Introduction

Ovarian cancer is the second deadliest gynecological cancer, with a 5-year survival rate of 50.8%1. Over 80% of these women are diagnosed as serous epithelial tissue2. The standard treatment involves cytoreduction surgery and chemotherapy. Primary debulking surgery (PDS) followed by platinum-based chemotherapy is the preferred initial treatment for advanced-stage ovarian cancer. Neoadjuvant chemotherapy (NACT) followed by interval debulking surgery (IDS) is an alternative option for those who cannot undergo PDS. While many randomized clinical trials have shown the non-inferiority of NACT in survival outcomes3,4, a recent study found that women in the NACT group experienced more relapses in the pelvis or upper abdomen compared to the PDS group5. Regardless of the treatment option, complete resection of the tumor through surgery remains crucial in reducing the relapse rate and improving survival.

Ovarian cancer recurs in 25% of early-stage and more than 80% of advanced-stage women6. Effective follow-up tests are needed to detect relapse or disease progression. Screening methods include physical examination, serum tumor biomarkers, and imaging techniques. The most widely used tumor biomarker is CA125, a glycoprotein found on the surface of ovarian cancer cells since 1981. Additional maker as HE4 have been introduced in Guideline, but CA125 seems to be the most reliable marker for disease monitoring7. The Gynecologic Cancer InterGroup (GCIG) has already accepted the research result by Rustin that CA125 levels more than twice the upper limit of normal (2 × ULN) predict tumor relapse8. However, with the significant improvement of the density resolution technology in medical imaging, using the traditional cut-off value of 2 × ULN U/mL for relapse is being questioned. Studies have shown that even within the normal range(< 35 U/mL), three progressively rising CA125 at 1- to 3-month intervals9 or an increase of more than 5 U/mL from 3 to 6 months after treatment10 are associated with a higher risk of relapse. Therefore, a more sensitive cut-off point is needed.

In fact, relying solely on CA125 levels is not sufficient to confirm relapse. A rise to more than the cut-off value of CA125 several months before the evidence of image-identified relapse or symptomatic clinical relapse is called biochemical relapse. Early treatment based on elevated CA125 levels alone can cause anxiety and side effects without improving survival11. Excessive imaging screening would cause unnecessary anxiety and financial burden. Undoubtedly, predicting image-identified relapse based on CA125 elevation is challenging but important. Additionally, initial treatment strategies can cause CA125 profiles to be worse in those having NACT than PDS as NACT patients have more advanced disease in the real clinical world. However, the tendency of significant deviation in CA125 will disappear following treatment, as CA125 will reduce to a low level after a combination of surgery and chemotherapy. Exploring the changing trends of CA125 profiles till relapse after the two distinct treatments in the real clinical world is expected to contribute more evidence to guide clinical practice in this area.

In this study, the aim is to explore the association between biochemical relapse (elevated CA125) and image-identified relapse of ovarian cancer to improve the prediction of relapsed lesions requiring treatment. Additionally, a subgroup analysis in relapsed patients was performed to clarify the effects of different initial treatments on the relapsed outcomes at the biomedical level.

Material and methods

Study population

All women diagnosed with International Federation of Gynecology and Obstetrics (FIGO, 2013) stage I-IV ovarian, tubal, and peritoneal cancers from January 2013 to June 2019 at the Department of Gynecology, The First Affiliated Hospital of Chongqing Medical University, China, were reviewed. We retrospectively collected clinical data of 162 women who achieved complete clinical response (CCR) following primary treatment and met the following inclusion criteria: (1) All pathological types were serous; (2) Relapse was confirmed by imaging and was treated accordingly; (3) Without other malignancy; (4) Complete clinical data; (5) Regular follow-up; (6) CCR. Those who achieved without of CCR, received previous treatment for biochemical relapse, and had irregular follow-up were excluded.

