Common Pitfalls of Head and Neck Research Using Cancer Registries

Patterns of Care for T1 Glottic Squamous Cell Carcinomas from 2004-2020

OBJECTIVE(S)

Describe recent national trends in overall treatment modalities for T1 glottic squamous cell carcinomas (SCC), and identify factors associated with treatment regimens.

METHODS

National Cancer Database from 2004-2020 was queried for all patients with glottic cT1N0M0 SCC. Treatment patterns over time were analyzed using the Cochran-Armitage test for trend. Multivariable logistic regressions were used to determine the factors associated with treatment regimens.

RESULTS

Of the 22,414 patients identified, most patients received RT only (57%), 21% received surgery only, and 22% received dual-modality treatment ("over-treatment"). Over the time period, there was a decreasing trend in rates of over-treatment for T1 glottic SCC (p < 0.001) and an increasing trend in surgery only (p < 0.001). Treatment in 2016-2018 (OR: 1.168 [1.004 to 1.359]), 2013-2015 (OR: 1.419 [1.221 to 1.648]), 2010-2012 (OR: 1.611 [1.388 to 1.871]), 2007-2009 (OR: 1.682 [1.450 to 1.951]), or 2004-2006 (OR: 1.795 [1.548 to 2.081]) versus 2019-2020 was associated with greater likelihood of over-treatment. T1b tumors were less likely to be over-treated (OR: 0.795 [0.707 to 0.894]) versus T1a tumors, and less likely to receive surgery first (OR: 0.536 [0.485 to 0.592]) versus T1a tumors.

CONCLUSION

Over-treatment for T1 glottic SCC has been declining, with increasing rates of surgery only. Year of treatment was significantly associated with the receipt of dual-modality treatment. Finally, patients with T1b disease were more likely to receive RT as the first and only treatment.

LEVEL OF EVIDENCE

3 Laryngoscope, 134:3633-3644, 2024.

Epidemiology and management of oral cavity squamous cell carcinoma in Ireland

INTRODUCTION

The epidemiology and management of oral cavity cancer have changed considerably in recent decades. This study examines epidemiological and management trends in oral cavity squamous cell carcinoma (OCSCC).

METHODS

A retrospective cohort study of data from the National Cancer Registry of Ireland between 1994 and 2014.

RESULTS

A total of 2725 patients were identified. The most common subsites were the tongue (34 %, n = 1025), lip (19 %, n = 575), floor of mouth (FOM) (18 %, n = 550), and retromolar trigone (RMT) (6 %, n = 189). The incidence of OCSCC remained largely unchanged (3.14 cases/100000/year) during the study period. 5-year disease-specific survival (DSS) was 58.6 % overall, varying between subsites (lip 85 %, RMT 62.9 %, tongue 54.7 %, and FOM 47.3 %). DSS improved over the study period (p = 0.03), in particular for tongue primaries (p = 0.007). Primary surgery significantly improved DSS versus radiotherapy (HR 0.28, p < 0.0001). Survival of T4 disease managed surgically was superior to that of T1 disease managed with radiotherapy. In node positive patients, chemotherapy improved overall survival (HR 0.8 p = 0.038) but not DSS (HR 0.87 p = 0.215).

CONCLUSION

Primary surgery remains the standard of care in the management of OCSCC. Prognosis has improved in line with an increase in the use of primary surgery in the same time frame, though the incidence remains unchanged.

Prognostic significance of head and neck spindle cell carcinoma

BACKGROUND

Our study investigates the prognostic significance of spindle cell histology on overall survival (OS) of conventional head and neck squamous cell carcinoma (HNSCC).

METHODS

The 2004 to 2017 National Cancer Database was queried for patients with head and neck spindle cell carcinoma (HNSpCC) (n = 1572) or HNSCC (n = 242 697) of the oral cavity, major salivary glands, sinonasal tract, oropharynx, hypopharynx, and larynx treated with curative intent.

