.—Laryngeal malignancy, “voice box” cancer, is uncommon with 12,620 estimated new cases and 3770 deaths in the United States in 2021,1 and represents only 6.2% of all respiratory system malignancies. The most significant risk factors are alcohol and tobacco consumption. Almost all cases (98%) of laryngeal cancer arise in the squamous epithelium, and in this analysis more than 75% are of well-or-moderately differentiated histopathology (Grades I&II). Local stage cancer (SEER Historic Staging) was more common than regional and distant stages combined (55.3% vs 44.7%). Tumors may arise above, below or at the level of the vocal folds and are described as supraglottic (encompassing the epiglottis, false vocal cords, ventricles, aryepiglottic fold and arytenoids), the glottis (encompassing the true vocal cords and the anterior and posterior commissures), and the subglottic region. In the National Cancer Institute’s Surveillance, Epidemiology, End-Results (NCI-SEER) Data Research, 9 Registries, Nov 2019 Sub (1975-2017),2 laryngeal cancer occurred more commonly in men than in women, 80.7% vs 19.3%, respectively with a 4.2 to 1 ratio. Additionally, there are racial disparities with African Americans presenting at a younger age and having a higher incidence and mortality than Caucasians. In the 1975-2017 period, overall median patient age was 64.4 years with White Americans-64.8 years and Black Americans-61.5 years. Unfortunately, the 5-year relative survival rate has declined 4%, and excess death rate has risen 13% since 1975 with overall incidence declining.
As a consequence, observed median survival is approximately 6.5-years for the total study-period pinpointing the need for further specialty research. This study follows the World Health Organization International Classification of Diseases for Oncology-3rd Edition (ICD-O-3)3 topographical identification, coding, labeling and listing of 43,103 patient-cases accessible for analysis in the United States National Cancer Institute’s Surveillance, Epidemiology and End Results program (NCI SEER Research Data, 9 Registries, 1975-2017). These are located in 6 primary anatomical sites: C32.0-Glottis, C32.1-Supraglottis, C32.2-Subglottis, C32.3-Laryngeal cartilage, C32.8-Overlapping lesion of larynx, C32.9-Larynx, NOS.
.—To update short- and long-term mortality and survival indices, and identify changing risk patterns for laryngeal cancer patients in a retrospective US population-based analysis, 1975-2017, using prognostic data stratified by ICD-O-3 Primary Site, age, sex, race, stage, histologic grade, two cohort entry time-periods (1975-1996 to 1997-2017), and disease duration to 20-years.
.—SEER*Stat v8.3.94 software (built March 12, 2021) was used to access SEER Research Data, 9 Registries, Nov. 2019 submission (1975-2017). For displaying risk, general methods and standard double decrement life table methodologies for converting and displaying ICD-O-3 coded laryngeal cancer primary site annual data to aggregate average annual mortality and survival units in durational-intervals of 0-1, 1-2, 2-5, 0-5, 5-10, 10-15, and 15-20 years were employed. The reader is referred to the “Registrar Staging Assistant (SEER*RSA)” for local-regional-distant Extent of Disease (EOD) sources used in the development of staging descriptions, and Summary Stage 2018 Coding Manual v2.0 released September 1, 2020. Cancer staging & grading procedural explanations, statistical significance and 95% confidence levels5 are described in previous Journal of Insurance Medicine articles6,7 and other publications.8,9 Poisson confidence intervals at the 95% level based on the number of observed deaths are used in this study but not displayed here to conserve space on the mortality tables. Excluded were all death certificate only and those alive with no survival time.
