JOURNAL OF ENDOUROLOGY

Volume 23, Number 9, September 2009 Experimental Endourology

ª Mary Ann Liebert, Inc.
Pp. &&&–&&&
DOI: 10.1089=end.2009.0411

Hypothermic Robotic Radical Prostatectomy:

Impact on Continence

David S. Finley, M.D.,1 Kathryn Osann, Ph.D.,2 Alexandra Chang, B.A.,3 Rosanne Santos, B.S.,3
Douglas Skarecky, B.S.,3 and Thomas E. Ahlering, M.D. 3

Abstract
Introduction: Radical prostatectomy undoubtedly causes inflammatory damage to surrounding neuromus-
cular tissues (i.e., bladder, urethra, and nerves) that may contribute to urinary incontinence. We report the use
of local hypothermia during robot-assisted laparoscopic prostatectomy to attenuate this injury.
Methods: Regional pelvic cooling was achieved using cold intracorporeal irrigation and an endorectal cooling
balloon (ECB). In all, 115 men undergoing hypothermic robot-assisted laparoscopic radical prostatectomy (hRLP)
(case #667–782) were prospectively compared with a historical cohort (case #1–666). Intracorporeal rectal and
neurovascular bundle temperatures (T) and intrarectal temperatures were measured. Continence was defined as
zero urinary pads. Kaplan–Meier analysis of time to zero pads and multivariate Cox proportional hazards
regression was used.
Results: Hypothermia was achieved in 112=115 patients; 6 were excluded (3 ECB malfunction, 2 prior radiation,
and 1 completion prostatectomy). Median endorectal T ¼ 18.78C (range 9.1–29.58C). Mean intracorporeal
T ¼ 25.588C (ECB þ irrigation, range 19.4–34.08C). Three and 12-month hRLP zero pad rates were 81% to 89%
and 100% for initial and extended cooling groups versus 65% and 89% for controls. Return to continence was
significantly faster for hRLP versus controls: median time to zero pad use was 39 days for hRLP versus 62 days
for controls. Multivariate analysis adjusting for American Urological Association (AUA) symptom score, nerve-
sparing surgery, learning curve, international index of erectile function-5, age, and prostate weight demonstrated
a significantly faster return to continence (hazard ratio ¼ 1.526; 95% CI 1.11, 2.09). Trends toward improved
continence were observed with colder temperatures and older patients.
Conclusions: Local hypothermia during prostatectomy resulted in a significant improvement in early postop-
erative zero pad continence rates. Longer and deeper cooling appears to be associated with improved conti-
nence, particularly among older patients.

Introduction oxia, acidosis, free radical production, and apoptosis.1–3 Both

modes of injury (primary and secondary) probably contribute

D uring radical prostatectomy dissection of the blad-

der neck, mobilization of the neurovascular bundles
(NVBs), and transection of the urethra–external sphincter
complex causes acute traumatic injury. The primary injury
results from the direct mechanical trauma of dissection, trac-
tion, and thermal energy that leads to neurapraxia or ax-
onotmesis of pelvic autonomic nerves as well as muscle
degeneration and damage. A secondary (indirect) wave of
damage ensues because of inflammation. The inflammatory
cascade includes neutrophil and macrophage infiltration with
subsequent release of proteolytic enzymes, activation of co-
agulation factors, proinflammatory cytokine formation, hyp-
to varying degrees to the postoperative functional deficits in
urinary and sexual function.
Theoretically, secondary inflammatory damage might be
blocked (or at least mitigated) with the use of local tissue
cooling (i.e., hypothermia). Application of hypothermia pre-
emptively (before dissection starts) should prepare tissues for
imminent damage by lowering their metabolic rate and oxy-
gen demands. With sufficient temperature reduction, the cell
enters into a quiescent state of low energy utilization. When
injury ensues, energy reserves are available for repair without
going into anaerobic metabolism. As a result, less lactate
formation occurs, protein synthesis is preserved, and most

1
Institute of Urologic Oncology, David Geffen School of Medicine at UCLA, Los Angeles, California.
Departments of 2Medicine and 3Urology, University of California Irvine, Orange, California.

