HYPOTHESIS: Prolotherapy, the injection of a growth promoting solution in injured ligaments and tendons of the shoulder is an effective treatment that decreases pain, increases functional capacity and promotes healing better and in less time than standard treatment with physiotherapy.

OVERVIEW: 75 subjects with rotator cuff tendinopathy proven by ultrasound will be recruited and assigned randomly into one of three groups of 25 to receive one of these three different treatments:

Group A (test): 25% dextrose with 0.1% lidocaine, injected into the tendons and ligaments Group B (control): 0.1% lidocaine injected in the rotator cuff tendons and ligaments Group C(control): 0.1% lidocaine injected subcutaneously above these structures All subjects will receive physiotherapy every other week for three months. To avoid placebo effects, patients, the radiologist and physiotherapist will not know to which treatment group the patients belong; the physician administering the injections will not be involved in assessing disability before or after treatment. (Note: The physician will know which patients belong to group C because it will be obvious: they are delivering a subcutaneous - versus a joint - injection).

There will be three sets of injections - one set per month for 3 months. The patients' condition will be tracked for nine months after the first treatment, to monitor changes in 3 outcome measures: pain (VAS and Rx #s), function (DASH and PESS), and tendon healing (as assessed by ultrasound).

SAMPLE SIZE: Sample size justification: Dr. Bertrand reviewed 144 cases treated with prolotherapy, 25 of whom fit the inclusion criteria (between the ages of 19 and 74 and not suffering from arthritis of the shoulder or from frozen shoulder). Among those patients, the difference between the pain score prior to treatment and that when they were given the questionnaire averaged 5.36 points over10 with a standard deviation of 3.28. Improvements from baseline to 6 months on the true placebo arm were projected to be about half those of the prolotherapy group, averaging 2.7 points of improvement with a standard deviation approximately equal to that of the prolotherapy group of 3.28. In order to achieve an 80% power of detecting the differences in mean improvements, given each group has a standard deviation of 3.3, and the average difference between groups is 2.7 points out of 10, each group should include 25 subjects.

RECRUITMENT AND CONSENT:

Recruitment: Ads were sent to all the North Shore and Burnaby physicians, physiotherapists, and chiropractors, posted in the Lions gate hospital ultrasound waiting room and North Shore recreation centers. A small adwords ad referred subjects to our website, www.prolotherapyhealing/shoulder.

Consent: At the initial appointment with Dr. Bertrand, the patient will be given information about the nature of this study and the procedures involved, including the fact that each of three treatments will consist of numerous shoulder injections. He will also be told that up to three monthly treatments may be needed to achieve relief from shoulder pain. If he decides to proceed, he will be given a copy of the consent form, the DASH questionnaire, and the VAS pain scale to fill out. The intake assessment includes socio-demographic information and medical history including the duration and severity of symptoms, smoking status, arm dominance and general health. If the patient manifests symptoms and signs of a rotator cuff tendinopathy which has lasted more than three months, and has no exclusion criteria, he will be scheduled for an ultrasound and an x-ray examination of the shoulder. The radiologist will rate the ultrasound using the USPRS. If no exclusion criteria are found, the patient will be referred to the physiotherapist for a baseline assessment of his shoulder using the (PESS), and scheduled for his first treatment session. The patient will be told to avoid all anti-inflammatory medications for the four weeks following each treatment. He will be given a prescription of Tylenol three or Tramacet as needed for pain control.

RANDOM ALLOCATION OF PATIENTS, AND STUDY GROUPS:

At the first treatment session the patient will be assigned his sequential number. A previously generated master list will have randomly allocated each number either to treatment (A) or to one of two control groups (B), (C) using a permuted block randomization scheme. (The pharmacist will generate and have custody of the master list, using dice to ensure a random allocation.) Dr. Bertrand's file will record the patient's number. The pharmacist will prepare all the mixtures in advance and label them with the patient's number including the letter C, if the patient is to receive lidocaine subcutaneously.

Group A - intervention (treatment) group. Will receive three, monthly, prolotherapy session using a mixture of 25% dextrose and 0.1% lidocaine (50% dextrose mixed with equal parts of 0.2% lidocaine solution (one part 1% lidocaine mixed with four parts normal saline)) in the rotator cuff ligaments and tendons .

Group B - modified prolotherapy control group. Will receive a mixture 0.1% lidocaine (one part 1% lidocaine mixed with nine parts normal saline solution) in the rotator cuff ligaments and tendons.

Group C - true control group. Will receive 0.1% lidocaine (one part 1% lidocaine mixed with nine parts normal saline solution) subcutaneously above these same locations.

Intention to treat: Patients will be included in this study if they have received one set of injections. If, after the first treatment, they fail to keep their physician appointment, the secretary will telephone the patients and administer the DASH questionnaire and the VAS by phone. She will do this according to the study timeline (i.e. one, two, three and six months following their first injection session).

