1 INTRODUCTION

Head and neck cancer (HNC) affects almost 900 000 individuals worldwide each year.¹ The heterogeneity in patient characteristics and tumor features and locations presents challenges to developing standardized, consensus-based treatment guidelines for HNC. Instead, highly individualized therapy programs are devised for patients with HNC by specialized interdisciplinary health care teams to achieve a singular goal: complete resection or destruction of tumor tissue with the highest possible preservation of existing anatomical structures and function as well as patient quality of life (QoL). To maximize their QoL when multiple therapeutic options are feasible, patients can co-determine their therapy program through shared decision-making, during which the patient and health care providers discuss treatment options and mutually decide upon the best course of action.²

Although shared decision-making, when possible, provides patients with a certain freedom of choice, the partial responsibility of choosing their own therapy program can lead to more regret upon reflection on their decision. As decision regret is associated with worse QoL after therapy,¹¹ it is vital to identify patient, tumor, and treatment characteristics and post-therapy symptoms that influence QoL to reduce the extent of decision regret that may be experienced.³

Reducing the extent of decision regret experienced by patients with HNC after radiotherapy could be achieved through different approaches. One approach is to reduce late treatment-related toxicity by using more sophisticated therapeutic technologies and considering different treatment modalities.⁴ In particular, techniques such as image-guided radiotherapy, intensity-modulated radiotherapy, forward and adaptive planning, and helical tomotherapy can reduce the intensity of radiation rays in peritumoral healthy tissue.^5-9 Indeed, one study shows that reductions in radiotherapy dose to organs-at-risk decreases dry mouth in patients with various types of HNC and improves QoL in patients with oropharynx cancer.¹⁰ Another approach is to select specific therapy parameters during the planning phase or provide different types of post-therapy care based on patient, tumor, and treatment characteristics that are predictive of post-therapy decision regret or reduced QoL. For instance, patient reports of “feeling sad,” “difficulty swallowing,” and “pain” as well as factors including tumor classification, particular therapeutic combinations, smoking status at diagnosis, and M. D. Anderson Symptom Inventory for head and neck cancer (MDASI-HN) symptom scores are associated with decision regret among patients with oropharyngeal cancer.¹¹ However, factors such as total radiation dose, single fraction dose, radiotherapy boost, or patient comorbidities were not included in the analysis.

To further inform the development of approaches to reduce the extent of decision regret experienced by patients with HNC after radiotherapy, the aim of this study was to identify patient, tumor, and treatment characteristics and post-therapy symptoms associated with decision regret among patients with HNC who underwent surgery and adjuvant radiotherapy.

2 MATERIALS AND METHODS

2.2 Patient selection

Patients who underwent surgery and adjuvant intensity-modulated radiotherapy with image guidance using helical tomotherapy for HNC between 1 January 2018 and 25 February 2021 were considered for inclusion in the study (Figure 1). However, patients with thyroid cancer were excluded due to its distinct histology and characteristics and availability of well-established, current treatment guidelines.^30-32 Although patients receiving definitive radiation therapy were initially included, due to the small number of patients receiving this singular treatment modality, they were ultimately excluded from the analysis to achieve a more homogenous patient group treated with combined therapies.

Information on patient age; sex (male vs. female); radiotherapy boost (yes vs. no); duration of radiotherapy course; total radiation dose (≤60 Gy vs. >60 Gy); single fraction dose (≤2.0 Gy vs. >2.0 Gy); inclusion of cervical lymphatic drainage pathway (CLDP) or lymph nodes (none, unilateral, or bilateral); CLDP total dose (<60 Gy vs. ≥60 Gy); CLDP single fraction dose (<2.0 Gy vs. ≥2.0 Gy); number of surgeries (1, 2, or ≥3); neck dissection (none, unilateral, or bilateral); adjuvant chemotherapy (yes vs. no); adjuvant immunotherapy (yes vs. no); tumor grade (G; 1 or 2 vs. 3 or 4), size or extent (T; Tis/1, 2, 3, or 4), node (N; 0 or 1 vs. 2–4), and metastasis (M; 0 or 1) classification; tumor histology; primary tumor site; human papillomavirus status (negative vs. positive); history of or current tobacco use (yes vs. no) and number of pack-years, if applicable; alcohol consumption (no, yes, or abuse); number of comorbidities; treatment regimen and need for post-surgery voice prosthetics (yes vs. no) was obtained from the digital documentation system of the University Medical Center.

