Veterinary Anaesthesia and Analgesia, 2016, 43, 44–54 doi:10.1111/vaa.

12275

RESEARCH PAPER

Stress-related biomarkers in dogs administered regional

anaesthesia or fentanyl for analgesia during stifle surgery

Marta Romano*,†, Diego A Portela*,‡, Gloria Breghi* & Pablo E Otero§

*Department of Veterinary Clinics, Veterinary Teaching Hospital ‘Mario Modenato’, University of Pisa, Pisa, Italy
†Clinica Veterinaria Apuana, Marina di Carrara, MS, Italy
‡Department of Clinical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY, USA
§Department of Anesthesiology, Faculty of Veterinary Science, University of Buenos Aires, Buenos Aires, Argentina

Correspondence: Marta Romano, Department of Clinical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853-
6401, USA. E-mail: marta.romano@me.com

Glasgow Composite Measure Pain Scale (Short-

Abstract
Objectives To compare the effects of regional anaes-
thesia and opioid administration on adrenocortical
and glycaemic responses, postoperative pain and
recovery quality in dogs undergoing stifle surgery.
Study design Prospective, blinded clinical study.
Animals Forty-five dogs anaesthetized for tibial
tuberosity advancement or tibial plateau levelling
osteotomy, and 15 healthy dogs undergoing non-
invasive orthopaedic diagnostic procedures.
Methods The baseline behaviour of each dog was
classified on a descriptive scale before anaesthesia.
Dogs were anaesthetized using a standard protocol.
Those undergoing surgery were randomly assigned
to one of three intraoperative analgesia protocols: 1)
peripheral (femoral and sciatic) nerve block (bupi-
vacaine 0.5%; 0.15 mL kg 1 in each nerve; group
PNB); 2) spinal anaesthesia (isobaric bupivacaine
0.5%; 0.05 mL kg 1; group SPI); and 3) intrave-
nous fentanyl variable rate infusion and postopera-
tive methadone (0.2 mg kg 1; group FEN). Blood
samples were collected for measurement of cortisol
and glucose concentrations on arrival (TAR), at
induction (TIND), at tracheal extubation (TEXT), and
1 hour post-extubation (TPEX). After extubation, a
researcher unaware of the dog’s group assignment
evaluated pain and recovery quality using the
Form) and a descriptive scale, respectively.
Results Median recovery quality was significantly
worse in group FEN (p < 0.0001) compared with all
other groups. Postoperative pain scores were lower
in group PNB compared with groups SPI and FEN.
Cortisol and glucose concentrations increased sig-
nificantly from TAR at TEXT and TPEX (p < 0.0001) in
group FEN, and were unchanged from TAR in the
other groups.
Conclusions and clinical relevance Analgesia with
a peripheral nerve block or spinal anaesthesia
prevented the glycaemic and cortisol responses to
surgery, promoted better recovery quality, and
decreased postoperative pain scores compared with
FEN. In the present study, the regional anaesthesia
techniques used were found to be excellent alterna-
tives to fentanyl administration.
Keywords dog, peripheral nerve block, regional
anaesthesia, spinal anaesthesia, stress.
Introduction
Trauma and surgery are potent triggers of a neuro-
humoral stress response in all animals (Kehlet
1989). This response developed in nature to allow
injured animals to survive by catabolizing their own
stored body fuels. However, it has been argued that

