Nonsurgical Treatment for Vocal Fold Leukoplakia: An Analysis of 178 Cases

Objective To assess the effectiveness and identify vocal fold leukoplakia types appropriate for nonsurgical treatment. Methods The vocal fold leukoplakia in 178 patients was divided by gross appearance into three subtypes: flat and smooth, elevated and smooth, and rough. All patients received nonsurgical treatment including smoking and drinking cessation, voice rest, omeprazole, and Chinese medication therapy. The clinical response of three subtypes was assessed after a 6-month follow-up. Results Vocal fold leukoplakia subtypes included flat and smooth (n = 66; 37.1%), elevated and smooth (n = 103; 57.9%), and rough (n = 9; 5.0%). The rate of complete response was 80.3%, 66.0%, and 0.0% for the 3 lesion types, respectively (rough versus flat and smooth, P < 0.001; rough versus elevated and smooth, P < 0.001, Fisher's exact test). The incidence of carcinoma in rough leukoplakia was significantly higher than that in smooth leukoplakia (44.4% versus 2.4%, P = 0.002, Fisher's exact test). Clinical type was the only significant factor for clinical response of nonsurgical treatment (P = 0.005, ordinal logistic regression). Conclusions The effectiveness of nonsurgical treatment for smooth vocal fold leukoplakia is better in comparison to rough vocal fold leukoplakia. Smooth leukoplakia could be managed with nonsurgical treatment; more aggressive treatments should be considered for rough leukoplakia.


Introduction
Vocal fold leukoplakia is clinically defined as white mucosal lesions that cannot be characterized as any other condition and is pathologically divided into two subtypes as follows: keratosis with nondysplasia and keratosis with dysplasia [1,2]. Leukoplakia without dysplasia does not convey premalignant potential and leukoplakia with dysplasia demonstrates premalignant potential [3]. However, benign and malignant lesions of vocal leukoplakia could not be discriminated clinically without a pathological biopsy; therefore, a consensus treatment strategy ranging from observation to complete resection for vocal fold leukoplakia has not been reached [4].
Vocal fold leukoplakia should be treated individually according to its benign or malignant possibility. A conservative therapy or observation strategy benefits those with a low risk of malignancy [1,5]. To data, surgical therapy remains the most widely studied modality of treatment. Ricci and Isenberg reported that approximately 50% of patients with clinical diagnosis of vocal fold leukoplakia do not have dysplasia [6,7], indicating that these patients received unnecessary surgical treatment.
Considering the macroscopic appearance, a classification and staging system of oral leukoplakia has been proposed [8]. However, besides a laryngoscopic imaging scoring system established by Fang et al. [9], there have been few reports about the clinical classification method of vocal fold leukoplakia to distinguish benign from malignant lesions. Thus, a method to classify the vocal fold leukoplakia can reflect the degree of lesions simply and comprehensively might be indispensable.
Some studies evaluated the effectiveness of nonsurgical intervention for oral leukoplakia [10,11], showing a significant effect of nonsurgical therapy. However, it is still unclear whether patients with vocal fold leukoplakia can benefit from nonsurgical treatment. Most studies have focused on the surgical treatment but ignored the nonsurgical treatment for vocal fold leukoplakia.
The purpose of this study was to propose a new classification method to vocal fold leukoplakia and to assess 2 BioMed Research International

Material and Methods
The protocol of this study was approved by the Institution  Figure 1. When vocal fold leukoplakia lesion had more than one morphological appearance on different locations, the lesion was categorized as elevated and smooth type if flat and smooth leukoplakia and elevated and smooth leukoplakia coexisted on vocal cords; rough leukoplakia was determined once rough lesion appeared on vocal cords.

Treatment.
Patients with rough leukoplakia were strongly recommended for vocal fold mucosal stripping by carbon dioxide (CO 2 ) laser. Nonsurgical treatment was conducted for patients with smooth vocal fold leukoplakia or patients with rough leukoplakia who had high-risk medical problems in surgery or strongly required to receive conservative treatment. Nonsurgical treatments included smoking and drinking cessation, strict voice rest, proton pump inhibitor (omeprazole 20 mg twice daily) therapy if accompanied with laryngopharyngeal reflux, and Chinese medication (Xuanboshuangsheng Granules 8 g twice daily). The main ingredients of Xuanboshuangsheng Granules are herbs including Radix Scrophulariae, Cortex Phellodendri, and Radix Glycyrrhizae (Drug Approval Number: Z05170495, Shanghai, China; Medical Institution: Ear, Nose, Throat Hospital of Fudan University, Shanghai, China; Associated Institution: Shanghai Liantang Pharmaceutical Corporation Limited).
Continuous nonsurgical therapy lasted for 6 weeks. Patients were seen at 2-to 4-week intervals in the first 3 months and 4-to 12-week intervals in the following months and evaluated by office-based laryngoscopic examination. Vocal fold mucosal stripping by CO 2 laser was performed for patients who had no improvements to previous nonsurgical treatment after a follow-up of 6 months.

