You are on page 1of 12

COCHRANE REVIEW

Surgical pelvic neuroablation for chronic pelvic pain:


a systematic review
Neil Johnson, Michelle Wilson and Cindy Farquhar
University Department of Obstetrics and Gynaecology, National Women's Hospital, Epsom, Auckland, New Zealand

Keywords ABSTRACT

endometriosis, LUNA, neuro Objective To determine, from the best available evidence, the effectiveness
ablation, pelvic pain, presacral of surgical pelvic neuroablative techniques as treatment for primary and
neurectomy. secondary dysmenorrhoea.
Design Systematic review of randomized controlled trials (RCTs).
Subjects Women with primary or secondary dysmenorrhoea.
Interventions Surgical neuroablative techniques including uterine nerve
ablation and presacral neurectomy.
Main outcome measures Pain relief (both short- and long-term) and
adverse effects.
Results Laparoscopic uterine nerve ablation (LUNA) has been shown by
one small study to be effective for primary dysmenorrhoea up to 12 months.
A larger study showed no significant difference between LUNA and laparo-
scopic presacral neurectomy (LPSN) in pain relief for primary dysmenor-
rhoea up to 6 months, but LPSN was significantly more effective beyond
6 months. For dysmenorrhoea secondary to endometriosis, two trials found
no significant difference in pain relief afforded by LUNA plus conservative
laparoscopic surgery for endometriosis vs. conservative laparoscopic surgery
alone. Presacral neurectomy (PSN) plus conservative surgery for endome-
triosis at laparotomy vs. conservative surgery for endometriosis at laparotomy
showed no significant difference in overall pain relief, although there was a
significant difference in relief of midline abdominal pain. Adverse events
Correspondence overall were significantly more common following PSN; these included
N. Johnson, University Department of constipation, urinary urgency, painless labour and one case necessitating a
Obstetrics and Gynaecology, National laparotomy for presacral haematoma 2 days after PSN.
Women's Hospital, Claude Road, Epsom,
Auckland 1006, New Zealand.
Conclusions There is insufficient evidence to recommend the use of
surgical pelvic neuroablation in the management of dysmenorrhoea, regard-
Accepted for publication 30 May 2000 less of cause. Further scientifically rigorous RCTs should be undertaken.

When menstrual pelvic pain is associated with an


INTRODUCTION
identifiable pathological condition, such as endome-
Dysmenorrhoea is the occurrence of painful menstrual triosis, adenomyosis or pelvic adhesions reflecting
cramps of uterine origin. It affects up to 50% of previous inflammation, it is considered to be secondary
women,1 with an impact not only on personal health dysmenorrhoea. In contrast, menstrual pain without
but also an economic impact through loss of working organic pathology is considered to be primary dysme-
hours2 along with reduced productivity, diminished norrhoea.5 The initial onset of primary dysmenorrhoea
work quality and more work-related accidents in those is usually at or shortly (612 months) after menarche,
who continue to work.3,4 when ovulatory cycles are established. The pain

