European Journal of Endocrinology (2013) 169 147–154 ISSN 0804-4643

CLINICAL STUDY

Subcutaneous hydrocortisone administration for emergency use

in adrenal insufficiency
Stefanie Hahner*, Stephanie Burger-Stritt* and Bruno Allolio
Endocrinology and Diabetes Unit, Department of Medicine I, University Hospital Wuerzburg, Oberdürrbacher Strasse 6, D-97080 Wuerzburg, Germany
(Correspondence should be addressed to S Hahner; Email: [email protected])
*(S Hahner and S Burger-Stritt contributed equally to this work)

Abstract
Objective: Evaluation of the pharmacokinetics and safety of s.c. hydrocortisone injection for use in
adrenal emergency.
Design: Single-center, open-label, sequence-randomized, crossover study in a tertiary care center.
Patients and methods: Twelve patients with chronic Addison’s disease. Comparison of hydrocortisone
pharmacokinetics after s.c. and i.m. injection (100 mg) and after s.c. administration of sodium chloride
(0.9%) respectively at three different visits.
Main outcome measure: maximum serum cortisol (Cmax), time to Cmax (tmax), and time to serum
cortisol O36 mg/dl (tserum cortisol O36 mg/dl) after s.c. administration compared with i.m. administration,
safety, and patient preference.
Results: Serum cortisol increased rapidly and substantially after both i.m. and s.c. injections (Cmax:
110G29 vs 97G28 mg/dl, PZ0.27, tmax: 66G51 vs 91G34 min, PZ0.17, and tserum cortisol O
36 mg/dl: 11G5 vs 22G11 min, PZ0.004 respectively). Both i.m. and s.c. injections were well
tolerated. Eleven (91.7%) patients preferred s.c. injection, whereas one patient did not have any
preference.
Conclusions: S.c. administration of 100 mg hydrocortisone shows excellent pharmacokinetics for
emergency use with only a short delay in cortisol increase compared with i.m. injection. It has a good
safety profile and is preferred by patients over i.m. injection.

European Journal of Endocrinology 169 147–154

Introduction In a prospective analysis of 455 patients, an even higher

incidence rate of 7.8 crises/100 patient years was
Patients with chronic adrenal insufficiency require observed (7). Importantly, in this study, in 3% the
lifelong replacement therapy with glucocorticoids. outcome of adrenal crisis was fatal. Furthermore,
However, even under stable replacement therapy, despite established replacement therapy and patient
patients remain at risk of life-threatening adrenal crisis education, mortality of patients with adrenal insuffi-
(1, 2). An increase in cortisol secretion is an important ciency is still increased (8, 9, 10) and the relative risk to
adaptive mechanism of the organism to deal with die from infectious disease was 6.6 in a Swedish study
stressful events like infectious disease, surgery, and also (8). Of note, adrenal insufficiency further represented
psychological distress (3, 4, 5, 6). Accordingly, adrenal the second most frequent cause of death in a Norwegian
crisis usually occurs under conditions of a relative analysis (10).
cortisol deficiency due to an increased glucocorticoid To avoid adrenal crises, patients are educated in
demand and inadequate replacement. In a retrospective adaptation of their glucocorticoid dose under conditions
analysis of 444 patients with chronic adrenal insuffi- of stress. In addition, they receive an emergency card
ciency, we found a frequency of 6.3 adrenal crises/100 and should be equipped with an emergency glucocorti-
patient years (2). Main precipitating factors were gastro- coid set. While under conditions of mild-to-moderate
intestinal and other infectious diseases. Rapid onset of stress glucocorticoid dose adaptation can be easily
symptoms was frequently reported (2). Similar obser- performed by the patients through an increase in the
vations have been made in a postal survey of Addison’s oral replacement dose. In case of vomiting and diarrhea
patients from the UK, Canada, Australia, and New or otherwise severe impairment of the clinical condi-
Zealand demonstrating that annually 8% of patients tion, parenteral administration of glucocorticoids by
experienced an adrenal crisis with gastrointestinal the attending physician is urgently required. As most
infection again being the most important trigger (1). physicians are not familiar with Addison’s disease, i.v.