RESULTS

Patients with HNSpCC presented more frequently with higher-grade tumors and cN0 disease than those with HNSCC (p < 0.001). In the oral cavity, the HR for death for SpCC compared with SCC was 1.33 (p < 0.001). In the oropharynx, the HR for death for SpCC compared with SCC was 1.47 (p = 0.028).

CONCLUSIONS

After adjusting for patient, tumor, and treatment characteristics, SpCC histology had an independent adverse prognostic effect on OS in the oral cavity and oropharynx. SpCC histology does not necessarily portend poorer survival in all HNSCC.

LEVEL OF EVIDENCE: 4

Pediatric Small Renal Masses: Can Tumor Size Predict Histology and the Potential for Nephron-sparing Surgery?

PURPOSE

The majority of children with unilateral renal masses suspicious for malignancy undergo radical nephrectomy, while nephron-sparing surgery is reserved for select cases. We investigated the impact of tumor size on the probability of histology. We hypothesized that pediatric small renal masses are more likely benign or non-Wilms tumor, thus potentially appropriate for nephron-sparing surgery.

MATERIALS AND METHODS

The SEER (Surveillance, Epidemiology, and End Results) database was analyzed for patients aged 0-18 years diagnosed with a unilateral renal mass from 2000-2016. Statistical analysis was performed to help determine a tumor size cut point to predict Wilms tumor and assess the predictive value of tumor size on Wilms tumor histology. Additionally, a retrospective review was performed of patients 0-18 years old who underwent surgery for a unilateral renal mass at a single institution from 2005-2019. Statistical analysis was performed to assess the predictive value of tumor size on final histology.

RESULTS

From the SEER analysis, 2,016 patients were included. A total of 1,672 tumors (82.9%) were Wilms tumor. Analysis revealed 4 cm to be a suitable cut point to distinguish non-Wilms tumor. Tumors ≥4 cm were more likely Wilms tumor (OR 2.67, P ≤ .001), but this was driven by the statistical significance in children 5-9 years old. From the institutional analysis, 134 patients were included. Ninety-seven tumors (72.3%) were Wilms tumor. Tumors ≥4 cm had higher odds of being Wilms tumor (OR 30.85, P = .001), malignant (OR 6.75, P = .005), and having radical nephrectomy-appropriate histology (OR 46.79, P < .001).

CONCLUSIONS

The probability that a pediatric unilateral renal mass is Wilms tumor increases with tumor size. Four centimeters is a logical cut point to start the conversation around defining pediatric small renal masses and may help predict nephron-sparing surgery-appropriate histology.

Evidence-Based Medicine in Otolaryngology Part 14: Falsehood and Bias

OBJECTIVE

Falsehood and bias can have tangible effects, whether related to the "hoax" of Corona virus disease/COVID-19 or the impact of personal protective equipment in city-wide news. The spread of false information requires the diversion of time and resources into rebolstering the truth. Our objective is thus to elucidate types of bias that may influence our daily work, along with ways to mitigate them.

DATA SOURCES

Publications are included which delineate specific aspects of bias or address how to preempt, mitigate, or correct bias, whether conscious or unconscious.

REVIEW METHODS

We discuss: (1) the background and rationale for proactively considering potential sources of bias, (2) relevant definitions and concepts, (3) potential means to limit effects of inaccurate data sources, and (4) evolving frontiers in the management of bias. In doing so, we review epidemiological concepts and susceptibility to bias within study designs, including database studies, observational studies, randomized controlled trials (RCTs), systematic reviews, and meta-analyses. We additionally discuss concepts such as the difference between disinformation and misinformation, differential or nondifferential misclassification, bias toward a null result, and unconscious bias, among others.

CONCLUSION

We have the means to mitigate sources of potential bias in database studies, observational studies, RCTs, and systematic reviews, beginning with education and awareness.

IMPLICATIONS FOR PRACTICE

False information may spread faster than true information, so it is beneficial to understand potential sources of falsehood we face, in order to safeguard our daily impressions and decisions. Awareness of potential sources of falsehood and bias forms the foundation for accuracy in our everyday work.