q2000 Blackwell Science Ltd Gynaecological Endoscopy 2000 9, 351361 351


352 N. JOH NSO N e t al .

duration is typically 4872 h and is associated with Uterine nerve ablation (UNA) involves the transec-
menstrual flow. In contrast secondary dysmenorrhoea tion of the uterosacral ligaments at their insertion into
is more likely to occur years after the onset of the cervix. Presacral neurectomy (PSN) involves resec-
menarche and occur premenstrually as well as during tion of the presacral nerves lying within the boundaries
menstruation. of the interiliac triangle (essentially the superior
Overproduction of uterine prostaglandins is probably hypogastric nerve plexus). These procedures both
involved in the aetiology of the typical pain of primary interrupt pelvic afferent sensory nerve fibres and
dysmenorrhoea.6 Prostaglandins are also implicated in would be expected to diminish uterine pain,14
secondary dysmenorrhoea; however, anatomical although PSN involves the interruption of a greater
mechanisms can also be identified, depending on the number of nerve pathways than UNA. Doyle (1955)15
type of accompanying pelvic pathology.4 Medical ther- described a technique of vaginal transection of the
apy includes oral contraceptive pills (OCPs) and non- uterosacral nerves apparently effective for dysmenor-
steroidal anti-inflammatory drugs (NSAIDs) which both rhoea. Uncontrolled studies have supported the use of
act by suppressing prostaglandin levels.7 Although the laparoscopic uterine nerve ablation (LUNA) for both
use of both OCPs and NSAIDs has been very successful, primary and secondary dysmenorrhoea with either
there is still a 2025% failure rate.8,9 complete relief or substantial reduction in menstrual
The ideal neuroablative surgical procedure for pelvic pain in the majority of subjects.1618 Earlier large
pain would transect all afferent sensory nerves from all retrospective studies provide data which have under-
the pelvic organs and leave all other nerves unaffected. lined the safety of LUNA.19 The use of PSN is also
Whilst pelvic neuroanatomy is complicated and still not supported by uncontrolled studies showing similar
completely understood, what is known makes it clear results to that of LUNA for both primary and
that no such `ideal neuroablative surgical procedure' secondary dysmenorrhoea,20,21 although PSN probably
exists (see Fig. 1). The body of the uterus is widely entails greater operative risk than UNA.22
considered to be innervated only by sympathetic However there are limitations to the usefulness of
nerves.10 The cervix has predominantly parasympa- UNA and PSN. A long-term study showed that success
thetic (but also sympathetic) innervation. The afferent rates declined rapidly over time from 72% in the first
sensory nerves from both the uterus and cervix traverse year to 39% in the fourth for LUNA.23 Others have
the cervical division of the LeeFrankenhauser plexus suggested a possible risk of anatomical distortion such
which lies within and around the site of attachment of as uterine prolapse24,25 and bladder dysfunction26
the uterosacral ligaments to the posterior aspect of the following LUNA. There is also the concern regarding
cervix.11,12 From the uterosacral ligament, the para- the effects of interruption of the pelvic nerves to the
sympathetic afferent nerves reach the dorsal root uterine muscles in subsequent pregnancies, such as
ganglia of the first to fourth sacral spinal nerves (S1 painless labour. Since both operations interrupt only
4) via the pelvic splanchnic nerves (nervi erigentes) some of the afferent sensory nerve fibres from the
and inferior hypogastric nerve plexus (also known as pelvis, neuroablative techniques may be less effective
the pelvic plexus) then the superior hypogastric nerve for dysmenorrhoea associated with additional pelvic
plexus (also known as the `presacral nerve' or pathology. For this reason these techniques are often
hypogastric plexus).13 The sympathetic afferent nerves combined with additional treatments such as vaporiza-
emerging from the LeeFrankenhauser plexus accom- tion of endometrial implants.
pany the uterine, iliac and inferior mesenteric arteries The objective of this review was to determine the
to the sacral sympathetic trunk via the sacral splanch- effectiveness of surgical pelvic neuroablative techni-
nic nerves, some of which bypass the superior ques (both laparoscopic and at laparotomy) for the
hypogastric nerve plexus. Afferent nerves accompany treatment of primary and secondary dysmenorrhoea.
both parasympathetic and sympathetic nerves from the
ovary: pain fibres bypass the uterosacral ligament and
M AT E R I A L S A N D M E T HO DS
course through corresponding plexuses to their cells of
origin in the dorsal root ganglia of the tenth and
Search strategy
eleventh thoracic spinal nerves (T1011); some of the
upper ovarian plexus afferent nerves course directly via This review has employed the search strategy devel-
renal and aortic plexuses and bypass the superior oped by the Menstrual Disorders and Subfertility
hypogastric nerve plexus. Group. Relevant trials were identified from the Trial

Gynaecological Endoscopy 2000 9, 351361 q2000 Blackwell Science Ltd


LUNA OR PRES ACRAL NEURECTOMY FOR CHRONIC PELVIC PAIN? 353

Figure 1 Afferent nerve supply of the


female pelvic organs. C, afferent nerve
supply of cervix (illustrated on right side
of diagram); O, afferent nerve supply of
ovary (illustrated on left side of diagram);
U, afferent nerve supply of uterus
(illustrated on right side of diagram);
superscript P, parasympathetic nerve;
superscript S, sympathetic nerve.

Register of the Review Group. The electronic databases 1. randomized controlled trials
Medline (from 1966 to 1998) and Embase (198098) 2. interventions involving surgical interruption of
were also searched using the following keywords: pelvic nerve pathways (both open and laparoscopic)
dysmenorrh$.tw where it is possible to draw conclusions regarding the
dysmenorrhea/ effect of the intervention vs. no treatment or vs.
painful menstruat$.tw another intervention
pelvic pain/ 3. primary outcome of pain relief specified.
surgery/ The other primary outcome was adverse effects. The
laparoscop$.tw selection of trials for inclusion in the review was
surgical procedures, laparoscopic performed independently by two reviewers (M.W. and
denervation C.F.) after employing the search strategy described
uterine nerve ablation.tw previously. Uncertainties or differences of opinion
presacral neurectomy.tw were resolved by consensus after discussion with the
The search was performed on titles, abstracts, and third reviewer (N.J.). The search strategy yielded seven
keywords of the listed articles. RCTs requiring closer scrutiny. From these, one was
In addition citation lists of relevant publications, excluded27 which compared laser laparoscopy (invol-
review articles, and included studies were searched, as ving LUNA and surgical treatment of endometrial
were relevant conference abstracts. implants) with no treatment (diagnostic laparoscopy
only) in women with minimal-to-moderate endome-
triosis. Inclusion criterion 2 above was not fulfilled
Identification of included trials
since there was no control group receiving only laser
Inclusion criteria were as follows: vaporization so that the effect of the intervention

