Download as pdf or txt
Download as pdf or txt
You are on page 1of 2

Editorial

Br J Sports Med: first published as 10.1136/bjsports-2019-101250 on 30 January 2020. Downloaded from http://bjsm.bmj.com/ on January 30, 2020 at Agence Bibliographique de l
Corticosteroid injections: glass half-­full, (PRP) to cortisone injections for tennis
elbow, the PRP injections have gener-
ally provided superior outcomes to CSI
half-­empty or full then empty? at 6–12 months6. Should a study such as
this conclude that blood product injection
John W Orchard  ‍ ‍ provides long-­term benefit or that corti-
sone injection leads to long-­term harm?
Although there are relatively few placebo-­
controlled RCTs that compare injected
Cortisone is the most common class of cortisone to injected blood products, the
injection in musculoskeletal medicine.
Box 1 Typic
usual pattern is superior long-­term results
Expert guidelines, such as the 2017 edition for the blood products.7
‘Patients with xxx are frequently treated
of Therapeutic Guidelines: Rheumatology
with corticosteroid injections…. RCTs
(Australia),1 generally recommend corti- Cortisone injections for knee OA
were identified …Almost all studies
costeroid injections (CSI) as a treatment I began with 2002 data relating corti-
had poor internal validity scores. For
for most tendon and joint pain conditions sone injections and tennis elbow because
short-­term outcomes (<or=6 weeks),
(including those listed in table 1). I quote, we may be able to learn from history
statistically significant and clinically
‘local corticosteroid injections should only when we consider cortisone injections
relevant differences were found for
be given by, or under the supervision of, and knee osteoarthritis (OA). After
corticosteroid injection compared to
clinicians with appropriate training and many low-­ quality studies suggested a
placebo. For intermediate (6 weeks-6
experience’.1 I find it remarkable that the short-­term benefit of cortisone injec-
months) and long-­term outcomes (>or=
guidelines do not spell out that cortisone tion in knee OA,8 a recent high-­ quality
6 months), no statistically significant

Enseignement Superieur (ABES). Protected by copyright.


injections should only be given where well-­powered placebo-­controlled RCT
or clinically relevant results in favour
high-­quality evidence shows that benefits showed no clinical difference in outcomes
of corticosteroid injections were found.
outweigh the harms; that is, the wide- between cortisone injection and placebo
More, better designed, conducted and
spread recommendations to use CSI for (saline) injection at any stage over the
reported RCTs with intermediate and
multiple joints and multiple tendons are 2-­year follow-­up.9 There was also a small
long-­term follow-­up are needed’.
‘eminence-­based’ (who should give them) (−0.1 mm) statistically significant addi-
rather than ‘evidence-­ based’ (when they tional deterioration in articular cartilage
should be given). as assessed by imaging in the cortisone
To be fair, I see that ‘appropriate to 12 months3–5. Cortisone injections injection group.9 In a propensity-­matched
training’ could cover ‘knowing the indica- provide short-­term pain relief but worsen cohort study,10 compared with non-­users,
tions’ for a treatment but at a time where the condition in the longer term compared HRs of joint structure worsening over 48
medical overtreatment is incontrovert- with no treatment or placebo treatment. months after cortisone injection was 3.02
ible—it exists and no one denies it exists—I The no-­ treatment or placebo groups (95% CI 2.19 to 4.16). Only 5% of control
argue that the case for ‘indication-­based improve in the longer term, whereas patients required knee replacement during
treatment’ should be explicit. cortisone causes temporary pain relief but follow-­up, compared with 22% of knees
seems to blunt the recovery associated in patients who had received cortisone
Do cortisone injections harm with the no treatment/placebo. injections.10
tendons? In studies that compared injected blood Although cortisone injections may alle-
Systematic reviews have examined products such as platelet-­ rich plasma viate symptoms (in the short term) for
whether cortisone injections are beneficial
or harmful. A typical systematic review
concludes as per the text in box 1. This Table 1  Comparison of short-­term (mainly benefit) versus long-­term effects (possible harm) of
conclusion was from a 2002 systematic corticosteroid injection for common indications
review on the use of CSI for tennis elbow,2
Injection indication Short-­term effects Long-­term effects
and there is strong similarity between
that conclusion and more recent ones for Tennis elbow (tendon) Definite benefit (multiple RCTs) Definite harm (vs placebo/conservative in multiple
up to 4–6 weeks RCTs) at 6 months
different conditions.
With respect to tennis elbow, research Knee OA (joint) Possible benefit up to 4–6 Likely harm (vs non-­use or placebo) over 2–4 years,
weeks cohort and some RCT studies
since 20022 has been decisive: five high-­
Rotator cuff tendons Possible short-­term benefit No long-­term benefit, insufficient data to conclude
quality randomised controlled trials
on long-­term harm
(RCTs) show that cortisone injection
Plantar fasciitis (tendon-­like) Possible short-­term benefit No long-­term benefit, insufficient data to conclude
outcomes are superior to placebo injec-
on long-­term harm
tion for tennis elbow in the short-­ term
Gluteal tendon insertions/ Possible short-­term benefit Unlikely to be a long-­term benefit, but not enough
(2–6 weeks) and that injection results are trochanteric bursitis data to conclude on long-­term harm
inferior compared with placebo injec- Hip OA (joint) Possible short-­term benefit No long-­term benefit, but emerging data suggest
tion or conservative treatment from 3 possible long-­term harm (eg, accelerated OA)11
De Quervain’s tenosynovitis Possible short-­term benefit No long-­term benefit, insufficient data to conclude
School of Public Health, University of Sydney, Sydney, (tendon-­like) on long-­term harm
New South Wales, Australia Achilles tendinopathy Possible short-­term benefit No long-­term benefit, lack of long- term safety
Correspondence to Dr John W Orchard, School of data but possible association with tendon rupture
Public Health, University of Sydney, Sydney, NSW 2006, Green shading indicates benefit; red shading indicates harm.
Australia; ​john.​orchard@​sydney.​edu.​au RCT, randomised controlled trial.