1
2 FINLEY ET AL.

Table 1. Multivariate Analysis of Baseline Characteristics, Nerve Sparing, and Cooling

95% Confidence interval

Variable Coefficient HR Lower Upper p-Value

Age 0.0328 0.9677 0.9553 0.9803 <0.0001

AUAss 0.0153 0.9848 0.9713 0.9984 0.0285
IIEF-5 0.0191 1.0193 1.0043 1.0346 0.0113
BMI 0.0273 0.9731 0.9497 0.9970 0.0278
Nerve sparing 0.3664 0.6932 0.4501 1.0676 0.0963
Stage (I vs. II vs. III) 0.0966 0.9079 0.7740 1.0651 0.2357
Prostate weight 0.0020 0.9980 0.9935 1.0025 0.3752
Learning curve 0.0002 0.9998 0.9993 1.0003 0.5158
Cooling vs. control 0.4230 1.5265 1.1127 2.0945 0.0088

HR ¼ hazard ratio; IIEF ¼ international index of erectile function; BMI ¼ body mass index; AUAss ¼ American Urological Association
symptom score.

importantly the inflammatory cascade is blunted. With
less proinflammatory molecules and free radical species
generated, the risk of apoptotic cell death is reduced.4,5 Tissue
damage from leukocyte infiltration is further reduced because
cooling also blocks adhesion molecule transcription and in-
hibits neutrophil adherence.5
We recently introduced the concept of hypothermic
prostatectomy and reported the outcomes of the first 50 pa-
tients demonstrating a significant improvement in time to
recovery of continence.6 Herein, we report our continence
results as our cooling technique has evolved over the past 115
patients.
Materials and Methods
An endorectal cooling balloon (ECB) was fabricated con-
sisting a 5@
2.5@ elliptical balloon fused to a 40 cm, 24F three-
way latex urethral catheter. The balloon distends at low
pressures and conforms to the rectal wall in an attempt to
maximize surface area for heat exchange. Anatomically,
the ECB extends from the membranous urethra to above the
seminal vesicles to induce loco-regional hypothermia of the
entire pelvis. The ECB is cycled continuously with 48C sa-
line or 08C polyethylene glycol–balanced electrolyte solu-
tion Golytely
(Braintree Laboratories Inc., Braintree, MA)
case #752–782 via gravity, at a height of approximately 40 cm
above the patient to a volume of approximately 200 cc; 48C
sterile water was used for intracorporeal irrigation to aug-
ment cooling and improve viewing—specifically, copious
active irrigation was used concurrent with cautery to
minimize collateral thermal injury. Intracorporeal tempera-
ture values were obtained from the anterior rectal surface
(TRECTAL), along the lateral surface of the NVB (TNVB), and
from within the lumen of the rectum using a 9F esophageal
temperature probe. Hypothermia was initiated just before
NVB dissection (case #667–739) or preemptively immediately
after intubation (case #740–782). Hypothermia was tempo-
rarily discontinued during most cases during rectal dissection
or creation of the anastomosis because of space limitations.
Cooling was reinstituted during lymph node dissection or
inguinal hernia repair (when indicated) and continued until
the patient was extubated and ready to be transported to
recovery.
Patient selection
After Institutional Review Board (IRB) approval, 115 con-
secutive men underwent hypothermic robot-assisted laparo-
scopic radical prostatectomy (hRLP) by a single surgeon
(T.A.). Standard pertinent clinical data such as age, abbrevi-
ated international index of erectile function (IIEF)-5 score,
American Urological Association symptom score (AUAss),
bother score, body mass index, prostate specific antigen
(PSA), Gleason score, clinical T-stage, prostate weight, and
nerve sparing were entered prospectively into an electronic
database.