q2000 Blackwell Science Ltd Gynaecological Endoscopy 2000 9, 351361


354 N. JOH NSO N e t al .

LUNA could not be ascertained. The six included trials of the aortic bifurcation, bilateral common iliac
were assessed independently by two reviewers for arteries, the right ureter and inferior mesenteric artery,
predefined quality criteria and methodological details and the superior haemorrhoidal vessels on the left.
(displayed in Tables 1, 2 and 3). The peritoneum overlying the sacral promontory and
1 cm caudal to the aortic bifurcation was elevated and
incised. The underlying adipose tissue was blunt
Description and methodological quality of studies
dissected. By successive cauterization and cutting, the
Six trials met the criteria for inclusion in the review. nerve plexus was identified and freed from the
Four trials had been fully published in peer-reviewed underlying tissue, left common iliac vein and middle
journals.5,28,31,32 Authors from five trials have been sacral vein. The neural tissue was excised with the aid
contacted to request additional information and/or of unipolar and/or bipolar electrocautery, removed
data. One trial has published only preliminary results and sent for histological examination. Candiani et al.
and the trial is currently being prepared for full (1992)31 and Tjaden et al. (1990)32 referred to the
publication with the possibility of further results being literature to describe their techniques of presacral
available at a later date;30 further information was neurectomy at laparotomy.33,34
supplied in correspondence regarding methodology. Double blinding was used in two studies;5,29 one was
One trial is currently in progress and unpublished data single-blind;32 Vercellini et al.30 used no blinding in
with an interim analysis from this trial were supplied in follow up, and for the other trial blinding was
correspondence.29 Details of methodology and alloca- unclear.31 It is surprising that few of these trials
tion were supplied for one further trial.32 Further employed double blinding since this is important
information has to date not been forthcoming for two when assessing a subjective outcome such as pain
trials.28,31 relief. However the area which can most seriously
A detailed description of the trials assessing women jeopardize the reliability of a trial is the randomization
with primary dysmenorrhoea is given in Table 1, of process; for only one trial29 was it clarified that true
those assessing UNA in women with endometriosis in randomization at the time of surgery, allowing ade-
Table 2 and those assessing PSN in women with quate allocation concealment, had taken place. One
endometriosis in Table 3. There were no trials asses- trial employed the use of case numbers in a process of
sing pathology causing secondary dysmenorrhoea quasi-randomization,5 but this trial was otherwise
other than endometriosis. scientifically rigorous and therefore included.
The surgery involved was performed using standard A power calculation was performed and adhered to
techniques. LUNA was performed by operative laparo- in one study.31 Tjaden et al. (1990)32 also included a
scopy in all four trials describing its use. Lichten & power calculation but the trial was stopped before the
Bombard (1987)5 described bilateral uterosacral liga- number of patients needed was reached. The monitor-
ment electrocautery at the ligamentous insertion into ing committee surprisingly stopped the trial after
the cervix until the tissues were blanched (generator evaluation of data from the first 26 patients (only
power 5.8 watts). The ligaments were then incised eight of whom had been randomized), as it was
using laparoscopic scissors and electrocautery reap- considered unethical to deprive patients with midline
plied to the base of the incision. Chen et al. (1996)28 dysmenorrhoea of the relief afforded by presacral
described a similar `coagulation and dissection' tech- neurectomy.
nique at the insertion of the uterosacral ligaments into
the cervix, using monopolar or bipolar electrocautery.
Data extraction
Dover et al. (1999)29 did not describe the technique in
detail, but previously this group have described their Linear analogue scales have been validated as the most
technique of laser laparoscopy involving uterine nerve suitable method of evaluating pain.35 Some authors
transection with the carbon dioxide laser.27 Vercellini et using this evaluation method have chosen to dichot-
al. (1997)30 described the use of electrosurgery and omize the data obtained; there is evidence to suggest
mechanical instruments; excision of a segment of at that dichotomous outcomes can be derived from
least 1 cm in depth and length from the uterosacral continuous data when interpreting pain relief data.36
ligaments at their uterine junction was performed. The outcome `pain relief' may be reported as
Chen et al. (1996)28 described the technique of dichotomous (pain relief or not) or continuous data.
LPSN in detail. Laparoscopic identification was made Meta-analysis with RevMan software requires data to be

