Factors Influencing or Salivary Fistula Formation Post-Parotidectomy Christopher J. Britt, MD, Andrew P. Stein, BA, and Gregory K. Hartig, MD University of Wisconsin School of Medicine and Public Health

Parotidectomy Volumetric and Extent of Resection Analysis Upon evaluation of patient characteristics of sialocele formation, we found no significant

INTRODUCTION differences in sidedness of parotidectomy, gender, age, or BMI between groups. Pathologically, the Patients were stratified into groups based on the extent of facial nerve dissection during rate of was statistically lower in the sialocele group. This may be partially explained by Parotidectomy with facial nerve dissection can be performed for several different reasons the case. If the lower facial nerve division was primarily traced, we described this as an inferior the volume of tissue removed. There was a larger volume of tissue removed in the malignant group with several different operative goals. The extent of parotidectomy can vary based on tumor superficial parotidectomy. If the upper division was traced primarily, we describe this as a compared to the benign group; however, this was not significant. location, tumor size, tumor , or need for lymphadenectomy. The range of parotid tissue superior superficial parotidectomy. If the dissection took place between the upper and lower removed can vary greatly depending on the surgery performed and size of the tumor. The extent divisions, we described this as a middle superficial parotidectomy. Groups for complete of gland resected also depends on whether the dissection is limited to the superficial aspect of the superficial parotidectomy, total parotidectomy, subtotal parotidectomy, and revision In our study, 72 (96%) patients who had sialocele, developed it within 1 month time and none gland or involves the deep lobe. parotidectomy. developed a sialocele after 6 months, and all patients with sialocele were free of this complication by 6 months. Complications of parotid surgery are diverse and are generally related to facial nerve injury, The volume of surgical specimens was calculated from the pathologic measurements. disruption of sensory or autonomic innervation, or wound complications. Sialocele or salivary Overall, the average volume of surgical specimens (n=67) was 37.4 cc (range 0.2–277 cc; 95% We did not see a statistically significant difference between volumes in any dissection group fistula is a wound complication that occurs between 5-39%. Sialocele is a collection of saliva at CI: (28.4 cc, 47.2 cc)). There were no measurements in 8 patients. The surgical specimen when compared against one another. This indicates that different classifications of surgical the surgical site, salivary fistula occurs when the fluid is no longer contained and drains onto an volume of patients who underwent total parotidectomy (n=6) was 44.8 cc, the volume of the dissection did not generate larger or smaller volumes of dissection, on average. Although there was epithelial surface. Sialoceles are thought to form secondary to a disruption of the salivary complete superficial parotidectomy (n=27) was 44.1 cc, the volume of inferior superficial no head to head comparison of the sialocele group surgical dissection volumes to the comparison parenchyma and subsequent spilling of saliva into the surrounding tissues. parotidectomy (n=21), the volume of the middle superficial parotidectomy (n=4) was 35 cc, the group, it is interesting to note that although the procedures were labelled differently in the operative volume of the superior superficial parotidectomy (n=6) was 24 cc. 2 patients who underwent note, the volumes of dissection were relatively close. We also had 6 patients who underwent total subtotal parotidectomy had an average volume of 18.33, the two patients who had recurrence parotidectomy. We argue that even with total parotidectomy, sialocele is still a possible complication Our objective was to determine if the extent of parotidectomy or other patient or tumor did not have pathologic measurements, there were 5 patients who had unknown extent of secondary to possible accessory or residual parotid tissue. characteristics influence the rate of sialocele or salivary fistula formation after parotidectomy. The resection. Groups with a n greater than 5 were compared and there was no statistical extent of parotidectomy was defined by the volume of tissue removed and by defined area. Age, significance in volume between any group. gender, sidedness, BMI, and pathology were also examined. We recognize that there are limitations to our study. It was carried out in a retrospective nature and diagnosis of sialocele is limited by documentation and patient follow up. The surgical Sialocele Comparison dissection volumes were compiled from data