Gynecologic Oncology 106 (2007) 362 – 369

www.elsevier.com/locate/ygyno

Treatment of (“bulky”) stage IB cervical cancer with or without neoadjuvant

vincristine and cisplatin prior to radical hysterectomy and pelvic/para-aortic
lymphadenectomy: A phase III trial of the gynecologic oncology group
Gary L. Eddy a,⁎, Brian N. Bundy b,1 , William T. Creasman c , Nick M. Spirtos d ,
Robert S. Mannel e , Edward Hannigan f , Dennis O’Connor g
a
Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Mercer University School of Medicine,
840 Pine Street, Suite 760, Macon, GA 31201, USA
b
GOG Statistical and Data Center, Roswell Park Cancer Institute, Buffalo, NY 14263, USA
c
Medical University of South Carolina, Charleston, SC 29425, USA
d
Division of Gynecologic Oncology, University of Nevada School of Medicine and the Women’s Cancer Center of Nevada,
Las Vegas, NV 89109, USA
e
Department of Obstetrics and Gynecology, University of Oklahoma Health Sciences Center, Oklahoma City, OK 73190, USA
f
Division of Gynecologic Oncology, University of Texas, Galveston Medical Branch, Galveston, TX 77555, USA
g
Norton Healthcare Inc., Clinical Pathology Associates, Louisville, KY 40207, USA
Received 27 September 2006
Available online 9 May 2007

Abstract

Objective. A randomized phase III trial was conducted to determine if neoadjuvant chemotherapy (NACT) prior to radical hysterectomy and
pelvic/para-aortic lymphadenectomy (RHPPL) could improve progression-free survival (PFS) and overall survival (OS), as well as operability,
with acceptable levels of toxicity. Adjuvant radiation therapy was prescribed for specific surgical/pathological risk factors for both regimens.
Methods. Eligible patients were required to have bulky FIGO Stage IB cervical cancer, tumor diameter ≥ 4 cm, adequate bone marrow, renal
and hepatic function, and performance status ≤2. Prospective random allocation was to either NACT (vincristine-cisplatin chemotherapy every
10 days for 3 cycles) before exploratory laparotomy and planned RHPPL (NACT + RHPPL), or RHPPL only.
Results. The study was closed prematurely, because of slow accrual, after 291 patients were enrolled, three were ineligible; thus 288 were
eligible and randomly allocated to RHPPL (N = 143) or NACT + RHPPL (N = 145). There were no notable differences between regimens with
regard to patient age, race, performance status, or tumor size. The median follow-up time is 62 months among living patients. The NACT +
RHPPL group had very similar recurrence rates (relative risk: 0.998) and death rates (relative risk: 1.008) when compared to the RHPPL group.
There were 79% that had surgery in the RHPPL group compared to 78% in the NACT RHPPL group. There were 52% who received post op RT in
the RHPPL group compared to 45% in the NACT + RHPPL group (not statistically significant).
Conclusion. There is no evidence from this trial that NACT offered any additional objective benefit to patients undergoing RHPPL for
suboptimal Stage IB cervical cancer.
© 2007 Elsevier Inc. All rights reserved.

Keywords: Suboptimal stage IB cervical cancer; Vincristine/cisplatin; Survival; GOG 141

Introduction

⁎ Corresponding author. Fax: +1 478 633 2175.

Tumor volume has a detrimental effect on survival in patients
E-mail address: eddy.gary@mccg.org (G.L. Eddy).
with Federation of Gynaecology and Obstetrics (FIGO) Stage IB
1
Dr. Bundy's current position is Associate Professor, Pediatrics Epidemio- cervical cancer. Plentyl and Friedman reported an 85% 5-year
logy Center, University of South Florida, Tampa, FL 33612, USA. survival in their retrospective review of 31 studies in which 3391
0090-8258/$ - see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.ygyno.2007.04.007
 G.L. Eddy et al. / Gynecologic Oncology 106 (2007) 362–369
women were treated for Stage IB disease [1]. In patients with
tumor diameter N 3 cm, Piver and Chung reported an 80% 5-year
survival rate [2], while Ballon et al. observed a 5-year survival
comparable to Stage IIIB patients (43% vs. 48%, respectively)
when lesion size N6 cm [3]. The compounding effect of stromal
invasion was noted by Gauthier and colleagues [4]. Patients with
small (b 2 cm diameter) lesions regardless of depth of stromal
invasion, and patients with intermediate sized (2.1 to 3.0 cm
diameter) lesions with stromal invasion ≤ 1.5 cm had a 97%
5-year survival rate. When lesion diameter was N3 cm but
stromal invasion remained ≤1.5 cm, the 5-year survival rate fell
to 7l%. In patients with both lesion diameter N 3 cm and stromal
invasion N 1.5 cm, the 5-year survival rate dropped further, to
31%. This combination of greater lesion size and deep stromal
invasion which occurs frequently in bulky FIGO Stage IB cervical
cancer, demands that new therapeutic approaches be explored to
improve outcome in this patient population.