44
Stress and regional anaesthesia in dogs M Romano et al.
the stress response may be unnecessary and even
harmful in patients undergoing surgery (Desbor-
ough 2000). In fact, studies in humans have shown
that a perioperative stress response has adverse
effects on immune function, which may increase
postoperative susceptibility to infections. Moreover,
stress can predispose to prolonged ileus and
hypercoagulability, increase the risk for ischaemia–
reperfusion injury and determine systemic inflam-
matory responses (Liu et al. 1995; Wolf 2012).
Fentanyl is commonly used to provide periopera-
tive analgesia in dogs (Pascoe 2000; Lamont &
Mathews 2007). However, this drug is associated
with several side effects that include bradycardia,
hypotension, hypoventilation, ileus, nausea, vomit-
ing and dysphoria (Lamont & Mathews 2007;
Becker et al. 2012; Keating et al. 2013).
A large number of studies have shown that
neuraxial anaesthesia can prevent the endocrine
and metabolic response to surgery (Wolf et al. 1993;
Liu et al. 1995; Meissner et al. 1997; Sibanda et al.
2006). Spinal (intrathecal) anaesthesia provides a
profound degree of blockade of both afferent
impulses from the surgical site and efferent auto-
nomic pathways to the liver and the adrenal
medulla, thus abolishing the adrenocortical and
glycaemic responses to surgery (Wolf et al. 1998;
Wolf 2012). A recent meta-analysis suggests that
regional anaesthesia in humans may improve post-
operative outcomes by reducing morbidity after
surgery, especially in critically ill patients (Kettner
et al. 2011).
In humans, severe regional anaesthesia-related
complications are rare (Auroy et al. 1997), but
higher incidences of complications, including nerve
damage and hypotension, are associated with spinal
anaesthesia. A prospective survey in children
revealed that overall complication rates were six
times higher after central than after peripheral nerve
blocks (Ecoffey et al. 2010). Urinary retention,
pruritus, respiratory depression and nausea are
associated with the intrathecal administration of
opioids as adjuvant analgesics (Auroy et al. 1997;
Sarotti et al. 2011).
Peripheral nerve blocks (PNB) and spinal anaes-
thesia are techniques that reliably provide effective
regional anaesthesia and analgesia during pelvic
limb surgical procedures in dogs (Campoy et al.
2012; Sarotti et al. 2012; Portela et al. 2013;
Vettorato et al. 2013).
No studies have been performed to assess the
effects of PNB on the adrenocortical and glycaemic
45 © 2015 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 44–54
responses after orthopaedic surgery in dogs. The aim
of the present study was to compare the effects of
three analgesic techniques (PNB, spinal anaesthesia
and opioid administration) on adrenocortical and
glycaemic responses in dogs undergoing stifle sur-
gery in a clinical setting to test the hypothesis that
PNB and spinal anaesthesia would reduce the
adrenocortical and glycaemic responses to surgery
in the selected population compared with opioid
administration.
Materials and methods
Animals
This prospective, blinded clinical study was con-
ducted in accordance with national legislation on
the protection of animals used for scientific purposes
and was approved by the Institutional Animal Care
and Use Committee of the University of Pisa (protocol
no. 1397).
Forty-five dogs admitted to the Clinica Veterinaria
Apuana (CVA) for tibial tuberosity advancement
(TTA) or tibial plateau levelling osteotomy (TPLO)
and 15 healthy dogs undergoing planned non-
invasive orthopaedic diagnostic procedures (com-
puted tomography or radiography) under general
anaesthesia were included in this study after written
informed consent had been obtained from their
owners. Physical, haematological and biochemical
evaluations were used to assess the health status of
each animal; dogs classified as having American
Society of Anesthesiologists (ASA) physical status of
class I or II were enrolled in the study. Animals
weighing <6 kg or >40 kg, aged <1 year or
>10 years, assigned a pre-anaesthetic behaviour
score of 4, receiving concurrent medications, with
clotting, neuromuscular or neurological disorders,
with skin infections, or with an ASA physical status
of class III or higher were excluded from the study.
Fifteen healthy dogs with no sign of pain or
lameness undergoing official hip and elbow dyspla-
sia diagnostic procedures were included as a control
group (group CTR). Forty-five dogs scheduled for
surgery were randomly assigned into three groups of
15 dogs each by withdrawing pieces of paper with
group identifications from a bag. The three groups,
respectively, were to be administered the following
analgesic protocols: 1) PNB of the femoral and sciatic
nerves (group PNB); 2) spinal (intrathecal) analgesia
(group SPI); and 3) fentanyl variable rate intrave-
nous (IV) infusion (group FEN).
 Stress and regional anaesthesia in dogs M Romano et al.
A descriptive scale, adapted from Becker et al.
(2012), was used to describe each dog on the basis of
its behaviour on arrival at the hospital (Table 1).
Anaesthetic management
Food, but not water, was withheld from all animals
for 12 hours prior to anaesthesia. A catheter (20 or
18 gauge; Delta-Ven 1; Delta Med SpA, Italy) was
aseptically placed in a cephalic vein, and all dogs
were administered methadone (0.1 mg kg 1; Epta-
done; Molteni & C Flli Alitti SpA, Italy) IV and
carprofen (4 mg kg 1; Rimadyl; Pfizer Italia Srl,
Italy) subcutaneously as preanaesthetic medication.
Five minutes after methadone administration, gen-
eral anaesthesia was induced with propofol (4–
6 mg kg 1; 10 mg mL 1, Propofol Kabi; Fresenius
Kabi Srl, Italy) IV and maintained with isoflurane
(Isoflo; Esteve SpA, Italy) in oxygen delivered
through a semi-closed rebreathing system after
orotracheal intubation. Lactated Ringer’s solution
was infused at 3–5 mL kg 1 hour 1 throughout
anaesthesia. A catheter (22 or 20 gauge; Delta-Ven
1) was placed in a dorsal pedal artery to facilitate
measurements of arterial pressure. Continuous elec-
trocardiography (ECG), heart rate (HR), respiratory
rate (fR), pulse oximetry, invasive systemic arterial
pressure (sAP), end-tidal carbon dioxide partial
Table 1 Pre-anaesthesia and recovery behaviour scoring
criteria (modified from Becker et al. 2012)
Score Behaviour
Pre-anaesthesia score
1 Dog is calm, may be alert and attentive (no
panting/pacing/barking)
2 Mildly excited when someone is present (barks/
paces/pulls on leash but can be controlled)
3 Excited or exuberant (jumps up/barks/paces/pants/
licks lips), but calms down with attention
4 Whines, attempts to escape from kennel, is
aggressive; personnel are unable to interact with
dog to calm it
Recovery score
1 Dog is quiet or raises head calmly, does not appear
agitated
2 Dog transiently pants or whines or gently paddles
with front feet immediately upon extubation, but
then settles
3 Dog occasionally pants/whines/whimpers
4 Dog is agitated, tries to bite or thrashes body in an
uncoordinated manner; non-responsive to
personnel
