Maxillary Sinus Mucosa Preservation Versus Mucosal Sacrifice: Radiographic Outcomes After Inferior Maxillectomy and Soft‐Tissue Free Flap Reconstruction

ABSTRACT

Background

This study evaluates the impact of maxillary sinus mucosa preservation on radiographic outcomes after an inferior maxillectomy and soft‐tissue free flap reconstruction.

Methods

A total of 90 patients in two cohorts (mucosal sacrifice versus mucosa preservation) were evaluated from an institutional dataset. Imaging was reviewed at set time points and sinus scores were allotted based on a modification of the Lund—MacKay system. Patients were evaluated for changes in sinus scores. A univariate (UVA) and multivariate (MVA) analysis was performed.

Results

In gp1 (mucosal sacrifice), mean postoperative sinus score was 1.21 ± 1.30 (p = < 0.001) at 24 months. On UVA, advanced T‐stage at 12 months (OR 7.143, 95% CI 1.484–34.384, p = 0.014) and 24 months (OR 5.500, 95% CI 1.219–24.813, p = 0.027), and on MVA, advanced T‐stage were associated with increased odds for higher sinus score (OR 4.993, 95% CI 1.084–22.995, p = 0.039). In Gp2 (Mucosa preserved), the mean postoperative sinus score was 0.31 ± 0.47 (p = 0.021) at 24 months. On UVA and MVA, no predictors were identified for a higher sinus score.

Conclusion

Following an inferior maxillectomy and soft‐tissue free flap reconstruction, notably reduced sinus scores were seen in patients with preservation of the remaining maxillary sinus mucosa.

1. Introduction

As per the latest global data, oral cavity cancer contributed to an estimated 377 713 new cases [1]. Notably, more than half of the patients are often identified with advanced‐stage disease [2, 3]. Extensive resections are required for such advanced‐stage malignancies which can lead to significant reconstructive defects, often with oroantral communications, particularly after an inferior maxillectomy. In Taiwan, such defects are relatively frequent given the advanced stage of disease in its betel nut chewing population [4, 5]. In current practice, these defects are routinely restored with the standard‐of‐care free flap reconstruction.

Even though free flaps have been commonly used for maxillary reconstruction, scant data exit about its impact on maxillary sinus function, radiographic changes, and clinical symptoms over time. Previously, sinonasal complaints have been reported in 43.1% of patients after maxillectomy and free flap reconstruction, with 19% of them subsequently requiring endoscopic sinus surgery [6]. On prior review of our institutional data, an increase in postoperative sinus score was seen in this cohort of patients [7]. In the past, after extirpative surgery, curettage and removal of remaining sinus mucosa had been suggested to decrease the risk of sinusitis. However, no study has evaluated the impact of preserving the residual sinus mucosa on postoperative sinusitis and sinus function.

The current study extends from our previous research work, with a focus on investigating the potential impact of maxillary sinus mucosa preservation on radiographic changes and postoperative sinus scores in patients who have undergone an inferior maxillectomy with soft‐tissue free flap reconstruction. The authors believe that by analyzing radiographic changes and postoperative sinus scores, comprehension of long‐term maxillary sinus function in these patients will be enhanced.

2. Materials and Methods

2.1. Study Population and Design

We conducted a retrospective analysis of the institutional head and neck cancer (oral cavity) database of patients from 2005 to 2021. Patients with primary disease who underwent an inferior maxillectomy as a part of the ablative procedure (by single surgeon S.F.H.) followed by soft‐tissue free flap reconstruction were included in the study. As a part of routine clinical practice, a preoperative CT/MRI scan was obtained followed by postoperative surveillance CT/MRI imaging within 3, 6, 12, and around 24 months. Exclusion criteria for the study included patients with positive margins, incomplete/missing imaging studies, and those requiring a bony free flap reconstruction to restore the maxillectomy defect. This study was performed abiding by the Declaration of Helsinki and approved by the Institutional Review Board at Chang Gung Memorial Hospital (#202300540B0).

