Cervical cancer (CC) is one of the most prevalent gynecologic malignancies globally. Radiotherapy serves as a cornerstone treatment for high-risk or inoperable cervical cancer patients (1, 2). Although CC primarily affects women aged 35–55 years, about 20% of cases occur in older adult patients (≥60 years) (3). Due to age-related physiological decline and comorbidities, older adult patients exhibit distinct radiotherapy tolerance and therapeutic responses in comparison with younger cohorts (4). Acute radiation enteritis (ARE) is a frequent complication of pelvic radiotherapy, manifesting as abdominal pain, diarrhea, mucoid/bloody stools, and increased infection risk. Severe cases may lead to life-threatening intestinal dysfunction complications (e.g., hemorrhage, perforation), which need treatment interruptions or surgical interventions. Given the high incidence of cervical cancer in older adult patients and their unique vulnerability to treatment complications, particular attention must be paid to ARE.

Current evidence indicates that ARE pathogenesis involves multifactorial interactions between clinical factors (e.g., age, hypertension, diabetes, prior abdominal/pelvic surgery) (4–6), dosimetric parameters (e.g., radiation dose/volume metrics) (7), and systemic inflammation markers (8). Some studies have also shown that the gut microbiota modulates the initiation and progression of radiation enteritis (RE) (9). However, no single factor reliably predicts ARE occurrence. Given the complexity of RE, it is really challenging for a single risk factor to accurately predict its occurrence. Therefore, multifactorial predictive models are urgently needed to guide personalized risk stratification and intervention strategies in older adult CC patients.

Existing risk prediction models are predominantly developed based on younger populations or mixed cohorts, lacking sophisticated evaluation tools tailored to older adult patients. This study retrospectively analyzed independent influencing factors of grade ≥2 ARE in older adult CC patients undergoing pelvic radiotherapy. Based on identified risk factors, we developed a nomogram model to aid in personalized risk assessment and provide a basis for further targeted intervention strategies.

In modern gynecological oncology radiotherapy, despite the implementation of advanced techniques such as intensity-modulated radiation therapy (IMRT) and volumetric modulated arc therapy (VMAT), which significantly reduce radiation doses to organs at risk (OARs) (10–12), the incidence of ARE remains high in clinical practice (35.5–75.0%) (13–15). This phenomenon is particularly prominent in older adult patients, whose unique physiological characteristics and underlying comorbidities lead to significantly reduced tolerance to radiotherapy and a markedly increased risk of ARE compared to younger patients (4). In contrast to previous studies, this article investigates the indicators for predicting the risk of RE in older adult patients by focusing on their specific risk factors.

Our study enrolled 251 older adult cervical cancer patients (age ≥60 years) who underwent radiotherapy or concurrent chemoradiotherapy. The results showed that the overall incidence of ARE was as high as 82%, with 61% of cases being grade ≥2. Compared to previous reports of ARE incidence (54–75%) and grade ≥2 ARE incidence (29–56%) (16–18), the toxicity rates in our cohort were significantly higher. Multivariate analysis confirmed that age was an independent risk factor for grade ≥2 ARE. Notably, a clear age gradient effect was observed even within the older adult patient population. The relationship between age and radiotherapy toxicity remains controversial (19). However, the results of this study support the use of age as an important reference indicator for assessing treatment risks and determining radiotherapy doses in clinical practice.

Hypertension and diabetes, common comorbidities in older adult patients, were identified as independent risk factors for grade ≥2 ARE in this study. The underlying pathological mechanisms may involve reduced gut microbiota diversity, abundance, and gene counts in hypertensive patients compared to healthy individuals, with a decrease in beneficial bacteria and an increase in pathogenic microorganisms (20). Dysbiosis of the gut microbiota can lead to intestinal endothelial dysfunction, promote vascular sclerosis, and impair local tissue perfusion, further exacerbating tissue ischemia (21). This delays the repair of radiation-induced mucosal damage, thereby increasing the risk of severe RE. In diabetic patients, abnormal glucose metabolism has been shown to correlate positively with inflammatory responses (22). This synergizes with the non-specific inflammation triggered by radiation-induced intestinal injury, exacerbating oxidative stress and impairing tissue repair, creating conditions conducive to RE (23). Additionally, diabetic microangiopathy increases the risk of post-radiation intestinal damage (24). These pathological changes collectively make patients with hypertension and diabetes more sensitive to radiation damage, significantly increasing the risk of ARE.

