Abstract
Objectives
To examine the epidemiology, subtypes, trends over time, and predictive factors for recurrence and malignant transformation of sinonasal papillomas.
Methods
A retrospective chart review of 118 patients with sinonasal papillomas from 2009 to 2019 was conducted at the University of California, Los Angeles. This study is a follow-up to a previously published study from 2000 to 2009 at the same academic center.
Results
The mean age was at presentation was 58.5 years, with a 2:1 male to female ratio, and average follow-up of 30.1 months. The rate of recurrence after complete resection was 19% with an average of 32.6 months to recurrence. The time to recurrence followed a bimodal distribution with 57% of cases recurring within 24 months (mean = 10) and 43% from 40 to 103 months (mean = 61). The proportion of the inverted papillomas rose from 38% in 2000-2004 to 89.6% in 2015-2019. Patients presenting at a younger age had a higher chance of recurrence (mean age 52 with recurrence vs. 61 without recurrence). Age did not correlate with histopathologic transformation in surgical pathology. Furthermore, histopathological transformation did not raise the chance of recurrence. Smoking, alcohol use, chronic rhinosinusitis, and allergic rhinitis were not associated with any of the outcome measures in this study. The most significant factor predicting recurrence, beside age at presentation, was the history of two or more prior sinus surgeries for papillomas or other reasons (OR = 3.52 and 5.81).
Conclusion
This study explored the features of sinonasal papillomas as well as the risk factors for recurrence and transformation. Younger age at presentation and two or more prior surgeries for papillomas were associated with recurrence. Time to recurrence followed a bimodal distribution, with late recurrences happenning from 40 to 103 months after surgery, emphasizing the importance of long-term follow-up for timely resection of tumors and prevention of malignancy.
Keywords
Introduction
Sinonasal papillomas are benign lesions of the nasal cavity and paranasal sinuses that exhibit high rates of recurrence. 1 These papillomas can exhibit significant extension to and destruction of nearby structures such as the orbit and skull base, leading to decreased quality of life. Additionally, they have about a 10% potential for transformation to dysplasia and squamous cell carcinoma.2,3 The etiology of these tumors remains unclear, but a number of genetic alterations and environmental factors have been implicated, including HPV infection, COX-2, EGFR, cell cycle regulators, angiogenic factors such as VEGF and osteopontin, and occupational exposures to a variety of insults such as welding fumes and organic solvents.4,5
Sinonasal papillomas are derived from the ectodermal Schneiderian epithelium of the nasal cavity and paranasal sinuses, which is distinct from the endodermal epithelium of the upper respiratory tract. There are three major histologic subtypes: inverted papilloma (IP), exophytic papilloma (EP), and oncocytic papilloma (OP). Inverted papillomas are the most widely studied of these subtypes. This study is a 10-year follow-up to a previously published study at the same academic center by Vorasubin et al, which sought to characterize the subtypes of sinonasal papillomas and analyze risk factors that predicted the subtypes in 70 patients from January 2000 to July 2011. 6 The purpose of this study is to clarify the epidemiology, subtype differences, predictive factors of recurrence and malignant transformation, and trends over time among sinonasal papillomas.
Materials and Methods
A retrospective chart review of 118 patients with sinonasal papillomas from 2009 to 2019 was conducted to collect data on papilloma subtype, location, patient age, sex, chronic rhinosinusitis, allergic rhinitis, malignant transformation, tobacco use, alcohol use, prior sinus surgeries, SNOT-22 scores, time to follow up, recurrence rates, time to recurrence, surgical method, method for clearing tumor margin, surgeon's gross assessment of obtained margins. Statistical analysis was conducted in GraphPad Prism v 9.3.1.
Age, SNOT-22 scores, months to follow up, and months to recurrence were treated as quantitative variables. Sex, subtype, chronic rhinosinusitis, allergic rhinitis, malignant transformation, tobacco use, alcohol use, prior sinus surgeries for papilloma, prior sinus surgeries for other reasons, location, surgical method, method for clearing tumor margin, surgeon's gross assessment of obtained margins, and recurrence were treated as categorical variables. Quantitative variables were analyzed using one-way analysis of variance among papilloma subtypes. Categorical variables were analyzed using chi-squared test or Fisher's exact test if there were fewer than five data points in a category. Logistic regression was used to determine significant factors predictive of recurrence and malignant transformation. Mixed subtypes were excluded from analyses comparing features of different subtypes for simplification. Patients with partial resections of papilloma were excluded from analyses relating to recurrence.
