Adverse events of intravenous glucocorticoid pulse therapy

in inflammatory diseases: a meta-analysis

N.A.M. Smits, N. Duru, J.W.J. Bijlsma, J.W.G. Jacobs

Department of Rheumatology and Clinical ABSTRACT and obstructive lung diseases (1). GCs
Immunology, University Medical Center Objective. To systematically analyse can be given via different routes includ-
Utrecht, Utrecht, The Netherlands. the literature on reported adverse ing intravenous, oral, intramuscular,
Noortje A.M. Smits, BSc events (AEs) of intravenous pulse glu- intra-articular and transdermal routes,
Nurten Duru, MD cocorticoids (GCs) (≥250 mg pred- and by inhalation. The dose and dura-
J.W.J. Bijlsma, MD, PhD, Professor
nisone equivalent) for inflammatory tion of therapy range from a high dose
J.W.G. Jacobs, MD, PhD, Assoc. Prof.
diseases. for short-term to a low dose for long-
Please address correspondence to:
J.W.G. Jacobs, MD, PhD,
Methods. A literature search was done term therapy. Characteristic schemes of
Department of Rheumatology and using PubMed, Embase, and Cochrane GC therapy are high dosages of ≥30 mg
Clinical Immunology, F02.127, databases. Studies were selected by prednisone equivalent daily, medium
University Medical Center Utrecht, two reviewers (NAMS and ND). Avail- dosages of 7.5–30 mg, low dosages of
Box 85500, 3508 GA Utrecht, able data on the prevalence of GC-re- ≤7.5 mg prednisone equivalent daily,
The Netherlands. lated AEs in patients with inflammatory and pulse therapy of ≥250 mg pred-
E-mail: [Link]@[Link] diseases were retrieved. nisone equivalent during one or more
Received and accepted on July 27, 2011. Results. In only 8 studies (344 pa- days (2).
Clin Exp Rheumatol 2011; 29 (Suppl. 68): tients), 4 placebo-controlled and 4 not The occurrence of adverse events (AEs)
S85-S92. placebo-controlled studies, intrave- with daily administration of medium
© Copyright CLINICAL AND nous pulse GC-related AEs had been doses of glucocorticoids is frequent (3).
EXPERIMENTAL RHEUMATOLOGY 2011. documented (in total 323 AEs), with an Therefore, in the 1960s, short-term in-
AE rate of 35/100 patient-years. In the travenous pulse dosing was introduced,
Key words: glucocorticoids, drug 4 placebo-controlled studies among RA hoping that GCs would still have their
toxicity, adverse events, infusions, and systemic sclerosis patients, most of beneficial effects but fewer AEs com-
intravenous, pulse therapy, drug, the odds ratios of individual AEs were pared to daily administration (1). In-
arthritis, rheumatoid, asthma, not statistically significant, except for travenously used GCs, such as methyl-
scleroderma, systemic flushing, heart rhythm disorder, distur- prednisolone (MP), have stronger anti-
bance of taste, lower respiratory infec- inflammatory and immunosuppressive
tion, and headache. In the 4 not place- potencies. For example, lymphopenia is
bo-controlled studies increased diasto- more marked in patients who received
lic blood pressure was most frequent, 80, 250, 500, or 1000 mg intravenous
followed by flushing and diabetes mel- MP than in patients who received oral
litus. Adverse events seen in more than prednisone in a range of 15 mg to 100
15% of patients of all included studies mg, with a dose-response relationship
were increased blood pressure, flush- (4). A decrease of approximately 75%
ing, headache, disturbance of taste, in the number of total peripheral cir-
tachycardia and hyperglycemia. culating lymphocytes was recorded in
Conclusion. GC pulse therapy results patients who received intravenous MP
in a high AE rate, i.e. 35/100 patient- (5). In addition, intravenous MP de-
years. Cardiovascular AEs are most creases lymphocyte blastogenesis and
frequently reported in the literature. lymphocyte activation (5, 6). Further-
Furthermore, flushing had the highest more, MP inhibits in a dose-dependent
odds ratio in the placebo-controlled manner the adherence of polymorpho-
studies and also a high event rate in the nuclear leukocytes (PMN) to endothe-
not placebo-controlled studies. lial cells and as a consequence the mi-
Competing interests: J.W.J. Bijlsma has gration of PMN from the bloodstream,
served as consultant and speaker for Nitec,
Introduction which results in granulocytosis in the
Mundipharma and Horizon;
J.W.G. Jacobs has been a speaker for Glucocorticoids (GCs) are used in treat- blood (7).
Mundipharma on two occasions; ment of various allergic and inflam- GCs act via three mechanisms; genom-
the other co-authors have declared no matory conditions, such as rheumatic ic mechanisms, specific non-genomic
competing interests. diseases, inflammatory bowel diseases, mechanisms and unspecific non-ge-

S-85
AEs of intravenous glucocorticoid pulse therapy / N.A.M. Smits et al.

