Thoracic Outlet Syndrome; are clinical tests useful?

Following on from the first blog that was unable to draw associations between anatomical variations and thoracic outlet syndrome (TOS); I thought it would be beneficial to review the currently used clinical tests, underpinned by a few questions, such as, are these tests valid? Does the research support using these tests in clinical practice? How much weighting can be put on these tests to provide clinically meaningful information in practice?

The overall incidence of TOS is rather elusive in the literature due to the difficulty in its diagnosis; however, when TOS is diagnosed, over 90% of those cases are reported to be neurogenic in nature, as little as 1% arterial and around 3% venous (Sanders et al 2007). Clinical features of true neurogenic TOS are described as; sensory loss particularly within the C8 and T1 dermatome, pronounced weakness and atrophy to the thenar eminence and possibly the medial aspect of the forearm; combined with a history of aching along the inner border of the upper limb (Ferrante 2012). Arterial or venous TOS has been reported to present with pallor or cyanosis to the hand, pins and needles, cold intolerance and pain associated with repetitive or strenuous upper limb activity (Hooper et al 2010). Within clinical practice, there are a number of differential diagnoses to consider, such as cervical radiculopathy and carpal tunnel syndrome, but this is beyond the scope of this review.

Are clinical tests used in practice?

There has been a lot of discussion amongst Physiotherapist and researchers over recent years with regards to musculoskeletal clinical tests, specifically around the shoulder. It has been identified that a lot of clinical tests can be highly sensitive, yet lack specificity and cannot accurately diagnose specific pathologies (Hughes et al 2008). Alternative methods of assessment have been proposed, such as the shoulder symptom modification procedure (Lewis 2009), which aims to provide some direction to the overall management of a person’s shoulder problem as opposed to just reproduce their pain. This made me think about clinical tests in relation to TOS; as mentioned above, it is already known that TOS is a difficult condition to diagnose and hence there is not a gold standard way of proving a person has this condition or not. So, how will clinical test help in this situation? What are we intending to prove? How will the management of the person change if the tests are positive? Or negative?

The general consensus…

There are few recent papers on TOS that provide an overview of the aetiology, clinical presentation, examination procedure and then discuss the copious amounts of differential diagnoses to consider (Hooper et al 2010, Ferrante & Ferrante 2016), but there are no up to date studies that have investigated the effectiveness of the clinical tests; therefore authors of the most recent papers for example, Hooper et al (2010) have incorporated older literature that have documented sensitivity and specificity values into their work.

Clinical tests highlighted in the literature are listed below: –

• Cyriax Release Test
• Adson’s Test
• Costoclavicular / Halsted Manoeuvre
• Elevated arm stress test (EAST) / Roos
• Wright’s Test/ Hyperabduction manoeuvre

After reading the articles on the clinical tests, I thought it would be best to summarise my thoughts on what the studies are suggesting, highlighting both the positive and negative aspects before drawing to a conclusion on whether they are useful or not.

Firstly, all of the clinical tests mentioned above are diagnostic / symptom provocation tests, looking to either reproduce pins and needles or look for a change in the radial pulse whilst the test is being conducted. Let’s start with the positives; unfortunately, there aren’t many, there have been a few studies which have reported using people with TOS (Gilliard et al 2001) or people with suspected TOS (Demirbag et al 2007) and attempted to provide sensitivity and specificity values. However, the inclusion criteria for these subjects have not been well documented or have been based on the researchers own bias views on what constitutes TOS (Gilliard et al 2001), or have included a wide variation of subjects with alternative diagnoses, such as Raynauds phenomenon or carpal tunnel syndrome and because they have positive clinical tests, they have been included in the study (Demirbag et al 2007).

This swiftly brings me onto the negatives; a lot of the research has been conducted on asymptomatic individuals, the aim of these studies has been to identify the high rate of false positive test results (Brismee et al 2004, Rayan & Jenson 1995 & Plewa & Delinger 1998). Plewa & Delinger (1998) found false positive results for the Adson’s test, the Costoclavicular manoeuvre and the East test; similar findings were made by Ryan & Jenson (1995) for the Adson’s test, Costoclavicular manoeuvre and also the Hyperabduction manoeuvre. Brismee et al (2004) found a total of 35 out of 119 allegedly asymptomatic participants reported pins and needles or paraesthesia within 15 minutes of holding the test position. Just to add more fuel to the fire, a study by Nord et al (2008) compared the outcome of the Adson’s test, Roos and costoclavicular manoeuvre in asymptomatic participants (n= 86) to people with carpal tunnel syndrome (CTS) (n=62); they found high false positive results in the asymptomatic group but even higher false positive results in the group with CTS.

Unfortunately, there were methodological flaws in all of the above studies due to a lack of standardised equipment used when measuring the radial pulse, the inclusion of people with other conditions such as carpal tunnel syndrome and diabetes in what was supposed to be an asymptomatic participant group and also the absence of a recognised reference standard when drawing conclusions on sensitivity and specificity values, to name a few.