Gynaecological Endoscopy 2000 9, 351361 q2000 Blackwell Science Ltd


LUNA OR PRES ACRAL NEURECTOMY FOR CHRONIC PELVIC PAIN? 355

Table 1 Trials assessing women with primary dysmenorrhoea

Lichten & Bombard (1987)5 Chen et al. 199628

Trial characteristics
Method of randomization By last digit of medical case number Not stated
on day of surgery
Allocation concealment Inadequate (owing to use of case numbers Unclear
in randomization process)
Timing of randomization On the day of surgery Unclear
Utilization of blinding Double-blind Unclear
Number of participants providing 21 68
data for analysis
Withdrawals/losses to follow up None None
Intention-to-treat analysis Apparently done by author since no women Apparently done by author since no women
withdrawn after randomization withdrawn after randomization
Power calculation Not done Not stated
Location of the trial USA Taiwan
Duration of follow-up 12 months 3 and 12 months
Baseline characteristics of
the studied groups
Age of women in the trial Range 1834 years Range 1840 years
Inclusion criteria Women with severe dysmenorrhoea, no Women with primary dysmenorrhoea
improvement with at least 2 NSAIDs and and/or chronic pelvic pain
an oral contraceptive concurrently
Exclusion criteria History of psychotherapy, major abdominal Pelvic pathology at laparoscopy
procedures, drug abuse, or demonstrable
pelvic pathology at laparoscopy
Interventions
Nature of intervention LUNA vs. diagnostic laparoscopy only LUNA vs. LPSN
Outcomes
Method of expressing pain ratings 5-point pain scale converted by authors 5-point pain scale, converted by authors
into pain relief or not into percentage pain relief, then into
success 50100%, or failure 050%
Timing of assessments Preoperatively; postoperatively at Preoperatively; postoperatively at 3 months
3 months and 12 months and 12 months
Other outcomes Adverse effects Adverse effects

LUNA, laparoscopic uterine nerve ablation; LPSN, laparoscopic presacral neurectomy

either dichotomous or to be presented as absolute formula (range  0.95)/4 was inappropriate). Vercel-
values of means with their standard deviations (allow- lini et al. (1997)30 also reported continuous data;
ing the calculation of a weighted mean difference). however, the reported figures were medians and
Three authors5,28,31 dichotomized their pain relief data interquartile ranges, making conversion to a mean
from the continuous outcome data reported by their and standard deviation also inappropriate.
trial participants; these reported figures were used in All data were extracted independently by two
this analysis. For the other three trials29,30,32 data were reviewers (M.W. and N.J.), entered into the Review
presented correctly but in a form considered unsui- Manager (RevMan 4.0.3) computer software (Update
table for inclusion in this meta-analysis; these data are Software, Oxford, UK) where appropriate, and double-
described in the results section but not included in the checked for accuracy. Statistical analysis was performed
meta-analysis. Tjaden et al. (1990)32 reported dichot- in accordance with the guidelines for statistical analysis
omous pain relief data as three subgroups: midline, developed by the Menstrual Disorders and Subfertility
lateral and back pain. Dover et al. (1999)29 correctly Group. For dichotomous data, odds ratios with 95%
presented their outcome data for pain relief as confidence intervals were calculated; there was no
medians with a range; these data were found to have pooling of results from different trials since each of the
a positive skew (hence use of the median as a `mean' three trials suitable for inclusion in a meta-analysis
and the standard deviation estimated using the examined different interventions (see Table 4).

q2000 Blackwell Science Ltd Gynaecological Endoscopy 2000 9, 351361


356 N. JOH NSO N e t al .

Table 2 Trials assessing uterine nerve ablation in women with endometriosis

Dover et al. 199929 Vercellini et al. 199730

Trial characteristics
Method of randomization Truly randomized (computer-generated) Truly randomized (computer-generated
randomization sequence and consecutively
numbered, opaque, sealed envelopes used)
Allocation concealment Adequate Adequate
Timing of randomization Intraoperatively Unclear
Utilization of blinding Double-blind Not used
Number of participants providing 46 (51 randomized) 81
data for analysis
Withdrawals/losses to follow up 1 withdrawn after randomization 4 withdrew after randomization (1 LUNA, 3
(pregnancy), 4 lost to follow-up conservative surgery only)
(2 LUNA, 3 controls)
Intention-to-treat analysis Not possible Done by author
Power calculation Not stated Not stated
Location of the trial Guildford, UK Milan, Italy
Duration of follow up 6 months Mean 9 months
Baseline characteristics of
the studied groups
Age of women in the trial Mean 28.1 years (range 2041 years) Not stated
Inclusion criteria Women with endometriosis stage 13 Women undergoing first-line operative
laparoscopy for symptomatic endometriosis
Exclusion criteria Pregnancy (current or planned), pelvic Previous surgical treatment for endometriosis,
pathology other than endometriosis, recent hypo-oestrogenizing medical therapy,
medicaltherapy, stage 4 endometriosis, presence of other pelvic abnormalities which
previous surgical treatment could cause pain (pelvic inflammatory disease,
genital malformation, myomas)
Interventions
Nature of intervention LUNA plus laser vaporization of LUNA plus conservative surgery vs.
endometriosis vs. laparoscopic laser laparoscopic conservative surgery
vaporization of endometriosis
Outcomes
Method of expressing pain ratings 10-cm visual analogue scale and 10-point 100-mm visual analogue scale
pain scale for dysmenorrhoea,
dyspareunia and pelvic pain
Timing of assessments Preoperatively; postoperatively at Preoperatively; postoperatively at 9 months
3 months and 6 months
Other outcomes Participant satisfaction with treatment results Adverse effects