nomic mechanisms (8). Genomic ef- dose intravenous GC therapy could be eases using the bibliographic databases
fects are mediated by cytosolic GC partly due to these unspecific non-ge- PubMed, Embase, and the Cochrane li-
receptors that alter the gene expression nomic effects (11). brary. The search consisted of relevant
by activating or repressing specific tar- Previously, the risk of AE of low to me- keywords for disease, e.g. rheumatic
get genes. Non-genomic effects are me- dium dose GCs have been systemically diseases, obstructive lung diseases, and
diated by steroid-selective membrane reviewed (3), but the risks of AEs of inflammatory bowel diseases, treatment
receptors (8, 9). Their effects occur in a administration of pulse glucocorticoids (intravenous pulse therapy), and AEs
few minutes, compared genomic effects have not been systematically analysed. which was checked by experts (JWGJ
which occur after at least 30 minutes. The aim of this study was to systemati- and JWJB). The search terms included
An example of a non-genomic effect is cally review the literature on reported keywords, words of title or abstract,
the rapid inhibition of p56lck (Lck) and AEs of intravenous pulse therapy of synonyms, and plurals. MESH terms
p59fyn (Fyn) kinases by GCs in human GCs and to quantify the risk of AEs, were added for the PubMed search.
T lymphocytes which leads to immu- independent of the underlying inflam- All search terms were combined using
nosuppression (10). Unspecific non-ge- matory disease. Boolean operators (see Box 1).
nomic mechanisms occur only at high
GC dosages. The mode-of-action could Methods Study selection
be direct interactions with the cellular Literature search Studies were selected by two reviewers
energy metabolism, which leads to an A literature search was performed by (NAMS and ND). The studies were in-
immunosuppressive effect within sec- the authors NAMS and ND, reviewing cluded if they met the following criteria
onds. The therapeutic benefits of high- GC-related AEs in inflammatory dis- to the title, abstract, and full text:

Box 1. General systemic literature search

Database Searchstring n° of studies

Pubmed (“asthma”[MeSH Major Topic] NOT (“asthma, aspirin-induced”[MeSH Terms] OR aspirin induced asthma[Title/ 64
Abstract] OR NSAID-induced asthma[Title/Abstract] OR aspirin-induced asthma syndrome[Title/Abstract])
OR “rheumatic diseases”[MeSH Major Topic] OR “rheumatic disease”[Title/Abstract] OR “rheumatic
diseases”[Title/Abstract] OR “rheumatoid arthritis”[Title/Abstract] OR “arthritis, rheumatoid”[MeSH Major
Topic] OR “polymyalgia rheumatica”[MeSH Terms] OR “lupus erythematosus, systemic”[MeSH Terms] OR
“polymyositis”[MeSH Terms] OR “dermatomyositis”[MeSH Terms] OR “giant cell arteritis”[MeSH Terms] OR
“takayasu arteritis”[MeSH Terms] OR “polyarteritis nodosa”[MeSH Terms] OR “wegener granulomatosis”[MeSH
Terms] OR “microscopic polyangiitis”[MeSH Terms] OR “churg-strauss syndrome”[MeSH Terms] OR “beh-
cet syndrome”[MeSH Terms] OR “sarcoidosis”[MeSH Major Topic] OR “polychondritis, relapsing”[MeSH
Terms] OR “shock, septic”[MeSH Terms] OR “inflammatory bowel diseases”[MeSH Major Topic] OR “in-
flammatory bowel disease”[Title/Abstract] OR “inflammatory bowel diseases”[Title/Abstract] OR “pulmonary
disease, chronic obstructive”[MeSH Major Topic] OR “chronic obstructive pulmonary disease”[Title/Abstract]
OR COPD[Title/Abstract] OR “scleroderma, systemic”[MeSH Terms])
AND
((“Glucocorticoids/therapeutic use”[Mesh] OR glucocorticoids[MeSH Terms] OR prednisolone[MeSH
Terms] OR prednisone[MeSH Terms] OR predniso*[Title/Abstract] OR dexamethasone[MeSH Terms]
OR methylprednisolone[MeSH Terms] OR hydrocortisone[MeSH Terms] OR cortisone[MeSH Terms] OR
solumedrol[Title/Abstract] OR “solu medrol”[Title/Abstract] OR depomedrol[Title/Abstract] OR “depo
medrol”[Title/Abstract]) AND (“40 mg”[Title/Abstract] OR “60 mg”[Title/Abstract] OR “80 mg”[Title/Ab-
stract] OR “100 mg”[Title/Abstract] OR “120 mg”[Title/Abstract] OR “200 mg”[Title/Abstract] OR “1000
mg”[Title/Abstract]) AND (intravenous[Title/Abstract] OR “pulse treatment”[Title/Abstract] OR “pulse
therapy”[Title/Abstract]))
AND
(“adverse effect”[Title/Abstract] OR “adverse effects”[Title/Abstract] OR “adverse event”[Title/Abstract] OR
“adverse events”[Title/Abstract] OR “side effect”[Title/Abstract] OR “side effects”[Title/Abstract] OR “side-
effect”[Title/Abstract] OR “side-effects”[Title/Abstract] OR “unwanted effect”[Title/Abstract] OR “unwanted
effects”[Title/Abstract] OR “osteoporosis”[MeSH Terms] OR “osteonecrosis”[MeSH Terms] OR “muscle
weakness”[MeSH Terms] OR “glucose intolerance”[MeSH Terms] OR “diabetes mellitus”[MeSH Terms] OR
“weight gain”[MeSH Terms] OR “hyperglycemia”[MeSH Terms] OR “menstruation disturbances”[MeSH
Terms] OR “dyslipidemias”[MeSH Terms] OR “atherosclerosis”[MeSH Terms] OR “hypertension”[MeSH
Terms] OR “edema”[MeSH Terms] OR “heart failure”[MeSH Terms] OR “water-electrolyte imbalance”[MeSH
Terms] OR “myocardial infarction”[MeSH Terms] OR “coronary artery disease”[MeSH Terms] OR “tach-
ycardia, sinus”[MeSH Terms] OR “hypokalemia”[MeSH Terms] OR “hypocalcemia”[MeSH Terms] OR
“hirsutism”[MeSH Terms] OR “alopecia”[MeSH Terms] OR “hypertrichosis”[MeSH Terms] OR “cushing