In the current study, CCR was achieved when the criteria were met: the absence of tumor-associated clinical symptoms, the negative signs of residual tumor on the physical examination and imaging study results, and a serum CA125 level below the ULN (35 U/ml) after primary treatment. Image-identified relapse was identified with the occurrence of any new measurable lesion as revealed through imaging studies.

Collection of clinical information of the enrolled women

(1) Women’s baseline characteristics and initial treatment: age at diagnosis, body mass index (BMI), FIGO stage, primary tumor type, CA125 level at diagnosis, the median number of total chemotherapy cycles, surgical procedures, operation time, blood loss, surgical outcome, PARP inhibitor use, and histologic grade; (2) CA125 level per treatment phase: at diagnosis, before IDS, at the end of primary therapy, at nadir during follow-up, at 2 × nadir, at the image-identified relapse, and time intervals between the key disease-progression points; (3) Women’ relapsed characteristics: imaging method, number of sites of relapse, specific sites of relapse (pelvis, upper abdomen/diaphragm, distant, lymphatic, carcinomatosis), and the extent of relapse (intra-abdominal, extra-abdominal, ascites/effusion). The censor date was March 2023.

Follow-up

Check-ups were performed every three months for the first 2 year post-treatment, every three to six months for the third to fifth years, and annual follow-ups for the sixth year. The follow-up items included: gynecological examination, gynecological ultrasound, serum CA125, and other serum tumor markers. The CA125 was measured by the same assay during follow-up in each patient. The nadir CA125 levels of individual patients were the lowest biomarker value after primary therapy during follow-up, and the 2 × nadir level is two times of the lowest CA125 value after primary therapy during follow-up. The highest level of CA125 was defined as the apex of the follow-up CA125 level whose measurement can be confirmed without relapsed disease. An annual imaging check-up consisted of a chest high-resolution computerized tomography (HRCT) scan and an abdomen and pelvic contrast-enhanced computerized tomography (CT) or magnetic resonance imaging (MRI). Whole-body positron emission tomography-computed tomography (PET-CT) was performed if a distant metastasis was highly suspected. The Response Evaluation Criteria in Solid Tumors (RECIST) and the World Health Organization (WHO) criteria were used to assess tumor therapy response and clinical relapse.

Statistical analyses

Continuous parameters were presented as the median with IQR (inter-quarter range) and analyzed by the Mann–Whitney test. The rank categorical variables were described as numbers (percentages) and analyzed with the Wilcoxon rank sum test. 35 U/ml and 2 × nadir of CA125 level were set as the cut-off values of relapse in all included patients, and comparisons of correlated sensitivity and specificity were made using the size of Youden’s index. Next, 35 U/ml and 2 × nadir of CA125 level were set as the key points in relapsed patients, and comparisons of correlated proportions were made using the McNemar test. The IBM SPSS Statistics 26.0 software was used for statistical analysis. Violin plots and ROC curve were created with GraphPad Prism 9. P values < 0.05 were considered statistically significant.

Ethical approval

This study was approved by the Ethical Committee of The First Affiliated Hospital of Chongqing Medical University (protocol code 2022-K404 and date of approval 1 September 2022) and conducted in accordance with the Declaration of Helsinki. Each patient had given written informed consent at the time of treatment for the future use of their clinical data. This study is compliant with all institutional and national guidelines and regulations for human subjects research.

