Keywords
Ovarian serous carcinoma - postchemotherapy changes - serous tubal intraepithelial
carcinoma
Introduction
Ovarian carcinoma represents 30% of all cancers of the female genital tract, of which
high-grade serous carcinomas (HGSCs) are predominant accounting for 70%.[1] Size of residual disease after surgery and International Federation of Gynecology
and Obstetrics (FIGO) stage at diagnosis are the two main prognostic indicators.[2] Most patients present at an advanced stage with a poor survival rate. The treatment
of such ovarian cancers is usually maximal surgical debulking followed by chemotherapy.
However, preoperative neoadjuvant chemotherapy (NACT) followed by intervention debulking
has been the recently followed trend in the management of ovarian cancer.[3] The present study emphasizes the clinicopathological findings as well as postchemotherapy
changes in ovarian serous carcinomas.
Materials and Methods
The present study is a 2-year retrospective study which includes cases from July 2015
to June 2017. The study was approved by the ethics committee of our institute. All
cases diagnosed as ovarian serous carcinoma were retrospectively reviewed. The study
group comprised 38 cases of serous ovarian carcinoma, of which 18 cases with advanced-stage
ovarian cancer were treated with preoperative chemotherapy. Clinical and gross findings
were collected, microscopic findings were reviewed, and tumor grade and stage were
assessed as per the 2014 World Health Organization Classification and the FIGO Surgical
Staging System, respectively.[2]
Clinicopathological data collected included age at diagnosis, chief complaints, any
significant past history or family history of malignancy, cytology of the peritoneal
fluid, any lymph node metastasis, FIGO stage, and preoperative chemotherapy regimen
and the methods of surgical treatment. Serum cancer antigen (CA)-125 values measured
immediately before and after chemotherapy had been noted. The chemotherapy regimen
varied between individual patients but usually included three cycles of carboplatin
and paclitaxel before debulking surgery followed by another set of three cycles. In
18 cases, diagnosis of adenocarcinoma was confirmed before chemotherapy by cytological
examination of the peritoneal fluid. Rest of the cases were diagnosed on tissue biopsies
before chemotherapy. Based on the chemotherapy response score (CRS) system criteria
proposed by the International Collaboration on Cancer Reporting, CRS was assessed
in cases which received NACT before debulking surgery.[3]
Results
A total of 156 ovarian tumors were encountered during the study period. Of these,
85 (54.5%) were malignant neoplasms. Serous carcinoma was the most common of all malignant
ovarian tumors, accounting to 38 cases (44.7%). Of these, six were low-grade serous
carcinomas (LGSC) and 32 were HGSCs.
The age at presentation ranged between 24 and 70 years for LGSC and 40 and 71 years
for HGSC. Among HGSC cases, two cases had past history of esophageal squamous cell
carcinoma (SCC) and one case had a history of SCC in the lung. Bilateral involvement
was seen in 16 (42%) cases.
Surface involvement with rupture of ovarian capsule was seen in seven cases. In 15
(39.47%) cases, omental deposits were seen.
All six cases of LGSC were not treated with preoperative NACT. Histopathology of LGSC
revealed a variety of architectural patterns, including single cells and irregularly
shaped small nests of cells haphazardly infiltrating stroma, extensive papillae formation
with variable fibrovascular stroma. The cells showed minimal nuclear atypia, and few
showed a single prominent nucleolus. Frequent psammoma bodies and very occasional
mitotic figures were noted [Figure 1]a],[Figure 1]b],[Figure 1]c]. Two cases showed a component of serous borderline tumor/atypical proliferative
serous tumor (SBT/APST). Necrosis was not identified. Ki-67 proliferation index was
significantly lower in LGSC when compared to HGSC.
Figure 1: (a-c) Low-grade serous carcinomas; (d-f) high-grade serous carcinomas; (c)
surface involvement; (f) omental metastasis (H and E, ×100)
Among 32 cases of HGSC, 14 cases were resected without prior NACT. Histopathology
revealed branching papillary fronds and solid masses of the cells with slit-like fenestrations
and stratifications. Focal areas showed cribriform and glandular patterns. Cells showed
marked pleomorphism with prominent nucleoli and frequent mitoses. Large areas of necrosis
and variable number of psammoma bodies were identified [Figure 1]d],[Figure 1]e],[Figure 1]f].