© 2015 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 44–54
pressure (PE′CO2) and end-tidal isoflurane concen-
tration (FE′Iso) were monitored throughout anaes-
thesia using a pre-calibrated multiparametric
monitor (Criticare POET Plus 8100; Criticare Sys-
tem, Inc., WI, USA). Body temperature was main-
tained above 37 °C using electrically heated and
thermal foil blankets. For sAP measurements, the
transducer was zeroed at the level of the right
atrium, assumed as the sternum manubrium, and
non-compliant tubing flushed with heparinized
saline solution was used to connect the arterial
catheter to the transducer. Volume-controlled ven-
tilation was provided to maintain PE′CO2 at 35–
45 mmHg (4.5–6.0 kPa).
Analgesic protocols
After induction of general anaesthesia, the hair of
the entire surgical limb and lumbar and gluteal
regions was clipped in all dogs submitted to surgery.
Dogs in group PNB were positioned in lateral
recumbency with the surgical limb uppermost and
the skin over the lumbosacral region was aseptically
prepared. The femoral and sciatic nerves were
blocked with lateral pre-iliac and parasacral
approaches, respectively (Portela et al. 2010,
2013) using unipolar insulated needles (Stimuplex;
B Braun Melsungen AG, Germany) connected to a
nerve stimulator (Stimuplex HNS12; B Braun Mels-
ungen AG). When the appropriate muscular
response was observed with 0.3–0.5 mA,
0.15 mL kg 1 of 0.5% bupivacaine (5 mg mL 1,
Bupisolver; Pierrel Farmaceutica SpA, Italy) was
injected at both the femoral and sciatic nerve sites.
Dogs in group SPI were positioned in lateral
recumbency with the pelvic limbs pulled cranially.
The lumbosacral region was aseptically prepared,
and a 22 gauge, 90 mm spinal needle (Terumo
spinal needle; Terumo Italia Srl, Italy) was inserted
using a paramedian approach at the L5–L6 lumbar
intervertebral space. When the needle was con-
firmed as being correctly positioned by the free flow
of cerebrospinal fluid (CSF) in the hub of the needle,
0.05 mL kg 1 0.5% isobaric preservative-free bup-
ivacaine (5 mg mL 1, Bupisolver; Pierrel Farma-
ceutica SpA) was injected over 30–40 seconds.
Dogs were then positioned for surgery, and FE′Iso
1.2% was maintained for 10 minutes. Baseline (TB)
measurements of HR, mean arterial pressure (MAP)
and FE′Iso were recorded. The surgical procedure
started 15 minutes later (T0). HR, MAP and FE′Iso
were recorded at T0, at 5 minutes after the
46
Stress and regional anaesthesia in dogs M Romano et al.
beginning of surgery (T5), at half-time of the
duration of surgery (T1/2) and at the end of surgery
(TEND). The half-time point of surgery was deter-
mined retrospectively from the anaesthetic record for
each dog.
Dogs in group FEN were administered a loading
dose of fentanyl (5 lg kg 1; Fentanest; Pfizer Italia
Srl) IV 15 minutes before the surgical incision (TB),
which was immediately followed by an infusion at
10 lg kg 1 hour 1. Intraoperatively, the infusion
rate was adjusted in response to changes in HR and
MAP to maintain the HR and MAP within 20% of
baseline values. The fentanyl infusion was reduced
to 3 lg kg 1 hour 1 at the start of wound closure,
and methadone (0.2 mg kg 1) was administered
intramuscularly (IM). Fentanyl administration was
stopped at the end of the surgery.
Dogs in group CTR received no medication other
than that described in the anaesthetic protocol. After
positioning for imaging, anaesthesia was stabilized
at FE′Iso 1.2% for 10 minutes and HR and MAP were
recorded as baseline (TB).
Intraoperative management
Dogs were monitored continuously to detect spon-
taneous movements or any change in cardiovascu-
lar variables in response to surgical stimulations.
Once the procedure began, FE′Iso was gradually
reduced by 0.1–0.2% every 10 minutes, until the
minimal concentration preventing movement was
reached. If spontaneous movements were detected,
propofol (0.5–1.0 mg kg 1) was administered IV
and FE′Iso was increased to the previous concentra-
tion. If HR or MAP increased by more than 20%
compared with TB, fentanyl (2.0 lg kg 1) was
administered IV over 30 seconds. If the variables
did not return to the pre-stimulation values or
increased again after 10–15 minutes, another bolus
of 2.0 lg kg 1 fentanyl was administered IV, fol-
lowed by a continuous infusion in SPI and PNB, or
adjustment of the fentanyl infusion rate in FEN.
Vaporizers for all animals were adjusted by the same
person (MR), who was aware of each dog’s group
assignment.
Postoperative evaluation
After tracheal extubation, each dog was positioned
in a cushioned kennel. Evaluation started when the
dog was observed to be aware of its surroundings.
Recovery quality scores were based on behaviours
47 © 2015 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 44–54
displayed shortly after tracheal extubation, using a
descriptive scale (Table 1). The postoperative eval-
uation was not blinded in CTR dogs because the hair
was not clipped on these dogs. Postoperative pain
evaluations in dogs that had undergone a surgical
procedure were performed every 15 minutes for
60 minutes by a researcher who was unaware of
the analgesic technique employed, using the short
form of the Glasgow Composite Measure Pain
Scale (GCMPS) (Reid et al. 2007). Methadone
(0.1 mg kg 1) IV was administered as rescue anal-
gesia if pain scores were greater than 5/20. How-
ever, if the pain score was 10/20 or higher, the
methadone rescue dose was 0.15 mg kg 1 IV (based
on the authors’ clinical experience with the GCMPS
Short-Form).
All dogs in SPI, PNB and FEN were administered
tramadol (0.3 mg kg 1; Altadol; Formevet Srl, Italy)
and carprofen (2 mg kg 1; Pfizer Italia Srl) orally
every 4–6 hours and every 24 hours, respectively,
for 96 hours after surgery. Dogs in PNB and SPI
were clinically evaluated during a follow-up period
of 30 days to identify any neurological deficits.
Sample collection
Blood samples were collected from all dogs on arrival
at the hospital (TAR), at induction (TINT), at tracheal
extubation (TEXT), and at 1 hour post-extubation
(TPEX) for measurement of serum cortisol and blood
glucose concentrations. The TEXT and TPEX samples
were collected after pain score evaluations, but
before the administration of methadone in the dogs
that required rescue analgesia at these time points.
Glucose concentration was measured immediately
after blood collection with a glucometer (Abbot
Freestyle Optium; Abbott Laboratories Ltd, UK).
A 5 mL aliquot of blood was dispensed into a serum
separator tube. Immediately after clot formation
occurred, the sample was centrifuged at 3000 g for
10 minutes, and serum was separated and stored at
80 °C in labelled Eppendorf tubes. Cortisol con-
centration was measured by a solid phase radioim-
munoassay (ST Aia Pack; Tosoh Bioscience, Japan).
The personnel measuring the serum cortisol and
blood glucose concentrations were unaware of the
anaesthetic protocol assigned to each dog.
Statistical analysis
Data were evaluated for normal distribution using
the Shapiro–Wilk normality test. The distributions of
 Stress and regional anaesthesia in dogs M Romano et al.

