A nomogram predicting dysphagia risk after adjuvant iodine-131 therapy in patients with papillary thyroid carcinoma Authors: Hanji Jiang, Jiayu Chen, Hui Zhu, Zhiyong Li Published: 2026 ABSTRACT ObjectivesTo identify clinical risk factors for dysphagia after iodine-131 (131I) therapy in patients with papillary thyroid carcinoma (PTC) and develop a validated risk-prediction nomogram.MethodsAdult PTC patients treated with 131I at the Department of Nuclear Medicine, Affiliated Hospital of Xuzhou Medical University, between January 2022 and October 2024 were retrospectively reviewed. All patients underwent total or partial thyroidectomy, had satisfactory postoperative wound healing, and had histopathologically confirmed PTC. Baseline clinical data, laboratory parameters, and pretreatment salivary gland function assessed by dynamic pertechnetate single-photon emission computed tomography/computed tomography were collected. Swallowing function was evaluated before treatment and at 6-month follow-up using the MD Anderson Dysphagia Inventory; dysphagia was defined as a total score less than 50. Variables significant in univariate analysis were entered into backward stepwise multivariate logistic regression. Independent predictors were used to construct a nomogram, which was assessed by area under the curve (AUC), bootstrap calibration, and decision curve analysis.ResultsAmong 201 patients, 68 (33.8%) developed dysphagia. Higher cumulative 131I dose, smoking history, and baseline salivary gland dysfunction were independent predictors of dysphagia. The nomogram incorporating these factors showed excellent discrimination, with an AUC of 0.913, good calibration with a bootstrap mean absolute error of 0.013, and favorable clinical net benefit.ConclusionHigher 131I dose, smoking history, and pre-existing salivary gland dysfunction independently predict dysphagia after 131I therapy in PTC patients. This nomogram may support individualized risk assessment and preventive counseling before treatment. Introduction Thyroid cancer is a common endocrine malignancy whose reported incidence has increased dramatically over recent decades, although much of the rise has been attributed to greater detection of small papillary tumors and changing diagnostic practices [1,2]. The vast majority of newly diagnosed differentiated thyroid cancers are papillary thyroid carcinoma (PTC), which accounts for roughly 80–90% of the cases in contemporary series [3,4]. Major guidelines endorse a stepped approach in the management of PTC, and the risk-adapted intervention spans from active surveillance to radical surgery [5,6]. Compared with patients with other cancers, patients with PTC generally have a much better prognosis [7]. After thyroidectomy, radioactive iodine (131I) therapy is typically used to ablate remnant thyroid tissue, treat microscopic residual disease, and manage known metastatic deposits, thereby reducing recurrence [8]. However, its side effects are not negligible. Its therapeutic efficacy relies on the sodium-iodide symporter (NIS), which is not only highly expressed in thyroid tissue but also present in nontarget tissues, including salivary glands, the gastric mucosa, and lacrimal glands [9,10]. Consequently, off-target radiation uptake in those sites causes sialadenitis (salivary gland inflammation), xerostomia (dry mouth), nausea, taste dysfunction, and so forth [11,12]. Among these complications, dysphagia (difficulty swallowing) is particularly troublesome. Robust evidence confirms that dysphagia can lead to reduced nutritional intake, an increased risk of malnutrition and anemia, a significant decline in quality of life, and a negative overall prognosis following surgery [13]. Postradiotherapy dysphagia is common in patients with head and neck cancer [14,15], although dysphagia can also occur in PTC patients before and after thyroidectomy [16,17]. Regarding 131I-induced dysphagia, existing research does not clearly explain specific mechanisms, although some risk factors have been proposed, including higher cumulative 131I activity, sex (female), older age, and a history of sialadenitis or autoimmune disease [18,19]. This lack of clarity presents a clinical challenge, and there are currently no validated tools to predict which PTC patients are more susceptible to dysphagia after receiving 131I. Hence, in this study, we sought to systematically analyze the incidence and risk factors of post-131I dysphagia. Our primary objective was to develop and validate a clinical nomogram to quantify the risk, thus providing a practical tool for early patient assessment, risk