Key points

Introduction

Laryngeal carcinomas account for approximately 2.4% of new malignancies worldwide each year. According to the American Cancer Society, 10,000 new cases of laryngeal cancer are diagnosed in the United States annually and result in 3900 yearly deaths. More than 95% of laryngeal cancers are the squamous cell carcinoma (SCC) type. The glottic larynx is the most common site of occurrence of laryngeal SCC. Laryngeal cancers that are considered “early” typically include carcinoma in situ (CIS), T1, and T2 lesions without metastasis. In CIS, malignant cells are present but have not penetrated the basement membrane. T1 lesions are limited to one (T1a) or both (T1b) vocal folds, with normal vocal fold mobility. T2 lesions extend to the supraglottis or subglottis and may impair vocal fold mobility without vocal fold fixation.

Early detection of pathologic tissue change is of utmost importance for effective treatment and the preservation of function in glottic malignancy. It can be difficult to find a balance between ensuring adequate resection and favorable oncologic outcome with preserving laryngeal structure and function. Imaging has traditionally been an important adjunct in the diagnosis, staging, and monitoring of glottic neoplasms. Although direct laryngoscopy and biopsy are the gold standard for definitive diagnosis of glottic cancers, radiologic imaging modalities have traditionally provided essential information regarding overall staging and prognosis, resectability, and the feasibility of subtotal surgical options. Accurate identification of tumor margins within the larynx is also paramount to maximize oncologic outcome, because leaving positive margins increases the risk of local recurrence by 32% to 80%. The goal of this article is to provide an overview of advanced techniques in laryngeal imaging and their application to the diagnosis, treatment, and long-term follow-up of glottic neoplasms.

Introduction

Laryngeal carcinomas account for approximately 2.4% of new malignancies worldwide each year. According to the American Cancer Society, 10,000 new cases of laryngeal cancer are diagnosed in the United States annually and result in 3900 yearly deaths. More than 95% of laryngeal cancers are the squamous cell carcinoma (SCC) type. The glottic larynx is the most common site of occurrence of laryngeal SCC. Laryngeal cancers that are considered “early” typically include carcinoma in situ (CIS), T1, and T2 lesions without metastasis. In CIS, malignant cells are present but have not penetrated the basement membrane. T1 lesions are limited to one (T1a) or both (T1b) vocal folds, with normal vocal fold mobility. T2 lesions extend to the supraglottis or subglottis and may impair vocal fold mobility without vocal fold fixation.

Early detection of pathologic tissue change is of utmost importance for effective treatment and the preservation of function in glottic malignancy. It can be difficult to find a balance between ensuring adequate resection and favorable oncologic outcome with preserving laryngeal structure and function. Imaging has traditionally been an important adjunct in the diagnosis, staging, and monitoring of glottic neoplasms. Although direct laryngoscopy and biopsy are the gold standard for definitive diagnosis of glottic cancers, radiologic imaging modalities have traditionally provided essential information regarding overall staging and prognosis, resectability, and the feasibility of subtotal surgical options. Accurate identification of tumor margins within the larynx is also paramount to maximize oncologic outcome, because leaving positive margins increases the risk of local recurrence by 32% to 80%. The goal of this article is to provide an overview of advanced techniques in laryngeal imaging and their application to the diagnosis, treatment, and long-term follow-up of glottic neoplasms.

Indirect laryngoscopy

Perhaps the most useful examination tool in the general otolaryngologic practice is indirect laryngoscopy, which can be performed by either laryngeal mirror examination or flexible fiber-optic endoscope. Irregularities of laryngeal mucosa may be concerning for malignancy. Because differentiating malignancy from benign processes is not predictable based on gross appearance, direct laryngoscopy and biopsy are warranted. However, the burning question in every patient’s mind when they are diagnosed with a concerning laryngeal mass is what is the likelihood of malignancy. The largest study currently available evaluating clinical leukoplakia was performed by Isenberg and colleagues in 2008 and combined their 15-year institutional experience with a review of the literature from the prior 50 years. They noted that there was no dysplasia in 54% of leukoplakias, mild to moderate dysplasia in 34%, and either CIS or SCC in 15%. SCC later developed in 4% of patients with no dysplasia at the time of biopsy, in 10% of patients with mild to moderate dysplasia at the time of biopsy, and in 18% with CIS at the time of biopsy. Although a general guideline can be elicited for patient counseling from these findings, all suspicious lesions require biopsy confirmation.

