Prevention and Treatment of Radiation-Induced Esophagitis With Oral Herbal Medicine: A Systematic Review and Meta-Analysis

Abstract

Background:

Radiation therapy is commonly used for the curative or palliative treatment of lung, esophageal, breast, or mediastinal tumors, with radiation-induced esophagitis (RIE) being a prevalent adverse event. Despite its high incidence, no cure has been identified for RIE, and validated treatment methods remain elusive. This systematic review explores global clinical research on herbal medicines for RIE treatment, assessing their efficacy in prevention and treatment.

Methods:

A comprehensive literature search across 5 databases targeted randomized controlled clinical trials (RCTs). RCTs published before June 2024 were eligible if they investigated herbal medicine use for RIE prevention or treatment. Using Cochrane methodology and RevMan 5.4 software, a meta-analysis was conducted, and the certainty of evidence was evaluated with the GRADE system.

Results:

Following the literature selection process, 81 RCTs with 7283 patients were included in the analysis. All studies administered herbal medicine to the treatment group and usual care with conventional medicine to the control group. Meta-analysis revealed that, in terms of overall incidence rate, herbal medicine demonstrated a significantly greater preventive effect compared to usual care (relative risk [RR]: 0.71; 95% CI: 0.65–0.78). Similarly, herbal medicine showed superior efficacy in improving RIE treatment outcomes compared to usual care (RR: 1.29; 95% CI: 1.21–1.38). Herbal medicine exhibited a comparable safety profile with fewer gastrointestinal adverse events and no significant liver toxicity.

Conclusions:

Herbal medicine demonstrated significant benefits in reducing RIE incidence, delaying onset, alleviating symptoms, and shortening duration in patients undergoing radiation therapy for thoracic malignancies. However, due to limited certainty of the evidence, well-designed, large-scale RCTs are essential to establish the clinical effectiveness of herbal medicine for prevention and treatment, ensuring high-quality evidence for incidence rate and treatment effective rates.

Introduction

The annual incidence of cancer is increasing worldwide, causing a growing global burden that poses a serious threat to human health and life. According to the 2020 GLOBOCAN estimates of cancer incidence and mortality (International Agency for Research on Cancer), 19.3 million new cancer cases were reported in 2020 with approximately 10.0 million cancer deaths. Among new cancer cases, female breast cancer was the most commonly diagnosed with 2.3 million cases (11.7%), followed by lung cancer (11.4%), colorectal cancer (10.0%), prostate cancer (7.3%), and stomach cancer (5.6%). Lung cancer accounted for the highest number of deaths (1.8 million, 18%), followed by colorectal cancer (9.4%), liver cancer (8.3%), stomach cancer (7.7%), and female breast cancer (6.9%). According to projections by the International Agency for Research on Cancer, the global cancer burden is expected to reach 28.5 million cases by 2040, indicating a 47% increase from the number of cases in 2020. ¹

Breast cancer and lung cancer, which rank first and second in incidence, are both malignant tumors in the chest, and high-dose radiation therapy is used as major standard treatment.^2,3 Radiation therapy is not only for curative purposes, but also for palliative purpose to relieve pressure or obstruction of the esophagus, bronchial tubes, and arteries. ⁴

The primary goal of radiation therapy is to deprive cancer cells of their multiplication potential by maximizing the radiation dose to abnormal cancer cells while minimizing exposure and damage to normal cells. ⁵ However, patients undergoing radiation therapy for thoracic cancers are inevitably exposed to a certain amount of radiation in the esophagus during treatment. Radiation therapy causes DNA damage, which consequently activates signaling pathways and induces stress-induced pro-inflammatory cytokines. This induces inflammation, edema, or erosion of the mucosal surface and irritation of the esophageal lining, leading to the onset of radiation-induced esophagitis (RIE). ⁶ The factors affecting the incidence rate and severity of RIE include the total dose of radiation therapy administered, fraction size, volume of the tissue irradiated in the esophagus, and the use of concomitant chemotherapy. ⁷ Symptoms of RIE include pain, heartburn, nausea, anorexia, dysphagia/odynophagia, substernal discomfort, and the sensation that food is stuck, severely impacting the quality of life of patients. In severe cases, esophageal stricture can occur, leading to dehydration, malnutrition, and weight loss and may become life-threatening.^5,6,8,9

In current clinical practice for RIE treatment, symptomatic management is applied only when symptoms manifest, and analgesics such as lidocaine or morphine, antibiotics and antifungal agents, glucocorticoid hormones, vitamin supplements, and topical anesthetics are used to reduce pain. ⁸ However, the effects of these methods are unclear, treatments are not effective for reducing RIE occurrence, and standard medication care may cause adverse events (AEs). Therefore, alternative treatments over conventional symptomatic management, specifically herbal medicine is gaining traction as an alternative to conventional symptomatic management.