This study was approved by the UCLA Institutional Review Board (IRB#13-000154), and the requirement for informed consent was waived. Our study was Health Insurance Portability and Accountability Act compliant.
Results
Between 2009 and 2019, 122 patients were identified with sinonasal papillomas in the pathology database. Four of these were excluded, three with prior diagnosis and surgery for papilloma at UCLA affiliated center before 2009 and one who passed from an unrelated cause prior to his sinonasal surgery. Table 1 summarizes the demographic information of the remaining 118 patients with the common study follow-up end date of March 2022. The male to female ratio was 2:1, with 79 males and 39 females. The histologic subtypes consisted of mostly inverted papillomas (85%), followed by oncocytic (7%), exophytic (5%), and mixed types (3%). Most were negative for malignant transformation (90%), with 7% undergoing various degrees of metaplasia and dysplasia, and 3% demonstrating squamous cell carcinoma. 36% of patients had undergone at least one prior surgery for papilloma prior to presentation to this institution, demonstrating a high rate of recurrence. 8% of cases went under partial resection of tumor due to proximity to nerves or dura and another 19% had recurrence after their surgery, with an average of 32.2 months to recurrence.
Demographics and Papilloma Characteristics
The average time to recurrence among the 23 cases that recurred after complete resection of tumor was 31.7 months. These cases appear to be divided into two groups: 13 cases of early recurrence within 24 months (mean 10.0 months), and 10 cases of late recurrence after 40 months (mean 61). The average time between the first surgery at this institution and the last clinic follow-up was 49 and 89 months for the two groups, respectively, compared to 19.4 months in the patients without identified recurrence.
Table 2 summarizes the primary attachment sites of sinonasal papillomas. 17% of inverted papillomas and 12.5% of oncocytic papillomas were attached to multiple sites. In inverted papillomas, the most common single site of attachment was the maxillary sinus (40%), followed by the skull base and sphenoid sinus (14.0%). The largest portion of oncocytic papillomas were attached to maxillary sinus, skull base and sphenoid sinus, as well as lamina papyracea (25% each). In contrast exophytic tumors were most attached to the nasal septum (67%) and inferior turbinate (33%).
Attachment Sites by Subtype
The data from this study were aggregated with previously published data by Vorasubin et al at the same medical center to elucidate changes in subtype distribution. 6 Table 3 shows the distribution of subtypes over 5-year periods from 2000 to 2019. The proportion of the inverted subtype increased steadily from 38.2% in 2000-2004 to 89.6% in 2015-2019, while the proportion of the exophytic subtype decreased from 44.1% in 2000-2004 to 4.5% in 2015-2019. The oncocytic subtype proportion fluctuated, beginning at 17.6% in 2000-2004 and dropping to 2.9% in 2005-2009, with intermediate values of 8.5% in 2010-2014 and 6.0% in 2015-2019.
Changes in Subtype Distribution Over 5-Year Intervals
Demographic characteristics of the inverted, oncocytic, and exophytic subtypes were compared to find variables that were significantly associated with specific subtypes (Table 4). Notably, gender distribution was different among various types with males comprising the majority of exophytic and inverted papillomas (100% and 68%, respectively), while females forming the majority of oncocytic papillomas (63%) (p = .045). Age, tobacco use, alcohol use, allergic rhinitis, chronic rhinosinusitis, prior sinus surgeries, and SNOT-22 scores were not significantly different among the various subtypes.