stratification, and the implementation of preventive strategies. Methods Study population This retrospective study included PTC patients who received 131I therapy at the Department of Nuclear Medicine, Affiliated Hospital of Xuzhou Medical University, from January 2022 to October 2024. The inclusion criteria were: (a) age greater than or equal to 18 years; (b) total or partial thyroidectomy with satisfactory postoperative wound healing; (c) histopathological confirmation of PTC; (d) baseline salivary gland function imaging performed 2 days before 131I treatment; and (e) complete clinical data. The exclusion criteria were: (a) presence of dysphagia before treatment, as determined by questionnaire screening; (b) history of radiotherapy for head or neck tumors; (c) long-term use of anticholinergic drugs; or (d) severe underlying disease of vital organs (e.g. heart, liver, and kidneys). Smoking history (yes/no) was defined, based on the medical record at admission, as any documented current or former tobacco smoking before 131I therapy. Detailed smoking exposure metrics (e.g. pack-years and time since cessation) were not consistently available for all patients. This study was approved by the Medical Research Ethics Committee of the Affiliated Hospital of Xuzhou Medical University (Approval No. XYFY2023-KL443-01). The committee waived the requirement for written informed consent as the study was retrospective and did not involve any change to clinical treatment. Treatment protocol and follow-up Before 131I administration, all patients stopped taking levothyroxine for 3–5 weeks and maintained a low-iodine diet to achieve a serum thyroid-stimulating hormone level of greater than or equal to 30 mU/L. The 131I dose was determined by the attending clinician based on an assessment of recurrence risk. Supportive medications (e.g. vitamin C) were administered to all patients later the same day of radioiodine therapy completion [20,21]. Patients were followed up for 6 months after therapy. At the follow-up assessment, the occurrence of symptoms including dry mouth, taste changes, parotid gland swelling, and dysphagia was recorded. During the follow-up period, no patients received dysphagia-specific interventions or new medications known to affect swallowing function. Swallowing outcomes were analyzed based on the 6-month follow-up assessment, which was the prespecified endpoint. Because serial instrumental or questionnaire assessments were not uniformly available at interim time points, the present study was not designed to distinguish transient from persistent dysphagia trajectories. Assessments Primary outcome: dysphagia Dysphagia was assessed before and after treatment using the MD Anderson Dysphagia Inventory (MDADI) [22,23], a 20-item instrument (Cronbach’s α = 0.90) evaluating general, emotional, social, and physiological domains. Each item is scored on a 5-point Likert scale, yielding a total score from 20 to 100. There is currently no universally accepted MDADI cutoff for PTC patients after adjuvant 131I therapy, and published thresholds (often derived from head and neck cancer populations) are not directly transferable to the present setting. Therefore, for this study, dysphagia was operationally defined as a total MDADI score of less than 50 points. This operational threshold was selected to identify patients with clinically meaningful swallowing-related quality-of-life impairment; it should only be interpreted as a pragmatic, study-specific definition for risk modeling pending future validation against objective swallowing assessments [e.g. videofluoroscopy (VFSS)/fiberoptic endoscopic evaluation of swallowing (FEES)]. Salivary gland function To determine the baseline salivary gland damage, dynamic salivary gland imaging was performed 2 days before 131I treatment; the instrument was a GE Infinia Hawkeye 4 single-photon emission computed tomography/computed tomography scanner, and the contrast agent was pertechnetate (10 mCi). Images were semiquantitatively interpreted by two senior nuclear medicine physicians. Regions of interest were drawn for the bilateral parotid and submandibular glands, and the adjacent temporal region was used for background correction. The following indicators were calculated: [24,25] • UptakeIndex(UI)=(Smax−B)B×100 % UptakeIndex(UI)=(Smax−B)B×100 % • Excretion Fraction(EF)=(SC,max−Smin)B×100 % Excretion Fraction(EF)=(SC,max−Smin)B×100 % where Smax is the maximum salivary gland uptake count, B is the background count, SC,max is the maximum uptake count after vitamin C stimulation, and Smin is the minimum count