One such approach is neoadjuvant chemotherapy (NACT). In
a preliminary study we reported an 88% objective response rate
and a 12% complete clinical response rate in 34 evaluable Stage
IB patients whose lesion diameter was ≥4 cm and who received
preoperative vincristine and cisplatin chemotherapy [5]. Thirty-
two patients (94%) were able to undergo radical hysterectomy and
pelvic lymphadenectomy (RHPPL). Among 31 patients evaluable
for survival, 81% were presumed disease-free with a minimum
follow-up of 24 months. Two subsequent phase III randomized
trials have suggested the superiority of neoadjuvant chemother-
apy prior to radical hysterectomy and pelvic lymphadenectomy
compared to radiation therapy in patients with bulky Stage IB/IIA
disease [6,7]. Noteworthy are the GOG study by Keys et al. [8]
and an NCI Clinical Announcement, released together in 1999,
which all but established concomitant radiation therapy and
cisplatin-based chemotherapy as the standard of care for this
disease. The current trial was undertaken in 1996 to determine if
NACT using vincristine-cisplatin prior to RHPPL could improve
progression-free survival (PFS) and overall survival (OS), as well
as operability, with acceptable levels of toxicity.
Methods
Eligibility
Patients with primary, previously untreated, histologically confirmed invasive
squamous cell carcinoma, adenocarcinoma, or adenosquamous carcinoma of the
uterine cervix with an exophytic lesion ≥4 cm (or whose cervix was expanded to
≥ 4 cm and presumed by the clinician to be principally involved with cancer), were
eligible for this Gynecologic Oncology Group (GOG) study (GOG 141). Also
required were adequate bone marrow function (WBC ≥3000/mcl and platelets
≥ 100,000/μl) and hepatic function (bilirubin ≤1.5 times institutional normal and
serum glutamic oxaloacetic transaminase (SGOT) and alkaline phosphatase ≤3
times institutional normal). Cancer was to be limited to the cervix by clinical exam,
according to the system adopted by the FIGO [9] and by either CT scan of the
abdomen and pelvis or intravenous pyelogram, as well as chest radiograph. GOG
performance status of 0–2 was required. Patients could not have concomitant
malignancy or prior invasive malignancy, other than non-melanoma skin cancer,
with evidence of disease within 5 years of the current cervical cancer diagnosis.
Other exclusion criteria included pregnancy, prior pelvic irradiation or cytotoxic
chemotherapy, septicemia or severe infection, prior hysterectomy, renal abnor-
malities requiring modification of radiation fields, intestinal obstruction or
gastrointestinal bleeding, or metastases to any extrauterine site other than the
363
pelvic nodes. Patients whose circumstances would not permit completion of the
study or required follow-up, or who were unsuitable candidates for RHPPL as
previously described by the GOG [10], were also ineligible. This study was
approved by participating institutions’ Internal Review Boards consistent with all
federal, state, and local requirements prior to enrolling any patient; and written
informed consent was provided by all patients prior to receiving any protocol
therapy.
Surgery with or without chemotherapy
All patients were to have radical hysterectomy and pelvic and para-aortic
lymphadenectomy only after surgical exploration with evaluation of possible
intraperitoneal disease, para-aortic lymph node metastases, extrauterine extension,
or unresectable pelvic lymph nodes (RHPPL), in accordance with the GOG
Surgical Manual. The RHPPL was comparable to a Type III hysterectomy as
described by Piver et al. [11]. Intraoperative histologic confirmation of any of these
findings required termination of surgery. Patients with intraperitoneal metastases
were treated on an individual basis at the discretion of their physician. All others
were treated with the chemoradiation therapy as described below. Half the patients
were randomly allocated to receive NACT followed within 2–4 weeks by RHPPL
(NACT+ RHPPL). NACT consisted of intravenous vincristine 1 mg/m2 and
cisplatin 50 mg/m2 every 10 days for 3 cycles.
Radiation therapy
Upon completion of the prescribed RHPPL, patients with positive pelvic
lymph nodes or parametrial margins were treated with standardized radiation
therapy delivering 4500 cGy to the whole pelvis. Patients whose para-aortic
nodes were positive on final surgical pathology following RHPPL received
extended field external beam therapy delivering 4500 cGy to the pelvis and para-
aortic nodes in continuous 180 cGy daily fractions. A four field technique with
anterior/posterior and two lateral opposed fields was preferable for all patients
receiving external beam therapy. Low Dose Rate brachytherapy using a standard
tandem and vaginal ovoids was given in one or two applications to raise the point
A dose to a total of 8500 cGy and point B dose to 6000 cGy. All radiation therapy
was to be completed within 7 weeks of its initiation.
Patients on NACT + RHPPL whose disease progressed beyond the cervix
during chemotherapy were treated with standardized pelvic chemoradiation
consisting of 8500 cGy to point A and 6000 cGy to point B with a combination of
4500 cGy external beam therapy delivered homogenously to the pelvis in 4–
5 weeks with 5 day/week fractions of 180–200 cGy with beam energy ≥4 meV
and intracavitary irradiation in one or two applications using standard
afterloading tandem and ovoid applicators. Cisplatin 40 mg/m2 I.V. to a
maximum dose of 70 mg I.V. was given weekly during external beam radiation
and once during intracavitary application, to a maximum of 6 doses.
Patients were physically examined prior to study entry, before each cycle of
chemotherapy, and immediately before planned RHPPL. Following surgery,
patients were examined every 3 months for 2 years, then every 6 months for the
next 3 years, and yearly thereafter.