gender and type of surgery among groups were
analysed using chi-squared tests. Age, weight, cor-
tisol and blood glucose concentrations, HR and MAP
were compared among groups using one-way
analysis of variance (ANOVA) and Bonferroni’s post
hoc test. Differences among data obtained at the
different time-points within each group were anal-
ysed using ANOVA for repeated measures and Bonfer-
roni’s post hoc test. FE′Iso, preoperative behaviour,
postoperative dysphoria and pain scores were com-
pared using the Kruskal–Wallis test and Dunn’s post
hoc test for multiple comparisons. Methadone doses
administered during the postoperative period were
compared using a Mann–Whitney test. Differences
were considered statistically significant at p < 0.05.
Parametric data were expressed as the
mean  standard deviation (SD) and non-paramet-
ric data as the median (minimum–maximum).
Statistical tests were performed using Prism Version
6.0 (GraphPad Software, Inc., CA, USA).
Results
Weight, gender, type of surgery and preanaesthesia
behaviour scores did not differ significantly among
groups (Table 2). The duration of procedures and
dog ages differed significantly (p < 0.0001) between
group CTR and the other groups (Table 2).
Appropriate muscular twitches after femoral and
sciatic nerve electrolocation were easily achieved in
all dogs in PNB. Freely flowing CSF was clearly seen
in the hub of the spinal needle in all dogs in SPI. No
adverse reactions were observed after injection of
bupivacaine. None of the dogs enrolled in PNB and
SPI showed signs of nociception during needle
insertion or local anaesthetic solution injection and
no complications associated with the PNB or SPI
protocols were observed. All dogs recovered motor
function after the offset of the local anaesthetic drug
and no neurological deficits were observed during
the 1 month follow-up period.
Two dogs in PNB and one dog in SPI required a
single IV bolus of fentanyl (2 lg kg 1) to treat
increases in HR and MAP of > 20% over baseline.
Within each group, no significant temporal change
in HR was detected at the different time points (p-
values were 0.51, 0.09, 0.05 and 0.99 in the PNB,
SPI, FEN and CTR groups, respectively) (Table 3).
Comparisons among groups showed that HR was
significantly lower in FEN at T1/2 compared with SPI
(p = 0.02) and at TEND compared with PNB and SPI
(p = 0.005) (Table 3). No significant differences
were obtained for MAP at the different time points
within each group (p values were 0.16, 0.14, 0.06
and 0.18 for PNB, SPI, FEN and CTR, respectively)
(Table 3). MAP was significantly lower in SPI
compared with CTR at T1/2 (p = 0.004) and at TEND
(p = 0.01) (Table 3). MAP remained above
65 mmHg in all animals in the study.
No significant differences in FE′Iso at the different
time points were detected in FEN and CTR (p = 0.91
and p = 0.82, respectively). In PNB and SPI, FE′Iso