S-86
 AEs of intravenous glucocorticoid pulse therapy / N.A.M. Smits et al.

syndrome”[MeSH Terms] OR “purpura”[MeSH Terms] OR “cataract”[MeSH Terms] OR “glaucoma”[MeSH

Terms] OR “peptic ulcer”[MeSH Terms] OR “pancreatitis”[MeSH Terms] OR “candidiasis”[MeSH
Terms] OR “depression”[MeSH Terms] OR “anxiety”[MeSH Terms] OR “irritable mood”[MeSH
Terms] OR “dizziness”[MeSH Terms] OR “tinnitus”[MeSH Terms] OR “carcinoma”[MeSH Terms] OR
“thrombocytopenia”[MeSH Terms] OR “leukopenia”[MeSH Terms] OR “leukocytosis”[MeSH Terms] OR
“proteinuria”[MeSH Terms] OR “arrhythmias, cardiac”[MeSH Terms] OR “hypernatremia”[MeSH Terms]
OR “bone loss”[Title/Abstract] OR “Vertebral deformity”[Title/Abstract] OR “Vertebral deformities”[Title/
Abstract] OR “fracture”[Title/Abstract] OR “fractures”[Title/Abstract] OR “bone mineral density”[Title/Ab-
stract] OR “bone density”[Title/Abstract] OR myopathy[Title/Abstract] OR “blood glucose”[Title/Abstract]
OR “fasting glucose”[Title/Abstract] OR “urine glucose”[Title/Abstract] OR “glycosuria”[Title/Abstract]
OR “adipositas”[Title/Abstract] OR “buffalo hump”[Title/Abstract] OR “hyperlipidemia”[Title/Abstract] OR
hyperlipidaemia[Title/Abstract] OR hypercholesterolaemia[Title/Abstract] OR “angina pectoris”[Title/Ab-
stract] OR “blood pressure”[Title/Abstract] OR oedema[Title/Abstract] OR “cardiac insufficiency”[Title/Ab-
stract] OR “fluid retention”[Title/Abstract] OR “facial fullness”[Title/Abstract] OR “facial swelling”[Title/Ab-
stract] OR “moon face”[Title/Abstract] OR “cutaneous atrophy”[Title/Abstract] OR “skin atrophy”[Title/Ab-
stract] OR “skin hemorrhage”[Title/Abstract] OR “skin bleeding”[Title/Abstract] OR striae[Title/Abstract] OR
“easy bruisability”[Title/Abstract] OR “easy bruising”[Title/Abstract] OR “wound healing”[Title/Abstract] OR
“hair loss”[Title/Abstract] OR “gastric ulcer”[Title/Abstract] OR “gastroduodenal ulcer”[Title/Abstract] OR
dyspepsia[Title/Abstract] OR dysfagia[Title/Abstract] OR “gastric hemorrhage”[Title/Abstract] OR “stomach
hemorrhage”[Title/Abstract] OR “gastroduodenal hemorrhage”[Title/Abstract] OR “viral infection”[Title/Ab-
stract] OR “fungal infection”[Title/Abstract] OR “bacterial infection”[Title/Abstract] OR “skin infection”[Title/
Abstract] OR “urinary infection”[Title/Abstract] OR “respiratory infection”[Title/Abstract] OR infection[Title/
Abstract] OR libido[Title/Abstract] OR infertility[Title/Abstract] OR palpitation[Title/Abstract] OR
psychosis[Title/Abstract] OR euphoria[Title/Abstract] OR seizures[Title/Abstract] OR tremor[Title/Abstract]
OR “mood disturbance”[Title/Abstract] OR “mood lability”[Title/Abstract])
Limits activated: Humans, All adults: 19+ years