q 2013 European Society of Endocrinology DOI: 10.1530/EJE-12-1057

Online version via www.eje-online.org
148 S Hahner, S Burger-Stritt and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2013) 169

glucocorticoid administration is often delayed (11), Further secondary end points were the safety of
putting the patient into danger, as serious clinical s.c. hydrocortisone administration and the response
deterioration may occur within hours. of plasma ACTH to glucocorticoid administration as a
To enable patients to better deal with imminent measure of glucocorticoid action. Finally, acceptability
adrenal crisis, physicians have started to educate of the different modes of glucocorticoid administration
patients and relatives about i.m. hydrocortisone auto- was assessed by a questionnaire.
injection. However, different from s.c. auto-injection,
self-injection into the muscle often poses a major hurdle
to patients and relatives. We, therefore, studied the Pharmacokinetic studies
feasibility and safety of s.c. hydrocortisone adminis-
tration for emergency use in adrenal insufficiency. Patients were investigated in the morning (started
between 0800 and 1100 h). The usual morning dose
of hydrocortisone was postponed until the end of the
respective visit, whereas fludrocortisone was taken as
Subjects and methods usual. After assessment of inclusion and exclusion
criteria, patients underwent a physical examination.
Patients An in-dwelling catheter was placed into a cubital vein.
All patients received sodium chloride 0.9% (divided
Twelve patients with primary adrenal insufficiency into 2!1 ml doses) as s.c. injection in the abdominal
currently registered with the outpatient department subcutaneous fat at the first visit and were then
of the University Hospital Wuerzburg agreed to randomized concerning the order of the two different
participate. Inclusion criteria were primary adrenal hydrocortisone administrations: 100 mg hydrocorti-
insufficiency under stable glucocorticoid replacement sone (Pfizer) in 2 ml solvent subcutaneously (divided
therapy due to autoimmune adrenalitis or bilateral
in 2!1 ml) and 100 mg hydrocortisone (Pfizer) in 2 ml
adrenalectomy (disease duration at least 12 months),
solvent intramuscularly in the thigh. Visits were
age R18 years, and ability to comply with the protocol
separated by a minimum of 7 days. At every study
procedures. Exclusion criteria were diabetes mellitus,
visit, blood samples were collected at K10, 0, 5, 10, 15,
current infectious disease with fever at the time of
investigation, known intolerance to the study drug or 20, 30, 45, 60, 75, 90, 120, 150, 180, 210, and
constituents of the study drug, oral contraception/ 240 min after injection. Saliva samples for determina-
oral estrogens for hormonal replacement therapy, preg- tion of salivary cortisol were collected at the same time
nancy or breastfeeding, and renal failure (creatinine points. In addition, at K10, 0, 30, 60, 90, 120, 150,
O2.5 upper limit of normal (ULN)). 180, 210, and 240 min, samples for determination of
The study was approved by the Ethics Committee plasma ACTH were collected.
of the University of Wuerzburg (permit no. 182/11_m)
and written informed consent was obtained from
all patients before participation. The study was Questionnaires and safety assessment
registered at clinicaltrials.gov (clinicaltrials.gov Patients received a diary to document any local or
identifier: NCT01450930); in addition, approval was systemic adverse events during the 3 days following the
obtained by the federal institute for drugs and intervention. After completion of all three study days,
medical devices (BfArM). EUDRACT-No.: EudraCT- patients further received a questionnaire to collect
Nr. 2011-002687-25. general information on their adrenal disease and a
The primary endpoint was defined as the time needed personal evaluation and rating of the different admin-
to reach cortisol levels of O36 mg/dl (O1000 nmol/l), istration modes. A photographic documentation of the
as such concentrations may be encountered in severe site of injection was performed after injection and at
critical illness (3, 12, 13, 14) and because 36 mg/dl is the end of the study visit.
twice the cutoff level required to pass the short
synacthen test representing a maximum ACTH stimu-
lus. The hypothesis was that after s.c. injection, such a Hormone measurements
serum level can be safely achieved within a time frame
of 30 min with higher patient acceptance compared For determination of serum cortisol, an automated
with i.m. administration. Further end points were luminescence assay was used, which is also used for
key pharmacokinetic data for hydrocortisone: Cmax, routine determination of serum cortisol levels (Cortisol
maximal serum concentration; Cmin, minimal serum Siemens Immulite 2000, Siemens Healthcare Diag-
concentration; Cav, average cortisol plasma concen- nostics, Eschborn, Germany). Intra- and interassay
tration; tmax, time to Cmax; AUCt, area under the plasma variations are !10%. Cross-reactivity for aldosterone
concentration curve from administration to last is 0.1%. Samples with values above the calibration
observed concentration at time t; mean residence time range of 50 mg/dl were reanalyzed after 1:20 dilution
(MRT); and t1/2, plasma concentration half-life. with standard solution.

www.eje-online.org
EUROPEAN JOURNAL OF ENDOCRINOLOGY (2013) 169 Subcutaneous hydrocortisone in adrenal emergency 149

