Early Target Identification for Symptomatic Intervention in Radiotherapy in Patients with Esophageal Cancer: A Longitudinal Cross-lagged Panel Dynamic Network Analysis

doi:10.21203/rs.3.rs-4961473/v1

Download PDF

Research Article

Early Target Identification for Symptomatic Intervention in Radiotherapy in Patients with Esophageal Cancer: A Longitudinal Cross-lagged Panel Dynamic Network Analysis

https://doi.org/10.21203/rs.3.rs-4961473/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Predictive identification of core symptoms in patients with esophageal cancer undergoing radiotherapy is essential for early symptom prevention. Few studies address dynamic symptom prediction using longitudinal network analysis.

Objectives

This study aimed to establish predictive relationships between symptoms and identify early intervention targets by constructing a dynamic symptom network.

Methods

This prospective longitudinal study enrolled 146 patients from June 2023 to May 2024. Data were collected using the M.D. Anderson Symptom Inventory Gastrointestinal Cancer Module and the Hospital Anxiety and Depression Scale at three points: T0(pre-radiotherapy), T1(mid-radiotherapy), and T2(1-month post-radiotherapy). R software was used to construct cross-lagged panel networks and calculate predictive and centrality indices.

Results

Distress had the most substantial outgoing influence. At T0→T1, distress predicted anxiety, depression, and disturbed sleep. At T1→T2, distress predicted loss of appetite. Anxiety and depression were most affected at T0→T1, while disturbed sleep and loss of appetite were most affected at T1→T2. Dry mouth and disturbed sleep were significantly aggravated at T1, with dry mouth showing the strongest bridging effect at T0→T1. Drowsiness had the highest centrality at T1→T2.

Conclusions

Distress should be targeted for early intervention to reduce anxiety and depression and improve sleep quality and physical condition post-radiotherapy. Clinicians should dynamically manage dry mouth and drowsiness during radiotherapy to prevent symptom crosstalk.

Implications for cancer survivorship

Managing psycho-emotional states before radiotherapy is critical to preventing other symptoms. Active pharmacologic and non-pharmacologic interventions are needed to improve dry mouth and drowsiness during radiotherapy.

Esophageal cancer

Radiotherapy

Cross-lagged panel network

Symptoms

Distress

Anxiety

The incidence and mortality rates of esophageal cancer have been increasing globally ¹. Due to its dietary and lifestyle habits (e.g., scalding food, hot tea, tobacco, and alcohol habits) ² and population aging ³, China has become the country with the highest incidence and mortality rates of esophageal cancer ¹. In recent years, with the popularization of early screening, treatment of digestive tract cancers, and the improvement of treatment levels in China, the incidence and mortality rates of esophageal cancer have declined, and the survival rate of patients has improved ⁴. However, the quality of survival related to the prognosis of the patients also needs extra attention from clinical practitioners, especially symptom management during patients’ treatment and nursing care after treatment, which has a more significant impact on the patient’s quality of life.

Radiotherapy, as a personalized adjuvant treatment for esophageal cancer, has been widely used in preoperative neoadjuvant, postoperative adjuvant, radical, and palliative treatments. It plays a significant role in preventing tumor recurrence and prolonging survival ⁵. Radiotherapy has many advantages, but its damage to normal tissues of the body and its psychological impact should not be ignored. Long-term ionizing radiation can lead to fatigue, sleep disorders, and distress ⁶; although some of the symptoms disappear immediately after the end of radiotherapy, some symptoms persist for a long time in the survival of the patients, causing irreversible damage to the patient's body, which seriously affects the patient’s quality of life ⁷. Therefore, understanding the dynamic changes in symptoms during radiotherapy in patients with esophageal cancer and early intervention for core symptoms is urgently needed.

Previous symptom–centered studies have focused on the postoperative or chemotherapy period of esophageal cancer, and the majority of cross-sectional studies have focused on single symptoms, total symptom scores, or clustered symptoms ^8–11, which have provided rich perspectives on the contemporaneous symptomatic experiences of patients with esophageal cancer under specific treatment modalities. However, different treatment modalities may result in significant symptomatic differences between patients. In addition, using a single symptom or cluster of symptoms as the object of study can overlook the complex interactions among multiple symptoms, resulting in information loss and even biased findings. It is necessary to include all symptoms and explore their interrelationships to understand the nature of symptom occurrence better. This can be done by constructing symptom networks, which is an emerging symptom research paradigm ^12,13. Symptom networks are categorized as contemporaneous ^14,15 and dynamic networks ^16–19. Contemporaneous networks can parse cross-sectional multi-symptom data to provide all symptom associations at one-time point. However, they cannot predict symptom occurrence at the next time point and causal relationships between symptoms. Dynamic networks make up for the shortcomings of contemporaneous networks, which can not only identify early core symptoms and tap early symptom intervention targets but also predict the symptom changes in the next stage through the causal relationship between symptoms, solving the problem from "what" to "why.”

