
I know of no references in the literature of arthritis which would be of any particular help in studies of the shoulder. If the reader wishes to inform himself as to the modern con.eption of what little is known about "arthritis" and "rheumatism" in general, he is referred to the number of the New England Journal of Medicine, for May 18, 1933, Vol. ccviii, no. 20, which prints several papers read at a recent symposium on the subject; with these he will find numerous references.