Clinical Response Evaluation.
The laryngoscopic images of each patient before and after treatment were compared by three experienced laryngologists. Complete response (CR) was defined as complete disappearance of the lesion for at least 4 weeks. Partial response (PR) was defined as reduction in lesion size of 50% or more for at least 4 weeks. No response (NR) was defined as no significant change for at least 4 weeks, including stable disease, reduction of less than 50%, and lesions with increase of less than 25%. Progressive disease (PD) was defined as appearance of any new lesions not previously identified or estimated increase of 25% or more in existent lesions or the progression from smooth lesion to rough lesion [13]. The time to complete response of patients was the time from patient's initial visit until complete disappearance of lesion.  [14]. Squamous cell hyperplasia with nondysplasia describes increased cell numbers but the architecture shows regular stratification and there is no cellular atypia. Mild dysplasia describes slight cytological atypia, most marked in the basal one-third of the epithelium. Moderate dysplasia describes more cytological atypia, changes presenting in the lower two-thirds of the epithelium. Severe dysplasia describes cytological atypia involving more than two-thirds of the epithelial thickness. Carcinoma describes full thickness architectural abnormalities in the viable cellular layers accompanied with cytologic atypia.

Statistical Analysis.
All statistical analyses were performed using SPSS software version 23.0 (IBM Corporation,

Characteristics of Patients.
A total of 178 patients with vocal fold leukoplakia treated with nonsurgical therapy were included in this study. The characteristics of baseline patient information are showed in Table 2. Of these case, 66 (37.1%) showed flat and smooth leukoplakia, 103 (57.9%) showed elevated and smooth leukoplakia, and 9 (5%) showed rough leukoplakia. 171 (96.1%) patients were male and only 7 (3.9%) patients were female, whose average age was 49.8 ± 8.9 years (ranging from 28 to 73 years). The mean age of the patients with rough leukoplakia was significantly older than the patients with smooth leukoplakia (rough versus flat and smooth, < 0.001; rough versus elevated and smooth group, < 0.001). Otherwise, the groups were well balanced.

Clinical Response after Nonsurgical Treatment.
Clinical response of patients with vocal fold leukoplakia is showed in   Figure 1.
The mean time to complete response was 55.3 ± 38.3 days of two smooth types. Complete response rate of smooth types was subsequently estimated by Kaplan-Meier survival curves and compared using log-rank test. Kaplan-Meier analysis showed that the 3-month complete response rate was 72.7% for flat and smooth type compared with 60.2% for elevated and smooth type. After 6 months, complete response rates were 80.3% and 66.0%, respectively ( = 0.075, log-rank test; Figure 2).  Proportion without complete response (%) The ordinal logistic regression analysis showed that clinical type ( = 0.005) was the only significant variable that influenced the clinical response. Other factors, including gender, age, smoking, alcohol consumption, laryngopharyngeal reflux, voice abuse, and site of lesions, appeared to be insignificant with clinical response (Table 4).

Postoperative Pathological Results.
A total of 51 patients who had no improvements to nonsurgical therapy after a follow-up of 6 months received surgery. Of these cases, 11 (21.6%) showed flat and smooth leukoplakia, 31 (60.8%) showed elevated and smooth leukoplakia, and 9 (17.6%) showed rough leukoplakia. The pathological results of vocal fold leukoplakia are listed in Table 5. The incidence of carcinoma in rough leukoplakia was significantly higher than that in smooth leukoplakia (44.4% versus 2.4%, OR, 32.8, 95% CI, 3.04 to 354.4, = 0.002, Fisher's exact test). The correlation coefficient of three morphological groups and pathological grades was < 0.001. The result of Kruskal-Wallis test followed by Nemenyi test noted that statistical significant differences between rough leukoplakia and smooth leukoplakia were observed, respectively (rough versus flat and smooth, < 0.001; rough versus elevated and smooth, = 0.008).

