Prognostic factors in pelvic exenteration for gynecological malignancies
G. Baiocchia,*, G.C. Guimaraesb, R.A. Rosa Oliveiraa, L.Y. Kumagaia, C.C. Faloppaa,
S. Aguiarb, M.D. Begnamic, F.A. Soaresc, A. Lopesb
aDepartment of Gynecologic Oncology, AC Camargo Cancer Hospital, Sao Paulo, Brazil
bDepartment of Pelvic Surgery, AC Camargo Cancer Hospital, Sao Paulo, Brazil
cDepartment of Pathology, AC Camargo Cancer Hospital, Sao Paulo, Brazil
Accepted 2 July 2012
Available online 18 July 2012
Objectives: Analyze morbidity, mortality and prognostic factors after pelvic exenteration (PE) for gynecological malignancies.
Methods: We reviewed a series of 107 individuals who underwent PE at A.C. Camargo Cancer Hospital from August 1982 to September
Results: Median age was 56.4 years. Primary tumor sites were uterine cervix in 73 cases (68.2%); vaginal, 10 (9.3%); endometrial, 14
(13.1%); vulvar, 7 (6.5%); and uterine sarcomas, 3 (2.8%). Median tumor size was 5.5 cm. Total PE was performed in 56 cases
(52.3%), anterior in 31 (29.9%), posterior in 10 (9.3%) and lateral extended in 10. Median operation time, blood transfusion and hospital
stay length were 420 min (range: 180e780), 900 ml (range: 300e4500) and 13 days (range: 4e79), respectively. There was no intra-
operative death. Fifty-seven (53.3%) and 48 (44.8%) patients had early and late complications, respectively. Five-year progression free sur-
vival (PFS), overall survival (OS) and cancer specific survival (CSS) were 35.8%, 27.4% and 41.1%, respectively. Endometrial cancer had
better 5-year OS (64.3%) than cervical cancer (23.1%). Lymph node metastasis negatively impacted PFS, CSS and OS. Presence of peri-
neural invasion negatively impacted PFS and CSS. No variable retained the risk of recurrence or death in the multivariate analysis.
Conclusions: PE has acceptable morbidity and mortality and may be the only method that can offer long-term survival in highly selected
? 2012 Elsevier Ltd. All rights reserved.
Keywords: Pelvic exenteration; Gynecologic neoplasm; Prognostic factors; Morbidity; Mortality
Pelvic exenteration (PE) refers to radical en bloc resec-
tion of multiple endopelvic and exopelvic organs, followed
by surgical reconstruction, to reestablish visceral and pari-
etal function.1This procedure was first reported by Alexan-
der Brunschwig2in 1948 for the treatment of persistent or
recurrent gynecological cancer. Improved perioperative
management has recently reduced perioperative mortality
rates to 2%e14%.3e14
The most common indication for PE is persistent or re-
current cervical cancer after radiation therapy.1Twenty-five
percent of patients with stage IbeIIa cervical cancer who
are treated with radiation therapy experience a recurrence,15
but only few are suitable candidates for PE.
tient selection have effected 5-year survival rates following
PE that range from 20% to 73%.7e9,12,14,16e22Other ad-
vances in PE have led to tailored surgery that includes organ
sparing when there is no evidence of tumor involvement and
to extended operations in which part of the parietal pelvis is
after PE, such as continent urinary diversion and confection
of neovagina, have also been described.23e26
Several prognostic factors for survival have been re-
ported, such as tumor size, presence of lymphovascular in-
vasion, and positive margins.12,27The aim of our study was
to retrospectively analyze morbidity, mortality, and the
prognostic factors that negatively impact recurrence and
survival in patients who underwent pelvic exenteration in
AC Camargo Cancer Hospital.
* Corresponding author. Departamento de Ginecologia, Hospital AC Ca-
margo, Rua Antonio Prudente 211, 01509-010 S~ ao Paulo, Brazil. Tel.: þ55
11 2189 5110; fax: þ55 11 2114 6072.
E-mail address: email@example.com (G. Baiocchi).
