Sialographic Analysis of Radioiodine-Associated Chronic Sialadenitis

Abstract

OBJECTIVES:

To apply a novel sialography classification system to identify parotid and submandibular ductal findings following I-131 therapy and to assess correlates to dose and duration of symptoms.

METHODS:

Patients who underwent sialography between February 2008 and February 2019 after previously receiving I-131 treatment were identified via a retrospective chart review. Their sialograms were systematically evaluated and scored by applying the Iowa parotid sialogram scale to also include submandibular gland analysis.

RESULTS:

From 337 sialograms, 30 (5 submandibular, 25 parotid) underwent analysis. Ductal stenosis was identified in all sialograms and was graded as moderate ( >50-75%) in 7/30 cases and severe ( >75%) in 15/30 cases. The distal (main) duct was narrowed in 23/30 cases. No association was identified between degree of ductal stenosis and I-131 dose (p=0.39), age (p=0.81), or time from I-131 therapy to sialogram (p=0.97).

CONCLUSIONS:

The Iowa parotid sialogram scale was successfully applied to report abnormalities of the parotid and submandibular ductal system. The most common manifestation of I-131-associated sialadenitis was a severe stenosis within the distal salivary duct. No statistically significant association was found between degree of ductal stenosis and dose of I-131, age, or duration of symptoms.

1. INTRODUCTION:

Sialography (retrograde injection of contrast dye into the ductal system of a major salivary gland) was initially described in the early 1900’s.¹ Technical modifications including the use of flexible catheters, routine stimulation of the parotid gland after instillation of dye, and use of water-soluble contrast media have improved the ease and success of the procedure. Technical challenges remain and, whereas some radiologists perform this procedure alone, involvement of a surgeon experienced in ductal cannulation and salivary gland surgery is considered by many to offer the highest success rates. Difficulties in coordinating efforts between services has been an impediment to the use of sialography as some institutions.

Over the past several decades, advances in computed tomography (CT) and ultrasound technology have provided alternatives to sialography for the diagnosis of salivary gland dysfunction. During the 1990’s, sialendoscopy was introduced as a diagnostic and therapeutic intervention for various salivary gland abnormalities.^2,3 This improved technology rekindled an interest in salivary imaging, including use of sialography, to assist in decision-making regarding the role for sialendoscopy.^4,5 Additionally, the intraductal irrigation done with bactericidal iodinated contrast in the course of performing a sialogram has helped with symptom control in selected cases.^6-9

Sialography has been reported as useful in characterizing the ductal and parenchymal changes associated with many salivary diseases, including sialosis, juvenile recurrent parotitis, Sjogren syndrome, and salivary gland calculi.^10-13 To our knowledge, sialography has yet to be similarly studied in assessing radioiodine-associated sialadenitis.

Salivary pain and swelling as well as xerostomia may manifest at any time after I-131 therapy.¹⁴ Salivary gland scintigraphy of I-131 recipients has revealed a dose-dependent association between I-131 and salivary gland dysfunction, primarily affecting the parotid glands.^15,16 Lee et al. published a review of 76 patients with salivary gland symptoms who underwent ultrasound or CT scan after receiving I-131 therapy, which demonstrated abnormal radiographic findings in 50% (38/76) of patients.¹⁷ Due to the superior resolution of the ductal system offered by sialography, a review of sialographic findings in I-131-associated salivary dysfunction is warranted.

The objective of this study was to analyze the sialographic findings from patients with I-131-associated ductal stenosis utilizing a novel classification system as previously described¹³, and to correlate these findings to I-131 dose and duration of symptoms.