Orchard JW. Br J Sports Med Month 2020 Vol 0 No 0    1


Editorial

Br J Sports Med: first published as 10.1136/bjsports-2019-101250 on 30 January 2020. Downloaded from http://bjsm.bmj.com/ on January 30, 2020 at Agence Bibliographique de l
patients awaiting Total Knee Replacement effects of CSI, for most indications there References
(TKR), the strongest cohort study to date is no demonstrated long-­ term benefit, 1 Therapeutic Guidelines: Rheumatology. Version 3.
shows that a patient may progress more and there are insufficient data to be able Melbourne: Therapeutic Guidelines Limited; 2017.
331 pages, 2017. Available: https://​tgldcdp.​tg.​org.​au/​
rapidly to knee replacement. Although to adequately assess long-­ term harm/ viewTopic?​topicfile=​limb-c​ onditions&​guidelineName=​
worse outcomes in the cohort study were risks. The medium-­ term and long-­ term Rheumatology#​toc_​d1e415
seen even with a single injection, repeated outcomes are now available for tennis 2 Smidt N, Assendelft WJJ, van der Windt DAWM, et al.
cortisone injections were associated with elbow and knee OA, and I believe corti- Corticosteroid injections for lateral epicondylitis: a
greater deterioration than a single injec- systematic review. Pain 2002;96:23–40.
sone injections should not be prescribed
3 Bisset L, Beller E, Jull G, et al. Mobilisation with
tion. A 2019 case series revealed that the for these two conditions as long-­ term movement and exercise, corticosteroid injection, or
risk of medium-­term to long-­term compli- results are consistently not as good as wait and see for tennis elbow: randomised trial. BMJ
cations for both knee and hip OAs treated comparators. 2006;333:939.
with cortisone injections may be quite The question remains: can we clinicians 4 Coombes BK, Bisset L, Brooks P, et al. Effect of
high.11 assume that CSI is benign and that there
corticosteroid injection, physiotherapy, or both on
clinical outcomes in patients with unilateral lateral
is no long-­term deleterious effect for the Epicondylalgia. JAMA 2013;309:461–9.
Systemic cortisone: short-term majority of indications? It might depend 5 Olaussen M, Holmedal Øystein, Mdala I, et al.
athletic performance versus on whether you have a glass-­half full or Corticosteroid or placebo injection combined with
long-term catabolic effects deep transverse friction massage, Mills manipulation,
a glass-­half empty mentality, or whether,
stretching and eccentric exercise for acute lateral
The systemic effects of cortisone roughly after looking at the data, it looks more epicondylitis: a randomised, controlled trial. BMC
follow a similar trajectory, with short-­ like the glass starts off full and ends up Musculoskelet Disord 2015;16:122.
term benefit followed, perhaps, by long-­ empty—the patient ends up worse off. 6 Branson R, Naidu K, du Toit C, et al. Comparison
term risks. In the short-­ term, cortisone of corticosteroid, autologous blood or sclerosant
in high doses is a performance-­enhancing Twitter John W Orchard @DrJohnOrchard injections for chronic tennis elbow. J Sci Med Sport

