In this study, we found that histological type was an independent prognostic variable in patients with LACC who were treated with concomitant chemoradiotherapy. Patients with AC had a worse prognosis than those with SCC (for DFS: HR = 1·46, 95% CI = 1·012 − 2·106; for OS HR = 1·723, 95% CI = 1·22 − 2·41). There is a considerable discrepancy in the literature regarding the prognostic value of histological types. On the one hand, some reports analyzed small patient cohorts and several variables from one or a few centers; on the other hand, some reports analyzed large data sets mainly from epidemiological records that included many patients, but they did not consider all the variables that could alter the results. Most studies on this variable in LACC conclude that histological type is an independent prognostic factor (Table 4).
Table 4
Comparison of histology impact between studies in locally advanced cervical cancer in the last decade
Authors
|
Years of patient inclusion
|
Type of study
|
Clinical stage
|
N
|
SSC/AC proportion %
|
OS
|
P-value
|
Galic et al (2012)
|
1988–2005
|
Retrospective
multicenter
|
IIB-IVA
|
10217
|
84.2/10.8/5.0†
|
32.5/17.9/29.2
|
0.014‡
|
Katanyoo et al (2012)
|
1995–2008
|
Retrospective
|
IIB-IVA
|
423
|
66.7/33.3
|
61.7/59.9
|
0.191
|
Intaraphet et al (2013)
|
1995–2011
|
Retrospective
|
I-IV
|
1978
|
82.5/17.5
|
60/54.7
|
< 0.001
|
Rose et al 2014
|
|
Retrospective 5 clinical trials
|
IB2-IVA
|
1671
|
89.1/10.9§
|
--------------
|
0.459
|
Chen et al (2014)
|
1995–2009
|
Retrospective
|
IIB-IVA
|
229
|
84.7/15.3§
|
58.1/41.3
|
0.090
|
Yun Lee et al (2015)
|
1993–2012
|
Retrospective
|
|
3156
|
85.0/15.0
|
---------------
|
0.003
|
Zhou et al (2017)
|
1988–2013
|
Retrospective
multicenter
|
I-IV
|
8751
|
86/10.6/3.4†
|
51.1/40.3
|
< 0.001
|
Yokoi et al (2017)
|
1993–2014
|
Retrospective
|
IIB-IVA
|
249
|
90.4/9.6
|
58.6/26.7
|
0.004
|
Seamon et a (2017)
|
1999–2012
|
Retrospective 3 clinical trials treated whit chemotherapy doublets
|
IVBξ
|
781
|
77/23
|
10.3/15.1
|
0.093
|
Cheng Yin et al (2018)
|
2004–2016
|
Retrospective
Single center
|
IB2-IVA
|
181
|
83.4/16.5
|
78.6/46.0
|
< 0.001
|
Hu et al (2018)
|
2011–2014
|
Retrospective
|
IB-IVA
|
815
|
91.2/8.7
|
85.2/74.4
|
0.005
|
Jonska-Gmyrek et al (2019)
|
2004–2012
|
Retrospective
|
IIB
IIIA-IIIB
|
161
|
67.7/32.3
|
81.7/62.8
73.1/33.6
|
0.03
0.005
|
Current Study
|
2005–2014
|
Retrospective
Single center
|
IB2-IVA
|
1291
|
1154/137
|
74.3/60.0
|
0.004
|
† Adenosquamous carcinoma ‡calculated only for IIIB clinical Stage §AC + adenosquamous carcinoma.
Galic et al. performed a multicenter retrospective study that included patients with stage IIB-IVA disease who were treated between 1988 and 2005. They concluded that women with locally advanced adenocarcinoma were 21% more likely to die than those with SCC (HR = 1·21, CI 95%=1·10 − 1·32).(26) Intaraphet et al., Yun Lee et al., Yokoio et al., Cheng Yin et al., Hu et al., and Jonska-Gmyrek et al. have also described statistically significant differences in the prognosis of AC and SCC.(30,33–37) Zhou et al. published one of the papers with the largest number of patients analyzed in the last decade. Using the SEER database, they assessed 8,751 patients and determined that AC had a worse prognosis than SCC.(28) However, Katanyoo et al., Rose et al., Chen et al., and Seamon et al. did not find differences.(25,31,38,39) It is worth noting Rose and Seamon compared the differences between histological types in prospectively recruited patients in controlled trials, which improved data quality. However, their main goal was not to find differences between histological types but to evaluate the safety and efficacy of a certain treatment. Rose et al. examined 1,671 patients with LACC from five different trials, whereas Seamon et al. evaluated 781 patients with recurrent or metastatic disease from three randomized trials conducted between 1999 and 2012. Neither of these studies found differences between histological types (p = 0·45 and 0·093, respectively).
Tumor grade is another relevant finding that is consistent with the literature. Our study shows that more than 23% of cases of adenocarcinoma are well differentiated, in contrast to < 1% of SCC. Since this grade has a good prognosis and is more common in adenocarcinoma, it is essential to consider it in the multivariate analysis.(20,22,26,28,31)
In addition to tumor grade, we found that age, clinical stage, LVSI, and parametrial involvement were independent prognostic factors in LACC. These findings have also been described in other series.(22,34)
We did not find differences among other variables, such as functional status, BMI, metastasis in the pelvic nodes, and tumor size, which are commonly described as risk factors. (38,40) Regarding tumor size, it is likely that in locally advanced stages when tumors invade neighboring structures, the size of the initial lesion loses its prognostic value.
The limitations of our study are its retrospective design and the limited information obtained from older CT scans (2005–2008). The strength of our study is the number of patients from a single center, which partly ensures the homogeneity of treatment and staging criteria.
Although AC and SCC are distinct entities at the histological and molecular levels, several factors could account for the literature discrepancy about the role of these as a prognostic factor. Mainly published reports have been retrospective so far. They exhibit the typical bias of such study design, especially the accuracy of the collected data, availability of information including all cofounding variables that could modify the results, such as total dose of radiation therapy or number of chemotherapy cycles. In some cases, the limitations include lack of imaging tests with the inability to detect lymph node disease, and the physician expertise to perform physical examinations and classify the disease.
CC incidence has decreased over the last 40 years. Incidence rates by stage at the time of diagnosis decreased from 2001 to 2015 for SCC, but those of AC remained stable or even increased.(18,41) Considering that the rates of AC could still increase, it is essential to determine whether the histological type is truly an independent prognostic factor requiring a special approach to improve the prognosis in patients with AC in LACC.