Professional Documents
Culture Documents
com
Review
Efficacy of neoadjuvant chemotherapy in patients with FIGO stage IB1 to IIA
cervical cancer: An international collaborative meta-analysis
H.S. Kim a, J.E. Sardi b, N. Katsumata c, H.S. Ryu d, J.H. Nam e, H.H. Chung a, N.H. Park a,
Y.S. Song a,f,g, N. Behtash h, T. Kamura i, H.B. Cai j, J.W. Kim a,*
a
Department of Obstetrics and Gynecology, Seoul National University College of Medicine, 28 Yeongeon-Dong, Jongno-Gu,
Seoul 110-744, Republic of Korea
b
Gynecologic Oncology Unit, Buenos Aires University, Buenos Aires, Argentina
c
Department of Medical Oncology, National Cancer Center Hospital, Tokyo, Japan
d
Department of Obstetrics and Gynecology, Ajou University School of Medicine, Suwon, Republic of Korea
e
Department of Obstetrics and Gynecology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Republic of Korea
f
Cancer Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea
g
Major in Biomodulation, World Class University, Seoul National University, Seoul, Republic of Korea
h
Department of Gynecologic Oncology, Vali-e-asr Hospital, Tehran University of Medical Sciences, Tehran, Iran
i
Department of Obstetrics and Gynecology, Kurume University School of Medicine, Kurume, Japan
j
Department of Gynecologic Oncology, Zhongnan Hospital, Wuhan University, Wuhan, China
Abstract
Background: The efficacy of neoadjuvant chemotherapy before surgery (NCS) has not been well-established in FIGO stage IB1 to IIA cer-
vical cancer when compared with primary surgical treatment (PST). Thus, we performed a meta-analysis to determine the efficacy of NCS
in patients with FIGO stage IB1 to IIA cervical cancer when compared with PST.
Methods: We searched Pubmed, Embase and the Cochrane Library between January 1987 and September 2010. Since there was a relative
lack of relevant randomized controlled trials (RCTs), we included 5 RCTs and 4 observational studies involving 1784 patients among 523
potentially relevant studies.
Results: NCS was related with lower rates of large tumor size (4 cm) (ORs, 0.22 and 0.10; 95% CI, 0.13e0.39 and 0.02e0.37) and lymph
node metastasis (ORs, 0.61 and 0.38; 95% CI, 0.37e0.99 and 0.20e0.73) than PST in all studies and RCTs. Furthermore, NCS reduced the
need of adjuvant radiotherapy (RT) in all studies (OR, 0.57; 95% CI, 0.33e0.98), and distant metastasis in all studies and RCTs (ORs, 0.61
and 0.61; 95% CI, 0.42e0.89 and 0.38e0.97). However, overall and loco-regional recurrences and progression-free survival were not dif-
ferent between the 2 treatments. On the other hand, NCS was associated with poorer overall survival in observational studies when com-
pared with PST (HR, 1.68; 95% CI, 1.12e2.53).
Conclusions: Although NCS reduced the need of adjuvant RT by decreasing tumor size and lymph node metastasis, and distant metastasis,
it failed to improve survival when compared with PST in patients with FIGO stage IB1 to IIA cervical cancer.
Ó 2012 Elsevier Ltd. All rights reserved.
0748-7983/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ejso.2012.09.003
116 H.S. Kim et al. / EJSO 39 (2013) 115e124
Potentially relevant
Identification citations/abstracts
indentified and screened Citations excluded (n=269)
for retrieval (n=523) Review articles (n=148)
Preclinical studies (n=48)
Other cancers (n=41)
Case reports (n=32)
Studies retrieved for more
Screening detailed information (n=254) Articles excluded (n=131)
Prognostic factors (n=37)
Drug-related (n=37)
Diagnostic tools (n=22)
Potentially appropriate Surgical skills (n=22)
Eligibility studies to be included in
Radiation therapy (n=13)
the meta-analysis (n=123)
Figure 1. Preferred reporting items for systematic reviews and meta-analyses flow diagram. PST, primary surgical treatment; NCS, neoadjuvant chemotherapy
before surgery.
preservation, whereas primary concurrent chemoradiation Although the efficacy of NCS should be confirmed by
(CCR) is effective to treat locally advanced cervical cancer more RCTs, it will take a long time to obtain relevant results.