Immunohistochemistry (IHC) with p53 (6/6) showed intense positivity in the lesional
cells of HGSC. Other markers that were significantly expressed in HGSC were cytokeratin
7 (4/4), CA-125 (1/1), Wilms tumor 1 (3/4), and epithelial membrane antigen (1/1).
Ki-67 proliferation index was significantly higher (67%) in HGSC when compared to
LGSC [Figure 2].
Figure 2: High-grade serous carcinomas: Positive staining by (a) calretinin; (b) epithelial
membrane antigen; (c) cytokeratin 7; (d) p53; (e) cancer antigen-125; (f) Ki67 – 67%
(IHC, ×100)
As depicted in [Table 1], two cases of LGSC and 13 cases of HGSC were in FIGO stage III at the time of presentation.
Among HGSC cases in stage III, one case involved the appendix, four cases involved
the colon with mesenteric deposits, and one case showed metastatic deposits in the
retroperitoneal lymph node.
Table 1
International Federation of Gynecology and Obstetrics stage of low-grade and high-grade
serous carcinomas
|
FIGO stage
|
LGSC, n (%)
|
HGSC, n (%)
|
|
FIGO – International Federation of Gynecology and Obstetrics; LGSC – Low-grade serous
carcinoma; HGSC – High-grade serous carcinoma
|
|
I
|
3 (50)
|
15 (46.9)
|
|
II
|
1 (16.7)
|
4 (12.5)
|
|
III
|
2 (33.3)
|
13 (40.6)
|
|
Total
|
6 (16)
|
32 (84)
|
Preoperative NACT was given in 18 cases of HGSC. Serous tubal intraepithelial carcinoma
(STIC) was identified in six (18.75%) cases of HGSC with distinct morphological features
such as epithelial stratification, moderate to marked nuclear pleomorphism, prominent
nucleoli, increased nuclear/cytoplasmic ratio, loss of polarity, and frequent mitotic
figures. All cases were ipsilateral. Abnormal p53 expression with high Ki-67 proliferation
index (60%) was seen in STIC lesions [Figure 3]. The incidence of STIC in cases not treated with prior NACT was 29% (four cases)
compared to 11% (two cases) in treated cases.
Figure 3: Serous tubal intraepithelial carcinoma: (a) (H and E, ×400); (b) p53 expression
(IHC, ×100); (c) Ki67 – 60% (IHC, ×100)
Of the 18 cases who received prior NACT, nearly all cases showed more than 50% reduction
in CA-125 levels, of which 33% (6 cases) showed below normal CA-125 levels postchemotherapy
with CRS-2 and CRS-3. [Table 2] shows the serum CA-125 values immediately before and after chemotherapy, the percent
drop in CA-125 concentrations, and the CRS suggesting a response to chemotherapy.
Table 2
Serum cancer antigen-125 values immediately before and after chemotherapy
|
Age (years)
|
Serum CA-125 (U/ml) prior chemotherapy
|
Serum CA-125 (U/ml) after chemotherapy
|
Percentage decrease in serum CA-125 values
|
CRS score
|
|
CA-125 – Cancer antigen-125; CRS – Chemotherapy response score
|
|
59
|
3200
|
85
|
97
|
3
|
|
45
|
237
|
12
|
95
|
3
|
|
40
|
107
|
25
|
76
|
2
|
|
58
|
482
|
18
|
96
|
3
|
|
52
|
600
|
124
|
79
|
1
|
|
70
|
4701
|
71
|
98
|
2
|
|
48
|
821
|
92
|
89
|
1
|
|
45
|
88
|
15
|
83
|
2
|
|
45
|
674
|
230
|
66
|
1
|
|
65
|
524
|
192
|
63
|
1
|
|
47
|
653
|
45
|
93
|
2
|
|
55
|
245
|
115
|
54
|
1
|
|
42
|
487
|
76
|
83
|
1
|
|
65
|
354
|
47
|
87
|
3
|
|
38
|
950
|
28
|
97
|
2
|
|
70
|
198
|
13
|
93
|
3
|
|
60
|
1827
|
83
|
94
|
2
|
|
60
|
2147
|
96
|
95
|
2
|
CRS was assessed in all the 18 cases who received NACT. Viable tumor with no or minimal
response (CRS-1) was seen in six cases, appreciable tumor response amid viable tumor
that is readily identifiable (CRS-2) was seen in seven cases, and complete or near-complete
response with no residual tumor or minimal irregularly scattered tumor foci (CRS-3)
was seen in five cases. In the current study, extensive fibrosis, necrosis with widespread
infiltration by chronic inflammatory cells in the tumor, sheets of foamy macrophages,
psammoma bodies, hemosiderin-laden macrophages, and marked anisonucleosis with the
presence of many bizarre tumor cell nuclei were significantly seen in cases with CRS-2
and CRS-3 when compared to CRS-1 [Figure 4].
