dr keith holt - sports medicine€¦ · frozen shoulder dr keith holt frozen shoulder is a...

Keith Holt - Perth Orthopaedic and Sports Medicine Centre © - 2017

What happens in frozen shoulder?

The shoulder joint, (the joint between the humeral head and the glenoid - the ball at the top of the arm bone and the socket of the scapula or wing bone) like all joints, is surrounded by a capsule, a thin but relatively strong sack which holds the normal joint fluid within the joint. In addition to that function however this capsule is responsible for firstly, holding the joint together so that it does not dislocate and secondly, for restricting the amount of motion in the joint. If this capsule is somewhat lax, there is a large range of motion available, but the joint in turn may dislocate. On the other hand, if the capsule is tight, the range of motion will be restricted but, the joint is very much held together and cannot dislocate.

The exact underlying cause of frozen shoulder is unknown but, suffice it to say, what happens is that the capsule of the shoulder joint becomes inflamed. That is, a true gleno-humeral (shoulder joint) capsulitis develops. Sometimes this is thought to be as a result of trauma but, more usually, it comes on without injury, probably being caused by a virus (although this has not been proven). There are some factors that may support a viral causation however, and these include the fact that, once one shoulder joint has had this condition, it almost never recurs in that same joint again: suggesting a local immunity. On the other hand however, it can occur in the opposite shoulder joint, although the incidence of that is not that high.

Another factor that may lead one to believe that this is an infective process is that it is perhaps more common in diabetics than in the general population, and diabetics are much more prone to infection than other members of the general population. Despite this circumstantial evidence however, and despite extensive research in the area, the exact aetiology has not been identified.

The clinical picture

Initially, when the inflammation begins, the shoulder becomes sore. This is sometimes noticed as a gradually increasing ache but, more often, it is noticed as a distinct pain when the shoulder is stretched or pushed, such as when lifting the arm up or reaching out suddenly. A very common

story is one of sudden onset pain when reaching into the back seat of a car for a bag or similar.

With time the pain becomes gradually worse. It is relatively constant in nature, it is always worse at night time and the shoulder is often very difficult to lie on. Over its course, the condition passes through 4 stages.

Stage 1: Initially there is pain present but the range of motion is relatively normal. This is the stage of early inflammation.

Stage 2: Following the initial stage, the capsule, which is becoming increasingly inflamed, starts to become swollen and thickened and, as a result, it begins to tighten up. As it tightens up, the range of motion in the shoulder becomes gradually restricted and often, within 1-2 months of the commencement of the disorder, this restriction of motion becomes apparent. The movement most sensitive to this restriction, and therefore most easily noted, is that of putting the arm behind the back. The loss however, is in every direction.

Once the condition becomes established the pain can become quite intense, and it is always very constant. As the capsule becomes tighter, not only is the humeral head jammed into its socket, but it is also pushed upwards towards

Frozen Shoulder

Dr Keith Holt

Frozen shoulder is a relatively common condition involving the shoulder joint, which causes pain and loss of motion in that joint, often for a substantial period of time. In most cases, the onset is without injury, and the cause is unknown. It can however, stem from trauma and, in particular, it is often seen in association with fractures that involve the shoulder joint itself. The natural history is for resolution with time, albeit that complete recovery may take up to 2 years. Incomplete recovery is unusual, except in insulin dependent diabetics.

CapsuleRight Shoulder

Note capsule extension down the biceps tendonH

umer

us

Scapula


the acromion. Between the top of the humeral head and the acromion lie the rotator cuff tendons, and these can become increasingly squashed (impinged), particularly so when the arm is taken up past shoulder height. This induces pain from the rotator cuff, which is typically down the outside side of the arm, often going down as far as the elbow or hand. It is this pain that frequently leads people to the diagnosis of impingement (of the rotator cuff tendons) as a primary problem. In this case however, the impingement occurs as a secondary problem, the main problem being within the gleno-humeral joint itself. For this reason, a standard injection into the sub-acromial space (bursa) and around the rotator cuff tendons, does not fully settle the problem. To be effective, the cortisone needs to be in the actual joint itself where it can deal with the inflammation in the capsule.

Stage 3: If the inflammatory condition within the shoulder joint is left alone it will continue unabated, often for some months, until gradually, the inflammatory cycle settles down and the pain starts to go. By that time however, the swelling in the capsule around the joint has turned into scarring, and thus, despite the fact that the pain gradually eases off, the restriction in motion may remain.