0748-7983/$ - see front matter ? 2012 Elsevier Ltd. All rights reserved.
Available online at www.sciencedirect.com
EJSO 38 (2012) 948e954
Patients and methods
This retrospective analysis included 107 individuals with
primary and recurrent gynecological malignancies who un-
derwent PE at AC Camargo Cancer Hospital from August
1982 to September 2010. The study was approved by the
Institutional Review Board. Eighty-seven patients (81.3%)
had PE from January 2000 to September 2010. All patients
were treated with curative intent. Patients with extrapelvic
metastatic disease, retroperitoneal lymph node metastasis
or invasion of the pelvic sidewall that was unsuitable for re-
section with free margins were excluded.
PE was classified as anterior (APE), posterior (PPE), to-
tal (TPE), and total with laterally extended endopelvic re-
section (LEER). APE refers to the removal of the
reproductive tract and bladder; PPE is the removal of the
reproductive tract and rectum; and TPE is the removal of
reproductive tract, bladder, and rectum. LEER refers to
TPE that includes resection of the obturator internus mus-
cle, iliococcygeus muscle, or pubococcygeus muscle. The
number of organs or structures (vagina, uterus, bladder, re-
tosigmoid colon, rectum, urethra, anal sphincter, vulva and
small bowel) resected and involvement was also evaluated.
Postoperative morbidity was considered to be early if it
occurred earlier than 30 days after the operation or before
hospital discharge and late if it occurred after 30 days.
Follow-up time was the interval between the date of surgery
and the last date for which information was available. Mor-
bidity was analyzed per the National Cancer Institute (NCI)
common toxicity criteria.
The database was generated in SPSS, version 16.0. The
association between parametric variables was assessed by
chi-square or Fischer’s exact test. Survival curves were con-
structed by KaplaneMeier life table analysis. For all tests,
an alpha error up to 5% (p < 0.05) was considered
Clinical and pathological data
The patients’ clinical and pathological data are summa-
rized in Table 1.
Median age was 56.4 years (range: 28e87). Of the 107
patients who underwent PE, 97 experienced persistent or
recurrent disease, with a median interval from the first
treatment and PE of 18.8 months and a mean interval of
47.2 months (range: 1e365). Ten cases (9.3%) underwent
PE as the primary treatment, all of whom had stage IVA
disease, and 2 had vaginal sarcomas, 3 had vaginal squa-
mous cell carcinomas (SCC), 4 had cervical cancers, and
1 had endometrial cancer.
The primary tumor sites in the series were uterine cervix
in 73 cases (68.2%); vaginal, 10 (9.3%); endometrial, 14
(13.1%); vulvar, 7 (6.5%); and uterine sarcomas, 3
(2.8%). Sixty-seven patients (62.6%) had SCC, 33
(30.8%) had adenocarcinomas, 6 had sarcomas (3 uterine
and 3 vaginal), and 1 had vulvar melanoma.
Of the 97 patients that had persistent or recurrent dis-
ease, 17 (15.9%) were asymptomatic, and the most com-
mon symptom was pain (39 cases, 36.5%), followed by
bloody discharge (32 cases, 29.9%). The diagnosis was sus-
pected or confirmed by clinical examination in 32 cases
(32.9%). In 6 (35.3%) of the 17 asymptomatic patients,
the diagnosis was made solely by clinical examination.
Twenty-seven patients (25.2%) were classified as
American Association of Anesthesiologists (ASA) grade
III. TPE was performed in 56 cases (52.3%), APE in 31
(29.9%), PPE in 10 (9.3%), and LEER in 10 (9.3%). A
median of 4 organs was resected (range: 2e7), and a me-
dian of 3 organs (range: 1e6) had tumor involvement in
the pathology specimen. Of the 56 patients who under-
went TPE, 12 (21.4%) had anal sphincter-sparing surgery
and colorectal anastomosis and 1 (1.7%) had both anal
and urinary sphincter-sparing surgery with an orthotopic
Of the 66 patients who underwent TPE or LEER, 44
(66.7%) had concomitant diversion with double-barreled
wet colostomy (DBWC), as reported by Guimaraes
et al.28Empty pelvic floor treatment after TPE or LEER
was performed with cecum rotation in 14 cases (21.2%),
temporary mechanical support in 13 cases (19.7%) e as
previously reported,29temporary mechanical support that
was associated with cecum rotation in 10 cases (15.1%),
marlex mesh in 1 case (1.5%), and a musculocutaneous
flap in 1 case (1.5%).