Figure 4: Chemotherapy-induced changes: (a) Bizarre tumor nuclei; (b) extensive fibrosis;
(c) tumor necrosis; (d) hemosiderin-laden macrophages; (e) psammoma bodies; (f) foamy
macrophages (H and E, ×100)
Discussion
Ovarian cancer is one of the leading causes of morbidity and mortality in females,
with a 5-year survival rate of 20%–40% in western nations.[1] Of these, 15%–20% are known to have germline mutations of BRCA1 or BRCA2. The lifetime
risk of developing ovarian cancer in women with germline mutations of BRCA is 16%–59%
and in women without germline mutations is only 1.4%.[4],[5] Serous carcinoma is the most common ovarian neoplasm which is now considered to
represent two separate diseases: LGSC and HGSC. This is based on the recent classification
of ovarian cancer into two broad categories, designated type I and II based on molecular
genetics and morphologic characteristics.[6]
HGSC is the most common subtype of epithelial ovarian cancers and the most aggressive
malignancy presenting at advanced stages (stage III or IV) at diagnosis. LGSC accounts
for only about 5% of all serous carcinomas and usually present one decade earlier
than HGSC.[7],[8] In our study, HGSC accounted for 84% (32 cases) and the youngest age of presentation
for LGSC was 24 years whereas it was 40 years for HGSC.
Factors associated with shorter lifetime number of menstrual cycles such as later
age at menarche, earlier age at menopause and oral contraceptive use,[9] greater numbers of births, tubal ligation,[10] especially among BRCA1/2 mutation carriers, have a protective effect, whereas infertility
is associated with a significant risk of ovarian carcinoma.[11],[12],[13]
Ovarian masses at the initial stages are usually asymptomatic and detected incidentally.
Presenting symptoms are relatively nonspecific which could delay the diagnosis, and
these include abdominal distension, bloating, pain, constipation, urinary frequency,
nausea, and anorexia.[14],[15] Cough and dyspnea are the common symptoms in cases with malignant pleural effusion.
Abdominal distension with pain, urinary frequency, and nausea were the predominant
presenting complaints in our study.
On imaging, ovarian masses are usually large, complex, cystic pelvic masses with areas
of septal thickening or may have solid nodular areas with increased vascularity. In
the present study, most of the neoplasms had a complex solid cystic appearance on
radiology, with two cases showing adherence and infiltration to rectum and sigmoid
colon, respectively. Ascites and omental/peritoneal nodules are identified in 15 (39.47%)
cases, indicating advanced stage disease with elevated serum biomarkers such as CA-125.[16]
LGSC exemplifies the classically held view of a stepwise progression, i.e., adenomacarcinoma
sequence, and in our study, two cases showed a component of SBT/APST.[6] As per the recent concepts, majority of HGSC originates from the fimbriated end
of the fallopian tube secretory epithelial cells and a small part from cortical inclusion
cyst.[17],[18] STIC is a high-grade preneoplastic noninvasive lesion, which was first reported
by Piek et al.,[19] in the fallopian tubes of women with a BRCA1 or BRCA2 mutation, who underwent risk-reducing
salpingo-oophorectomies.[20],[21],[22],[23] Review of literature shows that in asymptomatic BRCA mutation carriers, the incidence
of STIC is 0.6%–10%.[5],[24],[25],[26],[27] The frequency of STIC lesions increases with age and is lower with oral contraceptive
use. STIC is the earliest known manifestation of most pelvic serous cancers, with
an incidence of 11%–68% in HGSC.[28],[29],[30],[31],[32],[33] With thorough examination of fallopian tubes by Sectioning and Extensively Examining
the Fimbriated End of the Fallopian tube protocol, the incidence of STIC was 35% in
Koc et al.'s study.[34] In our study, STIC was seen in 6 (18.75%) cases of HGSC, of which cases not treated
with prior NACT showed significantly higher incidence (four cases) compared to treated
cases (two cases). The reduced incidence of STIC in treated cases could be due to
its response followed by regression to NACT.
