Therapeutic use of GCs in sports

GCs are widely used in sports medicine, mainly to treat musculoskeletal injuries and asthma. GCs administered by local injection can be an important and effective option in the treatment plan of a number of musculoskeletal conditions commonly encountered in sports medicine. For example, intra-articular GC injections are used to treat adhesive capsulitis, where it has been shown that a single GC injection provides faster pain relief and earlier improvement of shoulder function and motion compared with oral NSAIDs, and that they are effective for both short-term and long-term pain relief.23 24 Intra-articular GC injections are also commonly used for the treatment of bursitis in athletes.

Local GC injections also provide short-term pain relief and improvement in function in subacromial impingement syndrome, a condition commonly observed in overhead sports, where athletes perform frequent and quick upper limb actions causing alterations of scapular kinematics.25

Health risks associated with GC use

The proven or potential risk to health of the athlete is one of three criteria established by the Code to assess when determining whether a substance is prohibited. Despite their excellent therapeutic effects, GCs have a relatively high potential for triggering adverse effects, particularly when administered systemically.26 The potential for harmful effects ultimately depends on the duration of treatment, dose, route of administration and patient-specific factors.27 28 For these reasons, if athletes are taking GCs with the intention to improve performance, shorter periods of administration are likely to be preferred due to a lesser incidence of adverse effects such as muscle wasting, and without the need for tapering the dose at the end of the course.29 30

When administered as a local injection by the intra-articular or periarticular route, systemic absorption occurs to some extent due to the high vascularity of the joint and diffusion of the drug into the surrounding tissue after injection. This can lead to adverse effects including flushing of the face, blood glucose increase in diabetics, increase in blood pressure or palpitations, transient immunosuppression, and, rarely, water retention and intermittent gynaecological bleeding.

Locally occurring adverse effects after intra-articular GC injections are rare but may include reactive synovitis, local irritation and postinjection pain, septic arthritis, cutaneous depigmentation, tendinopathy and avascular necrosis.31–34 Adverse effects following local infiltrations into periarticular structures, tendon sheaths or bursae are also rare, but may include atrophy of the subcutaneous tissue and cutaneous depigmentation.

Mechanism of the performance-enhancing effect of GCs

During exercise, there is a strong stimulation of the hypothalamic-pituitary-adrenal (HPA) axis, which results in a rise in baseline blood cortisol concentration, according to the intensity and the duration of the activity.35 36 Administration of oral GCs has been shown to inhibit the HPA axis, resulting in a strong decrease in baseline blood cortisol concentration and an inhibition of exercise-induced cortisol secretion.37–42 Similar effects were not observed after administration via the inhaled route.43 44

The response of cortisol to endurance exercise results from both the central and peripheral actions of GCs by direct and/or indirect pathways, which can lead to effects including euphoria and decrease in fatigue. In addition, GCs increase energy mobilisation via proteolysis and gluconeogenesis while maintaining the blood glucose level, and have a marked effect on reducing inflammation.45 Any ergogenic effect produced by GCs probably includes a local muscular effect of the drug rather than any direct link to the anti-inflammatory response, but the complexity of the central and peripheral response after GC administration makes the exact causal effects of PE difficult to identify.

Studies of GC effects on performance

A number of recent studies show that the impact of GCs on athletic performance depends on the intensity and duration of exercise, the dose and duration of treatment and the route of administration.38–44 46–49 Results of all known studies looking at the effects of GCs on exercise performance are presented in table 1. Only the inhaled and oral routes of administration for GCs have been studied to assess their effect on PE, with evidence of PE only after oral administration.

