Protocol of the VICTORIA study: personalized vitamin D supplementation for reducing or preventing fatigue and enhancing quality of life of patients with colorectal tumor - randomized intervention trial

The primary objective of this trial is to test whether a personalized vitamin D₃ dosing regimen reduces or prevents fatigue (primary outcome) among stage I-III CRC patients with season-adapted vitamin D insufficiency or deficiency.

The secondary objectives are to assess the efficacy of the treatment on sub-domains of fatigue (physical, emotional and cognitive), QoL (overall and domain-specific), probable depression, functional well-being (FWB), infection frequencies, and various biomarkers (25(OH)D, HbA_1c, white blood cell count, leukocyte subtype counts, serum c-reactive protein (CRP), uric acid, creatinine, triglycerides, total, low-density lipoprotein (LDL) and high-density lipoprotein (HDL) cholesterol). The study further aims to evaluate the safety of the treatment regarding the occurrence of hypervitaminosis D, hypercalcemia, hypercalciuria, and renal dysfunction.

The trial is conducted as a national multicenter, randomized, double-blind, and placebo-controlled phase III trial using a confirmatory approach. By applying a parallel group design, the participants are randomly assigned in a 1:1 ratio to either the intervention group or to the control group. Recruitment is carried out in three large German rehabilitation clinics and is assumed to take approximately 24 months. The list of trial sites can be obtained from the German Clinical Trials Register (DRKS00019907).

The trial protocol and this manuscript have been developed in line with the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guideline [17].

The trial outcomes have been chosen in order to determine the efficacy and the safety of the intervention.

The primary outcome fatigue will be evaluated by using the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F) fatigue subscale, version 4.0, a commonly used and well-validated measure of fatigue in people with cancer and other chronic health conditions [18]. The tool assesses self-reported tiredness, weakness, and difficulty conducting common activities due to fatigue. Higher scores represent less fatigue. The primary endpoint will be measured as the mean difference in FACIT-F fatigue subscale between intervention and placebo group at trial week 13–16. A mean difference of ≥3 FACIT-F fatigue subscale points will be considered a clinically relevant difference [19]. Additionally, the mean difference in change of FACIT-F fatigue subscale from baseline to trial week 13–16 between intervention and placebo group will be determined as a secondary endpoint.

Subdomains of fatigue will be evaluated using the European Organization for Research and Treatment of Cancer – Cancer related Fatigue (EORTC QLQ-FA12) questionnaire. This is an alternative to the FACIT-F fatigue subscale without being a global fatigue assessment tool [20, 21], with higher scores representing more severe fatigue. Items which are relevant for the VICTORIA study include physical, emotional and cognitive fatigue domains. They will be assessed as the mean differences in EORTC QLQ-FA12 physical, emotional and cognitive fatigue scores between intervention and placebo group at trial week 13–16 as well as mean differences in changes in these scores from baseline to trial week 13–16 between intervention and placebo group.

Quality of life will be determined with the European Organization for Research and Treatment of Cancer – Core Quality of life questionnaire with 30 items (EORTC-QLQ-C30), version 3.0. The questionnaire is used to gauge the overall and domain-specific QoL in cancer patients [22]. Higher scores represent better functioning. Items which are relevant for the VICTORIA study include the five functional scales (assessing physical, role, emotional, cognitive, and social functioning) and one global health status/QoL scale. Endpoints are the mean differences in overall and domain specific quality of life scores of the EORTC QLQ-C30 questionnaire between intervention and placebo group at trial week 13–16 as well as mean differences in changes in these scores from baseline to trial week 13–16 between intervention and placebo group. Mean differences ≥5 points in the overall and domain specific scores of the EORTC QLQ-C30 will be considered clinically relevant differences [23].

Probable depression will be ascertained with the Geriatric depression scale (GDS-15) which has been developed for use among older adults [24, 25]. The focus on the elderly is crucial for this trial as the mean age of the CRC patients to be included is expected to be approximately 65 years. An overall score ≥ 5 points is considered as probable depression [25]. Endpoint will be the mean difference in GDS-15 scale between intervention and placebo group at trial week 13–16 as well as the mean difference in changes in this scale from baseline to trial week 13–16 between intervention and placebo group.