So, are clinical tests for TOS useful?

Based on the literature that is available, then unfortunately, they are not useful. In general, the research is of poor quality to start with, despite this, there are more studies looking at the high false positive ratio the tests produce than anything else, which speaks volumes.

One of the main issues on the topic of TOS is the lack of a gold standard; which means, even if clinical tests were used and they were positive, there would be no way of proving that the person had TOS anyway. It has also been illustrated that people with conditions such as carpal tunnel syndrome can have positive tests and I’m sure, if anyone one of us held our arms in one position for 15 minutes, we too, may begin to feel pins and needles.

 

So, in future, what advice do I give to colleagues surrounding clinical tests for TOS?

Aside from advising them not to spend time conducting the clinical tests mentioned above or rushing people off for scans and unnecessary investigations to see if they have any anomalous ribs, perhaps this highlights the need for clinicians to adopt more of an emancipatory approach to their assessment of individuals; instead of having an agenda to label or diagnose, try to understand the individuals view point and their experience of living with a health condition (Trede 2012). Giving someone a chance to be able to tell their story, actively listening but more importantly, understanding exactly what they are saying and demonstrating an appreciation of how much something is affecting a person’s life are all key factors that can positively influence the patient-therapist interaction (O’Keefe et al 2016). As clinicians, we are all familiar with the variety of responses seen between individuals following a total hip replacement, despite the identical procedure, it’s the individual’s biography that undoubtedly articulates their response to a given situation (Kirkengen & Thornquist 2012).

As Physiotherapists, a large part of our role involves trying to influence behaviour change and according to the health belief model, people are more likely to change their behaviour if they perceive this behaviour will have positive health benefits (Carpenter 2010); for some people, having a clear diagnosis may encourage this behaviour change to occur, but remember, one size does not fit all! Sometimes we are not always able to provide a clear diagnosis but it is essential that we do identify the nature of the presenting symptoms. Conducting a thorough assessment of the area and using appropriate, alternative clinical tests to help rule out sinister pathology and rule in neuromuscular dysfunction or being able to alleviate some of their symptoms during the physical assessment can assist in the clinical reasoning process.

Writing this blog has afforded me a greater understanding on an unfamiliar topic, but it has also made me reflect on what my aims are within an assessment. I used to try and attach a very biomedical label to every person that I assessed, they have “impingement” or a “disc bulge”, but I am not sure whether this is particularly helpful. There is qualitative research that has found some people with low back pain, that have a predominantly biomedical understanding of their condition, don’t understand why medical interventions don’t work (Snelgrove & Liossi 2009). This is completely understandable, if the medical encounters these people have experienced, have focused on their condition from a purely biomedical perspective.

For me, in my future clinical practice I will spend less time trying to fit people into boxes and spend more time listening, empathising and most importantly, focussing on trying to understand a person’s view point, instead of merely reducing them to a condition within 30 minutes.

 

 

Blog 2: Question 1

“Great blog Steph. Really interesting to read. With your findings that there is poor evidence for tests with high false positives then where does that leave us with physio assessment for TOS? Are there any alternative diagnostic tests to diagnose this condition?                                                              

Great work, Nicky.”

 

Great question Nicky. I think TOS, particularly the neurogenic type is difficult to diagnose generally. There are a lot of differentials to consider such as carpal tunnel syndrome, cervical radiculopathy, nerve root irritation or even something more sinister like a tumour which may produce similar findings when it comes to objective clinical testing. As the Illig et al (2016) article suggests, an in-depth work up seems like a sensible and systematic approach. I also think it may be better to use objective tests that have research behind them to support their use.

Regarding the arterial and venous presentation, Alan Taylor & Roger Kerry (2015) have devised a useful table of descriptors that are associated with vascular dysfunction in the peripheral regions, they separate it into upper and lower limb and then have documented both signs and symptoms for an arterial compared to more of a venous presentation. See table below for details. These types of signs and symptoms if present, would assist in the clinical reasoning process and also help with differential diagnoses.

 

Upper limb Signs/Symptoms	Arterial	Venous
Exercise induced pain	X	X
Numbness/tingling	X
Cold	X	X
Hot		X
Blue		X
Fatigue / Weak	X
Non-Dermatomal Pain	X	X
Cramp	X	X
Whitened skin / blanching	X
Swelling		X
Subjective swelling – non seen	X	X
Redness		X
Band of pain	X
Throbbing / pulsatile	X

 

(Table from Grieve’s Modern Musculoskeletal Physiotherapy Fourth Edition (2015) Chapter 35.2 Haemodynamics and Clinical Practice by Alan Taylor & Roger Kerry, Page 349.)

Additional vascular tests such a capillary refill may form another component to the clinical assessment, I appreciate this test has its own problems with variations of capillary refill time dependent upon, age, sex and environmental temperature (Lewin & Maconochie 2008) when used as a measure of central pressures, however when used for peripheral purposes and comparing to the person’s other limb, this may provide Physiotherapists with some objective measures.

