Introduction
Ovarian carcinoma (OC) ranks as the eighth leading
cause of both cancer incidence and mortality among women globally
(1). For epithelial OC, the
standard primary consists of a combination of maximal debulking
surgery and platinum-based chemotherapy (2). Both the European Society for Medical
Oncology (ESMO) and National Comprehensive Cancer Network (NCCN)
guidelines report that over 70% of patients with International
Federation of Gynecology and Obstetrics (FIGO) stages III-IV
disease experience recurrence within 3 years of initial treatment
(3,4). Therefore, these guidelines recommend
efficient and accurate follow-up observations after treatment
(3,4).
The ESMO guidelines recommend cancer antigen 125
(CA125) level monitoring and physical examinations as primary
surveillance methods, with computed tomography (CT) performed when
CA125 levels rise or when clinical symptoms indicative of relapse
occur (3). In comparison, the NCCN
guidelines suggest physical examinations and CA125 assessments
every 2-4 months for the first 2 years, every 3-6 months for the
following 3 years, and annually for up to 5 years (4). Both guidelines recommend CT imaging
when CA125 levels increased or when new symptoms suggestive of
recurrence emerge. These standardized approaches underscore the
clinical importance of CA125 monitoring for detecting OC
recurrence.
The tumor marker CA125, known as Mucin16, is widely
used for monitoring initial presentation and recurrent disease.
While approximately 85% of patients with OC demonstrate elevated
CA125 levels at initial diagnosis (3), significant variation exists across
histological subtypes (5).
Notably, high-grade serous carcinoma (HGSC) typically presents with
markedly elevated CA125 levels at initial diagnosis, whereas clear
cell carcinoma (CCC) often shows comparatively lower levels
(5,6). Consequently, although current ESMO
and NCCN guidelines recommend CA125 for OC surveillance, its
predictive value for recurrence detection potentially varies
substantially among histological subtypes.
This study examined the utility of CA125 for
recurrence detection across different histological subtypes of
epithelial OC.
Patients and methods
Patient selection and data
collection
This retrospective study analyzed patients with OC
treated at the National Defense Medical College Hospital
(Tokorozawa, Japan) between January 2006 and December 2022.
Inclusion criteria required that patients underwent complete
primary surgery and adjuvant chemotherapy, achieved either complete
or partial response after initial treatment, and had CA125 levels
<35 U/ml following initial treatment. Recurrence was defined as
disease progression based on the Response Evaluation Criteria in
Solid Tumors version 1.1 criteria using CT imaging (7). Patients diagnosed with recurrence
through non-CT were excluded.
Surveillance protocol and recurrence
definition
The follow-up schedule consisted of CA125 tests
every 3 months and CT scans every 6 months for the 1st year after
initial treatment and, then annually. Additional CT imaging was
performed when any symptoms developed, when CA125 values reached
≥35 U/ml without symptoms, or when CA125 values showed a twofold
increase from previous measurements. If CT imaging performed due to
CA125 elevation failed to confirm recurrence, monthly CA125 testing
was initiated followed by additional CT imaging as clinically
indicated. Histological subtypes were divided into two groups:
Group A comprised CCC and mucinous carcinoma (MC), while Group B
included HGSC, endometrioid carcinoma (EC), mixed carcinoma, and
carcinosarcoma. Staging followed FIGO classification (8), while pathological diagnosis adhered
to 2020 World Health Organization criteria (9). For recurrent cases requiring repeated
CT scans due to initial non-detection of recurrence when CA125
values ≥35 U/ml, we calculated the CA125-CT detection interval,
defined as the time between the first instance of CA125 values ≥35
U/ml without CT-detected recurrence and the subsequent CT
confirmation of recurrence.