Cross-lagged panel networks (CLPN) are a dynamic network analysis method that extends the combination of network analysis and cross-lag modeling to longitudinal data at discrete time points ²⁰. Using longitudinal panel data, we explored the interaction effects and predictability between network nodes over time.

Therefore, this study constructed a CLPN to understand the symptom dynamics, interactions between symptoms, and symptom predictability over time during radiotherapy in patients with esophageal cancer. By identifying early targets for intervention, this study aims to reduce the symptom burden on patients, accelerate their recovery, and improve their quality of life during the survivorship period.

2.1 Study design and participants

In this prospective longitudinal study conducted from June 2023 to May 2024, we recruited patients with esophageal cancer from a tertiary hospital in Shandong Province using convenience sampling. Written informed consent was obtained from all the patients who met the inclusion criteria. The inclusion criteria were esophageal squamous cell carcinoma or adenocarcinoma confirmed by tissue biopsy according to World Health Organization (WHO) diagnostic criteria, age > 18 years, completion of the prescribed course of radiotherapy, and no cognitive impairment or ability to complete the questionnaire independently. The exclusion criteria were previous radiotherapy for esophageal cancer, communication or comprehension deficits, and ongoing participation in other clinical intervention studies. We followed up with patients at three-time points: T0, one week before radiotherapy (baseline period); T1, mid-radiotherapy (median radiotherapy time, day 20); and T2, one month after the end of radiotherapy, with the total period of the three surveys ranging from approximately to 75–90 days. A total of 163 participants were eligible for recruitment, excluding missing data on symptoms (11), missing data on demographic and clinical characteristics (3), and dropouts (3), resulting in the inclusion of 146 participants in this study.

2.2 Measurement

Demographic and clinical characteristics

The patients self-reported demographic variables, including age, sex, educational background, marital status, average monthly income, smoking status, and alcohol history. Clinical characteristics, including tumor clinical stage, number of radiotherapy sessions, radiotherapy dose, and disease duration, were extracted from patients’ medical records.

Chinese version of the MD Anderson Symptom Inventory Gastrointestinal Cancer Module

The MD Anderson Symptom Inventory Gastrointestinal Cancer (MDASI-GI-C) scale consists of two parts: the first part consists of 13 core symptoms and five specific gastrointestinal symptoms for a total of 18 symptom entries, and the second part consists of six entries measuring the extent to which the symptoms interfere with life. All these elements measured how the patient felt about the symptoms in the past 24 h. This scale was very reliable ²¹.

Hospital Anxiety and Depression Scale

The Hospital Anxiety and Depression Scale is a self-rating scale for assessing patients’ anxiety and depression levels. It consists of 14 items, each with a score ranging from zero to three. It is divided into two subscales–Anxiety and Depression–each with seven items. The patients were asked to choose the level of description for the corresponding item based on how they felt during the past week. The Anxiety and Depression subscales have a total score range of 0–21. Normal range, no symptoms of anxiety or depression: 0–7; Mild symptoms, further assessment or attention is recommended: 8–10; Moderate symptoms, professional counseling, or treatment is recommended: 11–14; Severe symptoms, urgent professional psychotherapy is recommended: 15–21. This scale is widely used and has good reliability ^22,23.

2.3 Statistical analysis

Descriptive statistics were performed using SPSS 26.0. Normally distributed continuous variables were described as mean ± standard deviation, and non-normally distributed continuous variables were described as median ± quartiles (P25, P75); categorical variables were described as frequency (percentage).

Rhemtulla et al. ²⁰ designed the cross-lagged panel network analysis method, which compensates for the shortcomings of longitudinal analysis, which can only be performed on dense data, and realizes the analysis of discrete time-point data. To avoid biased results, we collected data from three time points because fewer time-point measurements (e.g., two time points) may amplify individual differences and lead to lagged effects between variables that are not causally related. This analysis was realized by R software (4.0.4) to construct the prediction paths for two dynamic networks, T0→T1 and T1→T2. In addition to estimating networks, predictive and network centrality metrics were critical components of this analysis. We estimated the networks by replacing the Lasso and Ridge regularization techniques with resilient networks, which ensured feature selection accuracy while maintaining the sparsity of the networks.

The predictive indicators include the outgoing expected impact (out-EI) and incoming expected impact (in-EI), to assess the predictive ability and sensitivity of the nodes. In clinical practice, out-EI can be regarded as a treatment target, and in-EI can be regarded as an important outcome indicator for evaluating the effectiveness of the intervention on the treatment target ²⁴. Centrality indicators mainly include in-intensity, out-intensity, and mediator centralities, which help to understand the importance of each node in the network.

