Randomization will be performed after patients are deemed eligible and have provided informed consent. Patients will be randomly allocated to the concentrated autologous platelet group (PRP group) or to the corticosteroid group (control group). A computer using block randomisation of 10 patients will be used to create a randomization schedule. Treatment assignments (placed in sequentially numbered opaque envelopes) will be assigned by the trial managers (WL, HS, HH, TG) who will also arrange the facilities needed for the procedure.
All patients with a plantar fasciitis who are admitted to one of the participating hospitals and meet the inclusion criteria are asked to join the study.
Plantar fasciitis was defined as pain at the point of the fascia plantaris origin at direct palpation. All affected patients were screened with X-ray of the calcaneus for bony abnormalities and to difentiate for subtalar arthritis. Sonography and MRI were not used standardly.
The Medical Ethical Committee of The Netherlands approved the study design, procedures and informed consent.
Trial registration number: http://www.clinicaltrials.gov NCT00758641.
The study will be conducted at the Orthopaedic Departments of the HAGA Ziekenhuis Den Haag, St. Antonius Ziekenhuis Nieuwegein, Diaconessenhuis Leiden and St. Elisabeth Ziekenhuis Tilburg between November 2008 and December 2010. Authors JP and TG will be responsible for the data and safety monitoring. Inclusion criteria: patients aged > 18 years, with plantar fasciitis (6-12 months duration), who failed conservative treatment are included. They have to be able to understand the informed consent and have a VAS pain score in the morning by first steps higher as 5 (0-10 scale).
Patients will be excluded from the study when they received local steroid injections within 6 months, physical/occupational therapies within 4 weeks, or non-steroidal anti-inflammatory within 1 week prior to randomization. Also patients will be excluded with the inability to fulfil follow-up criteria, significant cardiovascular, renal or hepatic disease, pregnancy, (local) malignancy, history of amenia (hemoglobin < 5.0), previous surgery for plantar fasciitis, active bilateral plantar fasciitis, diagnosis of vascular insufficiency or neuropathy related to heel pain, hypothyroidism and diabetics.
Platelet Concentrate Preparation
Fifty-five milliliters whole blood is collected from the uninvolved arm into a 60-mL syringe that contained 5 mL sodium citrate. A peripheral complete blood count is also collected at the time of the initial blood draw. The blood is then prepared according to the GPS System instructions (Cell Factor Technologies, Warsaw, Ind). This device is a desktop-size centrifuge with disposable cylinders for the blood approximately 0.05cc Platelet concentrate is obtained for each patient.
Autologous platelet concentrate contains concentrated white blood cells and platelets that are suspended in plasma. Since an acidic anticoagulant is introduced to the whole blood used to produce the platelet concentrate, the platelet concentrate must be buffered to increase the pH to normal physiologic levels. This is accomplished with 8.4% sodium bicarbonate solution added at a ratio 0.05cc of sodium bicarbonate solution to 1 cc of platelet concentrate.
The resulting buffered platelet concentrate contains approximately a 6 to 8 times concentration of platelets compared to baseline whole blood. No activating agent is used. The total time from blood draw to injection in the patients is about 30 minutes. No specialized equipment, other than the GPS machine, is required.
Initially, bupivacaine is infiltrated into the skin and subcutaneous tissue of both groups as a local field block. Approximately 0.05cc is also injected directly into the area of maximum tenderness. Then, either 5 to 6 cc platelet concentrate or 5 to 6cc corticosteroid is injected using a 22-g needle into the plantar fasciitis using a peppering technique. This technique involved a single skin portal and then 5 penetrations of the fascia.
Immediately after the injection, the patient is kept in sitting position without moving the foot for 15 minutes. Patients will go to the physiotherapist to obtain stretching exercises. Patients are sent home with instructions to limit their use of the feet for approximately 48 hours and use hydrocodone or acetaminophen for pain. The use of nonsteroidal medication is prohibited. After 48 hours, patients are given a standardized stretching protocol to follow for 2 weeks. A formal strengthening program is initiated after this stretching. At 4 weeks after the procedure, patients are allowed to proceed with normal sporting or recreational activities as tolerated. Any type of foot orthoses will not be allowed.