The FACIT-F FWB subscale will be used to assess the FWB by querying about limitations in the ability to work, coping with the disease, and general life satisfaction [18]. Higher scores represent better well-being. Endpoint will be the mean difference in the FACIT-F FWB score between intervention and placebo group at trial week 13–16 as well as the mean difference in changes in this score from baseline to trial week 13–16 between intervention and placebo group.

Finally, serum und urine parameters will be measured at every sample collection to evaluate the following secondary endpoints:

For the verum, 20,000 IU (Dekristol® 20,000 I.E., Mibe GmbH Arzneimittel) or 2000 IU colecalciferol (two tablets Dekristol® 1000 I.E., Mibe Arzneimittel GmbH) are encapsulated and filled up with a mixture of mannitol and colloidal silicon dioxide (Füllstoff DAC). The placebo consists solely of the hard gelatine capsules with filler. To ensure blinding, verum and placebo capsules are indistinguishable regarding appearance, quantity, weight, and packaging.

Patients, who have consented to participate in the trial, are randomly allocated to either vitamin D₃ or placebo for a total of 12 weeks (84 days).

The study flow is depicted in Fig. 1.

In total, 456 eligible CRC patients are being randomized in order to analyze 410 for the primary endpoint after an expected drop-out rate of 10%. The trial schedule is presented in Table 3.

During a first informative appointment, patients are briefed about the study and receive the printed study information. After having enough time to read the information materials, the potential study participants have the opportunity to ask questions before the official screening visit commences.

Data handling will be described in the trial data management plan. The Coordination Center for Clinical Trials, Heidelberg (KKS) provides and monitors a validated electronic case report form (eCRF) for remote data entry. The eCRF will be used for all data entry by study centers and the coordination center. All entries in the eCRF will be verifiable by source documents (if not explicitly exempt from this rule) and include:

All patient data are documented pseudonymously. All documents and data pertaining to the trial will be stored at the study centers in the investigator site file (ISF) and at the coordinating center in the trial master file for at least 10 years after the end of the trial in accordance with the applicable regulations.

Written informed consent will be obtained for each participant before enrollment in the trial and will be kept in the ISF at the respective study centers. The investigators will enter the date and quantities of distributed and returned study medication as well as all participant data collected during rehabilitation into the eCRF, except biomarker results from the second blood and urine sample collection to avoid unblinding. All data entries in the eCRF will undergo an automatic online check for plausibility and consistency, which is defined in the data validation plan. In case of implausibility, a warning message will be produced during data entry. A responsible investigator or a designated representative will be obliged either to correct the implausible data or to confirm its authenticity and give an appropriate explanation. The responsible data manager will check all explanations and will resolve the warning if the explanation is appropriate.

The responsible investigator has to confirm the correctness of all entries in the eCRF with a dated electronic signature. The time points and frequency are pre-defined in the eCRF specification.

All missing data or inconsistencies will be reported back to the study centers and have to be clarified by the responsible investigator before database lock. If no further corrections are to be made in the database, it will be declared locked and can be used for statistical analyses.

The coordinating center enters the pseudonymized laboratory results of the second and the final sample collection as well as study data from questionnaires and patient diaries in the eCRF. To ensure the integrity of the data, staff members of the coordinating center will have clearly defined roles and tasks which will be assigned via a delegation log. For example, coordinating center staff members who contact the study participants will not have access to the eCRF, the laboratory results or the sealed envelopes containing the information on the allocation of treatment. Staff members who open these envelopes or check the pseudonymized laboratory results of the study participants, in turn, will not have access to any patient-identifying data or other study data.

The data management will check the completeness, the validity, and the plausibility of the entered data using validation programs, which will generate queries where applicable. The head of the coordinating center or a designated representative will confirm the correctness of all entries in the eCRF by a dated signature as defined in the eCRF specification.

Each patient will complete a questionnaire in the rehabilitation clinic at baseline and in study weeks 13 to 16. The questionnaires will address lifestyle factors (smoking, alcohol consumption, diet, physical activity), CRC therapy (chemotherapy, radiotherapy, operation), and medical history (common diseases, family history of diseases) as well as the (validated) tools for assessment of the outcomes.