Clinical and pathological characteristics of the 107 patients with gyneco-
logical cancer submitted to pelvic exenteration (PE).
Variable No. of patients (%)
Median age (y)56.4 years
(range, 28e87 years)
Primary site Cervix
Anal and urinary
Type of PE
Organ sparing after
Lymph node metastasis41
949G. Baiocchi et al./EJSO 38 (2012) 948e954
The median operation time was 420 min (range:
180e780), and 95 patients (88.8%) received a blood trans-
fusion (median 900 ml, range: 300e4500). The median
length of hospital stay was 13 days (range: 4e79).
Fifty-seven (53.3%) and 48 (44.8%) patients had early
and late complications, respectively. However, based on
NCI toxicity criteria, 19 (17.8%) had grade III or IV early
complications. Fourteen patients (13,1%) had late compli-
cations that required surgical intervention.
The patients’ data on complications are summarized in
There were no intraoperative deaths. Thirteen patients
(12.1%) died postoperatively, 9 of whom died (8.4%) be-
fore 30 days after surgery. The postoperative deaths were
due to sepsis (4 cases, 30.7%), renal failure (2), abdominal
dehiscence (2), surgical site bleeding (2), gastric bleeding
(1), a cerebral vascular accident (1), and undetermined (1
Age and ASA were the only variables that correlated
with the risk of postoperative death. ASA III patients had
a 33.3% mortality rate versus 5% for ASA I and II
(p < 0.001), and patients aged over 70 years had
a 42.1% mortality rate compared with 5.7% for younger pa-
tients (p < 0.001). Further, 57.9% of patients aged over 70
years were also ASA III, for whom the risk of postoperative
death was 63.6%.
Thirty-three cases (30.8%) had grade 3 tumors. Median
tumor size was 5.5 cm (range: 1e24). Sixty-one patients
had pelvic and/or retroperitoneal lymph nodes that were re-
sected; a median of 9 lymph nodes were evaluated (range:
1e136). Twenty cases (32.8%) had lymph node involve-
ment. Sixty-eight patients were evaluated for vascular and
perineural invasion. Ten (14.7%) and 35 (51.5%) patients
had positive vascular and perineural invasion, respectively.
Twenty-six (39.4%) of 66 patients had lymphatic invasion.
Surgical margins were considered microscopically free of
disease (R0) in 93 patients (92.1%) and positive or involved
(R1) in 8 cases (7.9%).
The median follow-up time was 25.7 months (range:
1e122). Forty-eight (44.9%) patients experienced a recur-
rence. At the end of the follow-up, 30 patients (28%)
were alive with no evidence of disease, 39 (36.4%) had
died due to cancer, 18 (16.8%) died of other causes, and
7 (6.5%) were alive with evidence of disease.
Recurrence and survival
The 2- and 5-year progression-free survival (PFS) rates
were 50.5% and 35.8%, respectively. The 2- and 5-year
overall survival (OS) rates were 49.9% and 27.4%, respec-
tively. The 2- and 5-year cancer-specific survival (CSS)
rates were 68.2% and 41.1%, respectively.
Patients with endometrial cancer had a better 5-year OS
(64.3%)than those with
(p ¼ 0.016) (Fig. 1). When primary exenteration for cervi-
cal cancer was excluded, the 5-year OS was 24.7%. Re-
garding CSS, endometrial cancer patients had a 5-year
CSS of 79.6% versus 32.9% in cervical cancer patients
(p ¼ 0.017).