2.4 Sample size estimation

The MDASI-GI-C scale has 18 symptoms, and we included 13 symptoms after excluding those with an incidence of < 20% and a severity score of < 1. In addition, we included anxiety and depression symptoms to understand the interactions between physical and psychological symptoms. Therefore, the constructed symptom network model had a threshold parameter, n of 15, and the minimum overall correlation parameter required ²⁵ for a sample of 120 cases was estimated using the formula [n + n(n-1)/2]. In addition, considering a 15% sample dropout rate, the required sample size should be at least 141 and this study fulfilled this requirement.

2.5 Ethical consideration

The Academic Ethics Committee of the Second Affiliated Hospital of Shandong First Medical University approved this study(approval number: 2023–013; approval date: January 31, 2023).

3.1 Demographic and clinical characteristics of participants

Table 1 shows the statistical results of the demographic and clinical characteristics of the 146 patients who underwent radiotherapy for esophageal cancer at baseline. The mean age was 64.2 years (27.4%, female; 62.3%, rural population; 41.8%, smoked; 39%, drank alcohol; 6.8%, malnourished; 36.4%, overweight or obese). The majority of the patients (88.4%) were in the advanced clinical stage (stage III and IV), and the vast majority of the patients (93.2%) had received at least one course of chemotherapy. A total of 71.9% of patients had a disease duration of < 1 year.

Table 1

Demographic data and clinical characteristics of all participants at baseline (N = 146)
Baseline characteristics	N (%)
Gender
Male	106(72.6)
Female	40(27.4)
Age, years
Mean(SD)	64.2(8.3)
Residence
Rural area	91(62.3)
Town	37(25.4)
Urban area	18(12.3)
Smoking
Yes	61(41.8)
No	85(58.2)
Alcohol drinking
Yes	57(39.0)
No	89(61.0)
BMI, kg/m²
Mean(SD)	22.6(3.3)
<18.5	10(6.8)
18.5–23.9	83(56.8)
>23.9	53(36.4)
Clinical stage
Ⅰ	0(0.0)
Ⅱ	17(11.6)
Ⅲ	62(42.5)
Ⅳ	67(45.9)
Radiotherapy frequency
Mean(SD)	28.4(3.7)
Radiotherapy dose, gray
Mean(SD)	56.6(7.4)
surgical history
Yes	58(39.7)
No	88(60.3)
Chemotherapy history
Yes	136(93.2)
No	10(6.8)
Course of illness, years
<1	105(71.9)
≥ 1, <3	25(17.1)
≥ 3, <5	10(6.9)
≥ 5	6(4.1)

3.2 Symptom severity and prevalence

Table 2 summarizes the severity and frequency of symptoms at three time points. Before radiotherapy (T0), distress (3.96 ± 1.13,98.6%), sadness (3.27 ± 0.93, 99.3%), and fatigue (3.21 ± 1.22, 97.9%) were the most severe and common symptoms; as radiotherapy progressed, all patients at T1 experienced severe fatigue (4.42 ± 1.08, 100%), distress (4.38 ± 1.04, 100%) and loss of appetite (4.38 ± 1.28, 100%); 1 month after the end of radiotherapy (T2) distress (3.92 ± 1.07, 100%), fatigue (3.72 ± 1.36, 99.3%), and sadness (3.52 ± 0.88, 100%) remained the most severe and common symptoms. In comparison to T1, although the severity and incidence of these symptoms decreased, they were still higher than the baseline period levels. In addition, dry mouth (3.10 ± 1.33, 95.2%) and disturbed sleep (4.00 ± 1.55, 97.3%) were particularly severe and more frequent in T1 compared to dry mouth (0.16 ± 0.66, 8.2%) and disturbed sleep (2.64 ± 1.64, 91.1%) in T0, with radiotherapy. After radiotherapy, the severity of these two symptoms decreased, but the incidence remained high. The same trend of T1 > T2 > T0 was present for the incidence of anxiety and depression (≥ 8 points) at the three time points, with the incidence of anxiety and depressive symptoms reaching 62.3% and 78.8%, respectively, at T1. The incidence of symptoms 1 month after the end of radiotherapy remaining higher than the level at baseline.