group had significantly better pain relief scores


R E S U LT S
(OR 0.1). There was a highly significant difference
in the number of adverse effects reported by the
Primary dysmenorrhoea
participants: 94% of the PSN group reported constipa-
The intervention LUNA (see Table 4) effected a tion with no complications being reported in the
significant difference in pain relief for up to 6 months LUNA group (OR 0.02).
of follow-up (odds ratio, OR 15.5); both LUNA and
control groups were comparable prior to the interven-
Secondary dysmenorrhoea
tion. Pain relief assessed at 12 months also showed a
significant difference between the groups For UNA with surgical treatment of endometriosis vs.
(OR 10.9). surgical treatment of endometriosis alone, both trials
For LUNA vs. LPSN (see Table 4) there was no found no significant difference in pain relief.
significant difference in pain relief between the two Dover et al. (1999)29 reported comparable baseline
treatments up to 6 months of follow-up (OR 0.7). pain scores, and at 6 months postoperatively found no
However at more than 6 months' follow up, the LPSN significant difference between the experimental and

Gynaecological Endoscopy 2000 9, 351361 q2000 Blackwell Science Ltd


LUNA OR PRES ACRAL NEURECTOMY FOR CHRONIC PELVIC PAIN? 357

Table 3 Trials assessing presacral neurectomy (PSN) in women with endometriosis

Candiani et al. (1992)31 Tjaden et al. (1990)32

Trial characteristics
Method of randomization Randomization list used but unclear how 8 participants randomly allocated (centralized
randomization process with sealed, opaque,
sequentially numbered, identical envelopes) and
18 participants not randomly allocated
Allocation concealment Unclear Adequate
Timing of randomization Prior to surgery
Utilization of blinding Unclear Single blinding of randomly allocated
participants
Number of participants 78 recruited (71 analysed) 26 by authors, 8 for this review
providing data for analysis
Withdrawals/losses to follow up 7 withdrew after randomization, declining None amongst the 8 randomly allocated
the intervention to which they were assigned participants
(3 PSN, 4 conservative surgery)
Intention-to-treat analysis Not done (and not possible from the Apparently done by author since no
data given) women withdrawn after randomization
Power calculation Performed Not stated
Location of the trial Milan, Italy USA
Duration of follow-up 12 months 6 months
Baseline characteristics of
the studied groups
Age of women in the trial Mean 31.1 years (standard deviation Mean 31 years
3.6 years)
Inclusion criteria Women with endometriosis stage 3 or 4 Women with moderate to severe dysmenorrhoea
undergoing conservative surgery, moderate scheduled to undergo laparotomy for
or severe pelvic pain (dysmenorrhoea) conservative resection of endometriosis
Exclusion criteria Not stated Previous surgical treatment for endometriosis,
recent hypo-oestrogenizing medical therapy,
presence of other pelvic abnormalities which
could cause pain (pelvic inflammatory disease,
genital malformation, myomas)
Interventions
Nature of intervention PSN at laparotomy plus conservative PSN at laparotomy plus resection of
surgery vs. conservative surgery endometriosis vs. resection of endometriosis
at laparotomy at laparotomy
Outcomes
Method of expressing pain ratings 10-point visual analogue scale (but only Pain relief: yes or no
dichotomized data reported), and a
multidimensional pain scale
Timing of assessments Preoperatively; postoperatively at 12 months Preoperatively, postoperatively at 6 months
Other outcomes Adverse effects Incidence and location of pain
Adverse effects

control groups for dysmenorrhoea pain scores for both 9 months after surgery. The experimental group pain
the visual analogue scale (VAS) and the 10-point scale scores changed from a median of 75 (interquartile
(MannWhitney test, P 0.12 and P 0.21, respec- range 6785) to 38 (interquartile range 561), and the
tively). On the 10-cm VAS the experimental group pain control group scores changed from a median of 80
scores at 6 months had a median of 4.8 (range 19.0), (interquartile range 6587) to 22 (interquartile range
while the control pain scores had a median of 3.0 039). The authors interpreted these results as no real
(range 09.8). difference between the two groups. A 12-month
Vercellini et al. (1997)30 reported median and KaplanMeyer cumulative probability of recurrence
interquartile ranges for the participants' pain scores of moderate to severe dysmenorrhoea was 33.7% for
(100 mm VAS), preoperatively and at an average of the experimental group and 27.55% for the control

q2000 Blackwell Science Ltd Gynaecological Endoscopy 2000 9, 351361


358 N. JOH NSO N e t al .

Table 4 Summary of data from trials considered suitable for meta-analysis

Trial Experimental Control Odds ratio 95% confidence


group group interval
n/n (%) n/n (%)

Primary dysmenorrhoea
UNA vs. no UNA
Pain relief at less than 6 months' follow up A 9/11 (82) 1/10 (10) 15.5* 2.982.7
Pain relief at more than 6 months' follow-up A 5/11 (45) 0/10 (0) 10.9* 1.577.4
UNA vs. PSN
Pain relief at less than 6 months' follow-up B 29/35 (83) 29/33 (88) 0.67 0.182.6
Pain relief at more than 6 months' follow-up B 11/35 (31) 27/33 (82) 0.13* 0.050.35
Adverse effects B 0/35 (0) 31/33 (94) 0.02* 0.010.06
Secondary dysmenorrhoea (endometriosis)
PSN and endometriosis treatment vs.
endometriosis treatment
Pain relief at more than 6 months' follow-up C 29/35 (83) 27/36 (75) 1.59 0.154.9
Adverse effects C 18/35 (51) 0/36 (0) 14.6* 5.042.2