Being asymptomatic prior to PE was associated with
a better PFS (p ¼ 0.028). Lymph node involvement (5-
year PFS: 29.4% vs. 41.7%; p ¼ 0.017) and perineural in-
vasion (5-year PFS: 27% vs. 49.8%; p ¼ 0.041) negatively
impacted the risk of recurrence. Grade 3 disease (5-year
OS: 18.9% vs. 33.6%; p ¼ 0.045), lymph node involvement
(2-year OS: 44,9% vs. 64%; p ¼ 0.011) (Fig. 2), and in-
volvement of more than 3 resected organs (2-year OS:
34.3% vs. 53.2%; p ¼ 0.005) negatively impacted the
risk of death. The median OS for patients with lymph
Type of early and late complications of the 107 patients submitted to pelvic
TypeNo. of patients(%)
Deep venous trombosis
Pelvic floor infection
Reoperation (grade IV)
Deep venous trombosis
Pelvic floor infection
Reoperation (grade IV)
Figure 1. Overall survival curves for patients with primary endometrial and
cervical cancer (p ¼ 0.016).
950G. Baiocchi et al./EJSO 38 (2012) 948e954
node involvement was 23.1 months and no patient survived
more than 36 months. Lymph node involvement (2-year
CSS: 50% vs. 63.8%; p ¼ 0.001), perineural invasion (3-
year CSS: 32.5% vs. 73.7%; p ¼ 0.004), and involvement
of more than 3 resected organs (3-year CSS: 24.6% vs.
59.6%; p ¼ 0.048) also negatively impacted the risk of
death from cancer.
Tumor size >5 cm did not correlate with the risk of re-
currence (p ¼ 0.52), risk of death (p ¼ 0.19), or risk of
death from cancer (p ¼ 0.066).
Patients for whom the interval from the first treatment to
PE was greater than 24 months had a better OS (p ¼ 0.053)
and CSS (p ¼ 0.062). However, it did not achieve statistical
Negative margin (R0 resection) did not correlate with
a lower risk of recurrence (p ¼ 0.093), death (p ¼ 0.23),
or death from cancer (p ¼ 0.057). When patients with cer-
vical or vaginal cancer were analyzed separately, R0 resec-
tion positively impacted PFS (2-year CSS: 16.7% vs.
41.6%; p ¼ 0.026) and CSS (2-year CSS: 33.3% vs.
65.2; p ¼ 0.024) but not OS (p ¼ 0.21).
No variable retained the risk of recurrence or death in
the multivariate analysis.
Exenteration is a major surgical procedure, entailing
long operating times that range from 5 to 14 h,3,4,6,9,11,27
significant blood loss (2300e4000 cm3),3,6,11e13,27and pro-
longed hospital stays.3,9,27
We have noted 5-year overall and cancer-specific sur-
vival rates of 27.4% and 41.1%, respectively, consistent
with other series, in which median 5-year survival has
ranged from 30% to 60%.3,7,9,11,12,16,22,30e33Our data re-
flect a heterogeneous population with associated illnesses,
in which 25.2% of subjects were ASA III. We also included
extremely locally advanced tumorsdthe median tumor size
in our series was 5.5 cm, and 9.3% underwent exenterative
procedures that extended to the pelvic sidewall.
In contrast to other series, we observed better survival
rates for subjects with endometrial cancer (5-year OS
64.3%).22Despite its disparate mechanisms and metastatic
pathways compared with cervical cancer, we recommend
that PE be an option for select patients with recurrent endo-
metrial cancer with pelvic disease who are unsuitable for
other curative treatments.
The perioperative mortality within 30 days of PE is
0%e14%.3e14,22We noted a high postoperative death
rate that correlated with age over 70 years and preexisting
medical illness (ASA III). Matthews et al.19reported a se-
ries of 63 patients aged over 65 years who underwent PE,
observing an operative mortality rate of 11%. Our findings
support other reports3,34that have suggested that advanced
age and the presence of systemic disease render these pa-
tients unsuitable candidates for this extensive procedure.
However, Maggioni et al.22demonstrated no postoperative
mortality, in agreement with other reports that have con-
cluded that physiological, rather than chronological, age
should be considered in selecting patients.19,35
Our morbidity rate after PE is consistent with reported
values of 32%e84%,11,22,31with a reintervention rate of
26%e32%.36In our series, the rate of early and late com-
plications was 53.3% and 44.8%, respectively. However,
based on NCI toxicity criteria, only 17.8% had grade III
or IV early surgical complications.