Statistical methods
Categorical variables were compared using the
χ2 test or Fisher's exact test, as appropriate. We
evaluated the sensitivity, specificity, positive predictive value
(PPV), and negative predictive value (NPV) of CA125 for detecting
recurrence, both overall and by histological groups with 95%
confidence intervals. P<0.05 was considered to indicate a
statistically significant difference, and analyses were conducted
using JMP software (version 17.0; SAS Institute Inc.).
Evaluation of the CA125 ratio
Among patients with CA125 <35U/ml during
surveillance, we analyzed the CA125 ratio, defined as the maximum
CA125 value at CT surveillance divided by CA125 value at post
initial treatment. The formula was shown as follows.
The CA125 ratio=(The maximum CA125 value at time of
CT surveillance)/(The CA125 value at post initial treatment).
The optimal CA125 ratio cutoff was determined using
receiver operating characteristic (ROC) curve analysis for
recurrence detection in these patients. We calculated sensitivity,
specificity, PPV, and NPV using both CA125 ≥35 U/ml or the CA125
ratio, wither separately or in combination.
Results
Patient characteristics
A total of 333 patients with epithelial ovarian
cancer received initial treatment at our institution. Of these, 243
patients treated between January 2006 and December 2022 were
included in the analysis (Fig. 1).
The cohort comprised 98 patients with recurrence and 145 without
recurrence. As a supplemental analysis, we compared the clinical
backgrounds of these two groups. Patients with recurrence exhibited
significantly higher rates of advanced FIGO stages (P<0.01,
χ2 test), residual tumors (P<0.01, Fisher's exact
test), elevated CA125 levels (P<0.01, χ2 test), and
neoadjuvant chemotherapy administration (P<0.01, Fisher's exact
test).
Diagnostic performance of CA125 alone
in Group A and Group B
The study population was stratified into Groups A
(n=85) and Group B (n=158) (Table
I). Group A consisted of 63 (74.1%) patients with CCC and 22
(25.9%) with MC, while Group B comprised 94 (59.5%) with HGSC, 47
(29.7%) with EC, 14 (8.9%) with mixed carcinoma, and 3 (1.9%) with
carcinosarcoma. Compared to Group B, Group A had significantly more
patients at early FIGO stages (P<0.01), with no residual tumors
(P<0.01), normal CA125 levels (P<0.01), and no neoadjuvant
chemotherapy (P<0.01). Among 24 patients with recurrence, CT
failed to detect recurrence when initially raising CA125 values ≥35
U/ml required repeated testing. This occurred in two Group A
patients and 22 Group B patients (P=0.09). The median CA125-CT
detection interval was 3 months (range: 1-7) overall; 1 month
(range: 1-1) in Group A vs. 3 months (range: 1-7) in Group B. No
significant difference was observed in median CA125-CT detection
intervals between groups A and B (P=0.12). Using CA125 ≥35 U/ml as
the sole criterion, sensitivity were 70.4% in all patients, 28.6%
in Group A, and 81.8% in Group B, and specificity were 98.6% in all
patients, 98.4% in Group A, and 98.8% in Group B (Table II).
 | Table ICharacteristics of each group. |
Table I
Characteristics of each group.
| Variable | Group Aa (n=85) | Group Bb (n=158) | P-value |
|---|
| Age (%) | | | 0.17 |
|
≥61
years | 29 (34.1) | 68 (43.0) | |
|
<60
years | 56 (65.9) | 90 (57.0) | |
| FIGOc stage (%) | | | <0.01 |
|
I, II | 76 (89.4) | 65 (41.1) | |
|
III, IV | 9 (10.6) | 93 (58.9) | |
| Presence of residual
tumor (%) | | | <0.01 |
|
Yes | 4 (4.7) | 32 (20.3) | |
|
No | 81 (95.3) | 126 (79.7) | |
| Neoadjuvant
chemotherapy | | | <0.01 |
|
Yes | 1 (1.2) | 68 (43.0) | |
|
No | 84 (98.8) | 90 (57.0) | |
| Recurrence (%) | | | <0.01 |
|
Yes | 21 (24.7) | 77 (48.7) | |
|
No | 64 (75.3) | 81 (51.3) | |
| Number of recurrent
or non-recurrent patients with CA125 ≥35 U/ml (%) | | | |
|
Recurrence | 6 (85.7) | 63 (98.4) | 0.19 |
|
Non-recurrence | 1 (14.3) | 1 (1.6) | |
| Number of recurrent
or non-recurrent patients with CA125 <35 U/ml (%) | | | |
|
Recurrence | 15 (19.2) | 14 (14.9) | 0.45 |
|
Non-recurrence | 63 (80.8) | 80 (75.1) | |
| Table IIDiagnostic performance of CA125 alone
(≥35 U/ml) for predicting recurrence. |
Table II
Diagnostic performance of CA125 alone
(≥35 U/ml) for predicting recurrence.