2. METHODS:

2.1. DATA COLLECTION:

Approval to conduct this study was obtained from the University of Iowa Institutional Review Board (IRB) on November 25^th, 2019, IRB identification number 201909820. A waiver of consent was approved by the IRB for this retrospective review. 337 patients who underwent sialography at the University of Iowa between 2/8/2008 and 2/4/2019 were included. A chart review was performed, which identified 32 patients who were treated with I-131 administration prior to their sialogram. Clinical data was collected from these charts, including patient age at time of sialogram, gender, reason for I-131 therapy, I-131 dose, symptoms, duration of time from I-131 therapy to initiation of symptoms, duration of time from I-131 therapy to sialogram, BMI, type of symptoms, treatments performed after sialography, subjective outcome of symptoms at most recent follow-up, and duration of follow-up.

2.2. DESCRIPTION OF SIALOGRAPHY TECHNIQUE:

All sialograms were performed in the fluoroscopy suite by the senior investigator (HTH) using consistent technique, as previously described.¹⁸ All patients received routine antibiotic prophylaxis prior to their sialogram. A microscope was used to direct placement of a 22-gauge angiocatheter into the parotid or submandibular duct orifice. If necessary, a 0.015 or 0.018 inch guidewire was used to assist with placement of the angiocatheter. Isovue 370 contrast was instilled into the duct under direct fluoroscopic guidance, and radiographs were taken. The angiocatheter was subsequently removed. To evaluate contrast retention, the patient was given a small amount of lemon juice to swish and expectorate, and delayed radiographs were taken.

2.3. SIALOGRAPHY ANALYSIS:

The sialograms were systematically evaluated and scored by two neuroradiologists (BP and JEM) applying the Iowa parotid sialogram scale to classify ductal stenoses/strictures, peripheral ductal dilations, and patterns of contrast retention, as previously described (Table 1).¹³ We adapted this classification – initially developed to address the parotid gland – to additionally apply it to submandibular gland sialography. This classification system describes location of the ductal stenosis, degree of the ductal stenosis, retention of contrast, and peripheral duct dilation. Location of the stenosis is graded as 0 (no stenosis), 1 (tertiary and quaternary ducts), 2 (secondary ducts), or 3 (main duct). Degree of stenosis is graded as 0 (no stenosis), 1 (minimal; <25% stenosis), 2 (mild; 25-50% stenosis), 3 (moderate; >50-75% stenosis) and 4 (severe; >75% stenosis). Retention of contrast is graded as 0 (none), 1 (focal retention pockets), or 2 (indistinct retention of contrast). Peripheral duct dilation is graded as 0 (normal peripheral ducts), 1 (punctate dilation <1 mm in size), 2 (globular dilation 1-2 mm in size), 3 (coalescent, irregularly shaped globules >2 mm in size), or 4 (invasion/destruction of the gland parenchyma). The radiologists were informed that the patients were referred for a diagnosis of I-131 sialadenitis but were blinded to all other patient variables. The radiologists reviewed each image independently, followed by collaborative interpretation to determine the final grade.

Table 1:

Sialography grading scale used in this study, adapted from the parotid sialogram scale described by Foggia et al¹³.

Score	Location of ductal stenosis	Degree of ductal stenosis	Retention of contrast	Peripheral duct dilation
0	No stricture/stenosis	No stricture/stenosis	No retention	No dilation
1	Tertiary or quaternary duct	Minimal; <25% stricture/stenosis	Focal retention pockets	Punctate dilation; <1 mm in size
2	Secondary duct	Mild; 25-50% stricture/stenosis	Indistinct retention	Globular dilation; 1-2 mm in size
3	Main duct	Moderate; >50-75% stricture/stenosis	N/A	Coalescent, irregularly shaped globules; >2 mm in size
4	N/A	Severe; >75% stricture/stenosis	N/A	Invasion/destruction of gland parenchyma

2.4. PLOT GENERATION AND STATISTICAL ANALYSIS:

We performed statistical analysis to assess the significance of I-131 dosage, age, and time from I-131 therapy to sialogram versus severity of duct stenosis and peripheral duct dilation. Assuming a non-Gaussian distribution of these values, we used the Kruskal-Wallis one-way analysis of variance test (ANOVA). To visualize the distribution of these values, box-and-whisker plots were generated from the data. All analyses and plot generation were conducted using GraphPad Prism v8.1.2 software (San Diego, CA).