Patients will be asked to provide a blood and a spontaneous urine sample upon screening (baseline), at the end of rehabilitation (day 12–21) and once again in the study weeks 13 to 16. The first two samples will be collected in the rehabilitation clinic (inpatient) while the final sample will be collected by the patient’s GP (outpatient). For the sampling at the GP’s office, the coordinating center will send pseudonymized sampling kits to the participants after trial week 12. One part of the samples will be used to determine biomarkers immediately (as efficacy or safety outcomes). Another part will be stored at the coordinating center for potential future post-hoc analyses with novel biomarkers.

All blood samples drawn during the trial will be either collected in clotting activator containing tubes to analyze serum or in EDTA tubes to analyze whole blood. The urine samples will be donated in a collection cup of 40 ml from which a volume of 10 ml is extracted using Urin-Monovette®, Luer, Germany. Any residual volume will be discarded. The quantity of collected blood and urine samples is shown in Table 4.

To comply with the pre-analytical requirements, it is ensured that the transport time from sampling to analysis is less than 48 h and the temperature of the samples during transport is between 2 and 25 °C [36]. Serum tubes are inverted 3–5 times immediately after withdrawal and then stored at room temperature for 30 to 60 min to allow blood clotting. Tubes are subsequently centrifuged at 2500–3500 rpm for 10 min. EDTA tubes will be inverted eight to 10 times to ensure an even mixture of EDTA and blood.

The laboratories cooperating with the rehabilitation clinics will measure all parameters with standard state-of-the-art lab methods.

Participants will be asked about concomitant medication and diseases before inclusion. To minimize the risk of AEs and exacerbation of present conditions, patients will only be included if conditions that are listed as contraindications in the Summary of Product Characteristics of colecalciferol 20,000 IU or require special safety monitoring are not present (see exclusion criteria, Table 1). In the patient information document, patients will be informed about potential risks associated with the trial participation and instructed to contact the trial physician in case of serious medical problems. The observation period of AEs will begin with the first administration of the trial medication and end with its last administration. Events happening before the first administration are defined as medical history. AEs will be queried during every visit with the responsible investigator and additionally within the patient diary during the entire intervention phase. All AEs will be documented in the eCRF stating the participant’s randomization number, the start and end date, a description, the intensity, the seriousness, the relationship with the study medication, the measures taken, and the outcome.

All serious adverse events (SAEs) must be reported by the investigator to the pharmacovigilance department of KKS using a standardized form within 24 h after initial observation or awareness of the event. Every SAE will be subject to a second assessment by a designated person, who will be independent of the reporting investigator. All SAEs and their relevance for the risk-benefit assessment of the trial as well as the final report will be evaluated continuously during the trial.

A steering committee will be convened to ensure the ethical conduct of the trial and protect the rights as well as the welfare of the patients. The board consists of the coordinating investigator, the head of the coordinating center, the deputy head of the coordinating center, and the clinical pharmacology consultant. By periodically assessing the safety of the intervention and reviewing potential safety issues, amendments to the further trial conduct (modification, continuation, closure) will be decided and documented.

Internal standard operating procedures will be followed for the preparation, implementation, documentation, and the analyses of the clinical trial. All applicable regulations will be followed. The pharmacy of the University Hospital Heidelberg is holding a manufacturing license and is therefore authorized to produce the required study medication. Packing, labeling, and blinding will take place according to the applicable GCP and Good Manufacturing Practice regulations and standards.

As required by the German Drug Law (AMG) for multicenter trials, the sponsor has appointed a coordinating investigator (German: Leiter der klinischen Prüfung). Moreover, every trial site has appointed one principal investigator and at least one deputy investigator. All trial physicians and trial centers comply with the qualification requirements of the responsible Ethics Committee in terms of professional education, experience in clinical trials and equipment. Since the trial medication is well characterized, of low risk, and tested in a non-critical indication (fatigue), a Data Monitoring Committee has not been set up [37].

All data obtained in the course of the trial will be treated pursuant to the German Federal Data Protection Act (BDSG) and the European ordinance (EU) 2016/679. The individual participants will be exclusively identified by their patient identification and randomization numbers. The investigators will provide direct access to source data/documents for trial-related monitoring, audits, and regulatory inspection. Each participant will have to consent via written informed consent to direct access to their original medical records for trial-related monitoring, audit, and regulatory inspection.

