Facial paralysis after superficial parotidectomy: analysis of possible predictors of this complication.

INTRODUCTION: Salivary gland tumors represent 3-10% of all head and neck neoplasms. These tumors occur predominantly in major salivary glands. The parotid gland is affected most often, ranging from 36.6% to 83%. The pleomorphic adenoma comprises 45-60% of all salivary gland tumors. Several surgical approaches have been described to treat this tumor. Lesion of the facial nerve is one of the most serious complications that can occur after parotid gland surgery.
OBJECTIVES: To determine possible predictive factors related to the occurrence of peripheral facial paralysis (PFP) after superficial parotidectomy in the surgical treatment of the pleomorphic adenomas of the parotid gland.
METHODS: This was a primary, observational, case-control study performed through the revision of patients' charts and histopathological reports. Data was obtained from 1995 to 2014. The analyzed events were: tumor's length and depth; duration of the disease referred by the patient (more than 1, 5 or 10 years); primary or secondary surgical approach.
RESULTS: The analysis showed that tumor lengths equal or superior to 3.0cm were a risk factor of PFP with an odds ratio of 3.98 (p=0.0310). Tumor depths equal or superior to 2.0cm were also a risk factor with an odds ratio of 9.5556 (p=0.0049). When the tested event was secondary surgery to recurrent tumors we have found an odds ratio of 6.7778 (p=0.0029).
CONCLUSION: Tumors with 3.0cm or more in length and/or 2.0cm or more in depth have a significant higher risk of facial nerve injury. Secondary surgery to recurrent tumors also has a much higher risk of evolving with facial palsy after superficial parotidectomy.

Introduction

Salivary gland tumors represent 3–10% of all head and neck neoplasms. These tumors occur predominantly in major salivary glands.2, 3

The parotid gland is affected most often, ranging from 36.6% to 83%. Malignant tumors are a minority, occurring in 15–32% of cases.1, 4, 5

The most common malignant and benign tumors are the mucoepidermoid carcinoma and the pleomorphic adenoma, respectively.6, 7

The pleomorphic adenoma comprises 45–60% of all salivary gland tumors. Approximately 80% occur in the parotid gland, usually in the inferior pole of the superficial lobe; however, less frequently, it can occur at the deep lobe or in the accessory parotid tissue.

Several surgical approaches have been described to treat this tumor. In 1895, Senn described enucleation as the technique of choice. However, tumor removal was incomplete with an unacceptable rate of recurrence.

Total parotidectomy removes all gland tissue lateral and medial to facial nerve, whereas superficial parotidectomy removes parotid gland lateral to the facial nerve. In extracapsular dissection, it is not performed any dissection of the facial nerve.8, 10, 11, 12, 13, 14

Lesion of the facial nerve is one of the most serious complications that can occur in parotid gland surgery. It is estimated that 30–65% of all patients experience some sort of transient facial weakness, and 3–6% evolve with permanent dysfunction resulting in significant impact in the quality of life.

The main purpose of this study was to determine possible predictive factors related to the occurrence of peripheral facial paralysis after superficial parotidectomy in the surgical treatment of the pleomorphic adenomas of the parotid gland.

Methods

This was a primary, observational, case-control study performed through the revision of patients’ charts and histopathological reports. Data was obtained from 1995 to 2014. Patient's names were not retrieved; all information was obtained through hospital's registration numbers.

The analyzed events were: tumor's length and depth; duration of the disease referred by the patient (more than 1, 5 or 10 years); primary or secondary surgical approach. The dimensions evaluated were obtained from the histopathological reports and not from the clinical staging, to give a more precise value.

We divided the patients into groups according to the studied event:

Tumor length

Event analyzed: tumor length ≥0.5 cm, 1.0 cm, 1.5 cm, 2.0 cm, 2.5 cm or ≥3.0 cm.

Outcome: Peripheral facial paralysis.

Cases: Patients who underwent superficial parotidectomy with tumor length greater than previously referred measures.

Controls: Patients who underwent superficial parotidectomy with tumor length smaller than previously referred measures.

Tumor depth

Event analyzed: tumor depth ≥0.5 cm, 1.0 cm, 1.5 cm or 2.0 cm.