| Patients | Sensitivity, % (95%
CI) | Specificity, % (95%
CI) | PPV, % (95% CI) | NPV, % (95% CI) |
|---|
| All patients
(n=243) | 70.4 (60.3-79.1) | 98.6 (95.2-99.8) | 97.2 (90.3-99.7) | 83.1 (76.8-88.3) |
| Group Aa (n=85) | 28.6 (11.3-52.2) | 98.4 (91.7-99.9) | 85.7 (42.1-99.6) | 80.8 (70.3-88.8) |
| Group Bb (n=158) | 81.8 (71.3-89.8) | 98.8 (93.5-100) | 98.4 (91.6-100) | 85.1 (76.5-91.6) |
Evaluation of the CA125 ratio for
recurrence detection
Among 162 patients with CA125 <35 U/ml, 78
patients were in Group A, 94 patients were in Group B. ROC analysis
of the CA125 ratio yielded an AUC of 0.75 (Fig. 2). At a cutoff ratio of 1.5,
sensitivity were 58.6% in all patients, 53.3% in Group A, and 64.3%
in Group B, and specificity were 86.7% in all patients, 85.7% in
Group A, and 87.5% in Group B (Table
III).
| Table IIIDiagnostic performance of the CA125
ratio (≥1.5)a for
predicting recurrence in patients with CA125 <35 U/ml. |
Table III
Diagnostic performance of the CA125
ratio (≥1.5)a for
predicting recurrence in patients with CA125 <35 U/ml.
| Patients | Sensitivity, % (95%
CI) | Specificity, % (95%
CI) | PPV, % (95%
CI) | NPV, % (95%
CI) |
|---|
| All patients with
CA125 <35 U/ml (n=162) | 58.6
(38.9-76.5) | 86.7
(80.0-91.8) | 47.2
(30.4-64.5) | 91.2
(85.1-95.4) |
| Group
Ab (n=78) | 53.3
(26.6-78.7) | 85.7
(74.6-93.3) | 47.1
(23.0-72.2) | 88.5
(77.8-95.3) |
| Group
Bc (n=94) | 64.3
(35.1-87.2) | 87.5
(78.2-93.8) | 47.4
(24.4-71.1) | 93.3
(85.1-97.8) |
Optimization of recurrence detection
using the combined model
The combined approach (CA125 ≥35 U/ml or the CA125
ratio ≥1.5) demonstrated a sensitivity of 87.7% in all patients,
66.7% in Group A, and 93.5% in Group B, and a specificity of 85.5%
in all patients, 84.1% in Group A, and 86.4% in Group B (Table IV).
| Table IVDiagnostic performance of the
combined model (CA125 ≥35 U/ml or CA125 ratioa ≥1.5) for predicting
recurrence. |
Table IV
Diagnostic performance of the
combined model (CA125 ≥35 U/ml or CA125 ratioa ≥1.5) for predicting
recurrence.