poststimulation. Intake impairment was defined as an uptake index less than 50%, and excretion impairment as an excretion fraction less than 30%. A patient was considered to have salivary gland dysfunction if at least one of the four glands exhibited either abnormality [26]. Statistical analysis All statistical analyses were conducted using SPSS 27.0 (IBM Corp., Armonk, New York, USA) and R version 4.5.1 (R Foundation for Statistical Computing, Vienna, Austria). Continuous variables were assessed for normality; normally distributed data were expressed as mean ± SD and compared using independent t-tests, whereas nonnormally distributed data were expressed as median (P25, P75) and compared using the Mann–Whitney U test. Categorical variables were expressed as count (percentage) and compared using the χ2 test or Fisher’s exact test. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors for dysphagia. Variables with a P value of less than 0.050 in the univariate analysis were entered into the multivariate logistic regression model using a backward stepwise method. The independent predictors identified from the final multivariate model were used to construct a nomogram using the ‘rms’ package in R [27]. The model’s performance was assessed for discrimination, calibration, and clinical utility. Specifically, Discrimination (the ability to distinguish between patients with and without dysphagia) was assessed using the area under the receiver operating characteristic (ROC) curve (AUC). Discrimination (the ability to distinguish between patients with and without dysphagia) was assessed using the area under the receiver operating characteristic (ROC) curve (AUC). Calibration (the agreement between predicted probabilities and actual outcomes) was evaluated using a calibration curve generated via 1000 bootstrap resamples. Calibration (the agreement between predicted probabilities and actual outcomes) was evaluated using a calibration curve generated via 1000 bootstrap resamples. Clinical utility was assessed using decision curve analysis (DCA). Clinical utility was assessed using decision curve analysis (DCA). A two-sided P value of less than 0.05 was considered statistically significant. Results Patient characteristics A total of 201 PTC patients were included in the study, comprising 141 females (70.1%) and 60 males (29.9%). The mean age of the cohort was 46.92 ± 12.42 years. Six months after 131I treatment, 68 (33.8%) patients developed dysphagia (Table 1). No patients were lost to follow-up. Table 1Patient characteristicsCharacteristicaDysphagia (n = 68)Normal (n = 133)t/Z/χ2P valueaDemographics Age (years)50.0 (39.5–61.0)45.0 (35.0–54.5)−2.587 0.010 Sex (M/F)36 (52.9%)/32 (47.1%)24 (18.0%)/109 (82.0%)26.2 <0.001 BMI (kg/m2)26.04 ± 3.9725.26 ± 3.77−1.3370.181Personal and medical history Smoking44 (64.7%)52 (39.1%)14.293 <0.001 Hypertension10 (14.7%)18 (13.5%)0.0520.820 Diabetes7 (10.3%)7 (5.3%)1.7570.185Assessment Baseline salivary gland dysfunction20.887 <0.001 Yes62 (91.2%)80 (60.2%) No6 (8.8%)53 (39.8%) TNM stagingb2.4870.115 I–II61 (89.7%)127 (95.5%) III–IV7 (10.3%)6 (4.5%)Treatment 131I Dose (mCi)150 (150–150)c100 (100–120)−7.061 <0.001 131I, iodine-131; TNM, tumor–node–metastasis.aData are expressed as count (percentage), mean ± SD, or median (P25, P75). P values not exceeding 0.010 were typeset in boldface.bAccording to AJCC 8th edition.cRange: 69–150 mCi. Patient characteristics I, iodine-131; TNM, tumor–node–metastasis. Data are expressed as count (percentage), mean ± SD, or median (P25, P75). P values not exceeding 0.010 were typeset in boldface. According to AJCC 8th edition. Range: 69–150 mCi. No significant differences were observed between the dysphagia and nondysphagia groups regarding BMI, history of hypertension, history of diabetes, or DTC stage (all P > 0.05). However, patients who developed dysphagia were significantly older (median 50.0 vs. 45.0 years; P = 0.010) and more likely to be male (52.9 vs. 18.0%; P < 0.001). A history of smoking (64.7 vs. 39.1%; P < 0.001), the presence of baseline salivary gland dysfunction (91.2 vs. 60.2%; P < 0.001), and a higher 131I dose (median 150 vs. 100 mCi; P < 0.001) were also significantly associated with the dysphagia group. Risk factor analysis Variables significantly associated with dysphagia in the univariate analysis (Table 1) were entered into a multivariate logistic regression model. The analysis identified 131I dosage [odds ratio (OR) = 1.066, 95% confidence interval (CI): 1.043–1.090], smoking history (OR = 3.199, 95% CI: 1.716–5.967), and baseline salivary gland dysfunction (OR = 2.941, 