Although several clinical studies on herbal medicinal treatment for RIE exist, no prior systematic review has examined the therapeutic and prophylactic effects. Therefore, the first systematic review was conducted, guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA), to establish clinical evidence for using herbal medicines against RIE.

Methods

Data Sources and Searches

To identify and select eligible studies for inclusion in this review, a literature search was conducted across the following databases: 3 databases for publications in English, PubMed, Ovid-EMBASE, and Cochrane Central Register of Controlled Trials (CENTRAL); 3 portal site to databases in China: the China National Knowledge Infrastructure (CNKI), the Wanfang Database and the China Science and Technology Journal Database (VIP).

The keywords used for the literature search in English language-based databases were as follows: (neoplasm OR cancer OR carcino OR malignan* OR tumor OR tumour) AND (radiotherapy OR radiotherap* OR targeted radiotherap* OR radiation OR Irradiation OR external beam OR localized therapy) AND (traditional Korean medicine OR traditional Chinese medicine OR traditional oriental medicine OR Kampo medicine OR herb OR decoction OR botanic OR Drugs, Chinese Herbal OR Herbal Medicine OR Medicine, Kampo OR Medicine, Korean Traditional OR Medicine, Chinese Traditional OR Plants, Medicinal) AND (Esophagitis OR Esophageal stricture OR Esophagus OR Esophag*). This literature search was conducted on June 5, 2024. The search strategy was adjusted for Chinese and Japanese databases, using keywords related to “cancer” and “radiotherapy” while remaining within the basic search framework. A detailed description of the search strategy can be found in Supplemental Table 1.

Study Selection

Inclusion & Exclusion Criteria

This review included randomized controlled trials (RCTs) evaluating the clinical effects of herbal medicine on radiation-induced esophagitis (RIE) in patients with thoracic cancers. The study population consisted of patient who either began a radiation therapy schedule or developed RIE during radiation therapy.

Only RCTs using oral herbal medicine as an intervention were considered for analysis, specifically in the forms of decoctions and extracts. However, commercially available oral patent herbal medicines, including those containing insect-derived extracts such as Periplaneta americana, and injectable herbal medicines were excluded from the interventions. These products were excluded due to concerns regarding limited generalizability to international clinical practice, as insect-derived substances are rarely used in herbal medicine practices outside of China.

Control groups consisted of studies that employed usual care methods in conventional medicine for the treatment and prevention of RIE, which typically involve symptomatic management using proton pump inhibitors, analgesics, corticosteroids, or antibiotics, or no intervention for RIE prevention. Studies comparing herbal medicines alone (eg, herbal medicine vs herbal medicine) were excluded.

The outcomes evaluated included the incidence rate of RIE (≥grade 2), time to onset of RIE, duration of RIE symptoms, effective rate, and adverse events. The effective rate among patients with RIE was defined as the proportion classified as having a “markedly effective” or “effective” response to treatment according to the original study criteria.

Studies published in English, Chinese, Korean, or Japanese were eligible for inclusion.

The exclusion criteria were as follows: studies on non-human subjects, publications investigating adverse events of radiation therapy that did not include RIE, and interventions not involving herbal medicine.

Data Extraction

Two researcher (HJJ and LYJ) independently screened studies in 2 rounds, first by assessing the titles and abstracts for potential inclusion. EndNote was used for reference management and duplicate removal, and Excel was used to track inclusion and exclusion decisions during the screening process. Following this, both independently reviewed the full text of selected articles to confirm eligibility. Discrepancies between the 2 were resolved through discussion and consensus. After article selection, data such as the author, year of publication, sample size (male/female), interventions in the treatment group and control group, and types of cancers (Table 1 and Supplemental Table 2) were extracted from included studies using a standardized Excel spreadsheet. Based on the treatment strategies employed in the included studies, we further classified the interventions into 2 types for subgroup analysis: (1) “sole intervention,” where herbal medicine was used as the only treatment in the experimental group and compared with usual care in the control group; and (2) “add-on,” where herbal medicine was used in combination with usual care in the experimental group and compared with usual care alone. This classification allowed for clearer interpretation of the comparative effects of herbal medicine in different clinical contexts.