Results

Baseline characteristics of the present study population

The clinical baseline characteristics of the patients are given in Table 1. Of the 162 patients who achieved CCR following primary treatment included in the present study, 86 (53.1%) patients who were diagnosed with imaging-identified relapsed disease were defined as the relapse group, and 76 (46.9%) patients who were confirmed without relapsed disease as the relapse-free group. Most patients’ (143, 88.3%) tumors originated from the ovary. 52 (68.4%) and 75 (87.2%) patients had high-grade primary tumors in the relapse-free and relapse group, respectively (P = 0.006). Approximately 69.7% were classified in FIGO stage II/III in the relapse-free group and 75.6% were classified in FIGO stage III/IV in the relapse groups, respectively (P < 0.001). Likewise, the CA125 level at diagnosis in the relapse group was significantly higher than in the relapse-free group (P = 0.026). In primary treatment, both the proportion of NACT (52.3% vs 35.5%, P = 0.032) and distant metastases resection (15.1% vs 2.6%, P = 0.006) in the relapse group was significantly higher than that of the relapse-free group. In the surgical outcome, no residual tumors were achieved in more than half of the women in the relapse-free and relapse groups, respectively (71.1% vs 53.3%, P = 0.008). The two groups had similar baseline features, including age, BMI, tumor origin, Median number of chemotherapy cycles, pelvic and para-aortic lymphadenectomy. The median CA-125 level at image-identified relapse (64.0 U/ml; IQR, 29.4 U/ml to 134.1 U/ml) was significantly higher than that of the median pre-relapse highest CA-125 level (12.8 U/ml; IQR, 9.8 U/ml to 18.8 U/ml; P < 0.001; Fig. 1).

Table 1 Baseline characteristics of the present study population (162 women with ovarian, tubal, and peritoneal cancer).
Full size table
Figure 1
figure 1

Relapse levels of CA-125 in patients who developed image-identified relapse of serous ovarian, tubal, and peritoneal cancers (median, 64.0 U/ml; IQR, 29.4 U/ml to 134.1 U/ml) were significantly higher than the highest CA-125 level recorded during follow-up in patients who did not develop relapsed disease (median, 12.8 U/ml; IQR, 9.8 U/ml to 18.8 U/ml; P < 0.001).

Full size image

Prediction of imaging-identified relapse using follow-up CA-125 level

To get closer to the clinical reality, we evaluated two CA125 cut-off values that have significant predictive value for relapse: ULN (35 U/ml) and the 2 × nadir level. Of the 162 patients who achieved CCR following primary treatment, 64 (39.5%) had an increased CA125 level to 35 U/ml and 87 (53.7%) had an increased CA125 level to 2 × nadir during follow-up. When the cut-off value of CA125 was 35 U/mL, the sensitivity and specificity for predicting clinical relapse were 67.4% and 92.1% respectively; When the cut-off value of CA125 reached 2 × nadir, the sensitivity and specificity for predicting clinical relapse were 84.9% and 81.6% respectively. The difference was statistically significant, the Youden’s index of the 2 × nadir U/ml was higher than 35 U/ml (0.665 vs 0.595, Table 2). Of the 86 patients with imaging-identified relapse, 58 (67.4%) showed a CA125 level reaching 35 U/ml, while 73 (84.9%) showed a CA125 level getting 2 × nadir before relapse. The difference was statistically significant (P < 0.001, Table 3). Of 86 women experiencing relapse, 17(19.8%) whose follow-up CA125 reached its nadir × 2 level without reaching 35 U/ml could be identified as high-risk populations of omissions in clinical practice (Table 3). Moreover, as shown in Fig. 2, if the CA125 concentration at the image-identified relapse exceeded 35 U/mL, patients were defined as the 35 U/ml group; If the CA125 concentration at the image-identified relapse exceeded its 2 × nadir, patients were defined as the 2 × nadir group. The median time interval from CA125 reaching 35 U/mL to image-identified relapse was 0 months (IQR, 0 to 2.19 months), while that from CA125 reaching 2 × nadir to image-identified relapse was 2.67 months (IQR, 0.85 to 6.25 months), with a statistically significant difference (P < 0.001, Fig. 2).