3. RESULTS:

3.1. DEMOGRAPHIC AND CLINICAL DATA:

32 patients underwent chart review and review of sialogram images. Demographic and clinical information is summarized in Table 2 and Supplemental Table S1. The reasons for receiving I-131 therapy included thyroid cancer in 27 (84.4%), Graves’ disease in 4 patients (12.5%), and Hashimoto’s thyroiditis in 1 (3.1%). The I-131 dose was unknown in 5 of these patients. The mean I-131 dose for the remaining 27 patients was 144.2 millicuries (mCi) ranging from 20.5 to 403 mCi. The mean I-131 dose for the patients with thyroid cancer was 151.0 mCi ranging from 53 to 403 mCi. The dose of I-131 was only available in 2 of the 4 Graves’ disease patients; in these two patients, the doses were 20.5 and 98 mCi. For the one patient who received I-131 for Hashimoto’s thyroiditis, no records of the dose were available. Two patients received multiple doses of I-131 and for the purposes of calculating the total dose of I-131 therapy, a summation of the doses was used.

Table 2:

Clinical characteristics of the 32 included patients who underwent sialogram for radioiodine-associated sialadenitis. I-131 = radioactive iodine. BMI = body mass index.

All patients (n = 32):
Gender (male/total, percent male)	4/32 (12.5%)
I-131 dose (mean (mCi), range)	144.2 (20.5-403)
BMI at time of sialogram (mean, range)	31.7 (19.5-53.3)
Age at time of sialogram (mean (years), range)	48.7 (19-71)
Duration of follow-up (mean (days), range)	307.2 (0-1436)
Time between I-131 therapy and symptoms developing (mean (days), range)	1034.9 (0-9375)
Time from I-131 therapy to sialography (mean (days), range)	1400.2 (157-9479)
Duration of symptoms (mean (days), range)	365.3 (25-2876)

All the patients in this study reported salivary symptoms prior to their sialogram. 30 patients (93.8%) reported salivary gland pain, 30 (93.8%) had salivary gland swelling, 21 (65.6%) had xerostomia, 7 (21.9%) had altered taste, 4 (12.5%) had dry eyes, and 2 (6.3%) reported infections of a salivary gland. Treatment subsequent to sialography included in-clinic Kenalog infusion in 18 (56.3%) patients, Kenalog infusion at the time of sialendoscopy under general anesthesia in 9 (28.1%) patients, and submandibular gland resection in 2 (6.3%) patients. No patients required parotidectomy.

7 patients (21.9%) did not return for a follow-up appointment after their sialogram. Of the remaining 25 with reported follow-up, the mean duration of follow-up was 393.2 days, ranging from 36 to 1,436 days. At their most recent follow-up visit, 21 of these 25 patients (84.0%) reported subjective improvement of their symptoms and 4 (16.0%) reported no change in symptoms. No patients reported worsening of their symptoms. Of the 25 patients who reported improvement, 5 (24.0%) received a sialogram as their only procedure, and the remaining 21 (76.0%) received additional interventions including steroid insufflation, sialendoscopy, and submandibular gland excision. Of the 4 patients who reported no change, 1 (25.0%) received a sialogram as their only procedure, 2 (50.0%) received a sialogram with intraductal Kenalog infusion, and 1 (25.0%) received a sialendoscopy with intraductal Kenalog infusion.