Data submission and review
Histologic confirmation of diagnosis was performed on central review by the
GOG Pathology Committee. Participating institutions were required to submit
copies of pathology reports and slides, operative reports, discharge summaries,
and follow-up data to the GOG Statistical and Data Center (SDC). Eligibility was
confirmed by the GOG GYN Committee. Data relative to protocol treatment,
toxicities, recurrence, and survival were reviewed by the Study Chair (GE) to
assure adherence to the protocol and accuracy and quality of submitted data.
Measuring outcomes
Adverse effects of chemotherapy and surgery together were graded according
to standard GOG toxicity criteria [12]. All eligible patients were considered
evaluable for adverse effects to preserve comparability between regimens.
Progression-free survival (PFS) was defined as the date from protocol registration
to the date of reappearance of disease, clinical progression of existing disease, or
364 G.L. Eddy et al. / Gynecologic Oncology 106 (2007) 362–369
death, whichever comes first. Clinical progression was defined as a 50% or greater
increase in the cross product of the tumor compared to the baseline product. Overall
survival (OS) was defined as the length of life from entry onto the study to death, or
to the date of last contact. Response was recorded for those on the NACT+ RHPPL
regimen and presented in this report to compare with previous clinical trials of this
experimental regimen. Surgical outcome was based on intraperitoneal gross
assessment and surgical pathologic findings with emphasis on parametrial, vaginal,
cervical tissue and pelvic, and para-aortic lymph nodes. Sites of recurrence/
progression were grouped as local and distant. Local was considered to be confined
to the pelvis, cervix and/or vagina. Distant included any site other than local,
including lung, abdomen, vulva, bone and brain, and groin, supraclavicular,
scalene, mediastinal and axillary lymph nodes.
Statistical design
The recurrence hazard rate for the RHPPL group was based on a previous
GOG study (GOG 49). A 33% decrease (Δ = 1.5) in the recurrence hazard rate by
NACT was considered a clinically important improvement. This difference
would have required a total of 167 recurrences [13] (96 in the RHPPL group) to
produce a statistical power of 0.835 using the log rank test [14] and setting the
probability of type I error at 0.05 (one-sided test). The slightly greater statistical
power (over the nominal level of 80%) was set to compensate for interim futility
analysis. We anticipated accruing 415 patients over 5½ years with 3 years of
additional follow-up to observe the indicated number of recurrences.
Statistical analysis
For the interim analysis, the predetermined stopping rules for efficacy were
set to test at the 0.005 level at the interim (after observing 50 recurrences/deaths)
and at the 0.0475 level at the final analysis to control the overall type I error at 0.05
Fig. 1. CONSORT flow sheet.
(both one-sided) [15]. The futility stopping rule was to apply if the NACT + RHPPL
group showed any excess risk of progression compared to the RHPPL group, as
determined by the relative risk of progression estimate. Using computer
simulations, we estimated that the futility analysis reduced the statistical power
to 80% from the original fixed sample design above.
All eligible patients were included in the analysis of OS and PFS (intent-to-
treat principle). All causes of death were included in the calculation of survival,
and the estimates of the cumulative proportions of OS and PFS were based on
Kaplan–Meier procedure [16]. Adjusted and unadjusted relative risk (RR)
estimates and confidence intervals of progression and death by treatment were
accomplished using the Cox model [17]. The cumulative incidence curve was
used to evaluate the rate of site-specific recurrences between the regimens [18].
This method properly handles censored cases and determines the incidence rate
of local failure in the presence of distant (with or without local relapse) failure,
and of death due to intercurrent disease. The Wilcoxon Rank Sum test adjusted
for ties was used to test the independence of risk of severe and life-threatening
toxicity (grade 0–2 vs. grade 3 vs. grade 4) and the assigned treatment group
[19]. The p-values and confidence intervals included in this report are not
adjusted for the interim analysis conducted due to the inconsequential
difference.
Results
The study was activated in December, 1996. There were 291
patients enrolled onto this study, of whom 144 were randomly
allocated to RHPPL alone and 147 to NACT + RHPPL. In
January of 1999, the Data Monitoring Committee (DMC) met to
consider the interim analysis conducted on 52 recurrences.
Since neither stopping boundary was crossed, the DMC voted to
 G.L. Eddy et al. / Gynecologic Oncology 106 (2007) 362–369 365

Table 1 Table 3
Patient characteristics High risk surgical-pathologic findings
RHPPL NACT + RHPPL RHPPL NACT + RHPPL All
(N = 143) (N = 145) (N = 143) (N = 145) (N = 288)
No. (%) No. (%) No. (%)
No. (%) No. (%)
Extrauterine
Age at entry
Positive pelvic nodes 56 (39) 47 (32) 103 (36)
≤ 30 13 (9) 21 (14)
Positive para-aortic nodes 20 (14) 15 (10) 35 (12)
31–40 49 (34) 45 (31)
Parametrial involvement 28 (20) 24 (17) 52 (18)
41–50 47 (33) 47 (32)
Intraperitoneal disease 9 (6) 7 (5) 16 (6)
51–60 21 (15) 22 (15)
Uterine
≥ 61 13 (9) 10 (7)
N1/3 cervical penetration 26 (18) 31 (21) 57 (20)
Race
CLS+ involvement 62 (43) 57 (39) 119 (41)
White 88 (62) 90 (62)
Positive surgical margin 15 (10) 13 (9) 28 (10)
Black 16 (11) 19 (13)
+
Hispanic 26 (18) 28 (19) Capillary-lymphatic space.