Table 2. Multivariate Analysis of Baseline Characteristics, Nerve Sparing,

and Cooling (Initial Versus Recent)

95% Confidence interval

Variable Coefficient HR Lower Upper p-Value

Age 0.0330 0.9675 0.9551 0.9802 <0.0001

AUAss 0.0148 0.9853 0.9719 0.9989 0.0346
IIEF-5 0.0216 1.0218 1.0067 1.0373 0.0046
BMI 0.0286 0.9718 0.9483 0.9960 0.0225
Nerve sparing 0.3901 0.6770 0.4395 1.0430 0.0769
Stage (I vs. II vs. III) 0.0890 0.9148 0.7796 1.0737 0.2760
Prostate weight 0.0022 0.9978 0.9933 1.0024 0.3489
Learning curve 0.0005 0.9995 0.9989 1.0001 0.1332
Cooling (case #668–743) vs. control 0.4204 1.5226 1.1089 2.0905 0.0093
Cooling (case #744–782) vs. control 0.4729 1.6046 0.9576 2.6891 0.0726
HYPOTHERMIC ROBOTIC RADICAL PROSTATECTOMY 3

Table 3. All Controls Versus Early Cooling Versus Optimal Cooling: Continence as Dichotomous Variable
at 3 and 12 months (Excludes Those Who Are Incontinent but Have Less Than 90 and 365 Days Follow-Up)

n (%)

All controls Case #668–743 Case #744–782 Pearson chi-square

3 Months 13.4301
No 216 (35) 13 (19) 3 (11) p-Value
Yes 405 (65) 56 (81) 25 (89) 0.0012
Excluded 10 0 4
12 Months 10.4463
No 62 (11) 0 (0) 0 (0) p-Value
Yes 520 (89) 63 (100) 26 (100) 0.0054
Excluded 49 6 6

Continence
After hRLP, urethral catheters were removed after 7 days.
Postoperative continence was assessed by a patient-reported
short form of the expanded prostate cancer index-5 (EPIC-5)
questionnaire and the self-reported number of pads used
per day. In addition, patients mailed in continence land-
marks as they were achieved (e.g., landmark 1, 1 pad on
an average day; landmark 2, pad free). Continence was
strictly defined as the use of zero pads. All patients re-
quiring a security pad or liner were not considered conti-
nent for the purposes of this study.
Complications
Patients were asked about postoperative bowel symptoms
(i.e., proctalgia, tenesmus, hematochezia, or fecal inconti-
nence) via EPIC questions 22 to 30. All other complications
were documented.
Statistical analysis
Statistical analysis was completed by a clinically unin-
volved third party (K.O.). Baseline characteristics between the
hypothermia group and the normothermia control group
were analyzed with unpaired two-tailed t-tests. Multivariate
modeling (Cox proportional hazards regression) was per-
formed to examine the impact of baseline characteristics such
as IIEF score, AUAss, and age on continence between groups
as well as a learning curve bias (Tables 1 and 2). The Kaplan–
Meier method was used to calculate the time to zero pads; the
log-rank statistic was used to assess for differences. The pri-
mary outcome was time to zero urinary pads. In addition, we
evaluated the proportion of men achieving zero pad status at
defined time points (1, 3, and 12 months) as a dichotomous
variable (Tables 3 and 4).
A subset analysis was done to compare the hypothermia
group to all men who underwent normothermic RLP with
posterior urethral plate reconstruction (Rocco group) (case
#593–666). The effect of a potential learning curve was further
examined by Kaplan–Meier and multivariate analysis compar-
ing our most recent normothermic cohorts of patients consisting
of case #442 to 554, #555 to 666 (Rocco) to the hypothermia
group (case #667–782) Table 2. The analysis software used was
SAS (SAS, Cary, NC) with statistical significance set at p < 0.05.
Results
Hypothermia was successfully achieved and maintained in
112=115 patients. In three patients cooling was poor because of
the technical difficulties associated with the ECB. Six patients
were excluded; three because of the cooling balloon mal-
function and three for clinical reasons (two had prior radiation
therapy and one previous incomplete prostatectomy). There
were no complications or adverse side effects associated with
hRLP. There were no significant changes in core body tem-
perature during any of the cases. Mean endolumenal rectal
temperature was 18.78C (range 9.1–29.58C, standard deviation
4.8). Mean intracorporeal temperature was 25.588C (standard