3.2. QUALITATIVE RESULTS:

Sialography was performed in all 32 patients; 26 (81.3%) patients had a parotid sialogram and 6 (18.8%) had a submandibular sialogram. A representative parotid sialogram is shown in Figure 1. If salivary gland symptoms were bilateral, the procedure was performed on the more symptomatic gland. The grades of the sialograms based on the Iowa sialogram classification system are summarized in Table 3. A total of 30 sialograms were graded. Two (6.25%) sialograms were excluded from review due to inadequate image quality precluding analysis of the ductal system. In one patient (patient 15, Supplemental Table S1), a large accessory parotid gland was identified with stenosis of the accessory duct. This patient received two measurements for stenosis location and degree, though only the findings from the main parotid duct were used for the purposes of data analysis.

Figure 1:

Representative sialogram images under fluoroscopy of the left parotid gland in a patient who had previously underwent radioactive iodine therapy. (A) shows the initial cannulation of the left parotid duct without injection of dye. No calculi are visualized. Lateral (B) and antero-posterior (AP) (C) sialogram images after injection of 3 cc of Isoview-370 dye show a marked stenotic portion (>75% stenotic) of the distal parotid duct. Magnified AP view (D) demonstrates the intraglandular parotid ductules with an irregular contour and paucity of normal branching pattern. Lateral (E) and AP (F) sialogram images obtained after the patient was given a small amount of lemon juice to swish and expectorate demonstrate no significant washout of contrast, consistent with the significant distal obstruction.

Table 3:

Characteristics of the sialograms from the 32 included patients with radioactive sialadenitis based on the Iowa parotid sialogram scale, adapted to also include submandibular glands. Location of the stenosis is graded as 0 (no stenosis), 1 (tertiary and quaternary ducts), 2 (secondary ducts), or 3 (main duct). Degree of stenosis is graded as 0 (no stenosis), 1 (minimal; <25% stenosis), 2 (mild; 25-50% stenosis), 3 (moderate; >50-75% stenosis) and 4 (severe; >75% stenosis). Retention of contrast is graded as 0 (none), 1 (focal retention pockets), or 2 (indistinct retention of contrast). Peripheral duct dilation is graded as 0 (normal peripheral ducts), 1 (punctate dilation <1 mm in size), 2 (globular dilation 1-2 mm in size), 3 (coalescent, irregularly globules >2 mm in size), or 4 (invasion/destruction of the gland parenchyma). PDD = peripheral duct dilation. † = unable to evaluate remaining patients using the scoring system due to inadequate quality of images. ‡ = unable to classify remaining patients due to extravasation of contrast. § = unable to classify remaining patients due to no contrast progressing past the stenosis.

Sialogram score	Location (n = 30†)	Degree (n = 30†)	Retention (n = 23†,‡,§)	PDD (n = 24†,‡,§)
0	0 (0.0%)	0 (0.0%)	7 (29.2%)	12 (50.0%)
1	4 (13.3%)	4 (13.3%)	6 (26.0%)	4 (16.7%)
2	3 (10.0%)	3 (10.0%)	10 (43.5%)	2 (8.3%)
3	23 (76.7%)	7 (23.3%)	N/A	1 (4.2%)
4	N/A	16 (53.3%)	N/A	5 (20.8%)

Abnormal ductal stenosis was identified in all 30 included sialograms. The stenoses were located in the tertiary or quaternary ducts in 4 (13.3%) glands, in the secondary ducts in 3 (10.0%), and in the primary (peripheral) duct in 23 (76.7%). The degree of stenosis was minimal (<25% stenosis) in 4 (13.3%) glands, mild (25-50% stenosis) in 3 (10.0%), moderate (>50-75% stenosis) in 7 (23.3%), and severe (>75% stenosis) in 16 (53.3%). Retention of contrast and peripheral duct dilation could not be adequately assessed in 2 glands and 1 gland, respectively, due to extravasation of contrast. Retention of contrast and peripheral duct dilation could not be assessed in 5 sialograms due to the presence of a primary ductal stenosis preventing any passage of contrast proximally. Of the 23 sialograms evaluated for contrast retention, 7 (30.4%) demonstrated no contrast retention, 6 (26.0%) had focal pockets of contrast retention, and 10 (43.4%) had indistinct contrast retention. Of the 24 sialograms evaluated for peripheral duct dilation, 12 (50.0%) had normal-appearing peripheral ducts, 4 (16.7%) had diffuse punctate dilation of the peripheral ducts, 2 (8.3%) had globular uniform peripheral duct dilation, 1 (4.2%) had a coalescence of irregularly dilated globules, and 5 (20.8%) had destruction of gland parenchyma.