ACTH levels were also determined with an automated years. BMI was 26.3 (20.8–29.8). All patients were on
luminescence assay (ACTH Siemens Immulite 2000, stable adrenal replacement therapy with hydrocortisone
Siemens Healthcare Diagnostics). Intra-assay variation 22.5 (range 15–39) mg and fludrocortisone 0.09
was !10%, and interassay variation was %10%. (0.05–0.15) mg. Eight patients further received
For determination of salivary cortisol levels, a levothyroxine replacement due to autoimmune
luminescence immunoassay was used (IBL Inter- thyroiditis (125 mg; 50–175). No patient received any
national GmbH, Hamburg, Germany). Intra- and medication known to induce cortisol-metabolizing
interassay variations were !5%. Detection range was hepatic cytochrome P450 3A4.
0.005–4 mg/dl. Levels above the detection range were
diluted 1:10 with standard solution.
Pharmacokinetics of hydrocortisone after s.c.
and i.m. administration
Pharmacokinetic analysis
Cortisol levels were undetectable or remained continu-
For pharmacokinetic analysis, the maximum (Cmax), ously below the normal range after administration of
minimum (Cmin), and average (Cav) serum and salivary sodium chloride. Pharmacokinetic parameters are given
cortisol concentrations were assessed. In addition, the in Tables 1 and 2. Maximum serum cortisol levels after
time until the maximum serum or salivary concen- administration of 100 mg hydrocortisone did not differ
tration (tmax) and the time to a serum concentration of significantly after i.m. vs s.c. injection (110G29 and
36 mg/dl (tO36 mg/dl) were determined. The area under 97G28 mg/dl respectively). Maximum serum cortisol
the concentration–time curve from zero to last sampling concentrations were reached after 66G51 and 91G
time (AUC0–240), MRT, which indicates the average 34 min respectively (PZ0.17). No statistical difference
amount of time that a compound spends in a particular was observed between the two administration routes
system, and elimination half-life (t ⁄ ) were calculated.
1
2
regarding Cav, AUC0–240, MRT, and half-life. However,
Cmax, Cmin, tmax, and tO36 mg/dl were obtained directly time until reaching serum cortisol levels above 36 mg/dl
from the observed values. Cav was calculated as mean of significantly differed (11G5 min after i.m. adminis-
all values between 5 and 240 min. AUC0–240 was tration vs 22G11 min after s.c. administration,
assessed by the trapezoidal rule. MRT was calculated as PZ0.004). Salivary cortisol levels also increased after
the area under the moment curve divided by the AUC0– both i.m. and s.c. administrations of hydrocortisone.
240. Elimination half-life was calculated using the Maximum levels did not show statistical difference
formula: t ⁄ Zln (2)/ke. The elimination rate constant
1
2 (59G26 mg/dl after i.m., 41G17 mg/dl after s.c.
was calculated as follows: keZ(ln (C1)Kln (C2))/Dt). administration, PZ0.065). Cav was significantly higher
after i.m. hydrocortisone administration (28G11 vs
20G9 mg/dl, PZ0.046). Furthermore, tmax was shorter
Statistical analysis after i.m. hydrocortisone administration (51G28 vs
Statistical analysis was performed by PASW Statistics 20 74G22 min, PZ0.036). MRT was significantly longer
(IBM SPSS, IBM Corp.). For comparison of the after s.c. hydrocortisone administration (110G12 vs
pharmacokinetics of i.m. vs s.c. administration of 91G18 min after i.m. administration, PZ0.006). Half-
100 mg hydrocortisone, a one-way ANOVA was used. life and AUC0–240 did not differ significantly. ACTH levels
The influence of BMI on pharmacokinetic parameters continuously decreased from 714G420 at baseline to
was examined by calculation of Pearson’s correlation 15G7 ng/l after i.m. hydrocortisone administration
coefficient. c2 test was used for comparison of the and 984G712 at baseline to 18G7 ng/l 240 min after
patients’ preference of administration mode. Data are s.c. hydrocortisone administration with no differences
presented as meanGS.D. or median and range. The 95% between the two modes of hydrocortisone adminis-
CI, minimal (MIN) and maximal (MAX) values, and tration (Fig. 1C).
P values were calculated. Differences were considered as A significant correlation was observed between BMI
statistically significant when P!0.05. and Cmax (rZK0.728, PZ0.007), Cav (rZK0.745,
PZ0.005), AUC (rZK0.637, PZ0.026), and MRT
(rZ0.776, PZ0.003) after s.c. administration of
hydrocortisone. No significant correlation with BMI
Results was found for tmax (rZ0.492, PZ0.104). No corre-
lation between BMI and pharmacokinetic data was seen
Study cohort
after i.m. hydrocortisone administration. Patients were
The 12 patients (five females and seven males) further divided into a group with BMI O25 kg/m2
completed all study visits. Ten patients had been (nZ7) and patients with a BMI %25 kg/m2 (nZ5).
additionally diagnosed with autoimmune thyroiditis Comparing pharmacokinetic parameters of s.c. and i.m.
and two females suffered from premature ovarian injection in the patients with BMI %25 kg/m2, no
failure. Median age was 47.5 (29–62) years, and significant differences were observed. In contrast, in the
median duration of Addison’s disease was 11 (1–42) group with BMI O25 kg/m2, significant differences