Table 2 Image-identified relapse prediction using the CA125 cut-off values of 35 U/ml and 2 × nadir U/ml in 162 women with ovarian, tubal, and peritoneal cancer.
Full size table
Table 3 The classification of 86 relapsed women based on different CA125 cutoff values.
Full size table
Figure 2
figure 2

The time interval characteristics of the 35 U/m group and the 2 × nadir group in 86 relapsed women. If the CA125 concentration at the image-identified relapse exceeded 35 U/mL, patients were defined as the 35 U/ml group; If the CA125 concentration at the image-identified relapse exceeded its 2 × nadir, patients were defined as the 2 × nadir group. (A) 35 U/ml group: violin plot of time intervals (median, 0 months; IQR, 0 months to 2.19 months) between firstly reaching 35 U/mL of CA125 and image-identified relapse in fifty-eight women (58, 67.4%) ; (B) 2 × nadir group: violin plot of time intervals (median, 2.67 months; IQR, 0.85 months to 6.25 months) between firstly reaching 2 × nadire U/mL of CA125 and image-identified relapse in seventy-three women (73, 84.9%). Inside each violin plot is a box plot showing the quartile, median, and range distribution. There are significant differences between red violin and blue violin distributions in each plot.

Full size image

Baseline characteristics and relapsed characteristics of relapsed patients

Among the 86 women with image-identified relapse, 41 (47.7%) women undergoing PDS followed by ACT were defined as the PDS group, and 45 (52.3%) women undergoing NACT followed by IDS were defined as the NACT group, for the initial treatment. The baseline and relapsed characteristics of 86 relapsed women are listed in Table 4. At initial diagnosis, 68.3% of the women in the PDS arm had a serum CA125 level lower than 1000 U/mL, while 57.8% of the women in the NACT arm showed a CA125 over 1000 U/mL (P = 0.003). Approximately 95.5% and 53.7% were classified in FIGO stage III/IV in the NACT and PDS groups, respectively (P < 0.001). Likewise, the median number of chemotherapy cycles in the PDS arm is six compared to eight in the NACT arm (P = 0.019). The two groups with relapse had similar baseline features, including age, BMI, and primary tumor type. High-grade serous histology was observed in nearly 90% of women in each group. Both complete and optimal surgeries were achieved in more than half of the women (53.7% vs. 53.3% in the PDS and NACT groups, respectively). No differences were observed in the surgical parameters between the two groups with relapsed disease.

Table 4 Baseline and relapsed characteristics of 86 women with relapsed ovarian, tubal, and peritoneal cancer.
Full size table

As shown in Table 4, most women were detected to have multiple relapsed lesions by the examination of PET-CT and MRI. The pelvis was the most common site of relapse (62.8%) and was detected with a higher percentage in the NACT group than in the PDS group (71.1% vs. 53.7%, P = 0.094). The incidence of distant relapses (brain/chest/liver/spleen parenchyma) in the NACT group was significantly higher than in the PDS group (46.7% vs. 17.1%, P = 0.003). An intra-abdominal relapse was far more common than an extra-abdominal one in all women (82.6% vs. 17.4%), with the same trend in both groups.

CA125 values and time intervals between the key disease-progression points of relapsed patients

Serum CA125 values at various time points and time intervals were presented in Table 5. For all the relapsed women, the median CA125 level at diagnosis was 929.9 U/mL, IQR from 495.8 to 2256.6 U/mL. Relapsed women in the NACT group showed a higher CA125 than those in the PDS group at diagnosis (1523.6 vs. 663.1 U/mL, P = 0.003). After initial treatment, the median CA125 value of 86 relapsed women decreased to nearly 10 U/mL, IQR from 7 to 13.8 U/mL. During follow-up, the median CA125 level at nadir was 7.8 U/mL in 86 relapsed women, and there was no difference between the two groups (7.6 vs. 8.8 U/mL, respectively). Subsequently, the disease progressed until imaging-relapsed evidence was detected. At this period, the median CA125 level in the PDS group (60.5 U/mL) was slightly higher than that in the NACT group (54.2 U/mL), but there was no significant difference between the two groups with relapsed disease. In the PDS group, it took 7.5 months from surgery to the CA125 level reaching the nadir and then 11.6 months from the nadir level to image-identified relapse. The NACT group took 6.6 months from surgery to the CA125 nadir level and 9.6 months from the nadir level to the image-identified relapse. Although the PDS group is slightly longer than the NACT group in each timespan, there was still no statistical difference between the two groups.