2.2. Study Groups and Imaging Evaluation

The study cohort consisted of two distinct patient groups; group 1: with mucosal sacrifice (from 2005 to 2018) and group 2: with mucosal preservation (from 2019 to 2021). In group 1 patients, as a part of surgical technique, after tumor extirpation, all visibly remaining maxillary sinus mucosa was removed or scraped off using curettes, whereas in group 2, the remaining sinus mucosa was preserved/not compromised.

As previously used by the authors, a modification of the Lund—MacKay staging system was employed for image evaluation [7]. Ipsilateral (congruent with disease laterality) maxillary sinus was graded from 0 to 3 based on the degree of opacification. Score 0 was allotted for no sinus opacification, score 1 for partial opacification, score 2 for complete opacification, and score 3 for sinus completely obliterated by free flap. Scores were allotted (by S.F.H.) for all preoperative and follow‐up imaging.

2.3. Study Variables

Demographic variables included age at the time of diagnosis (years) and gender. Cancer‐related variables included TNM staging, tumor subsite, disease laterality, and comorbidities. Treatment‐related variables included the type of free flap reconstruction and need for adjuvant therapy. Outcome variables were postoperative sinus scores and the difference in sinus scores between the two cohorts on modified Lund–MacKay staging.

2.4. Statistical Analysis

Frequencies were computed to describe demographic and clinicopathologic characteristics. The primary outcome of interest was changes in sinus scores over time and differences in sinus scores between patient groups. Given the ordinal nature of the data, the Wilcoxon signed‐rank test was used to compare scores within each group, and the Kruskal–Wallis test was used to analyze differences in scores between groups. A univariate (UV) and multivariate analysis was performed to evaluate sinus score (at 3, 6, 12, and 24 mos) and its relationship to age, radiation, comorbidities, and tumor stage (T). A logistic regression analysis was also performed using the mucosal sacrifice cohort as a variable. Significance was determined at an alpha level of 0.05. The statistical software R language version 4.1.1 (R core team, 2021) was used to perform all data analyses.

3. Results

3.1. Population Sample and Characteristics

A total of 90 patients met the inclusion criteria with 57 in group 1 (mucosa sacrificed) and 33 in group 2 (mucosa preserved). Within this population, patients were predominantly male (98%; n = 88), with buccal mucosa being the most common subsite (74%, n = 67). The median age for the entire cohort was 55.8 years with an interquartile range (IQR) of 14.6 years. Additionally, 63% (n = 57) of patients had T3/T4 classification disease (advanced T‐stage) and 73% (n = 66) received surgical excision with adjuvant therapy (radiation only or chemoradiotherapy). Complete demographic and clinicopathologic characteristics for both groups are included in Table 1.

TABLE 1.

Demographic, clinical, and pathologic characteristics of the patient cohort.

Group	Group 1 mucosa sacrifice	Group 2 mucosa preserved	p a
Sample size	N = 57	N = 33
	Median (IQR)	Median (IQR)
Age (years)	54.8 (14.1)	57.6 (14.6)	0.23
	n (column percent)	n (column percent)
Sex			1.00
Males	56 (98.2%)	32 (97.0)
Females	1 (1.8%)	1 (3.0)
T category			0.02
T1–T2	26 (45.6%)	7 (21.2%)
T3–T4	31 (54.4%)	26 (78.8%)
Adjuvant treatment b			0.16
No	18 (31.6%)	6 (18.2%)
Yes	39 (68.4%)	27 (81.8%)
Sub‐sites			0.19
Buccal	44 (77.2%)	22 (66.7)
Mandibular gingiva	2 (3.5%)	4 (12.1%)
Maxillary gingiva	7 (12.3%)	7 (21.2%)
Retromolar fossa	3 (5.3)	0 (0.0%)
Soft palate	1 (1.8%)	0 (0.0%)

Abbreviations: IQR, interquartile range; n, number of patients; N, total number of patients.

^a P values obtained from two‐sample t‐test for age and chi‐square/Fisher's exact test for other variables.

^b Postoperative radiation or chemoradiation therapy.