Table 4. Controls by Subgroup Versus Cooling: Continence as Dichotomous Variable at 3 and 12 Months
(Excludes Those Who Are Incontinent but Have Less Than 90 and 365 Days Follow-Up)

n (%)

Case #1–250 Case #251–450 Case #451–592 Rocco (case #593–667) Cooling (case #668–782) Pearson chi-square

3 Months 20.4833
No 68 (29) 75 (40) 49 (39) 24 (34) 16 (16) p-Value
Yes 170 (71) 111 (60) 78 (61) 46 (66) 81 (84) 0.0004
Total 238 (100) 186 (100) 127 (100) 70 (100) 97 (100)
Excluded 1 3 4 2 4
12 Months 13.4161
No 28 (12) 20 (12) 10 (9) 4 (6) 0 (0) p-Value
Yes 201 (88) 152 (88) 107 (91) 60 (94) 89 (100) 0.0094
Total 229 (100) 172 (100) 117 (100) 64 (100) 89 (100)
Excluded 10 17 14 8 12
4 FINLEY ET AL.
FIG. 1. Kaplan–Meier analysis of time to zero pads of control group versus cooling. hRLP ¼ hypothermic robot-assisted
laparoscopic radical prostatectomy.
deviation 3.6). For case #667 to 693 cooling was initiated just
before bladder neck incision and continued until extubation.
For case #744 to 782 cooling was initiated at the time of intu-
bation and discontinued when extubated.
There was no difference in estimated blood loss, hospital
stay, or complications between the hRLP and control groups.
Of the 109 patients, there was one complication requiring
observation in the ICU for 24 hours because of an antiemetic
side effect. There were no patient-reported complaints of
rectal pain, urgency, fecal incontinence, or bloody bowel
movements. In addition, all respondents reported either ‘‘no
problem’’ or ‘‘a very small problem’’ with overall bowel
habits, in line with responses from the control group.
Urinary continence
Recovery of urinary control was significantly faster for the
hypothermia group compared with the control group as a
whole. The median time to pad-free status was 39 days for
hRLP versus 62 days for the control ( p ¼ 0.0003, Fig. 1).
Compared with the most recent normothermic control, the
FIG. 2. Control versus cooling: continence as dichotomous
variable at 3 and 12 months (excludes those who are incon-
tinent but have less than 90 and 365 days follow-up).
Rocco group, the hRLP group achieved a more rapid return to
continence, p ¼ 0.0159, chi-square 5.8163 (data not shown).
Continence rates at 1, 3, and 12 months for the hRLP and
control group were 39% (Pearson chi-square 2.41, p ¼ 0.119),
83% (Pearson chi-square 12.83, p ¼ 0.003), and 100% (Pearson
chi-square 10.45, p ¼ 0.0012) versus 31%, 65%, and 89%,
respectively (Fig. 2).
There was no trend in time to continence across groups of
patients (i.e., 1–250, 251–500) to suggest a learning effect. In
addition, continence was not improved with experience when
learning curve was measured as a continuous variable by
consecutive patient number. After adjusting for all potential
covariables, the hazard ratio for hRLP was 1.52 (95% CI: 1.11,
2.09; p ¼ 0.0088) indicating faster return to continence in the
hRLP group relative to the control group.
We further evaluated the potential effect of a learning curve
by comparing our most recent normothermic cohorts (case
#451–592) and (case #593–666) to our hypothermia group.
The normothermic cohorts did not have a significantly different
Kaplan–Meier time to zero pads ( p ¼ 0.554, Fig. 3) indicating a
lack of a learning curve effect. A Cox proportional hazards
regression adjusting for baseline covariates again found no
evidence of a learning curve effect ( p ¼ 0.762) (Table 5).
Initial hRLP group versus extended
cooling hRLP group
A subset analysis of patients who underwent ‘‘extended
cooling’’ (case #744–782) (beginning of the case until extuba-
tion) revealed a trend toward improved early continence rates
(Table 2 and Fig. 4). For case #668 to 743, at 3 and 12 months,
81% and 100% were continent (Pearson chi-square 13.43,
p ¼ 0.0012) compared with 89% and 100% for case #744 to 782
(Pearson chi-square 10.44, p ¼ 0.0054), versus 65% and 89% for
controls, respectively (Table 3 and Fig. 4). This difference held
up when cases were divided into subgroups to assess for
learning curve effects (Tables 4 and 5; Fig. 5).
Continence as a function of temperature
Among men 65 years or older, when temperature was
separated into quartiles (158C, 15.1–208C, 20.1–258C, and
HYPOTHERMIC ROBOTIC RADICAL PROSTATECTOMY 5