Videostroboscopy

Videostroboscopy is a well-established method of imaging the vocal folds that allows the examiner to assess vocal fold vibration. It is a noninvasive in-office examination that requires only occasional topical anesthesia, results in minimal patient discomfort, and is an integral part of the standard workup for dysphonia. A synchronized flashing light is directed onto the vocal folds via a rigid or flexible endoscope, effectively providing a still photo of the vocal folds in the midst of vibratory motion. By synchronizing the stroboscopic light to the frequency of the repetitive vocal fold vibration, the viewer perceives vocal fold vibration at a much slower rate than the actually vibratory speed. Videostroboscopy is a useful tool in the workup and diagnosis of hoarseness. The visual effect of slower vocal fold oscillation is particularly helpful in the evaluation of abnormalities of laryngeal structure, vibratory asymmetry, and decreased or absent vibration. It can also characterize abnormalities in glottic closure and allows measuring of glottal gap. Stroboscopy has been shown to be able to accurately guide the diagnosis of benign midmembranous vocal fold lesions such as nodules, polyps, and cysts. Because of its ability to detect variations in vocal fold vibration, videostroboscopy may be used to characterize vocal fold epithelial lesions. Vocal fold hyperkeratosis has been shown to decrease the amplitude and inhibit the mucosal wave of the vocal fold on videostroboscopic examination. The mass effect on the vocal fold causes a lowered fundamental frequency. It has also been suggested that complete loss of vibration suggests early invasive carcinoma. The vocal ligament and intermediate layer of the lamina propria are largely responsible for vocal fold vibration amplitude. Amplitude will be reduced as lesions become more infiltrative because the freedom of the epithelium to vibrate is compromised. Colden and colleagues, however, showed that reduced amplitude of vocal fold vibration and mucosal wave propagation in vocal fold keratosis did not reliably predict the presence of malignancy or depth of invasion into the lamina propriae. They did suggest that an intact mucosal wave likely indicates that there is not extensive invasion into the vocal ligament. In reality, a cancerous lesion may demonstrate normal vibration amplitude and mucosal wave propagation if there is enough superficial lamina propria (SLP) underlying the lesion to allow for pliability of the epithelium. Alternatively, a lesion with normal amplitude and mucosal wave cannot be assumed to be benign. A benign lesion may cause a significant decrease in mucosal wave and amplitude perhaps because of compensatory phonotrauma confounding the examination. In addition, a cancerous lesion may cause complete loss of vibration and mucosal wave with no invasion into the vocal ligament. The ability of stroboscopy to detect abnormalities in the mucosal wave propagation, while not inherently diagnostic of malignancy, can alert the examiner of potentially cancerous lesions and lead to further workup.

High-speed imaging

High-speed imaging (HSI) is a laryngeal imaging technique that allows for thousands of images of the vibrating vocal folds to be taken per second, usually via a rigid endoscope. This technique overcomes some of the shortcomings of videostroboscopy, such as its dependence on periodic vibration and a minimum requisite phonation time of 2 seconds. Because stroboscopy requires periodicity to produce the strobe effect, it cannot accurately reveal vibratory patterns in cases of dysphonia caused by aperiodic vibration. HSI captures at least 2000 frames per second, or approximately10 to 20 frames per vibratory cycle depending on the fundamental frequency. Thus, it is not dependent on periodic vibratory motion and can reveal more about vibratory behavior than videostroboscopy. For example, HSI has been shown to be able to detect subtle features that indicate vocal fold paresis that are not evident on fiber-optic laryngoscopy or videostroboscopy. HSI is not yet widely available in clinical practice, however, and its practical applications continue to be defined. A study comparing the diagnostic accuracy of HSI compared with videostroboscopy found no difference between the 2 modalities in the diagnosis of 28 patients with dysphonia. The authors do, however, endorse the utility of HSI in challenging diagnostic cases such as vocal fold scar. Other studies have shown HSI to be more accurate and interpretable than stroboscopy in patients with vocal pathologic abnormality resulting in aperiodic voices. Patel and colleagues advocate that in cases of severe dysphonia with values exceeding 0.87% jitter, 4.4% shimmer, and a signal-to-noise ratio of less than 15.4 dB on acoustic analysis, HSI may aid in clinic decision-making. Like stroboscopy, HSI does not directly diagnose malignant vocal fold lesions, but can detect subtle abnormalities in the mucosal vibratory properties that may spur further workup of potentially cancerous laryngeal lesions. An example of a series of images created with HSI in a patient with vocal fold scar is shown in Fig. 1 .

A high-speed video montage demonstrating 5 glottal cycles from a patient with left vocal fold scar. There is asynchrony of cycle-to-cycle vibration. (Photo courtesy of Peak Woo, MD.)