3.3. STATISTICAL RESULTS:

Of the 32 patients who underwent sialogram, 2 were excluded from statistical analysis due to the sialogram images being of poor quality and unable to be interpreted. Four additional patients were excluded due to there being no available record of their I-131 dosage. The 26 remaining patients underwent the Kruskal-Wallis ANOVA test comparing I-131 dose, patient age, and time from I-131 therapy to sialogram against degree of duct stenosis. These findings are summarized in Table 4. Box-and-whisker plots of these data are shown in Figure 2.

Table 4:

Results from statistical analysis via Kruskal-Wallis one-way analysis of variance of mean I-131 dose, age, and time from I-131 therapy to sialogram based on degree of ductal stenosis identified on sialogram from minimal to severe. I-131 = radioactive iodine.

	Minimal	Mild	Moderate	Severe	P-value
n	2	2	7	15	*
Mean I-131 dose (mCi) (standard deviation)	102.4 (6.2)	124.2 (100.7)	176.9 (108.8)	126.7 (44.4)	0.39
Age (years) (standard deviation)	50.5 (12.0)	44.5 (13.4)	43.3 (15.0)	48.2 (12.0)	0.81
Time from I-131 therapy to sialogram (days) (standard deviation)	363.5 (208.6)	472.5 (365.6)	698.1 (999.3)	1038.9 (1238.0)	0.97

Figure 2:

Radioactive iodine (I-131) dose (A), patient age (B), and time from I-131 therapy (C) stratified by severity of ductal stenosis via box-and-whisker plot.

4. DISCUSSION:

I-131 therapy is an important adjunctive or primary treatment for patients with disorders of the thyroid gland.¹⁹ Administration of I-131 can result in significant salivary morbidity due to radiation damage from accumulation of radioactive iodine within the salivary glands.^20,21 The capacity of the salivary glands to concentrate iodine is an evolutionarily conserved trait across mammals, resulting in a continual circulation of iodides between the salivary glands and the gastrointestinal tract.²² In fact, iodine will concentrate more rapidly in the salivary glands than any other tissue in the body. Salivary dysfunction from radiation damage can lead to acute and chronic complications including xerostomia²¹, recurrent sialadenitis, dental caries^23,24, and dysphagia²⁵. Oral sequelae from I-131 have also been linked to poorer quality of life.^26,27 Although all salivary glands are involved in the transport of iodine, serous-secreting cells are more susceptible to damage from radioactive iodine than mucous-secreting cells.²¹

This study identified common patterns of salivary gland damage resulting from I-131 administration. In this selective group of patients with symptomatic radioiodine-associated chronic sialadenitis, ductal stenosis was identified in all patients. In the majority of cases, the stenosis was located in the main duct and identified with sufficiently distal involvement to permit treatment with dilation including use of sialendoscopy. The degree of ductal stenosis was classified as severe in more than half of the sialograms studied. In 5 out of 32 cases, this stenosis was sufficiently severe as to preclude passage of the instilled contrast past the narrowing. In these cases the sialogram was unable to assess the status of the peripheral (promixal) ducts. By performing a sialogram before considering operative sialendoscopy, the surgeon can identify specific areas of stenosis to target. The sialogram may also identify obliterative stenoses that preclude use of sialendoscopy and warrant consideration for alternative interventions. Previous studies have demonstrated that salivary dysfunction and xerostomia after I-131 administration occur in a dose-dependent fashion.^26,28,29 The failure to identify an association between I-131 dose and degree of abnormalities on sialography likely reflects a selection bias to assess only patients with symptoms sufficiently severe to warrant this evaluation.