Stage 4: With the subsequent passage of time this scarring gradually stretches up and, in better than 95% of cases, a full range of motion eventually returns. The time course for this condition to go through all four stages is somewhere between six months and two years depending on the severity. Overall however, the natural history of this problem is towards resolution and, the majority of people, regain either full motion of their shoulder or very nearly full motion. The exception to this is in the insulin dependent diabetic, where, unfortunately, full resolution does not always occur.

Making the diagnosis

This is usually straightforward as there are only really two conditions which cause a global loss off motion, that is, a loss of motion in every direction. Both are caused by the capsule getting thick and tight and restricting all movements. The most common is frozen shoulder and the other is arthritis of the shoulder (gleno-humeral joint). It turns out that arthritis is relatively uncommon and, by the time it starts to restrict shoulder movement, it can be seen on a plain x-ray. Hence, a restriction of motion in every direction, associated with a normal x-ray, is highly likely to be frozen shoulder.

The critical finding is that of a loss of external rotation. That is, a loss of the ability to turn the arm out. Whereas internal rotation (putting the arm up behind the back) and elevation (lifting it right up) can be restricted by scarring in the sub-acromial bursa (as occurs in chronic impingement syndrome), loss of external rotation can only be caused by capsular tightness. Because frozen shoulder can lead to secondary impingement, it is important to check this motion. By doing this, the primary problem will be elucidated.

The diagnosis can be more difficult in the very earliest stages of the disease, before there is a significant motion loss. If there is some doubt as to whether the diagnosis is one of frozen shoulder or impingement, and if the x-ray is relatively normal, then one should generally assume that this is frozen shoulder and not impingement: and treat it as such. The risk for missing this diagnosis is that, if the pain is bad enough to warrant surgery, and if sub-acromial (bursa) injections have failed, then sub-acromial decompression may be considered. This intervention, even though arthroscopic, has the potential to stir up the capsulitis (frozen shoulder) and make it worse,

The Normal Shoulder JointThe arm is in neutral position and the capsule

is loose.

The normal joint in external rotationThe arm is turned out, stretching the capsule

to its limit.

Frozen ShoulderThe capsule is thick and contracted. External

rotation is prevented.

Biceps tendon

Biceps tendon

Biceps tendon

Capsule

Capsule

Capsule

Humera

lHea

d

Glenoid

Humera

lHea

d

Glenoid

Hum

eral

Hea

d

Glenoid


and often for a very prolonged period.

Investigations

As mentioned above, the critical investigation is a plain x-ray, and this should always be done. An ultrasound may show evidence of impingement, but it will not differentiate between primary impingement and secondary (due to frozen shoulder or capsular restriction). The other problem with the ultrasound is that it may suggest a tendon tear when there isn't one. This is because secondary impingement can be very pronounced, leading to a lot of swelling in the tendons. If bad enough, this swelling can look like a hole on the ultrasound, the tendon fibres just looking like fluid. Therefore, such a report should always be interpreted in the light of clinical examination, and not be acted on independently.

MRI scanning has a place, but usually not as a primary investigation. It is good to look for other problems when the diagnosis is unclear, and it is helpful following surgery when an expected recovery path is not being followed. Unfortunately however, the MRI findings in this condition are soft and not always present. Things to look for however are, a thickened, tight, inferior capsule and, sometimes, some fluid in the joint which extends into the biceps sheath (see the diagram of the capsule on page 1 to see why this happens). This latter finding can also be seen on an ultrasound, something that is again highly suggestive of the diagnosis.

Treatment

There are very few things that can be done for frozen shoulder. It is an inflammatory condition. The capsule itself is inflamed and sore, and therefore, anything that stretches or pulls on it, will aggravate it. As such, during the inflammatory phases, physiotherapy and exercise do not help, and indeed, may actually make things worse. Exercises and stretching are therefore not recommended and, similarly, because of the potential for aggravation, any surgery should be avoided.

The one treatment that does seem to offer some help in this condition, is an injection of cortico-steroid (cortisone) into the shoulder joint itself. It is thought that the inflammatory cycle within the shoulder joint somehow eventually becomes self perpetuating, probably causing a continuation of the problem long after the instigating cause has gone. The aim then, is to break this cycle and to try and get the inflammation to resolve. It is uncertain as to why this cycle, once set up, continues along relentlessly but, for whatever reason, this does seem to explain the long time course of this problem.