(IIB to IVA).1,2 Although primary radiotherapy (RT) or Thus, we designed an international collaborative meta-
CCR is considered, primary surgical treatment (PST) with analysis using most of the relevant studies to better understand
or without adjuvant RT or CCR is also an alternative the impact of NCS on clinical outcomes, including intermedi-
method to treat stage IA2 to IIA disease.3,4 ate- or high-risk factors, the need for adjuvant RT, pattern of
The advantage of PST is that patients can avoid RT-related disease recurrence, and survival in patients with FIGO stage
complications including menopause, and preserve fertility, if IB1 to IIA cervical cancer when compared with PST.
desired. However, adjuvant RT or CCR should also be consid-
ered according to the final pathologic reports, including inter- Methods
mediate- or high-risk factors after PST for improving clinical
outcomes in patients with stage IB1 to IIA cervical cancer.1 To Search strategy
overcome the limitations of PST, the efficacy of neoadjuvant
chemotherapy before surgery (NCS) has been investigated For this meta-analysis, we searched PubMed, Scopus,
in patients with cervical cancer over the past two decades.5,6 Embase, and Cochrane Central Register of Controlled Tri-
The potential advantage of NCS has been suggested to de- als (CENTRAL) in the Cochrane Library between January
crease tumor volume, and to increase the possibility of obtain- 1987 and September 2010. For this search, we used com-
ing a wider uninvolved resection margin, thereby avoiding mon keywords for this meta-analysis as follows: cervical
adjuvant RT, whereas the delay in surgery, prolonged duration cancer; cervical carcinoma; and neoadjuvant chemotherapy.
of treatment, and accelerated repopulation of resistant cancer We also reviewed the bibliographies of the relevant articles
cells should be considered as disadvantages of NCS. to locate additional publications.
Nevertheless, the efficacy of NCS has not been deter-
mined when compared with PST in previous randomized Selection criteria
controlled trials (RCTs),7e11 and observational stud-
ies.5,12e14 In contrast, two relevant meta-analyses have We included previous relevant studies which met all of
shown the efficacy of neoadjuvant chemotherapy in cervi- the following criteria: cervical cancer; stage IB1 to IIA dis-
cal cancer.15,16 However, there were some limitations in ease by 1994 FIGO criteria17; and a comparison of clinical
the meta-analyses, including a lack of the evaluation of sur- outcomes, including intermediate- or high-risk factors, the
gical efficacy due to RT as the main treatment, heterogene- need of adjuvant RT, pattern of disease recurrence, and sur-
ity among relevant studies by including patients with vival between PST and NCS. The exclusion criteria were
locally advanced cervical cancer, no investigation for the FIGO stage IA or locally advanced cervical cancer (stage
efficacy of NCS affecting intermediate- or high-risk factors, IIB to IVA), no comparison of the clinical outcomes be-
and the need for adjuvant RT when compared with PST. tween PST and NCS, and due to a lack of accessibility
Table 1
Demographic characteristics of 9 included studies.
Author Design Duration Follow-up (months) FIGO stage Histology No. of patients Regimen of NCS Cycles of NCS Indication of radiotherapy
NCS PST
Sardi7 RCT 1987e1993 67 IB SCC 102 103 $V: 1 mg/m2 (D1) Every 10 days $All cases
$B: 25 mg/m2 for 3 cycles
IB1: 41 IB1: 74 (D1eD3)
IB2: 61 IB2: 56 $P: 50 mg/m2 (D1)
Napolitano8 RCT 1986e1995 49 IB-IIA SCC 70 56 $V: 1 mg/m2 (D1) Every 21 days $Parametrial invasion
$B: 25 mg/m2 for 3 cycles $Lymph node metastasis
IB1: 24 IB1: 22 (D1eD3)
IB2: 36 IB2: 29 $P: 50 mg/m2 (D1)
IIA: 10 IIA: 5
Cai9 RCT 1999e2002 62 IB SCC, AC 52 54 $F: 24 mg/kg Every 21 days $Deep stromal invasion
(D1eD5) for 2 cycles $Parametrial invasion
IB1: 14 IB1: 24 $P: 75 mg/m2 (D1) $Lymph node metastasis
IB2: 38 IB2: 30
Eddy10 $V: 1 mg/ $Parametrial invasion
117
(continued on next page)
118 H.S. Kim et al. / EJSO 39 (2013) 115e124
AC, adenocarcinoma; ASC, adenosquamous carcinoma; B, bleomycin; C, carboplatin; F, 5-fluorouracil; FIGO, International Federation of Gynecology and Obstetrics; NCS, neoadjuvant chemotherapy before
and difficulty with critical reading, non-English literature.