Table 2 Demographic data, type of surgery and duration of anaesthesia of dogs undergoing non-invasive diagnostic
procedures [control (group CTR)] or orthopaedic surgery with the following analgesic protocols: peripheral (femoral and
sciatic) nerve blocks (group PNB); spinal (intrathecal) analgesia (group SPI), or variable rate intravenous infusion of fentanyl
(group FEN)

Group

Variable PNB SPI FEN CTR

Dogs (n) 15 15 15 15
Age (years, mean  SD) 6.2  2.4 6.0  2.5 5.6  2.5 1.2  0.2*
Weight (kg, mean  SD) 25.1  9.3 27.3  10.0 26.5  8.8 25.7  10.8
Males (n) 8 9 7 11
Females (n) 7 6 8 4
Pre-anaesthesia behaviour score (median and range) 1 (1–2) 1 (1–2) 1 (1–2) 1 (1–2)
TPLO (n) 12 14 11
TTA (n) 3 1 4
Duration of anaesthesia (minutes, mean  SD) 151  45 160  47 173  51 75  9*

SD, standard deviation; TPLO, tibial plateau levelling osteotomy; TTA, tibial tuberosity advancement. *Significantly different from other
groups (p < 0.0001).

© 2015 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 44–54 48
Stress and regional anaesthesia in dogs M Romano et al.