INTERVENTION PROTOCOL: Injection procedure. After patients are randomized to Groups A, B, and C, they will receive three sets of injections (one set per month for 3 months). The entire shoulder area will be sterilized with 70% alcohol. In order to decrease the pain of the actual injections, the areas to be treated will be injected intradermally, using a number 32G one half inch needle, with a mixture of 1% Xylocaine buffered by an equal volume of sodium bicarbonate solution for both subjects and controls. The intervention will consist of the following: For groups A and B, the solution will be injected, in 1 mL amounts, using a 27G 1½ to two inch needle in the following locations:

Supraspinatus tendon: on the anterior-superior part of the greater tuberosity, this tendon is generally tender to palpation over an area of about 2 to 3 cm in height and 1 cm in width. Unless the tendon is tender to palpation, 1 mL of solution is injected. All tender parts of the tendon are injected at 1 cm intervals, each with .5 mL of solution, a total of 1 to four injections.

Infraspinatus tendon: immediately posterior to the supraspinatus tendon, it is in line with the spine of the scapula on the greater tuberosity. Unless the tendon is tender to palpation, 1 mL of solution is injected. All tender parts of the tendon are injected at 1 cm intervals, each with .5 mL of solution, a total of 1 to four injections.

Teres minor tendon: attaches on the posterior superior surface of the greater tuberosity. Unless the tendon is tender to palpation, 1 mL of solution is injected. All tender parts of the tendon are injected at 1 cm intervals, each with .5 mL of solution, a total of 1 to four injections.

Teres minor and triceps insertion on the outer edge of the scapula. To avoid the risk of injecting the lungs, the injections are made with the subject's arm in extreme adduction, as if to scratch his back. This brings the scapula forward, distancing it from the rib cage. These areas are injected only if the subject complains of tenderness to palpation. All tender areas are injected at 1 cm intervals, each with .5 mL of solution, a total of 1 to four injections.

Coracoid process: insertion of the short tendon of the biceps. It is the bony pro-eminence under the clavicle medial to the head of the humerus. 1 mL of solution is injected in the area.

Long tendons of biceps: immediately medial to the acromioclavicular joint, with the arm in slight external rotation a 2 inch needle is inserted vertically with a 15° anterior tilt until bone is reached. 1 mL of solution is injected in the area.

Subscapularis tendon: inserts on the lesser tuberosity of the humerus posterior to the long tendon of the biceps. With the arm in full external rotation and adduction, a 2 inch needle is inserted 1 cm lateral to the coracoid process until it reaches the humerus. 1 mL of solution is injected in the area. Additional injections can be performed 1 cm above and 1 cm below this area with .5 mL of solution if the subscapularis is tender to palpation.

Inferior glenohumeral ligament: with the arm externally rotated and abducted 90°, a 2 inch needle is inserted posteriorly 1 cm below the coracoid process until the inferior part of the glenohumeral joint is reached. 1 mL is inserted in the area and the needle is moved medially and laterally. Another milliliter is injected in each of those two areas, corresponding to the scapular and humeral insertion of the inferior glenohumeral ligament. 22G, 3 or 4 inch needles are used for obese subjects.

All needle insertions are made until bone is reached. No injection is done otherwise.

At the end of the procedure, the entire shoulder joint should be pain free and exhibit full or almost full range of motion. The patient will be warned that this pain free status will only last as long as the local anesthetic is in place.

The control group (Group C) will be treated in all respects identically to groups A and B, with the following exceptions: (1) they will receive injections of 0.1% lidocaine combined with normal saline (unlike group A), and (2) the depth of the injections will be different, subcutaneous rather than inside the ligaments, but the same volume of solution will be used as for groups A and B.

To assess if the subjects are truly blind as to their treatment, they will be asked to state which treatment (true prolotherapy, modified prolotherapy, placebo, don't know) they think they have received after each treatment, and this will be recorded on the patient's chart and database. Neither the patient nor the physiotherapist or radiologist will be aware to which group the patient has been randomized.

Outcome measures assessment: At each of the 3 treatment sessions, we will: (1) record side effects experienced during the session or following the previous injection session, and (2) administer the VAS pain scale, and DASH (disabilities of the arm, shoulder and hand) questionnaire prior to being treated. If requested, patients will receive a prescription for either Tylenol number three or Tramacet to deal with post-treatment pain. The amounts prescribed and the amounts used will be recorded.

Throughout the 3-month injection intervention, subjects will also receive physiotherapy every 2 weeks. One month and 3 months after the third set of injections patients will be reassessed by the physiotherapist using the (PESS) and also fill out the VAS pain scale and answer the DASH questionnaire. Finally, a follow-up ultrasound (USPRS) will be administered 3 months after completing the injections.

