Ultrasound guided injection of dexamethasone versus placebo for treatment of plantar fasciitis: protocol for a randomised controlled trial

The La Trobe University Human Ethics Committee has approved the trial (Application Number: 09-062) and all participants will provide written informed consent prior to enrolment. Ethical standards will adhere to the National Health and Medical Research Council (NHMRC) National Statement [25] and the World Medical Association's Declaration of Helsinki [26]. Publications associated with the trial will be formatted according to the CONSORT Statement [27].

The trial will be conducted at the La Trobe University Podiatry Clinic and will recruit local community-dwelling participants by multiple newspaper advertisements. Participants will be required to have a clinical history of pain beneath the heel for at least eight weeks prior to enrolment, and to report a minimum heel pain magnitude of 20 mm on a 100 mm visual analogue pain scale. On clinical examination, participants will also be required to report sensitivity to palpation of the medial calcaneal tubercle and/or the proximal plantar fascia. To confirm the diagnosis of plantar fasciitis, the dorso-plantar thickness of the plantar fascia will be measured by ultrasound at a standard location where the fascia crosses the anterior aspect of the inferior calcaneal border. Participants will be required to have a plantar fascia thickness value of 4.0 mm or greater [28]. Finally, participants will be required to attend the second visit (during which a tibial nerve block and heel injection will be given) with a friend or family member who is able to provide transportation from the clinic.

Applicants will be excluded from the study if they have received a corticosteroid injection for plantar fasciitis within the previous six months, or if they have any of the following: a known hypersensitivity to lignocaine hydrochloride or corticosteroids, current skin or soft tissue infection near the injection site, posterior heel pain, current pregnancy, systemic inflammatory disease, diabetes mellitus, previous local surgery, or a history of local trauma. Applicants will also be excluded if they are unable to walk household distances without the use of an aid, or if they have commenced any treatment regimen for plantar fasciitis within four weeks prior to enrolment.

Screening of applicants according to these criteria will occur by a preliminary telephone interview, followed by clinical examination at the initial visit. After a detailed explanation of the study protocol, eligible applicants will be invited to participate in the trial. However, prior to enrolment, all applicants will be assessed for competence to give consent by use of the Evaluation to Sign Consent (ESC) tool [29]. The ESC is specifically designed for use in clinical research and has been shown to have good inter-rater reliability (Pearson r = 0.81) [30]. A range of descriptive characteristics will also be collected at the initial visit after enrolment (Table 1). Participants will then be scheduled to attend a second appointment (approximately one week later) where baseline measurements will be taken and the trial intervention performed (Figure 1).

Participants will be randomly allocated to one of two treatment arms: (i) ultrasound-guided injection of the plantar fascia with corticosteroid (experimental group), or (ii) ultrasound-guided injection of the plantar fascia with normal saline (control group). Prior to plantar fascia injection, participants in both groups will be given an ultrasound-guided posterior tibial nerve block with 2% lignocaine hydrochloride. All participants will be positioned prone on a treatment table with their knees extended. Corticosteroid injection will be performed with a 25 gauge (38 mm) needle and a 1 mL Luer-lock syringe containing 1 mL of 4 mg/mL dexamethasone sodium phosphate. Saline injection will be performed with a 25 gauge (38 mm) needle and a 1 mL Luer-lock syringe containing 1 mL normal saline (0.9% sodium chloride). For both injections, the needle will be inserted through the medial heel perpendicular to the long axis of the ultrasound transducer, and will be advanced under continuous guidance into the proximal plantar fascia [31]. Infiltration will occur near the calcaneal enthesis in the region of maximal fascia thickening. An aseptic technique will be used for all injections to minimise infection risk, including use of sterile gloves, sterile transducer covers and sterile transmission gel. Participants with bilateral plantar fasciitis will have both feet treated (with their allocated intervention) during a single appointment. Following treatment, participants will be advised to avoid all running and other high impact activities for at least 2 weeks.

Throughout the initial 8 weeks of enrolment, all participants will be required to complete a daily stretching program shown to decrease pain associated with plantar fasciitis [32]. The stretching technique requires participants to cross their affected leg over the contralateral knee in a seated position, then pull back on the toes until they feel a stretch in the arch of their foot. Participants will be instructed to repeat the stretch 10 times, with each stretch lasting for 10 seconds. All participants will be asked to complete the stretching program three times per day and to record their stretching frequency in a diary.

Treatment allocation will be performed according to a computer-generated randomised number sequence. Allocation will be concealed in a password protected computer file accessible by an investigator not involved in measuring outcomes (ADM). This investigator will also prepare the syringe prior to heel injection, thereby ensuring the investigator giving injections, measuring outcomes and analysing data (AMM) is blinded throughout the trial duration. As both treatment solutions (i.e. dexamethasone sodium phosphate and normal saline) appear in the syringe as clear liquids free from visible particulate matter, the syringe contents will not require masking. This protocol will also ensure that trial participants are blinded to their treatment allocation.