95% CI: 1.763–4.932) as independent predictors (P < 0.001 for all) (Table 2). Table 2Analysis of factors related to dysphagia in DTC patients after 131I treatmentVariable B SEWaldOR (95% CI)P valueAge0.2020.2610.5991.223 (0.734–2.041)0.439Sex0.0270.0201.8181.027 (0.988–1.068)0.178Smoking history1.1630.31813.3943.199 (1.716–5.967) <0.001 Baseline salivary gland dysfunction1.0790.26316.7802.941 (1.763–4.932) <0.001 131I dosagea0.0640.01131.9321.066 (1.043–1.090) <0.001 CI, confidence interval; DTC, differentiated thyroid carcinoma staging; 131I, iodine-131; OR, odds ratio.aOR per 1 mCi increase. Analysis of factors related to dysphagia in DTC patients after 131I treatment CI, confidence interval; DTC, differentiated thyroid carcinoma staging; 131I, iodine-131; OR, odds ratio. OR per 1 mCi increase. Nomogram construction and performance A nomogram predicting posttreatment dysphagia was constructed based on the three independent predictors: 131I dosage, smoking history, and baseline salivary gland dysfunction (Fig. 1). The model generates a total score by summing the points for each variable, which corresponds to a predicted probability of developing dysphagia. The nomogram model for predicting dysphagia after iodine-131 therapy in patients with papillary thyroid carcinoma. The model demonstrated strong discriminative ability. The ROC curve analysis yielded an AUC of 0.913 (95% CI: 0.714–0.941) (Fig. 2). At the optimal cutoff value of 0.226, the model had a sensitivity of 0.941 and a specificity of 0.714. ROC curve of the model. AUC, area under the curve; ROC, receiver operating characteristic. Model validation and clinical utility Internal validation was performed using 1000 bootstrap resamples. The calibration curve showed high agreement between the nomogram-predicted probabilities and the actual observed frequencies (mean absolute error = 0.013), which indicates good model calibration (Fig. 3). Calibration curve of the model. DCA was performed to assess the model’s clinical utility (Fig. 4). Using the nomogram to guide clinical decisions (red line in Fig. 4) provides a net benefit across a wide range of high-risk thresholds, superior to the strategies of treating all patients (gray line) or treating no patients (black horizontal line). DCA curve of the model. DCA, decision curve analysis. Discussion In this study, the incidence of dysphagia in PTC patients receiving adjuvant 131I therapy reached 33.8% (68/201). Our analysis identified three independent risk factors: baseline salivary gland dysfunction, smoking history, and a higher administered 131I dose. We successfully developed and validated a nomogram incorporating these variables that demonstrated strong predictive performance. As the long-term prognosis for most PTC patients is excellent, the focus of clinical care has increasingly shifted to long-term quality of life and the mitigation of treatment-related side effects [28]. Our finding that pre-existing salivary gland dysfunction is a major predictor is a key advancement. 131I accumulates in salivary glands via NIS [29] and can cause acute and chronic inflammation, leading to obstructive sialadenitis and xerostomia [30]. Our data suggest a cumulative-damage model, where patients whose glands are already compromised before therapy are significantly more vulnerable to developing clinically relevant swallowing dysfunction. This highlights the need for baseline salivary function assessment, especially in high-risk patients [31]. It must be noted that the proposed predictor may be less applicable in centers where pretherapy salivary scintigraphy is not routinely performed. In such settings, the model may require adaptation using more widely available surrogate markers (e.g. baseline xerostomia symptoms, salivary flow testing, or ultrasound-based salivary gland assessment), and future studies are necessary. Similarly, the strong association with smoking history aligns with known biological mechanisms. Tobacco use is a potent irritant that causes mucosal inflammation, microvascular damage, and impaired tissue healing [32,33]. It is plausible that smoking compromises the ability of the oral and pharyngeal mucosa to withstand and repair radiation-induced damage, thereby amplifying the effects of xerostomia and contributing to dysphagia. The present finding provides a strong, evidence-based rationale for smoking cessation counseling before initiating 131I therapy. In our cohort, a higher administered 131I dose was associated with an increased dysphagia risk, which suggested a dose–response relationship. This may be explained by greater salivary gland