Embase ((‘asthma’/exp NOT (‘asthma, aspirin-induced’/exp OR ‘aspirin induced asthma’/exp OR ‘nsaid-induced asth- 69
ma’/exp OR ‘aspirin-induced asthma’/exp) OR ‘rheumatic disease’/exp OR ‘rheumatic diseases’/exp OR ‘rheu-
matoid arthritis’/exp OR ‘arthritis, rheumatoid’/exp OR ‘polymyalgia rheumatica’/exp OR ‘lupus erythemato-
sus, systemic’/exp OR ‘polymyositis’/exp OR ‘dermatomyositis’/exp OR ‘giant cell arteritis’/exp OR ‘takayasu
arteritis’/exp OR ‘polyarteritis nodosa’/exp OR ‘wegener granulomatosis’/exp OR ‘microscopic polyangiitis’/
exp OR ‘churg-strauss syndrome’/exp OR ‘behcet syndrome’/exp OR ‘sarcoidosis’/exp OR ‘polychondritis,
relapsing’/exp OR ‘shock, septic’/exp OR ‘inflammatory bowel disease’/exp OR ‘inflammatory bowel dis-
eases’/exp OR ‘pulmonary disease, chronic obstructive’/exp OR ‘chronic obstructive pulmonary disease’/exp
OR ‘copd’/exp OR ‘scleroderma, systemic’/exp) AND ([article]/lim OR [article in press]/lim OR [conference
abstract]/lim OR [conference paper]/lim OR [conference review]/lim OR [review]/lim OR [short survey]/lim)
AND ([adult]/lim OR [aged]/lim) AND [humans]/lim)
AND
((‘glucocorticoids’/exp OR ‘glucocorticoid’/exp OR ‘prednisolone’/exp OR ‘prednisone’/exp OR predniso*
OR ‘dexamethasone’/exp OR ‘methylprednisolone’/exp OR ‘hydrocortisone’/exp OR ‘cortisone’/exp OR ‘sol-
umedrol’/exp OR ‘solu medrol’/exp OR ‘depomedrol’/exp OR ‘depo medrol’/exp) AND (‘40 mg’ OR ‘60 mg’
OR ‘80 mg’ OR ‘100 mg’ OR ‘120 mg’ OR ‘200 mg’ OR ‘1000 mg’) AND (‘intravenous’/exp OR ‘drug pulse
therapy’/exp OR ‘short course therapy’/exp) NOT (‘oral’/exp OR ‘intramuscular’/exp OR ‘rectal’/exp OR
‘transdermal’/exp OR ‘topical’/exp OR ‘intranasal drug administration’/exp) AND ([article]/lim OR [article in
press]/lim OR [conference abstract]/lim OR [conference paper]/lim OR [conference review]/lim OR [review]/
lim OR [short survey]/lim) AND ([adult]/lim OR [aged]/lim) AND [humans]/lim)
AND
((‘adverse effect’/exp OR ‘side effect’/exp OR ‘side-effect’/exp OR ‘osteoporosis’/exp OR ‘osteonecrosis’/exp
OR ‘muscle weakness’/exp OR ‘glucose intolerance’/exp OR ‘diabetes mellitus’/exp OR ‘weight gain’/exp OR
‘hyperglycemia’/exp OR ‘menstruation disturbances’/exp OR ‘dyslipidemias’/exp OR ‘atherosclerosis’/exp OR
‘hypertension’/exp OR ‘edema’/exp OR ‘heart failure’/exp OR ‘water-electrolyte imbalance’/exp OR ‘myocar-
dial infarction’/exp OR ‘coronary artery disease’/exp OR ‘tachycardia, sinus’/exp OR ‘hypokalemia’/exp OR
‘hypocalcemia’/exp OR ‘hirsutism’/exp OR ‘alopecia’/exp OR ‘hypertrichosis’/exp OR ‘cushing syndrome’/
exp OR ‘purpura’/exp OR ‘cataract’/exp OR ‘glaucoma’/exp OR ‘peptic ulcer’/exp OR ‘pancreatitis’/exp OR
‘candidiasis’/exp OR ‘depression’/exp OR ‘anxiety’/exp OR ‘irritable mood’/exp OR ‘dizziness’/exp OR ‘tin-
nitus’/exp OR ‘carcinoma’/exp OR ‘thrombocytopenia’/exp OR ‘leukopenia’/exp OR ‘leukocytosis’/exp OR
‘proteinuria’/exp OR ‘arrhythmias, cardiac’/exp OR ‘hypernatremia’/exp OR ‘bone loss’/exp OR ‘vertebral
deformity’/exp OR ‘fracture’/exp OR ‘fractures’/exp OR ‘bone mineral density’/exp OR ‘bone density’/exp
OR ‘myopathy’/exp OR ‘blood glucose’/exp OR ‘urine glucose’/exp OR ‘glycosuria’/exp OR ‘adipositas’/exp
OR ‘hyperlipidemia’/exp OR ‘hyperlipidaemia’/exp OR ‘hypercholesterolaemia’/exp OR ‘angina pectoris’/exp
OR ‘blood pressure’/exp OR ‘oedema’/exp OR ‘cardiac insufficiency’/exp OR ‘fluid retention’/exp OR ‘face

S-87
AEs of intravenous glucocorticoid pulse therapy / N.A.M. Smits et al.

edema’/exp OR ‘moon face’/exp OR ‘cutaneous atrophy’/exp OR ‘skin atrophy’/exp OR ‘skin hemorrhage’/

exp OR ‘skin bleeding’/exp OR ‘striae’/exp OR ‘easy bruisability’/exp OR ‘restlessness’/exp OR ‘wound
healing’/exp OR ‘hair loss’/exp OR ‘gastric ulcer’/exp OR ‘gastroduodenal ulcer’/exp OR ‘dyspepsia’/exp
OR ‘dysphagia’/exp OR ‘gastric hemorrhage’/exp OR ‘stomach hemorrhage’/exp OR ‘gastroduodenal hemor-
rhage’/exp OR ‘viral infection’/exp OR ‘fungal infection’/exp OR ‘bacterial infection’/exp OR ‘skin infection’/
exp OR ‘urinary infection’/exp OR ‘respiratory infection’/exp OR ‘infection’/exp OR ‘libido’/exp OR ‘infertil-
ity’/exp OR ‘palpitation’/exp OR ‘psychosis’/exp OR ‘euphoria’/exp OR ‘seizures’/exp OR ‘tremor’/exp OR
‘mood disturbance’/exp OR ‘mood lability’/exp) AND ([article]/lim OR [article in press]/lim OR [conference
abstract]/lim OR [conference paper]/lim OR [conference review]/lim OR [review]/lim OR [short survey]/lim)
AND ([adult]/lim OR [aged]/lim) AND [humans]/lim)