The most common early complications were related to
wound (29.8% of 57 patients) or pelvic infections
(26.3%), but they were nearly always treated conserva-
tively. The most threatening complications were those
that affected the urinary and gastrointestinal systems. Eight
(7.4%) patients had intestinal fistulas, and 5 had urinary fis-
tulas (4.6%), similar to Maggioni et al. (4.6%).22
We reported29an alternative method of treating the
empty pelvic floor with a temporary mechanical support,
as performed in 23 cases. Briefly, a silicone expander de-
vice previously indicated as a urinary bladder modelator
was used. The 300 ml capacity apparatus presents a filiform
tube for filling/emptying. After total PE, we placed the de-
vice on the pelvic floor and the filling tube was externalized
through the perineum or abdominal wall. The device was
then filled with a saline solution to settle in the small pelvis.
Due to its size and mobility, whenever possible we dis-
sected the caecum from the right parietocolic space at about
5 cm from parietal peritoneum, turned it inferomedially and
placed it carefully and tensionless on the silicone device.
After the procedure we did a peritoneal reflection restora-
tion by suturing the pericecal peritoneum to the remaining
parietal peritoneum on the silicone expander. We removed
the silicone device with median of 10 days after PE. After
patient’s sedation, we emptied the device through its exter-
nalized filiform tube and then pulled and removed it.
Figure 2. Overall survival curves for patients with positive and negative
lymph node (LN) metastasis (p ¼ 0.011).
951G. Baiocchi et al./EJSO 38 (2012) 948e954
The patients who underwent TPE or LEER using a me-
chanical device were more likely to develop early or late
pelvic infectionsd53.8% of such cases experienced a pelvic
infection versus 18.6% of those without the device
(p ¼ 0.002). Further, our overall early pelvic infection
rate was 15.8%, and we believe that a musculocutaneous
flap, such as Vertical Rectus Abdominis Myocutaneous
(VRAM) flap, when associated with cecum transposition
is the ideal choice for most patients. The musculocutaneous
flap may substitute permanently the mechanical support
provided by the silicone expander as previously described.
The majority of late complications in our series was re-
lated to urinary diversions (33.3%), the most frequent of
which was hydronephrosis. Most patients were managed
conservatively or by positioning ureteral stents when mod-
erate hydronephrosis was found.
In our data, 30.8% of patients underwent any surgical in-
tervention (including stents) during follow-up, because
33.3% and 29.2% of early and late complications, respec-
tively, required surgery. Maggioni et al.22
a 23.3% rate of reoperation.
Margin status appears to be the factor that is most con-
sistently associated with prognosis.34,37Fotopoulou et al.30
observed that complete tumor resection correlated with bet-
ter overall and progression-free survival. Numa et al.38re-
ported a 5-year survival of 51.3% for an R0 resection
compared with 0% for an R1 resection. Further, in series
by Berek et al.3and Shingleton et al.,12no patient survived
3 years following exenteration when the margins of the sur-
gical specimen were tumor-involved. Lawhead et al.7re-
ported that 8 of 9 patients with tumor-involved resection
margins died from the disease. The series in Marnitz
et al.39included only cervical cancer patients, and the 2-
year survival rate was 10.2% for patients with positive mar-
gins and 55.2% (5-year survival rate of 44.8%) for those
who had negative margins (p ¼ 0.0057).
However, microscopic margin involvement can be de-
tected during the final histopathological review and cannot
be determined preoperatively in all cases.38In the Mag-
gioni series,22when cervical and vaginal cancer were re-
ported separately, 27 of 83 (32.5%) patients had positive
margins in the final pathological examination, and survival
rates differed between patients with involved margins (25%
at 5 years) and disease-free resection margins (60% at 5
years) (p ¼ 0.043). Nevertheless, in Rutledge and McGuf-
fee’s study,40negative margin status was associated with
longer disease-free survival (p ¼ 0.01) but not with CSS
We found that 7.9% of our patients had R1 margins and
that survival did not differ from patients who underwent PE
with microscopic negative resection margins (R0) on the
surgical specimen. This finding may be attributed to the
heterogeneity of the group. However, as reported in other
series, when we analyzed only cervical and vaginal cancers,
R0 resection impacted PFS (p ¼ 0.026) and CSS
(p ¼ 0.024) positively.