Experimental group refers to the first intervention and control group refers to the second intervention listed in the left-hand column.
Trials: A, Lichten B, Chen et al.28 (1992); C, Candiani et al.31 (1992).
UNA, uterine nerve ablation; PSN, presacral neurectomy.
*Significant difference between experimental and control groups.

group. An intention-to-treat analysis on subject satisfac- (Fisher exact test, P 0.03). However for back pain or
tion showed that 68% of the experimental group and lateral pain associated with dysmenorrhoea there were
73% of the control group were very satisfied or satisfied no significant differences between the groups. These
with treatment, while 32% of the experimental group results are based on only the eight randomized
and 27% of the control group were uncertain, participants. One woman experienced transient urin-
dissatisfied, or very dissatisfied with treatment. No ary urge incontinence for 2 weeks following PSN; no
adverse effects were reported for either group. complications occurred in the group who did not
For PSN with surgical treatment of endometriosis vs. undergo PSN.
surgical treatment of endometriosis alone, Candiani et
al. (1992)31 found no significant difference between
DISCUSSION
the experimental and control groups overall
(OR 1.6) (see Table 4). However the authors origin- This systematic review has drawn on the results of six
ally collected information on the incidence, site and RCTs examining surgical pelvic neuroablation for
severity of pain and in analysis split their results into chronic pelvic pain. Data were analysed from a total
separate areas of pain. They interpreted their findings of 295 women who were randomly allocated treatment
as showing a marked reduction in the recurrence of in these trials. Of these, 115 women underwent LUNA
midline abdominal dysmenorrhoeic pain; however, this (46 with primary dysmenorrhoea, 69 with endometrio-
was at the borderline of statistical significance sis), 33 underwent LPSN (all of whom had primary
(P 0.06). There was a highly significant difference dysmenorrhoea) and 39 underwent PSN at laparotomy
in the proportion of patients with adverse effects from (all of whom had endometriosis). Laparoscopic and
the treatment: the control group reported none but in open techniques could be combined for these inter-
the PSN group 13 women reported constipation, three ventions, as there is evidence to suggest they have
urinary urgency, and two experienced a painless first similar ranges of pain relief for dysmenorrhoea.18,20
stage of labour (OR 14.6). LUNA is effective for pain relief in primary
Tjaden et al. (1990)32 also originally collected dysmenorrhoea in the short term but its effectiveness
information on incidence and location of pain and may decline over time.5 Chen et al. (1996)28 confirm
classified pain relief into areas. They found that the this suggestion and demonstrate that LPSN may retain
experimental and control groups were significantly its effectiveness for longer. The importance of long-
different in pain relief for midline abdominal pain term follow up for future trials must be emphasised.