A significant performance-enhancing effect has been demonstrated after administration of oral prednisolone (PRED) and prednisone (PSONE) at 50–60 mg/day over a 7-day period, respectively, in female and male recreationally trained subjects, during prolonged endurance exercise lasting more than 40 min.38–40 A small improvement in performance was also observed after administration of PSONE 60 mg/day over a 7-day period during repeated exercise leading to muscle fatigue and exhaustion.48

However, no studies have demonstrated an improvement in performance following acute systemic GC administration (such as a single dose of 20 mg of PRED) at any intensity of exercise,41 42 or which examine the effect on performance by combining different GCs by any route of administration. In addition, an impaired cycling performance has been demonstrated 12 hours after acute PRED administration, suggesting that any ergogenic effect obtained in the 2–3 hours after initial administration may be reversed with GC elimination.49

A significant PE effect has been confirmed after oral dexamethasone (DEX) administration of 4 mg/day over a 5-day period during knee extensor and maximal graded exercises,46 47 and a small improvement in performance observed with the same doses during repeated exercise leading to muscle fatigue and exhaustion.47 However, no PE was observed at lower doses of oral DEX of 1–3 mg/day over 4.5 days.46 47

No PE has been demonstrated after inhaled GCs have been administered. Although only inhaled budesonide has been studied, this finding is expected to apply to all inhaled GCs at recommended therapeutic doses due to their relatively low systemic exposure compared with oral administration.43 44

These few studies demonstrate that under certain conditions and routes of administration GCs can enhance sport performance, while under other conditions no PE effect could be measured. It is conceivable that in certain circumstances, other GCs in other doses or other sports could also potentially influence performance in a positive way. This concept of potential PE is the basis for defining the status of these drugs on the Prohibited List.

Novel approach for defining permitted and prohibited GC use

The naturally occurring GC hormones and their synthetic analogues possess a wide range of potencies and pharmacokinetic properties.50 The body produces a daily output of endogenous GC (cortisol), typically 18–22 mg/day51 and hence this amount of GC should be regarded as the normal physiological baseline for cortisol levels in the body. The upper end of this basal physiological range can be defined as 26.4 mg/day which is 20% above the upper typical output of 22 mg/day. The supraphysiological threshold can be defined as >32 mg/day which is >20% above the 26.4 mg/day upper basal physiological range. Although there is no reference range for daily cortisol production, these ranges are similar to that seen for morning serum cortisol levels. The 20% margins allow for additional variability recognising that less than a 20% difference in exposure is not usually considered clinically relevant compared with expected biological variability.

Administering GC drugs can result in a total GC exposure (exogenous + endogenous) that exceeds the supraphysiological threshold for daily physiological GC exposure when expressed in cortisol equivalents. Based on the studies presented above it is generally accepted that GCs have PE potential, but it is not known exactly at what level total GC exposure is required to elicit a meaningful PE effect as there are insufficient data to construct a dose-response relationship. However, like other pharmacological actions of GCs, PE is not expected to be an all-or-nothing effect but rather gradually increase with increasing GC exposure, following a non-linear sigmoid dose-response curve. On this basis, PE, even though initially very small, may occur once the supraphysiological threshold for daily physiological GC exposure is exceeded. Whereas GC drug use that does not exceed the supraphysiological threshold of 32 mg/day, in cortisol equivalents, can reasonably be regarded as not having significant PE potential. This gradual onset of effect is represented by the gradual shading from green to red in figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Physiological, supraphysiological and GC exposure with low risk and high risk of performance-enhancing effects in mg/day shown in cortisol equivalents. Column 1 shows the physiologically normal basal cortisol daily production rate typically 18–22 mg/day. Column 2 shows the upper end of basal physiological range defined as 26.4 mg/day which is 20% above 22 mg/day. Column 3 depicts the supraphysiological range of >32 mg/day which is defined as >20% above the 26.4 mg/day upper basal physiological range; exceeding this represents a continuum for increasing the risk for performance-enhancing and adverse health effects as shown in columns 4 and 5. The 20% margins (indicated with an asterix * in column 4) allow for additional variability recognising that a 20% difference in exposure is not usually clinically relevant compared with biological response variability. Column 4 shows the impact of acute administration of exogenous GC to upper basal cortisol (26.4 mg/day) exceeding the supraphysiological range of >32 mg/day in the acute administration scenario (eg, oral dexamethasone 0.65 mg, oral prednisolone 3.3 mg, intra-articular triamcinolone acetonide 8.5 mg). Column 5 shows the impact of the same exogenous GC doses shown in column 4 but in a chronic administration scenario where basal endogenous GC levels are suppressed by chronic administration of high doses of exogenous GC. GC, glucocorticoid.