Videokymography

Videokymography is an additional method to measure the vibratory capabilities of the vocal folds. Images from a single transverse line perpendicular to the glottal line are recorded, and successive images are shown in real-time on a monitor along with the time dimension. Once the pixel lines are extracted, they are configured consecutively side-by-side based on frame number to create a kymogram. The kymogram visualizes the motion of the mucosal wave, displaying the open and closed phases, periodicity, left to right symmetry, phase difference, and amplitude. Thus videokymography allows for the assessment of left to right asymmetries, open quotient, and propagation of mucosal waves. Both standard and high-speed modes are available, with the former capturing 50 to 60 images per second, and the latter providing nearly 8000 images per second. The earliest videokymography systems were only capable of generating one kymogram per examination. In addition, because the scanning camera uniquely displayed time on one axis and the single line of the video image on another axis, the 2-dimensional video imaging of the larynx was sacrificed. However, newer versions simultaneously provide a laryngoscopic and kymographic image simultaneously. The laryngoscopic image is used to select the position for capture of the kymographic image. A digital kymograph generated from a patient with a vocal fold polyp is displayed in Fig. 2 .

A digital kymograph of voice onset in a patient with a vocal fold polyp. The onset of vocal fold vibration that is delayed with reduced amplitude on the right.

(Photo courtesy of Peak Woo, MD.)

Unlike videostroboscopy in which many images are needed to analyze the vibrational pattern of the vocal folds, videokymography depicts the pattern in one image. The vibratory pattern may change along the length of the vocal fold, and therefore, multiple points may need to be evaluated. Videokymography has the potential to serve as an adjunct in the diagnosis of early glottic cancer. Schutte and colleagues described the case of a patient who had previously undergone partial cordectomy and radiation therapy for laryngeal cancer who presented with persistent dysphonia. Vocal fold mobility was limited and stroboscopic evaluation revealed low-amplitude vibration. Videokymography revealed no mucosal waves, suggesting tumor infiltration, which was later confirmed with biopsy.

Optical coherence tomography

Optical coherence tomography (OCT) is an imaging technology that provides cross-sectional images of subsurface tissue structure at approximately 10-μm resolution to a depth of 1.2 mm using backscattered light. The tissue sample is probed with infrared light, and interferometric methods are used to detect light reflected from within the tissue. Images are formed by dividing the light into 2 paths, one of which is directed at the tissue sample and another directed to a reference mirror. The light returning from the 2 paths is then compared. If the 2 paths have the same length and refractive index, the beams are identical or coherent and the signal will be above the threshold for detection. If the reference and probe beams have traveled different optical distances, the beams will not be coherent and there is destructive interference that prevents detection. This interference results in selective detection of light from fixed depth within the tissue, which can be used to produce a 2-dimensional image. In the larynx, the image formed can provide information regarding the structure of the vocal folds that is analogous to a vertical histologic section. Both benign and malignant vocal fold lesions may disrupt the mucosal layer of the vocal fold, leading to dysphonia. OCT offers a noninvasive method to characterize the structure of the epithelium and SLP in both normal vocal folds and in pathologic conditions and may be a valuable tool in both diagnosis and treatment planning.

OCT has been shown to provide information regarding the thickness of the epithelium, integrity of the basement membrane, and the structure of the lamina propria. Maturo and colleagues showed that OCT may be used to quantitatively analyze the layers of the lamina propria of the vocal fold. This feature of OCT has the potential to help to characterize the subepithelial development of the pediatric vocal fold and help determine the need for operative intervention in children with dysphonia. OCT has also been shown to have utility in guiding subepithelial injections into the vocal folds in real-time.

Polarization-sensitive optical coherence tomography (PS-OCT) is an additional form of OCT imaging that measures the intensity and polarization of state change of reflected light within the tissue in order to simultaneously characterize tissue structure and birefringence. Collagen fibers are birefringent and can change the polarization state of reflected light; thus, PS-OCT can detect the collagen content of the vocal ligament that differs from the overlying SLP, which contains much less collagen. Normal vocal fold tissue has a well-defined junction between the epithelium, which has low signal intensity, and the SLP, which has higher signal intensity and lacks significant structure, creating a light-dark-light banding pattern at the epithelium-SLP junction on PS-OCT images that indicates the presence of collagen in the SLP.