Table 5. Multivariate Analysis Examining the Effect of Learning Curve

95% Confidence interval

Variable Coefficient HR Lower Upper p-Value

Age 0.044 0.957 0.931 0.984 0.0021

AUAss 0.015 0.985 0.956 1.016 0.3403
IIEF-5 0.006 1.006 0.978 1.035 0.6635
BMI 0.027 0.973 0.930 1.019 0.2483
Nerve sparing 1.026 2.789 0.350 22.222 0.3328
Stage (I vs. II vs. III) 0.189 0.828 0.590 1.162 0.275
Prostate weight 0.004 1.004 0.997 1.012 0.2616
Learning curve: 0.054 0.947 0.666 1.346 0.7618
case #442–554 vs. case #555–666
Age 0.037 0.964 0.944 0.984 0.0004
AUAss 0.012 0.988 0.966 1.010 0.2863
IIEF-5 0.013 1.013 0.991 1.037 0.2508
BMI 0.012 0.988 0.952 1.025 0.5211
Nerve sparing 1.017 2.766 0.366 20.924 0.3244
Stage (I vs. II vs. III) 0.149 0.862 0.678 1.094 0.2216
Prostate weight 0.002 1.002 0.996 1.008 0.5558
Learning curve: 0.353 1.424 1.083 1.872 0.0114
case #442–666 vs. case #667–782

>258C), we found a trend toward improved continence with
lower temperatures (Cochran’s linear trend 2.788, p ¼ 0.095;
Pearson p ¼ 0.11, Fig. 5). A similar trend in continence
recovery as a function of temperature was not observed in
younger men.
Discussion
Surgical trauma results in acute injury to the involved tis-
sues such as nerves, muscle, soft tissue, and blood vessels.2
During radical prostatectomy, direct damage to the pelvic
nerves, bladder, urethra, and rhabdosphincter as well as
secondary inflammatory effects is undoubtedly a major con-
tributing factor to the postoperative urinary incontinence and
sexual dysfunction associated with the operation. Peripheral
nerve injury is a major contributing factor to these co-
morbidities. Peripheral nerve injury can be graded according
to Sir Herbert Seddon’s7 classification into three categories of
severity: (1) neurapraxia: a mild compression, blunt impact, or
stretch injury to the nerve with no structural damage. A
concussion-like state results in a transient conduction block
from which full recovery is likely to occur; recuperation may
take hours to weeks, (2) axonotmesis: a moderately severe
injury, which results in axonal disruption and Wallerian de-
generation; the nerve can regenerate or regrow from the point
of injury to the end organ at approximately 1 inch per