Discussion
Vocal fold leukoplakia can be histologically diagnosed as squamous cell hyperplasia, mild dysplasia, moderate dysplasia, severe dysplasia, and carcinoma according to pathological classification systems [14]. However, there have been few reports about clinical classification of vocal fold leukoplakia. Oral leukoplakia was divided into two subtypes as nonhomogenous and homogenous [8]. Lee et al. divided vocal fold leukoplakia into three morphological groups: superficial type, exophytic type, and ulcerative type [15]. Fang et al. proposed a method to categorize the vocal fold leukoplakia based on morphologic characteristics scoring, which included thickness, texture, color, hyperemia, size, and symmetry [9]. Similarly, this study proposed a new morphological classification of vocal fold leukoplakia. In the last decades, new endoscopic tools, especially narrow band imaging, have been used for clinical classification of vocal leukoplakia based on microvascular changes [16], whereas the present classification according to macroscopic appearance provides a valuable source of laryngoscopic examination, which is more commonly applied in clinical practice.
There is still no agreement on the management of vocal fold leukoplakia. To data, surgical treatment has been suggested as an option [1,4]. Although the disappearance and reduction of oral leukoplakia with nonsurgical therapy have been documented in the past [17,18]. To our knowledge, there have been few records about the effectiveness of nonsurgical therapy for vocal fold leukoplakia. Xu et al. found that a complete response up to 85% was observed in vocal leukoplakia with andrographolide therapy with a follow-up of 12 months [19]. In our study, 127 of 178 patients (71.3%) with nonsurgical treatment had complete or partial response. Additionally, we analyzed the time to complete response (mean ± SD, 55.3 ± 38.3 days) of nonsurgical treatment in vocal fold leukoplakia for the first time. These findings demonstrated that some lesions of vocal fold leukoplakia might disappear or decrease in size without surgical therapy and these lesions might benefit from nonsurgical intervention.
A study of oral leukoplakia without surgical treatment demonstrated 32.5% of homogenous lesions and 24.3% of nonhomogenous lesions, respectively, disappeared or reduced [20]. Likewise, the analyses of data (Table 3) noted that the effectiveness of smooth leukoplakia was better in comparison to rough leukoplakia. Nonsurgical treatment exhibited significant curative effects to smooth leukoplakia. Result of Kaplan-Meier analysis noted that there was no significant difference of complete response rate for two smooth types. Elevated and smooth leukoplakia behaves similar to flat and smooth leukoplakia and therefore should be managed similarly.
The risk factors including tobacco smoking, alcohol intake, voice abuse, and laryngopharyngeal reflux might be related to vocal fold leukoplakia [14]. It was reported that tobacco smoking is the most important factor that could increase the disappearance of oral leukoplakia [17]. However, we made comprehensive analysis based on various clinical factors including clinical type, gender, age, smoking, alcohol use, voice abuse, laryngopharyngeal reflux, and site of lesions. The only significant factor associated with clinical response was clinical type of vocal fold leukoplakia. Based on our multivariate analysis regression models, patients who present with a smooth vocal fold leukoplakia would best be served by nonsurgical treatment and patients who present with a rough vocal fold leukoplakia would need aggressive therapy. Additionally, it remains unknown whether pathological grade of vocal fold leukoplakia would affect the clinical response of nonsurgical treatment. In the present study, this issue on vocal fold leukoplakia is unable to be investigated since pathological results cannot be determined without a biopsy which might deteriorate the quality of voice.
Vocal fold leukoplakia should be managed on its benign and malignant possibilities. Hyperplasia with nondysplasia or mild dysplasia could not be regarded as a precancerous lesion of larynx and should be managed with no surgical intervention; the lesion with more than moderate dysplasia should be managed more aggressively [5,21,22]. In the present study, vocal fold mucosal stripping by CO 2 laser was performed in 51 patients who showed no improvements upon previous nonsurgical treatment. The results of the pathological diagnosis showed that smooth lesions mainly presented with nondysplasia and mild dysplasia in pathology, whereas rough lesions mainly presented with severe dysplasia and carcinoma. We believed that this classification method was useful for differentiating between benign and malignant lesions. Additionally, following clinical data of three groups compared (Table 2), the mean age of the patients with rough leukoplakia was significantly older than those with smooth leukoplakia; therefore, age was an important factor to consider when we identify vocal fold leukoplakia types appropriate for nonsurgical treatment based on this classification method.
The first limitation of this study is lack of a control group of patients receiving surgery. Secondly, the number of patients with rough leukoplakia was only 9 due to our recommendation that patients with rough leukoplakia should be treated with surgery in consideration of malignant risk. Thirdly, the relationship between clinical type and pathological grades needs to be studied with a larger sample size. Lastly, a prospective cohort study is required to validate the usage of the classification method and recognize the effect of nonsurgical treatment.

Conclusion
The effectiveness of nonsurgical treatment for smooth vocal fold leukoplakia is significantly better in comparison to rough vocal fold leukoplakia. The classification method is recommended to guide the decision-making about indications for management. In general, smooth leukoplakia could be managed with nonsurgical treatment; more aggressive treatments should be considered for rough leukoplakia.