Asian/Pacific islander 11 (8) 6 (4)
Other 2 (1) 2 (1)
Performance status
0 121 (85) 118 (81) were deemed ineligible: one in each regimen with the wrong
1 22 (15) 27 (19) primary and one in the NACT + RHPPL group with the wrong
Tumor diameter
cell type. This report is based on the observation of 113 patients
4.0 25 (17) 21 (14)
4.5 14 (10) 8 (6) with disease progression (plus 12 patient deaths without
5.0 40 (28) 53 (37) documented recurrence) and a total of 104 deaths.
5.5 3 (2) 6 (4) The CONSORT flow chart is presented in Fig. 1. Patient and
6.0 35 (24) 36 (25) disease characteristics for the eligible patients (RHPPL, 143;
6.5 2 (1) 2 (1)
NACT + RHPPL, 145) appear in Table 1; there were no note-
7.0 14 (10) 9 (6)
7.5 3 (2) 1 (1) worthy differences between the two regimens. The frequency
≥ 8.0 7 (5) 9 (6) of performing RHPPL is similar between regimens (Table 2,
Histology RHPPL: 79%; NACT +RHPPL: 78%). Six patients in the NACT
Squamous cell 110 (77) 113 (78) +RHPPL group compared with one patient in the RHPPL group
Adenosquamous 14 (10) 13 (9)
did not receive RHPPL in violation of the protocol (did not meet the
Adenocarcinoma 12 (8) 9 (6)
criteria for contraindication). Surgical–pathologic high risk factors
are reported in Table 3. There are no significant differences between
continue the trial. However, poor accrual and a chronic problem the two arms.
with off-protocol use of radiation therapy led the GOG Protocol
Committee to close the trial in September of 2001, having
achieved only 70% (291/415) of the accrual goal. Three patients Table 4
Radiation therapy status and reason indicated ⁎
RHPPL NACT + RHPPL
Table 2
(N = 143) (N = 145)
Hysterectomy status and reason not performed
No. (%) No. (%)
RHPPL NACT + RHPPL
(N = 143) (N = 145) None 69 (48) 80 (55)
Protocol authorized-RT ▪
No. (%) No. (%)
Extension beyond cervix 26 (18) 20 (14)
Performed reason abandon ⁎ 113 (79) 113 (78) Unresectable pelvic nodes 2 (1) 0 (0)
Extension beyond cervix 25 † (17) 18 (12) Positive para-aortic nodes 7 (5) 5 (3)
Unresectable pelvic nodes 2 (1) 0 (0) Positive pelvic nodes 18 (13) 21 (14)
Positive para-aortic nodes 2 (1) 1† (1) Positive surgical margins 2 (1) 1 (1)
Parametrial involvement 0 (0) 1† (1) Clinical progression § – (–) 2 (1)
Clinical progression ‡ – (–) 2 †† (1) RT not authorized ▪
Chemotherapy adverse effects ‡ – (–) 1† (1) Microscopic parametrial involvement 13 (9) 10 (7)
Patient noncompliant 0 (0) 3† (2) Intermediate risk factors † 6 (4) 5 (3)
Positive pelvic nodes § 0 (0) 5 ††† (3) No explanation provided 0 (0) 1 (1)
No reason given § 1† (1) 1† (1) Received RT total 74 (52) 65 (45)
Not done total 30 (21) 32 (22) ⁎ Patients with multiple reasons are included in the category highest on the
⁎ Patients with multiple reasons were included in the category highest on the list.
†
list. Either depth of cervical invasion was greater than 1/3 or capillary-lymphatic
§
Abandoning the hysterectomy based solely on this reason was considered a space involvement with a tumor diameter of ≥5 cm (based on GOG 92
protocol violation. eligibility criteria).
† §
Represents one patient who had no surgical procedures performed. A total of Applicable to NACT + RHPPL regimen only.
▪
2 in the RHPPL group and 10 in the NACT + RHPPL group. Authorized: protocol explicitly prescribed radiation therapy with the
‡
Applicable to NACT + RHPPL regimen only. indicated risk factors.
366 G.L. Eddy et al. / Gynecologic Oncology 106 (2007) 362–369

Table 5
Grade 3 and 4 adverse events
RHPPL NACT + RHPPL
(N = 143) (N = 145)
No. (%) No. (%)
Hematologic 15 (10) 22 (15)
GI 17 (12) 18 (12)
GU † 10 * (7) 1 ** (1)
CV 1 (1) 1 (1)
Pulmonary 3 (2) 1 (1)
Neurologic 2 (1) 2 (1)
Metabolic 0 (0) 3 (2)
Hepatic 1 (1) 1 (1)
Pain 0 (1) 2 (1)
Cutaneous 1 (1) 0 (0) Fig. 3. Overall survival by treatment group.
Lymphatics 0 (0) 3 (2)
Allergic 1 (1) 1 (1)
Other *** 2 (1) 2 (1)
†
p-value b0.01.