www.eje-online.org
150 S Hahner, S Burger-Stritt and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2013) 169

Table 1 Pharmacokinetic parameters after s.c. and i.m. administration of hydrocortisone (serum cortisol concentrations).

MeanGS.D. 95% CI MIN MAX P

Cmax (mg/dl)
I.m. 109.9G29.2 91.4–128.5 67.3 177.0 0.270
S.c. 96.8G27.5 79.34–114.3 50.9 153.0
Cav (mg/dl)
I.m. 74.8G20.0 62.1–87.5 48.8 117.1 0.078
S.c. 60.8G17.1 79.3–71.7 34.4 101.9
tmax (min)
I.m. 65.8G51.3 33.3–98.4 20 180 0.165
S.c. 91.3G33.5 70.0–112.5 45 150
tO36 mg/dl (min)
I.m. 10.8G5.2 7.6–14.1 5 20 0.004
S.c. 22.1G11.2 15.0–29.2 5 45
AUC0–240 (mg/dl per min)
I.m. 17 997.3G4332.3 15 244.7–20 749.9 12 789 28 207 0.318
S.c. 16 335.5G3608.2 14 043.0–18 628.1 10 208 24 244
MRT (min)
I.m. 112.1G10.5 105.5–118.8 100.4 130.8 0.116
S.c. 119.1G10.2 112.6–125.6 99.7 131.1
t ⁄ (h)
1
2

I.m. 2.2G1.5 1.3–3.1 1.0 6.4 0.900

S.c. 4.7G4.7 1.7–7.7 0.8 16.6

Cmax, maximum concentration; Cav, average concentration; tmax, time to maximum concentration; tO36 mg/dl, time to O36 mg/dl (for conversion in nmol/l multiply
with 27.6); AUC0–240, area under the concentration–time curve from zero to last sampling time; MRT, mean residence time; t ⁄ , half-life time; MIN, minimum;
1
2

MAX, maximum. Analyses were performed by one-way ANOVA. For conversion of cortisol concentration from mg/dl to nmol/l multiply with 27.6.

between s.c. and i.m. injection were observed for Cav Patient questionnaires
(53G9 vs 73G18 mg/dl for s.c. and i.m., respectively,
PZ0.02), tO36 mg/dl (27G11 vs 10G5 min, Six (50%) patients reported that they had already
PZ0.002), and MRT (125G6 vs 113G12 min, experienced an adrenal crisis since diagnosis of
PZ0.036; Fig. 2). Similarly, no differences between Addison’s disease. All patients were equipped with an
the BMI groups were observed after i.m. injection, emergency card and had received instructions regard-
whereas values for tmax and MRT differed significantly ing dose adaptation of hydrocortisone. In general, the
between the BMI groups after s.c. injection (Supple- patients rated their coping with the disease as ‘very
mentary Tables 1 and 2, see section on supplementary good’ (nZ3) or ‘good’ (nZ9). However, five patients
data given at the end of this article). indicated that they felt restricted in their leisure time

Table 2 Pharmacokinetic parameters after s.c. and i.m. administration of hydrocortisone (salivary cortisol concentrations).

MeanGS.D. 95% CI MIN MAX P

Cmax (mg/dl)
I.m. 58.7G26.0 42.2–75.2 24.5 111.2 0.065
S.c. 41.3G16.9 30.6–52.1 16.3 69.3
Cav (mg/dl)
I.m. 27. 9G10.5 21.2–34.5 13.7 49.4 0.046
S.c. 19. 6G8.5 14.2–25.0 8.1 37.0
tmax (min)
I.m. 51.3G27.5 33.8–68.7 30 120 0.036
S.c. 73.8G21.7 60.0–87.5 45 120
AUC0–24 (mg/dl per min)
I.m. 6177.9G2096. 5 4845.8–7509.9 3379.2 10 503.5 0.238
S.c. 5196.4G1862.8 4012.8–6380.0 2335.6 8329.3
MRT (min)
I.m. 91.3G17.8 80.0–102.6 65.8 128.5 0.006
S.c. 110.2G12.3 102.4–118.0 92.0 131. 3
t ⁄ (h)
1
2