FIG. 3. Kaplan–Meier analysis of time to zero pads for the

most recent normothermic subgroup and the hypothermic
group to examine the potential effect of learning curve.
The hypothermia group was significantly different from
the normothermic groups ( p ¼ 0.0005, Breslow–Gehan). The
normothermic groups were not statistically different FIG. 4. Kaplan–Meier analysis of time to zero pads of
( p ¼ 0.554, Breslow–Gehan). control group versus cooling subgroups.
6 FINLEY ET AL.
FIG. 5. Continence as a function of tem-
perature achieved (Pearson chi-square
5.8819, p ¼ 0.1175; Cochran’s Linear Trend
2.788, p ¼ 0.095).
month—recovery takes 8–24 months, and (3) neurotmesis:
occurs after severe injury or laceration that transects the nerve
completely with no capacity for regrowth. A neuroma or scar
may form resulting in a permanent injury with a potential for
only partial recovery. During radical prostatectomy, injury to
the pelvic nerves and NVBs (i.e., wide excision, partial exci-
sion, transaction, or incision of the NVB) occurs according
to Seddon’s model along a spectrum of nerve injury.
Hypothermia is one strategy to minimize or down-grade this
type of nerve injury.
Hypothermia has been demonstrated to have a dramatic
protective impact in numerous experimental injury models of
the central and peripheral nervous systems. The use of mild to
moderate hypothermia (i.e., 33–288C) has been shown to be
effective in shielding neurons from damage.8,9 In a rabbit
model of spinal cord ischemia, Isaka and colleagues10 ap-
plied trans-vertebral cold packs and infused cold saline into
a cross-clamped aorta to produce spinal cord cooling. A
modest reduction in spinal cord temperature of just 4.38C
completely prevented paraplegia compared with complete
paraplegia in all the control rabbits. Sun and coworkers11
studied the effects of hypothermia on protecting guinea
pig optic nerves from stretch injury to the optic nerve.
Significant axonal ultrastructural changes occurred under
room temperature conditions. In contrast, optic nerves that
were maintained under moderate hypothermia (328C) for
2 hours showed no difference compared with sham-operated
animals.
Hypothermia is not only neuroprotective but also can
shield other tissues from severe injury. Local tissue hypo-
thermia is commonly used. It is well established that applying
an ice-pack to an injured extremity greatly reduces pain and
edema after closed soft-tissue injury.12,13 Schaser and
coworkers14 quantified this effect by assessing microvascular
permeability after controlled striated muscle injury in rats
with or without superficial cold therapy for 20 minutes. The
cold therapy group was found to have significantly decreased
interstitial fluorescent-labeled albumin levels compared with
sham animals. In addition, cold therapy was found to pre-
serve microcapillary density, reduce leukocyte adhesion and
chemotaxis, and myonecrosis. Kelly and colleagues15 showed
that regional hypothermia to 48C protected against ischemic
peripheral nerve injury after prolonged application of a
tourniquet to the hind limb in rats.
The use of thermal energy on or near the nerves also con-
tributes to neural damage. Temperatures as low as 418C to
438C can produce irreversible injury.16–18 Importantly, the use
of thermal energy can have effects beyond the site of cautery.
Mandhani and colleagues19 measured temperature changes at
the NVB during robot-assisted radical prostatectomy (RARP)
with monopolar and bipolar cautery. The average tempera-
ture rise with monopolar and bipolar cautery at the NVB
during distant (>1 cm from the NVB) anterior black neck in-
cision was 43.68C and 38.88C, respectively, after approxima-
tely 60 seconds of cautery. During NVB dissection itself using
both cautery modalities, the mean temperatures within the
NVB measured within 1 cm of the cautery rose to 53.68C and
60.98C, respectively. The damaging effects of thermal energy,
however, can be reduced dramatically with the use of hypo-
thermia and adjunctive irrigation. Donzelli et al20 studied the
thermoprotective effects of simultaneous irrigation during
bipolar cautery to protect rat sciatic nerves. With room tem-
perature normal saline irrigation, they found simultaneous
irrigation significantly reduced the temperature response to
cautery. Blinded review of the cauterized nerves found pres-
ervation of myelin and axons. Rats that received irrigation
showed enhanced functional recovery and less paresis ac-
cording to the Sciatic Functional Index score. In the present
study, we used copious cold irrigation simultaneously with
monopolar and bipolar cautery to lower target tissue tem-
peratures and minimize collateral thermal injury.
Although most of the incontinence after radical prostatec-
tomy (RP) is thought to be because of intrinsic sphincter de-
ficiency (ISD) and detrusor hyperreflexia,21,22 to date, efforts
have focused on techniques to preserve the integrity of the
external striated sphincter complex and minimize the impact
of neural trauma. Virtually no urologic literature exists on the
impact of inflammation (or its prevention) from surgical ex-
cision of the prostate on these processes. An interesting study
by Callsen-Cencic and Mense23 induced detrusor hyperre-
flexia in rats by infusing caustic mustard oil in their bladders.
Application of hypothermia to the spinal cord was found to
significantly reduce detrusor instability. They observed
marked suppression of detrusor contraction frequency at
HYPOTHERMIC ROBOTIC RADICAL PROSTATECTOMY
218C to 258C and complete elimination at <208C. A contrib-
uting factor to new onset detrusor instability after RP may be
