Systematic use of patient reported outcome during radiotherapy for head and neck cancer: study protocol for the national DAHANCA 38 trial

Introduction

Standard treatment in the curative setting of head and neck squamous cell carcinoma (HNSCC) includes radiotherapy, with or without surgery, and/or combined chemotherapy [1–5]. Irrespectively of treatment intensity, the patients suffer from multiple acute side effects including pain, dysphagia, mucositis and xerostomia [6–9]. This negatively affects the patients’ quality of life during radiotherapy [10,11]. The standard approach for assessing side effects are observer-based scoring systems [12–14]. It has been shown in Danish and US head and neck cancer populations, that there is a discrepancy between observer-based scoring systems and the patient’s subjective assessment [15,16]. Patient Reported Outcome (PRO) can be used as a tool to obtain the patients’ subjective symptoms with a systematic approach [17]. There is an increased focus on using PROs in head and neck cancer trials [18,19]. PRO can be used proactively, meaning that the clinicians actively reviews the patients’ PRO answers during therapy, and uses the feedback from patients to optimize the treatment e.g. management of side effects. Basch et al. have shown that patients with metastatic cancer receiving chemotherapy and proactive PRO counseling had improved Health-Related Quality of Life (HRQOL) [20]. In the radiotherapy setting, proactive PROs has the potential to become a relevant tool to detect acute side effects in order to improve management and avoid patient deterioration. This is it yet to be shown in clinical trials.

In Denmark all centers treating HNSCC are part of the multidisciplinary Danish Head and Neck Cancer Group (DAHANCA). The healthcare system is government-funded for all citizens. This means that the study population represents all patients despite socioeconomic background [21].

This paper describes the organization and methods behind the national DAHANCA 38 trial comparing systematic use of PRO during radiotherapy for HNSCC with standard clinical counseling. The hypothesis is that proactive use of PRO during radiotherapy will lead to improvements in the patients’ quality of life, based on more precise monitoring and management of side effects.

Material and methods

Patients

Inclusion criteria are age ≥18 years, a diagnosis of HNSCC, primary or post-operative curative radiotherapy with or without concomitant cisplatin, able to read and speak Danish, no serious cognitive deficits, and signed informed consent to participate in the study. Exclusion criteria is prior radiotherapy in the head and neck area, and proton therapy.

Design

The trial is designed as a prospective nation-wide, sequential cohort study, clinicaltrials.gov ID No. NCT03918382, DAHANCA 38 protocol version 1.3, 2^nd July 2019 [22], SPIRIT checklist, Supplementary material 1.

The study will be performed in a public health care system including all six centers treating HNSCC. Enrolled patients will receive radiotherapy ± cisplatin according to DAHANCA guidelines [23]. The intervention in the trial is proactive use of PRO. The sequential design consists of two phases, see Figure 1. In the first phase, the patients will be enrolled into the control group. The control group will receive standard clinical counseling according to current procedure. The second phase will be initiated when the last patient in the control group at the specific center has finished radiotherapy. In the second phase (the PRO group), the patients will receive standard clinical counseling plus proactive use of electronic PRO (ePRO) until 2 months after completion of radiotherapy, Table 1. Centers will be continuously accruing for 4 months starting in June 17^th, 2019 at center 1. Each center has committed to include eligible patients and are responsible identification log. A screening log will be held of eligible patients.

Figure 1. Timeline for the DAHANCA 38 trial.

Table 1. Overview of DAHANCA 38 assessment times, measures and tasks.

			Radiotherapy +/- chemotherapy
	Task/measures	Baseline	Week 1	Week 2	Week 3	Week 4	Week 5	Week 6	End of RT	14 days post RT	2 months post RT
All patients (control group and PRO group)	Out-patient visits	ᴏ	ᴏ	ᴏ	ᴏ	ᴏ	ᴏ	ᴏ	ᴏ	ᴏ	ᴏ
	Informed consent	ᴏ
	Background information^a	ᴏ
	EORTC QLQ-C30	ᴏ				ᴏ			ᴏ		ᴏ
	EQ-D5-L5	ᴏ				ᴏ			ᴏ		ᴏ
PRO group (only patients in the PRO group)	ePRO	/	/	/	/	/	/	/	/	/	/
	PREM								/
Clinicians (control group and PRO group)	DAHANCA toxicity score	x	x	x	x	x	x	x	x	x	x
	Case Report Form^b	x	x	x	x	x	x	x	x	x	x
	Off study registration										x

ePRO: electronic Patient Reported Outcome; EORTC QLQ-C30: European Organization for Research and Treatment of Cancer Quality of life; EQ-D5-L5: EuroQol Group’s five-level EuroQol five-dimensional questionnaire; PREM: Patient Reported Experience Measure; RT: radiotherapy.

^a Background information (education level, work situation, living situation).

^b CRF: Case Report Form (opioid treatment, feeding tube, hospital admissions).

Electronic PRO questionnaire

We have previously shown that four validated head and neck cancer-specific PRO questionnaires are equally preferred by the patients [24,25]. Based on these results and the recommendations from Chera et al., the ePRO questions consist of 33 head and neck relevant items from the Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE™) library, 7 head and neck specific items from the European Organization for Research and Treatment of Cancer Quality (EORTC) item library and free text fields for additional symptoms, Supplementary material 2 [18,26]. The generic clinical PRO system, AmbuFlex, will be used for collection of ePROs. AmbuFlex has been used for research purposes in clinical epidemiological studies since 2004 [27].