The ability of OCT and PS-OCT to noninvasively provide information about vocal fold structure gives it the potential to be a valuable adjunct in the diagnosis of early glottic malignancies. OCT provides valuable information about the structure of the vocal folds in benign, premalignant, and malignant conditions. It has the potential to guide biopsies and treatment as well as monitor disease progression and response to therapy. As previously described, PS-OCT and OCT can differentiate between the vocal ligament and SLP because of their differences in collagen content. This difference in tissue content allows these modalities to potentially identify epithelial lesions that disrupt the normal collagen pattern, such as invasive carcinoma. Burns and colleagues used PS-OCT and OCT to characterize the cross-sectional structure of the vocal folds of patients undergoing microlaryngoscopy for both benign and malignant lesions. Compared with normal glottic tissue, scar tissue displayed a more intense birefringence pattern, whereas cancer showed disruption or absence of the layered structure of the vocal fold and the birefringence pattern. In the vocal fold, biopsies and resections of lesions can lead to scarring that can have adverse effects on the voice of varying severity. PS-OCT and OCT could guide more accurate biopsies and resections by distinguishing potentially malignant areas from scar tissue or inflammation, thus preserving maximal normal tissue and improving functional outcomes.

Autofluorescence

Autofluorescence (AF) is the natural fluorescence emission of tissue arising from endogenous fluorophores after exposure and activation by radiation of a suitable wavelength. Its clinical utility stems from the fact that premalignant and malignant lesions can be differentiated from normal tissue because of decreased AF. Autofluorescence endoscopy (AFE) has been applied in several medical specialties to detect malignant mucosal changes. The different fluorescence emissions in AFE are due to neoplasia-induced changes in tissue morphology, optical properties, and the concentration of endogenous fluorophores. The intracellular fluorophores, nicotinamide adenine dinucleotide plus hydrogen and flavin adenine dinucleotide, are found in all tissue layers, but their concentration is nearly 100 times lower in malignant tissue than in benign tissue. Collagen and elastin are structural proteins and extracellular fluorophores that are found in subepithelial layers. Epithelial thickening of malignant and premalignant tissue also inhibits the penetration of exciting light into submucous layers, which also accounts for the reduced AF in neoplastic tissue.

As described, the differences in AF properties of malignant and benign tissues make this technology extremely useful in the diagnosis of early glottic cancers. Several studies have demonstrated the utility of AF in the larynx. Harries and colleagues in 1995 applied the lung imaging fluorescence endoscopy system, which uses a helium-cadmium laser light source (442-nm wavelength), to laryngeal lesions. By comparing biopsy specimens of areas of laryngeal mucosa with decreased AF (reddish-brown coloration) to specimens taken from mucosa with normal AF (green coloration), they established that this technology could effectively identify malignant lesions of the larynx. Zargi and colleagues compared AF to standard white light microlaryngoscopy in 108 patients and found sensitivities of 86.9% and 71%, respectively, for identifying malignant lesions in the larynx. Specificities were 82.8% and 80.6%, respectively. Combining the 2 methods yielded a sensitivity of 97.1% for cancerous lesions and 61.5% for precancerous lesions, with an overall specificity of 71.8%. They determined that incorrect assessments were due to bleeding, which can cause a false positive impression of malignancy because surface blood can diminish AF of underlying tissue; hyperemia, due to increased presence of blood within the tissue; and leukoplakia, which emits strong AF and can result in a false negative assessment in some instances. False positives may also be due to mild dysplasia with inflammatory reactions or vocal fold scarring. AF has also been applied through indirect laryngoscopy and shown to be a useful modality in the identification of early laryngeal cancers. Arens and colleagues generated AF images via a 70° rigid-angled endoscope and found 89% concordance with histopathology resulted in the identification of laryngeal malignancy. AF has been applied during transoral laser resection of early laryngeal cancers. Succo and colleagues performed a prospective cohort study with 73 patients undergoing CO ₂ laser resection of early glottic cancers. The use of AF was associated with superficial disease-free margins in 97.2% of cases and superficial close margins in 2.8%. Diagnostic accuracy was improved in 16.4%, and 8.2% of cases were upstaged as a result of AF use. They reported a sensitivity of 96.5% and a specificity of 98.5%. They concluded that AF can help identify positive superficial margins intraoperatively, leading to improved local control and disease-specific survival.