related to inflammation such as this from mobilizing the
bladder, cutting the bladder neck, and manipulating pelvic
autonomic and=or cavernous nerves. We hypothesized that
the use of local hypothermia during radical prostatectomy
would attenuate the damage to the sphincteric complex, au-
tonomic pelvic nerves, and detrusor muscle.
We recently published the initial report describing the im-
pact of hypothermia on continence during RP.6 Using multi-
variate Cox regression analysis, we found hypothermia had
an independent effect of recovery of continence. The present
study represents an update of these data to include our most
recent cohort of patients in which our hypothermic technique
improved—deeper hypothermia was achieved for longer
periods of time. Our most recent group underwent cooling for
3 to 4 hours from intubation to extubation compared with
our initial cohort in which hypothermia was started at blad-
der neck incision. We were also able to achieve consistently
colder temperatures in the more recent group of men by cy-
cling the ECB with polyethylene glycol that has a lower
freezing point than that of saline. The overall continence rate
at 3 months was significantly higher for the extended cooling
group compared with the initial cooling group when re-
stricted to only those patients who reached continence (89%
vs. 81%; p ¼ 0.0012). In addition, among men 65 years or older,
we found a trend toward improved continence with lower
temperatures (Cochran’s linear trend 2.788, p ¼ 0.095; Pearson
p ¼ 0.11). This finding suggests that better urinary outcomes
are achieved with ‘‘better cooling,’’ particularly in older men
who may be more susceptible to inflammatory injury and less
likely to recover.
We are currently in the process of testing a second-
generation ECB that is capable of achieving colder tempera-
tures more efficiently. Although these initial results are
encouraging, a blinded multicenter randomized controlled
trial of hypothermia versus no hypothermia is needed to
validate our findings. Further studies to characterize the op-
timal degree and depth of cooling are also underway. Finally,
the underlying principles of hypothermia predict that a sim-
ilar therapeutic effect should be seen on recovery of potency.
Conclusion
With improved cooling technique, our continence re-
sults have continued to improve providing compelling evi-
dence to support the use of hypothermia during radical
prostatectomy.
Acknowledgment
We thank Professor Ralph V. Clayman for his brilliant
input and tireless support.
Source of Funding
None.
Disclosure Statement
Thomas Ahlering and David Finley are co-inventors of a
University of California owned patent on a device related to
7
this or the subject of this publication. The University of Cali-
fornia currently has a licensing agreement with Innercool
Therapies.
References
1. Ni Choileain N, Redmond HP. The immunological conse-
quences of injury. Surgeon 2006;4:23–31.
2. Osborn L. Leukocyte adhesion to endothelium in inflam-
mation. Cell 1990;62:3–6.
3. Weiss SJ. Tissue destruction by neutrophils. N Engl J Med
1989;320:365–376.
4. Westermann S, Vollmar B, Thorlacius H, Menger MD. Sur-
face cooling inhibits tumor necrosis factor-alpha-induced
microvascular perfusion failure, leukocyte adhesion, and
apoptosis in the striated muscle. Surgery 1999;126:881–889.
5. Yeh CH, Wang YC, Wu YC, Lin YM, Lin PJ. Ischemic pre-
conditioning or heat shock pretreatment ameliorates neur-
alnal apoptosis following hypothermic circulatory arrest.
J Thorac Cardiovasc Surg 2004;128:203–210.
6. Finley DS, Osann K, Skarecky D, Ahlering T. Hypo-
thermic nerve sparing radical prostatectomy: Rationale,
feasibility, and impact on early continence. Urology 2009;73:
691–696.
7. Seddon HJ. A review of work on peripheral nerve injuries in
Great Britain during World War II. J Nerv Ment Dis 1948;
108:160–168.
8. Kawamura N, Schmelzer JD, Wang Y, Schmeichel AM, Low
PA. The therapeutic window of hypothermic neural pro-
tection in experimental ischemic neuropathy: Protection in
ischemic phase and potential deterioration in later reperfu-
sion phase. Exp Neurol 2005;195:305–312.
9. Yang XF, Kennedy BR, Lomber SG, Schmidt RE, Roth-
man SM. Cooling produces minimal neuralpathology in
neocortex and hippocampus. Neuralbiol Dis 2006;23:637–
643.
10. Isaka M, Kumagai H, Sugawara Y, Okada K, Orihashi K,
Ohtaki M, Sueda T. Cold spinoplegia and transvertebral
cooling pad reduce spinal cord injury during thor-
acoabdominal aortic surgery. J Vasc Surg 2006;43:1257–
1262.
11. Sun X, Tang W, Zheng L. Ultrastructural observation of ef-
fect of moderate hypothermia on axonal damage in an ani-
mal model of diffuse axonal injury. Chin J Traumatol 2002;5:
355–360.
12. Deal DN, Tipton J, Rosencrance E, Curl WW, Smith TL.
Ice reduces edema. J Bone Joint Surg Am 2002;84:1573–
1578.
13. Laing DR, Dalley DR, Kirk JA. Ice therapy in soft tissue
injury. N Z Med J 1973;78:155–158.
14. Schaser KD, Disch AC, Stover JF, Laufer A, Bail HJ, Mittl-
meier T. Prolonged superficial local cryotherapy attenuates
microcirculatory impairment, regional inflammation, and
muscle necrosis after closed soft tissue injury in rats. Am J
Sports Med 2007;35:93–102.
15. Kelly C, Creagh T, Grace PA, Bouchier-Hayes D. Regional
hypothermia protects against tourniquet neuropathy. Eur J
Vasc Surg 1992;6:288–292.
16. Wondergem J, Haveman J, Rusman V, Sminia P, VanDijk JD.
Effects of local hyperthermia on the motor function of the rat
sciatic nerve. Int J Radiat Biol 1988;53:429–438.
17. Hoogeveen JF, Troost D, Wondergem J, van der Kracht AH,
Haveman J. Hyperthermic injury versus crush injury in
the rat sciatic nerve: A comparative functional, histopath-
8 FINLEY ET AL.

