1.1

Introduction

At its 1993 meeting in Orlando, the International Association for the Study of Pain (IASP) defined the characteristics of complex regional pain syndrome (CRPS) types 1 and 2 . The criteria for CRPS type 1 are an initiating noxious event and continuing pain, allodynia or hyperalgesia for which the pain is disproportionate to the inciting event (associated or not with dystrophic phenomena). This definition assimilates adhesive capsulitis of the shoulder “frozen shoulder” to CRPS type 1 .

Management of this shoulder syndrome remains problematic and treatment outcomes are often inconsistent and only partially satisfactory . The natural history of CRPS is usually considered in three stages . In the acute stage, neuropathic-type pain and vasomotor dysfunction (with oedema and sweating disorders) predominate. In the dystrophic stage (after 3 to 6 months of disease progression), motor and trophic phenomena intensify, in parallel with an increase in pain levels. Lastly, in the atrophic stage, pain levels decrease but trophic and motor disorders persist. In fact, the boundaries between these three stages are not usually so clear and it is now accepted that the presence and intensity of the different symptoms can vary . Stage II CRPS of the shoulder usually presents as neuropathic-type pain and greater shoulder stiffness during external rotation and abduction. This stage often provides the physician with a dilemma: lack of mobilization accelerates the shoulder stiffness but mobilization during rehabilitation is often prevented by the induction or accentuation of pain.

We have developed a protocol that aims at both alleviating pain and increasing joint range of motion for patients with CRPS stage II of the shoulder and for whom previous treatments have not provided sufficient relief.

This was a multidisciplinary patient management process involving physicians from the Lorient-Quimperlé Pain Clinic, orthopaedic surgeons and rehabilitation physicians.

The patient was fitted with an interscalene catheter for one week, in order to implement a continuous brachial plexus block for at least partial pain control. Rehabilitation sessions were initiated concomitantly (with a view to gaining joint range of motion) and continued for 2 to 3 weeks after catheter removal.

From 2003 to 2008, 59 patients were prospectively evaluated 1, 6 and 12 months after the block procedure. All patients were interviewed by telephone at the end of 2008.

One month after treatment, we noted significant improvements in both pain and the joint range of motion. These improvements were still present after 6 months.

1.2

Patient and methods

1.3

Results

1.3.2

Pain levels

Over the first month, the initial pain VNRS score fell from 7.4 to 3.6 ( P < 0.001). There was a very slight rebound at 6 months (with a VNRS score of 4.2; P < 0.001) but the results were stable over time (with a pain VNRS score of 3.8 at 12 months and 2.1 at the time of the phone survey; P < 0.001) ( Fig. 1 ).

Pain on a verbal numeric rating scale from 0 to 10.

1.3.3

The Constant score

The Constant score rose from 21.7 to 56.5% at 1 month ( P < 0.001). This improvement persisted over time (54.4% at 6 months and 62.2% at 12 months; P < 0.001) ( Fig. 2 ).

Change over time in the Constant score.

The improvement concerned all four items of the Constant score, which were evaluated individually ( Fig. 3 ). Strength gains were modest but remained statistically significant ( P < 0.001).

Change over time in the various Constant score items.

Pain: +217% ( P < 0.001); activities of daily living: +170% ( P < 0.001); range of motion: +184% ( P < 0.001); strength: +265% ( P < 0.001).

1.3.4

Passive mobility

The shoulder joint’s external rotation range rose from 5.4° before treatment to 29.9° at 1 month, 31.5° at 6 months and 31.4° at 12 months. The joint’s abduction range increased from 38.6° at baseline to 74.2° at 1 month, 79.4° at 6 months and 83.3 at 12 months ( P < 0.001 at all time points) ( Fig. 4 ).

Change over time in the shoulder joint’s range of motion: abduction: +116% ( P < 0.001); external rotation: +481% ( P < 0.001).

1.3.5

Medication use

At the l-month evaluation, 50 of the 59 patients were only taking step 1 analgesics. The remaining nine were on step 2 analgesics. Twenty-eight patients had been able to abandon analgesic treatment after three years and 14 others had been able to reduce their dose or move to a lower WHO step ( Fig. 5 ).

Medication use after 3 years.

1.3.6

The telephone survey

Fifty-two of the 56 patients were contacted; of these, 86% said that they felt better or much better as a result of the treatment. Almost 70% were ready to renew treatment with the nerve block catheter, if required. In terms of professional activity, 24 patients had been able to return to work and seven had retired in the meantime. Thirteen of the 51 patients remained on long-term sick leave ( Fig. 6 ).

Professional situation.

1.4

Complications

During the treatment period, we noted 13 complications:

• •
  

  five catheter occlusions;

  
  
• •
  

  two cases of partial paresis;

  
  
• •
  

  one case of pneumothorax;

  
  
• •
  

  one catheter displacement;

  
  
• •
  

  one case of deep vein thrombosis of the leg;

  
  
• •
  

  ne case of haematoma at the puncture point;

  
  
• •
  

  one episode of major anxiety and;

  
  
• •
  

  one case of a catheter-related increase in pain.