Patterns of structural disease identified on sialography can assist in differentiating I-131 induced salivary gland destruction from other causes of salivary dysfunction. In contrast to the patterns identified in I-131 sialadenitis, our previous work demonstrated that Sjogren’s syndrome more commonly affects the proximal parotid ducts relative to the distal ducts, which may become involved later in the disease process.¹³ The capacity to distinguish etiology of ductal pathology is particularly important for patients with autoimmune disease who undergo I-131 therapy, as autoimmune thyroid disease and Sjogren’s syndrome are the most commonly concomitant polyautoimmunities.³⁰ Cases have been described in the literature in which lip biopsy has been used to distinguish I-131-associated salivary gland dysfunction from autoimmune salivary gland dysfunction.³¹ Sialography may represent a minimally invasive alternative test to biopsy for differentiating these disease entities. In contrast to our previous work which exclusively employed this classification system for parotid sialograms, this study also included imaging of the submandibular glands. The criteria were readily applicable to the submandibular glands without modification, and these sialograms demonstrated similar patterns of ductal stenosis as the parotid glands.

Our study has implications for counseling patients regarding the effects of I-131 therapy on salivary gland function and the available treatment options. In this study, all patients reported salivary symptoms with symptom onset occurring one month after I-131 treatment on average. This suggests that the first few weeks after I-131 administration may represent a window for therapeutic intervention aimed at preventing progression of salivary dysfunction. Use of sialogogues and vigorous hydration are conservative preventative strategies with limited success reported.¹⁹ More invasive interventions for ductal obstruction (e.g., stenosis, stricture) and sialadenitis resulting from I-131 administration include intraductal steroid insufflation and sialendoscopy with mechanical duct dilation.^32-34 In this study, over half of the patients received treatment with Kenalog infusion or sialendoscopy with mechanical dilation in the weeks to months following their sialogram, and the majority of patients reported subjective improvement of their symptoms after undergoing these targeted therapeutic interventions. More research is needed to determine the rate of success of these procedures in reducing morbidity associated with I-131 administration. A prospective trial examining the effects of steroid infusion into the salivary glands or other conservative measures in all patients who undergo I-131 therapy would be valuable for determining if the risk of sialadenitis is modifiable by early intervention.

The small sample size, lack of a healthy control group, and retrospective design are limitations of this study. The finding that dose was not associated with the extent of salivary pathology identified on sialography may be attributed both to selection of patients with severe symptoms and also the small sample size. However, the absence of an association between dose and abnormal radiographic findings may also represent the fact that salivary dysfunction is multifactorial. Other factors contributing to radiographic abnormalities on sialography may include gland uptake factors including variable I-131 absorption and clearance, variable ductal anatomy, use of xerostomia medications, comorbid conditions, and use of preventative measures. Additionally, all data in this study was collected from a single institution; further validation of these results will be needed with a larger sample size at other centers utilizing the same salivary duct scaling system to demonstrate reproducibility of the findings. Nevertheless, this study represents the largest cohort of sialograms characterizing I-131-associated sialadenitis to date, with all sialograms performed with consistent technique and interpretation by neuroradiologists experienced in the evaluation of sialograms.

5. CONCLUSION:

All patients in this study who reported salivary gland symptoms after I-131 therapy were found to have abnormalities of the salivary ductal system. Ductal stenosis was found to primarily affect the distal salivary duct and was not associated with peripheral (proximal) duct dilation. No statistically significant association was noted between degree of stenosis identified on sialography and I-131 dose received, patient age, or duration of symptoms. Sialography may be a useful modality to identify the location and extent of I-131-associated ductal injury to help direct therapy.

Supplementary Material

Supplemental Table S1:

detailed characteristics of the individual 32 patients who underwent sialography for radioiodine-associated chronic sialadenitis.