-Lymphovascular invasion
-Parametrial invasion
Indication of radiotherapy
Studies identified
surgery; O, observational; P, cisplatin; PST, primary surgical treatment; RCT, randomized controlled trial; SCC, squamous cell carcinoma; T, paclitaxel; V, vincristine.
for 2 or 3 cycles
Cycles of NCS
Every 21 days
or
IIA: 28
IB2: 9
Statistical analysis
NCS
61
SCC, AC
tors and the need for adjuvant RT, and for improving
disease recurrence and survival when compared with PST.
This meta-analysis was performed in line with the recom-
IB1, IB2, IIA
Figure 2. Comparison of intermediate-risk factors between primary surgical treatment (PST) and neoadjuvant chemotherapy before surgery (NCS) in patients
with FIGO stage IB1 to IIA cervical cancer; (A) large tumor size (4 cm): (B) lymphovascular invasion: (C) deep stromal invasion (1/2).
to represent substantial heterogeneity, and we used the ran- 997 patients were included in 5 RCTs,7e11 and 4 observational
dom effects model using the DerSimonian and Laird method. studies, respectively.5,12e14 Clinical characteristics of the 9 in-
However, the fixed effect model using the ManteleHaenszel cluded studies are depicted in Table 1. For surgical treatment,
method was used in this meta-analysis when the I2 was 50% type II or III hysterectomy with pelvic lymphadenectomy
because it indicated no heterogeneity. was performed in all studies, whereas para-aortic lymphade-
For identifying publication bias, funnel plots were repre- nectomy was performed if indicated except 2 studies.9,10 As
sented, which were scatter plots of HRs or ORs of individ- chemotherapeutic regimens, vincristine/bleomycin/mitomy-
uals studies on the X axis against the SE of the log HR or cin/cisplatin,11 vincristine/bleomycin/cisplatin,7,8 vincristine/
log OR of each study on the Y axis. HRs or ORs of small cisplatin,10,12 5-fluorouracil/cisplatin,5,9 5-fluorouracil/carbo-
scale studies scattered widely at the bottom of the graph platin,5 paclitaxel/cisplatin,14 and paclitaxel/carboplatin were
with the spread narrowing among large-scale studies, which administered.5,14
resembled symmetric inverter funnels, suggesting that there Furthermore, the indications for adjuvant RT after sur-
was no publication bias in this meta-analysis. This meta- gery were different among all studies, which included lym-
analysis was performed using Review Manager Version phovascular invasion,12 deep stromal invasion,9,11,12
5.0 (the Nordic Cochrane Centre, Copenhagen, Denmark), parametrial invasion,5,8e14 lymph node metastasis,5,8e14
and a p <0.05 was considered statistically significant. and positive resection margins.5,12e14 Patients with 2
intermediate-risk factors (large tumor size 4 cm, lympho-
Results vascular invasion, and deep stromal invasion 1/2) or 1
high-risk factor (positive resection margins, parametrial in-
Characteristics of relevant studies vasion, and lymph node metastasis) received adjuvant RT
after surgery in 3 observational studies,5,13,14 and all pa-
After the final screening of 9 relevant studies, 1784 patients tients in 1 RCT received adjuvant RT regardless of interme-
were enrolled for this meta-analysis. Among them, 787 and diate- or high-risk factors.7 In all of the studies, 2 or 3
120 H.S. Kim et al. / EJSO 39 (2013) 115e124
Figure 3. Comparison of high-risk factors between primary surgical treatment (PST) and neoadjuvant chemotherapy before surgery (NCS) in patients with
FIGO stage IB1 to IIA cervical cancer; (A) positive resection margin: (B) parametrial invasion: (C) lymph node metastasis.
cycles of chemotherapy were administered every 10 or 21 when compared with PST in all studies, whereas NCS
days in patients who received NCS. showed only a reduction of lymph node metastasis in
RCTs [82/324 (25.3%) vs. 137/333 (41.1%); OR, 0.38;
95% CI, 0.20e0.73] without a decrease in the observation
Intermediate- or high-risk factors
studies (Fig. 3).
