Cancer and Anesthesia: Survival After Radical Surgery - a Comparison Between Propofol or Sevoflurane Anesthesia

The study will be completed in accordance with Declaration of Helsinki, Good Clinical Practice (GCP) and applicable regulatory requirements. The study will include approximately 8,000 pts from about 10 sites in Sweden and probably also from one or two international sites. The population will be pts that are scheduled for radical breast- or colorectal cancer surgery in general anesthesia with or without additional regional anesthesia.

A clinically relevant absolute difference in five-year-survival would be 5%. Based on data from our retrospective study, we will have 80% power to detect a difference of 5% with a P-value of <0.05 if including 7,378 pts (see below). By adding 8.4% more to the inclusion, we will have a reasonable safety margin for loss of pts or technical errors.

Breast cancer, expected 5-year survival of 87% vs 82%: 1,650 pts Colon cancer, expected 5-year survival of 60% vs 55%: 3,000 pts Rectal cancer, expected 5-year survival of 70% vs 65%: 2,728 pts Margin for loss of patients or technical errors 622 pts Total 8,000 pts.

During the preoperative anesthetic procedures, the pts will be screened and consented for the study. Demographic data will be collected in connection with screening/enrolment. The pts will be randomized to either propofol- or sevoflurane anesthesia.

The only study specific procedure for this protocol is the randomization. The products used in this study are well established and registered since many years. The manufacturers of the products used are responsible for safety monitoring for pharmacovigilance purposes according to regulations.

It is the responsibility of the Investigator at each site that all pts, considered as candidates for the study, are listed in the "Screening and Enrolment log". Pts will receive a consecutive screening number when signing the informed consent and are thereafter considered to be study patients. Once they have been randomized, they will also get a patient-specific randomization number and they are considered as enrolled in the study. It is the responsibility of the investigator to keep a patient identification list, identifying each individual study patient.

Data will be collected from the normal clinic (demographics, anesthesia- and surgical related data) and the quality registries used within oncology (tumor-specific and survival data) respectively. The source of data will be the pts' Medical Records.

The following variables noted during the care of the pts will be recorded:

• Demographic variables including habits and general health
• Age, Gender
• Place of residence, Zip code (proxy for socioeconomics)
• Length, Weight
• Smoking (pack years; 1 pack year = 20 cigarettes/day for a year) and Alcohol (standard drinks; 1 standard drink = 12 g pure ethanol)
• ASA-grades (American Society of Anesthesiologists)
• Co-morbidity
• On-going medication
• Other anesthetics last year (date, duration, type of)
• Study specific variable
• Randomization, i.e. allocation to both propofol- or sevoflurane-group and confirmation of hypnotic product used during surgery
• Anesthesia related variables
• Duration
• Doses of intraoperative opioids during current anesthesia
• Other adjuvant intraoperative treatment such as inotropic drugs
• Accumulated time with mean arterial pressure (MAP) under 65 or over 130 mmHg
• Accumulated hydration balance
• Bleeding volume
• Transfusions (red blood cells, plasma)
• Pre-and postoperative laboratory-analyses
• Complementary regional blockade (kind of block, type and doses of local anesthetics)
• Doses of postoperative morphine
• Re-anesthesia (type of)
• Surgical variables
• Cancer location
• Date of surgery
• Duration
• The following complications will be collected up to 30 days post-surgery
• Surgical complications (intra- and postoperative) that receive an International Classification of Diseases number (ICD)
• Occurrence of following diagnoses during 30 days after surgery: myocardial infarction, stroke, pulmonary embolism, pneumonia and renal failure

The following tumor specific variables will be transferred from the cancer quality registries

• Tumor data (stage, proliferation, hormone status)
• Complementary therapy (radiation, chemotherapy, anti-hormone, antibody, angiogenesis inhibitors)
• Recurrence/metastasis (localization, time point)
• Re-surgery (date, duration, indication)
• Date of death and cause of death, when applicable

No coding will be performed as complications will be added into the database with ICD numbers.

Uppsala Clinical Research Center (UCR) will be responsible for the Data Management of the clinical database and will write a study specific Data Management Plan (DMP) where further details will be specified. These activities will be made in accordance with the Standard Operation Procedures (SOP) at UCR.

An electronic CRF (eCRF) will be used in this study which also will serve as the clinical database for the study. UCR will be responsible for set-up, support and management of this eCRF. All above mentioned data will be collected in this database. This system will also include handling of Data Clarification Forms (DCFs) to resolve any inconsistencies detected by the quality control procedures.