Gynaecological Endoscopy 2000 9, 351361 q2000 Blackwell Science Ltd


LUNA OR PRES ACRAL NEURECTOMY FOR CHRONIC PELVIC PAIN? 359

Neuroablative procedures appear less effective when (whether performed laparoscopically or at laparotomy)
used as an adjunct to standard conservative surgery for bring into question whether further work is required.
endometriosis. However PSN may be effective for The view of the authors is that clinicians should
midline pain associated with endometriosis, but not consider the situation very carefully before performing
for other types of pain such as non-menstrual pain, this operation, even in the context of a RCT. This
lateral pain and deep dyspareunia, and so is unlikely to review suggests that women with endometriosis are a
modify substantially the subjective and social impact of different population from women with primary dysme-
the condition. Careful study of Fig. 1 clarifies why PSN norrhoea when pain relief from LUNA is considered.
might be more effective than LUNA, since transection Ideally these populations should be studied separately.
of the superior hypogastric nerve plexus would ablate a For primary dysmenorrhoea, in order to have 80%
higher proportion of relevant afferent nerves than power at 95% confidence level to detect benefit in 50%
transection of the LeeFrankenhauser nerve plexus. It of women, assuming `benefit' in 10% controls, at least
also explains why UNA could be particularly ineffective 48 participants would be required for analysis following
for pain arising from the ovary or adjacent tissues, as randomization. For endometriosis, in order to have
could be the case in ovarian or paraovarian endome- 80% power at 95% confidence level to detect benefit in
triosis, since all ovarian afferent nerves bypass the Lee 90%, assuming benefit in 60% controls undergoing
Frankenhauser plexus and many of the pain fibres conventional endometriosis surgery,27 at least 76
bypass both of these nerve plexuses. Unless a periarter- participants would be required for analysis following
ial sympathectomy of the iliac, inferior mesenteric and randomization.
ovarian vessels is performed, a number of afferent The authors are aware of two further ongoing trials:
fibres will always be left intact. 1. A multicentre RCT to assess the efficacy of
More convincing is the high incidence of adverse laparoscopic uterosacral nerve ablation (LUNA) in
effects following PSN. Complications were reported in the treatment of chronic pelvic pain (University of
women who underwent PSN in all trials. Out of 33 Birmingham Clinical Trials Unit, UK). The trialists aim
women who underwent LPSN, 31 experienced con- to randomly allocate 120 women with minimal endo-
stipation, which was very severe in some cases.28 metriosis, minor adhesions or no visible pelvic pathol-
Amongst 35 women who underwent PSN, Candiani et ogy on laparoscopy.
al. (1992)31 subsequently reported 13 with constipa- 2. LUNA for chronic pelvic pain: an RCT at the
tion, three with urinary urgency, two with an asympto- University Department of Obstetrics and Gynaecology,
matic first stage of labour and one woman who National Women's Hospital, Auckland, New Zealand.
required a laparotomy 48 h following PSN, for a The trialists aim to randomly allocate 50 women with
presacral haematoma. Tjaden et al. (1990)32 reported primary dysmenorrhoea and 80 women with endome-
transient urinary urge incontinence in one out of four triosis.
women randomly allocated to PSN. No complications
were reported in women randomly allocated to LUNA!
LPSN is a surgical procedure which requires a high CONCLUSION
degree of skill from an experienced pelvic laparoscopic
There is insufficient evidence to recommend the use of
surgeon trained specifically in this retroperitoneal
surgical pelvic neuroablation in the management of
operation. The presacral region may be highly vascular
dysmenorrhoea, regardless of cause. Further scientifi-
and the procedure carries major potential hazards for
cally rigorous RCTs should be undertaken and the
the unwary or inadequately trained surgeon. The Royal
results of the above trials are eagerly anticipated.
College of Obstetricians and Gynaecologists classifies
LPSN as a level 4 procedure, requiring advanced skills.
Conversely, although LUNA must be performed pre-
A C K N O W L E DG E M E N T S
cisely to avoid complications, it should be within the
scope of all competent pelvic laparoscopic surgeons. The authors acknowledge the helpful comments of
There is definitely a place for further well-designed those who proofread previous versions of this review,
RCTs examining the intervention of LUNA, which and the authors of included trials who supplied extra
could be realistically available to all women. The high information and/or data. Special thanks are also due
degree of specialized skill required to perform LPSN to Owen Sinclair for help with data extraction, Anne
and the high likelihood of complications from PSN Lethaby for advice on statistics, Sarah Hetrick for help