Survival differences between women and men in the non-reproductive cancers: Results from a matched analysis of the surveillance, epidemiology, and end-results program

Background

Men with non-reproductive cancers have a discrepant outcome compared to women. However, they differ significantly in the incidence of cancer type and characteristics.

Methods

Patients with single primary cancer who were 18 years or older and whose data were gathered and made accessible by the Surveillance, Epidemiology, and End Results (SEER) program were included in this retrospective analysis. Kaplan-Meier curves and Cox regression before and after propensity score matching were performed to analyze the risk survival by sex.

Results

Among the 1,274,118 patients included [median (range) age, 65 year (18-85+) years; 688,481 (54.9%) male]. The median follow-up was 21 months (0-191). Substantial improvements in survival were observed for both sexes during the years of inclusion analyzed, with no difference between them, reaching a reduction of almost 17% of deaths in 2010, and of almost 28% in 2015, compared to 2004. The women had a median survival of 74 months and overall mortality of 48.7%. Males had a median survival of 30 months (29.67-30.33) with an overall mortality of 56.2%. The PSM showed a reduced difference (6 months shorter median survival and 2.3% more death in men), but no change in hazards was observed compared to the unmatched analysis [adjusted HR: 0.888 (0.864-0.912) vs. 0.876 (0.866-0.886) in unmatched].

Conclusions

The discrepancy in survival between men and women is not explained only by the incidence of more aggressive and more advanced cancers in the former.

Establishment of prognostic nomogram for T1N0M0 glottic squamous cell carcinoma: an SEER database analysis

OBJECTIVES

The study aimed to construct prognostic models for OS and CSS in patients with T1N0M0 glottic SCC. In addition, we used PSM to re-assess the effect of surgery alone and radiation alone.

METHODS

The Surveillance, Epidemiology, and End Results database was searched for patients with confirmed T1N0M0 glottic SCC. Patients with complete data were randomly divided into the training and the validation cohort (7:3), Cox-regression analysis was performed to identified significant predictors of OS and CSS. PSM was used to mimic randomized controlled the trials. Kaplan-Meier survival method and log-rank tests were utilized for survival analysis.

RESULTS

A total 1827 patients met the inclusion criteria. Survival analysis indicated that the patients who underwent the primary site surgery had a better OS (P = 0.002) and CSS (P = 0.008), compared with non-surgery patients. Cox-regression analysis proved that age, marital status, T1 stages, surgery, radiation, sequential treatments, and chemotherapy had significant effects on OS. While age, marital status, histologic grade, surgery, radiation, sequential treatments, and chemotherapy were substantially associated with CSS. Patients who received primary site surgery had a better OS and CSS, compared with non-surgical patients. Patients receiving radiation had a better CSS than non-radiation patients. However, patients who received sequential treatments or chemotherapy had a worse OS and CSS, compared with controlled groups. Predictive nomogram models were established to predict patients' prognosis with good consistency between the actual observation and the nomogram prediction. Before PSM, patients who underwent surgery alone had a better OS and CSS than those who received radiation alone. After PSM, patients receiving surgery still had a better OS than those receiving radiation. However, there were no statistically significant differences in CSS.

CONCLUSIONS

Nomogram models were developed to predict OS and CSS in patients with T1N0M0 glottic SCC. Primary site surgery could definitely increase OS and CSS, while radiation could significantly increase CSS. Using PSM, surgery alone could significantly enhance OS, as compared to radiation alone. Chemotherapy should not be recommended for early glottic carcinoma.

The current state of glioma data registries

Background

The landscape of glioma research has evolved in the past 20 years to include numerous large, multi-institutional, database efforts compiling either clinical data on glioma patients, molecular data on glioma specimens, or a combination of both. While these strategies can provide a wealth of information for glioma research, obtaining information regarding data availability and access specifications can be challenging.

Methods

We reviewed the literature for ongoing clinical, molecular, and combined database efforts related to glioma research to provide researchers with a curated overview of the current state of glioma database resources.

Results

We identified and reviewed a total of 20 databases with data collection spanning from 1975 to 2022. Surveyed databases included both low- and high-grade gliomas, and data elements included over 100 clinical variables and 12 molecular data types. Select database strengths included large sample sizes and a wide variety of variables available, while limitations of some databases included complex data access requirements and a lack of glioma-specific variables.