3.2. Sinus Scores Based on Imaging: Modified Lund—MacKay Staging System

3.2.1. Group 1: Mucosa Sacrifice

A modified Lund—MacKay staging system was used to evaluate all scans [7]. The overall ipsilateral preoperative mean sinus score was 0.27 ± 0.44, and the postoperative mean sinus score at 24 months was 1.21 ± 1.30 (p = < 0.001). In comparison to the preoperative state, a significant increase in sinus score was noted at all postoperative time points (6, 12, and 24 months) (p = < 0.001). Complete results are presented in Table 2.

TABLE 2.

Sinus scores based on imaging at various time points between the two groups.

Sinus score analysis between groups
Study time points	Mucosa sacrifice		Mucosa‐preserving		Mucosa sacrifice versus mucosa preservation
	Mean sinus score ± SD	p a	Mean sinus score ± SD	p a	p b
Pre‐op	0.28 ± 0.45	ref	0.61 ± 0.50	ref	—
6 mos	1.25 ± 1.31	< 0.001	0.45 ± 0.63	0.197	0.014
12 mos	1.20 ± 1.30	< 0.001	0.31 ± 0.47	0.021	0.005
24 mos	1.21 ± 1.30	< 0.001	0.30 ± 0.47	0.034	0.007

Abbreviation: SD, standard deviation.

^a p value presented the significance of difference between each time point and pre‐op for 2 groups, respectively, using Wilcoxon signed‐rank test.

^b p value presented the difference between 2 groups at each time point using Kruskal–Wallis test.

3.2.2. Group 2: Mucosa Preservation

The overall ipsilateral preoperative mean sinus score was 0.61 ± 0.50, and the postoperative mean sinus score at 24 months was 0.31 ± 0.47 (p = 0.021). In comparison to the preoperative state, a significant decrease in sinus score was noted at postoperative time points 12 and 24 months (p = 0.021 and p = 0.034, respectively). Complete results are presented in Table 2.

3.2.3. Comparison in Sinus Scores Between Groups: Mucosa Sacrifice Versus Mucosa Preservation

Patients in the mucosa preservation cohort demonstrated a significantly lower sinus score on follow‐up imaging (6, 12, and 24 months) (p = 0.014, p = 0.005, and p = 0.007, respectively) in comparison to those where the sinus mucosa was sacrificed (Table 2).

3.3. Impact of Predictor Variables: UV, Multivariate, and Regression Analysis

3.3.1. Group 1: Mucosa Sacrifice

On UV analysis, advanced T‐stage at 12 months (OR 7.143, 95% CI 1.484–34.384, p = 0.014) and 24 months (OR 5.500, 95% CI 1.219–24.813, p = 0.027) was associated with significantly higher sinus score (Table 3). On MV analysis, advanced T‐stage continued to be associated with increased odds for higher sinus scores (OR 4.993, 95% CI 1.084–22.995, p = 0.039) (Table 4).

TABLE 3.

Association between sinus score and covariates at each time point: simple logistic regression model.

Sinus score correlation
Odds ratio (95% CI), p
	3 Mos	6 Mos	12 Mos	24 Mos
Group 1: Mucosal sacrifice group (n = 57)
Age	0.22 (0.01–2.6), 0.23	2.0 (0.33–11.9), 0.44	0.9 (0.19–5.0), 0.98	0.78 (0.18–3.4), 0.75
Comorbidities	0.19 (0.015–2.5), 0.2	0.6 (0.15–2.7), 0.55	0.59 (0.14–2.3), 0.45	0.3 (0.08–1.2), 0.09
Tumor stage	5.9 × 109 (0–NA), 1.00	2.2 (0.5–9.6), 0.29	7.1 (1.4–34.3), 0.01	5.5 (1.2–24.8), 0.02
PORT	3.6 (0.28–46.3), 0.32	1.8 (0.3–8.9), 0.44	2.3 (0.4–10.8), 0.28	2.4 (0.54–11.0), 0.24
Group 2: Mucosa‐preserving group (n = 33)
Age	0.54 (0.03–8.2), 0.66	1.02 (0.19–5.5), 0.97	1.1 (0.17–7.7), 0.88	1.1 × 109 (0–NA), 0.99
Comorbidities	0.14 (0.01–1.9), 0.14	2.1 (0.42–10.7), 0.35	2.4 (0.37–15.2), 0.35	1.9 (0.21–25.6), 0.44
Tumor stage	1.8 (0.15–23.3), 0.62	2.8 (0.27–29.5), 0.37	9.9 × 108 (0–NA), 0.99	1 × 109 (0–NA), 0.99
PORT	1.8 (0.15–23.3), 0.62	2 (0.18–22.5), 0.57	9.9 × 108 (0—NA), 0.99	9.4 × 108 (0–NA), 0.99