Potential Predictors of Papilloma Subtype
Analysis was performed to determine demographic characteristics that were significantly associated with papilloma recurrence (Table 5). We excluded nine patients for whom operation reports specifically noted a partial or incomplete resection. In most cases this was due to proximity of the tumor attachment to dura or a nerve. The average age of patients with recurrent papillomas was significantly lower than those without recurrence (52.0 vs. 61.0, p = .009). Having two or more prior surgeries for papilloma was also a risk factor for further recurrence (OR = 3.52, 95% CI 1.22-9.71). However, for patients with only one prior surgery for papilloma the odds ratio for recurrence was 0.85 (95% CI 0.17-3.85). Similarly, two or more prior sinus surgery for other indications besides sinonasal papillomas was also correlated with higher recurrence (OR = 5.81, 95% CI 1.39-24.4). The presence of histopathologic transformation (either dysplasia, metaplasia, or carcinoma) in the samples from first surgery or prior, did not increase the risk of recurrence (p = .437).
Potential Predictors of Recurrence
The surgeons in this center used electrocautery and burring with diamond drills for obtaining margins for most of the cases, followed by blunt removal of tissue with forceps or microdebrider. Other techniques such as CO2 laser were rarely used and are not included in the analysis. Most surgeries were performed via endoscopy with assistance of navigation, especially when tumors involved the skull base or were close to a nerve pathway. In most of these cases close margins were obtained with combination of drilling and cautery. We found no correlation between the surgical method or method for obtaining margins on the recurrence rate. Similarly, surgeon's gross assessment of the margins at the end of surgery did not appear to predict recurrence.
Table 6 compares demographics and papilloma features to analyze significant risk factors for precancerous and cancerous transformation, which included dysplasia, metaplasia, and carcinoma. Patients with transformation were slightly older at an average age of 66.5 compared to 57.9 in those without transformation, but this was not statistically significant (p = .058). None of the other demographic and papilloma features were significantly associated with malignant transformation.
Potential Predictors of Dysplasia, Metaplasia, and Carcinoma
Discussion
This study characterized the demographic and histopathologic features of sinonasal papillomas across a 10-year period at a single academic center to elicit characteristics that were associated with specific papilloma subtypes, recurrence, and malignant transformation. Overall, the demographics, including the mean age, male to female ratio, rate of recurrence, and average time to recurrence in this population are very similar to those reported in the literature, such as study by Kim et al 7 10% of specimens in this study were positive for transformation when including lesions demonstrating dysplasia and metaplasia. Of these, however, only 3.4% of the specimens demonstrated carcinoma on surgical pathology at the time of resection, which is lower than the rate of about 10% reported in the literature. 2
The aggregated data from this study and the previous study by Vorasubin at the same institution, demonstrates significant changes in the subtype distribution across time. The proportion of inverted papilloma rose significantly over time, while the exophytic subtype decreased. Other studies have shown a wide range in the distribution of subtypes from 47-78% inverted, 6-50% exophytic, and 2-26% oncocytic. 8 More recent literature supports the incidence of IP to be closer to 80-96%.1,9 We believe the significant rise in the proportion of inverted papilloma in our institution is mostly due to the increased awareness of surgical treatment for IPs, affecting referral practices as well as increased identification in pathology analyses in the recent cohort. It is possible that HPV prevalence could have partially contributed to the increased proportion of inverted papillomas, mirroring the rise in other HPV-positive head and neck cancers in the past three decades. However, the role of HPV in development of inverted papilloma remains unclear.4,5,10,11 Detection rates of HPV in sinonasal papilloma tissues range from 0 to 72% in the literature, averaging around 25%. 4 HPV may be more prevalent in papillomas with severe dysplasia or associated carcinoma, and yet HPV has not been found to be a reliable parameter to detect malignancy or recurrence.5,12,13 Limited viral transcriptional activity of HPV in papillomas with detected HPV DNA also argues against HPV playing an active role in the development of sinonasal papillomas. 10 In this institution HPV serotyping is not commonly performed on sinonasal papilloma specimens, and as such we are not able to provide additional information on this topic.
Vorasubin et al demonstrated a significant association between subtype and attachment location, which was also demonstrated in this study. 6 In both studies, inverted papillomas were predominantly attached to the maxillary sinus, and to a lesser degree skull base, and exophytic papillomas were predominantly attached to the nasal septum, nasal floor, and inferior turbinate, both of which are consistent with the literature.7,8 The previous study showed oncocytic papillomas with a propensity for attachment to the ethmoid sinus, whereas this study had a much more balanced distribution between the lamina papyracea, maxillary sinus, skull base attachment sites.