| Patients | Sensitivity, % (95%
CI) | Specificity, % (95%
CI) | PPV, % (95%
CI) | NPV, % (95%
CI) |
|---|
| All patients
(n=243) | 87.8
(79.6-93.5) | 85.5
(78.7-90.8) | 80.4
(71.6-87.4) | 91.2
(85.1-95.4) |
| Group
Ab (n=85) | 66.7
(43.0-85.4) | 84.1
(72.7-92.1) | 58.3
(36.6-77.9) | 88.3
(77.4-95.2) |
| Group
Bc (n=158) | 93.5
(85.5-97.9) | 86.4
(77.0-93.0) | 86.7
(77.5-93.2) | 93.3
(85.1-97.8) |
Discussion
This study evaluated the clinical utility of CA125
in detecting epithelial OC recurrence by histological subtype.
While the CA125 test was inadequate for CCC and MC, it demonstrated
higher utility for the group mainly comprising HGSC and EC.
Furthermore, combining the CA125 test with CA125 ratio improved
sensitivity for CCC and MC, though this approach remained limited.
However, this combined analysis was more effective for the HGSC and
EC groups.
Among OC, CCC and MC subtypes, CCC and MC
represented minor subtypes worldwide; however, CCC is a relatively
more prevalent in Asia, accounting for 30% of cases (10,11).
Therefore, HGSC is widely recognized as the major histological
subtype (2). Compared to HGSC, CCC
and MC exhibited greater chemoresistance, are diagnosed at a
younger age and earlier stage, and are associated with a higher
potential for complete resection (no residual tumor) and a lower
rate of Neoadjuvant chemotherapy (12-14).
Since our results align with previous studies, the study groups
included in our study reflect common clinical presentations.
Previous studies reported a median CA125-CT detection interval of
3-5 months in HGSC (15).
Consistent with this, our study found a median interval of 3 months
in both groups, with no significant difference across histological
subtypes.
In our study, the CA125 test demonstrated low
sensitivity for detecting recurrence in patients with CCC and MC,
whereas HGSC it showed high sensitivity and specificity for HGSC,
consistent with ESMO and NCCN guidelines (3,4).
This discrepancy arises because CA125 levels are potentially
reflect HGSC disease burden at recurrence or initial diagnosis
(16), while CCC and MC typically
exhibit lower CA125 responsiveness (5). Therefore, alternative methods may be
required to detect recurrence in patients with CCC and MC.
Detecting recurrence in patients with CA125 levels
of <35 U/ml remains challenging (17,18).
Therefore, we evaluated the CA125 ratio, the rate of CA125
increase. Although the CA125 ratio was only moderately efficient
for all patients, the combined CA125 ≥35 U/ml or the CA125 ratio
model had a higher sensitivity, particularly for patients with the
group including HGSC as the major types. Therefore, this method
might be useful in reducing the number of CT scans required for
patients with HGSC and EC as the major types. However, its utility
was limited to patients with CCC and MC, highlighting that CT scans
cannot be omitted during surveillance in patients with CCC and
MC.
The difference of the utility of CA125 between two
Groups in our study might be affected with the imbalance of FIGO
stage. In fact, most cases were discovered at early-staged diseases
in Group A, while most cases in Group B were found at
advanced-staged diseases. Generally, several reports showed CCC and
MC are diagnosed at an early stage and HGSC is often found at an
advanced stage because of these biological characteristics itself
(10-13).
Therefore, this distribution reflected the actual clinical
situation in daily practice. Because our study analyzed the utility
of CA125 on the perspective of histology, we considered this
imbalance as not confounding factor but histological
characteristics. Thus, our study emphasized the specific follow-up
strategies using CA125 for different histological subtypes.
This study had several limitations. First, our
findings are based on a single-institution retrospective design
with a relatively small sample size. Second, we did not perform the
analysis with factors effected with CA125 such as tumor size or
sites of recurrence. Finally, this real-world, non-standardized
imaging protocol could influence the calculated sensitivity and
specificity compared to a theoretical study where all patients
received imaging at fixed intervals regardless of CA125. Further
large-scale studies are needed to evaluate the biological influence
of histology and to optimize follow-up protocols for these
subtypes.