Table 2

Severity and prevalence of symptoms during radiotherapy in patients with esophageal cancer
Symptom Item	T0		T1		T2
Symptom Item	Average Severity Score Mean(SD)	Occurrence Frequency (%)	Average Severity Score Mean(SD)	Occurrence Frequency (%)	Average Severity Score Mean(SD)	Occurrence Frequency (%)
Distress	3.96(1.13)	98.6	4.38(1.04)	100	3.92(1.07)	100
Fatigue	3.21(1.22)	97.9	4.42(1.08)	100	3.72(1.36)	99.3
Lack of appetite	2.76(1.58)	91.8	4.34(1.28)	100	3.42(1.33)	97.9
Disturbed sleep	2.64(1.64)	91.1	4.00(1.55)	97.3	2.70(1.50)	93.8
Sadness	3.27(0.93)	99.3	3.75(0.89)	100	3.52(0.88)	100
Dry mouth	0.16(0.66)	8.2	3.10(1.33)	95.2	1.34(0.90)	81.5
Memory problem	2.93(1.70)	90.4	2.89(1.68)	91.8	2.90(1.69)	91.1
Drowsy	1.53(1.11)	82.2	2.17(1.05)	95.2	1.82(1.10)	88.4
Pain	0.81(1.36)	35.6	1.53(1.22)	70.5	0.68(1.14)	35.6
Difficulty swallowing	1.09(1.64)	40.4	1.43(1.59)	53.4	1.01(1.35)	45.9
Change in taste	0.91(1.09)	52.1	1.36(1.08)	71.9	1.02(1.01)	62.7
Numbness	0.82(1.31)	34.2	0.86(1.31)	37.7	0.86(1.35)	37
Nausea	0.42(0.92)	21.5	0.43(0.81)	24.7	0.21(0.62)	11.6
Diarrhea	0.36(1.00)	13.0	0.42(0.96)	17.8	0.12(0.48)	6.8
Shortness of breath	0.36(0.88)	16.4	0.30(0.77)	15.1	0.24(0.65)	13.0
Feeling bloated	0.17(0.66)	8.2	0.12(0.46)	6.2	0.05(0.28)	4.1
Vomiting	0.10(0.60)	4.1	0.05(0.34)	2.1	0.04(0.26)	2.7
Constipation	0.04(0.31)	2.1	0.02(0.25)	0.7	0.05(0.44)	1.4
HADS- Anxiety	6.85(2.98)	-	9.40(3.27)	-	7.92(3.14)
0–7	5.21(1.32)	67.8	6.05(1.04)	37.7	5.54(1.10)	52.1
8–10	9.00(0.74)	23.3	9.22(0.79)	22.3	8.91(0.80)	30.1
11–14	12.33(0.87)	6.2	12.31(1.04)	30.8	12.00(1.08)	12.3
15–21	16.75(1.50)	2.7	16.00(1.58)	6.2	15.88(1.13)	5.5
HADS- Depression	8.25(3.03)	-	9.67(2.74)	-	8.82(2.76)	-
0–7	5.85(1.34)	40.4	6.39(0.99)	21.2	6.28(1.16)	32.2
8–10	8.70(0.73)	43.2	8.88(0.77)	45.2	8.70(0.77)	47.3
11–14	11.80(0.95)	13.7	12.22(1.11)	28.1	12.17(1.01)	16.4
15–21	17.25(2.63)	2.7	15.88(1.73)	5.5	16.83(2.14)	4.1

3.3 Network structure and predictive indicator results

Figures 1 and 2 show the CLPN visualization results and predictive metrics for the 15 symptom nodes of patients with esophageal cancer T0→T1 (pre-radiotherapy to mid-radiotherapy), respectively. Green arrows indicate positive predictions, and red arrows indicate negative predictions. Figures 3 and 4 show the CLPN visualization results and predictive indicators for T1→T2 (mid-radiotherapy to one month after the end of radiotherapy).

During T0→T1, the distress group had the highest out-EI (V5, out-EI = 2.74). It mainly predicted anxiety (V5→V14, edge weight = 1.43), disturbed sleep (V5→V4, edge weight = 0.53), fatigue (V5→V2, edge weight = 0.26) and depression (V5→V15, edge weight = 0.24). Anxiety (V14, in-EI = 2.08) had the highest in-EI, which was mainly influenced by distress (V5→V14, edge weight = 1.43), sadness (V10→V14, edge weight = 0.22), and fatigue (V2→V14, edge weight = 0.21); and depression (V15, in-EI = 1.88) had the second highest in-EI after anxiety, which was mainly influenced by sadness (V10→V15, edge weight = 0.62) and memory difficulties (V6→V15, edge weight = 0.38).

During T1→T2, distress still had the highest out-EI (V5, out-EI = 0.62), mainly predicting sadness (V5→V10, edge weight = 0.19) and loss of appetite (V5→V7, edge weight = 0.16). Depression (V15, out-EI = 0.53) had the second highest ranked out-EI and mainly predicted anxiety (V15→V14, edge weight = 0.21) and disturbed sleep (V15→V4, edge weight = 0.13). From mid-radiotherapy to 1 month after the end of radiotherapy, disturbed sleep had the highest in-EI value (V4,in-EI = 0.60), mainly influenced by fatigue (V2→V4, edge weight = 0.16) and depression (V15→V4, edge weight = 0.13); loss of appetite (V7,in-EI = 0.54) had the second-highest in-EI, mainly influenced by distress ( V5→V7, edge weight = 0.16), dysphagia (V12→V7, edge weight = 0.15) and sadness (V10→V7, edge weight = 0.12).