Outcome: Peripheral facial paralysis.

Cases: Patients who underwent superficial parotidectomy with tumor depth greater than previously referred measures.

Controls: Patients who underwent superficial parotidectomy with tumor depth smaller than previously referred measures.

Recurrence of the disease

Event analyzed: secondary surgery to treat recurrent tumor.

Outcome: Peripheral facial paralysis.

Cases: Patients who underwent superficial parotidectomy with a recurrence of the tumor from previous surgery.

Controls: Patients who underwent superficial parotidectomy for a primary tumor.

Duration of the disease

Event analyzed: Time referred by the patient from the moment the lesion was noticed until surgical treatment.

Outcome: Peripheral facial paralysis.

Cases: Patients with duration of the disease superior to 1, 5 or 10 years.

Controls: Patients with duration of the disease inferior to 1, 5 or 10 years.

Facial paralysis was classified and graduated according to House-Brackmann's scale (HBS). Patients were evaluated at the first and seventh postoperative days; first, third and sixth months after surgery. And, in the first year after surgery.

All patients whose charts and histopathological reports did not fulfill these criteria were excluded from the study. Patients who have had any kind of facial motricity impairment previous to surgery were also excluded.

Data was then analyzed to calculate odds ratio, confidence interval, z-statistic, and significance level (p < 0.05).

Results

During the studied period 219 patients were submitted to surgical approach in the parotid gland; 157 were due to pleomorphic adenoma.

After excluding those submitted to total parotidectomy and those whose information was incomplete there was a total of 84 patients who underwent superficial parotidectomy.

Males were 28 patients and females, 56. The average age was 56.7 years (17–89). Average length was 2.892 cm (0.8–6.0 cm) and average depth was 1.880 cm (0.5–3.2 cm). The right side was affected in 46 cases, and the left side, in 38 cases.

Facial paralysis occurred in 14 out of 84 patients (16.66%), four males and ten females.

Facial paralysis was graded according to HBS. Five patients had an HBS II, seven HBS III and two HBS IV. None was classified as HBS V or VI.

Time elapsed from initial paralysis to complete recovery was evaluated in weeks. One patient with HBS II had full recovery within two weeks. On the other hand, an HBS IV patient had full recovery only on the 36th week. No patient remained with facial paralysis after 36 weeks. Medium time until full recovery was 19.19 weeks (Fig. 1).

Figure 1

Facial paralysis grade versus elapsed time until complete recovery.

To evaluate tumor's dimensions several values were sequentially tested in a 0.5 cm interval. Regarding tumor's length we have found no statistical significance when tumor's length was less than 2.5 cm; however, when a 3.0 cm cutoff point was set up we have found an odds ratio of 3.98 with a confidence interval of 1.1345–13.9728 (p = 0.0310) (Table 1).

Table 1

Case-control table – tumor's length.

Facial paralysis
Length	Yes	No
≥3.0 cm	10	27
<2.9 cm	4	43

The same was done with tumor's depth, and with a 2.0 cm depth we have found an odds ratio of 9.5556 with a confidence interval of 1.9832–46.0422 (p = 0.0049) (Table 2).

Table 2

Case-control table – tumor's depth.

Facial paralysis
Depth	Yes	No
≥2.0 cm	12	27
<1.9 cm	2	43

When the tested event was secondary surgery to recurrent tumors we have found an odds ratio of 6.7778 with a confidence interval of 1.9227 to 23.8928 (p = 0.0029) (Table 3).

Table 3

Case-control table – primary versus secondary surgery.

Facial paralysis
	Yes	No
Secondary	7	9
Primary	7	61

Considering the duration of the disease, the evaluated parameters were more than 1, 5 and 10 years. We have found odds ratios of 1.3684 with a confidence interval of 0.2725–6.8709 (p = .7032), 0.1784 with a confidence interval of 0.0099–3.2072 (p = 0.2423), and 0.5096 with a confidence interval of 0.0260–9.9975 (p = 0.6571), respectively (Table 4, Table 5, Table 6).

Table 4

Case-control table – duration of the disease (1 year).

Facial paralysis
	Yes	No
≥1 year	12	57
<1 year	2	13

Table 5

Case-control table – duration of the disease (5 years).