Table 1.

Characteristics of Included Studies for the Treatment of Radiation Induced Esophagitis.

No.	Study	Cancer type	Age	Treatment Group (M,F)	Control group (M,F)	Treatment group intervention*	Control group intervention	Outcome
Sole
1	Wang 2021	Acute RIE	TG: 68.00 ± 13.10	30 (20,10)	30 (22,8)	Qingwei powder (清胃散)	Oral intake (lidocaine, dexamethasone, gentamicin, saline)	Effective rate
			CG: 69.20 ± 34.02
2	Huang 2020	Acute RIE	59. 62 ± 4. 39	40 (25, 15)	40 (26, 14)	Herbal decoction (自拟益气养阴解毒汤)	Gargle (Lidocaine, dexamethasone, gentamicin, Vit B6)	Effective rate
3	Lu 2021	Acute RIE	TG: 61.43 ± 6.27	32 (17, 15)	32 (15, 17)	Herbal decoction	Oral intake (Gentamicin, dexamethasone, lidocaine, saline)	RIE duration
			CG: 65.86 ± 13.99
4	Wang 2007	RIE	Total 55.62	72 (NR)	72 (NR)	Herbal decoction based on pattern differentiation	Oral intake (prednisolone) or intravenous (dexamethasone)	Effective rate
5	Tian 2007	RIE	TG: 55.18	38 (23, 15)	38 (21, 13)	Herbal decoction (滋阴解毒汤)	Oral intake (dexamethasone, gentamicin, 25% glucose)	Effective rate
			CG: 56.02
6	Huang 2006	RIE	TG: 46.1	32 (25, 7)	30 (25, 5)	Herbal decoction (血府逐瘀汤)	Cisapride 10 mg, Oral intake (gentamicin, dexamethasone, saline)	Effective rate
			CG: 48
7	Zhang 2006	RIE	NR	30 (20, 10)	25 (15, 20)	Herbal decoction (滋阴清热止痛汤)	Oral intake (saline, dexamethasone, lidocaine, gentamicin)	Effective rate
8	Zhou (a) 2004	Acute RIE	NR	80 (NR)	78 (NR)	Herbal decoction	Oral intake (saline, gentamicin, fluocinolone, procaine, anisodamine, ephedrine)	Effective rate
9	Zhang 1999	RIE	NR	30 (NR)	30 (NR)	Herbal power (双料喉风散)	Oral intake (Aluminum hydroxide, prednisolone)	Effective rate
10	Zhou 2016	RIE	TG: 61.47 ± 16.52	40 (22, 18)	40 (21, 19)	Herbal decoction	Dioctahedral smectite, gentamicin	Effective rate
			CG: 61.08 ± 16.11
11	Zhang 2009	Acute RIE	TG: 61.8	54 (31, 23)	51 (35, 16)	Herbal decoction	Intravenous (cefuroxime sodium, dexamethasone)	Effective rate
			CG: 60.2
Addon
1	Zhao 2009	Acute RIE	Total: 58	69 (NR)	69 (NR)	Herbal decoction (银翘青黛汤) + CG treatment	Dioctahedral smectite	Effective rate
2	Cui 2016b	Acute RIE	TG: 57	50 (26, 24)	50 (27, 23)	Herbal decoction + CG treatment	Oral intake (Saline, gentamicin, lidocaine, anisodamine)	Effective rate
			CG: 56
3	Ren 2014	Acute RIE	NR	32 (NR)	31 (NR)	Herbal decoction (增液汤) + CG treatment	Oral intake (saline, lidocaine, dexamethasone, gentamicin)	Effective rate

Abbreviations: NR, not reported; M, male; F, female

^*Detailed herbal composition were presented in Supplemental Table 3.

Assessment of Risk of Bias (ROB)

Assessment of risk of bias (ROB) for studies included in this review was performed based on 6 domains of criteria specified in the Cochrane risk-of-bias tool. ¹⁰ Each domain of ROB was assessed into levels of “low,” “unclear,” or “high.” Two independent researchers (HJJ and LYJ) individually assessed the ROB. Disagreements between the researchers in the assessment process were resolved through a re-discussion of the assessment results.