Qualified staff will regularly monitor every study site by reviewing source documents, entries into the eCRFs, and essential documents to ensure that the trial complies with the protocol and regulatory requirements. Monitoring includes an on-site initiation visit, regular on- and off-site visits during the recruitment phase and a close-out visit. Before the study start, the participating sites will be personally trained and introduced to all study specific procedures during the on-site initiation visits. After each visit, the monitor will prepare a report for the sponsor and a follow-up letter with findings and eventual necessary measures for the sites. All procedures will be pre-defined in the monitoring manual.

All planned substantial changes of the clinical trial need to be signed by the sponsor, the coordinating investigator, the biometrician and the clinical pharmacology consultant. According to §10 of the GCP ordinance (GCP-V), protocol amendments are submitted in writing to the responsible Ethics Committee and the national competent authority.

All analyses will be done using SAS software version 9.4 or higher. Homogeneity of the treatment groups will be described by comparison of the demographic data and the baseline values of key variables. All statistical tests will have a two-sided significance level of 0.05. Besides, 95% confidence intervals will be estimated for all outcomes in the placebo and verum group.

The primary analysis will test the null hypothesis.

H₀: The FACIT-F fatigue subscale at week 13–16 is the same in the two groups.

versus the alternative hypothesis.

H₁: The FACIT-F fatigue subscale at week 13–16 is different in the two groups.

The primary endpoint FACIT-F fatigue subscale will be analyzed with an intention-to-treat approach. The intention-to-treat population will include all randomized study participants with data for the primary endpoint and who did not withdraw consent during the trial. A per-protocol analysis will be done additionally as a sensitivity analysis. The per-protocol population will exclude study participants who meet the following criteria:

We will test on normal distribution for the outcome FACIT-F fatigue subscale with the Shapiro-Wilks test. If normal distribution can be assumed, the primary test statistic will be the two-sample t-test for the mean difference. If normal distribution cannot be assumed for the FACIT-F fatigue subscale, we will log-transform the FACIT-F fatigue subscale and test for normal distribution once more with the Shapiro-Wilks test. If normal distribution can neither be assumed for FACIT-F fatigue subscale nor log-transformed FACIT-F fatigue subscale, a Wilcoxon Rank-Sum test will be applied. In addition, appropriate 95% confidence intervals will be estimated for the FACIT-F fatigue subscale in the two groups and for the difference in FACIT-F fatigue subscale between the two groups.

A priori defined subgroup analyses will be conducted for groups defined by age (< 65 / ≥ 65 years), sex (male / female), CRC stage (I or II / III), 25(OH)D levels at screening (< 30 / ≥ 30 nmol/L), season at screening (Dec-Feb/Mar-Mai/Jun-Aug/Sept-Nov), BMI at screening (< 30 / ≥ 30 kg/m²), FACIT-F fatigue subscale at screening (≤ 34 / > 34 points), mild to moderate anemia at screening (hemoglobin 8–10 mg/dl / > 10 mg/dl), chemotherapy and/or radiotherapy in 9 months before screening (yes / no), chemotherapy and/or radiotherapy during trial (yes / no), geriatric depression scale (GDS-15) score at follow-up (< 5 / ≥ 5 points), insomnia at follow-up (EORTC QLQ-C30 insomnia item < 3 / ≥ 3 points), pain at follow-up (EORTC QLQ-C30 pain scale < 6 / ≥ 6 points), use of strong opioids (ATC codes N02AB03, N02AA01, N02AG01, N02AA51, N02AA03, N02AG04, N02AA53, N02AA05, N02AJ18, N02AJ19, N02AA55, N02AA56, N02AJ17, N02AX06, N02AE01, and N02AA25) at follow-up (yes / no), use of psycholeptics (ATC code N05) at follow-up (yes / no), and use of corticosteroids for systemic use (ATC code H02) at follow-up (yes / no).

In addition, a subgroup analysis will be conducted for study participants, who most likely profit from the vitamin D₃ intervention because the following conditions apply combined (yes / no): No protocol deviations (patients did not withdraw consent during the trial, were adherent to the trial medication (defined as taking at least 80% of all capsules), and did not take vitamin D₃ or vitamin D analogs in addition to the trial medication), FACIT-F fatigue subscale ≤34 at screening, and neither use of strong opioids nor psycholeptics at follow-up.

A mean difference ≥ 3 FACIT-F fatigue subscale points will be considered a clinically relevant difference [19].

