Vascularity Outcomes of Lingual Artery Ligation in Transoral Robotic Base of Tongue Resections

Methods

Patient Selection

The UAMS institutional review board (#260316) approved a retrospective review on 25 patients (21 male, 4 female) who underwent TORS for base of tongue (BoT) OPSCC at our tertiary academic institution (May 2011 to December 2019). Hounsfield units (HU) of the intrinsic muscles (IMs) and genioglossus muscles (GGs) were measured by a staff radiologist as a surrogate marker for tongue vascularity in postoperative contrasted computed tomography (CT) images (range, 2-9 months; mean, 4 months).^11-14 In iLAL patients (n = 15), the values from the ligated/resected side were compared with the nonligated, contralateral hemi-tongue ( Figure 1 ), as well as with patients with nonligated vessels (nLAL, n = 10) ( Figure 2 ). Comparisons were also performed in nLAL patients between the side of resection and the contralateral hemi-tongue and between individual GG and IM values as well as averaged IM and GG values, using independent samples t test and paired t test. Categorical variables were analyzed using χ² and paired t test.

OPEN IN VIEWER

Figure 1.

Comparison of side of tumor resection to contralateral hemisphere.

OPEN IN VIEWER

Figure 2.

Comparison of side of tumor resection between ligated and nonligated patients.

Results

No significant differences were found in demographic, pathologic, or preoperative variables among cohorts ( Table 1 ). One significant, nonfatal bleeding event occurred intraoperatively in the nonligation group.

OPEN IN VIEWER

Table 1.

Patient Demographics and Clinicopathologic Data Comparison.

Characteristic	Nonligated (n = 10)	Ligated (n = 15)	P value
Age at surgery, mean ± SD (range), y	59.78 ± 5.07 (55-69)	55 ± 9.79 (38-71)	.17
Sex, No.			.13
Male	7	14
Female	3	1
Race, No.
White	9	11	.32
African American	1	2	.81
Asian	0	2	.24
Time from surgery to postoperative CT, mean ± SD, d	125.1 ± 42.34	129.73 ± 66.24	.85
XRT prior to CT, No. (%)	7 (70)	13 (87)	.31
Major bleeding, No. (%)	1 (10)	0	.22
T-stage, No.			.28
Tx	2c	0
T1	7	10
T2	1	5

Abbreviations: CT, computed tomography; XRT, radiation therapy.

Comparison of the IMs and GGs on the side of ligation/resection to the contralateral side revealed no significant difference in HU (P = .662 and .618, respectively; Figure 1 ). Comparison of the side of resection between cohorts demonstrated significantly higher HU in nonligated patients on IM measurements but no difference between GG values (P = .050 and .818, respectively; Figure 2 ). No significant difference was appreciated, however, upon comparison of combined muscle groups (average of IM + GG) between cohorts with respect to laterality (P = .212; Figure 2 ). Neither gross tongue atrophy nor complications in the neck were incurred as a result of the performance of iLAL during the study.

Discussion

As utilization of TORS for OPSCC increases, optimization of the surgical technique to minimize complications is paramount. iLAL has been shown to significantly decrease the severity of OPH in several studies.^5-8 Our investigation adds to the literature as reported on the effect of iLAL on human tongue vascularity is not reported.

Using HU, we quantitatively examined the effects of iLAL on tongue vascularity following intervention (mean 4 months). The nLAL IM on the side of resection demonstrated significantly higher HU measurements and thus vascularity than all other IM groups. While this may be an expected finding compared to the iLAL/resected IM (Figure 2;P = .050), a finding of significantly lower vascularity of the contralateral nLAL/nonresected IM is certainly interesting (Figure 1;P = .039). This may be attributable to ongoing healing/remodeling within the surgical bed on the side of resection and thereby increased blood flow at the time of postoperative imaging.

Considering the iLAL/resection IM showed no significant difference from the contralateral IM (P = .662) and significantly lower vascularity in comparison to the nLAL/resection IM (P = .050), it is reasonably inferred that performance of iLAL was effective. This finding correlates well with other studies demonstrating higher incidence of severe OPH in nLAL patients. Further studies are needed to explore the above hypotheses concerning the etiology and consequence of observed trends in vascularity.

Although we did find significant differences in vascularity between IM muscle groups, none were appreciated when combining muscle groups (IM + GG) with respect to laterality in contralateral comparison of iLAL patients or iLAL vs nLAL comparisons, respectively (P = .626, P = .212; Figures 1–2). These findings suggest no significant change in vascularity between each hemi-tongue as a result of iLAL at 4 months postoperatively. The absence of statistically significant differences within these comparisons may be the most clinically relevant findings in the study.

While this investigation serves well as a pilot study, it is not without important limitations. A post hoc power analysis was performed for all significant findings, demonstrating inadequate study power to definitively reject the null hypothesis. Due to small cohort size and limited power, the findings of this study should be interpreted with caution. Further limitations exist in the relatively short-term follow-up (~4 months), preventing conclusive assessment of long-term vascular changes. Large-volume studies with longer follow-up are warranted to characterize long-term outcomes in this patient population.

Conclusion

iLAL may have no significant effect on tongue vascularity with respect to laterality in the early postoperative period. While these novel findings are encouraging and support the iLAL performance, further studies are warranted to corroborate our findings as well as demonstrate safety and efficacy, particularly with regard to long-term effects.