The risk was assessed as “low” when the method of random sequence generation was clearly described. Similarly, the risk was assessed as “low” for allocation concealment when methods such as the use of opaque and sealed envelopes was explicitly specified. For blinding of participants, the risk was assessed as “low” when blinding was explicitly stated in the text of the selected literature and when the study design indicated that blinding was likely implemented. The risk was assessed as “low” when blinding of the assessor was ensured the outcome assessment was undertaken by an independent third-party assessor. For incomplete outcome data, the risk was assessed as “low” if dropouts or follow-up losses were explicitly reported and addressed adequately, “unclear” if these issues were not mentioned, and “ high” if data omissions due to follow-up losses were described (eg, missing data were excluded from the analysis without appropriated handling). For selective reporting, the risk was assessed as “low” if safety outcomes related to herbal medicine, such as AEs were reported. If AEs related to herbal medicine were not reported, the risk was classified as “unclear.” For other biases, studies with concurrent chemotherapy and radiation therapy were assessed as “unclear” due to potential confounding effects on RIE and other outcomes. Among the 20 studies that included concurrent chemotherapy, the risk of bias was rated as “unclear,” whereas studies in which radiation therapy was the sole intervention were assessed as having a “low” risk of bias.

Data Analyses

All values from the results of the included studies were extracted and presented as the mean and standard deviation or the total number of patients and number of patients for each case. Outcome measurements were analyzed at each time point after treatment. For dichotomous variables, relative risk (RR) was calculated, while for continuous variables, the mean difference (MD) was used in pooled analyses. Meta-analysis was performed using a random-effects model with 95% confidence intervals, applying the Mantel-Haenszel or inverse variance methods. Heterogeneity across the included studies was assessed using the chi-square test and Higgins and Thompson I² statistic. ¹¹ Subgroup analyses were conducted to address and resolve the identified heterogeneity.

Certainty of Evidence

The certainty of evidence in the meta-analysis was evaluated according to the methodology and criteria specified in the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) approach. The certainty of evidence was categorized into 4 tiers: high, moderate, low, and very low. GRADE methodology used the results from assessments across the following domains: ROB, inconsistency, indirectness, imprecision, publication bias, substantial effect, and dose response.^{12 -14}

Results

Study Selection

In the first and second rounds of literature searches across the databases, 2341 published articles were identified as follows: PubMed (70), EMBASE (56), Cochrane (100), CNKI (844), WanFang (494), and VIP Journals Database (777). After excluding 107 duplicates, 2234 articles were included in the abstract review stage, of which 2094 publications were irrelevant to RIE and were excluded from the analysis. After reviewing 140 full-text articles for eligibility, 59 were excluded for the following reasons: they were non RCTs, oral herbal medicine was not used as the intervention in the treatment group, or herbal medicine was used in the control group instead of usual care. Therefore, 81 articles were included in the systematic review (Figure 1). Figure 1 presents the PRISMA-based flow diagram of the literature search process. Further details regarding the study characteristics included in this review are summarized in Table 1 and Supplemental Table 2.

Figure 1.

PRISMA flow diagram of the systematic search.

Study Characteristics

A total of 7283 patients were included in 81 RCTs included in this review.^{15 -95} Among these, 67 RCTs included focused on the prophylactic effects on herbal medicine in patients undergoing radiation therapy for thoracic cancers, such as lung or esophageal cancer. Fourteen RCTs specifically investigated its therapeutic effects in patients who had already developed RIE as a result of radiation therapy.

Regarding outcome measures used in the selected studies, common indicators were employed due to the consistent clinical context of RIE. These indicators included the incidence rate of RIE, time to incidence, duration of RIE, and effective rate. Among these, the incidence rate was reported in 63 studies, covering RIE grades 1 to 4. Thirteen RCTs measured the time to incidence of RIE during radiation therapy in patients with thoracic cancers, and 1 RCT assessed the duration of RIE. The effective rate was reported in 13 RCTs.

All 81 selected RCTs used herbal medicine as an intervention. Among these, Yinqiao Qingdai Decoction (银翘青黛汤) and Zhuye Shigao Decoction (竹叶石膏汤) were the most commonly employed. However, due to the wide variety of prescriptions included, their individual proportions were relatively low. All studies administered herbal medicines orally, prepared as decoctions or extractions or pastes for patients. Commonly used medicinal herbs included Ophiopogon japonicus, Glycyrrhiza uralensis, Rehmannia glutinosa, Scrophularia ningpoensis, and Astragalus membranaceus among others. Details are available in Supplemental Tables 3 and 4.