Primary outcomes will be pain and plantar fascia thickness at 4, 8 and 12 weeks. Pain will be measured by the foot pain domain of the Foot Health Status Questionnaire (FHSQ), which has been shown to have a high degree of internal consistency (Cronbach's α = 0.88) and test-retest reliability (intra-class correlation coefficient = 0.86) [33]. Participants treated for bilateral plantar fasciitis will be asked to describe symptoms without specific reference to an individual foot (i.e. bilateral pain will be evaluated as one independent sample). Plantar fascia thickness will be measured by ultrasound at a standard location where the fascia crosses the anterior aspect of the inferior calcaneal border (Figure 2). This measurement technique has been shown to have good intra-rater reliability, with the 95% limits of agreement ranging from -0.7 mm to 0.5 mm [34]. Participants treated for bilateral plantar fasciitis will have thickness measurements taken for each individual foot. However, for bilateral cases, the mean change in plantar fascia thickness for the two feet will be calculated at each follow-up in order to evaluate data as one independent sample. Plantar fascia measurements and ultrasound guided injections will be performed with a variable frequency (5-10 MHz) linear array transducer (Acuson Aspen, Siemens Medical Solutions, Pennsylvania, USA).

Secondary outcomes will be function, use of oral analgesic medication, and 'first-step' pain at 4, 8 and 12 weeks. Function will be measured by the foot function domain of the FHSQ, which has been shown to have a high degree of internal consistency (Cronbach's α = 0.85) and test-retest reliability (intra-class correlation coefficient = 0.92) [33]. Participants treated for bilateral plantar fasciitis will be asked to describe foot function without specific reference to an individual foot (i.e. bilateral function will be evaluated as one independent sample). Use of oral analgesic medication (including paracetamol, aspirin, opioids and non-steroidal anti-inflammatory drugs) will be recorded at each visit. 'First-step' pain, experienced when initially rising from bed in the morning, will be measured on a 100 mm visual analogue scale. Participants treated for bilateral plantar fasciitis will be asked to indicate the magnitude of symptoms without specific reference to an individual foot (i.e. bilateral pain will be evaluated as one independent sample).

Prospective sample size calculation indicates a sample of 40 participants per group (i.e. 80 total) would provide 80% power to detect a minimal important difference of 13 points [35] on the pain domain of the FHSQ (SD = 20, alpha = 0.05, 5% loss to follow-up). The extra precision provided by covariate analysis was conservatively ignored when performing this calculation.

Complications and adverse events associated with the intervention (e.g. infection, nerve injury or plantar fascia rupture) will be recorded in participant files and reported in the final manuscript.

Statistical analyses will be undertaken using SPSS software (version 14.0 or later, SPSS Corporation, Chicago, IL, USA). Continuous data will be explored for normality using standard tests to satisfy the assumptions of parametric statistics. All analyses will be conducted on an intention-to-treat basis and missing follow-up data will be replaced with baseline observations carried forward (BOCF) (i.e. baseline data will be carried forward in circumstances where follow-up observations are missing). In comparison to a last observation carried forward (LOCF) approach, BOCF has been shown to provide a more conservative estimate when analysing the treatment effect of pain-relief medication [36]. Continuous outcomes with a normal distribution will be analysed using a linear regression technique with baseline measurements adjusted for by the analysis of covariance model (ANCOVA) [37]. If data is found to be not normally distributed, transformation will be attempted. However, if data is still not normally distributed after transformation, non-parametric statistical tests will be used. Other data (nominal or ordinal) will be analysed using appropriate non-parametric statistical tests. Statistical significance for hypothesis tests will be set at the conventional level of 0.05.

In the treatment of musculoskeletal disorders, corticosteroid injection is typically used to inhibit synthesis of arachidonic acid from membrane phospholipids, thereby suppressing prostaglandin-mediated inflammation and pain [38]. However, as described previously, histological studies indicate that plantar fasciitis is predominantly a degenerative disorder, with limited involvement of chronic inflammatory processes. Consistent with these findings, absence of inflammation is also reported in the tendinopathy literature [14], and the precise mechanism of pain in tendinopathy remains uncertain [39]. Several alternatives to prostaglandin-mediated pain have been suggested in relation to tendon models, including neurovascular in-growth [40], up-regulation of excitatory neurotransmitters (e.g. substance P, glutamate and acetylcholine) [14, 39‐41], and increased presence of biochemical irritants (e.g. chondroitin sulfate) [39].The action of corticosteroids on these mechanisms is currently unclear [42], however, corticosteroids have been shown to inhibit fibroblast proliferation and expression of ground substance proteins [14, 38]. It is possible that these known effects may be of benefit in the treatment of plantar fasciitis, as increased fibroblast proliferation and excessive secretion of proteoglycans are commonly reported features of the condition. Based on similar reasoning, corticosteroids have been suggested as potentially beneficial for treatment of early-stage tendinopathy [14].