In cervical cancer radiotherapy, the rectum and small intestine are the primary dose-limiting organs. When the cumulative dose of pelvic external irradiation exceeds 45–50 Gy, the risk of acute and late intestinal toxicity increases significantly (25). Therefore, strict control of rectal dose and volume in radiotherapy planning is crucial for reducing gastrointestinal toxicity. Compared to 3D conformal radiotherapy, IMRT has been proven to significantly reduce the incidence of grade 2 ARE in patients with late-stage rectal cancer (26). The small intestine, being highly sensitive to radiation, exhibits a close correlation between radiation dose and the occurrence of ARE. Studies have shown that prone positioning during gynecological radiotherapy can effectively reduce small intestine exposure, as this position allows the small intestine and parts of the colon to naturally shift away from the target area (18). Additionally, Chen et al. found that bladder filling status is closely related to the volume of small intestine within the planning target volume (PTV) (27). However, there is no consensus on the relative importance of various predictive factors. Through systematic analysis of dosimetric parameters for the rectum and small intestine, our study first confirmed that mean rectal dose (Dmean_R) is an independent risk factor for grade ≥2 ARE in older adult cervical cancer patients, consistent with the findings of Huang et al. (28). Notably, our study found no significant differences in small intestine dosimetric parameters between the two groups, aligning with Ma et al.’s conclusions (29). However, this result contrasts with some literature, which may be attributed to differences in radiotherapy techniques, bladder filling status, and patient positioning across studies. Therefore, in optimizing radiotherapy plans for older adult cervical cancer patients, priority should be given to rectal dose control, particularly limiting the mean dose (Dmean_R). At the same time, despite the lack of statistical significance for small intestine dosimetric parameters in this study, given the heterogeneity among studies, attention should still be maintained on small intestine dose parameters in clinical practice.

This study systematically evaluated the predictive value of various serum biomarkers for grade ≥2 ARE, with the lactate dehydrogenase to albumin ratio (LAR) demonstrating the best predictive performance (AUC = 0.694). As a composite indicator of the ratio of LDH to albumin, LAR integrates two key indicators to reflect the patient’s radiation tolerance. LDH, a critical enzyme in anaerobic glycolysis, directly reflects the degree of cellular damage and local hypoxia. Specifically, reactive oxygen species (ROS) released by inflammatory cells exacerbate oxidative damage, and ROS further enhance the inflammatory response. These interactions promote cellular damage and cell membrane rupture, which subsequently leads to the release of LDH. Studies have shown that elevated serum LDH levels correlate positively with the severity of tissue inflammatory injury. Serum LDH and lactate are risk factors for mortality in patients with severe inflammatory diseases (30). Albumin, an essential nutritional marker, also possesses anti-inflammatory and microcirculation-stabilizing properties (31). Hypoalbuminemia exacerbates radiation-induced microvascular damage and inflammatory responses. Thus, LAR levels reflect systemic inflammatory status. When radiation causes intestinal injury, released inflammatory factors disrupt the gut microenvironment and inhibit repair processes. Existing research indicates that LAR is an independent risk factor for poor prognosis in patients with lower respiratory tract infections (32), severe infections (33), and non-small cell lung cancer (34). To date, the role of LAR as a potential biomarker for radiation enteritis has not been explored. Our results demonstrated that LAR exhibited superior predictive performance (AUC = 0.694), better than other indicators such as diabetes (AUC = 0.600), although among the 54 diabetic patients, 45 exhibited higher-grade ARE. This may be because diabetes is a binary variable—while strongly associated with severe ARE (45/54 cases), it may not fully distinguish between the two groups. In contrast, LAR is a continuous biomarker that integrates LDH and albumin, enabling more sensitive risk stratification and capturing the inflammatory and nutritional status of all patients. This means that this simple and easily accessible indicator can effectively identify high-risk patients and provide important references for clinical decision-making.

There are many factors that influence ARE, and even experienced clinicians find it difficult to predict it early. Therefore, our nomogram, which includes clinical characteristics, dosimetric parameters, and novel biomarkers, was constructed based on five independent risk factors. We found that our predictive model demonstrated favorable predictive performance, with improved sensitivity and accuracy compared to individual predictors. Meanwhile, the calibration curve and decision curve analysis (DCA) indicated that the nomogram has good clinical applicability, thus providing more accurate risk stratification for older adult patients and avoiding overtreatment or undertreatment.

However, our study has several limitations. First, as a single-center retrospective study, the model was only internally validated within the enrolled cohort. Second, we only focused on the most clinically significant and common parameters. The limited sample size may introduce selection bias and affect the study’s reliability. Future research with larger sample sizes and external validation are warranted to further verify the model’s performance and enhance its generalizability.

In conclusion, age, hypertension, diabetes, Dmean_R, and LAR are independent risk factors for grade ≥2 ARE in older adult cervical cancer patients who received radiotherapy. A nomogram prediction model was established based on these factors. The developed nomogram integrates routine clinical indicators, balancing predictive accuracy and practicality, and serves as a valuable tool for improving radiotherapy safety in this vulnerable population.