3 RESULTS

Of the 211 patients identified, 57 (27.0%) died since undergoing radiotherapy. Of the remaining patients, 108 (70.1%) successfully completed the questionnaire. One patient withdrew during the interview for emotional reasons, but available data were included in the analysis to avoid drop-out bias. Among the 108 included patients, median age at diagnosis was 67 years (range, 34–96), and time since treatment was a median of 1.58 years (range, 0.18–3.21). Median total radiotherapy dose was 60.9 Gy (range, 20–72), and median single fraction dose was 2.1 Gy (range, 1.8–4.4).

The mean DRS score was 23.94 with a standard deviation calculated as 32.36. Following a previously published range for DRS score interpretation,³² 40.5% of patients reported no regret (DRS = 0), 30.1% reported mild regret (DRS between 1 and 25), and 29.4% reported moderate to strong regret (DRS > 25). Cronbach's alpha for DRS questionnaire items was 0.945, reflecting high internal consistency within this patient group.

3.1 Correlations between patient, tumor, and treatment characteristics and decision regret

Single fraction dose was the only characteristic that was significantly associated with DRS score after Bonferroni correction (p = 0.044, r = 0.291) (Tables 1–3). Specifically, patients who received a single fraction dose >2.0 Gy had lower DRS scores than patients who received a single fraction dose ≤2.0 (Figure 2).

TABLE 1. Mann–Whitney U tests for associations between patient, tumor, and treatment characteristics and DRS score

	N	Middle ranka	Rank sum	Mann–Whitney U	Z	p-value
Sex
Female	37	57.58	2130.50	1199.50	−0.764	0.445*
Male	71	52.89	3755.50
Radiotherapy boost
No	23	60.43	1390.00	841.00	−1060	0.289*
Yes	85	52.89	4496.00
Radiation dose
≤60 Gy	62	60.48	3749.50	1055.50	−2.381	0.289
>60 Gy	46	46.45	2136.50
Single fraction dose
≤2.0 Gy	63	61.94	3902.50	948.50	−3.024	0.044
>2.0 Gy	45	44.08	1983.50
CLDP dose
≤60 Gy	40	47.41	1896.50	803.50	−1.201	0.230*
>60 Gy	47	41.10	1931.50
CLDP single fraction dose
<2.0 Gy	38	47.49	1804.50	798.50	−1.171	0.241*
≥2.0 Gy	49	41.30	2023.50
Chemotherapy
No	66	56.34	3718.50	1264.50	−0.792	0.428*
Yes	42	51.61	2167.50
Immunotherapy
No	82	53.73	4405.50	1002.50	−0.472	0.637*
Yes	26	56.94	1480.50
G classification
1 or 2	51	58.39	2978.00	1000.00	−2.230	0.442
3 or 4	52	45.73	2378.00
N classification
0 or 1	57	48.29	2752.50	1099.50	−0.093	0.926*
2–4	39	48.81	1903.50
M classification
0	62	31.97	1982.00	29.00	−1.299	0.194*
1	2	49.00	98.00
Histology
Squamous	101	54.02	5456.00	305.00	−0.626	0.531*
Other	7	61.43	430.00
HPV status
Negative	83	55.58	4613.50	947.50	−0.678	0.498*
Positive	25	50.90	1272.50
Tobacco use
No	11	28.09	309.00	243.00	−1.072	0.284*
Yes	55	34.58	1902.00
Voice prosthetic
No	100	53.58	5358.00	308.00	−1.117	0.264*
Yes	8	66.00	528.00

• Abbreviations: CLDP, cervical lymphatic drainage pathway; DRS, Decision Regret Scale; G, tumor grade; HPV, human papillomavirus; N, nodal status.
• ^a The middle rank is a statistical value describing the average rank position of a subgroup within a population based on a specific attribute (i.e., decision regret).
• * p-value before Bonferroni correction.