.—Total SEER annual age-adjusted incidence rates from 1980 to 2017 have diminished from 5.25 patient-cases/100,000/year to 2.59/100,000 per year, and in the same period annual age-adjusted US death rates declined from 1.61 deaths/100.000/year to 0.91 deaths/100,000/year (Ref. 10, CSR Tables 12.5-6), However, in the 0-5-year disease durational interval for all staged cases in both cohort time-periods (Table 5), excess death rates (EDR) rose from 80 per 1000 persons per year in the 1975-96 cohort, to 89 per 1000 persons per year in the 1997-17 cohort, (a 10% rise in excess mortality in 42 years). Further, in the 5-10-year disease durational interval, EDR rose from 39 per 1000 persons per year to 45 per 1000 persons per year with corresponding cohort declines in cumulative survival ratios (SR), and overall declines in median observed and relative survival times in the later cohort (not shown). The epidemiologic burden of malignancy is >4-fold higher in males and increases in parallel with aging, peaking after 65 years. The most significant risk factors for laryngeal cancer are tobacco and alcohol consumption.
.—Although annual incidence and mortality rates from 1980 to 2017 have diminished, there is no concomitant improvement in larynx cancer survival (SR) and mortality (EDR) indices, with rising mortality and diminishing survival in all staged cases at 5-years disease duration between the 1975-96 and 1997-2017 analytic cohorts. Larynx cancer remains a burdensome clinical, social, and public health challenge.
INCIDENCE, TRENDS AND PREVALENCE
From 1980 to 2017, overall larynx cancer annual age-adjusted US incidence rates diminished from 5.25 cases per 100,000 per year to 2.59 cases per 100,000 per year (a 49% decrease), and associated with diminishing incidence trends (average annual percent change [AAPC] -2.4} in the same time period. Parenthetically, age-adjusted US death rates in the same period simultaneously dropped from 1.61 deaths per 100,000 per year to 0.91 per 100,000 per year, a 43% decline.10
Chart 1. Incidence Rates: Age-Adjusted Rates by Race, Ethnicity & Sex (CSR Table 12.7). Incidence rates of invasive cancer of the larynx in the United States, 2013-2017, SEER 21 areas, was 2.9 cases per 100,000 per year, all races, both sexes combined. Incidence rates increase with age and vary by sex and race in the US Incidence is higher in males than in females, higher in blacks than in whites and higher with increasing age.
Chart 2. Incidence Trends: (CSR Table 12.15). Total average annual percent change in invasive cancer of the larynx, 2008-2017, SEER 21 areas, was -3.0 per 100,000 per year in all races. The trends in males, females and Hispanics were -3.1, -2.2 and -2.8, respectively.
Prevalence: the estimated complete number of United States invasive larynx cancer prevalence counts on January 1, 2017, by race, ethnicity, sex and years since diagnosis totaled 96,231 cases (CSR Table 12.21).
SEER CASE STATISTICS
Chart 3 displays overall demographic data for 43,103 patients with cancer of the larynx present in the SEER 9 Registries, 1975-2017, distributed by age, sex, race, stage, and grade. Chart 4 indicates the frequency and mean age (FAMA) statistics for each laryngeal cancer ICD-O-3 topographic subsite. Glottic and supraglottic subsites contained the most patient cases, 37,226 (86.4%) and the remainder of larynx ICD-O-3 primary locations, 5877 cases (13.6%). After exclusion of cases with no follow-up data, 35,507 (82.4%) patients remained for survival and mortality analysis. The ethnic distribution of patients with larynx cancer in the SEER frequency database was: White-82.7%, Black-13.4%, Other-3.7% and Unknown-0.25%. More than 80% of cases, both sexes combined were diagnosed after age 55 years. Mean ages for males and females were 64.5 years and 63.5 years, respectively, Whites 64.8, Blacks 61.5, Other 66.1 and Unknown 60.9 years. The SEER historic stage A selection code is used for staging (local, regional or distant). Overall, there were 27,922 staged cases and 1782 (6.4%) unstaged cases. Local cases comprised 55.3%, regional-37.7%, distant-7.0%, and regional-distant staging-44.7%. Parenthetically, but not included here were 2295 in-situ cases. A total of 32,615 cases, approximately 76%, were graded and 10,415 (24%) were of unknown grade. Well and moderately differentiated grade larynx cancers (Grades I&II) accounted for 76% of graded malignancies and poorly differentiated and undifferentiated histopathology (Grades III&IV) accounted for 24% of graded cancers.