With a median follow-up time of 62 months among living
⁎ Includes UTI, urinary incontinence, hydronephrosis, radiation cystitis, patients, the primary endpoints of PFS (RHPPL: 60.4 and 53.8,
bladder atony, ureteral injury, and ureteral stent placement for hole in bladder. and NACT + RHPPL: 59.7 and 56.2, at 3 and 5 years,
⁎⁎ Bilateral ureteral injury.
respectively) and over-all survival (RHPPL: 69.3 and 60.7, and
⁎⁎⁎ Includes fatigue, abscess, hypertension, and lymphocyst (requiring
NACT + RHPPL: 67.7 and 63.3, at 3 and 5 years, respectively) are
surgery).
very similar between regimens (Figs. 2 and 3). This is further
substantiated by the hazard ratio (relative risk) estimates of
Although not a statistically significant difference, 45% of progression and death, which are essentially equal to 1.
patients in the NACT + RHPPL group received post-operative Regardless of the prognostic factors used for adjustment including
radiotherapy compared to 52% of patients in the RHPPL group. tumor diameter, age, aortic node status, and gross tumor extension
A substantial percentage of patients in both regimens (13%, at surgery, there was no evidence that the NACT + RHPPL group
RHPPL; 11%, NACT + RHPPL) received radiotherapy absent experienced a reduction in risk. Also, no difference was noted by
of any protocol-specified indication for radiation therapy. One treatment group when the progression rate was partitioned into
of these indications (intermediate risk factor) was included as an local vs. distant sites of recurrence (Table 6). Among patients with
eligibility criterion in a recent GOG study and was first reported tumor diameter ≥5 cm, there were no significant differences in
by Sedlis et al. in 1999 [20] and updated by Rotman et al. [21] RHPPL operability or adjuvant radiotherapy received. The hazard
(Table 4), but there was no provision for it in the present trial. ratio of progression and death was tested for consistency across
Levels of toxicity were acceptable in this study and were tumor size (4–5 cm vs. ≥6 cm). This test for interaction was not
generally comparable between regimens. The frequency of significant for either endpoint (PFS, p = 0.75; OS, p = 0.50). The
grade 3–4 genitourinary adverse effect was statistically higher clinical partial and complete response frequency was 37% and
in the RHPPL group (7%, RHPPL; 1%, NACT + RHPPL). No 15%, respectively, among those in the NACT + RHPPL group. An
other category was significantly different; in fact, all other additional 38% of NACT + RHPPL patients had shrinkage of
categories with the exception of hematologic were very similar
(Table 5).
Table 6
Cumulative incidence of local and distant sites of recurrence
RHPPL NACT + RHPPL
(N = 143) (N = 145)
Site Year No. (Incidence rate †) No (Incidence rate †)
Local * only 1 19 (0.135) 21 (0.146)
2 23 (0.164) 27 (0.190)
3 24 (0.173) 29 (0.207)
Distant § 1 11 (0.079) 13 (0.091)
(with or without local) 2 22 (0.161) 21 (0.150)
3 26 (0.196) 22 (0.158)
Unknown site 1 0 (0.0) 1 (0.007)
2 0 (0.0) 1 (0.007)
3 0 (0.0) 1 (0.007)
⁎ Local includes cervix, vagina, and pelvic.
§
Distant includes lung, abdomen, vulva, bone, brain, and groin,
supraclavicular, scalene, mediastinal and axillary lymph nodes.
†
Cumulative incidence estimate of the site of progression at the indicated
Fig. 2. Progression-free survival by treatment group. time point.
 G.L. Eddy et al. / Gynecologic Oncology 106 (2007) 362–369
tumor but did not have the ratio of the final tumor cross-product
area to the initial tumor cross-product b 50%. Only 4 patients
(2.8%) had progressive disease in our study.
Discussion
The clinical objective response of 52% reported in this study
is substantially lower than the 88% reported in our pilot study of
34 evaluable patients using the same NACT regimen [5]. The
latter rate compares favorably with the 95% rate reported in
2003 using paclitaxel and cisplatin in a phase II trial of 43
patients receiving NACT prior to RHPPL and adjuvant cisplatin
chemoradiation [22]. The clinical complete response rate in this
trial compared favorably with our pilot study, 15% vs. 12%,
respectively. It was not anticipated that our clinical partial
response rate with vincristine and cisplatin NACT would be so
much lower in the group-wide study. It is possible that clinical
assessment of the cervix following NACT overestimated
residual disease. The dimensions of the entire cervix, not just
the portion of the cervix with residual disease may have been
reported by examiners less experienced comparing the imme-
diate pre-operative exam to the residual disease present on the
gross pathologic exam. An additional 38% of patients had
tumor shrinkage of 1–49% with 28% having 26–49% tumor
shrinkage as measured by the ratio of final tumor cross-product
area to the initial tumor cross-product area. If only the latter
patients were shifted to the partial response category the clinical
objective response would have been 80%. This may explain
why the recent SNAP01 trial of 219 patients randomized to
either ifosfamide and cisplatin or ifosfamide, cisplatin, and
paclitaxel did not measure clinical response rates but relied on
optimal pathologic response rate to show the advantage of
adding paclitaxel [23]. In our study, there was no gross clinical
disease in 17 out of 113 patients in the NACT + RHPPL group
(15%) and the complete pathologic response rate was 5%.