In accordance with the principles of GCP, monitoring of the study will be arranged by the Sponsor. UCR has been appointed by the Sponsor to monitor this study. During the study, the Monitor will have regular contacts with the study sites, including visits to ensure that the study is conducted and documented properly in compliance with the protocol, GCP and applicable regulatory requirements. The Monitoring activities will be made in accordance with the SOP at UCR. Both centralized monitoring activities and on-site monitoring activities will be used in this project. It has been decided that the major part of the monitoring will be made by centralized methods together with continuous telephone contacts with site but with limited on-site activities. This decision is based on the nature of the study - observational with no other study specific procedures than the randomization. Additionally, the investigational products used are well established within clinical practice since many years and with documented safety profile. Each site will have a site initiation (can be organized as multi-center initiation meeting or web-based training) with focus on providing information about study objectives, study procedures and eCRF-training. Each site will then have at least one on-site visit during the study. Based on the outcome of centralized monitoring activities, the frequency of on-site monitoring activities may be increased. The Monitor will review source documents for verification of consistency with the data recorded in the eCRFs for a random sample of pts. In addition, selected key variables will be proofread. All inconsistencies detected during these procedures will be resolved through the DCF's, being issued to the monitor or investigational site personnel. Completed DCFs will be filed and entered into the database accordingly. The Monitor will also provide information and support to the Investigators. The centralized monitoring will be used to detect each site´s level of compliance in terms of completion of eCRF generally but also with focus on critical data (e.g. compliance in randomization to propofol or sevoflurane and date of surgery). The outcome of the statistical checks that will be used will indicate whether increased monitoring activities are needed on a specific site. In terms of review of completed informed consent, only a random sample of the pts will be checked by the monitor in connection with on-site visit. Site will log all completed informed consents on on-going basis. This log needs to be distributed to the monitor on on-going basis and this will serve as monitoring tool of the consent-process. If there are signs of that the site does not comply with the specified consent-process, extended site-specific training/ increased monitoring activities will be needed. All monitoring procedures, including both centralized methods and on-site activities, will be described in the Monitoring Plan which will be compiled in collaboration with the Sponsor. The study center may also be subject to quality assurance audit by the Sponsor as well as inspection by the Medical Procurement Agency. The Investigator and other responsible personnel must be available during the monitoring visits, audits and inspections and should devote sufficient time to these processes.

Generally, the Medical Records/laboratory reports/anesthetic reports will serve as source data. CRF can also be used as source data. A site-specific Origin of Source Data-log will be compiled at each site to verify the source of data.

The following minimum amount of information should be recorded in the hospital records:

• Clinical study number.
• Subject identification.
• Date when patient information was given and when signed Informed Consent was obtained.
• Diagnosis.
• Fulfillment of eligibility criteria.

Tumor-specific variables including data about survival will not be part of this database but will be transferred from the quality registries used within oncology. These quality registries are routinely used within oncology as part of the quality assurance of the care of these patients (the INCA-platform is used). The registries are handled by six Regional Cancer Centers (RCCs). Complete oncologic- and outcome data will be available for all types of cancer included in the study.

In order to be able to link the information in these databases, the pts' full personal identification number (ID) will be collected and noted. It will be deleted after data cleaning and merge of the databases to ensure anonymity for each person. All analysis will be made after ID removal.

When all pts have been completed (i.e. completion of follow-up 30 days post-surgery), all data have been entered into the database, and all queries solved, the Database Closure procedures will start. Decisions will be made how to classify pts into analysis populations, and how to handle protocol violations and deviating or missing data. All decisions will be dated and documented in a Database Closure document. After that the database will be locked. Any changes in the database thereafter will be documented.

Statistical analysis The clinical database will be transferred to the Sponsor by UCR. Tumor-specific data will be transferred to the Sponsor from the RCC quality registers. Sponsor will then be responsible for linking these databases and will also do the statistical analysis.

In the present study all endpoints will be evaluated by descriptive methods. All variables will be presented as aggregated data. Categorical variables will be summarized in frequency tables (presenting frequencies and proportions) by type of anesthesia. The quantitative variables will be summarized by number of observations (n), mean, standard deviation (SD), median, minimum (min) and maximum (max) by anesthesia. If applicable, separate summaries will be presented for patients receiving one, two or more anesthesia, as well as pooled summaries. Graphical methods may be used wherever it is regarded as appropriate.

Overall survival and time to progress will be presented as Kaplan-Meier curves together with median survival time and time to progression, respectively. The estimates for cumulative 1- and 5-year overall survival will be compared between pts given sevoflurane or propofol. In a next step, Cox Proportional hazard models will be calculated to assess the risk of death adjusted for potential effect modifiers and confounders. There will also be stratifications for different types of tumors, cardio-pulmonary status, ASA-class etc. Hazard ratios with 95 % confidence intervals will be presented for all models.