Selection of a particular corticosteroid agent for local injection varies across disciplines [43] and geographic regions [44], with limited evidence available to assist in decision-making. In relation to treatment outcomes, systematic reviews of randomised trial data have revealed no difference in clinical efficacy between various corticosteroid types [45, 46]. Nonetheless, when selecting a corticosteroid for treatment of soft tissue disorders, guidelines in the rheumatology literature recommend use of an agent with high tissue solubility [42], and avoidance of fluorinated compounds [47]. When corticosteroids are injected locally, duration of action is inversely related to solubility, with high solubility agents (phosphates) having a shorter duration of action, and low solubility agents (acetates) a longer duration of action [42, 48]. Accordingly, high solubility corticosteroids (e.g. dexamethasone phosphate) are thought to reduce the risk of post-injection flare and soft -tissue atrophy [49, 50].

Fluorination of a corticosteroid molecule improves anti-inflammatory (glucocorticoid) action and decreases sodium retaining (mineralocorticoid) activity, thereby improving the anti-inflammatory potency of the drug while reducing systemic side effects [51]. Despite these advantages, comparisons of fluorinated versus non-fluorinated corticosteroids have demonstrated various effects by which fluorinated agents increase collagen degradation [52‐54]. In addition, case-series reports of adverse events following corticosteroid injection, such as rupture of the plantar fascia, have largely involved the fluorinated corticosteroid triamcinolone [55, 56]. However, this drug is also the most insoluble (and therefore longest acting) injectable corticosteroid available [48, 57], and adverse events following use of shorter-acting fluorinated agents are less frequently reported. With this in mind, it is likely that injection of a corticosteroid that is both fluorinated and relatively insoluble should be avoided when treating soft tissue disorders.

Accordingly, two corticosteroids were considered for use in this trial: (i) methylprednisolone acetate (non-fluorinated, moderate acting) and (ii) dexamethasone sodium phosphate (fluorinated, shorter acting). However, as normal saline solution appears as a clear and colourless liquid, blinding of the investigator performing the injections was considered unachievable with use of an acetate compound. Therefore, dexamethasone sodium phosphate was considered the most appropriate corticosteroid for use in the trial.

Clinician surveys have revealed that combining corticosteroid and local anesthetic solutions prior to soft tissue injection is a widely adopted practice [19, 44, 58]. Reported benefits of this include provision of temporary pain relief, dilution of potentially harmful corticosteroid crystals (acetates only), and confirmation of accurate solution deposit [59]. Despite this common practice, mixing of corticosteroid solution will not occur in this trial as regional anaesthesia will be performed prior to plantar fascia injections.

Fusiform thickening of the plantar fascia is a well established feature of plantar fasciitis. According to a meta-analysis of diagnostic imaging studies [28], people with plantar heel pain are over 100 times more likely to have an abnormally thickened (> 4.0 mm) plantar fascia compared with asymptomatic controls (odds ratio = 105.11, 95% confidence interval = 3.09 to 3577.28, P = 0.01). Abnormal thickening is also reported in the tendinopathy literature, and is thought to be the result of increased secretion of ground substance proteins (e.g. proteoglycans) and subsequent tissue oedema [14]. These changes are considered a response to acute tensile overload and may lead to reduced tissue stress by increasing cross-sectional area [14]. However, it is also noted that increased thickening is not a feature of normal tendon adaptation to chronic loading regimens [14, 60], thereby suggesting that changes in tendon thickness are related to abnormal tissue substance.

In response to corticosteroid injection, plantar fascia thickness values have been shown to decrease significantly as early as two weeks [24] and one month [61] following treatment. One of these studies also reported a correlation between decreased plantar fascia thickness and pain relief (Pearson r = 0.61, P < 0.001) [61]. However, as neither of these studies made comparisons to a control group, it is possible that the findings were partly due to the condition's natural course. Nonetheless, a recent randomised trial reported a significant decrease in pain and plantar fascia thickness following injection with botulinum toxin, when compared to placebo injection [62]. This evidence suggests that plantar fascia thickness measurements can provide useful objective data, and may assist in identifying overall improvement in the condition.

Studies investigating the clinical benefit of regional anaesthesia given prior to plantar fascia injection have produced inconsistent findings. Two trials report no difference in overall procedure comfort between plantar fascia injections given after tibial nerve blockade, and plantar fascia injections given alone [22, 63]. Explanations for these counterintuitive findings include: (i) pain and paraesthesia commonly occur during a nerve block from contact between the needle tip and nerve fascicle [22, 64] and (ii) successful anaesthesia of the posterior tibial nerve is difficult to achieve, especially within a short period of time [22, 65]. In contrast, one study found that a posterior tibial nerve block effectively reduces pain during the plantar fascia injection itself [66]. Furthermore, in comparison to a landmark-based technique, use of ultrasound guidance during regional anaesthesia has been shown to reduce the occurrence of paraesthesia and inadvertent intravascular injection, while improving block onset time and success rates [64, 67, 68]. With this in mind, we have chosen to perform an ultrasound-guided posterior tibial nerve block prior to plantar fascia injection.

Combination of placebo treatment and double-blinding will be undertaken to control for intervention effects not due to the pharmacological action of corticosteroids. Prescription of a standardised stretching program has been introduced in order to compensate for the presence of a placebo group, thereby ensuring every participant receives treatment for their condition.