Table 3 Intraoperative heart rate (HR), mean arterial blood pressure (MAP) and end-tidal isoflurane (FE′Iso) in dogs
undergoing non-invasive diagnostic procedures [control (group CTR)] or orthopaedic surgery with different analgesic
protocols: peripheral (femoral and sciatic) nerve blocks (group PNB); spinal (intrathecal) analgesia (group SPI), or variable
rate intravenous infusion of fentanyl (group FEN) (n = 15 dogs in each group)

Time point

Variable Group TB T0 T5 T1/2 TEND

HR (beats minute 1)
PNB 82  23 81  21 82  22 81  22 85  22
SPI 84  17 85  16 87  18 86  17 89  17
FEN 69  14 68  14 70  15 66  16* 65  14†
CTR 79  18 79  15 81  15 79  16 78  16
MAP (mmHg)
PNB 80  9 79  9 78  9 77  8 79  8
SPI 74  8‡ 73  9§ 73  8 75  9‡ 75  10§
FEN 77  9 77  11 76  10 77  10 79  8
CTR 85  7 84  7 85  6** 87  7 86  7
F E′Iso (%)
PNB 1.2  0a 1.2  0a 1.2  0a 0.9  0.2††b 0.9  0.2††b
SPI 1.2  0a 1.2  0a 1.2  0a 1  0.1††b 0.9  0.2††b
FEN 1.2  0 1.2  0 1.2  0 1.2  0.1 1.1  0.1
CTR 1.2  0 1.2  0 1.2  0 1.2  0.1 1.1  0.1

TB, 15 minutes before the beginning of the surgical or diagnostic procedure; T0, beginning of the procedure; T5, 5 minutes after the
beginning of the procedure; T1/2, half time of the duration of the procedure; TEND, end of the procedure. Different superscript letters
indicate significant differences within the same group (p < 0.0001). *Significantly different from SPI at the same time point (p = 0.02).
†Significantly different from PNB and SPI at the same time point (p = 0.005). ‡Significantly different from CTR at the same time point
(p = 0.004). §Significantly different from CTR at the same time point (p = 0.01). **Significantly different from SPI and FEN at the same
time point (p = 0.002). ††Significantly different from FEN and CTR at the same time point (p < 0.0001).

was significantly lower at T1/2 and TEND compared
with TB, T0 and T5 within the same group
(p < 0.0001) and at T1/2 and TEND compared with
values recorded in FEN and CTR at the same time
points (p < 0.0001) (Table 3). The mean fentanyl
infusion rate in FEN was 18.8  6.3 lg kg 1
hour 1.
Median recovery scores were 1 (range: 1–3) in
PNB, 1 (range: 1–4) in SPI, 2 (range: 2–4) in FEN
and 1 (range: 1–2) in CTR, and thus were
significantly higher in FEN (p < 0.0001) compared
with all other groups. No differences in recovery
scores were observed among PNB, SPI and CTR
(p = 0.21).
Median pain scores at the first postoperative
evaluation were 2 (range: 1–4) in PNB, 4 (range:
1–9) in SPI and 3 (range: 1–12) in FEN, and thus
were significantly lower in PNB than in SPI and FEN
(p = 0.0064) (Fig. 1). Median pain scores were
significantly lower in PNB than in SPI at 45 minutes
[3 (range: 1–4) versus 4 (range: 1–12); p = 0.021]
and at 60 minutes [3 (range: 1–4) versus 6 (range:
1–11); p = 0.024] (Fig. 1). No significant differences
in postoperative pain scores emerged between SPI
and FEN (p = 0.15). Analyses of pain scores at the
different time points within the groups detected no
significant differences.
None of the dogs in PNB were administered
methadone during the 1 hour postoperative evalu-
ation period. All dogs in SPI and FEN were admin-
istered methadone at different time points within this
period. Significantly more methadone was adminis-
tered to dogs in FEN [median: 0.3 mg kg 1 (range:
0.2–0.4 mg kg 1)] than to dogs in SPI [median:
0.1 mg kg 1 (range: 0.1–0.35 mg kg 1)] (p =
0.0002).
Mean  SD blood glucose concentrations at TAR
in PNB and SPI were 85  10 mg dL 1 and
91  8 mg dL 1, respectively, and were signifi-
cantly increased at TEXT to 99  9 mg dL 1
(p < 0.0001) and 99  12 mg dL 1 (p = 0.0016),
respectively (Fig. 2). In FEN, glucose concentrations
at TAR were 89  12 mg dL 1 and significantly
increased to 114  14 mg dL 1 at TEXT and

49 © 2015 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 44–54
 Stress and regional anaesthesia in dogs M Romano et al.

Figure 1 Postoperative pain scores evaluated for 60 minutes starting from the regaining of consciousness (T0) using the
short form of the Glasgow Composite Measure Pain Scale (maximum score possible: 20) in dogs submitted to femoro-tibial
joint surgery (tibial tuberosity advancement or tibial plateau levelling osteotomy) and administered femoral and sciatic nerve
blocks (group PNB), spinal (intrathecal) anaesthesia (group SPI) or a variable rate intravenous infusion of fentanyl (group
FEN) as analgesic protocols (n = 15 per group). Boxes represent interquartile ranges; the horizontal line within the box
represents the median value, and the upper and lower whiskers represent maximum and minimum values. *Significantly
lower compared with SPI and FEN (p = 0.0064). †Significantly lower compared with SPI (p = 0.021). ‡Significantly lower
compared with SPI (p = 0.024).

Figure 2 Peri-anaesthetic blood glucose (mg dL 1) concentrations in dogs undergoing anaesthesia for non-invasive
diagnostic procedures (group CTR) or orthopaedic surgery (tibial plateau levelling osteotomy or tibial tuberosity
advancement) with different analgesic protocols (n = 15 dogs per group). Group PNB: peripheral femoral and sciatic nerve
blocks; group SPI: spinal analgesia; group FEN: variable rate intravenous infusion of fentanyl. Time points: TAR, before
anaesthesia; TINT, after endotracheal intubation; TEXT, after extubation; TPEX, 1 hour after extubation. *Significant
differences compared with PNB, SPI and CTR (p < 0.0001). †Significant differences compared with TAR (p < 0.0001).
‡Significant differences compared with TAR (p = 0.0016).