For any exogenous GC dose or route of administration, where the estimated systemic exposure exceeds these limits, there is an increased risk of inducing PE effects (figures 1 and 2). Using this approach, two categories were defined:

1. Low-risk GC use

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Relationship between approved therapeutic dose ranges (blue bars) for various GC drugs and their corresponding doses estimated to either produce systemic exposures above the upper physiological range for GC exposure (green bars) or estimated to have a high risk of performance-enhancing effects (red bars) during acute administration.* *For the intra-articular route of administration different extents of systemic absorption in 24 hours were assumed where F is the fraction of the dose absorbed. GC, glucocorticoid.

Low-risk use is defined as a dose of exogenous GC (cortisol equivalent dose ≤5.28 mg/day) that when added to normal physiological daily GC exposure (26.4 mg/day) results in a total exposure (endogenous + exogenous) that does not reach supraphysiological levels (cortisol equivalent dose >32 mg/day) (figure 1). In terms of adverse effects on health the extent of HPA-axis suppression is expected to result in <50% cortisol suppression.

2. High-risk GC use

High-risk use is defined as a dose of exogenous GC that is equivalent to or greater than a dose demonstrated to be performance enhancing (eg, DEX 4 mg, equivalent to 32.6 mg cortisol) (figure 1). This is also ≈6 times the 5.28 mg/day acceptable dose defined above and results in a total GC exposure (26.4 mg endogenous + 32.6 mg exogenous) of ≈60 mg/day. In terms of adverse effects on health the extent of HPA-axis suppression is expected to result in ≥80%–90% cortisol suppression.

Between these two categories, there is an intermediate risk of adverse health effects but there are insufficient data to experimentally derive the potential for PE. However, the level of risk is anticipated to be on a continuous dose-response scale, starting with little or no PE effects rising to a high and known PE risk. Despite these limitations, this framework still allows clear guidance on acceptable low-risk use versus unacceptable high-risk use since many of the widely used GC doses and formulations fall into one of these categories.

Applying the approach to GC doses with known performance-enhancing effects

Clinical data confirming PE are only available for 4 mg oral DEX, and 50 mg and 60 mg oral PRED when given over 5–7 days, based on published data as described above.38–40 46 47 Using the methodology described in this article (see below) the corresponding cortisol equivalent doses were estimated to correspond to 32.6 mg, 80 mg and 96 mg, respectively (figure 1), all of which exceed the 32 mg/day threshold set to determine low-risk versus high-risk use of a GC drug in sport. These results validate that the new framework to determine acceptable or unacceptable GC use in sport applies appropriately to drugs and doses with known PE effects and corroborates that their use is high risk and therefore unacceptable in sport.

Methodology for determining cortisol equivalent doses

For any GC the dose can be expressed in terms of a cortisol-equivalent dose and thereby the dose which exceeds 32 mg/day total cortisol-equivalent exposure for any GC and route of administration can be determined. This approach was applied to define the GC doses and routes of administration that should be prohibited, or not prohibited in sport.

To make this assessment it is first necessary to convert the administered exogenous GC dose into a cortisol-equivalent dose (figure 1). This was calculated as follows:52

Equation 1: Cortisol equivalent dose = CL_cort . (P_r . ((F_GC . Dose_GC) / CL_GC))),

where, F=the bioavailability of the systemically absorbed fraction of the dose for the route of delivery and formulation; P_r=relative GC potency in terms of the GC-receptor binding or GC activity (cortisol=1); CL=the rate at which active GC is cleared from the body via metabolism and/or excretion; CL_cort is the plasma clearance of cortisol (L/h); and CL_GC is the plasma clearance of the exogenous GC (L/h).50

These parameters are mostly available for commonly used GC formulations and routes of administration.50 52–77 Although the bioavailability for some less common topical delivery routes (eg, skin, eye, ear) and older molecules is not known, this was not a significant issue since even when hypothesising 100% bioavailability the upper physiological exposure threshold was not exceeded within the approved therapeutic dose ranges (figure 2). Similarly, for parenteral injections other than intravenous, for example, intradermal, periarticular and intralesional, bioavailability was assumed to be 100%.