Cortico-steroids are very powerful anti-inflammatory agents and they attack the inflammatory cycle at several points simultaneously. Because of this, they are much more powerful, and much more effective, than anti-inflammatory tablets (which generally have very little effect in this condition). If the cortico-steroid can be placed into the shoulder joint itself then, in over 80% of cases, the inflammatory cycle can be broken. This happens over a few days following injection and generally, by 1-2 weeks, the constant ache of the inflammation is gone. This usually means that the night ache resolves quite quickly. What is left however, is an ache or pain that occurs at the extremes of motion, when the capsule is pulled on or stretched. Given time however, this can also be expected to resolve.

Once the inflammation starts to subside, the swelling in the

capsule gradually decreases. As a result of this, the range of motion starts to improve and, in general, one would expect a return of around 10-15% of the range within two weeks of injection. To some degree however, this does depend on just how long the problem has gone on and by how much the range of motion has been restricted.

Once the inflammatory cycle has been broken, the body itself mops up the remaining inflammation. The capsule then starts to gradually stretch up, thus allowing motion to return again. After injection, because the inflammation has been treated, the condition passes through the stages towards resolution much more quickly. In essence therefore, the cortisone just advances the condition from the painful restriction of stage 2, to the less painful, but still somewhat restricted, stage 3. Once the night ache and overall pain is better however, most people can then easily put up with the restriction in range until the problem finally resolves.

At the time when the shoulder joint is injected, the sub-acromial space can also be injected to decrease any secondary rotator cuff impingement pain. This often dramatically reduces the inflammation in the rotator cuff tendons, thus reducing the arm pain that it causes.

MRI of a Frozen Shoulder showing a tight, thickened inferior capsule

MRI of a Frozen Shoulder showing swelling extending into the biceps sheath (fluid = white)


Continuing management

Once the inflammatory cycle has been broken it still takes a couple of months for everything to settle down and, during that couple of months, it is very important not to stretch the shoulder or push it around. This also means that activities such as sport should not be undertaken in that time frame. Subsequent to that however, once the inflammatory cycle has definitely resolved, a gradual stretching program can be embarked on and, at this stage, physiotherapy can be very helpful.

If during the recovery phase there is a flare up of the inflammation, then this should be treated immediately by re-injection to try and stop the inflammatory cycle reforming. Having said that, it is unusual to have to inject an affected shoulder more than once, and extremely unusual to have to inject it more than twice.

Full recovery of all motion may still take a year or so, depending on the initial restriction, degree of inflammation, time delay to treatment, etc. Nevertheless, the statistics favour a good result, with full or near full recovery eventually ensuing.

Insulin dependent diabetics

For some reason, insulin dependent diabetics can get a version of this condition that becomes protracted, may be severe in terms of both restricted range and pain, and does not respond well to conservative treatment. Cortisone is still tried in these individuals, but cortisone acts to oppose insulin, hence, for a couple of days after injection, the blood glucose levels may rise, requiring increased doses of insulin for control. To some extent, this does depend on how much of the cortisone is systemically absorbed, but that will be unknown at the time of injection. This then requires some vigilance by the diabetic individual concerned and, if not good at self managing insulin, then help may be required.

Other treatment

Injection under image guidance

If injection in the office does not work or cannot be done, then injection by a radiologist under image guidance may be helpful. Certainly, anyone who does not respond to injection in the office will be considered for this, and this is to ensure that the cortisone has been put into the joint.

Whereas injections into the sub-acromial space (bursa) are usually done under ultrasound guidance, those into the joint tend to be done under CT or x-ray control. Hence, these need to be organised by making an appointment with the radiology department. It is also recommended that it be done in a centre that does this a lot of this type of work, thereby increasing the chance of accuracy, and decreasing the number of problems. If it is deemed to be a consideration, we will recommend centres of expertise for you.

Some radiology departments do what is called hydro-dilatation, whereby the capsule of the joint is distended or ruptured by forcing fluid into it. This can make it very sore, and in fact, the evidence would suggest that it is no more effective than just putting cortisone in the joint. For this reason, we do not advocate this over just a simple cortisone injection, and we think that, in hydro-dilatation, the improvement is caused by the cortisone that is put in at the time and not by any dilatation. Given how thick and tough the capsule becomes in this condition, it stands to reason

that it cannot be stretched in any case: it just ruptures at a weak spot and the fluid leaks out.

Systemic Medication

If the above does not work, and the inflammation cannot be brought under control, then sometimes there is the option of using systemic drugs to help. Cortisone tablets are reasonably safe for short periods of time and may be helpful where an injection has not been. Usually this is started at a medium dose and then, when an effect is seen, the dose is gradually reduced. If this takes some weeks, then a gradual reducing dose to wean ff the medication may be necessary.