The treatment of advanced-stage ovarian cancer is usually maximal surgical debulking
followed by chemotherapy. However, recently, there has been a trend toward preoperative
chemotherapy followed by intervention debulking. In our study, 18 (47.4%) cases received
prior NACT. As per Hynninen et al., visualization of residual disease following NACT
is difficult and less reliable. Therefore, prognostic effect of complete resection
is less influential after NACT than after primary surgery.[35]
CA-125 levels pre- and post-NACT were assessed in our study. CA-125 levels together
with cross-sectional imaging are routinely used to assess response to NACT to determine
the suitability for interval debulking surgery (IDS).[36] IDS is not suitable for those who show no CA-125 response to NACT and usually have
a poor prognosis.[37] A study by Pelissier et al. highlights the role of CA-125 as a useful marker in
prediction of disease progression post-NACT.[38]
[Table 2] depicts that, in our study, nearly all cases showed more than 50% reduction in CA-125
levels, of which 33% showed below normal levels postchemotherapy with a CRS-2 and
CRS-3. A study by Rustin et al. demonstrated that CA-125 response does not discriminate
between the different grades of pathologic response.[39] There was a correlation between the biochemical response to chemotherapy and the
morphological response in our study, emphasizing its role as a useful marker for detecting
disease progression. In a study by McCluggage et al., there was no obvious correlation
between CA-125 levels postchemotherapy and the morphological response.[40]
CRS score is assessed in the omentum as the assessment of adnexal disease after chemotherapy
was difficult and less reproducible to score and showed no significant correlation
with the outcome. Omentum is proved to be the more prognostically relevant disease
site for chemotherapy response assessment as it responds least to chemotherapy.[41] CRS system is easy to apply, reproducible, and prognostically relevant. The grading
system recommends the selection of a single block of involved omental tissue that
shows the least response to chemotherapy for CRS scoring, and the amount of viable
tumor should be assessed on a single hematoxylin and eosin-stained section.[3] Patients with CRS-1 or -2 have a high probability of platinum-refractory disease
compared to those with CRS-3. Addition of novel agents such as bevacizumab and/or
combinations of agents can improve outcomes in such patients with long-term disease
control.[42],[43] Whether CRS can be applied to different treatment regimens is not yet determined,
but it stands as a potential step toward individualized treatment modification in
patients with HGSC postchemotherapy. The present study highlights the importance of
CRS system and its incorporation into routine reporting for its prognostic significance.
Morphological features of carcinoma ovary following NACT differ markedly from those
of the native neoplasm. Boehm et al. developed a new simple three-tier CRS scoring
system for assessing histopathological response to NACT.[3] It has been used in ovarian cancer reporting by the International Collaboration
on Cancer Reporting as reported by McCluggage et al.[40] Assessment of CRS has to be done in the omentum as it responds least to chemotherapy.[41] Extensive fibrosis, necrosis with widespread infiltration by chronic inflammatory
cells in the tumor, sheets of foamy macrophages, psammoma bodies, hemosiderin-laden
macrophages, and marked anisonucleosis with the presence of many bizarre tumor cell
nuclei were significantly seen in cases with CRS-2 and CRS-3 when compared to CRS-1.
Conclusion
HGSC is the most common and the most aggressive ovarian carcinoma that usually presents
at an advanced stage. There is correlation between the biochemical and morphological
response to chemotherapy. Pathologist should be aware of chemotherapy-induced morphological
features, and proper sampling of the entire tumor area followed by assessment of chemotherapy
response is highly recommended due to its prognostic significance.