Tumor diameter at the preoperative clinical evaluation
has been reported to be a prognostic factor, although the ap-
propriate cutoff remains to be defined. Some groups have
proposed a cutoff of 3e5 cm.1,12,37Shingleton et al.12
and H€ ockel at al.4suggested that recurrent tumor diameter
can predict the prognosis, whereas Morley et al.,9Sym-
monds et al.,14and Maggioni et al.22failed to conclude
such a correlation.
In our series, the median tumor size was larger than in
other series (5.5 cm), and neither extreme (3 cm or 5 cm)
was a significant factor with regard to survival. In contrast,
tumor invasivenessdie, the involvement of more than 3 or-
gansdnegatively impacted survival; thus, these data may
be more significant than tumor size itself.
The prognostic significance of the interval between pri-
mary treatment and PE is controversial. It appears that tu-
mors that take longer to recur are more indolent and
more likely to be controlled or cured. Some studies have
shown that longer intervals are associated with a better
prognosis,3,9,12,13,39whereas others have not observed
such a correlation.11We noted better overall survival for
patients who recurred after 24 months of treatment, albeit
insignificantly. Notably, PFS after PE was better for asymp-
tomatic patientsdthe first such finding. It is unclear
whether asymptomatic patients are more likely to have
more indolent tumors or whether earlier disease develop-
ment is more amenable to salvage therapy.
Several groups have described lymph node metastasis as
an important negative prognostic factor,17,22,34,37,41and
others consider it a contraindication for PE.9,16,42Yet, Rut-
ledge and McGuffee40concluded that lymph node status
was unrelated to survival and that long-term survival can
be achieved, even with positive lymph node involvement,
if they are completely resected. In patients with positive
lymph nodes, 36% remained alive after 3 years of follow-
up versus 26% at 5 years. In addition, Marnitz et al.39
and Ketcham et al.6failed to describe lymph node involve-
ment as a significant factor of survival.
We consider lymph node status an important prognostic
factor and propose that lymph node evaluation might take
part of the surgical procedure to better stratify those pa-
tients who are more likely to recur and benefit from adju-
vant treatment trials. In our series no patient with lymph
node involvement had prolonged survival.
The preoperative determination of lymph node status re-
mains problematic, but preoperative radiological studies by
elvic metastasis,42should be mandatory for restaging. Since
2005, all patients who underwent PE in our institution have
had a PET-CT scan with negative findings for distant disease.
Overall, our series have a comparable sample size to the
most important studies of this kind and may add important
information to literature. Nevertheless, our data may help to
select patients to PE and better stratify those with higher
risk of recurrence and death after PE.
952G. Baiocchi et al./EJSO 38 (2012) 948e954
In summary, lymph node involvement and presence of
perineural invasion negatively impacted the risk of recur-
rence and death from cancer. Number of organs involved
rather than tumor size should be considered as a prognostic
factor. For cervical and vaginal cancer, R0 resection corre-
lates to better outcome. However, no variable retained the
risk of recurrence or death in the multivariate analysis. Fi-
nally, PE should not be offered to patients classified as ASA
III and also aged over 70 years.
Our series suggests that PE is a feasible therapeutic op-
tion that avoids intraoperative death and has acceptable
morbidity and mortality. Careful patient selection, meticu-
lous operative techniques, multiple-surgeon effort, intensive
postoperative management, and the creation of specific and
adequate referral centers remain the cornerstones of reduc-
ing the complication rates and maximizing the success of
this aggressive operation. PE might be the only approach
that effects long-term survival in highly select patients
with persistent or recurrent gynecological malignancies.
Conflict of interest statement
All authors declare that there is no conflict of interest.
