Aim, study design, endpoints and sample size calculation

The aim of the SURVIVE study is to evaluate whether an updated and individualized breast cancer surveillance in medium and high-risk patients can improve the prognosis of these patients.

After enrollment, all participants will follow guideline based standard follow-up care in addition to blood sample collection every 3 months for the first 3 years and every 6 months for the consecutive 2 years. In case of detection of distant metastases, contralateral or local recurrence of disease, the patients will receive guideline based diagnostic measures and therapy (Fig 1, 2). Quality of Life (QoL) will be monitored every 6 months by standardized questionnaires (QLQ-C30 and PA-F12) (Fig 1, 3). Randomization will be in a 1:1 ratio to the intensive versus standard surveillance group and is blinded for patients and participating study centers (Fig 2). In the intensive surveillance group, the blood will be tested for tumor markers CA27.29 (comparable to CA15−3), CA125, carcinoembryonic antigen (CEA), CTCs and ctDNA (Fig 1, 2). For ctDNA testing, a tumor informed approach will be used, using primary tumor tissue to create an individualized primer set for each patient. Regarding the use of the tumor marker CA 27.29 instead of CA 15.3, it has been shown that they are usually concordant, but it is not advised to use them interchangeably [19]. Therefore, one could argue if the results of this trial are transferable to CA 15.3. However, the immunoassays for CA15.3 and CA27.29 target the epitopes on the same MUC1 – which is associated with higher tumor burden – and are considered identical [20,21]. In the SOC group, the blood samples will be stored in a biobank for translational research. If there is a positive biomarker result in the Intensive Surveillance group, the treatment centers and therefore the patients will be notified of the result (and consequently unblinded), and imaging in terms of CT of chest and abdomen and bone scan is advised. If distant metastases are detected, patients should receive therapy in accordance with national guidelines. If there is no evidence of disease, the patient will continue to receive blood testing in the scheduled intervals.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. SPIRIT schedule of enrollment, interventions, and assessments.

a: Results of screening tests or examinations performed as standard of care prior to obtaining informed consent but within the timelines may be used rather than repeating required test; b: Premature study intervention discontinuation due to either of the following: 1. via imaging verified recurrence, 2. Pregnancy, 3. safety or medical reasons, 4. patients’ withdrawal of consent; c: Primary tumor tissue according to the requirements stated in the study protocol section 11.2. Important note: in case of bilateral breast cancer, tissue from both sides is needed. For multifocal breast cancer: tissue sample of one focus is sufficient, if the tumor biology among the foci does not differ. However, if the tumor biology among the foci differs, tissue of the therapy-defining focus, according to the investigator, should be used; d:Results of serum creatinine and renal clearance performed within 84 days prior to obtaining informed consent may be used rather than repeating required test; e: Materials for blood collection and transportation devices will be provided by the Central Study Center at Dep. Ob/Gyn University Hospital Ulm; f: Mandatory at Month 12, additionally required at visit after first verified tumor marker increase, whenever that occurs; g: Highly endorsed in case of detection of recurrence/metastatic disease; h: Referral to radiologist from investigator or aftercare specialist (gynecologist/family physician/oncologist) – Can be triggered by either of the following: 1. clinical indication (both arms), 2. increase in tumor markers (only intensive surveillance arm), 3. if presence of ctDNA was detected (only intensive surveillance arm); i: If CT and SPECT not possible due to, e.g., allergy to the contrast agent: PET scan or whole-body MRI, Complementary imaging tests if indicated; j: Available on paper, handed out at study visits; k: Staging examinations or Breast Imaging obtained during the standard aftercare visits should be documented in the eCRF (e.g., Bilateral mammography and supplementary bilateral breast ultrasound with axillary lymph nodes (where applicable) and/or MRI of the breasts.

https://doi.org/10.1371/journal.pone.0331203.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Study design. ctDNA = Circulating free tumor DNA, CTC = circulating tumor cells, M0 = no evidence of distant metastases, M1 = evidence of distant metastasis.

https://doi.org/10.1371/journal.pone.0331203.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Timeline of the SURVIVE study.

QoL = Quality of life.

https://doi.org/10.1371/journal.pone.0331203.g003

Co-primary endpoints will be 5-year OS rate and the overall lead time effect. The latter is defined as the median time from molecular to via imaging verified distant recurrence Lead Time (which can be calculated only for patients in the liquid-biopsy guided arm) plus the difference in time to distant recurrence between the two study groups.