Note: NA was due to no cases in cells within 2 × 2 cross‐table, hence odds ratio approaching infinity.

Abbreviations: CI, confidence interval; Mos, months; NA, not accessible; PORT: postoperative radiation therapy.

TABLE 4.

Multivariate regression analysis of sinus score.

Sinus score at 24 months
OR (95% CI), p
	Mucosa sacrificed	Mucosa preserved
PORT	1.859 (0.371–9.305), 0.451	2.072 × 108 (0–NA), 0.999
T‐stage	4.993 (1.084–22.995), 0.039	3.587 ×108 (0–NA), 0.999

Note: NA was due to no cases in cells within 2 × 2 cross‐table, hence odds ratio approaching infinity.

Abbreviations: NA, not accessible; OR, odds ratio; PORT, postoperative radiation therapy.

3.3.2. Group 2: Mucosa Preservation

On UV analysis, none of the selected predictor variables including age, underlying comorbidities, advancing T‐stage, or PORT were associated with a significantly higher risk of sinusitis (Table 3). Similarly, in this cohort, on MV analysis, there were no significant odds for a higher sinus score associated with advanced T‐stage or PORT (Table 4).

3.3.3. Logistic Regression Analysis

Using the mucosal sacrifice group as our variable, a regression analysis was performed, and no significant difference was noted between the groups at any time point; marginal significance impacting sinus score was noted in the mucosa preservation group at 12 months (p = 0.095).

4. Discussion

Our main objective was to assess the potential impact of preserving the remaining sinus mucosa on radiographic changes and postoperative sinus scores, in patients who underwent an inferior maxillectomy with soft‐tissue free flap reconstruction for oral cancer. By assessing these radiographic changes, we aimed to shed light on the effects these surgical interventions may have on long‐term maxillary sinus function.

The key findings of this study demonstrate that patients whose remaining sinus mucosa is not sacrificed, after reconstruction, achieve a significantly lower postoperative sinus score as compared to patients whose remaining sinus mucosa was entirely sacrificed/scraped off. Additionally, within the cohort with mucosal preservation, there were no significant changes in sinus scores over time (preoperative vs. postoperative), and no identifiable predictor variables were noted on UV or MV analysis.

The existing literature lacks sufficient data to perform a comparative analysis of the outcomes of our study. A recent study by Lerner and colleagues investigated sinonasal issues following maxillectomy with free flap reconstruction [6]. Their findings revealed that 43% of patients experienced postoperative sinonasal complaints, with 19% eventually necessitating endoscopic sinus surgery. The type of flap used and radiation exposure did not exhibit statistically significant impacts within their cohort. Notably, the study primarily concentrated on clinical complaints and did not provide information on the correlation or progression of disease based on radiographic findings after free flap reconstruction. Previously, most authors have evaluated the effects of radiation on sinus mucosa, particularly in nasopharyngeal carcinoma patients [8, 9]. Their work has revealed that the effects of radiation on sinus disease appear to be most significant at 3 months and then showcase a gradual improvement with time [8, 9]. The data of this study further builds upon some of the work that the authors have previously performed with regard to sinus changes post soft‐tissue free flap reconstruction [7]. Earlier our study had focused entirely on patients whose remaining sinus mucosa was completely removed/scraped off after resection and prior to a soft‐tissue free flap inset. An overall increase in postoperative sinus score was seen, with advancing T‐stage being a key predictor for higher scores.