1. H€ ockel M, Dornh€ ofer N. Pelvic exenteration for gynaecological tu-
mours: achievements and unanswered questions. Lancet Oncol 2006;
2. Brunschwig A. Complete excision of pelvic viscera for advanced car-
cinoma. Cancer 1948;1:177–83.
3. Berek JS, Howe C, Lagasse LD, et al. Pelvic exenteration for recurrent
gynaecologic malignancy: survival and morbidity analysis of the 45-
year experience at UCLA. Gynecol Oncol 2005;99(1):1539.
4. H€ ockel M. Laterally extended endopelvic resectiondNovel surgical
treatment of locally recurrent cervical carcinoma involving the pelvic
side wall. Gynecol Oncol 2003;91:369–77.
5. Houvenaeghel G, Moutardier V, Karsenty G, et al. Major complica-
tions of urinary diversion after pelvic exenteration for gynecologic
malignancies: a 23-year mono-institutional experience in 124 patients.
Gynecol Oncol 2004;92:680–3.
6. Ketcham AS, Deckers PJ, Sugarbaker EV, et al. Pelvic exenteration for
carcinoma of the uterine cervixdA 15-year experience. Cancer 1970;
7. Lawhead RA, Clark DG, Smith DH, et al. Pelvic exenteration for re-
current or persistent gynaecologic malignancies: a 10- year review of
(1972e1981). Gynecol Oncol 1989;33:279–82.
8. Magrina JF, Stanhope CR, Waever AL. Pelvic exenterations: supraleva-
tor, infralevator, and with vulvectomy. Gynecol Oncol 1997;64:130–5.
9. Morley GW, Hopkins MP, Lindenauer SM, et al. Pelvic exenteration,
university of Michigan: 100 patients at 5 years. Obstet Gynecol 1989;
10. Saunders N. Pelvic exenteration: by whom and for whom? Lancet
11. Sharma S, Odunsi K, Driscoll D, et al. Pelvic exenteration for gyneco-
logical malignancies: twenty-year experience at Rosewell Park Cancer
Institute. Int J Gynecol Cancer 2005;15:475–82.
12. Shingleton HM, Soong SJ, Gelder MS, et al. Clinical and histopatho-
logic factors predicting recurrence and survival after pelvic exentera-
tion for cancer of the cervix. Obstet Gynecol 1989;73:1027–34.
Cancer Center experience
13. Stanhope CR, Symmonds RE. Palliative exenterationdWhat, when,
and why? Am J Obstet Gynecol 1985;152:12–6.
14. Symmonds RE, Pratt JH, Webb MJ. Exenterative operations: experi-
ence with 198 patients. Am J Obstet Gynecol 1975;121(7):907–18.
15. Landoni F, Maneo A, Colombo A, et al. Randomised study of radical
surgery versus radiotherapy for stage IbeIIa cervical cancer. Lancet
16. Rutledge F, Smith J, Wharton J, et al. Pelvic exenteration; analysis of
296 patients. Am J Obstet Gynecol 1977;129:881–92.
17. Roberts W, Cavanagh D, Bryson P, et al. Major morbidity after pel-
vic exenteration: a seven year experience. Obstet Gynecol 1987;69:
18. Hatch KD, Gelder MS, Soong SJ, et al. Pelvic exenteration with low
rectal anastomosis: survival, complications, and prognostic factors.
Gynecol Oncol 1990;38:462–7.
19. Matthews C, Morris M, Burke T, et al. Pelvic exenteration in the el-
derly patient. Obstet Gynecol 1992;79:773–7.
20. Morris M, Alvarez RD, Kinney WK, et al. Treatment of recurrent ad-
enocarcinoma of the endometrium with pelvic exenteration. Gynecol
21. Miller B, Morris M, Rutledge F, et al. Aborted exenterative procedures
in recurrent cervical cancer. Gynecol Oncol 1993;50:94–9.
22. Maggioni A, Roviglione G, Landoni F, et al. Pelvic exenteration: ten-
year experience at the European Institute of Oncology in Milan. Gyne-
col Oncol 2009;114(1):64–8.