Our findings would be useful for the follow-up
methods at clinical setting. Because the sensitivity of the
follow-up methods by the CA125 value and the CA125 ratio of
patients with CCC and MC was not satisfactory, CT scan was often
performed in order to miss the recurrence. Hence, because the
sensitivity of this follow-up method for patients with HGSC was
sufficient, we could reduce the times of CT scan to discover
recurrence. Thus, our results might minimize medical costs and
medical radiation exposure.
In conclusion, our findings demonstrate that CA125
is clinically useful for detecting recurrence in patients with HGSC
but shows limited utility in CCC and MC. These results highlight
the need for subtype-specific recurrence detection strategies,
particularly for CCC and MC, warranting further investigation into
surveillance methods.
Acknowledgements
Not applicable.
Funding
Funding: No funding was received.
Availability of data and materials
The data generated in the present study may be
requested from the corresponding author.
Authors' contributions
KK, MM and MT conceptualized the study. KK and MM
were involved in data curation. KK, SN, RT, TI, NK, JS, YO, TH, HS,
KO and YH performed the formal analysis. KK wrote the original
draft. MM and MT reviewed and edited the manuscript. KK and MM
confirm the authenticity of all the raw data. All authors read and
approved the final manuscript.
Ethics approval and consent to
participate
The present study was approved by the Institutional
Review Board of National Defense Medical College (approval no.
4925; Tokorozawa, Japan). All patient records and information were
anonymized and de-identified before analysis. The present study was
exempt from obtaining informed consent from all participants.
Patient consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing
interests.
References
|
1
|
Bray F, Laversanne M, Sung H, Ferlay J,
Siegel RL, Soerjomataram I and Jemal A: Global cancer statistics
2022: GLOBOCAN estimates of incidence and mortality worldwide for
36 cancers in 185 countries. CA Cancer J Clin. 74:229–263.
2024.PubMed/NCBI View Article : Google Scholar
|
|
2
|
Kuroki L and Guntupalli SR: Treatment of
epithelial ovarian cancer. BMJ. 371(m3773)2020.PubMed/NCBI View Article : Google Scholar
|
|
3
|
González-Martín A, Harter P, Leary A,
Lorusso D, Miller RE, Pothuri B, Ray-Coquard I, Tan DSP, Bellet E,
Oaknin A, et al: Newly diagnosed and relapsed epithelial ovarian
cancer: ESMO Clinical Practice Guideline for diagnosis, treatment
and follow-up. Ann Oncol. 34:833–848. 2023.PubMed/NCBI View Article : Google Scholar
|
|
4
|
NCCN Guidelines: Version 4.2026:
Epithelial Ovarian Cancer/Fallopian Tube Cancer/Primary Peritoneal
Cancer. https://www.nccn.org/professionals/physician_gls/pdf/ovarian.pdf.
|
|
5
|
Liu H, Xu Y, Ji J, Dong R, Qiu H and Dai
X: Prognosis of ovarian clear cell cancer compared with other
epithelial cancer types: A population-based analysis. Oncol Lett.
19:1947–1957. 2020.PubMed/NCBI View Article : Google Scholar
|
|
6
|
Charkhchi P, Cybulski C, Gronwald J, Wong
FO, Narod SA and Akbari MR: CA125 and ovarian cancer: A
comprehensive review. Cancers (Basel). 12(3730)2020.PubMed/NCBI View Article : Google Scholar
|
|
7
|
Eisenhauer EA, Therasse P, Bogaerts J,
Schwartz LH, Sargent D, Ford R, Dancey J, Arbuck S, Gwyther S,
Mooney M, et al: New response evaluation criteria in solid tumours:
Revised RECIST guideline (version 1.1). Eur J Cancer. 45:228–247.