For the reasonable concordance of intermediate- or high-
risk factors among the 9 included studies, we defined inter- Disease recurrence and adjuvant radiotherapy
mediate- or high-risk factors as in 3 previous studies.5,13,14
As a result, NCS was associated with lower rates of large There were no differences in overall and loco-regional
tumor size (4 cm) [total, 134/450 (29.8%) vs. 407/673 recurrences between the two treatments in all studies,
(60.5%); OR, 0.22; 95% CI, 0.13e0.39], lymphovascular RCTs, and observational studies. However, NCS was asso-
invasion [185/576 (32.5%) vs. 354/779 (45.4%); OR, ciated with a lower rate of distant metastasis in all studies
0.49; 95% CI, 0.29e0.84], and deep stromal invasion [49/630 (7.8%) vs. 86/793 (10.8%); OR, 0.61; 95% CI,
(1/2) [237/558 (42.5%) vs. 506/790 (64.1%); OR, 0.45; 0.42e0.89] and RCTs [33/321 (10.3%) vs. 51/323
95% CI, 0.23e0.89] than PST in all studies. These finding (15.8%); OR, 0.61; 95% CI, 0.38e0.97], whereas there
were similar in the observational studies, whereas NCS was was no difference in distant metastasis in the observational
only related to a lower rate of large tumor size (4 cm) studies (Fig. 4). Furthermore, NCS decreased the need for
compared to PST in RCTs [23/132 (17.4%) vs. 119/190 adjuvant RT when compared with PST in all studies [214/
(62.6%); OR, 0.10; 95% CI, 0.02e0.37] (Fig. 2). Further- 620 (34.5%) vs. 483/912 (53%); OR, 0.57; 95% CI,
more, NCS decreased parametrial invasion [79/630 0.33e0.98]. Nevertheless, when we performed sub-
(12.5%) vs. 145/858 (16.9%); OR, 0.63; 95% CI, analyses according to the design of the study, there was
0.43e0.86] and lymph node metastasis [158/570 (27.7%) no difference in the need for adjuvant RT in RCTs and ob-
vs. 288/825 (34.9%); OR, 0.61; 95% CI, 0.37e0.99] servational studies (Fig. 5A).
H.S. Kim et al. / EJSO 39 (2013) 115e124 121
Figure 4. Comparison of disease recurrence between primary surgical treatment (PST) and neoadjuvant chemotherapy before surgery (NCS) in patients with
FIGO stage IB1 to IIA cervical cancer; (A) overall recurrence: (B) loco-regional recurrence: (C) distant metastasis.
Figure 5. Comparison of the need of adjuvant radiotherapy and survival between primary surgical treatment (PST) and neoadjuvant chemotherapy before
surgery (NCS) in patients with FIGO stage IB1 to IIA cervical cancer; (A) the need of adjuvant radiotherapy: (B) progression-free survival: (C) overall
survival.