Vorasubin et al also showed a significant association of age and history of chronic rhinosinusitis with subtype, which were not replicated in this study. In the previous study, exophytic papillomas were found in significantly younger patients with a mean age of 44.2 compared to inverted papillomas (56.9) and oncocytic papillomas (59.6). In this study, the average age was similar between inverted papillomas (57.7) and exophytic papillomas (56.8), although the population with oncocytic papillomas was nearly 10 years older with a mean age of 67.8. In the previous study, the minority of patients (21%) with exophytic papillomas exhibited symptoms of chronic rhinosinusitis, in contrast to the majority in this study (87%). The comparison of data for exophytic and oncocytic papillomas between these two studies may be skewed due to the much lower number of exophytic and oncocytic papillomas in this study. In this study, the only other significant difference in demographics among subtypes was a higher proportion of females in the oncocytic subtype, whereas the other two subtypes are more heavily skewed toward males. This is consistent with the literature demonstrating a more balanced gender distribution in the oncocytic subtype. 8
Although age was not significantly different among the various subtypes, age was shown to be an important factor for recurrence. Patients presenting at a younger age had a higher chance of recurrence, which may be because more aggressive tumors can present at an earlier age. Of note, we observed no correlation between age of diagnosis and length of follow up. We also did not observe correlation between any of the studied factors and histopathologic transformation. This is likely attributable to the low power of the study with only 12 cases of transformation. Multicenter aggregate data, enabling larger sample size can better help identify factors that affect recurrence and malignant transformation.
Smoking and alcohol use were not significantly associated with any of the outcome measures in this study. Similarly, chronic rhinosinusitis and allergic rhinitis were not significant factors for subtype, recurrence, or transformation. We did not observe a higher recurrence rate among papillomas with histopathologic transformation. The most significant factor predicting recurrence, beside younger age at resection, was two or more prior surgery for papillomas (OR = 3.52) reflecting the aggressively recurrent nature of some of these tumors. The correlation between recurrence of sinonasal papilloma and history of two or more surgeries for reasons other than papilloma (OR = 5.81) could be due to undiagnosed papilloma in the prior surgery.
The time to recurrence followed a bimodal curve with late recurrent cases happenning on average at 61 months after surgery and early recurrences happenning on average 10 months after surgery. This was mirrored by the bimodal distribution of total months of follow up from surgery (49 and 89 months for early and late recurrent cases), compared to 19.4 months in the patients without identified recurrence. This makes physician's follow-up practices, and patient's tolerance for symptoms, the most likely contributors the timing of the identification of recurrence. Although the quality of the surgical resection versus de novo formation of papilloma after complete resection, can also be a contributor. 8 of 13 early recurrent cases, and 9 of 10 late recurrent cases underwent repeat surgery to remove tumor within 3 months of diagnosis of recurrence. Considering the malignant potential of the sinonasal papillomas, long-term follow-ups appear to be necessary to identify and resect recurrent tumors in a timely manner.
Conclusion
This study explored the demographic, clinical, and pathologic features of sinonasal papillomas as well as their associated risk factors for recurrence and transformation. These benign lesions exhibit high rates of recurrence and a chance for malignant transformation. While recurrence after two years is considered to be uncommon in the literature, in this 10-year study, we observed that a significant number of recurrences occurred after 40 months, which suggests that more frequent and long-term monitoring after removal of sinonasal papillomas could improve the identification of recurrences at an earlier stage. Factors significantly associated with recurrence included younger age at presentation and multiple prior sinus surgeries for papillomas or other reasons. Future studies should further characterize the etiologic factors that can be targeted to reduce the incidence and recurrence of these lesions.
Footnotes
Acknowledgments
The authors would like to thank the UCLA Head and Neck Surgery department for providing the patient data. The authors would also like to thank the UCLA Statistics Department for providing guidance on statistical analysis.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