Comparative male and female larynx cancer diagnostic frequency with advancing age is illustrated in the Figure. The zenith of male diagnostic frequency occurs at quinquennial ages 60-64 years and then declines; male mean age at diagnosis, 64.5-years. Female diagnostic frequency crests at ages 65-69 years and then declines; female mean age at diagnosis was 63.5 years.
FOLLOW-UP (FU)
Standard procedures were used by the 9 SEER registries in the follow-up (FU) of patients and in the confirmation of death data. Losses to FU were very low, less than 0.5% of all entrants at 5 years. Losses were higher in Hispanics than in black and white patients.
RESULTS
Table 1 shows comparative mortality and survival for the local stage of cancer of the larynx in the 1975-96 cohort. Results are given to duration 20 years in 4 age groups, male and female and all histopathologic grades combined, and by sex for all ages combined. EDR values (extra deaths per 1000 per year in all groups containing males (except for ages 75 up) are higher in the second than in the first year of follow-up, then tend to decrease. In the total female group EDR is at a maximum in the first year (52.4 per 1000), but the usual trend for EDR to decrease progressively with duration is not evident: all EDR values after 2 years are above the minimum of 35 per 1000 at duration 2-5 years. In males EDR also shows less evidence of the usual trend for excess mortality as EDR to decrease with duration: the minimum EDR occurs at duration 10-15 years and is still above 25 per 1000 per year. There is little sex difference in EDR. By age group, EDRs are higher over age 65, but Mortality Ratios (MRs) show the usual decrease with advancing age from a maximum of 745% (age <55, duration 1-2) to a minimum of 120 (age 75 up, duration 10-15). The matching pattern of Survival Ratios (SRs) shows a range of 85% to 78% at 5 years (depending on age), and a further decrease to 34% at 20 years.
Table 2 shows for the 1975-96 cohort, comparative mortality and survival in advanced cancer of the larynx, male and female combined. EDR increases with age, in the regional stage from an initial 129 per 1000 at ages under 55 to 262 per 1000 at ages 75 and up. Initial EDR is even higher in the distant stage (only 7% of all staged cases): 311 per 1000 under 65 and 440 at ages 65 up. With these very high levels of initial excess mortality in advanced cancer of the larynx, EDR does tend to decrease with duration, but remains very high at durations over 10 years: of sixteen 15-year regional survivors at 15 years, 15 had died prior to 20 years. The maximum MR of 2,414% was found in the second year of patients under 55 in the regional stage. MR decreased both with advancing age and duration. SR also decreased with advancing age and duration: at 5 years in the regional stage SR ranged from 58% to 41%, and in the distant stage was 34% to 24%. In the regional stage at 20 years, the range was from 33% to less than 5%.
In the 1997-2017 cohort results are shown similar tabular design for local cancer in Table 3 and advanced cancer in Table 4. Contrary to cohort comparisons in most other cancer sites, the EDRs in matching stage/age groups do not show a consistently lower excess mortality in the 1997-2017 cohort. The patterns of EDR, MR and SR described in the 1975-96 cohort are generally observed in the later cohort. Excess mortality remains relatively high in localized cancer of the larynx, and very high in the advanced stages.