3.4 Centralized indicator results

Figures S1 and S2 (See supplementary document for details) show the centrality indicators of symptoms in the T0→T1 and T1→T2 phases, respectively. In the T0→T1 phase, the symptom node with the highest out intensity was distress (V5, Rs = 2.80); anxiety had the highest in intensity (V14, Rs = 3.63), followed by depression (V15, Rs = 3.03); and the mediator centrality indicator with the highest centrality indicator was dry mouth (V9, Rb = 63), followed by fatigue (V2, Rb = 60). At T1→T2, the symptom node with the highest out intensity remained distressed (V5, Rs = 0.62), followed by depression (V15, Rs = 0.53); the nodes with higher in intensity were lethargy (V8, Rs = 0.68), loss of appetite (V7, Rs = 0.66), and disturbed sleep (V4, Rs = 0.60); the highest indicator of mediated centrality was drowsy (V8, Rb = 86).

This study explored the dynamic symptom network during radiotherapy in patients with esophageal cancer using the CLPN. Our findings revealed symptom predictive ability and interactions between symptoms at two adjacent time points, emphasizing the need for early and precise symptom interventions.

Distress had the strongest predictive power during T0→T1. Distress at T0 predicted multiple changes in symptoms at T1. Of these, anxiety had the strongest predictive ability. Cancer-related distress is a psychologically, socially, and spiritually multifactorial unpleasant experience that interferes with an individual's ability to cope effectively with illness and treatment. However, it is a relatively mild, normal expression of negative psychological emotions ²⁶. Anxiety disorders are an extension of the distress continuum, which begins with normal feelings of common psychological brittleness, sadness, and fear and can lead to the emergence of psychiatric disorder symptoms when induced by risk factors or under chronic stress. Anxiety differs from distress in that it is excessive, unfounded, and often illogical fears and worries ²⁶. Once the distress progresses to anxiety or depression, the patient is exposed to increased safety risks, such as suicidal and self-injurious behaviors ²⁷, and fewer effective interventions, even requiring medication for severe anxiety disorders. The fact that distress predicts anxiety may be related to the patient's psychological fragility, fear of radiotherapy, and experience of physiological symptoms such as fatigue and pain during radiotherapy, which exacerbates distress symptoms. Radiotherapy exacerbates anxiety and depressive symptoms in patients ^28,29. The majority of the participants in this study were clinically advanced, and almost all of them experienced varying degrees of distress at baseline. Notably, clinically advanced and psychologically vulnerable patients with cancer experience anxiety symptoms immediately after treatment ³⁰. Considering the high prevalence of anxiety and depression at T1 and the patient safety risks that may result, the use of distress at baseline as an early target for symptomatic intervention and the provision of psychosocial support to patients as early as possible are essential for preventing the onset of serious psychiatric disorders.

In addition, distress at T0 predicted symptoms such as disturbed sleep and fatigue, consistent with the findings of several studies ^24,27 that negative emotions exacerbate disturbed sleep and lead to physical fatigue. Therefore, addressing the symptoms of distress in patients with esophageal cancer at baseline can have a positive effect on preventing the development of anxiety disorders, disturbed sleep, fatigue, and depression during radiation therapy.

During T0→T1, anxiety and depression had higher in-EI. They were susceptible to other symptoms at T0, indicating that they could be used as mid-radiotherapy outcome indicators to assess the effectiveness of symptomatic interventions at T0. As previously described, anxiety and depression act as extensions of a continuum of distress, and the assessment of and intervention for psychological symptoms in patients before radiotherapy are key to preventing them from developing psychiatric disorders during radiotherapy.

During the T1→T2 period, distress remained the highest out-EI symptom, but unlike the T0→T1 period, distress in T1 predominantly predicted loss of appetite and sadness in T2. This result was consistent with the findings of Shang et al. ²⁴. Patients' psychological symptoms worsen in the middle stage of radiotherapy, and psychological stress can activate the gut-brain axis to affect the normal functioning of the gastrointestinal tract by inhibiting the vagus nerve from sending signals, which in turn leads to symptoms such as bloating, nausea, and loss of appetite ^31–33.

The onset of depression at T1 predicted anxiety and disturbed sleep at T2, and the onset of depression at T1 was influenced by distress at T0. Disturbed sleep at T2, the symptom with the highest in-EI, could be prevented by managing symptoms of distress at baseline, thus alleviating disturbed sleep. Loss of appetite, the second-highest in-EI, was affected by psychological symptoms and dysphagia at T1. In patients with esophageal cancer, the radiation site is often located in the chest, which may lead to swelling of the throat and esophagus and consequently dysphagia, and the exacerbation of this symptom is cumulative with the duration of radiotherapy. The patient may experience symptoms after 1 week of radiotherapy and continue until 2 weeks after the end of radiotherapy ³⁴. However, some studies have shown that radiotherapy for esophageal cancer can alleviate dysphagia symptoms, which may depend on the cause of the symptoms; if dysphagia is due to tumor compression of the esophagus or pharynx, radiotherapy can alleviate these symptoms ³⁵.