q2000 Blackwell Science Ltd Gynaecological Endoscopy 2000 9, 351361


360 N. JOH NSO N e t al .

in coordinating proofreading of this paper and Sue 19 Feste JR. Laser laparoscopy: a new modality. Journal of
Furness for assistance with identifying trials. Reproductive Medicine 1985; 30: 4137.
20 Nezhat C, Nezhat F. A simplified method of
laparoscopic presacral neurectomy for the treatment of
R E F E R E N C ES central pelvic pain due to endometriosis. British Journal
of Obstetrics and Gynaecology 1992; 99: 65963.
1 Sobczyk R, Dysmenorrhea. The neglected syndrome. 21 Chen FP, Soong YK. The efficacy and complications of
Journal of Reproductive Medicine 1980; 25: 200. laparoscopic presacral neurectomy in pelvic pain.
2 Dawood MY. Nonsteroidal anti-inflammatory drugs and Obstetrics and Gynecology 1997; 90: 9745.
changing attitudes towards dysmenorrhea. American 22 Daniell JF, Lalonde CJ. Advanced laparoscopic
Journal of Medicine 1988; 84: 239. procedures for pelvic pain and dysmenorrhoea.
3 Lumsden MA. Dysmenorrhoea. In: Studd JWW, ed. Bailliere's Clinical Obstetrics and Gynecology 1995; 9: 795
Progress in Obstetrics and Gynaecology, Vol. 5. Edinburgh: 807.
Churchill Livingstone, 1985: 27692. 23 Papasakelariou C. Long-term results of laparoscopic
4 Dawood MY. Dysmenorrhea. Clinical Obstetrics and uterosacral nerve ablation. Gynaecological Endoscopy 1996;
Gynecology 1990; 33: 16878. 5: 1779.
5 Lichten EM, Bombard J. Surgical treatment of primary 24 Davis GD. Uterine prolapse after laparoscopic
dysmenorrhea with laparoscopic uterine nerve ablation. uterosacral transection in nulliparous airborne trainees:
Journal of Reproductive Medicine 1987; 32: 3741. a report of three cases. Journal of Reproductive Medicine
6 Rosenwaks Z, Seegar-Jones G. Menstrual pain: its origin 1996; 41: 27982.
and pathogensis. Journal of Reproductive Medicine 1980; 25 Good MC, Copas PR Jr, Doody MC. Uterine prolapse
25: 20712. after laparoscopic uterosacral transection: a case report.
7 Furniss LD. Nonsteroidal anti-inflammatory agents in Journal of Reproductive Medicine 1992; 37: 9956.
the treatment of primary dysmenorrhea. Clinical 26 Fitzpatrick CC, Flood H, Punch M, Hilgers TW, Elkins
Pharmacology 1982; 1: 32733. TE, McGuire EJ. Bladder dysfunction after repeat
8 Dawood MY. Dysmenorrhea. Pain and Analgesia 1985; 1: laparoscopic uterine nerve ablation (LUNA).
20. International Urogynecology Journal 1995; 6: 313.
9 Henzl MR. Dysmenorrhea: achievements and challenge. 27 Sutton CJG, Ewen SP, Whitelaw N, Haines P. Prospective
Sexual Medicine Today 1985; 9: 812. randomized double-blind controlled trial of laser
10 Owman C, Rosenbren E, Sjoberg NO. Adrenergic laparoscopy in the treatment of pelvic pain associated
innervation of the human female reproductive organs: a with minimal, mild and moderate endometriosis. Fertility
histochemical and chemical investigation. Obstetrics and and Sterility 1994; 62: 696700.
Gynecology 1967; 30: 76373. 28 Chen FP, Chang SD, Chu KK, Soong YK. Comparison of
11 Frankenhauser G. Die Bewegungenerven der laparoscopic presacral neurectomy and laparoscopic
Gebarmutter. Zentralblatt Medizin Nature Wissenschaft uterine nerve ablation for primary dysmenorrhea.
1864; 1: 35. Journal of Reproductive Medicine 1996; 41: 4636.
12 Cleland JP. Paravertebral anaesthesia in obstetrics. 29 Dover RW, Pooley P, Haines P, Sutton CJG. Prospective,
Surgery, Gynecology, and Obstetrics 1933; 57: 51. randomised, double-blind controlled trial of
13 Williams PL, Warwick R. Reproductive organs of the laparoscopic laser uterine nerve ablation in the
female. In: Williams PL, Warwick R, eds. Gray's Anatomy, treatment of pelvic pain associated with endometriosis.
36th edn. Edinburgh: Churchill Livingstone, 1980: Unpublished (ongoing).
142334. 30 Vercellini P, Aimi G, Busacca M, Uglietti A, Viganali M,
14 Sutton C, Whitelaw N. Laparoscopic uterine nerve Crosignani PG. Laparoscopic uterosacral ligament
ablation for intractable dysmenorrhoea. In: Sutton C, resection for dysmenorrhea associated with
Diamond M, eds. Endoscopic Surgery for Gynecologists. endometriosis: results of a randomized controlled trial
London: WB Saunders, 1993: 3965. (Abstract). Fertility and Sterility 1997; 68 (Suppl. 1): 3.
15 Doyle JB. Paracervical uterine denervation by 31 Candiani GB, Fedele L, Vercellini P, Bianchi S, Di Nola
transection of the cervical plexus for the relief of G. Presacral neurectomy for the treatment of pelvic
dysmenorrhoea. American Journal of Obstetrics and pain associated with endometriosis: a controlled study.
Gynecology 1955; 70: 11. American Journal of Obstetrics and Gynecology 1992; 167:
16 Wiborny R, Pichler B. Endoscopic dissection of the 1003.
uterosacral ligaments for the treatment of chronic 32 Tjaden B, Schlaff WD, Kimball A, Rock JA. The efficacy
pelvic pain. Gynaecological Endoscopy 1998; 7: 335. of presacral neurectomy for the relief of midline
17 Ewen SP, Sutton CJG. A combined approach to painful dysmenorrhoea. Obstetrics and Gynecology 1990; 76: 89
heavy periods: laparoscopic laser uterine nerve ablation 91.
and endometrial resection. Gynaecological Endoscopy 33 Buttram VC Jr, Reiter RC, eds. Endometriosis. In:
1994; 3: 1678. Surgical Treatment of the Infertile Female. Baltimore:
18 Perez JJ. Laparoscopic presacral neurectomy. Results of Williams and Wilkins, 1985: 89147.
the first 25 cases. Journal of Reproductive Medicine 1990; 34 Rosenshein NB, Rock J, eds. Presacral neurectomy. In:
35: 62530.

Gynaecological Endoscopy 2000 9, 351361 q2000 Blackwell Science Ltd


LUNA OR PRES ACRAL NEURECTOMY FOR CHRONIC PELVIC PAIN? 361

Surgery in the Retroperitoneal Space. Philadelphia: JB 36 Moore A, McQuay H, Gavaghan D. Deriving


Lippincott, 1988: 349. dichotomous outcome measures from continuous data
35 Revill SI, Robinson JO, Rosen M, Hogg MIJ. The in randomised controlled trials of analgesics:
reliability of linear analogue for evaluating pain. verification from independent data. Pain 1997; 69: 127
Anaesthesia 1976; 31: 11918. 30.

q2000 Blackwell Science Ltd Gynaecological Endoscopy 2000 9, 351361


Copyright of Gynaecological Endoscopy is the property of Wiley-Blackwell and its content may not be copied
or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.

You might also like