injury and reduced salivary secretion at higher radioiodine exposure, which can worsen xerostomia and impair swallowing-related quality of life. Our findings are directionally consistent with the broader head and neck radiotherapy literature, in which greater radiation exposure to swallowing-related structures is associated with worse swallowing outcomes; however, because the exposure pattern of 131I therapy differs substantially from external beam radiotherapy, direct comparisons should be made with caution [34]. While 131I radiation is more targeted, its effects are not confined to the thyroid bed, and a higher dose may deliver sufficient collateral radiation to adjacent pharyngeal structures to induce similar, although milder, fibrotic changes or nerve dysfunction. The primary innovation of this study is the integration of the identified risk factors into a single predictive nomogram. To our knowledge, this is the first nomogram specifically designed to predict dysphagia in PTC patients receiving 131I. The model is a reliable clinical tool with strong discrimination (AUC = 0.913) and good calibration, and it can be used to stratify patients based on their risk profile. For example, a patient identified as high-risk (e.g. a smoker with pre-existing dysfunction receiving a high dose) could be prioritized for preventive measures, such as salivary-sparing agents, aggressive hydration, or oral moisturizers [31]. This study also has several limitations. First, its retrospective, single-center design may restrict the generalizability of the results. Specifically, our dataset did not fully capture surgery-related factors (such as the exact extent of thyroidectomy, the scope of central or lateral neck dissection, and the occurrence of recurrent laryngeal nerve injury) that are known to influence swallowing outcomes. Consequently, residual confounding from these surgical variables cannot be entirely excluded. Second, we assessed dysphagia using the MDADI, a subjective, patient-reported inventory. While the MDADI effectively captures swallowing-related quality of life, a metric highly relevant to survivor care, it cannot distinguish between underlying physiological mechanisms, such as xerostomia versus primary muscle dysfunction. The results from MDADI should be viewed as complementary to, rather than a substitute for, objective physiological assessments. Third, the 6-month follow-up period may be insufficient to characterize the long-term trajectory of the condition. We could not reliably determine whether dysphagia at the study endpoint represented transient or persistent dysfunction, and late-onset symptoms occurring beyond 6 months may have been missed. Finally, although internal validation using 1000 bootstrap resamples demonstrated good optimism-corrected performance, we did not perform external validation in an independent cohort. Such validation is a necessary prerequisite for broader clinical implementation of our nomogram. Future prospective multicenter studies should incorporate objective diagnostic tools, such as VFSS or FEES, to validate this nomogram and further elucidate the pathophysiology of 131I-induced dysphagia [35]. In conclusion, this study identifies significant, modifiable risk factors for post-131I dysphagia and provides the first validated nomogram to predict its occurrence. This nomogram can help clinicians identify at-risk patients, guide personalized treatment planning, and set realistic expectations, ultimately serving to optimize long-term quality of life for PTC survivors. Acknowledgements Conceptualization, data curation, software, methodology, formal analysis, validation, writing – original draft preparation: H.J. Methodology, software, formal analysis, validation, writing – original draft preparation: J.C. Resources, data curation, validation, writing – review and editing: H.Z. Conceptualization, supervision, validation, writing – review and editing: Z.L. All authors have read and approved the final version of the article. This study was approved by the Medical Research Ethics Committee of the Affiliated Hospital of Xuzhou Medical University (Approval No. XYFY2023-KL443-01). The committee waived the requirement for written informed consent as the study was retrospective and did not involve any change to clinical treatment. All methods were carried out in accordance with relevant guidelines and regulations or the Declaration of Helsinki. The datasets used and analyzed during the current study are available from the corresponding author on reasonable request. Conflicts of interest There are no conflicts of interest.