© 2015 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 44–54 50
Stress and regional anaesthesia in dogs M Romano et al.

Figure 3 Peri-anaesthetic serum cortisol concentrations (lg dL 1) in dogs undergoing anaesthesia for non-invasive
diagnostic procedures (group CTR) or orthopaedic surgery (tibial plateau levelling osteotomy or tibial tuberosity
advancement) with different analgesic protocols (n = 15 dogs per group). Group PNB: peripheral femoral and sciatic nerve
blocks; group SPI: spinal analgesia; group FEN: variable rate intravenous infusion of fentanyl. Time points: TAR, before
anaesthesia; TINT, after endotracheal intubation; TEXT, after extubation; TPEX, 1 hour after extubation. *Significant
differences compared with TAR (p < 0.0001). ‡Significantly lower compared with TINT (p = 0.008). †Significantly higher
compared with CTR, PNB and SPI (p < 0.0001).

112  12 mg dL 1 at TPEX (p < 0.0001). No

significant differences were detected at any time
point in CTR (p = 0.11). When blood glucose
concentrations were compared among groups, con-
centrations were significantly higher in FEN than in
PNB, SPI and CTR at TEXT and TPEX (p < 0.0001)
(Fig. 2).
Cortisol concentrations did not increase signifi-
cantly from TAR to TINT in any group, and not at any
time point in either PNB (p = 0.18) or SPI (p = 0.16)
(Fig. 3). In FEN, cortisol concentrations were signif-
icantly increased at TEXT (9.1  1.6 lg dL 1) and
TPEX (10.8  1.4 lg dL 1) (p < 0.0001). Cortisol
concentration was significantly decreased in CTR at
TPEX compared with TINT (p = 0.008) (Fig. 3). A
comparison among groups revealed that serum
cortisol concentration in FEN was significantly
higher at TEXT and TPEX than in PNB, SPI and CTR
(p < 0.0001) (Fig. 3).
In the two animals in PNB in which adminis-
tration of fentanyl for intraoperative rescue anal-
gesia was required, neither glucose nor cortisol
concentrations changed significantly at any time
point.
Discussion
The results obtained in the present study show that
blood glucose and serum cortisol concentrations did
not change in association with surgery when PNB or
spinal anaesthesia analgesic protocols were used in
dogs undergoing pelvic limb orthopaedic surgery,
and concentrations were similar to those measured
in the control group. However, when fentanyl
infusion was administered intraoperatively, blood
glucose and serum cortisol concentrations were
significantly higher at extubation and 1 hour later
compared with those in all other groups.
Nociception is a potent trigger for the stress
response and can therefore cause remarkable
increases in serum cortisol and blood glucose con-
centrations (Kehlet 1989), with two- to four-fold
increases measured in dogs (Naitoh et al. 2002;
Devitt et al. 2005; Sibanda et al. 2006). Increases in
HR and MAP in response to surgical manoeuvres are
commonly used to identify nociception in patients.
In this study, the analgesia protocols and provision
of rescue analgesia provided control of HR and MAP
increases in all groups. However, significant