Secondary endpoints are:

• Invasive disease-free survival (IDFS)
• Distant disease-free survival (DDFS)
• Distant recurrence-free survival (DRFS)
• Breast cancer specific survival (BCSS)
• Invasive breast cancer free survival (IBCFS)
• 10-year OS rate
• Molecular to via Imaging verified Distant Recurrence Lead Time in the Interventional arm
• Quality of life (QoL) with questionnaires: EORTC QLQ-C30 and PA-F12
• Liquid biopsy sensitivity (CA27.29, CEA, CA125, CTC and ctDNA)
• Liquid biopsy specificity (CA27.29, CEA, CA125, CTC and ctDNA)
• Liquid Biopsy False-Positive Rate (CA27.29, CEA, CA125, CTC and ctDNA)
• Liquid Biopsy False-Negative Rate (CA27.29, CEA, CA125, CTC and ctDNA)
• Rate of liquid biopsy positivity (CA27.29, CEA, CA125, CTC and ctDNA)

The study is designed as a two-arm parallel, (partially) double-blinded randomized superiority study. According to the sample size calculation, approximately 3500 randomized patients (1750 in both arms) will be required to achieve 80% power at a 2-sided significance level of 5% to detect a 2.35% improvement in 5-year OS rate from 92.65% in the standard surveillance arm to 95.0% in the tumor marker, ctDNA and CTC guided surveillance arm (equivalent to a hazard ratio of 0.672) using a log-rank test. The reference value of 92.65% 5-year OS rate is based on the results from the randomized phase III Success A study that reported 5-year OS rates for 3754 intermediate-to-high-risk early breast cancer patients receiving standard chemotherapy regimen [22]. Furthermore, this sample size calculation is based on the assumptions of a uniform accrual pattern (accrual time 2 years), a dropout rate of 15% in both arms and a total study duration of 10 years.

Study population

Participants must have histologically confirmed, intermediate to high-risk breast cancer without any evidence of distant disease. The primary antineoplastic therapy should have been completed (adjuvant chemotherapy, surgery or radiotherapy, whichever occurs last) no more than 24 months prior to study enrollment in triple-negative or HER2 positive breast cancer (Table 1). For hormone receptor positive disease, enrollment can take place up to 60 months after the end of primary anti-tumor therapy (Table 1). Enrollment of patients during any kind of adjuvant therapy except chemotherapy (e.g., but not limited to endocrine therapy, antibody therapy, CDK4/6-inhibitors, PARP inhibitors, PI3K inhibitors, antibody-drug conjugates and other novel agents) is allowed. Sufficient amount of primary tumor tissue must be available to be able to perform the tumor-informed ctDNA test (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Main inclusion and exclusion criteria.

https://doi.org/10.1371/journal.pone.0331203.t001

Study interventions and randomization

In total, 3500 patients shall be enrolled in the SURVIVE trial and will be randomized in a 1:1 ratio in either the intensive surveillance or the standard surveillance group. Stratification factors will be hormone receptor status (negative vs positive), HER2 status (negative vs positive) and histological lymph node status at surgery ((y)pN0 vs (y)pN+).

Peripheral blood samples will be drawn from patients in both groups every 3 months for the first 3 years after study inclusion, thereafter every 6 months for another 2 years. In the Standard of care group, the blood will be stored in a biobank for further translational studies and retrospective analyses. In the Intensive Surveillance group, the blood will be analyzed regarding tumor markers (CA27.29, CEA and CA125) and ctDNA levels. CTCs analysis will be performed in every patient of the Intensive Surveillance group one year after enrollment and will furthermore be triggered after the first detection of abnormal tumor marker levels. Based on evidence for associations with tumor load, an abnormal increase in serum tumor marker levels as well as detection of ctDNA or CTCs will initiate further staging examinations for patients in the Intensive Surveillance group.

An abnormal increase in serum tumor markers is defined as a delta in increase from baseline (CA27.29 + 75% and/or CEA + 100% and/or CA125 + 150%), because current available data suggest that using kinetics improve sensitivity and specificity for the detection of distant metastases compared to static cut-offs [23–25]. One of the hypothesized reasons is heterogeneity of serum levels on an inter- and intraindividual level due to several possible interferences such as renal function, current health status (in particular infections) or comorbidities. In contrast to serum tumor markers, where a defined increase compared to individual baseline values is necessary to trigger imaging, a single positive CTC or ctDNA result is sufficient to trigger the performance of staging examinations.

Given the possibility of several subsequent positive liquid biopsy results without detection of metastases in the corresponding imaging, certain rules have been established to restrict the radiation exposure for patients. Generally, and used independently from the liquid biopsy marker, the minimum interval between two staging examinations should be at least 3 months for CT scans and at least 6 months for bone scans. In addition, there are the following biomarker-specific algorithms: If there is a serum tumor marker elevation at subsequent time points without any evidence of metastatic disease as confirmed by imaging, a maximum of three consecutive negative staging examinations will be performed before a further increase in tumor marker levels will be required (according to the deltas stated above) to trigger another imaging. If a patient is CTC-positive and the performed corresponding imaging was negative, the regular standard liquid biopsy testing will commence without further CTC testing. In case of ctDNA positivity and negative imaging the patients will continue liquid biopsy-based surveillance. If the ctDNA persists a minimum interval of three months between two CT-scans is mandatory.