The type of steroid that is used during the study is kenacort 40 mg/ml triamcinolon acetonide.
Main study pain
Pain will be measured using a visual analogous scale at all time points. The VAS score of the Foot Function Index will be used.
The score records the patient's reported pain using a scale of 0-10, where 0 is pain-free and 10 is the worst pain imaginable. The scale will be a 10 centimeter line beginning with 0 and ending with 10, the score will be marked at the point on the line that corresponds with the patient response.
Each patient randomly assigned to a treatment and with at least one non-missing pre- and post-baseline measurement, will be classified at each visit, as either a treatment success or failure. Patients with a score reduction of 25% compared to baseline, did not require pain medication beyond the protocol defined allowable period, and did not require escape therapy will be considered a treatment success.
The absolute change from baseline to endpoint means the baseline value is subtracted from the endpoint value. Percent change is defined as the absolute change multiplied by 100 divided by the baseline value. For patients whose pain improves, these values will be less than zero.
The treatment is defined as successful if the pain reduction after 6 months is over 25%. If the patient is lost to follow-up, the last measurement will be carried forward. If the patient has obtained a different treatment, the subject is classified as unsuccessful.
Function and quality of life
The function and satisfaction will be measured using the AOFAS foot questionnaire [18, 19], the Foot Function Index (FFI) [20–22] and WHOQOL quality of life questionnaire [23, 24] questionnaires.
All patients will be followed up at 4, 8, 12, 26 and 52 weeks. All patients will complete AOFAS and VAS scores at all follow up moments
Determination of sample size
This study has two main parameters: number of patients with 25% pain reduction after 6 months and the average pain reduction;
Corticosteroids and stretching have a recurrence of pain after 6 months [2, 25]. Previous studies with platelet rich plasma by tendonitis showed that the platelet rich plasma group has no recurrence. These results showed 63 percent success rate (VAS reduction over 25 percent after 6 months) for the platelet rich plasma group compared to 27 percent success rate for the corticosteroid group. If these percentages are comparable for plantar fasciitis 38 patients are needed in all groups to achieve a 25% pain reduction (alfa 0,05 and power 0,9).
Conservative treatment with taping or stretching shows a 30 ± 24 points decrease in VAS pain at walking [26, 27]. Corticosteroid injections have shown to have comparable results [2, 25]. 25% increase in pain reduction is seen as clinical relevant. To prove a difference in pain reduction of 15 ± 24 points and a α of 0,05 and a power of 0,9, 55 patients are needed in each group.
Needed number of patients
To be able to measure a difference in average pain reduction, the largest number of patient is needed 55 patients in both groups. To compensate for possible lost to follow-up, 60 patients will be included in each group (Figure 1).
This is a double blind randomized and prospective study. Patients will be randomized in equal proportions to either autologous platelet concentrate or corticosteroid injections. The primary endpoint of the trial is treatment success at 6 months. The other measurements are to measure the speed of recovery and the re-occurrence of pain.
Statistical analyses will test the null hypotheses of no differences between patients treated with autologous platelet concentrate and those treated with corticosteroid in the proportion of the protocol defined, successfully treated patients.
All tests of treatment effect will be made using a one-sided alpha level of 0.025, unless otherwise specified. Tests for the interaction between treatment and a blocking factor will be made at the alpha level of 0.10. All p-values will be rounded then quoted to 3 decimal places, unless the true p-value is less than 0.001 whereby the notation (p < .001) will be used or if the true p-value is greater than 0.999, in which case the notation (p > .999) will be used. Confidence intervals, where indicated for the difference between the two treatment groups, will be calculated at the 95% level. All data will be analyzed by a blinded researcher.