51 © 2015 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 44–54
 Stress and regional anaesthesia in dogs M Romano et al.
increases in blood glucose and serum cortisol con-
centrations were detected at extubation in dogs in
FEN. All animals were administered methadone and
carprofen for preanaesthetic medication, and all
group FEN animals were administered methadone at
the start of surgical closure. One possible explana-
tion for the increased blood glucose and cortisol
concentrations in FEN could be that cardiovascular
response may not be a good indicator of nociception,
and although no sympathetic response was evoked
by surgical stimulation, nociception, and therefore
the activation of the stress response, may still be
present. Another possible explanation may refer to
the different mechanism of action of the analgesic
effect of opioids compared with the mechanism of
regional anaesthesia on afferent pain transmission
(Abram & Olson 1994; Lamont 2008). In fact, local
anaesthetic agents administered perineurally or
intrathecally are able to block nerve conduction of
afferent activity, thereby stopping afferent neural
traffic before it can reach higher levels in the central
nervous system (CNS), whereas opioids modulate
synaptic neurotransmission of nociceptive traffic
once it has reached the CNS (i.e. the spinal cord
dorsal horn) (Lamont 2008). The less complete
blockade of afferent traffic at this level may allow
nociceptive signalling to reach the hypothalamic–
pituitary–adrenal (HPA) axis. This axis may still be
activated despite fentanyl administration, which
may explain the increases in cortisol and glucose
concentrations observed in the present study (Liu
et al. 1995).
In humans, opioids suppress the stress response to
surgery only at very high doses (e.g. morphine
4 mg kg 1 or fentanyl 50–100 lg kg 1), which are
also associated with a high incidence of side effects
(Liu et al. 1995; Desborough 2000). In the present
study, at the clinical doses used, fentanyl, combined
with methadone and carprofen, failed to control
glycaemic and adrenal responses to surgery.
The complete suppression of adrenal and glycae-
mic responses found in the present study in the
groups receiving peripheral or neuraxial regional
anaesthesia is in accordance with previous reports
on spinal anaesthesia in humans (Wolf et al. 1998)
and on epidural analgesia in dogs (Kona-Boun et al.
2006). The suppression of the adrenal response that
can be achieved with spinal anaesthesia is explained
by its double-action mechanism: afferent nociceptive
neural traffic is blocked before it reaches the CNS,
and sympathetic efferent fibres that innervate the
adrenal glands are also blocked, which may explain
© 2015 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 44–54
why spinal anaesthesia is very effective in control-
ling the stress response (Desborough 2000; Wolf
2012). However, no glycaemic or adrenal responses
occurred in group PNB. In the absence of a direct
effect of PNB on the sympathetic efferent branches to
the adrenal glands, suppression of the glycaemic and
adrenal responses is more likely to represent the
result of interrupted afferent nociceptive neural
activity to the CNS.
Although this study was not designed to evaluate
the isoflurane-sparing effect of the analgesic proto-
cols, it was apparent that less isoflurane was
required when it was administered in association
with regional anaesthesia in comparison with the
administration of a fentanyl infusion. Regional
anaesthesia will completely block the nociceptive
afferents so that less isoflurane is required to achieve
unconsciousness and avoid movements (Zhang
et al. 2007).
The hospital environment can be a potential cause
of stress in dogs (V€
ais€anen et al. 2005). However, in
the study reported here, serum cortisol and blood
glucose concentrations measured in dogs on arrival
at the hospital were within the normal limits
described in dogs (Nelson 2006) and were not
increased immediately after induction of anaesthe-
sia.
The GCMPS is a validated tool for determining
acute pain in dogs (Reid et al. 2007) and was used to
establish the need for rescue analgesia in the present
study. The postoperative pain score was significantly
lower in PNB and none of the dogs in this group
required rescue analgesia in the first hour post-
surgery. A higher dose of methadone was adminis-
tered for rescue analgesia in FEN than in SPI in the
first hour after endotracheal extubation.
The pain scale used in this study indicated that
postoperative pain was clinically controlled in all
animals, despite increases in cortisol and glucose
concentrations in FEN. It could be hypothesized that
the pain scale used was insufficiently sensitive to
detect pain associated with these surgical procedures
and, therefore, that the analgesic therapy adminis-
tered may have been inadequate. In humans, opioid-
induced hyperalgesia is associated with increased
sensitivity to nociceptive stimuli in the postoperative
period (Bekhit 2010; Fletcher & Martinez 2014).
Although data specific to dogs are lacking, the
occurrence of this phenomenon could potentially
impact on the stress response to surgery.
All dogs in SPI were administered methadone in
the first hour after endotracheal extubation,
52
Stress and regional anaesthesia in dogs M Romano et al.
whereas none of the dogs in PNB were. One
explanation for this difference may refer to the
variability of duration of bupivacaine according to
the route of administration. The average duration of
anaesthesia in SPI was 160 minutes and, conse-
quently, the effect of spinal anaesthesia was likely to
have diminished, which is in agreement with the
duration of spinal anaesthesia in dogs described by
Sarotti et al. (2012). The median time to first rescue
analgesia in dogs after PNB of the pelvic limb is
reported to be about 14 hours (range: 6–24 hours)
(Campoy et al. 2012).
Glucose concentrations rise in proportion to the
intensity of surgical stimulation as a result of insulin
resistance and the gluconeogenic effects of cortisol
(Desborough 2000). Hyperglycaemia has been asso-
ciated with a worse outcome in critical patients
because it results in higher sensitivity to infection
and impaired wound healing (van den Berghe et al.
2001).
Increased cortisol concentrations have been
reported to have remarkable effects on glucose
concentrations and on immune function and thus
its secretion can worsen the postoperative outcome
and increase the risk for wound infection (van den
Berghe et al. 2001). For these reasons, limiting wide
swings in cortisol and glucose concentrations during
and after anaesthesia may benefit the patient.
Further studies should be conducted to assess the
effect of the stress response on postoperative out-
comes in dogs undergoing surgery.
The present study has several limitations. Firstly,
it would have been ideal to evaluate the animals for
a longer time after surgery (e.g. 24 hours) in order
to assess further changes in the selected variables.
Secondly, measuring cortisol and glucose concen-
trations during surgery would have identified
whether these increases occurred before extubation.
Lastly, the study was performed in a clinical setting
and the study population was a non-homogeneous
group of dogs of various breeds, ages and both
genders, although it does reflect a real population of
dogs presenting for surgery.
In conclusion, PNB and spinal anaesthesia con-
trolled the glycaemic and adrenocortical responses
to surgery in the study population, promoted better
quality of recovery, and decreased postoperative
methadone consumption compared with IV fentanyl
administration in dogs. Regional and spinal anaes-
thesia should be considered as valid alternatives to
fentanyl administration in dogs undergoing pelvic
limb surgeries.
53 © 2015 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 44–54
References
Abram SE, Olson EE (1994) Systemic opioids do not
suppress spinal sensitization after subcutaneous
formalin in rats. Anesthesiology 80, 1114–1119.
Auroy Y, Narchi P, Messiah A et al. (1997) Serious
complications related to regional anesthesia: results of
a prospective survey in France. Anesthesiology 87, 479–
486.
Becker WM, Mama KR, Rao S et al. (2012) Prevalence of
dysphoria after fentanyl in dogs undergoing stifle
surgery. Vet Surg 42, 302–307.
Bekhit MH (2010) Opioid-induced hyperalgesia and
tolerance. Am J Ther 17, 498–510.
van den Berghe G, Wouters P, Weekers F et al. (2001)
Intensive insulin therapy in critically ill patients. N Engl J
Med 345, 1359–1367.
Campoy L, Martin-Flores M, Ludders JW et al. (2012)
Comparison of bupivacaine femoral and sciatic nerve
block versus bupivacaine and morphine epidural for stifle
surgery in dogs. Vet Anaesth Analg 39, 91–98.
Desborough JP (2000) The stress response to trauma and
surgery. Br J Anaesth 85, 109–117.
Devitt CM, Cox RE, Hailey JJ (2005) Duration, complications,
stress, and pain of open ovariohysterectomy versus a simple
method of laparoscopic-assisted ovariohysterectomy in
dogs. J Am Vet Med Assoc 227, 921–927.
Ecoffey C, Lacroix F, Giaufre E et al. (2010) Epidemiology
and morbidity of regional anesthesia in children: a
follow-up one-year prospective survey of the French-
Language Society of Paediatric Anaesthesiologists
(ADARPEF). Paediatr Anaesth 20, 1061–1069.
Fletcher D, Martinez V (2014) Opioid-induced hyperalgesia
in patients after surgery: a systematic review and a meta-
analysis. Br J Anaesth 112, 991–1004.
Keating SCJ, Kerr CL, Valverde A et al. (2013)
Cardiopulmonary effects of intravenous fentanyl
infusion in dogs during isoflurane anesthesia and with
concurrent acepromazine or dexmedetomidine
administration during anesthetic recovery. Am J Vet
Res 74, 672–682.
Kehlet H (1989) Surgical stress: the role of pain and
analgesia. Br J Anaesth 63, 189–195.
Kettner S, Willschke H, Marhofer P (2011) Does regional
anaesthesia really improve outcome? Br J Anaesth 107
(Suppl 1), 90–95.
Kona-Boun JJ, Cuvelliez S, Troncy E (2006) Evaluation of
epidural administration of morphine or morphine
and bupivacaine for postoperative analgesia after
premedication with an opioid analgesic and
orthopedic surgery in dogs. J Am Vet Med Assoc 229,
1103–1112.
Lamont LA (2008) Multimodal pain management in
veterinary medicine: the physiologic basis of
pharmacologic therapies. Vet Clin North Am Small
Anim Pract 38, 1173–1186.
 Stress and regional anaesthesia in dogs M Romano et al.
Lamont LA, Mathews KA (2007) Opioids, nonsteroidal
anti-inflammatories, and analgesic adjuvants. In: Lumb
& Jones’ Veterinary Anesthesia and Analgesia (4th edn).
Tranquilli WJ, Thurmon JC, Grimm KA (eds). Blackwell
Publishing, USA. pp. 241–271.
Liu S, Carpenter RL, Neal JM (1995) Epidural anesthesia
and analgesia: their role in postoperative outcome.
Anesthesiology 82, 1474–1506.
Meissner A, Rolf N, Van Aken H (1997) Thoracic epidural
anesthesia and the patient with heart disease: benefits,
risks, and controversies. Anesth Analg 85, 517–528.
Naitoh T, Garcia-Ruiz A, Vladisavljevic A et al. (2002)
Gastrointestinal transit and stress response after
laparoscopic vs conventional distal pancreatectomy in
the canine model. Surg Endosc 16, 1627–1630.
Nelson R (2006) Patologie della ghiandola surrenale. In:
Medicina Interna del Cane e del Gatto (3rd edn). Nelson
R, Couto CG (eds). Elsevier, Italy. pp. 797–835.
Pascoe PJ (2000) Opioid analgesics. Vet Clin North Am
Small Anim Pract 30, 757–772.
Portela DA, Otero PE, Tarragona L et al. (2010) Combined
paravertebral plexus block and parasacral sciatic block in
healthy dogs. Vet Anaesth Analg 37, 531–541.
Portela DA, Otero PE, Briganti AG et al. (2013) Femoral
nerve block: a novel psoas compartment lateral pre-iliac
approach in dogs. Vet Anaesth Analg 40, 194–204.
Reid J, Nolan A, Hughes J et al. (2007) Development of the
short-form Glasgow Composite Measure Pain Scale
(CMPS-SF) and derivation of an analgesic intervention
score. Anim Welf 16(Suppl 1), 97–104.
Sarotti D, Rabozzi R, Corletto F (2011) Efficacy and side
effects of intraoperative analgesia with intrathecal
bupivacaine and levobupivacaine: a retrospective study
in 82 dogs. Vet Anaesth Analg 38, 240–251.
© 2015 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 44–54
Sarotti D, Rabozzi R, Franci P (2012) A retrospective study
of efficacy and side effects of intrathecal administration of
hyperbaric bupivacaine and morphine solution in 39
dogs undergoing hind limb orthopaedic surgery. Vet
Anaesth Analg 40, 220–224.
Sibanda S, Hughes JML, Pawson PE et al. (2006) The
effects of preoperative extradural bupivacaine and
morphine on the stress response in dogs undergoing
femoro-tibial joint surgery. Vet Anaesth Analg 33, 246–
257.
V€ais€anen MAM, Valros AE, Hakaoja E et al. (2005)
Pre-operative stress in dogs – a preliminary
investigation of behavior and heart rate variability in
healthy hospitalized dogs. Vet Anaesth Analg 32,
158–167.
Vettorato E, De Gennaro C, Okushima S et al. (2013)
Retrospective comparison of two peripheral lumbosacral
plexus blocks in dogs undergoing pelvic limb orthopaedic
surgery. J Small Anim Pract 54, 630–637.
Wolf AR (2012) Effects of regional analgesia on stress
responses to pediatric surgery. Paediatr Anaesth 22, 19–
24.
Wolf AR, Eyres RL, Laussen PC et al. (1993) Effect of
extradural analgesia on stress responses to abdominal
surgery in infants. Br J Anaesth 70, 654–660.
Wolf AR, Doyle E, Thomas E (1998) Modifying infant stress
responses to major surgery: spinal vs extradural vs opioid
analgesia. Paediatr Anaesth 8, 305–311.
Zhang J, Zhang W, Li B (2007) The effect of epidural
anesthesia with different concentrations of ropivacaine
on sevoflurane requirements. Anesth Analg 104, 984–
986.
Received 16 August 2014; accepted 28 January 2015.
54