Nine months after treatment started: follow-up telephone survey, to determine whether improvements are retained.

STATISTICAL ANALYSIS:

With the exception of 'total amount of pain medication used', the main assessment for each of the outcome measures (VAS, DASH, PESS, USPRS), will be based on the difference in each patient's scores from their first (baseline) visit to their last visit six months later. For each outcome, the three groups will be compared using ANOVA. In cases where the ANOVA indicates differences between the three groups, subsequent comparisons of the prolotherapy group to each of the control groups will be carried out. The overall primary assessment is the comparison of the prolotherapy group (A) to the true control group (C) on the VAS improvements from baseline to six months; all other analyses are secondary. All statistical tests will be two-sided and differences will be considered statistically significant if the p-value is less than 0.05.

To further explore the pattern of the treatment effect over time, the improvements from baseline to other time points will be analyzed similarly. Methods of longitudinal data analysis will be used to carry out more comprehensive exploratory analyses.

Other variables, collected at the first patient visit, such as duration of pain prior to randomization, the patient's age and gender, their general health and smoking status, whether the tendinopathy affects the dominant arm, and the severity of the tendinopathy can influence the duration and intensity of pain, shoulder function, and ultrasound evidence of healing. The relationships of such prognostic factors with the outcome measures will be examined. In cases where strong relationships are apparent, the above analyses will be repeated with adjustment for such factors. That is, the analysis of covariance (ANCOVA) will be used in place of ANOVA, with the corresponding subsequent comparisons of the prolotherapy group to each of the control groups.

SIGNIFICANCE:

Rotator cuff tendinopathy is known to be notoriously resistant to standard therapy using NSAIDs, physiotherapy, steroid injections and surgery. After three years only 46% are pain-free. In my experience, using prolotherapy, 75% are more than 50% improved after three months, at which time 42% are pain-free. If this experience is corroborated by the current research project, this may augur a radical change in the way rotator cuff tendinopathy is treated. The materials needed for prolotherapy ( syringes, needles, dextrose solution and local anesthetic) are inexpensive and available to all family practitioners regardless of their location. The knowledge needed is relatively easy for family physicians to acquire, making this treatment potentially accessible even in the remotest rural locations where surgery and even physiotherapy are unavailable. There are no major nerves or blood vessels in the areas injected, so the risks of this treatment are extremely low (e.g. bruising and soreness for a few days following the injections. These complications compare favorably with the 10% complication rate following shoulder surgery, which also requires a prolonged postoperative recovery period, and the, at times serious, potential side effects of NSAIDs and corticosteroid injections. Getting people to return to work sooner, and for much less cost than the current therapies (3x $150 compared to $24,300) would benefit both the healthcare system and those who suffer with this problem.

Inclusion Criteria:

• consecutive patients over 19 and less than 75 years of age

Exclusion Criteria:

• allergy to corn, as the dextrose solution is corn-based
• allergy to local anesthetic
• immune deficiency
• conditions requiring anti-inflammatory medications including prednisone, corticosteroid injection less than eight weeks prior to the first set of injections
• use of immune suppressants
• symptomatic osteoarthritis of the gleno-humeral or acromio-clavicular joint
• age over 75 or under 19
• adhesive capsulitis based on a thorough physical examination, where shoulder flexion or abduction is below 100 °, horizontal adduction is below 30 °, the hand behind the back is below the waist, external rotation is less than 30 °
• full thickness tear greater than 1.2 cm as seen on ultrasound
• autoimmune disorders such as lupus or rheumatoid arthritis
• neurological disorders including Parkinson, seizures, and dementias are excluded for patient safety during the procedures
• HIV, viral hepatitis and other blood borne communicable diseases, to protect the investigators
• calcium deposits greater than 8 mm in diameter
• type III acromion as seen on x-ray
• painful condition elsewhere in the body likely to cloud the subject's assessment of his shoulder pain
• no evidence of tendinopathy as seen on ultrasound
• uncontrolled diabetes: A1C > 7

The de-identified database Is now available on request to Doctor Helene Bertrand at heleneb@mail.UBC.ca

The data will be available from January 2015 to January 2020

This study was published in Archives Of Physical Medicine And Rehabilitation 2016 volume 97 pages 17 - 25 under the title: Dextrose Prolotherapy Versus Control Injections In Painful Rotator Cuff Tendinopathy

1 year

For the 25% dextrose and .1% lidocaine and the .1% lidocaine enthesis injections, only the pharmacist knew the contents of each participant's numbered bottle. For the .1% subcutaneous injections, both the pharmacist and the physician where aware of the contents. There was a small c on the label. Only one participant asked the meaning of this letter, and he was told it stood for "contents".