Dry mouth has the highest mediator centrality during T0→T1, and they assume a vital role in a dynamic network that transmits symptom changes and connects symptoms at different time points, and intervention in these symptoms is a critical way to prevent the occurrence of other symptoms. The severity and incidence of dry mouth were significantly higher at T1 than at T0, which is consistent with the study of Pinna et al. ³⁶. Patients with esophageal cancer can experience dry mouth after radiotherapy because the radiation produced by radiotherapy spreads to a certain extent to the surrounding tissues; therefore, even though the main site of radiotherapy is the esophagus, salivary glands in the adjacent oral tissues may also be involved ³⁶, and the occurrence of dry mouth symptoms can widely affect the changes in other symptoms in the T1 stage.

In the T1→T2 phase, drowsiness was considered the symptom with the strongest bridging effect. The symptoms are affected by multiple symptoms from mid-radiotherapy to 1 month after the end of radiotherapy, and they also influence the development of multiple symptoms. Notably, dry mouth predicts drowsiness, which may be because dry mouth causes patients to wake up frequently or fail to reach the deep sleep stage, often making them feel tired or sleepy during the day. Drowsiness affects multiple psychological symptoms in patients' T2 stage, such as distress and depression, and also affects appetite, so focusing on the sleepy symptoms in T1→T2 is warranted.

To the best of our knowledge, this is the first study to use cross-lagged panel network analysis to identify targets for early symptomatic intervention in radiotherapy for patients with esophageal cancer. The significance of this study is, firstly, that we provide new methodology that enables the identification of effective symptom prediction targets in longitudinal data, and secondly, that we have found the importance of intervening on psycho-emotional symptoms before radiotherapy in patients, which not only alleviates the occurrence of anxiety and depression during radiotherapy, but also improves many physical symptoms.

This study has some limitations. First, this study fulfills the CLPN requirements for time points; however, the fact that only three-time points in close proximity were tracked limits our understanding of the long-term experience of patients' symptoms after radiotherapy. Future studies could consider incorporating more time points. Second, the study’s sample size is limited, and more extensive studies are needed to validate the results of the present study. Since CLPN is an emerging research study of discrete time points method of dynamic symptom interactions, including multiple time points of tracking and a rich sample size is essential to provide reliable findings.

In conclusion, our study revealed a dynamic interaction between symptoms and their predictive power during radiotherapy in patients with esophageal cancer. We emphasize the importance of actively addressing patients' distressing symptoms before radiotherapy to focus on preventing the risk of anxiety and depression in the mid-radiotherapy period, as well as alleviating disturbed sleep and fatigue; dry mouth symptoms, which are highly prevalent in the mid-radiotherapy period, have a solid bridging role in the network and should be assessed and intervened by clinical practitioners early on to alleviate sleepiness and psychological symptoms by reducing the interactions between symptoms.

Conflict of interest statement:

The authors have no funding or conflicts of interest to disclose.

Acknowledgements

We thank all the study participants and those who contributed their time and effort to this study.

Funding

This study was funded by Shandong First Medical University [Grant No. SKX340031, 2022].

CRediT authorship contribution statement:

Ke Wang was involved in conceptualization, methodology, data curation, formal analysis, and writing the original draft. Mengjia Liu was involved in data investigation, data analysis, and writing the original draft. Lin Yang was involved in data investigation and analysis. Min Diao, Hong Li, and Yaxin Chang were involved in data investigation. Jordan Tovera Salvador was involved in writing, review, and editing. Zhaoxia Yang was involved in conceptualization, formal analysis, and supervision.

Data availability statement

The data that support the findings of this study are available from the corresponding author [ZX, Yang], upon reasonable request.