Table 5 gives the usual aggregate averaged comparisons for duration 0-5 and 5-10 years, by combinations of cohort, stage, stage, grade and race, all ages and both sexes combined. It is evident that grade 3&4 malignancy of the tumor cells is associated with a significantly higher EDR and MR than in cases graded 1&2. In both local and regional stages EDR is slightly higher in 1997-2017 than in 1975-96, in 2 of 4 comparisons at duration 0-5 years. The numbers of deaths in the comparison cells are large, ranging from 140 to 2028. At 0-5 years, all staged cases combined, the EDR was 80 per 1000 per year in 1975-96 and 89 in 1997-2917. EDR tends to be somewhat higher in nonwhite than in white patients. Also in this duration interval, both cohorts combined, EDR was 78 per 1000 per year in patients of the white race, and higher, 110 per 1000 per year in all patients who were nonwhite. By race the 5-year SR values were 65% for whites and 56% for nonwhites. In all staged cases, 5-year relative survival (SR) was higher in the 1975-96 cohort (67%) than in the later 1997-2017 cohort (64%).
Of the six ICD-O-3 topographical laryngeal cancer subsites shown in Table 6, cancer of the glottis had the “better” prognosis, with an EDR 35 per 1000, 30% deaths at 5 years duration, and SR of 80%. Cancer of the supraglottis had the worst prognosis with an EDR of 103 per 1000 per year, 55% of deaths at 5 years, and SR of 48%. Glottic and supraglottic tumors constituted 52% and 34% of all laryngeal malignancies respectively. Subglottic and laryngeal cartilage cancers displayed intermediate mortality and survival results at 0-5 years duration but constituted only 2.3% of all laryngeal cancer subsites. For all laryngeal cancer subsites combined (C32.0-9), 1975-2017, EDR was 86 per 1000 per person at 5 years duration, Observed and relative survival (P & SR) 56% and 65% respectively. Median observed and relative survival approximately 6.5 years and 11 years, respectively. Forty-two percent of deaths occurred in the first 5 years, and at 20 years duration EDR was above 40 per 1000 per person per year.
COMMENTARY
The majority of larynx cancer burden was observed in men, especially among males aged 40-74 years. Almost 98% of laryngeal cancer cases are of epidermoid origin, so no effort has been made to analyze mortality by histologic type. Fortunately, from 1980 to 2017, annual laryngeal cancer age-adjusted US. incidence and death rates fell synchronously approximately 49% and 43% respectively. However, comparing the 1975-1996 and 1997-2017 excess death rates at 0-5 year’s duration for all staged cases, EDR rose to 93 per 1000 individuals, a 13% increase with an associated reduction in cumulative survival (SR) from 66% to 63% indicating a challenging need to improve diagnostic and treatment effectiveness. Nevertheless, clinical and demographic factors of improved prognosis and survival in patients with laryngeal cancer include: Stage; local vs advanced stage at diagnosis; Histologic grade; well differentiated vs poorly differentiated tumors. Additionally, among patients with the most advanced disease (SEER historic stage-distant, AJCC stage IV), total laryngectomy (TL) was associated with increased survival compared with chemo-radiotherapy (RT), whereas both TL and chemo-RT improved survival over RT among patients with stage III cancer. Insurance type and black race also showed significant associations with survival, which may reflect barriers in access to care.11
Almost all cancers of the larynx arise in the glottis (vocal cord area) or supraglottis, more than half of laryngeal cancers are localized. When invasive spread occurs, it is generally to the lymph nodes, or by direct extension to soft tissues adjacent to the larynx; distant metastasis is rare (7% in this SEER database). As shown in Table 5, EDR averaged over 30 per 1000 persons per year at duration 0-5 years in the best local cases (grade 1&2), more than twice this rate in grade 3&4, and 112 to 181 per 1000 persons per year in regional cancer. There is no survival or mortality improvement from the 1975-1996 to the 1997-2017 cohort. These data are for all ages and both sexes combined. SR values at 5 years range from 85% to 68% in the local stage, and 62% to 42% in the regional stage. Excess mortality is at a much lower level in cancer of the larynx than it is in cancer of the trachea and lung.
Richard Bunker Singer, MD, March 22, 1914–February 19, 2010. Consummate gentleman, dear friend, scholar, and esteemed colleague.