Cochrane ((asthma NOT (“asthma, aspirin-induced” OR “aspirin induced asthma” OR “NSAID-induced asthma” OR “as- 25
pirin-induced asthma syndrome”)) OR “rheumatic diseases” OR “rheumatic disease” OR “rheumatic diseases”
OR “rheumatoid arthritis” OR “arthritis, rheumatoid” OR “polymyalgia rheumatica” OR “lupus erythematosus,
systemic” OR polymyositis OR dermatomyositis OR “giant cell arteritis” OR “takayasu arteritis” OR “polyar-
teritis nodosa” OR “wegener granulomatosis” OR “microscopic polyangiitis” OR “churg-strauss syndrome”
OR “behcet syndrome” OR sarcoidosis OR “polychondritis, relapsing” OR “shock, septic” OR “inflamma-
tory bowel diseases” OR “inflammatory bowel disease” OR “inflammatory bowel diseases” OR “pulmonary
disease, chronic obstructive” OR “chronic obstructive pulmonary disease” OR COPD OR “scleroderma, sys-
temic”):ti,ab,kw
AND
(glucocorticoids OR prednisolone OR prednisone OR predniso* OR dexamethasone OR methylprednisolone
OR hydrocortisone OR cortisone OR solumedrol OR “solu medrol” OR depomedrol OR “depo medrol”):
ti,ab,kw AND (”40 mg” OR “60 mg” OR “80 mg” OR “100 mg” OR “120 mg” OR “200 mg” OR “1000 mg”):
ti,ab,kw AND (intravenous OR “pulse treatment” OR “pulse therapy”):ti,ab,kw
AND
(“adverse effect” OR “adverse effects” OR “adverse event” OR “adverse events” OR “side effect” OR “side
effects” OR “side-effect” OR “side-effects” OR “unwanted effect” OR “unwanted effects” OR osteoporosis
OR osteonecrosis OR “muscle weakness” OR “glucose intolerance” OR “diabetes mellitus” OR “weight gain”
OR hyperglycemia OR “menstruation disturbances” OR dyslipidemias OR atherosclerosis OR hypertension
OR edema OR “heart failure” OR “water-electrolyte imbalance” OR “myocardial infarction” OR “coronary
artery disease” OR “tachycardia, sinus” OR hypokalemia OR hypocalcemia OR hirsutism OR alopecia OR
hypertrichosis OR “cushing syndrome” OR “purpura” OR “cataract” OR “glaucoma” OR “peptic ulcer” OR
pancreatitis OR candidiasis OR depression OR anxiety OR “irritable mood” OR dizziness OR tinnitus OR car-
cinoma OR thrombocytopenia OR leukopenia OR leukocytosis OR proteinuria OR “arrhythmias, cardiac” OR
hypernatremia OR hypernatraemia OR “bone loss” OR “Vertebral deformity” OR “Vertebral deformities” OR
“fracture” OR “fractures” OR “bone mineral density” OR “bone density” OR myopathy OR “blood glucose”
OR “fasting glucose” OR “urine glucose” OR glycosuria OR adipositas OR “buffalo hump” OR hyperlipidemia
OR hyperlipidaemia OR hypercholesterolaemia OR “angina pectoris” OR “blood pressure” OR oedema OR
“cardiac insufficiency” OR “fluid retention” OR “facial fullness” OR “facial swelling” OR “moon face” OR
“cutaneous atrophy” OR “skin atrophy” OR “skin hemorrhage” OR “skin bleeding” OR striae OR “easy bruis-
ability” OR “easy bruising” OR “wound healing” OR “hair loss” OR “gastric ulcer” OR “gastroduodenal ulcer”
OR dyspepsia OR dysphagia OR “gastric hemorrhage” OR “stomach hemorrhage” OR “gastroduodenal hemor-
rhage” OR “viral infection” OR “fungal infection” OR “bacterial infection” OR “skin infection” OR “urinary
infection” OR “respiratory infection” OR infection OR libido OR infertility OR palpitation OR psychosis OR
euphoria OR seizures OR tremor OR “mood disturbance” OR “mood lability”):ti,ab,kw
Total number of studies minus duplicates: 158 –27= 131

• Study population: adults with in- led trials, prospective trials, cohort munosuppressive drug. Only full text
flammatory diseases treated with studies, and observational studies. available papers were included.
GCs. If only a subgroup of the total study
• Intervention: patients who received population received intravenous pulse Data extraction
≥250 mg prednisone equivalent in- GCs and their stratified data were re- The characteristics of included stud-
travenous pulse therapy for one or ported, this subgroup was included in ies were recorded, like number of pa-
more days. the analysis. Exclusion criteria were tients, gender, age, diagnosis, dose of
• Outcome: numbers of AEs on pa- case reports, non-English language, GC, study duration, patients dropping
tient’s level caused by GC treat- and pulse therapy started together out from the study (missing data), and
ment. with more than one disease-modifying deaths. The following information
• Study designs: (randomised) control- antirheumatic drug (DMARD) or im- about the reported AEs was collected:

S-88
 AEs of intravenous glucocorticoid pulse therapy / N.A.M. Smits et al.

type and number of AEs, the frequency

of monitoring, blinding of investigator,
and study methodology. AEs were clas-
sified in groups, according to classifica-
tions used in literature. (3, 12).