When aminolevulinic acid (ALA) is applied topically to laryngeal mucosa, it preferentially induces fluorescence within neoplastic cells; this is due to its role in heme synthesis: 2 ALA molecules condense to form porphobilinogen, which is then metabolized to fluorophore protoporphyrin IX (PTP). The enzyme ferrochelatase incorporates Fe+ into PTP to form heme, which is not a fluorophore. Neoplastic cells are more permeable to ALA and thus accumulate PTP, whereas ferrochelatase production is downregulated. Thus, when mucosa is treated with ALA, neoplastic cells emit the orange-red color of PTP, whereas healthy mucosa fluoresces green. Several studies have compared ALA-induced fluorescence to AF and found similar accuracy between the 2 modalities in the diagnosis of laryngeal dysplasia and invasive carcinoma. However, the addition of ALA-induced fluorescence to AF does not improve diagnostic accuracy. ALA-induced fluorescence may also be better than AF in differentiating recurrent cancer from scar tissue after laser surgery. Like AF, ALA-induced fluorescence cannot be used to determine histologic detail and thus cannot characterize grade of dysplasia or identify invasive carcinoma.

Biologic endoscopy techniques

Narrow Band Imaging

Narrow band imaging (NBI) is an optical technique that illuminates the intraepithelial papillary capillary loop using narrow bandwidth filters in a red-green-blue sequential illumination system. In order to support their growth requirements, all kinds of tumors require the recruitment of surrounding blood vessels and vascular endothelial cells. Tumors promote the growth of new blood vessels from pre-existing ones, and these new vessels have characteristic features including chaotic blood flow, tortuous and dilated structure, and excessive branches and connections. Their walls have many openings, widened interendothelial junctions, and a discontinuous or absent basement membrane. The NBI filter sets are 415 nm and 540 nm to provide images of the microvascular structure. Blue light, with a wavelength of 415 nm, is the hemoglobin absorption band, and therefore, capillaries on the surface of mucosa can be clearly visualized at this wavelength. The wavelength for green light is 540 nm, which penetrates the deeper tissues to enhance subepithelial vessels. Fig. 3 shows an example of a laryngeal image generated with NBI compared with a stroboscopic image from the same patient. NBI displays capillary patterns and can identify boundaries between different types of tissues, which can aid in the early identification of tumors. Superficial mucosal lesions that may be missed by white light endoscopy can be identified by their neoangiogenic pattern. In the head and neck, this technology has also been applied to the oropharynx, hypopharynx, and oral cavity. In these areas, superficial carcinoma appears as brown dots in a well-demarcated brownish area under NBI. This appearance is due to the microvascular proliferation pattern. Piazza and colleagues applied NBI coupled with a high-definition television camera both preoperatively and intraoperatively to 279 patients either undergoing workup for laryngeal SCC or who had previously undergone treatment of the condition. The findings obtained with NBI were compared with histologic analysis of biopsy samples, and high-definition NBI showed an overall sensitivity of 98% with a specificity of 90%. Kraft and colleagues compared NBI to conventional white light endoscopy in patients with suspected laryngeal malignancies and compared the findings of each modality with biopsy results. They found sensitivities of 97% versus 79% for NBI and white light endoscopy, respectively. Accuracies were 97% versus 90%, respectively, and specificities were 96% and 95%, respectively. Ni and colleagues described a classification system for vascular patterns observed in the larynx using NBI. They designated 5 types (I–V) based on the vascular features of the intraepithelial papillary capillary loop, with types I–IV corresponding to nonmalignant lesions and type V being malignant. They found that this classification correlated well with the histologic examinations of the laryngeal lesions they studied. Bertino and colleagues applied NBI using the Ni classification to 248 patients with pharyngolaryngeal lesions and found sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV) of 97.4%, 84.6%, 92.7%, 91.6%, and 95.1%, respectively. Ninety-eight percent of malignant lesions by histologic examination corresponded to a type V NBI pattern, and 84.8% of benign lesions corresponded to a type I–IV pattern. NBI has also been applied more specifically to the follow-up of patients treated for laryngeal and hypopharyngeal carcinoma, conditions in which early detection of recurrent disease is often difficult. Zabrodsky and colleagues applied NBI via transnasal flexible videoendoscopy in an ambulatory setting to 66 patients previously treated for laryngeal or hypopharyngeal cancer with radiation or chemotherapy. Suspicious lesions identified by NBI were then biopsied, and they found an accuracy of 88%, sensitivity of 92%, specificity of 76%, PPV of 96%, and a NPV of 91%. The investigators also asserted that many of the lesions identified were not seen with white light endoscopy. In addition to aiding in the diagnosis and treatment of laryngeal malignancies, NBI has also been shown to be a useful adjunct in the management of recurrent respiratory papillomatosis.

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Functional Anatomy and Oncologic Barriers of the Larynx
2. Voice Outcomes of Transoral Laser Microsurgery of the Larynx
3. Quality of Life After Conservation Surgery for Laryngeal Cancer
4. Transcervical Conservation Laryngeal Surgery

Stay updated, free articles. Join our Telegram channel

Join