These complications all occurred during the first month of treatment and did not have long-term, harmful consequences for the patients.

The occlusions prompted catheter repositioning and did not require the treatment to be interrupted. There were no negative effects on the outcome.

The two cases of partial paresis (in the C5 and C6 segments) were related to the catheter’s positioning and bupivacaine’s neurotoxic effects. Full motor recovery took just 2 months for one patient but over a year for the second.

The case of pneumothorax during catheter placement did not require drainage but did prompt the operation to be postponed.

Movement of the catheter during rehabilitation required repositioning and did not cause the treatment to be interrupted.

There was no clear causative link between our treatment protocol and the case of deep vein thrombosis of the leg.

The hematoma at the puncture point did not have any impact other than a week-long increase in pain levels.

Lastly, it should be noted that the episode of major anxiety and increased pain in one patient throughout the period of catheter use led to a disappointing treatment outcome.

1.5

Discussion

It is important to note that the patients included in this study presented CRPS type 1 of the shoulder with intense pain (a baseline VNRS score of 7.4) and significant stiffness (external rotation in position 1 of 5.4° and abduction of 38.6°), despite multidisciplinary follow-up in a pain clinic and the initiation of standard treatments for this pathology (as specified in the inclusion criteria) . Our study had the advantage of only including patients who were refractory to other therapies. Despite these “difficult cases”, our method produced significant decreases in pain and increases in joint range of motion within the first month. The results were slightly less good at 6 months but remained stable thereafter. This type of change over time has also been reported in the literature . Even though improvement is usually significant and rapid, it is only partial in most cases – regardless of the type of therapy . Manipulation under anaesthesia , arthroscopic release , hydraulic capsule distension and non-surgical treatments all fail to resolve residual pain or loss of mobility. Massoud et al. compared manipulation under anaesthesia, arthroscopic release and open arthroscopy and found residual pain in half of the cases . In a series of 45 patients treated with arthroscopy, Gerber et al. observed that most experienced some residual pain and loss of mobility and that a return to work was often delayed or prevented . Doddendorff et al. found that after more than 5 years, pain and stiffness persisted in 53% of his 75 patients treated with manipulation under general anaesthesia . The improvement observed in the present study had professional repercussions because of the 52 patients on long-term sick leave before the treatment, only 13 were professionally inactive 3 years afterwards. This outcome appears to be better than in other studies . Almost all the patients had been able to reduce their analgesic use after removal of the catheter and over half had stopped using analgesics after 3 years. Unfortunately, a lack of full data on the medication intake at 6 and 12 months prevented us from performing a valid analysis for these time points.

It is noteworthy that neither the initial professional situation (workplace accident, work-related disease, etc.) nor the occurrence of complications during the study period had an effect on the treatment outcome. The only statistically significant, negative predictive factor was the absence of a decrease in pain during the first phase of treatment (i.e. when the nerve block catheter was in place).

To the best of our knowledge, only four groups have reported on studies in which bupivacaine was used to treat painful, stiff shoulders . The nerve blocks were suprascapular in two studies, interscalene in the third study and intra-articular in the fourth study. However, these results are difficult to interpret because the series were small and very heterogeneous and did not difference between and evaluate CRPS types 1 and 2 or between the various disease phases. Continuous perfusion was only used in one case.

Bupivacaine is known for its preferential effect on the smallest fibres, including C fibres and sensory fibres. At the concentrations used in our treatment protocol, bupivacaine mainly targets the C fibres. It is interesting to note that in the great majority of cases, interscalene block enables pain control and, above all, that the pain level hardly increases when the block is removed.

This could be due to continuous, prolonged blockade of the C fibres by bupivacaine.

The lack of a control group means that our results must be interpreted with a degree of caution. Although the significant progress made by treatment-refractory patients in the first month post-block emphasizes the efficacy of our treatment programme, the assessments are at 1 and 3 years are more delicate, given the natural history of CRPS type 1. Moreover, we were only able to assess 23 patients at 1 year post-block. However, during the telephone interview at 3 years (with 52 of the 59 patients), it is interesting to note that the great majority said that their condition had been improved or very much improved by our treatment programme. Despite the burdensome nature of the protocol, 70% said that they would undergo the same treatment again, if needed. Patients who failed to respond adequately to the continuous interscalene brachial plexus block were provided with follow-up treatment in the Lorient-Quimperlé Pain Clinic.

1.6

Conclusion

Although the long-term outcomes in the present series do not appear to differ fundamentally from those reported in the literature, our treatment protocol significantly decreased pain and increased shoulder joint range of motion (from the first month of treatment onwards) in CRPS stage II patients who were refractory to conventional therapies. Despite a slight deterioration at 6 months, these initial gains were stable over time.

Conflict of interest

The authors have not declared any conflict of interest.