I.m. 4.6G9.8 K1.7–10.8 0.4 35.3 0.666

S.c. 3.3G7.2 1.14–5.4 0.7 10.7

Cmax, maximum concentration; Cav, average concentration; tmax, time to maximum concentration; AUC0–240, area under the concentration–time curve from
zero to last sampling time; MRT, mean residence time; t ⁄ , half-life time; MIN, minimum; MAX, maximum. Analyses were performed by one-way ANOVA. For
1
2

conversion of cortisol concentration from mg/dl to nmol/l multiply with 27.6.

www.eje-online.org
EUROPEAN JOURNAL OF ENDOCRINOLOGY (2013) 169 Subcutaneous hydrocortisone in adrenal emergency 151

(a) 140 S.c. hydrocortisone

the mode of administration of hydrocortisone, 11
I.m. hydrocortisone patients preferred s.c. administration (P!0.01),
120 S.c. NaCl whereas one patient did not have any preference.
Serum cortisol (µg/dl)

100
Adverse events
80
Both the i.m. and s.c. injections were well tolerated. In
60 total, 21 adverse events were documented: after i.m.
administration of hydrocortisone, pain in the whole leg
40 of the side of injection (nZ2), local pain at the injection
site (nZ2), burning at the injection site (nZ2), and
20
local feeling of pressure at the injection site (nZ7); after
0 s.c. administration of hydrocortisone, local pain at the
–10 0 15 30 45 60 75 90 120 150 180 240 injection site (nZ1), itching at the injection site (nZ1),
Time (min) and burning at the injection site (nZ2); and after s.c.
administration of sodium chloride 0.9%, burning at
(b) 90 S.c. hydrocortisone
injection site (nZ2), local pain at the injection site
I.m. hydrocortisone
(nZ1), and itching at injection site (nZ1). All
80
S.c. NaCl symptoms were mild, did not require any intervention,
70 and resolved within 15–240 min after onset.
Salivary cortisol (µg/dl)

60

50
Discussion
40

30 Despite careful patient education, provision of an

emergency card, and an emergency steroid self-
20
injection kit, patients with adrenal insufficiency on
10 stable replacement are at risk of adrenal crisis (1, 2)
0 with evidence of increased mortality due to such crises
–10 0 15 30 45 60 75 90 120 150 180 240 (7, 8, 9, 10). This observation indicates that current
Time (min) prevention strategies need to be improved. Patients are
educated to adapt their oral hydrocortisone dose in case
(c) 1800 of stressful events. In general, high cortisol levels may be
S.c. hydrocortisone
I.m. hydrocortisone
obtained after a sufficient dose of oral hydrocortisone.
1600
S.c. NaCl However, in case of incipient adrenal crisis, these levels
1400 should be achieved within a very short time frame
1200 and enteral absorption of hydrocortisone may be
ACTH (ng/l)

1000
largely impaired, e.g. in case of gastroenteritis. Self-
administration of parenteral hydrocortisone in case of
800 impending crisis holds great potential to reduce crisis-
600 associated morbidity and mortality. Accordingly, in
400 some centers, training of i.m. self-injection of hydro-
cortisone has become part of patient education (11, 15).
200
However, many patients are reluctant to use i.m. self-
0 injection, while s.c. self-injection seems to be more
–10 0 10 45 75 120 150 180 240 acceptable to patients. Accordingly, our study provided
Time (min) clear evidence that patients prefer s.c. over i.m.
injection.
Figure 1 Comparison of s.c. and i.m. administration of hydrocorti- The major finding of our study is the high efficacy
sone; s.c. administration of sodium chloride was used as control.
(a) Curves represent mean serum cortisol concentrations vs time. of s.c. hydrocortisone administration to safely reach
Dotted line indicates the 36 mg/dl (1000 nm/l) threshold. (b) Curves target serum cortisol concentrations. When planning
represent mean salivary cortisol concentrations vs time. (c) Curves this study, we hypothesized that the time to reach
represent plasma ACTH concentrations vs time. a cortisol concentration of 36 mg/dl would allow a
reasonable evaluation of the clinical utility of s.c.
activities. Eleven of the 12 study participants were of the hydrocortisone administration. This assumption is
opinion that being able to self-inject hydrocortisone based on the established cutoff levels of the short
would make them feel safer. Nine patients indicated that synacthen test, a maximum ACTH challenge, where a
this would also improve their quality of life. Regarding cortisol increase to 18 mg/dl (500 nmol/l) after 30 min

www.eje-online.org
152 S Hahner, S Burger-Stritt and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2013) 169

(a) 160 It may be possible that in case of shock status, s.c.