TABLE 2. Kruskal-Wallis tests for associations between patient, tumor, and treatment characteristics and DRS score

	N	Middle ranka	Kruskal–Wallis H	df	p-value
Primary tumor site
Nasopharynx	2	22.00	8.116	7	0.322*
Oropharynx	17	46.59
Hypopharynx	7	51.29
Larynx	9	61.39
Oral cavity	51	57.91
Nose/sinuses	9	39.67
Skin	11	64.82
Salivary glands	2	57.50
CLDP
None	21	53.60	1.093	2	0.579*
Unilateral	29	59.47
Bilateral	58	52.34
Number of surgeries
One	77	54.62	2.231	2	0.328*
Two	21	48.60
Three or more	10	65.95
Neck dissection
None	19	48.89	1.094	2	0.579*
Unilateral	26	52.94
Bilateral	63	56.83
Treatment regimen
RT	53	56.98	2.248	3	0.522*
RT + CT	29	47.78
RT + IMT	13	53.73
RT + ICT	13	60.15
Alcohol consumption
No	19	24.29	0.408	2	0.815*
Yes	11	25.73
Abuse	17	22.56

• Abbreviations: CLDP, cervical lymphatic drainage pathway; CT, chemotherapy; df, degrees of freedom; DRS, Decision Regret Scale; ICT, immunochemotherapy; IMT, immunotherapy; RT, radiotherapy.
• ^a The middle rank is a statistical value describing the average rank position of a subgroup within a population based on a specific attribute (i.e., decision regret).
• * p-value before Bonferroni correction.

TABLE 3. Spearman's correlations between patient, tumor, and treatment characteristics and DRS score

	Spearman's rho	p-value	N
Age	−0.052	0.595*	108
Radiotherapy duration	−0.085	0.384*	108
T classification	−0.093	0.343*	105
Smoking pack-years	0.143	0.361*	43
Comorbidities	−0.135	0.165*	108

• Abbreviations: DRS, Decision Regret Scale; T, tumor size or extent.
• * p-value before Bonferroni correction.

4 DISCUSSION

We investigated decision regret among patients with HNC who underwent surgery and adjuvant radiotherapy and its association with patient, tumor, and treatment characteristics and the burden of specific symptoms. Over half (59.5%) of patients reported experiencing some degree of decision regret, and over a quarter (29.4%) reported experiencing moderate to strong decision regret.

Symptom burden has long been suspected to be a major driver of decision regret,¹¹ but few studies have distinguished between different symptoms.³⁷ Also, some studies only assessed patients' priorities before therapy or correlated symptoms with QoL rather than decision regret.^{9, 27, 28} In the present study, we found that five specific post-therapy symptoms (restriction in everyday life, trouble speaking, trouble swallowing, pain in irradiated areas, dissatisfaction with appearance), two emotional states (sadness and worry over future health), and one treatment characteristic (single fraction dose) were associated with decision regret among patients with HNC. Correlations between decision regret and pain, sadness, and feeling distressed or upset were reported by a previous study that analyzed some therapy characteristics and symptoms using items from the MDASI-HN.¹¹ However, somewhat different from our results, this previous study found very strong correlations between decision regret and dry mouth, difficulty swallowing, teeth/gum problems, and coughing. As this previous study included only patients with oropharyngeal cancer, their symptoms were possibly more prevalent and severe due the particular tumor localization.¹¹ Although our regression model explained a high amount (36%) of variance in decision regret, the remaining variance must be attributed to other factors that were not investigated in the present study. Furthermore, as correlation is not a unidirectional concept, it remains possible that sadness and worry about future health, for example, may not be the cause but rather the outcome of decision regret. Thus, further research on the psychological aspects of decision regret is needed to advance our understanding of how to properly manage decision regret in post-therapeutic care. Patients' social situation as a potentially stabilizing factor after radiotherapy should also be investigated.