Facial paralysis
	Yes	No
≥5 year	0	11
<5 year	14	59

Table 6

Case-control table – duration of the disease (10 years).

Facial paralysis
	Yes	No
≥10 year	0	4
<10 year	14	66

Discussion

This was a retrospective, case-control study that analyzed a significant number of patients and used the pleomorphic adenoma's clinical and morphological features to determine possible predictive factors of facial paralysis after superficial parotidectomy.

To the best of our knowledge, we have found no previous similar study in the reviewed literature.

The exclusion of doubtful data and the usage of measures obtained from high quality clinical and histopathological reports made the evaluation very precise and reliable.

In this study, the surgeries were performed by residents under the strict supervision of a senior Head and Neck surgeon. Patients with larger and/or deeper or recurrent tumors were operated by senior surgeons.

This may be considered a bias of the study since more than one surgeon performed the surgeries; however, this fact reflects best the reality of most medical residence centers in Brazil. Therefore, this information could be useful to predict and best prevent this undesirable outcome by the hands of training residents.

Intraoperative neural monitoring (IONM) was not routinely used due to hospital's limitation quota. Therefore, its use was reserved to recurrent tumors.

Females were 56 cases, and males, 28. Of the 14 patients with postoperative facial paralysis, 10 were females and four were males. No permanent lesions were reported, and all patients were fully recovered within 36 weeks. No severe facial paralysis (HBS V or VI) was also reported during the studied period.

Transient facial palsy remains the most common complication in parotid gland tumor and an important factor of distress both to patients and the surgical team. Permanent palsy is seen in 0–19% of the cases.15, 16, 17, 18, 19, 20, 21

If the lesion is permanent the consequences are tremendous, resulting in possible cornea ulcers, facial asymmetry, dysphasia, and drooling.

Superficial parotidectomy remains the most efficient technique yet available, allowing the surgeon through the complete facial nerve dissection, with better chances of preserving its function.

It is still a matter of debate whether the incidence of facial paralysis is higher after malignant tumors resection, due to a more aggressive surgical approach, or after benign lesions, usually with a longer duration of the disease, associated with tumor's adherence and adjacent inflammatory process.

It is quite intuitive to relate the increase of tumor's dimension to a higher incidence of complications. In this study, we have demonstrated that tumors with 3.0 cm or more in length and 2.0 cm or more in depth have a significant risk of evolving with facial paralysis. This should be taken into account in the pre-operative evaluation, and consequently, demand a much more careful technique during nerve dissection.

All of the 16 cases of recurrence occurred after enucleation of the lesion performed elsewhere, since this approach is not preconized at our institution. As seen in this study, the secondary surgery has a much higher risk of resulting in facial palsy, surely due to perilesional inflammation, fibrosis and lack of anatomical landmarks.

Surgery performed after a previous approach of the parotid gland has an odds ratio of 6.7778 with a confidence interval of 1.9227–23.8928 (p = 0.0029). This should be recognized when performing the primary surgery, demanding a curative approach to avoid a secondary and, therefore, more risky surgery and during the secondary surgery itself, since the risks of facial nerve injury are significantly higher.

On the other hand, duration of the disease had no correlation with a higher risk of facial nerve injury during superficial parotidectomy. The calculated odds ratios were 1.3684 with a confidence interval of 0.2725–6.8709 (p = 0.7032); 0.1784 with a confidence interval of 0.0099–3.2072 (p = 0.2423), and 0.5096 with a confidence interval of 0.0260–9.9975 (p = 0.6571), with 1, 5 or 10 years of disease, respectively.

One might expect to find a higher risk of facial nerve injury with longer disease period. The explanation why that was not found in this study remains unclear. Further studies focusing at how the microscopic ultrastructure of tumor's boundaries changes over time might raise more information on this matter.

Conclusion

Facial nerve injury during superficial parotidectomy remains the most common and most feared complication.

In this study, we have found that tumors with 3.0 cm or more in length and/or 2.0 cm or more in depth have a significant higher risk of facial nerve injury.

Secondary surgery to recurrent tumors also has a much higher risk of evolving with facial palsy after superficial parotidectomy.

Conflicts of interest

The authors declare no conflicts of interest.