Table 5 Serum CA125 concentrations at various time points and time intervals between the key disease-progression points of 86 relapsed women.
Full size table

Discussion

The 5-year and 12-year survival rates of relapsed ovarian cancer were less than 30% and 5%, respectively12. Monitoring relapse is a key aspect of the overall management of women with EOC. With CA125 levels frequently rising several months before image-identified relapse11, determining the imaging examination time based on the fluctuation of CA125 values could alleviate anxiety and reduce the economic burden for women with asymptomatic relapse by avoiding multiple negative imaging examinations. Therefore, defining the CA125 cut-off that signifies the need for diagnostic imaging and therapy to relapse is critical to ensure effective management.

In 1996, RUSTIN et al. posited that the criterion for tumor progression in patients whose CA125 levels fell to within the normal range is a doubling of CA125 level from the upper limit of normal (2 × ULN)8. In 2001, they confirmed that doubling of CA-125 from its nadir level has now been shown to accurately define progression in patients with persistently elevated levels over the normal range13. Based on the research results of RUSTIN’s team, the GCIG accepted the CA125 criteria into the Response Criteria in Solid Tumor (RECIST) to predict ovarian cancer relapse. The criterion of EOC relapse has remained relatively stable over the past 20 years; in comparison, the density resolution technology in medical imaging has improved significantly, and more timely and accurate detection of smaller lesions was found. Using a CA125 level of 2 × ULN (70 U/ml) as part of the criterion remains uncertain. The difference between RUSTIN’s study and our study is that we investigated whether 35 U/ml and 2 × nadir of CA125 level can increase the accuracy of progression instead of traditional 2 × ULN of CA125. Compared to the population of RUSTIN’s study, in our study, we selected relatively stable patients whose CA125 decreased to normal (< 35 U/ml) and showed complete clinical response after primary treatment with regular follow-up. The CCR patients who were detected with an increased CA125 level of 2 × ULN during follow-up were only 37 (22.8%), which was significantly lower than that of 35 U/ml and 2 × nadir.

However, using 35 U/ml to replace 2 × ULN as the CA125 abnormal threshold for all CCR women in general also may result in omissions or delays in relapse detection. In this study, by using the sensitivity and specificity of the 2 × nadir U/ml and 35 U/ml to compare the size of Youden’s index, we found that CA125 reaching 2 × nadir during the follow-up process might be a more sensitive and early relapse signal in women having completed initial treatments for serous ovarian cancer. Moreover, more than half of the women (17, 60.71%) who never reached the upper limit of the normal range of 35 U/mL at the time of relapse showed a CA125 level over 2 × nadir. The combination of 2 × nadir and 35 U/mL predicted outcomes better than 35 U/mL alone, which could significantly improve the detection rate of relapse.

Many studies have redefined the value of CA125 in predicting relapse. A rise of 5 U/mL or 10 U/mL from the nadir CA125 level has been associated with relapse in women who had a complete response to therapy14. Wang et al. proposed that the CA125 level of 1.68 × nadir was defined as the indicator of relapsed disease15. In a French multicenter study, nadir CA125 values below 20 kU/L were associated with more prolonged Overall Survival (OS) and Disease-free Survival (DFS) (P < 0.0001)16. However, these studies aimed to reveal the association between CA125 elevation and the ultimate relapse without proposing a specific follow-up strategy. In the only randomized trial ever performed of CA125 monitoring after the initial treatment, patients who only experienced the increase of CA125 were assigned to early treatment, and patients with clinical or symptomatic relapse were assigned to delayed treatment11. The results showed no evidence of the survival benefit of early treatment. This finding challenged the widespread belief by denying routine recommend of CA125 follow-up and the early treatment of relapse based on a raised CA125 concentration alone. Actually, monitoring CA125 has been a routine in conventional follow-up, and effective but slightly expensive imaging examinations, such as computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography/computed tomography (PET/CT), were used as an additional check when CA125 is abnormal17,18. Therefore, our study wants to analyze and investigate the correlation between early and delayed treatment groups in the study mentioned above, whether the raised CA125 concentration alone can predict the clinical relapse of the delayed treatment group. It is essential but challenging for asymptomatic women to determine the optimal imaging examination timing based on the fluctuation of CA125. Some gynecologic oncologists advocate a wait-and-see approach, but the duration of the waiting period and the relapse surveillance are unclear19. Asymptomatic women waiting for symptoms may suffer from anxiety, leading to decreased quality of life and unnecessary imaging examinations11. Despite the SGO and NCCN guidelines recommending that imaging only be used when clinically indicated, more than 75% of ovarian cancer women were still receiving routine surveillance for non-indication imaging20,21. Hence, a clear cut-off of the CA125 level to start the diagnostic imaging was required. According to our results, when the CA125 level reaches 2.095 × nadir or 35 U/mL, an imaging examination at this follow-up visit or the subsequent follow-up would be recommended. More specifically, in clinical practice, imaging examinations should be performed once the CA125 level exceeds 35 U/mL or within three months for women whose CA125 level is consistently lower than 35 U/mL but reaches 2 × nadir.