If clinical symptoms arise that could indicate distant metastases, patients in both groups should undergo further staging examinations in accordance with current national guidelines. If distant metastases are verified by imaging, patients must be treated according to the current standard of care, thereby ending the intervention phase of the SURVIVE study.

QoL is assessed every six months by using the two questionnaires EORTC QLQ C30, which is well established regarding general quality of life, and the PA-F12, which focuses on fear of disease progression.

A follow-up of five years with yearly checks on patients’ well-being, QoL and status of recurrence will take place after the end of the 5-year intervention phase.

Statistical analysis and data management plan

The primary endpoint overall survival will be analyzed based on the intention-to-treat (ITT) set, estimated by the Kaplan Meier product limit method, and 5-year OS rates, 95% confidence intervals and survival plots will be provided. Overall survival will be compared between patients in the two randomization arms using the log-rank test, and univariable cox regression as well as additional multivariable analyses adjusted for other factors will be performed using suitable regression models (cox proportional hazard regression model). Factors to be adjusted for will comprise the stratification factors hormone receptor status, HER2 status and lymph node status at surgery as well as other potentially prognostic clinic-pathological parameters such as age, body mass index, tumor size, tumor grade and tumor histology. The final variable list for the multivariable analysis will be set forth in a statistical analysis plan (SAP), which will be finalized prior to data base lock. Hazard ratios and the corresponding 95% confidence intervals will be reported as obtained by both univariable and adjusted multivariable analyses. The co-primary endpoint Overall Lead Time Effect will be calculated as described above. As it is a purely descriptive composite measure consisting of two unrelated median values calculated for different patient cohorts, no confidence intervals will be provided. Given that the co-primary endpoint Overall Lead Time Effect is purely descriptive and does not involve a statistical comparison between the two randomization arms, there is no need for adjusting the significance level for multiplicity for the primary endpoint analyses.

A three-step futility analysis to assess ctDNA positivity rate and molecular as well as overall lead time at pre-defined timepoints is planned to facilitate decisions as to whether or not the trial will be continued in case of unexpected low ctDNA positivity rates or unexpected short lead times. While all analyses regarding secondary objectives will have exploratory and hypothesis-generating character only, they will provide a wealth of important data regarding the usability of different liquid biopsy markers in the context of breast cancer surveillance and MRD detection following primary therapy in breast cancer patients.

No formal efficacy interim analysis with specifically defined stopping rules is planned. An informal, non-binding interim analyses of QoL will be performed one year after 50% of patients (n = 1750) have been randomized. Based on the individual data from the PA-F12 and EORTC QLQ C30 questionnaires obtained from these patients 12 months after randomization, differences in QoL between the two randomization arms will be evaluated (with the main focus being fear of progression and depression) to ensure that patients in the Intensive Surveillance arm do not suffer from unacceptable emotional burden as compared to the standard arm. There will be no prespecified rules for stopping the trial due to concerns regarding emotional burden for patients in the experimental arm. However, the results of the QoL interim analysis will be presented to the Independent Data Monitoring Committee (IDMC) to evaluate whether a premature termination of the trial due to any concerns regarding major psychological or physical harms is deemed necessary.

Electronic data capture is performed using secuTrial® and statistical analyses will be conducted using the statistical software package IBM SPSS Statistics (IBM Corp., Armonk, N.Y., USA).

Regulatory, dissemination and timeline

The trial has initially been approved by the Ethics Committee of the Ulm University (Nr 341/22) on 19^th October 2022, and two amendments have been approved in March 2023 and May 2024 comprising among several smaller changes and modifications the extension of the enrollment period of hormone receptor positive breast cancer patients. The study will be conducted at over 100 centers which are distributed all over Germany to assure the availability to the majority of breast cancer survivors.

First patient in was Q4/2022, and with an estimated recruitment period of four years, last patient in is predicted for Q4/2026. Accordingly, end of intervention is predicted for Q4/2031. As there will be a follow-up of further five years after the end of intervention, last data will be collected approximately Q4/2036. Final results will be available after this time point.

Limitations and risks of the study design

In the clinical routine, serum tumor marker testing during aftercare is still in discussion [30] and in some countries like Germany there are well established guidelines advising against this approach in the absence of symptoms [31]. Nevertheless, patients as well as doctors are frequently inclined to ask for or perform tumor marker testing. This is the reason why we integrated serum tumor markers in addition to CTCs and ctDNA in the liquid biopsy marker panel used for the assessment of molecular disease in the SURVIVE trial.