Among the 67 studies investigating the prevention of RIE, 32 predominantly compared radiotherapy combined with herbal medicine as the intervention group versus radiotherapy alone as the control group. Additionally, 16 studies frequently utilized oral administration of dexamethasone, lidocaine, or gentamicin in the control group, while other approaches included gargling and the use of dioctahedral smectite.

In contrast, among the 14 studies focusing on the treatment of RIE, 9 commonly employed oral administration of dexamethasone, lidocaine, or gentamicin in the control group (Table 1).

ROB Assessment

Detailed information on the ROB assessment is shown in Figure 2. In the domain of random sequence generation, 29 RCTs (35.8%) reported adequate random sequence generation processes, and these studies were assessed as low risk. The remaining 52 did not provide clear description of this domain, and their ROB was assessed as unclear. In term of allocation concealment, 3 RCTs used appropriate allocation methods and were assessed as low risk, while the remaining studies did not provide clear descriptions for this domain, resulting in an unclear risk assessment. Due to the nature of the study designs, only 1 RCT successfully implemented participant and personnel blinding and was assessed as low risk. The remaining studies had designs that made blinding impossible, and thus, they were all assessed as high risk. For outcome assessment, most studies were assessed as high risk. However, 1 study merely mentioned blinding without specifying the methods used, resulting in an unclear risk assessment. Another study with detailed blinding methods was assessed as low risk. In the domain of incomplete outcomes, 3 studies appropriately addressed issues such as dropouts and were assessed as low. Two studies failed to properly handle dropouts or missing data, resulting in a high ROB assessment. The remaining studies did not provide sufficient details on how missing data or dropouts were managed and were therefore assessed as unclear ROB. In the domain of selective reporting, 29 studies that addressed adverse events were assess as low ROB. The remaining studies did not mention AEs, leading to an unclear ROB assessment. Regarding the domains of other biases, none reported funding; however, evaluation with clear evidence was lacking, resulting in “unclear” ROB assessments for all studies.

Figure 2.

Risk of bias of included studies.

Quantitative Data Synthesis

Effect of Herbal Medicine Versus Control on RIE Prevention

To compare the effects of herbal medicine and usual care of conventional medicine (usual care) in RIE prevention, the overall incidence rate of RIE was analyzed in 63 RCTs. Meta-analysis indicated that herbal medicine was associated with a greater preventive effect on RIE compared to usual care, but heterogeneity was also confirmed in the analysis (RR: 0.69; 95% CI 0.63-0.76: heterogeneity: P < .00001, I² = 96%; Figure 3(3.1)). Subgroup analyses were performed based on addon versus sole interventions and chemotherapy use, but heterogeneity persisted across all analyses.

Figure 3.

Effect of herbal medicine versus control on prevention outcomes.

In the 38 RCTs comparing herbal medicine and usual care, the incidence rate of RIE ≥ Grade 2 was analyzed comparatively. A meta-analysis comparing the incident rates of moderate to severe RIE between herbal medicine and usual care demonstrated a significantly superior effect of herbal medicine in preventing the development of severe esophagitis; however, heterogeneity was also identified (RR: 0.51; 95% CI: 0.44–0.60, heterogeneity: P < .00001, I² = 75%; Figure 3(3.2)).

Thirteen RCTs measured the time to RIE incidence. The meta-analysis showed that herbal medicine had a significantly extended the time to incidence; however, heterogeneity was also identified (MD: 5.09; 95% CI: 3.96–6.21, heterogeneity: P < .00001, I² = 98%; Figure 3(3.3)).

Effect of Herbal Medicine Versus Control on Treatment (effective rate)

Fifteen studies measured the effective rates for the treatment of RIE. A meta-analysis confirmed the significant superiority of herbal medicine in the effective rate compared to usual care; however, heterogeneity was identified (RR: 1.29, 95% CI: 1.21–1.38, heterogeneity: P = 0.27, I² = 17%; Figure 4(4.1)).

Figure 4.

Effect of herbal medicine versus control on treatment outcomes.