23. Penalver M, Bejany D, Averette H, et al. Continent urinary diversion
in gynecologic oncology. Gynecol Oncol 1989;34:274–88.
24. Husain A, Curtin J, Brown C, et al. Continent urinary diversion and
low-rectal anastomosis in patients undergoing exenterative proce-
dures for recurrent gynecologic malignancies. Gynecol Oncol
25. Berek JS, Hacker NF, Lagasse LD. Vaginal reconstruction performed
simultaneously with pelvic exenteration. Obstet Gynecol 1984;63:
26. Hatch KD. Construction of a neovagina after exenteration using the
vulvobulbocavernosus myocutaneous graft. Obstet Gynecol 1984;63:
27. Hatch KD, Berek JS. Pelvic exenteration. In: Berek JS, Hacker NF, ed-
itors. Berek & Hacker’s gynecologic oncology. 5th ed. Philadelphia:
Lippincott Williams and Wilkins; 2010, p. 803–17.
28. Guimaraes GC, Ferreira FO, Rossi BM, et al. Double-barreled wet co-
lostomy is a safe option for simultaneous urinary and fecal diversion.
Analysis of 56 procedures from a single institution. J Surg Oncol
29. Guimaraes GC, Baiocchi G, Rossi BM, et al. The use of silicone ex-
pander and cecal transposition after pelvic exenteration. Eur J Surg
30. Fotopoulou C, Neumann U, Kraetschell R, et al. Long-term clinical
outcome of pelvic exenteration in patients with advanced gynecolog-
ical malignancies. J Surg Oncol 2010;101:507–12.
31. Pawlik TM, Skibber JM, Miguel AR. Pelvic exenteration for advanced
pelvic malignancies. Ann Surg Oncol 2005;13(5):612–23.
32. Goldberg GL, Sukumvanich P, Einstein MH, et al. Total pelvic exen-
teration: the Albert Einstein College of Medicine/Montefiore Medical
Center Experience (1987e2003). Gynecol Oncol 2006;10:261–8.
33. Averette HE, Lichtinger M, Sevin BU, et al. Pelvic exenteration: a 15
year experience in a general metropolitan hospital. Am J Obstet Gyne-
34. Fleisch MC, Pantke P, Beckmann MW, et al. Predictors for long-term
survival after interdisciplinary salvage surgery for advanced or recur-
rent gynecologic cancers. J Surg Oncol 2007;95:476–84.
35. Bladou F, Houvenaeghel G, Delpero JR, et al. Incidence and manage-
ment of major urinary complications after pelvic exenteration for gy-
necological malignancies. J Surg Oncol 1995;58:91–6.
36. Soper JT, Berchuck A, Creasman WT, et al. Pelvic exenteration:
factors associated with major surgical morbidity. Gynecol Oncol
953 G. Baiocchi et al./EJSO 38 (2012) 948e954
37. Park JY, Choi HJ, Jeong SY, et al. The role of pelvic exenteration and Download full-text
reconstruction for treatment of advanced or recurrent gynecologic ma-
lignancies: analysis of risk factors predicting recurrence and survival.
J Surg Oncol 2007;96(7):560–8.
of gynecological malignancies. Arch Gynecol Obstet 1997;259:133–8.
39. Marnitz S, K€ ohler C, M€ uller M, et al. Indications for primary and sec-
ondary exenterations in patients with cervical cancer. Gynecol Oncol
40. Rutledge FN, McGuffee VB. Pelvic exenteration: prognostic signifi-
cance of regional lymph node metastasis. Gynecol Oncol 1987;26:
41. Barakat RR, Goldman NA, Patel DA, et al. Pelvic exenteration for re-
current endometrial cancer. Gynecol Oncol 1999;75(1):99–102.
42. Husain A, Akhurst T, Larson S, et al. A prospective study of the accu-
(18FDGPET) in identifying sites of metastasis prior to pelvic exenter-
ation. Gynecol Oncol 2007;106(1):177–80.
954 G. Baiocchi et al./EJSO 38 (2012) 948e954