There have been few large series of stage IB cervical cancer
patients with tumor diameter ≥ 4 cm, the standard used for
inclusion in this study. 83% of RHPPL and 86% of NACT +
RHPPL patients in this study, 188 and 124 patients, res-
pectively, met the FIGO stage IB2 requirements. Although
Delgado et al. reported on 732 stage I squamous cancer of the
cervix patients treated with RHPPL in a prospective GOG study,
only 152 patients (21%) had tumor diameter N 3 cm [10]. No
further details were provided regarding tumor diameter N 3 cm.
Buda et al. reported on 204 locally advanced squamous cancer
of the cervix patients treated with NACT + radical surgery in the
SNAP01 trial, but only 97patients had stage IB2 disease [23].
Yessian et al. reported on a series of 600 patients treated with
RHPPL at a single institution over 14 years, but only 58 patients
(9.6%) had stage IB2 disease [24]. Thus, our trial which
included 242 patients represents one of the largest surgical trials
of stage IB2 cervical cancer. It is hard to establish definite
benchmarks for pathology findings, complications, and adju-
vant RT rates given the relatively small number of stage IB2
patients evaluated separately in other trials. It is reasonable to
assume they could be substantially different than those found in
Stage 1A1, 1A2, and 1B1 patients.
367
The post-surgical RT rate reported in this study was 45% in
NACT + RHPPL patients and 52% in RHPPL patients. If only
patients who had RHPPL performed are considered 34% of
NACT + RHPPL patients and 40% of RHPPL patients received
RT. In the SNAP01 trial 79 of 152 evaluable patients (52%)
received post-surgical RT [23]. Thirty-five out of 58 stage 1B2
patients (60%) in the Yessian et al. study received post-RHPPL
radiation therapy [24]. Despite 35 cases of unauthorized post-
surgical RT in our study, 12% of all patients, the overall post-
surgical RT rate of 48% was lower. When Stage 1B2 patients are
treated with RHPPL, it should be anticipated that at least half
will also receive radiation therapy, and if more recently
proposed guidelines for adjuvant RT are used, as recommended
by Sedlis et al. [20], the rate would be approximately 90% as
shown in Table 3. The tendency to broaden the criteria for post-
operative radiotherapy inpatients with bulky cervical cancer is
not unique to the GOG; a phase III trial of stage IIB patients
involving neoadjuvant chemotherapy conducted by Sardi and
associates between May 1988 and January 1993 at the
University of Buenos Aires, incorporated pelvic radiotherapy
in all 4 treatment arms [25].
The early closure of this study due to slow accrual was at least
partially attributable to a study by Sedlis et al. published in 1999
[20]. That randomized trial of pelvic radiation vs. no further
treatment in Stage IB cervical cancer (with free margins) after
radical hysterectomy and pelvic lymphadenectomy included the
following combinations of intermediate risk factors: any lesion
size with deep stromal invasion and capillary lymphatic space
involvement (CLS); middle 1/3 cervical stromal invasion with
CLS and tumor ≥2 cm; deep or middle 1/3 stromal invasion with
no CLS but tumor ≥4 cm; or superficial 1/3 stromal invasion with
no CLS but tumor ≥5 cm. That phase III study demonstrated a
47% reduction in risk of recurrence in patients who received
adjuvant pelvic radiotherapy compared to those in the surgery
only group. Clearly, the eligibility criteria in the Sedlis et al. study
were substantially different than those used in the current study,
where initial tumor volume was markedly greater with greater risk
of occult metastases, both pelvic and distant. Of note, only 84/277
(30%) of eligible patients in the study by Sedlis and colleagues
had cervical cancers ≥4 cm. Nonetheless, simultaneous with
publication of that study, we began to observe with greater
frequency, RT protocol violations for adverse findings in the
primary tumor, which had not been an indication for RT in this
study, and rapidly declining accrual to the present study.
A consequence of the early closure of the current study was
the reduction in statistical power to detect a difference in PFS or
OS, having met only 75% of the proposed number of
recurrences/deaths. For a 33% reduction in the rate of
progression, the statistical power is 73% rather than the planned
84%. Notwithstanding, there was striking similarity between the
treatment groups with respect to operability and adjuvant
radiotherapy, with nearly super-imposable PFS curves. A more
pertinent probability than the designed power is what some refer
to as conditional power. Conditional power is the probability of
rejecting the null hypothesis (i.e., significant difference in favor
of the NACT + RHPPL group) if the study was completed as
planned with all future progressions (experienced by new and
368 G.L. Eddy et al. / Gynecologic Oncology 106 (2007) 362–369
existing patients) governed by the alternative hypothesis on
which the study was designed (a 33% reduction in risk in the
NACT + RHPPL group), and given the data as reported. Using
computer simulation, this probability was estimated to be only
4%; that is, the data presented in this report support the null
hypothesis to such a degree that it could not be outweighed,
except very rarely, by additional data that would have been
collected if the study was completed (assuming it would support
the alternative hypothesis).
After 24 years of reported Phase II trials of NACT in cervical
cancer, results from recently completed Phase III trials have been
eagerly awaited. Unfortunately, few trials have evaluated the
benefit of adding NACT to RHPPL in bulky stage IB cervical
cancer. In 2003, a systematic review and meta-analysis of
individual patient data from 21 randomized trials using NACT for
locally advanced cervical cancer included nearly 3000 patients,
but approximately 70% of the patients were treated with RT with
or without NACT [26]. Only 5 trials involving 872 patients
studied NACT + radical surgery vs. RT. Only two of these trials
studied stage 1B2 patients. Benedetti-Panici et al. reported on 441
patients, but only 174 patients had 1B2 to IIA N4 cm disease [27].