suggesting that NCS might help in avoiding complications a previous result where the efficacy of NCS was observed
by adjuvant RT.5,11,13,14 less frequently in early-stage disease.5
Furthermore, the paradoxical finding that NCS failed to im-
Survival prove survival in spite of the decrease of intermediate- or high-
risk factors when compared with PST can be explained by
NCS did not improve PFS and OS when compared with “Concealing effect”.24 It means that neoadjuvant chemotherapy
PST in patients with FIGO stage IB1 to IIA cervical cancer. could make a small metastasis hard to detect at histological ex-
In particular, these findings were not consistent with the amination by making a tumor smaller. Since tumor detection on
previous meta-analysis where PFS was significantly im- histological examination depends on the thickness of tissue sec-
proved with NCS (HR ¼ 0.76; 95% CI, 0.62e0.94).16 tion, small tumors could be undetected if they were located be-
However, only early-stage disease (FIGO stage IB1 to tween tissue sections. In particular, this effect could be observed
IIA) was included in this meta-analysis, while locally ad- more definitely in FIGO stage IIA than IB disease.5
vanced disease (FIGO stage IIB to IIIB) was also included
in the previous study.16 Because tumor vascularity and per- Adjuvant treatment
fusion have been demonstrated to increase in advanced cer-
vical cancer,22,23 chemotherapeutic drugs are expected to This meta-analysis demonstrated that NCS may have the
have better delivery in advanced-stage disease, which re- possibility to reduce the need of adjuvant RT by decreasing
sults in greater efficacy of NCS. Conversely, it would be risk factors, in spite of no difference in survival between
less effective to treat FIGO stage IB1 disease which shows the two treatments. However, NCS showed the possibility
lymphovascular tumor emboli or deep stromal invasion less of poor OS when compared with PST in observational studies
frequently. This hypothesis can be also supported by (HR, 1.68; 95% CI, 1.12e2.53), suggesting that post-
H.S. Kim et al. / EJSO 39 (2013) 115e124 123
Korea: results of a multicenter retrospective Korean study (KGOG- 20. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsis-
1005). Int J Gynecol Cancer 2007;17:132–6. tency in meta-analyses. BMJ 2003;327:557–60.
14. Cho YH, Kim DY, Kim JH, Kim YM, Kim YT, Nam JH. Comparative 21. Guirguis-Blake J, Calonge N, Miller T, Siu A, Teutsch S,
study of neoadjuvant chemotherapy before radical hysterectomy and Whitlock EU.S. Preventive Services Task Force. Current processes
radical surgery alone in stage IB2eIIA bulky cervical cancer. of the U.S. Preventive Services Task Force: refining evidence-based
J Gynecol Oncol 2009;20:22–7. recommendation development. Ann Intern Med 2007;147:117–22.
15. Neoadjuvant Chemotherapy for Locally Advanced Cervical Cancer Meta- 22. Alcazar JL, Jurado M, Lopez-Garcıa G. Tumor vascularization in cer-
analysis Collaboration. Neoadjuvant chemotherapy for locally advanced vical cancer by 3-dimensional power Doppler angiography: correla-
cervical cancer: a systematic review and meta-analysis of individual pa- tion with tumor characteristics. Int J Gynecol Cancer 2010;20:393–7.
tient data from 21 randomised trials. Eur J Cancer 2003;39:2470–86. 23. Testa AC, Ferrandina G, Distefano M, et al. Color Doppler velocime-
16. Rydzewska L, Tierney J, Vale CL, Symonds PR. Neoadjuvant chemo- try and three-dimensional color power angiography of cervical carci-
therapy plus surgery versus surgery for cervical cancer. Cochrane Da- noma. Ultrasound Obstet Gynecol 2004;24:445–52.
tabase Syst Rev 2010;1. CD007406. 24. Kim K, Kim MJ, Chung HH, et al. Inadvertent potential risk of neo-
17. Quinn MA, Benedet JL, Odicino F, et al. Carcinoma of the cervix adjuvant chemotherapy in cervical cancer. Med Hypotheses 2009;73:
uteri. FIGO 26th annual report on the results of treatment in gyneco- 1005–7.
logical cancer. Int J Gynaecol Obstet 2006;95(Suppl. 1):S43–S103. 25. Matsumura M, Takeshima N, Ota T, et al. Neoadjuvant chemotherapy
18. Liberati A, Altman DG, Tetziadd J, et al. The PRISMA statement for followed by radical hysterectomy plus postoperative chemotherapy but
reporting systematic reviews and meta-analyses of studies that evalu- no radiotherapy for stage IB2eIIB cervical cancer-irinotecan and plat-
ate health care interventions: explanation and elaboration. PLos Med inum chemotherapy. Gynecol Oncol 2010;119:212–6.
2009;6:e1000100. 26. Kim HS, Kim JH, Chung HH, et al. Significance of numbers of met-
19. Kim HS, Ju W, Jee BC, et al. Systematic lymphadenetomy for survival astatic and removed lymph nodes in FIGO stage IB1 to IIA cervical
in epithelial ovarian cancer: a meta-analysis. Int J Gynecol Cancer cancer: primary surgical treatment versus neoadjuvant chemotherapy
2010;20:520–8. before surgery. Gynecol Oncol 2011;121:551–7.