For intra-articular injections the situation is more complex since they can be absorbed slowly from the injection site. Although 100% bioavailability may eventually be achieved, we estimated the fraction absorbed in each 24-hour period postdose from the absorption half-life which is known in most cases. A sensitivity analysis was also performed using different bioavailability scenarios (figure 2).

The approved dose ranges for all the commonly used GC drugs, formulations and routes of administration78 were converted into cortisol equivalents using equation 1. These were compared with the cortisol-equivalent doses corresponding to normal and supraphysiological GC thresholds (figure 1). These values are not based on circulating cortisol concentrations, which can vary widely throughout the day due to the pulsatile nature of cortisol secretion, but nevertheless they still reflect the variability seen in morning serum cortisol concentrations (5–25 µg/dL) that are used in clinical chemistry.

Higher or lower cortisol production or secretion rates and circulating concentrations can occur during stress, disease or exercise. Consequently, some athletes may have elevated circulating cortisol levels during exercise,79 80 but there is no evidence that this alters the daily cortisol production rate and therefore was not considered relevant to the argument for using a single supraphysiological threshold of >32 mg/day for total GC exposure (endogenous + exogenous). This is a conservative position, because if an athlete generates increased cortisol levels during extreme exercise and has higher endogenous GC exposure, taking exogenous GC will only result in a higher risk of PE.

Consideration of cortisol suppression

We also considered that cortisol suppression can occur following exogenous GC administration. With acute and single doses, this is not important because although the body immediately starts to reduce cortisol production/secretion the amount of cortisol already present in the body is not immediately removed and hence is added to exogenous dose (figure 2). With chronic administration cortisol suppression is relevant, but not in the range that has a high risk of PE because the contribution of the endogenous cortisol to the total GC exposure is very small (figure 2). For example, chronic administration of 20 mg/day PRED or 4 mg/day of DEX results in 80%–90% cortisol suppression and the 32 mg/day total GC threshold is still exceeded with 20 mg/day PRED and 4 mg/day DEX despite cortisol suppression (figure 2).

Results of the application of the new methodology

Oral and injectable routes (eg, intravenous, intramuscular, subcutaneous, intra-articular) when used at their approved doses are likely to produce total GC exposures in excess of 32 mg/day except at the lowest doses that are seldom used clinically (figure 1). For example, doses of oral PRED (3.3 mg) and oral DEX (0.65 mg) and intra-articular triamcinolone acetonide (TA, 1.4 mg) are estimated to produce total GC exposures at the threshold of 32 mg/day (figure 2).

However, none of the inhaled, intranasal, dermal or other topical GCs, when used at their maximum licensed approved doses, would exceed the 32 mg/day threshold.

Using this new methodology, the dose that corresponds with unacceptable use in sport was determined for most routes of known licensed GCs. This formed the basis by which the revised specific RLs were set, which subsequently guided the establishment of washout periods.

Establishing improved urinary RLs for GC use

Up to now, a general urinary RL of 30 ng/mL was used to differentiate between permitted and prohibited administrations of all GCs. However, different studies have shown the need to establish compound-specific RLs given the diversity of administration routes and doses, as well as pharmacokinetic and pharmacodynamic properties between the different GCs.1–20

The ideal RL should not produce AAFs after permitted administrations or negative results after prohibited use. When testing athletes, the potential for AAFs after permitted administrations for therapeutic purposes have to be precluded, therefore the RL must be based on concentrations obtained after permitted administrations.