This was the only subgroup that demonstrated an overall and
progression-free survival advantage for the 87 patients treated
with NACT + radical surgery. In contrast, Chang and coworkers
could show no benefit for 68 patients treated with NACT + radical
hysterectomy and pelvic lymphadenectomy compared to 52
patients treated with conventional RT [28].
Few studies have evaluated the effectiveness of NACT added
to radical surgery for bulky Stage I cervical cancer patients. In
two GOG trials evaluating RHPPL in cervical cancer reported
17 years apart, the progression-free rate of 63% in both arms of
this study is remarkably similar to the 67% 5-year survival
reported by Delgado and coworkers in 152 patients with stage
1B N 3 cm disease [29]. In 2006, a poster presentation by
Katsumata and others failed to demonstrate any advantage for
NACT in 134 patients with stage 1B2, IIA (N 4 cm), or IIB
disease randomly assigned to either NACT + radical hysterecto-
my or radical hysterectomy alone [30]. With 288 evaluable
patients in our study, a comparable outcome was noted with the
exception of fewer Grade 3 and 4 adverse urologic adverse
events in the NACT + RHPPL group. This may be a reflection of
improved operability allowing more precise dissection of the
ureter from the cardinal and uterosacral ligaments. This occurred
despite the greater use of post-operative RT in the NACT +
RHPPL group (52% vs. 45%) and the greater number of patients
with tumor size N 5 cm in this group (116 vs. 104). With a
difference of only 9 severe urologic adverse events in the 143
patients (6%), this has to be weighed against the expense and
delay in subsequent surgery NACT required of all 143 RHPPL
patients. Although not definitive, the results of this study have
led the GOG to recommend against adding NACT to RHPPL in
future randomized trials of stage IB2 cervical cancer patients.
Acknowledgments
This study was supported by National Cancer Institute grants to
the Gynecologic Oncology Group (GOG) Administrative Office
(CA 27469) and the GOG Statistical and Data Center (CA 37517).
The following GOG institutions participated in the related
treatment study: University of Alabama at Birmingham, Duke
University Medical Center, Abington Memorial Hospital, Walter
Reed Army Medical Center, University of Minnesota Medical
School, University of Mississippi Medical Center, Colorado
Gynecologic Oncology Group PC, University of Pennsylvania
Cancer Center, Milton S. Hershey Medical Center, University of
Cincinnati, University of North Carolina School of Medicine,
University of Iowa Hospitals and Clinics, University of Texas
Southwestern Medical Center at Dallas, Indiana University
Medical Center, Wake Forest University School of Medicine,
Albany Medical College, University of California-Irvine, Tufts-
New England Medical Center, State University of New York-
Downstate, University of Kentucky, Cleveland Clinic Foundation,
Washington University School of Medicine, Cooper Hospital/
University Medical Center, Columbus Cancer Council, University
of Massachusetts Medical School, Fox Chase Cancer Center,
Medical University of South Carolina, Women’s Cancer Center,
University of Oklahoma, Tacoma General Hospital, Thomas
Jefferson University Hospital, Mayo Clinic, Tampa Bay/H. Lee
Moffitt Cancer Center, and Gynecologic Oncology Network.
References
[1] Plentyl AA, Friedman EA. Clinical significance of cervical lymphatics.
Lymphatics system of the female genitalia. Philadelphia: Saunders; 1971.
p. 85–115.
[2] Piver MS, Chung WS. Prognostic significance of cervical lesions size and
pelvic node metastases in cervical carcinoma. Obstet Gynecol 1975;46:
507–10.
[3] Ballon SC, Berman ML, Lagasse LD, et al. Survival after extraperitoneal
pelvic and para-aortic lymphadenectomy and radiation therapy in cervical
carcinoma. Obstet Gynecol 1981;57:90–5.
[4] Gauthier P, Gore I, Shingleton HM, et al. Identification of histopathologic
risk groups in Stage IB squamous carcinoma of the cervix. Obstet Gynecol
1985;66:569–73.
[5] Eddy GL, Manetta A, Alvarez RD, et al. Neoadjuvant chemotherapy with
vincristine and cisplatin followed by radical hysterectomy and pelvic
lymphadenectomy for FIGO Stage 1B bulky cervical cancer: a Gynecologic
Oncology Group pilot study. Gynecol Oncol 1995;57:412–6.
[6] Benedetti-Panici P, Greggi S, Colombo A, et al. Neoadjuvant chemotherapy
and radical surgery versus exclusive radiotherapy in locally advanced
squamous cell cervical cancer: results from the Italian multicenter randomized
study. J Clin Oncol 2001;20:179–88.
[7] Chang TC, Lai CH, Hong JH, et al. Randomized trial of neoadjuvant
cisplatin, vincristine, bleomycin and radical hysterectomy versus radiation
therapy for bulky Stage IB and IIA cervical cancer. J Clin Oncol 2000;18:
1740–7.
[8] Keys HM, Bundy BN, Stehman FB, et al. Weekly cisplatin during
radiation therapy versus irradiation alone each followed by adjuvant
hysterectomy in bulky stage IB cervical carcinoma: a randomized trial of
the Gynecologic Oncology Group. N Engl J Med 1999;340:1154–61.