gathered from pathologic reports which may not MATERIALS AND METHODS Gender accurately depict in vivo surgical volume. Pathologic data is limited by the individual pathologist and Male 38 (52.8%) 335 (54.0%) inter-observer variability as are the dimensions of the specimens. Finally, any fluid collection that We obtained Institutional Review Board (IRB) approval to analyze the medical records of all Female 34 (47.2%) 285 (56.0%) formed post operatively that was no consistent with a hematoma was considered a sialocele. Some of these fluid collections may not have been caused by salivary fluid production. patients who underwent parotidectomy at the University of Wisconsin between 1994 and 2013. Facial Nerve Function Specifically, the records were methodically evaluated in order to extract the following information: No weakness 55 (83.3%) 408 (72.5%) sex, age, surgical pathology, side, sialocele or salivary fistula formation, and body mass index CONCLUSIONS (BMI). The size of the mass was also recorded from surgical pathology. Once all medical records Weakness 11 (16.7%) 155 (27.5%) had been reviewed, we examined patient charts to identify all individuals with sialocele or salivary Lesion side Upon evaluation of patient characteristics of sialocele formation, we found no significant differences fistula formation status post parotidectomy. Patients who did not have a sialocele that had all Right 32 (45.1%) 296 (48.0%) pertinent data for comparison were used in the comparison group. Univariate analysis was in sidedness of parotidectomy, gender, age, or BMI between groups. Pathologically, the rate of Left 39 (54.9%) 321 (52.0%) employed to compare the variables between parotidectomies performed at our institution with and malignancy was statistically lower in the sialocele group. We did not see a statistically significant without sialocele or salivary fistula. Volumes from each dissection group were also compared to Pathology difference between volumes in any dissection group when compared against one another. We one another using univariate analysis. Benign 62 (86.1%) 444 (71.6%) present the largest series of sialoceles in the literature to date. Malignant 10 (13.9%) 176 (28.4%) RESULTS Continuous variables: Mean (95% CI) Age 55.9 (52.1, 59.6) 57.5 (56.3, 58.8) SELECTED REFERENCES Electronic Medical Record Review and Patient Selection BMI 29.3 (27.7, 30.9) 29.0 (28.4, 29.5) Herbert HA, Morton RP. Sialocele after parotid surgery: assessing the risk factors. Otolaryngology-- head and neck surgery : official journal of American Academy of Otolaryngology-Head and Neck Table 1. Patient and tumor characteristics among patients diagnosed with sialocele or A total of 771 patients underwent parotidectomy at the University of Wisconsin between Surgery 2012;147:489-92. salivary fistula and the comparison group 1994 and 2013. 75 patients (9.7%) developed a sialocele or salivary fistula. 620 patients without sialocele had sufficient data for use in the comparison group. Witt RL. The incidence and management of sialocele after parotidectomy. Otolaryngology--head DISCUSSION and neck surgery : official journal of American Academy of Otolaryngology-Head and Neck Surgery Patient and Tumor Characteristics of Patients who developed Sialocele or Salivary Fistula 2009;140:871-4. Currently, there is disagreement in the literature about the relationship between the extent of For age, gender, pathology, BMI, sidedness, and facial nerve symptoms see table 1. There was surgery and sialocele formation. Witt et al. and Tuckett et al. concluded that the greater the extent of Tuckett J, Glynn R, Sheahan P. Impact of extent of parotid resection on postoperative wound no statistical difference between the age groups (p=0.621). The male to female ratio was 1.14:1. resection, the less likely sialocele formation because the less tissue there is to create saliva. complications: A prospective study. Head & neck 2013. This was not significant when compared to the parotidectomy group as a whole (p=0.467). BMI Conversely, Herbert et al. did not find extent of dissection to be associated with sialocele formation. was significant when compared to the overall group (p=0.343). No study has objectively looked at differences in extent of dissection.

Pathologically, 64 patients (85.3%) had benign and of these 23 patients (30.7%) had and 20 patients (26.7%) had Warthin’s tumors. 11 patients (14.7%) had malignancy and of these 3 were squamous cell carcinoma (4%), and 2 where (2.7%). When compared to the all patients who had undergone parotidectomy (28.4%), the rate of malignancy was lower in the sialocele group (p=0.029). There was no difference between groups with comparison to facial nerve weakness(p=0.945). Finally, there was no statistical difference between sidedness of sialocele (p=0.260)