The compound-specific RL needs to be based on urinary concentrations obtained after administration studies and so the RLs for GCs for which these data are available were reviewed. All relevant studies are presented in table 2.1–20 For other GCs, with limited or no data available, the RL has been maintained at 30 ng/mL until further data are generated in the future.

In most of the studies, concentrations obtained after enzymatic hydrolysis (free and glucuronide fractions) were described. The distribution of concentrations and the maximum concentrations in urine (Cmax) after permitted administrations, obtained after the maximum daily doses recommended by manufacturers, were the most important data taken into consideration to define the compound-specific RLs. When studies at those doses were not available or the number of volunteers was low, a conservative value was applied to the highest individual urinary concentration. The proposed RLs are summarised in table 3.

Regarding the marker of discrimination, the parent drug was suitable for the majority of GCs studied,5 6 13 15 17 except for budesonide and triamcinolone hexacetonide (THA). For budesonide, the metabolite 6β-hydroxy-budesonide (6β-OHBUD) showed better balance between specificity and sensitivity compared with the parent drug or other metabolites.2 3 THA, a pro-drug of TA, was not detected in urine after administration and the active drug was selected as the marker of discrimination.7

For all GCs, urinary concentrations after dermatological applications were very low (table 2) and, in general, data obtained after inhaled or intranasal administrations were used to define the RLs.

For 6β-OHBUD, data after intranasal and inhaled administrations were available (table 2). Concentrations after intranasal administration were very low. However, concentrations obtained after inhalation steered the decision to increase the RL to 45 ng/mL to reduce the possibility of an AAF after inhalation at high doses.

For TA, dermatological and intranasal administrations were studied. After the maximum daily intranasal dose, the mean and median concentrations were very low, however a Cmax of 7 ng/mL was obtained. Due to these results and because concentrations of TA after intramuscular administrations were very low, especially after low doses,5 6 a reduction of the RL to 15 ng/mL was proposed to increase the sensitivity of detection after intramuscular use.

For BET, dermatological and intranasal studies were available. The Cmax obtained after repeated intranasal administration at doses below the maximum recommended daily doses suggested an increase of the RL to 60 ng/mL to avoid an AAF after higher intranasal doses. For DEX, the same criterion was proposed due to the similar pharmacological and pharmacokinetic properties with BET.

For PRED and PSONE, dermatological, intraocular and intranasal studies were available. The Cmax obtained after repeated intraocular or intranasal administration using doses below the maximum daily dose suggested an increase of the RL of PRED and PSONE to 100 ng/mL and 300 ng/mL, respectively, to avoid an AAF after these administration routes at higher doses.

For methylprednisolone (MP), only studies after dermatological application were available. Despite the low concentrations obtained after dermatological application, the RL was maintained at 30 ng/mL because the sensitivities after oral, intramuscular and intra-articular administrations and the washout periods did not significantly change using lower RLs.

Establishing washout periods following administration of GCs

After GCs administration, urinary concentrations which could result in an AAF can be reached for different periods of time after administration depending on the GC administered and the dose.

Guidance relating to clinical use according to the manufacturer’s licensed doses was deemed necessary to reduce the risk of AAF after medical use during an out-of-competition period and, therefore, washout periods for oral and injectable routes were defined. The washout period refers to the time taken from the last dose to the time of the in-competition period (midnight beginning the evening of the in-competition period) in order to reduce the GC concentration in the urine below the RLs. The washout periods were established taking into account the time of detection of the drugs using the new RLs established for each. The maximum detection times obtained in each study are indicated in table 2.

It is worth highlighting the long detection times obtained after intramuscular use of TA (table 2). For triamcinolone, the recommended washout period is 10 days to ensure specificity after the maximum oral dose (table 2).

In order to simplify the recommendations and facilitate the therapeutic use of GCs in out-of-competition periods, the washout periods were unified by administration routes and the recommended values are described in table 3.