ological and morphological study. J Neurol Sci 1992;108: Address correspondence to:
55–64. David S. Finley, M.D.
18. Xu D, Pollock M. Experimental nerve thermal injury. Brain Institute of Urologic Oncology
1994;117:375–384. David Geffen School of Medicine at UCLA
19. Mandhani A, Dorsey PJ Jr., Ramanathan R, Salamanca JI, Rao 924 Westwood Blvd., Suite 1050
S, Leung R, Berryhill R Jr., Tewari AK. Real time monitoring Los Angeles, CA 90024
of temperature changes in neurovascular bundles during
robotic radical prostatectomy: Thermal map for nerve-
E-mail: dfinley@mednet.ucla.edu
sparing radical prostatectomy. J Endourol 2008;22:2313–2317.
20. Donzelli J, Leonetti JP, Wurster RD, Lee JM, Young MR.
Neuroprotection due to irrigation during bipolar cautery. Abbreviations Used
Arch Otolaryngol Head Neck Surg 2000;126:149–153. AUA ¼ American Urological Association
21. Hammerer P, Huland H. Urodynamic evaluation of changes BMI ¼ body mass index
in urinary control after radical retropubic prostatectomy. ECB ¼ endorectal cooling balloon
J Urol 1997;157:233–236. HR ¼ hazard ratio
22. Ficazzola MA, Nitti VW. The etiology of post prostatectomy hRLP ¼ hypothermic robot-assisted laparoscopic
incontinence and correlation of symptoms with urodynamic radical prostectomy
findings. J Urol 1998;160:1317–1320. IIEF ¼ international index of erectile function
23. Callsen-Cencic P, Mense S. Control of the unstable urinary NVB ¼ neurovascular bundle
bladder by graded thermoelectric cooling of the spinal cord. RP ¼ radical prostatectomy
BJU Int 1999;84:1084–1092.