[9] American Joint Committee on Cancer. Manual for staging of cancer.
4th ed. Philadelphia: JB Lippincott Company; 1992. p. 155–7.
[10] Delgado G, Bundy BN, Fowler Jr WC, et al. A prospective surgical
pathologic study of Stage I squamous carcinoma of the cervix: a
Gynecologic Oncology Group study. Gynecol Oncol 1989;35:314–20.
[11] Piver MS, Rutledge F, Smith JP. Five classes of extended hysterectomy for
women with cervical cancer. Obstet Gynecol 1974;44:265–72.
[12] Blessing JA. Design, analysis and interpretation of chemotherapy trials in
gynecologic cancer. In: Deppe G, editor. Chemotherapy of gynecologic
cancer. 2nd ed. New York: AR Liss, Inc; 1990. p. 63–97.
 G.L. Eddy et al. / Gynecologic Oncology 106 (2007) 362–369
[13] Schoenfeld D. Sample sizes for the proportional hazards regression model.
Biometrics 1983;39:499–503.
[14] Mantel N. Evaluation of survival data and two new rank order statistics
arising in its consideration. Cancer Chemo Rep 1966;50:163–70.
[15] Lan KKG, DeMets DL. Discrete sequential boundaries for clinical trials.
Biometrika 1983;70:659–63.
[16] Kaplan EL, Meier P. Nonparametric estimation from incomplete observations.
J Am Stat Assoc 1958;53:457–81.
[17] Cox DR. Regression model and life tables (with discussion). J R Stat Soc
1972;34:187–220.
[18] Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data.
New York: John Wiley and Sons, Inc; 1980.
[19] Hollander M, Wolfe DA. Nonparametric statistical methods. Wiley series
in probability. New York: John Wiley and Sons, Inc; 1999.
[20] Sedlis A, Bundy BN, Rotman MZ, et al. A randomized trial of pelvic
radiation therapy versus no further therapy in selected patients with Stage IB
carcinoma of the cervix after radical hysterectomy and pelvic lymphade-
nectomy. a Gynecologic Oncology Group study. Gynecol Oncol 1999;73:
177–83.
[21] Rotman M, Sedlis A, Piedmonte MR, Bundy B, Lentz SS, Muderspach LI,
Zaino RJ. A phase III randomized trial of postoperative pelvic irradiation
in stage IB cervical carcinoma with poor prognostic features: follow-up of
a Gynecologic Oncology Group study. Int J Radiat Oncol Biol Phys
2006;65:169–76.
[22] Duenas-Gonzalez A, Lopez-Graniel C, Gonzalez-Enciso A, et al. A phase
II study of multimodality treatment for locally advanced cervical cancer:
neoadjuvant carboplatin and paclitaxel followed by radical hysterectomy
and adjuvant cisplatin chemoradiation. Ann Oncol 2003;14:1278–84.
[23] Buda A, Fossati R, Colombo N, et al. Randomized trial of neoadjuvant
chemotherapy comparing paclitaxel, ifosfamide, and cisplatin with
369
ifosfamide and cisplatin followed by radical surgery in patients with
locally advanced squamous cell cervical carcinoma: the SNAP01 (Studio
Neo-Adjuvante Portio) Italian Collaborative Study. J Clin Oncol 2005;23:
4137–45.
[24] Yessaian A, Magistris A, Burger RA, et al. Radical hysterectomy followed
by tailored postoperative therapy in the treatment of stage IB2 cervical
cancer: feasibility and indications for adjuvant therapy. Gynecol Oncol
2004;94:61–6.
[25] Sardi JE, Sananes CE, Gioaroli AA, et al. Neoadjuvant chemotherapy in
cervical carcinoma Stage IIB: a randomized controlled trial. Int J Gynecol
Cancer 1998;8:441–50.
[26] Neoadjuvant Chemotherapy for Cervical Cancer Meta-analysis Collaboration.
Neoadjuvant chemotherapy for locally advanced cervical cancer: a systematic
review and meta-analysis of individual patient data from 21 randomised trials.
Eur J Cancer 2003;39:2470–86.
[27] Benedetti-Panici P, Greggi S, Colombo A, et al. Neoadjuvant chemotherapy
and radical surgery versus exclusive radiotherapy in locally advanced
squamous cell cervical cancer: results from the Italian multicenter
randomized study. J Clin Oncol 2002;20:179–88.
[28] Chang T, Lai C, Hong J, et al. Randomized trial of neoadjuvant cisplatin,
vincristine, bleomycin, and radical hysterectomy versus radiation therapy
for bulky Stage IB and IIA cervical cancer. J Clin Oncol 2000;18:1740–7.
[29] Delgado G, Bundy B, Zaino R, et al. Prospective surgical–pathological study
of disease-free interval in patients with Stage IB squamous cell carcinoma of
the cervix. A Gynecologic Oncology Group study. Gynecol Oncol
1990;38:352–7.
[30] Katsumata N, Yoshikawa H, Hirakawa T, et al. Phase III randomized trial
of neoadjuvant chemotherapy (NAC) followed by radical hysterectomy
(RH) versus RH for bulky stage I/II cervical cancer. J Clin Oncol 2006
[Proc ASCO; 24:18S (abst 5013)].