Effect of Herbal Medicine Versus Control on Treatment (Duration)

One RCT reported the duration of RIE symptoms, showing that herbal medicine significantly reduced symptom duration compared to usual care (MD: −4.53, 95% CI: −7.93 to −1.13; Figure 4(4.2)).

Adverse Events

Among the 81 included RCTs, 28 reported adverse events (AEs) related to herbal medicine treatment, while the remaining studies either did not report AEs or only documented toxic reactions associated with radiation therapy or concurrent chemotherapy. Of the 28 studies, 13 reported no AEs, including three ⁷² that specifically stated the absence of elevated hepatic or renal markers. Ren et al ⁶⁰ reported 1 case each of elevated liver enzymes and creatinine in the herbal medicine group, compared to 2 and 1 cases, respectively, in the control group. Similarly, Zhang et al ⁸⁸ observed elevated liver enzymes in 11 experimental and 16 control cases. Among the 13 studies that reported gastrointestinal AEs, most found fewer gastrointestinal symptoms in the herbal medicine group compared to the control group. However, Chen and Liang ²⁴ reported similar rates between groups, with 49 cases in the herbal medicine group and 47 in the control group, indicating no significant difference in gastrointestinal AEs.

Certainty of Evidence

The certainty of evidence for the outcome measures of the studies included in the systematic review and the meta-analysis was evaluated using the GRADE method. When comparing herbal medicine with usual care, the certainty of evidence for the overall RIE incidence rate, the incidence rate of RIE incidence ≥grade 2, and the time to incidence was downgraded by 2 levels due to ROB and substantial inconsistency. The certainty of evidence was rated “Moderate” for effective rate of RIE, owing to ROB. For the duration of RIE, the certainty of evidence was rated as “very low” due to ROB and very serious imprecision, primarily because it was based on a single RCT. The certainty of evidence for each outcome is summarized in Table 2.

Table 2.

Meta-analysis of Outcomes and Certainty of Evidence.

Variable	Overall effect						Studies (N)	Sample size (N)	Quality of evidence
	RR or MD	95% CI	P (Heterogeneity)	I 2	P (Overall effect)	Statistical method
Herbal medicine versus control (RIE prevention)
Overall incidence rate (Grade 1-4)	0.69	0.63, 0.76	<.00001	96	<.00001	Random Mantel-Haenszel	63	5849	Lowa,b
Gr 2 + incidence rate	0.51	0.44, 0.60	<.00001	75	<.00001	Random Mantel-Haenszel	38	3186	Lowa,b
Time to incidence (days)	5.09	3.96, 6.21	<.00001	98	<.00001	Random Inverse Variance	13	915	Lowa,b
Herbal medicine versus control (treatment)
Effective rate	1.29	1.21, 1.38	.27	17	<.00001	Random Mantel-Haenszel	13	1181	Moderate a
Duration of RIE (days)	-4.53	−7.93, −1.13	-	-	.009	Random Inverse Variance	1	64	Very lowa,c

Abbreviations: Gr, grade; MD, mean difference; RIE, radiation induced esophagitis.

^aRisk of bias.

^bInconsistency.

^cVery serious imprecision.

Discussion

This systematic review evaluated the effect of herbal medicine on the prevention, treatment and safety of RIE. A meta-analysis was conducted on RCTs comparing herbal medicine to usual care. The results confirmed that herbal medicine is effective in reducing the incidence rate of RIE, delaying its onset, improving treatment efficacy, and shortening its duration.

RIE is one of the most common AEs in patients undergoing radiation therapy for thoracic cancers such as lung, esophageal, and breast cancers. ⁶ In severe cases, it may lead to esophageal stricture or fibrosis, causing symptoms such as pain, indigestion, nausea, and dysphagia/odynophagia. These symptoms can contribute to reduced food intake, leading to dehydration, malnutrition, weight loss, and can even lead to death.^5,6,8,9 Currently, symptomatic treatment is the standard approach for RIE management. However, it does not reduce disease incidence and has limited efficacy, leading to increasing interest in alternative treatments, including herbal medicine.