Enseignement Superieur (ABES). Protected by copyright.


2017;20:528–33.
substance, with five out of seven RCTs Contributors  Single author paper. 7 Fitzpatrick J, Bulsara MK, O’Donnell J, et al. Leucocyte-­
placebo-­controlled athletic trials showing Rich platelet-­rich plasma treatment of gluteus Medius
Funding  The author has not declared a specific grant
improved performance (mainly time to for this research from any funding agency in the public, and Minimus tendinopathy: a double-­blind randomized
exhaustion).12 In the longer term, there commercial or not-­for-­profit sectors. controlled trial with 2-­year follow-­up. Am J Sports Med
is an association of oral corticosteroid use Competing interests  JWO is a Sport and Exercise 2019;47:1130–7.
8 Juni P, Hari R, Rutjes AW, et al. Intra-­Articular
with increased risk of sepsis, deep venous Medicine Physician affiliated with Australasian College corticosteroid for knee osteoarthritis. Cochrane
thrombosis (DVT)and fractures.13 This is of Sport and Exercise Medicine Physicians. Database Syst Rev 2015;10:Cd005328.
consistent with what every medical student Patient consent for publication  Not required. 9 McAlindon TE, LaValley MP, Harvey WF, et al. Effect of
knows about Cushing’s syndrome: that intra-­articular triamcinolone vs saline on knee cartilage
Provenance and peer review  Not commissioned; volume and pain in patients with knee osteoarthritis.
excess of corticosteroid eventually leads externally peer reviewed.
JAMA 2017;317:1967–75.
to catabolic effects (tissue weakening) in
© Author(s) (or their employer(s)) 2020. No commercial 10 Zeng C, Lane NE, Hunter DJ, et al. Intra-­Articular
multiple systems, including muscle, bone re-­use. See rights and permissions. Published by BMJ. corticosteroids and the risk of knee osteoarthritis
and joints in the musculoskeletal system. progression: results from the osteoarthritis initiative.
A picture emerges (table 1) of consistent Osteoarthritis Cartilage 2019;27:855–62.
likely short-­term benefit for CSI perhaps 11 Kompel AJ, Roemer FW, Murakami AM, et al. Intra-­
Articular corticosteroid injections in the hip and
for most musculoskeletal conditions, To cite Orchard JW. Br J Sports Med Epub ahead of
knee: perhaps not as safe as we thought? Radiology
although it should be noted that even print: [please include Day Month Year]. doi:10.1136/ 2019;293:656–63.
bjsports-2019-101250
many RCTs are at high-­risk of bias. This 12 Tacey A, Parker L, Garnham A, et al. The effect of
means that some of the reported benefits Accepted 15 January 2020 acute and short term glucocorticoid administration
from CSI in RCTs could be from placebo/ Br J Sports Med 2020;0:1–2. on exercise capacity and metabolism. J Sci Med Sport
2017;20:543–8.
non-­specific effects if the trial design was doi:10.1136/bjsports-2019-101250
13 Waljee AK, Rogers MAM, Lin P, et al. Short term use of
not able to adequately blind participants ORCID iD oral corticosteroids and related harms among adults in
or assessors from group allocation, for John W Orchard http://​orcid.​org/​0000-​0003-​3530-​ the United States: population based cohort study. BMJ
example. With respect to the long-­ term 1711 2017;357:j1415–j15.

2 Orchard JW. Br J Sports Med Month 2020 Vol 0 No 0

You might also like