One of the challenges in the study is the lack of routine sinonasal endoscopic exams and missing records of clinical symptoms to correlate the radiographic findings; as some may question whether a lower sinus score implies the absence of disease or asymptomatic mild sinusitis. However, what becomes more evident is that in such patients the development of sinusitis is related to the integrity of the remnant mucosa and its mucociliary function. Herein we demonstrate a case of buccal cancer (pT2N0M0) that initially received an inferior maxillectomy (with mucosa preservation) and ALT free flap reconstruction (Figure 1a). After 1 year, there was an insite recurrence, which necessitated a salvage resection. Nonetheless, on intraoperative endoscopic examination, the mucosa was noted to be near normal in appearance (Figure 1b). This further strengthens the research findings that, with preserved mucosa during inferior maxillectomy, the sinus can eventually regain its intactness in both appearance and function. Additionally, none of the patients in our entire cohort (group 1 and group 2) received endoscopic sinus surgery during the follow‐up period. This was because of asymptomatic nature or lack of symptoms such as purulent rhinorrhea or severe nasal obstruction. We also acknowledge that evidence correlating Lund‐Mackay scores to subjective sinus assessment tools such as SNOT‐22 has previously been reported [10]; however, the authors believe that results of the study are an objective assessment that could perhaps be closely related to patient symptoms. Additionally, no significant difference was noted during a regression analysis; this is in part due to a small sample size in the mucosa preservation group with empty cells during the early T‐stage and PORT analysis which can bias the outcome of regression analysis and make it difficult to calculate HR. Hence, the authors believe that the optimal way to demonstrate the difference in sinusitis in mucosa preservation or sacrifice groups is by calculating of score difference as demonstrated in Table 2. Finally, the ability to completely obliterate the sinus is based on several factors such as patient's BMI, choice of free flap, size of the defect, and pliability of the free flap used. Hence, it may not be feasible to obliterate the sinus in all cases for the reconstructive surgeon. This is wherein the utility of this study lies, in evaluating sinus outcomes with preservation versus sacrifice of the sinus mucosa despite limitations and challenges that may exist with reconstructive surgery.

FIGURE 1.

(A) CT scan of a patient with a history of left buccal SCC after inferior maxillectomy and ALT free flap reconstruction with local recurrence after 1 year. (B) On intraoperative endoscopic exam, the maxillary sinus mucosa was normal in appearance and covered with a thin layer of mucus. Asterisk indicates the maxillary sinus ostium. [Color figure can be viewed at wileyonlinelibrary.com]

When taking these research findings into consideration, the readers must account for the heterogeneity between the two groups (mucosa‐preserving vs. mucosa sacrifice). Despite best efforts, given the paucity of such surgical cases and challenges with such strict imaging follow‐ups, it is hard to control patients between groups based on initial preoperative sinus scores to matched tumor stages to predict their impact on outcomes. Additionally, the mucosal sacrifice group also had a larger sample size which can skew the findings. Finally and most importantly, this is a radiographic study; the true cause‐effect of these findings on postoperative clinical sinus symptoms was not examined in this study. Despite a vital need for correlation between radiographic and clinical findings, as our institutional database did not uniformly consist of reliably recorded sinus symptoms or endoscopic examination, this could not be accomplished by our team. We anticipate that our findings will enhance comprehension of maxillary sinus changes after such surgical interventions, furnishing clinicians with valuable data to advance their understanding of postoperative sinus function.

5. Conclusion

In patients with oral cancer after an inferior maxillectomy and soft‐tissue free flap reconstruction, notably lower postoperative sinus scores were seen when residual maxillary sinus mucosa was preserved. This is in contrast to cases where the remaining sinus mucosa was entirely sacrificed. Conducting additional research that correlates such radiographic findings with clinical symptoms/findings may facilitate a more in‐depth exploration of long‐term maxillary sinus function in these patients.

Ethics Statement

This study was performed abiding by the Declaration of Helsinki and approved by the Institutional Review Board at Chang Gung Memorial Hospital (#202300540B0).

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.