International Agency for Research on Cancer. Cancer today. https://gco.iarc.who.int; 2024[accessed 08 February 2024].
Yu C, Tang H, Guo Y, et al. Hot Tea Consumption and Its Interactions With Alcohol and Tobacco Use on the Risk for Esophageal Cancer: A Population-Based Cohort Study [published correction appears in Ann Intern Med. 2018 May 1;168(9):684]. Ann Intern Med. 2018;168(7):489-497. doi:10.7326/M17-2000
Xia C, Dong X, Li H, et al. Cancer statistics in China and United States, 2022: profiles, trends, and determinants. Chin Med J (Engl). 2022;135(5):584-590. Published 2022 Feb 9. doi:10.1097/CM9.0000000000002108
Feng X, Zhu J, Hua Z, et al. Satisfaction in population-based cancer screening in a Chinese rural high-risk population: the Yangzhong early diagnosis and treatment of upper gastrointestinal cancer. BMC Health Serv Res. 2022;22(1):675. Published 2022 May 19. doi:10.1186/s12913-022-08076-1
Cellini F, Manfrida S, Casà C, et al. Modern Management of Esophageal Cancer: Radio-Oncology in Neoadjuvancy, Adjuvancy and Palliation. Cancers (Basel). 2022;14(2):431. Published 2022 Jan 15. doi:10.3390/cancers14020431
Dilalla V, Chaput G, Williams T, Sultanem K. Radiotherapy side effects: integrating a survivorship clinical lens to better serve patients. Curr Oncol. 2020;27(2):107-112. doi:10.3747/co.27.6233
Stiegelis HE, Ranchor AV, Sanderman R. Psychological functioning in cancer patients treated with radiotherapy. Patient Educ Couns. 2004;52(2):131-141. doi:10.1016/s0738-3991(03)00021-1
Watt E, Whyte F. The experience of dysphagia and its effect on the quality of life of patients with oesophageal cancer. Eur J Cancer Care (Engl). 2003;12(2):183-193. doi:10.1046/j.1365-2354.2003.00376.x
Gillman A, Hayes M, Sheaf G, Walshe M, Reynolds JV, Regan J. Exercise-based dysphagia rehabilitation for adults with oesophageal cancer: a systematic review. BMC Cancer. 2022;22(1):53. Published 2022 Jan 10. doi:10.1186/s12885-021-09155-y
Zerbib F, Omari T. Oesophageal dysphagia: manifestations and diagnosis. Nat Rev Gastroenterol Hepatol. 2015;12(6):322-331. doi:10.1038/nrgastro.2014.195
Lagergren P, Johar A, Rosenlund H, et al. Severe reflux, sleep disturbances, and health-related quality of life after esophageal cancer surgery. J Cancer Surviv. 2021;15(6):818-824. doi:10.1007/s11764-020-00974-9
Zhu Z, Xing W, Hu Y, Wu B, So WKW. Paradigm shift: Moving from symptom clusters to symptom networks. Asia Pac J Oncol Nurs. 2021;9(1):5-6. Published 2021 Dec 25. doi:10.1016/j.apjon.2021.12.001
Schweren L, van Borkulo CD, Fried E, Goodyer IM. Assessment of Symptom Network Density as a Prognostic Marker of Treatment Response in Adolescent Depression. JAMA Psychiatry. 2018;75(1):98-100. doi:10.1001/jamapsychiatry.2017.3561
Wang K, Diao M, Yang Z, Salvador JT, Zhang Y. Identification of Core Symptom Cluster in Patients With Digestive Cancer: A Network Analysis. Cancer Nurs. Published online October 26, 2023. doi:10.1097/NCC.0000000000001280
Zhu Z, Sun Y, Kuang Y, et al. Contemporaneous symptom networks of multidimensional symptom experiences in cancer survivors: A network analysis. Cancer Med. 2023;12(1):663-673. doi:10.1002/cam4.4904
Bringmann LF, Lemmens LH, Huibers MJ, Borsboom D, Tuerlinckx F. Revealing the dynamic network structure of the Beck Depression Inventory-II. Psychol Med. 2015;45(4):747-757. doi:10.1017/S0033291714001809
Groen RN, Snippe E, Bringmann LF, et al. Capturing the risk of persisting depressive symptoms: A dynamic network investigation of patients' daily symptom experiences. Psychiatry Res. 2019;271:640-648. doi:10.1016/j.psychres.2018.12.054
Chavez-Baldini U, Verweij K, de Beurs D, et al. The interplay between psychopathological symptoms: transdiagnostic cross-lagged panel network model. BJPsych Open. 2022;8(4):e116. Published 2022 Jun 27. doi:10.1192/bjo.2022.516
Ross EJ, Cassisi JE, Joseph D, Dunn ME, Jex S. Cross-lagged analyses between gastrointestinal symptoms, psychological distress, and disability in emerging adults. Appl Psychol Health Well Being. 2022;14(3):920-936. doi:10.1111/aphw.12358
Wysocki, Anna, et al. “Cross-lagged Network Models.” PsyArXiv, 1 Oct. 2022. Web.
Wang XS, Williams LA, Eng C, et al. Validation and application of a module of the M. D. Anderson Symptom Inventory for measuring multiple symptoms in patients with gastrointestinal cancer (the MDASI-GI). Cancer. 2010;116(8):2053-2063. doi:10.1002/cncr.24920
Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the Hospital Anxiety and Depression Scale. An updated literature review. J Psychosom Res. 2002;52(2):69-77. doi:10.1016/s0022-3999(01)00296-3
Yang Y, Ding R, Hu D, Zhang F, Sheng L. Reliability and validity of a Chinese version of the HADS for screening depression and anxiety in psycho-cardiological outpatients. Compr Psychiatry. 2014;55(1):215-220. doi:10.1016/j.comppsych.2013.08.012
Shang B, Bian Z, Luo C, et al. Exploring the dynamics of perioperative symptom networks in colorectal cancer patients: a cross-lagged panel network analysis. Support Care Cancer. 2023;32(1):62. Published 2023 Dec 27. doi:10.1007/s00520-023-08288-z
Papachristou N, Barnaghi P, Cooper B, et al. Network Analysis of the Multidimensional Symptom Experience of Oncology. Sci Rep. 2019;9(1):2258. Published 2019 Feb 19. doi:10.1038/s41598-018-36973-1
Li X, Wang XS, Huang H, et al. National survey on the availability of oncology palliative care services at tertiary general and cancer hospitals in China. BMC Palliat Care. 2023;22(1):144. Published 2023 Sep 28. doi:10.1186/s12904-023-01259-5
Tomaso CC, Johnson AB, Nelson TD. The effect of sleep deprivation and restriction on mood, emotion, and emotion regulation: three meta-analyses in one. Sleep. 2021;44(6):zsaa289. doi:10.1093/sleep/zsaa289
Deshields T, Tibbs T, Fan MY, Bayer L, Taylor M, Fisher E. Ending treatment: the course of emotional adjustment and quality of life among breast cancer survivors immediately following radiation therapy. Support Care Cancer. 2005;13(12):1018-1026. doi:10.1007/s00520-005-0801-z
Takahashi T, Hondo M, Nishimura K, et al. Evaluation of quality of life and psychological response in cancer patients treated with radiotherapy. Radiat Med. 2008;26(7):396-401. doi:10.1007/s11604-008-0248-5
Henry M, Sargi E, Frenkiel S, et al. Longitudinal study indicating antecedent psychosocial vulnerability as predictor of anxiety disorders post-treatment in people with head and neck cancer. Psychooncology. 2021;30(11):1910-1919. doi:10.1002/pon.5760
Bajic JE, Johnston IN, Howarth GS, Hutchinson MR. From the Bottom-Up: Chemotherapy and Gut-Brain Axis Dysregulation. Front Behav Neurosci. 2018;12:104. Published 2018 May 22. doi:10.3389/fnbeh.2018.00104
Halvorson CS, Sánchez-Lafuente CL, Johnston JN, Kalynchuk LE, Caruncho HJ. Molecular Mechanisms of Reelin in the Enteric Nervous System and the Microbiota-Gut-Brain Axis: Implications for Depression and Antidepressant Therapy. Int J Mol Sci. 2024;25(2):814. Published 2024 Jan 9. doi:10.3390/ijms25020814
Maunsell R, Sodergren S, Hopkinson J, Shaw C, Foster C, Wheelwright S. Nutritional care in colorectal cancer-what is the state of play?. Colorectal Dis. 2021;23(12):3227-3233. doi:10.1111/codi.15933
Che Z, Suhail A, Hainc N, et al. The Quantification of Radiation Fibrosis Using Clinically Indicated Magnetic Resonance Imaging for Head and Neck Cancer Patients. Dysphagia. Published online March 27, 2024. doi:10.1007/s00455-024-10678-2
Welsch J, Kup PG, Nieder C, et al. Survival and Symptom Relief after Palliative Radiotherapy for Esophageal Cancer. J Cancer. 2016;7(2):125-130. Published 2016 Jan 1. doi:10.7150/jca.13655
Pinna R, Campus G, Cumbo E, Mura I, Milia E. Xerostomia induced by radiotherapy: an overview of the physiopathology, clinical evidence, and management of the oral damage. Ther Clin Risk Manag. 2015;11:171-188. Published 2015 Feb 4. doi:10.2147/TCRM.S70652

No competing interests reported.

Supplementarydocument.docx

Download PDF

Reviewers invited by journal
18 Sep, 2024
Editor assigned by journal
23 Aug, 2024
Submission checks completed at journal
23 Aug, 2024
First submitted to journal
23 Aug, 2024

You are reading this latest preprint version

Early Target Identification for Symptomatic Intervention in Radiotherapy in Patients with Esophageal Cancer: A Longitudinal Cross-lagged Panel Dynamic Network Analysis

Status:

Version 1

Abstract

Figures

1 Introduction

2 Materials and Methods

2.1 Study design and participants

2.2 Measurement

2.3 Statistical analysis

2.4 Sample size estimation

2.5 Ethical consideration

3 Results

3.1 Demographic and clinical characteristics of participants

3.2 Symptom severity and prevalence

3.3 Network structure and predictive indicator results

3.4 Centralized indicator results

4 Discussion

5 Implications

6 Limitations

7 Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1