BMI ≤25 kg/m2
BMI >25 kg/m2
absorption might be diminished. However, the aim of
140
training patients in performing self-injection is rather to
120 prevent shock in incipient adrenal crisis and to shorten
Serum cortisol (µg/dl)

the time interval until parenteral glucocorticoids are

100 provided by easy-to-perform and timely self-adminis-
80 tration of hydrocortisone than to treat adrenal crisis
in a status that needs professional health care with i.v.
60 glucocorticoids. The critical time frame in which we
40
believe that s.c. self-administration of hydrocortisone
has its value is incipient adrenal crisis at a time when
20 the patient is still able to treat himself/herself until
professional health care (and i.v. hydrocortisone) can be
0
–10 0 15 30 45 60 75 90 120 150 180 240 provided. Our aim is not to fully replace i.m. self-
Time (min) after s.c. injection injection, which still has its value. However, patients
may now choose their most preferred administration
mode. In our study, most patients preferred s.c. injection
(b) 160
BMI ≤25 kg/m2 over i.m. injection, although s.c. injection was given
BMI >25 kg/m2 as two single injections compared with one single i.m.
140
injection.
120
Furthermore, it might be possible that administration
Serum cortisol (µg/dl)

100 of two single doses might have changed the pharmaco-

kinetic profile in comparison with a single adminis-
80 tration. As 100 mg hydrocortisone was only available as
60 a 2 ml solution and as the volume of fluid administered
subcutaneously as a single bolus should not exceed
40 1 ml, we had to perform two simultaneous injections.
20
Nevertheless, a good absorption of hydrocortisone
after s.c. administration was observed in our study.
0 We, therefore, believe that a potential further delay of
–10 0 15 30 45 60 75 90 120 150 180 240 the increase in cortisol levels after s.c. administration of
Time (min) after i.m. injection a single injection would still by far be outweighed by the
advantages of higher patient acceptance and higher
Figure 2 Comparison of patients with a BMI %25 kg/m2 vs patients
with a BMI O25 kg/m2 after s.c. (a) and i.m. (b) administration of
probability to be used in case of emergency, which
hydrocortisone. Dotted line indicates the 36 mg/dl (1000 nm/l) may be life-saving. In fact, the cortisol concentrations
threshold. after both i.m. and s.c. hydrocortisone administrations
clearly reach supraphysiological levels even if compared
is considered to show an adequate adrenal response. with the most severe stress in patients with intact
By doubling this value, a high safety margin concerning adrenal function and remain supraphysiological for
sufficient glucocorticoid availability can be assumed. several hours, although cortisol concentrations start to
Furthermore, patients in intensive care with such decline after 120 min. However, in incipient adrenal
cortisol concentrations are considered sufficiently crisis, it may be better to provide excess glucocorticoids
protected from cortisol deficiency, and in general, the for a few hours rather than to under-replace the patient.
average serum cortisol concentration remains This approach is also a common practice in periopera-
below these concentrations in patients with critical tive glucocorticoid coverage, as current replacement
illness (3, 12, 13, 14, 16). With a mean time of 22 min strategies lead to greatly increased cortisol concen-
(95% CI 15–29 min) to reach this target concentration, trations compared with mean cortisol concentrations in
s.c. administration of hydrocortisone, therefore, clearly patients without adrenal disorders undergoing surgery.
fulfills the requirements for rapidly covering the However, unexpected complications like postoperative
increased glucocorticoid needs in case of emergency. septic episodes or severe bleeding may require high
However, compared with i.m. hydrocortisone glucocorticoid availability, which may not be adequately
administration, the time to reach this target concen- covered with a lower replacement dose and carry the
tration was significantly delayed by 11 min. The risk that the higher need in an individual patient is not
disadvantage of such a delay seems to be minor detected in time.
compared with higher ease and acceptance of s.c. It may be expected that patients would prefer one
hydrocortisone administration, which may facilitate an single injection over two injections. Further develop-
earlier self-injection compensating for the short delay ments should, therefore, aim at the development of
in serum cortisol increase. a hydrocortisone preparation, which is suitable for a

www.eje-online.org
EUROPEAN JOURNAL OF ENDOCRINOLOGY (2013) 169 Subcutaneous hydrocortisone in adrenal emergency 153