2009.PubMed/NCBI View Article : Google Scholar
|
|
8
|
Prat J: FIGO Committee on Gynecologic
Oncology. FIGO's staging classification for cancer of the ovary,
fallopian tube, and peritoneum: Abridged republication. J Gynecol
Oncol. 26:87–89. 2017.PubMed/NCBI View Article : Google Scholar
|
|
9
|
World Health Organization Classification
of Tumours Editorial Board, Female Genital Tumours: WHO
Classification of Tumours. Vol 4, 5th edition. International Agency
for Research on Cancer, Lyon, 2020.
|
|
10
|
Kim SI, Lim MC, Lim J, Won YJ, Seo SS,
Kang S and Park SY: Incidence of epithelial ovarian cancer
according to histologic subtypes in Korea, 1999 to 2012. J Gynecol
Oncol. 27(e5)2016.PubMed/NCBI View Article : Google Scholar
|
|
11
|
Okamoto A, Glasspool RM, Mabuchi S,
Matsumura N, Nomura H, Itamochi H, Takano M, Takano T, Susumu N,
Aoki D, et al: Gynecologic Cancer InterGroup (GCIG) consensus
review for clear cell carcinoma of the ovary. Int J Gynecol Cancer.
24 (9 Suppl 3):S20–S25. 2014.PubMed/NCBI View Article : Google Scholar
|
|
12
|
Oliver KE, Brady WE, Birrer M, Gershenson
DM, Fleming G, Copeland LJ, Tewari K, Argenta PA, Mannel RS, Secord
AA, et al: An evaluation of progression free survival and overall
survival of ovarian cancer patients with clear cell carcinoma
versus serous carcinoma treated with platinum therapy: An NRG
Oncology/Gynecologic Oncology Group experience. Gynecol Oncol.
147:243–249. 2017.PubMed/NCBI View Article : Google Scholar
|
|
13
|
Takano M, Kikuchi Y, Yaegashi N, Kuzuya K,
Ueki M, Tsuda H, Suzuki M, Kigawa J, Takeuchi S, Tsuda H, et al:
Clear cell carcinoma of the ovary: A retrospective multicentre
experience of 254 patients with complete surgical staging. Br J
Cancer. 94:1369–1374. 2006.PubMed/NCBI View Article : Google Scholar
|
|
14
|
Borella F, Mitidieri M, Cosma S, Benedetto
C, Bertero L, Fucina S, Ray-Coquard I, Carapezzi A and Ferraioli D:
Update on prognostic and predictive markers in mucinous ovarian
cancer. Cancers (Basel). 15(1172)2023.PubMed/NCBI View Article : Google Scholar
|
|
15
|
Rustin GJ, van der Burg ME, Griffin CL,
Guthrie D, Lamont A, Jayson GC, Kristensen G, Mediola C, Coens C,
Qian W, et al: Early versus delayed treatment of relapsed ovarian
cancer (MRC OV05/EORTC 55955): A randomised trial. Lancet.
376:1155–1163. 2010.PubMed/NCBI View Article : Google Scholar
|
|
16
|
Rustin GJ, Nelstrop AE, Tuxen MK and
Lambert HE: Defining progression of ovarian carcinoma during
follow-up according to CA 125: A North Thames Ovary Group Study.
Ann Oncol. 7:361–364. 1996.PubMed/NCBI View Article : Google Scholar
|
|
17
|
Gupta D and Lis CG: Role of CA125 in
predicting ovarian cancer survival-a review of the epidemiological
literature. J Ovarian Res. 2(13)2009.PubMed/NCBI View Article : Google Scholar
|
|
18
|
Plotti F, Guzzo F, Schirò T, Terranova C,
De Cicco Nardone C, Montera R, Luvero D, Scaletta G, Lopez S,
Capriglione S, et al: Role of human epididymis protein 4 (HE4) in
detecting recurrence in CA125 negative ovarian cancer patients. Int
J Gynecol Cancer. 29:768–771. 2019.PubMed/NCBI View Article : Google Scholar
|