Intensified surveillance strategies inherently carry certain risks, including increased cumulative radiation exposure, the potential for treatment-related morbidity resulting from invasive diagnostic procedures, and the considerable psychological burden associated with recurrent false alarms and ongoing uncertainty.

While the biomarkers used in this trial have varying rates of specificity and sensitivity (see above), an imaging is triggered by any positive liquid biopsy result. Thus, for example even a rather unspecific rise in serum tumor markers could lead to several subsequent CT scans with no detectable metastatic disease. In addition, given that the lead time of ctDNA has been shown to be as long as three years, a positive ctDNA result can also lead to several CT scans that have to be performed regularly, before distant metastases are detected via imaging.

Several studies have shown that scan-associated anxiety can cause significant distress among cancer survivors [32]. These concerns were thoroughly discussed with patient representatives to ensure that their lived experiences and preferences were incorporated into the study design. Ultimately, a consensus emerged that the potential benefits of intensified surveillance—particularly the psychological reassurance it provides and the reduction in anxiety associated with less frequent follow-up—were judged to outweigh the associated risks. Moreover, we tried to minimize unnecessary imaging by integrating several biomarker-specific algorithms defining a maximum number of CT scans per time allowed and monitor fear and stress levels of the patients using the fear of progression questionnaire PA-F12.

Increasing the number of CT scans inevitably exposes patients to a higher cumulative radiation dose, thereby elevating the risk of various long-term complications, including, among others, radiation-induced secondary malignancies [33]. However, it is important to emphasize that the staging investigations in this study are not performed routinely but are instead triggered exclusively by abnormal findings in the liquid biopsy. This is therefore not a study based on regular imaging intervals. Rather, it is anticipated that, within a defined lead time, metastatic disease will become radiologically detectable following a pathological liquid biopsy result. A key objective of the study is to demonstrate that liquid biopsy can facilitate the early identification of patients with incipient metastatic disease, thereby enabling timely intervention—potentially with a curative intent.

There is evidence that early treatment of oligometastatic breast cancer with curative intention demonstrates no survival benefit in terms of OS or PFS [34]. While other studies involving different cancer types have demonstrated a survival benefit of additional radiotherapy targeting oligometastatic disease, they have also highlighted the potential risks, particularly treatment-associated morbidity and even mortality [35,36]. However, in the context of an increasingly dynamic therapeutic landscape—characterized by the continuous development of novel and more precisely targeted treatments—further investigation of this therapeutic concept remains highly relevant and warranted.

This trial is only partially blinded, as patients in the intensive surveillance arm with a positive liquid biopsy result are informed and advised to undergo staging imaging (patients in the intensive surveillance arm without any positive liquid biopsy result and all patients in the standard surveillance arm will remain blinded throughout the study). While blinding is important to reduce and prevent bias, one could argue that patients in the standard surveillance group might have a false sense of security, believing in having negative liquid biopsy results as no information to the patient regarding the results takes place. After extensive discussions with patients’ representatives during the phase of protocol development, we reached the consensus that this is an acceptable risk the patients should we willing to take, given both the detailed information regarding this aspect in the informed consent form and a thorough education and clarification during the enrollment process.

Another potential limitation of the study lies in the heterogeneity of the study population. Patients with different breast cancer subtypes—each associated with distinct biological behavior and recurrence risk—are included at various time points during follow-up. This variability may introduce confounding factors and limit the comparability of outcomes across subgroups, potentially affecting the generalizability of the findings.

In conclusion, the SURVIVE trial investigates whether a liquid biopsy guided intensified surveillance following primary therapy improves overall survival in breast cancer patients. It is currently the only recruiting study in this setting and extent that has the potential to revolutionize current breast cancer aftercare. Furthermore, the concept of initiating therapeutic interventions based on the detection of molecular disease prior to radiologically confirmed metastases is gaining increasing attention across various malignancies. For example, in colorectal cancer, ctDNA is being used to identify patients at high risk of relapse and to guide adjuvant therapy decisions, as demonstrated by the DYNAMIC trial [37]. In breast cancer, upcoming studies such as the SURVIVE HERoes trial (NCT06643585) and the TREAT ctDNA trial (NCT05512364) will investigate the efficacy of different agents such as Trastuzumab Deruxtecan and Elacestrant in patients with MRD detected by ctDNA, aiming to prevent the development of metastatic disease before it becomes visible on imaging. These promising approaches highlight the potential of molecular surveillance to transform early cancer management and improve patient prognosis. Should the study demonstrate a survival benefit, it could revolutionize breast cancer follow-up care and significantly improve the long-term prognosis for women diagnosed with breast cancer.