Depending on the availability of future funding for genotyping of all randomized trial participants, further subgroup analyses are planned that stratify trial participants by genetic susceptibility for low 25(OH)D levels or cancer related fatigue [38, 39].

As a sensitivity analysis, a linear regression model will be carried out with FACIT-F fatigue subscale as the dependent variable and treatment group as the independent variable. The linear regression model will be adjusted for all variables named in the subgroup analysis while using some variables continuously instead of categorically (age, hemoglobin levels at screening, 25(OH)D levels at screening, BMI at screening, FACIT-F fatigue subscale at screening, insomnia scale at follow-up and pain scale at follow-up). If appropriate, missing covariate values will be imputed with multiple imputation.

All secondary endpoints with a continuous scale will be analyzed with the same statistical methods as described for the primary endpoint. All dichotomous secondary endpoints will be tested with a Chi² test. AEs will be analyzed with descriptive statistics including frequencies of SAEs.

Lastly, the success of the personalized vitamin D₃ intervention in raising the serum 25(OH)D level will be evaluated in the intervention group at trial day 12–21. Furthermore, the success of the maintenance dose to maintain sufficient vitamin D status in the intervention group will be evaluated in trial week 13–16. To be successful, the mean 25(OH)D levels in the group with personalized vitamin D₃ intervention should be higher than 50 nmol/L. This translates into the following test hypotheses:

To test these hypotheses, we will conduct a one-sample t-test on the mean 25(OH)D levels at a one-sided significance level of 0.025. 25(OH)D levels may be log-transformed if this improves the approximation of a normal distribution (to be tested with a Shapiro-Wilks test).

Exploratory sub-project: “Determinants of the achieved 25(OH)D levels by the personalized vitamin D ₃ intervention”

In the following, we describe the statistical methods of an observational sub-project for the secondary outcome “25(OH)D level”. This solely observational research project shall be addressed in those 228 VICTORIA trial participants who received the personalized vitamin D₃ intervention and adhered to the trial medication (defined by taking at least 80% of the trial medication capsules), which will result in a final sample size n < 228. If 25(OH)D levels or log-transformed 25(OH)D levels are normally distributed (to be tested with a Shapiro-Wilks test), a linear regression model will be carried out with a continuous 25(OH)D level variable. In addition, a logistic regression model will be used with a dichotomized 25(OH)D level variable (< 50 nmol/L / ≥ 50 nmol/L) as the dependent variable.

Covariates for the linear and logistic regression model will be study center, age, sex, baseline 25(OH)D, compliance up to time of blood sampling, difference between exactly calculated and rounded supplied loading dose, number of days between last intake of trial medication and blood sampling, season at recruitment, baseline BMI, baseline waist circumference, baseline smoking, baseline physical activity, baseline physical functioning (EORTC QLQ-C30 subscale), baseline Charlson Comorbidity index, baseline frailty, baseline FACIT-F fatigue subscale, GDS-15 total score ≥ 5 points (at baseline), baseline anxiety (GAD-7 score), baseline pain (EORTC QLQ-C30 pain scale), CRC stage, time since CRC tumor surgery, stoma, concomitant use of vitamin D products (at follow-up), chemotherapy and/or radiotherapy in 9 months before trial (yes / no), chemotherapy and/or radiotherapy during trial (yes / no), concomitant use of laxatives (at baseline and/or follow-up), concomitant use of other drugs limiting vitamin D₃ bioavailability at baseline and/or follow-up (phenytoin, barbiturates, systemic glucocorticoids, rifampicin, isoniazid, cholestyramine, orlistat, dactinomycin or systemic azole-antimycotics), nutritional vitamin D intake (at follow-up), appetite loss (at baseline and/or follow-up), sun exposure in last summer before baseline, skin type, solarium use in last 2 months before baseline, frequency of diarrhea (at baseline and/or follow-up), vitamin D binding protein levels and single nucleotide polymorphisms (SNPs) previously shown to be associated with low 25(OH)D levels. If appropriate, missing covariate values will be imputed with multiple imputation.

The analyses will be carried out both for the 25(OH)D measurement from day 12–21 (end of loading dose consumption) and for the 25(OH)D measurement from week 13–16 (end of maintenance dose consumption).

At the time of submission, recruitment for the VICTORIA study is planned to start in summer 2020.