Quality assessment
The selected studies were judged for the
quality of their assessing and reporting
of AEs, using the following criteria (3):
• Predefined AE: yes/no; whether the
study applied predefined AEs.
• Standardised AE scoring protocol:
yes/no; whether in the study an AEs
protocol, e.g. questionnaire had been
used.
• Missing data: yes/no; whether the
study reported the number of miss-
ing data and why data was missing,
e.g. drop outs. Fig. 1. Number of adverse events of intravenous GC pulse therapy per 100 exposed patients. Total
A study could score 0 to 3 quality exposed patients n=344 (rheumatoid arthritis (RA): 6 studies, n=303; systemic sclerosis (SS): 1 study,
points (1 point per criterion), where 3 n=17; asthma: 1 study, n=24).
points reflects the highest quality.
Adverse events of all 8 included 3 months to 4.5 years approximately.
Data analysis studies Three of the studies were randomised
AEs of GCs versus placebo were ana- In total, 344 patients received intrave- double-blinded controlled trials and
lysed calculating odds ratios. Studies nous GCs and 323 AEs were recorded, one was a non-randomised controlled
without placebo group were analysed with an AE rate of 35/100 patient-years. trial.
with events/exposed patients. Analyses Cardiovascular AEs were most often The overall odds ratio of reported AEs
were done by using the Comprehen- noted, followed by infectious AEs. in these placebo-controlled studies was
sive Meta-Analysis software (Biostat, Dermatological, gastrointestinal, and 1.83 with an 95% confidence inter-
Engelwood, New Jersey, USA). neurological AEs were only reported in val (CI) of 0.98–3.40 (p-value 0.06).
RA patients. In patients with systemic Most of the odds ratios of individual
Results sclerosis, only infections were report- AEs were not statistically significant,
Literature search ed, and one asthma patient developed except for flushing, heart rhythm dis-
Using the specified search strategy (on- diabetes mellitus (Fig. 1). Only one order, disturbance of taste, lower res-
line Appendix), 158 studies were ob- AE was classified as severe by the au- piratory infection, and headache (Table
tained, consisting of 64, 69, and 25 hits thors, which was nocturia and frequent II), showing an increased risk associ-
in PubMed, Embase, and Cochrane, re- voiding in an RA patient (17). Adverse ated with intravenous GCs, compared
spectively. Doubles (n=27) were filtered events seen in more than 15% of pa- to placebo. The odds ratio of flushing
by loading the studies into Reference tients of all included studies were in- was highest: 15 (95% CI 5.3–40), fol-
Manager 11, an electronic bibliograph- creased blood pressure, flushing, head- lowed by that for headache: 6.2 (95%
ic management system. After screening ache, disturbance of taste, tachycardia CI 2.3–16).
the titles and abstracts, and reading the and hyperglycemia.
full texts, 5 articles met the inclusion The 4 not placebo-controlled
criteria. Using the reference lists of the The 4 placebo-controlled studies studies
full text available studies another 3 re- The study characteristics of these 4 Of the 4 not placebo-controlled studies,
lated articles were included. In total, 8 studies are listed in Table I. In total, 3 had been performed among patients
articles described AEs of intravenous 220 patients received high pulse intra- with RA and 1 among patients with
GCs. The studies included patients with venous GCs and 218 patients received asthma (17-20). In total 124 patients
RA, asthma, and systemic sclerosis. placebo. One study reported on sys- received intravenous pulse GCs and
Relevant AE data were extracted from temic sclerosis patients receiving intra- in total 39 different AEs were docu-
these studies. Only 4 studies compared venous dexamethasone (14). In other mented. The top 20 most reported AEs
GCs with placebo (13-16). The other studies, patients with RA or asthma are listed in Table III. The event rate of
studies were not controlled or compared received MP. The quality of reporting increased diastolic blood pressure was
various dosages of GCs. The different AEs of these studies was 1, 2, 3, and 2, highest, 88%, followed by event rates
GC groups were analysed separately. and the length of the study ranged from of 20% and 24% of flushing and dia-

S-89
AEs of intravenous glucocorticoid pulse therapy / N.A.M. Smits et al.

Table I. Characteristics of the 4 placebo-controlled studies on intravenous pulse GC. with inflammatory diseases. It is strik-
ing that after approximately 50 years
Hansen 1990 Williams 1982 Williams 1988 Sharada 1994
of intravenous pulse GC use in clinical
Patients (N) 97 20 286 35 practice, the prevalence of AEs is not
Age (years) 60.0 56.0 52.3 33.7 well known. Therefore, this study was
Gender (% female) 73 90 71 91 undertaken to report intravenous GC-
Disease RA RA RA SS related AEs present in the literature.
Kind of GC MP MP MP D It was remarkable that only 8 studies
Cumulative dose# 7875 1250 2308 4000
met our criteria. Frequently studies had
Study duration (days) 365 84 1689 180
to be excluded because they focused
Blind scoring of AE Yes Yes No Yes
Quality¥ 1 2 3 2
on treatment effects of GCs and not on
Predefined AE No Yes Yes Yes AEs. In almost all studies concomitant
AE scoring per protocol Yes Yes Yes Yes drugs were used, such as DMARDs
Missing data* No No Yes No and NSAIDs, which could interfere
with reported AEs. Nowadays, rheuma-
#
Prednisone equivalent (mg) tologists tend to start earlier concomi-
*Missing data: if the study noted that patients dropped out of the study
¥
Quality of reporting adverse events: studies could score 1 point per criteria with a maximum of 3 tant drugs, for example DMARDs in
points. The three criteria consist of predefined AE, AE scoring per protocol, and missing data. early RA. We found only studies of the
RA: rheumatoid arthritis; SS: systemic sclerosis; MP: methylprednisolone; D: dexamethasone. 1980s that reported AEs of intravenous
GCs without concomitant drugs. We in-
betes mellitus, respectively. Diabetes Discussion cluded studies with patients using con-
mellitus, heart rhythm disorders, and To our knowledge, this is the first study comitantly more than one DMARD or
osteonecrosis were reported in 2 stud- that systematically analysed reported immunosuppressive drug, but only if
ies; all other AEs had been reported AEs of intravenous pulse GCs (≥250 these drugs had not been initiated at the
only in one study. mg prednisone equivalent) in patients same time as the GC.