In the study of Wang et al., comparing with a CA125 level > 1.68 × nadir at relapse, women with CA125 level ≤ 1.68 × nadir at relapse can extend the overall and progression-free survival durations15. In the present study, the sensitivity attained using a CA125 level of 2 × nadir was relatively high (84.9%). However, our findings did not reveal a relationship between the CA125 levels at image-identified relapse and overall survival. The possible reason was that not every asymptomatic woman start treatment immediately after observing image-identified relapse. Given the selection bias inherent to retrospective studies, attempts to translate the above-mentioned favorable outcome to all CCR OvCa women should be approached with caution. Our study aimed not only to predict relapse, but also to clarify the time point to start image examination and appropriate therapy according to the CA125 levels. Based on the current results, our team would further verify the improvement effect of intervention at 2 × nadir CA125 value on survival outcomes by performing prospective trials.

Several differences were noticed between the relapse-free and relapse groups, including those in the FIGO stage, CA125 at diagnosis, primary treatment, surgical procedures, surgical outcome, and histologic grade. The higher clinical stage, CA125 level, and histologic grade frequently represented a sign of a more serious malignant condition and far-spread range of the disease, which tends to result in extending the application of neoadjuvant chemotherapy and debulking surgery. Besides, several differences were noticed between the NACT and PDS groups in the relapsed patients, including those in the FIGO stage, CA125 at diagnosis, the number of chemotherapy cycles, and the incidence of distant metastasis. These differences might be related to the severity of the disease itself in the two groups of women. Gynecologic oncologists would like to use platinum-based chemotherapy several times in advance to reduce the tumor size of women with advanced OC, aiming to achieve optimal surgery3,4,22,23.

Focused on relapse disease in our study, the tendency of significant deviation in CA125 disappeared following treatment, as CA125 decreased to nearly 10 U/mL, IQR from 7 to 13.8 U/mL. It is not what we expected that the changing trends of follow-up CA125 profiles till relapse after the two distinct treatments in the real clinical world was still no statistical difference. Additionally, there is evidence that most EOC and primary peritoneal cancers originated from the umbrella end of the fallopian tube. The standard treatment for these three types of cancers is almost identical to clinical guidelines. Therefore, five women with fallopian tube cancer and five women with peritoneal cancer in this study were not excluded. Limitations of this study include its retrospective and unicentric nature and the relatively small sample size. Another drawback is that we mainly focused on CA125 without considering the predictive function of other tumor biomarkers24.

Conclusion

To sum up, our results showed that the CA125 level of 2 × nadir indicates the need to initiate imaging examination within three months to detect relapsed serous ovarian, tubal, and peritoneal cancers. As the combination of 2 × nadir and 35 U/mL predicted outcomes better than either alone, it may be a practical and low-cost method for oncologists to manage women with serous OvCa.