single s.c. injection. However, in our study, s.c. injection be a future goal to effectively improve crisis manage-
was still preferred over i.m. injection, although two ment in adrenal insufficiency. To better standardize the
simultaneous injections had to be performed. pharmacokinetic studies, in our study, the injections
The findings for salivary cortisol are similar to those were performed by the same investigators and not by the
for serum cortisol, indicating a rapid rise in free patients. However, in general practice, it is crucial that
bioavailable cortisol. Furthermore, suppression of patients not only receive prescription of hydrocortisone
plasma ACTH was not significantly different between ampoules but also receive structured training on how
i.m. and s.c. hydrocortisone administrations, supporting to perform self-injection.
the view of similar glucocorticoid action due to early In conclusion, our data indicate that s.c. emergency
ceiling effects after both administrations. administration of hydrocortisone is feasible and effica-
A number of pharmacokinetic parameters differed cious with regard to target serum cortisol concen-
significantly between i.m. and s.c. administrations, and trations in incipient adrenal crisis. Future studies
for other parameters, the small number of patients may should clarify how both i.m. and s.c. injections perform
have contributed to a lack of significance. This is not under conditions of manifest circulatory insufficiency.
unexpected as a slightly faster systemic availability after Subcutaneous hydrocortisone may become an import-
i.m. administration has been described for a number ant new tool to improve the self-management of
of drugs, which is mainly due to a higher perfusion patients with adrenal insufficiency in the ambulatory
of muscle compared with subcutaneous fat (17). setting.
A significant correlation was observed between BMI
and Cmax (rZK0.728, PZ0.007) but not between BMI Supplementary data
and tmax. However, tmax significantly differed between
patients with BMI O25 and %25 kg/m2. In the This is linked to the online version of the paper at http://dx.doi.org/10.
1530/EJE-12-1057.
patients with BMI %25 kg/m2, pharmacokinetic para-
meters after s.c. and i.m. injection were comparable.
In contrast, significant differences were observed for Declaration of interest
Cav, tO36 mg/dl and MRT in the overweight patients The authors declare that there is no conflict of interest that could be
(BMI O25 kg/m2). Thus, the slightly delayed cortisol perceived as prejudicing the impartiality of the research reported.
increase after s.c. hydrocortisone administration in
overweight/obese patients should be taken into account Funding
if using s.c. hydrocortisone as a crisis prevention
measure. Nevertheless, supraphysiological cortisol This work has been supported by the Else Kröner-Fresenius Stiftung
(grant no. 2010_EKES.29 to S Hahner).
levels were rapidly and consistently obtained also in
these participants after s.c. administration.
Self-administration of hydrocortisone may not only Author contribution statement
improve management of adrenal insufficiency but also S Hahner was involved in design of study, preparation of study protocol
be associated with significant risks, as the patient may and applications for authorities, conduct of the study, data analysis,
delay or cancel emergency care by a physician wrongly and preparation of the manuscript. S Burger-Stritt was involved in
believing that self-management will be sufficient. Thus, preparation of applications for authorities, conduct of the study,
laboratory measurements, and data analysis. B Allolio was involved
it is recommended that self-injection of hydrocortisone in design of the study and preparation of the manuscript.
should invariably trigger a contact to a physician for
further assessment. However, it is our experience and
that of others (11) that the time from contacting an Acknowledgements
emergency physician to the injection of hydrocortisone The authors thank Walter Rüger and the whole team of the Endocrine
frequently exceeds several hours, even if the emergency Outpatient Clinic, Department of Internal Medicine I, University
card is shown and an emergency set is provided by the Hospital Wuerzburg for their excellent cooperation.
patient, as physician attendance may be delayed and
as many physicians are reluctant to use the emergency
set for still unknown reasons. Thus, self-injection holds
the potential to greatly shorten the time to sufficient References
hydrocortisone coverage.
No ready-for-use hydrocortisone preparation is avail- 1 White K & Arlt W. Adrenal crisis in treated Addison’s disease: a
predictable but under-managed event. European Journal of Endo-
able so far; thus, preparation of hydrocortisone for crinology 2010 162 115–120. (doi:10.1530/EJE-09-0559)
injection takes time and patients may struggle to handle 2 Hahner S, Loeffler M, Bleicken B, Drechsler C, Milovanovic D,
glass ampoules with powder and separate solvent and Fassnacht M, Ventz M, Quinkler M & Allolio B. Epidemiology of
needles in case of emergency. Thus, development of adrenal crisis in chronic adrenal insufficiency: the need for new
prevention strategies. European Journal of Endocrinology 2010 162
an easy-to-handle prefilled syringe, preferably a pen, 597–602. (doi:10.1530/EJE-09-0884)
enabling the patient to quickly administer a sufficient 3 Kyle UG, Jolliet P, Genton L, Meier CA, Mensi N, Graf JD,
amount of hydrocortisone as a single injection should Chevrolet JC & Pichard C. Clinical evaluation of hormonal stress