Table II. Odds ratios of adverse events recorded in the 4 placebo-controlled studies on intravenous pulse GC.

Adverse event Study/subgroup Study Number of Odds ratio p-value

methodology patients (95% CI)

Musculoskeletal
Musculoskeletal Williams 1988 2 143 vs. 143 0.59 (0.14-2.52) 0.48
Osteonecrosis Williams 1988 2 143 vs. 143 4.09 (0.45-37.02) 0.21
Dermatological
Dermatological Williams 1988 2 143 vs. 143 1.21 (0.36-4.05) 0.76
Flushing Hansen 1990 1 50 vs. 47 14.69 (5.34-40.46) 0.00
Cardiovascular
Cardiovascular Williams 1988 2 143 vs. 143 0.50 (0.24-1.06) 0.07
Heart rhythm disorder Hansen 1990 1 50 vs. 47 2.93 (1.03-8.36) 0.04
Gastrointestinal
Gastrointestinal Williams 1988 2 143 vs. 143 0.92 (0.40-2.08) 0.83
Disturbance of taste Hansen 1990 1 50 vs. 47 5.06 (1.55-16.54) 0.01
Endocrine and metabolic
Endocrine and metabolic Williams 1988 2 143 vs. 143 0.16 (0.02-1.35) 0.09
Infectious
Lower respiratory tract infection Sharada 1994 1 17 vs. 18 5.62 (1.18-26.85) 0.03
Skin infection Sharada 1994 1 17 vs. 18 1.07 (0.13-8.56) 0.95
Dental infection Sharada 1994 1 17 vs. 18 1.06 (0.06-18.45) 0.97
Neurological
Neurological Williams 1988 2 143 vs. 143 2.03 (0.37-11.26) 0.42
Headache Hansen 1990 1 50 vs. 47 6.19 (2.33-16.43) 0.00
Ophthalmological
Glaucoma Williams 1982 1 10 vs. 10 3.32 (0.12-91.60) 0.48
Others
Haematological Williams 1988 2 143 vs. 143 5.07 (0.24-106.56) 0.30
Genitourinary Williams 1988 2 143 vs. 143 0.38 (0.12-1.25) 0.11
Others Williams 1988 2 143 vs. 143 3.02 (0.12-74.78) 0.50

Study methodology: 1: Randomised controlled trial; 2: controlled study without randomisation; others: not specified by authors.

S-90
 AEs of intravenous glucocorticoid pulse therapy / N.A.M. Smits et al.

Table III. Top 20 most frequently reported adverse events in the 4 not placebo-controlled in patients who have a heart disease.
studies.* So, acquiring a detailed cardiovascular
Adverse event Event rate Events/exposed Number of studies/
history is recommended, with atten-
(%) patients subgroups reporting tion for high blood pressure and signs
the AE of angina pectoris, or a familial car-
diovascular risk profile. Independently
1 Increased diastolic blood pressure 88 44/50 1
of these results, before the treatment
2 Flushing 24 12/50 1
3 Diabetes mellitus 20 15/74 2
with intravenous GCs, an ECG should
4 Angina pectoris 20 1/5 1 be taken and blood pressure should be
5 Headache 20 10/50 1 monitored in each patient during ther-
6 Heart rhythm disorders 18 10/55 2 apy. Formulating recommendations for
7 Disturbance of taste 18 9/50 1 monitoring for diabetes mellitus based
8 Moon face 18 9/50 1 on this literature search is not possible,
9 Gastroduodenal ulcer 14 7/50 1
other than to check blood glucose lev-
10 Osteonecrosis 11 8/73 2
els during the therapy.
11 Cataract 10 5/50 1
12 Dizziness 10 5/50 1 In conclusion, intravenous GC pulse
13 Pollacisuria 9 2/22 1 therapy results in a high AE rate, name-
14 Urinary tract infection 8 4/50 1 ly 35/100 patient-years. Cardiovascular
15 Mood disturbance 8 4/50 1 AEs are most frequently reported in the
16 Skin changes 6 3/50 1 literature, in particular increased di-
17 Osteoporosis 4 2/5 1 astolic blood pressure and heart rhythm
18 Urticaria 4 2/50 1
disorders. Furthermore, flushing had
19 Fatigue 4 2/50 1
20 Sleep disturbance 4 2/50 1
the highest odds ratio in the placebo-
controlled group and a high event rate
*Please note: included studies are characterised by a wide range of study duration and study popula- in the not placebo-controlled group.
tion, influencing the ranks in this top 20.
References
Often, AEs were not registered nu- paired glucose intolerance (19). Infec- 1. BARRY M: The use of high-dose pulse meth-
ylprednisolone in rheumatoid arthritis. Un-
meral but reported as the mean level of tions could be linked to the adminis- proved therapy. Arch Intern Med 1985; 145:
the study group. This was done for in- tered GCs but also to the inflammatory 1483-4.
stance with glucose levels, weight, and disease (21); for instance patients with 2. BUTTGEREIT F, DA SILVA JA, BOERS M et al.:
blood pressure. These AEs could not systemic sclerosis are prone to develop Standardised nomenclature for glucocorti-
coid dosages and glucocorticoid treatment
be included because mean levels give skin ulcer infections and dental infec- regimens: current questions and tentative
no insight into the prevalence of AEs, tions. Finally, because the primary answers in rheumatology. Ann Rheum Dis
as the number of patients who suffered aim of the studies most frequently had 2002; 61: 718-22.
from the AE cannot be calculated from not been analysis of AEs, it was less 3. HOES JN, JACOBS JW, VERSTAPPEN SM, BI-
JLSMA JW, VAN DER HEIJDEN GJ: Adverse
means. straightforward to identify studies re- events of low- to medium-dose oral gluco-
Limitation of our meta-analysis is that porting on AEs from their title and ab- corticoids in inflammatory diseases: a meta-
the included studies were character- stract; studies may have been missed. analysis. Ann Rheum Dis 2009; 68: 1833-8.
4. BADSHA H, EDWARDS CJ: Intravenous puls-
ised by a wide range of study duration, Most of the reported AEs were mild.
es of methylprednisolone for systemic lupus
from 2 days to 4.5 years. One study For example, the lower respiratory erythematosus. Semin Arthritis Rheum 2003;
had no follow up, and due to this they tract infections in the systemic scle- 32: 370-7.
only reported short-term AEs (19). An- rosis patients were minor, not associ- 5. WEBEL ML, RITTS RE, JR., TASWELL HF,
DANADIO JV, JR., WOODS JE: Cellular im-
other study reported both short-term as ated with chest x-ray abnormalities or munity after intravenous administration of
long-term AEs (20). Also the range of positive sputum cultures. Furthermore, methylprednisolone. J Lab Clin Med 1974;
the study population was wide, rang- flushing and headache generally are 83: 383-92.
ing from only 5 patients per group to considered to be not severe and tem- 6. FAN PT, YU DT, TARGOFF C, BLUESTONE R:
Effect of corticosteroids on the human im-
143 patients per group, but the overall porary. Though, patients often experi- mune response. Suppression of mitogen-in-
number of patients per group was low, ence these AE as very inconvenient and duced lymphocyte proliferation by “pulse”
approximately 20 patients per group. patients who receive intravenous GCs methylprednisolone. Transplantation 1978;
Another limitation is that some studies should be aware of the occurrence of 26: 266-7.
7. YOSHIDA N, YOSHIKAWA T, NAKAMURA Y et
reported AEs which could be (partly) these AEs. al.: Methylprednisolone inhibits neutrophil-
also attributed to the patients’ history Heart rhythm disorders induced by endothelial cell interactions induced by inter-
or the underlying disease. For instance, high-dose GC treatment, like palpi- leukin-1beta under flow conditions. Life Sci
1997; 60: 2341-7.
diabetes mellitus was reported as an tation and tachycardia are more of a 8. BUTTGEREIT F, WEHLING M, BURMESTER
AE in an asthma patient, while this clinical problem. They might spontane- GR: A new hypothesis of modular glucocor-
patient was already known to have im- ously recover, but could be dangerous ticoid actions: steroid treatment of rheumatic