Herbal medicines are administered to RIE patients based on Traditional Chinese Medicine (TCM) and Korean Medicine (KM) principles, which emphasize heat-clearing, toxicity elimination, body fluid production, esophagus strengthening, phlegm resolution, and dampness removal. Ophiopogon japonicus, the most commonly included herbal component in the reviewed prescriptions, aligns with TCM principles by clearing heat and nourishing body fluids. Its primary components exhibit anti-inflammatory properties, contributing to the treatment of various inflammatory diseases and enhancing immunomodulatory functions.^{96 -98} Glycyrrhiza uralensis, the second most frequently included herb, is widely used to enhance the efficacy of other herbal formulations. ⁹⁹ Modern pharmacological studies have found that its major component, saponin, possesses strong anti-inflammatory and immunomodulatory properties. ¹⁰⁰ Rehmannia glutinosa, the third most common herb in the reviewed prescriptions, is traditionally known for its heat-reducing, blood-cooling, and body fluid-nourishing effects. ¹⁰¹ Modern pharmacological studies have reported its ability to inhibit pro-inflammatory gene expression and cool blood.^102,103

Previous studies have explored the use of herbal medicine for radiation-related injuries. For instance, Jaungo has been applied externally for radiation dermatitis, ¹⁰⁴ while oral herbal medicine has been used to alleviate symptoms of radiation-induced lung injury by reducing inflammation and fibrosis in lung tissue.^105,106 Several studies have demonstrated the efficacy, including Astragalus membranaceus, Glycyrrhizae uralensis, Ophiopogogon japonicus, and Rehmannia glutinosa.^105,106

No severe AEs related to herbal medicines were reported in the 81 RCTs included in the review. Among them, 28 documented AEs, while others either did not report AEs or instead attributed toxic reactions to radiation or chemotherapy. Of the 28 studies, 13 reported no AEs, and most found fewer gastrointestinal symptoms in the herbal medicine group compared to the control group. However, some studies reported similar occurrences in both groups. A few studies observed cases of elevated liver enzymes or creatinine, however, these occurrences were comparable between groups, suggesting that they were not specifically attributable to herbal medicine. Overall, the risk of AEs in the herbal medicine group was generally lower or similar to that in the control group, with mild AEs such as vomiting and nausea resolving quickly.

Despite these findings, this review identified significant methodological limitations in the included studies. Only 35.8% of RCTs reported adequate random sequence generation, and allocation concealment was clearly described in only 3 studies, introducing potential selection bias. Blinding of participants and personnel was largely unfeasible due to study designs, leading to a high risk of bias. Similarly, blinding of outcome assessment was inadequate, further increasing the risk of bias. These limitations highlight the need for well-designed RCTs to strengthen the evidence base for the effectiveness of herbal medicines in RIE.

Another limitation is the geographical concentration of the studies. Despite a comprehensive search across global databases, all included studies were conducted in China. This reflects the predominant use of herbal medicine in East Asian countries. However, to enhance generalizability of findings, future RCTs should be conducted in diverse settings.

Additionally, heterogeneity observed in some meta-analyses presents a challenge in interpreting the results. The studies were analyzed under the broad category of herbal medicine; however, heterogeneity arose due to variations in herbal composition and treatment regimens. Moreover, because standard treatment for RIE primarily focuses on symptom management rather than a fully standardized protocol, there were variations in control group treatments, particularly in studies conducted in China where oral steroids and antibiotics were frequently used. It remains uncertain whether these control treatments align with globally accepted usual care, raising concerns about generalizability.

This study represents the first systematic review in herbal medicine for RIE. Using rigorous methodology and meta-analyses, it evaluated key outcomes, including incidence rate, time to onset, effective rate, and duration. However, the overall certainty of evidence remains limited due to risk of bias in multiple domains, highlighting the need for well designed, large-scale RCTs. Future studies with lower risk of bias are essential to generate high-quality evidence, enabling a clear verification of herbal medicine’s effectiveness for RIE. This research highlights a potential alternative treatment for RIE, an increasingly prevalent condition due to rising cancer incidence.

Conclusion

This systematic review demonstrated that herbal medicine appears to be effective in reducing the incidence of RIE, delaying its onset, alleviating symptoms, shortening its duration in patients undergoing radiation therapy for thoracic malignancies, including esophageal, lung, mediastinal, and breast cancers. Additionally, herbal medicine was associated with fewer adverse events compared to usual care.

However, the overall certainty of evidence remains low due to methodological limitations, including risk of bias. To establish the clinical efficacy for herbal medicine in both preventing and treating RIE, well-designed, large-scale RCTs with rigorous methodologies and lower risk of bias are necessary Future research should focus on generating high-certainty evidence regarding incidence rates and treatment effectiveness to support the integration of herbal medicine into clinical practice.