We found that the strongest contributor to decision regret among patients with HNC was restriction in everyday life. Therefore, patients should be questioned before therapy to assess their individual priorities and social situation, which should be accommodated through a suitable therapy choice. Pain in the irradiated areas also contributed to decision regret, suggesting that pain management after radiotherapy is important for reducing decision regret and improving patients' QoL. Similar to a previous study,¹¹ we also found that symptoms such as trouble speaking or swallowing as well as feeling sad are associated with decision regret. However, post-radiotherapy symptoms such as dry mouth and loss of taste were not strongly associated with decision regret despite their frequency and severity among patients with HNC.

Considering patient, tumor, and treatment characteristics, only single fraction dose was associated with decision regret. Specifically, a single fraction dose >2.0 Gy was associated with less decision regret, suggesting a possible benefit of hypofractionated radiation courses. The exact fraction doses comprising this group (>2.0 Gy) in our study were 2.1, 2.2, and 3.0 Gy. Hypofractionation is considered to be well-tolerated in patients with HNC, contributing to better local control and reducing early toxicity.^{38, 39} As improved tumor control likely leads to increased survival and reduced symptom burden, these findings support the use of hypofractionation in situations where multiple therapeutic strategies are viable.

This study has some additional limitations that must be considered. First, the study involved a small number of patients with different disease sites and severities, which could serve to mask potential differences among patient subgroups or limit the generalizability of our findings. Second, as we assessed relationships between patient- and treatment-related factors and decision regret between 2 months and 3.3 years after radiotherapy, it is unclear whether these relationships persist over a longer period of survivorship. However, a previous study of patients with breast cancer reports that the degree of treatment-related decision regret as well as relationships among treatment characteristics and decision regret remain largely stable between ~9 months post-diagnosis and 4 years later.⁴⁰ Third, although we took measures to keep the questionnaire brief, to ensure that questionnaire items would be clearly understood by patients, and to standardize questionnaire and interview experiences across patients, it remains possible that the interviews were prone to biases influenced by factors such as patients' communication ability, attention span, and desire to report positive feelings. Lastly, as our questionnaire was self-constructed, it may be less valid than other widely used standardized questionnaires. However, content validity was achieved through strict definition of item goals and expert consensus on both the construction and selection of content, and construct validity was achieved by using pre-existing validated questionnaires and item pools as inspiration and guidance. Nevertheless, independent validation of our questionnaire would further help confirm our findings.

5 CONCLUSION

This study is one of the first to evaluate decision regret in patients with HNC who underwent surgery and adjuvant radiotherapy and to explore its association with patient, tumor, and treatment characteristics and symptom burden. The two factors most strongly associated with decision regret were restriction in everyday life and a low single fraction dose (≤2.0 Gy). We recommended that health care providers identify a patient's priorities in everyday life and plan multimodal therapy accordingly. When confronted with several potential therapy options, hypofractionation may be preferred. This study makes an important contribution to the current literature, as highly individualized therapy programs that optimize medical care and fit patients' personal needs form the current standard for HNC. Future investigations of psychological aspects of decision regret, its social context, and the impact of combined therapies will help contextualize these results and further enhance our understanding of the origin of and interactions between decision regret and its causes.

ETHICS STATEMENT

All procedures performed in studies involving human participants, including participants provision of an informed consent prior to enrolment in the study, were in accordance with the ethical standards of the institutional Ethics committee approval number 024/21 dated 22 January 2021 and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.