www.eje-online.org
154 S Hahner, S Burger-Stritt and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2013) 169

state in medical ICU patients: a prospective blinded observational 11 Wass JA & Arlt W. How to avoid precipitating an acute adrenal
study. Intensive Care Medicine 2005 31 1669–1675. (doi:10.1007/ crisis. BMJ 2012 345 e6333. (doi:10.1136/bmj.e6333)
s00134-005-2832-9) 12 Sprung CL, Annane D, Keh D, Moreno R, Singer M, Freivogel K,
4 Salluh JI, Bozza FA, Soares M, Verdeal JC, Castro-Faria-Neto HC, Weiss YG, Benbenishty J, Kalenka A, Forst H et al. Hydrocortisone
Lapa ESJR & Bozza PT. Adrenal response in severe community- therapy for patients with septic shock. New England Journal of
acquired pneumonia: impact on outcomes and disease severity. Medicine 2008 358 111–124. (doi:10.1056/NEJMoa071366)
Chest 2008 134 947–954. (doi:10.1378/chest.08-1382) 13 Ray DC, Macduff A, Drummond GB, Wilkinson E, Adams B &
5 Desborough JP. The stress response to trauma and surgery. British Beckett GJ. Endocrine measurements in survivors and non-
Journal of Anaesthesia 2000 85 109–117. (doi:10.1093/bja/85.1.109) survivors from critical illness. Intensive Care Medicine 2002 28
6 Kudielka BM, Buske-Kirschbaum A, Hellhammer DH & 1301–1308. (doi:10.1007/s00134-002-1427-y)
Kirschbaum C. HPA axis responses to laboratory psychosocial 14 Goodman S, Sprung CL, Ziegler D & Weiss YG. Cortisol changes
stress in healthy elderly adults, younger adults, and children: among patients with septic shock and the relationship to ICU and
impact of age and gender. Psychoneuroendocrinology 2004 29 hospital stay. Intensive Care Medicine 2005 31 1362–1369.
83–98. (doi:10.1016/S0306-4530(02)00146-4) (doi:10.1007/s00134-005-2770-6)
7 Hahner S, Spinnler C, Beuschlein F, Fassnacht M, Lang K, 15 Allolio B, Lang K & Hahner S. Addisonian crisis in a young man
Quinkler M & Allolio B. Adrenal crisis and general morbidity in with atypical anorexia nervosa. Nature Reviews. Endocrinology
chronic adrenal insufficiency prospectively assessed in 472 2011 7 115–121. (doi:10.1038/nrendo.2010.211)
patients. Endocrine Abstracts 2011 26 OC1.5. 16 Oppert M, Reinicke A, Graf KJ, Barckow D, Frei U & Eckardt KU.
8 Bergthorsdottir R, Leonsson-Zachrisson M, Oden A & Plasma cortisol levels before and during “low-dose” hydrocorti-
Johannsson G. Premature mortality in patients with Addison’s sone therapy and their relationship to hemodynamic improve-
disease: a population-based study. Journal of Clinical Endocrinology and ment in patients with septic shock. Intensive Care Medicine 2000
Metabolism 2006 91 4849–4853. (doi:10.1210/jc.2006-0076) 26 1747–1755. (doi:10.1007/s001340000685)
9 Bensing S, Brandt L, Tabaroj F, Sjoberg O, Nilsson B, Ekbom A, 17 Voortman G, Mannaerts BM & Huisman JA. A dose proportionality
Blomqvist P & Kampe O. Increased death risk and altered cancer study of subcutaneously and intramuscularly administered
incidence pattern in patients with isolated or combined auto- recombinant human follicle-stimulating hormone (Follistim*/
immune primary adrenocortical insufficiency. Clinical Endo- Puregon) in healthy female volunteers. Fertility and Sterility
crinology 2008 69 697–704. (doi:10.1111/j.1365-2265.2008. 2000 73 1187–1193. (doi:10.1016/S0015-0282(00)00542-2)
03340.x)
10 Erichsen MM, Lovas K, Fougner KJ, Svartberg J, Hauge ER,
Bollerslev J, Berg JP, Mella B & Husebye ES. Normal overall
mortality rate in Addison’s disease, but young patients are at risk Received 5 December 2012
of premature death. European Journal of Endocrinology 2009 160 Revised version received 15 April 2013
233–237. (doi:10.1530/EJE-08-0550) Accepted 14 May 2013

www.eje-online.org