The same principles could be applied for assessing the radial or brachial pulses and comparing sides, so long as the technique during the assessment is adequate, then absent or diminished pulses can be indicators of occlusion due to a thrombus or embolus (Nicholson 2014).

With regards to the neurogenic type, there is some recent research that highlights the difference in sensory testing in terms of loss or gain of nerve function that you may see between different subgroups, such as cervical radiculopathy, fibromyalgia and people with nonspecific arm pain (Tampin et al 2012). So far, the research on TOS provides a wide range of signs and symptoms we could see in people affected. However, in order to assist Physiotherapists in their differential diagnoses and to provide some objective markers, establishing whether a person has a gain or loss of nerve function will enable us to gauge physiologically what is happening to the large and small nerve fibres. In a clinical setting, this sensory testing would include, light touch, pink prick, vibration and thermal testing and nerve palpation.

There is also evidence to support the use of a standard upper limb neurological examination in people with a musculoskeletal presentation, using myotomal muscle testing (C4 -T1), reflexes (C5-6 /C7-8), nerve palpation (ulnar, median and radial) and upper limb neurodynamic tests (ulnar, median and radial), demonstrating moderate to substantial reliability (Schmid et al 2009).

The Upper limb tension test may also have a place in the assessment, with evidence to support its use when used in conjunction with additional nerve function tests as described above (See response to Alex’s question for details and references). Despite these clinical tests, I feel it would still be difficult to officially “diagnose”. Depending on the findings from the objective assessment, it may be that additional investigations or an onward referral may be appropriate. If there’s nothing concerning the clinician following their assessment, it may be that a course of Physiotherapy is indicated and providing a diagnosis may not necessarily be required.

 

Blog 2: Question 2

“Hey Steph,

I really enjoyed the blog, and funnily enough good timing, I am not sure if you saw the article in frontline (July’s edition) on TOS but it was interesting. In that article, they refer to an article by Dr Illig et al (2016) about a way to diagnose TOS accurately. (see July’s Frontline p 16-17 for article)

I agree with you that it does seem difficult to assess the specificity and sensitivity of tests when there is no gold standard to compare them to. Do you think then that we should do as Jeremy Lewis has done with the shoulder and refer to the condition on a general term, and then use a symptom modification method for treatment? (obviously taking into account the patient’s situation and beliefs as you refer to). I have read, some of the work by Karl Illig and Hugh Gelabert (who include the EAST test in their documentation) and they are trying to build on the research in an attempt to develop more consistency in diagnosis, treatment and assessment results thus allowing more data to be put to the test of academic rigor and scrutiny therefore shedding more light on this topic.

What are your thoughts? Forget these special tests completely? From the blog, it would imply you would but then what about how you might relay this to colleagues and consultants when discussing patients.”

 

Hi Alex,

Thank you for your question.

I certainly have read the article by Illig et al (2016) which was posted in the current front-line edition. I did too find this interesting, because it goes into such a lot more depth in their work up and they also take in to consideration psychosocial factors which will definitely be a contributing factor to a person’s symptoms.

As you mention, they do still include the EAST test for neurogenic type only, this does surprise me considering the literature available; their justification for using this particular test is because they feel it causes narrowing to the interscalene triangle and therefore stresses the brachial plexus. Within the article, they do not provide a reference to support this statement. I did however, find an article by Fried & Nazarian (2013) who have used dynamic ultrasound imaging to correlate whether they can visualise compression at the brachial plexus that coincides with physical symptoms during arm elevation. They use case study examples so the level of evidence is low and there is no certainty that this compression is directly linked to the person’s symptoms.

On balance, I don’t believe there is enough robust evidence to support the use of the EAST test, however if a person’s symptoms were occurring with arms above their head, all Physiotherapists would look at the person in their provocative position. So, although there may not be evidence to support the test specifically, you could probably justify the arm abduction / elevation component when assessing function, if consistent with their aggravating factors.

The upper limb tension test (ULTT) is also used in the neurogenic type work up within the Illig et al (2016) article, this test is more promising and has evidence to support its use.  The reference used within the article is the study by Kleinrensink et al (2000); using cadavers, the study found the upper limb tension test 1 to be a valid clinical test and does put tension on the cords of the brachial plexus. This is supported by biomechanical research that highlights specific nerves are subject to increased pressures when placed in provocative positions, i.e. the median nerve in carpal tunnel syndrome (Werner et al 1997). However, the most recent research advises neurodynamic tests should not be used in isolation when trying to identify nerve dysfunction, as negative tests can actually indicate more severe dysfunction (Baselgia et al 2017).

In summary, I think there is more substantive research to support the use of the ULTT than the EAST test to help identify TOS but there is also a wider picture to consider, which the paper by Illig et al (2016) has illustrated. I think when discussing clinical cases with colleagues, it would be more beneficial if additional neurovascular testing was used in combination with the patient history, thorough clinical reasoning and to show an appreciation that alternative diagnoses have been considered. (See response to Nicky)