S-91
AEs of intravenous glucocorticoid pulse therapy / N.A.M. Smits et al.

diseases revisited. Arthritis Rheum 1998; 41: treatment with methylprednisolone combined Pulsed methylprednisolone in active early
761-7. with disease modifying drugs in rheumatoid rheumatoid disease: A dose-ranging study.
9. WEHLING M: Specific, nongenomic actions arthritis. BMJ 1990; 301: 268-70. Br J Rheumatol 1988; 27: 211-4.
of steroid hormones. Annu Rev Physiol 1997; 14. SHARADA B, KUMAR A, KAKKER R et al.: 18. TVEDE N, NIELSEN LP, ANDERSEN V: Brady-
59: 365-93. Intravenous dexamethasone pulse therapy in cardia after high-dose intravenous methyl-
10. LOWENBERG M, TUYNMAN J, BILDERBEEK diffuse systemic sclerosis. A randomized pla- prednisolone therapy. Scand J Rheumatol
J et al.: Rapid immunosuppressive effects of cebo-controlled study. Rheumatol Int 1994; 1986; 15: 302-4.
glucocorticoids mediated through Lck and 14: 91-4. 19. MARQUETTE CH, STACH B, CARDOT E et al.:
Fyn. Blood 2005; 106: 1703-10. 15. WILLIAMS IA, BAYLIS EM, SHIPLEY ME: High-dose and low-dose systemic corticos-
11. BUTTGEREIT F: Mechanisms and clinical A double-blind placebo-controlled trial of teroids are equally efficient in acute severe
relevance of nongenomic glucocorticoid methylprednisolone pulse therapy in active asthma. Eur Respir J 1995; 8: 22-7.
actions. Z Rheumatol 2000; 59 (Suppl. 2): rheumatoid disease. Lancet 1982; ii: 237-40. 20. WEUSTEN BL, JACOBS JW, BIJLSMA JW:
II/119-II/123. 16. WILLIAMS IA, MITCHELL AD, ROTHMAN W, Corticosteroid pulse therapy in active rheu-
12. DA SILVA JA, JACOBS JW, KIRWAN JR et al.: TALLETT P, WILLIAMS K, PITT P: Survey of matoid arthritis. Semin Arthritis Rheum
Safety of low dose glucocorticoid treatment the long term incidence of osteonecrosis of 1993; 23: 183-92.
in rheumatoid arthritis: published evidence the hip and adverse medical events in rheu- 21. AU K, REED G, CURTIS JR et al.: Extended
and prospective trial data. Ann Rheum Dis matoid arthritis after high dose intravenous report: high disease activity is associated
2006; 65: 285-93. methylprednisolone. Ann Rheum Dis 1988; with an increased risk of infection in patients
13. HANSEN TM, KRYGER P, ELLING H et al.: 47: 930-3. with rheumatoid arthritis. Ann Rheum Dis
Double blind placebo controlled trial of pulse 17. SHIPLEY ME, BACON PA, BERRY H et al.: 2011; 70: 785-91.

S-92