If that is not enough, and the joint is still very stiff, then a drug like methotrexate can be used. This is an agent that remains the mainstay of treatment for rheumatoid arthritis and it is a powerful anti-inflammatory agent. It is a once a week drug which, for this indication, is only given in low to medium dose. It has a large list of side effects but, fortunately, most people tolerate it well. It does however require monitoring of liver function on a monthly or so basis and, as the dose is increased, folate is required to reverse some of its effects at the end of the week. That is, the folate is taken 6 days after the dose. One day before the next dose.

The advantage of methotrexate is that it tends to reduce tissue swelling and can be used over extended periods of time, without the side effects of long term cortisone. Thus, when the problem is protracted, it may be a useful adjunct to other treatments.

Capsular release

If at the end of this process there is still a restriction of motion, then the remaining capsule may have to be released. This traditionally has been be achieved by manipulation under anaesthesia, but this causes an uncontrolled capsular tear which can be quite painful. It can also flare up the inflammation and cause a significant scarring response, leading to more delay in recovery. For this reason, the current treatment option for end stage loss of range, is an arthroscopic release of the capsule, literally cutting it, right around the joint near the socket. Whilst this would be easy in a normal shoulder, it can actually be nearly impossible in a very tight, restricted joint. It can be very difficult just getting an arthroscope into the joint, let alone getting an instrument into the joint to begin cutting the capsule. Hence, in trying to get good visualisation, some damage to the joint surfaces may occur. This is thus another reason not to do this to early in the peace.

For the patient who has had a major restriction of motion for a long period of time and is functionally disabled, a release may just help restore motion and relieve residual pain. Whilst in some centres there has been a recent tendency to do this earlier rather than later, the likelihood of success in most individuals is no different to conservative care (with injection and time). Also, there is a suggestion that early release may lead to some small but long term problems that are not seen in those treated without release. Only time will delineate this but, because of that, we regard this as last resort surgery and do not advocate its use until all other options have failed.

For the diabetic who has not improved with conservative care, release may be the option of choice. Unfortunately however, in these individuals, it does not always work, and sometimes, the end result is no better. Despite this, there are no other current options available.


Following release, return of motion can be slow, and further injections maybe required. Resolution may still take months.

Sub-acromial decompression

In the patient who has protracted impingement (possibly with sub-acromial spurs) which has been made worse by the concomitant onset of frozen shoulder, there may be a case to be made for sub-acromial decompression (removal of the spurs to stop impingement). Because of the risk of stirring up the inflammation in the capsule and making the restriction of motion worse however, it is generally recommended that the decompression be delayed as long as possible, and preferably, until the range of motion has either returned or has been static for a period of longer than three months.

Frozen shoulder as a complication of sub-acromial decompression is now thought to occur in up to 10% of cases. In general however, this is mild and responds well to injections into the shoulder and sub-acromial space as described. The problem however, is that the restriction of motion may take a while to be noticed over and above the normal post operative restriction, and hence, this can easily be mistaken for slow, but still normal, post operative progress. There may thus be a delay in making the diagnosis, which in turn delays appropriate treatment. If there is doubt, it is generally better to inject the shoulder anyway, just to be certain. There is no evidence that this causes harm.

Conclusion

Frozen shoulder is an inflammatory condition of the shoulder joint capsule. The treatment of the condition is directed towards the resolution of that inflammation. This is achieved by rest and local cortico-steroid injection into the shoulder joint itself. Any secondary impingement is also initially treated by cortico-steroid injection, but into the sub-acromial space (bursa). This relieves pain, by decreasing inflammation and swelling, in the capsule and tendons. The shoulder joint becomes less sore quite quickly, but restoration of full motion takes more time. If there is still some symptomatic impingement remaining after full motion has been regained, then sub-acromial decompression may help this.

It must be remembered that, because this is an inflammatory condition, the initial treatment is not one of stretches and exercises, but rather one of rest. It should also be remembered that, no matter how long the problem has persisted, the natural history for most, is one of full resolution.

Questions and concernsPlease contact Dr Holt’s office

Phone: +61 8 92124200Fax: +61 8 94815724Email: [email protected]

Further information can also be obtained on this and other related topics, such as:Impingement and rotator cuff tears

at: www.keithholt.com.au

dr keith holt - sports medicine€¦ · frozen shoulder dr keith holt frozen shoulder is a...

Documents