Trial flow

The patients have scheduled out-patient visits weekly during radiotherapy, two weeks and two months after completion of radiotherapy. At the scheduled out-patient visits, the patients are seen by a doctor and/or a nurse (standard clinical counseling). The number of visits depends on the individual patient´s treatment plan but is in general between 8-10 times from baseline until two months after completion of radiotherapy. The patients will go off study two months after radiotherapy and continue standard follow-up, Table 1.

In the standard consultation, the patients’ side effects are discussed and graded (DAHANCA toxicity score) [14] by the clinician, and a focused physical examination is performed. The toxicity score and objective findings are reported to the DAHANCA database. The DAHANCA database contains information about tumour-related conditions, treatment, and demographic patient data.

All questionnaires in the trial are in Danish. At baseline (before start of radiotherapy), all participants complete a paper questionnaire about their level of education, their current work situation and their living situation. The EORTC QLQ-C30 [28] and EuroQol Group’s five-level EuroQol five-dimensional questionnaire (EQ-5D-5L) [29] will be completed by both groups at baseline, week four of treatment, at completion of radiotherapy and two months after completion of radiotherapy. These two PRO questionnaires will be used as passive collection of PRO in relation to our predefined outcomes and are completed on paper.

A case report form will be completed by the clinician after each scheduled out-patient visit to ensure endpoint data are captured.

Patients in the PRO group will complete the ePRO at baseline, weekly during radiotherapy until end of radiotherapy and two weeks and two months after radiotherapy completion. The patients will complete the ePRO on a tablet available in the waiting room before their consultation or at home. It is mandatory for the clinicians to review the ePRO answers prior to or during the consultation.

Patients in the PRO group will receive a Patient Reported Experience Measure (PREM), which is a feedback form to evaluate patient satisfaction with the ePROs [30,31].

Endpoints

Clinical data from all patients will be collected and will include age, stage, treatment regime, information on hospital admissions, treatment adjustments, and opioid treatment. The statistical analysis will include descriptive statistics of the two groups regarding number of enrolled patients, age, stage, treatment regime. A statistical analysis plan is available for details on planned data acquisition and analysis [32]. Data will be captured in RedCap. The primary endpoint is difference in global HRQOL at the end of radiotherapy using the EORTC QLQ-C30 tool. Secondary and explorative endpoints are DAHANCA toxicity score [14], time to start opioid treatment, time to tube-feeding/other feeding, weight loss, compliance to treatment (radiotherapy, chemotherapy, nimorazole), hospitalization due to toxicity except tube-feeding, quality-adjusted life years (QALYs), and PREM.

DAHANCA toxicity score is an objective grading of the patient’s symptoms by the clinician. It employs a categorical scale 0-4 in which 0 is no/nothing and 4 is severe. Differences between DAHANCA toxicity score in the PRO group and ePRO scores will be reported. Differences in time to opioid treatment, time to time to tube-feeding/other feeding, weight loss, number of hospitalizations reported and number of patients completing planned treatment between the control group and the PRO group will be reported. The EQ-5D-5L will be used to measure QALYs. The PREM questionnaire at end of radiotherapy covers the patient’s experience of the toxicity reporting, usefulness of the electronic PRO tool to cover symptoms and the patient’s experience of the communication with doctor/nurses and handling of the patient’s side effects [31]. Descriptive statistics of QLQ-C30 domains will be performed.

Missing data will be handled according to EORTCs recommendations [33]. We will investigate if data is missing at random and whenever possible cross-reference with national registries to acquire insights toward missing data.

Sample size calculation

The sample size calculation is based on the primary research question: difference in HRQOL at end of radiotherapy by EORTC QLQ-C30 (scale range 0-100, a higher score indicating better quality of life). We assume that the response is normally distributed with a standard deviation of 20 [10,11]. There will be two patients in the PRO group per patient in the control group. If the true difference in the PRO group and the control group means is 7, we will need 194 patients in the PRO group and 97 patients in the control group to yield 80% power to detect this assumed difference, using a two-sided test and a significance level of 0.05. Centers enrolling less than five patients in the control group will be excluded from the analysis. The sample size was determined from PS software (http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize).

Conclusion

This is the first national prospective study investigating proactive use of PRO during radiotherapy in a head and neck cancer population. The study will generate unique detailed information on quality of life, the patient´s perception of frequencies, severity and the development over time of acute side effects from radiotherapy. The study will explore if a systematic approach to the patients’ experience of symptoms will lead to optimized handling of side effects and better quality of life emphasizing the value of a more personalized care.

Ethical approval

The patients will receive verbal and written information and must give written informed consent which they can withdraw at any time.

The study will follow General Data Protection Regulation (GDPR) and is registered at the Capital Region of Denmark (ID. P-2019-24) and the Central Denmark Region (ID. 1-16-02-175-19). In a Danish setting the project does not need approval from the National Committee on Health Research.

Disclosure statement

Varian Medical Systems has funded part of the salary for CHM, but do not have influence on the scientific work.

Additional information

Funding

This work was supported by The Health Foundation; under Grant number 17-B-0257; Danish Comprehensive Cancer Center Radiotherapy and Danish Cancer Society under Grant number 9511-WP7 RH; Faculty of Health and Medical Sciences, University of Copenhagen under Grant number 18-10, Varian Medical Systems and The Oncological Research Fund Rigshospitalet. CJ is supported by the Danish Cancer Society Research Center.