Conclusions

This review highlights current databases and registries and their potential utility in clinical and genomic glioma research. While many high-quality resources exist, the fluid nature of glioma taxonomy makes it difficult to isolate a large cohort of patients with a pathologically confirmed diagnosis. Large, well-defined, and publicly available glioma datasets have the potential to expand the reach of glioma research and drive the field forward.

Geographic Access to Radiation Therapy Facilities in the United States

BACKGROUND

The current distribution of radiation therapy (RT) facilities in the US is not well established. A comprehensive inventory of US RT facilities was last assessed in 2005, based on data from state regulatory agencies and dosimetric quality assurance bodies. We updated this database to characterize population-level measures of geographic access to RT and analyze changes over the past 15 years.

METHODS

We compiled data from regulatory and accrediting organizations to identify US facilities with linear accelerators used to treat humans in 2018-2020. Addresses were geocoded and analyzed with Geographic Information Services (GIS) software. Geographic access was characterized by assessing the Euclidian distance between zip code tabulation areas (ZCTA)/county centroids and RT facilities. Populations were assigned to each county to estimate the impact of facility changes at the population level. Logistic regressions were performed to identify features associated with increased distance to RT and associated with regions that gained an RT facility between the two time points studied.

RESULTS

In 2020, a total of 2,313 US RT facilities were reported compared to 1,987 in 2005, representing a 16.4% growth in facilities over nearly 15 years. Based on population attribution to ZCTA centroids, 77.9% of the US population lives within 12.5 miles of an RT facility, and 1.8% of the US population lives more than 50 miles from an RT facility. We found that increased distance to RT was associated with non-metro status, less insurance, older median age, and less populated regions. Between 2005 and 2020, the population living within 12.5 miles from an RT facility increased by 2.1 percentage points, while the population living furthest from RT facilities decreased 0.6 percentage points. Regions with improved geographic RT access are more likely to be higher income and better insured.

CONCLUSION

1.8% of the US population has limited geographic access to radiation therapy. We found that people benefiting from improved access to RT facilities are more economically advantaged, suggesting disparities in geographic access may not improve without intervention.

Internal Staging Discordance in National Cancer Databases for Non-Small-Cell Lung Cancer

BACKGROUND

The Surveillance, Epidemiology and End Results (SEER) and the National Cancer Database (NCDB) are databases for cancer analysis which may be subject to error in data reporting. We examined rates and impact of discordant data for non-small cell lung cancer.

METHODS

NCDB and SEER were queried for non-small cell lung cancer pathologic Tumor, Node, Metastasis data (NCDB) or "derived" data (SEER). Discordancy between descriptors with stage and impact of outlier data were analyzed.

RESULTS

Incomplete staging was noted in 71.5% of NCDB and 10.3% of SEER. 174,829 patients from NCDB and 117,114 from SEER were analyzed. NCDB had 97 cases with ≥20 positive lymph nodes recorded vs. 27 in SEER (p<0.001). Mean and median sampled lymph nodes were skewed with inclusion of these data-points (p<0.001). NCDB misclassified 0.99% tumors >5cm as stage I vs. 0.04% in SEER (p<0.001). NCDB mis-staged positive lymph nodes as pathologic N0 (0.59%) or Stage 0/Stage I (0.65%). NCDB misclassified pathologic N1 as < Stage II (0.91%) or N2 as < Stage III (0.36%). NCDB misclassified Stage I with documentation of pathologic N1-N3 disease (0.24%) or Stage II with evidence of N2 or N3 disease (0.50%). NCDB misclassified pathologic M1 as pathologic Stage <IV in 0.9% of cases and misclassified 19.8% of stage IV as pathologic M0. SEER collaborative staging had no discordancy (p<0.001).

CONCLUSIONS

NCDB and SEER are two powerful cancer databases. However, cumulative discordancy rate was 4.9% for NCDB and 0.008% for SEER with more mistaging and outliers in NCDB.

Evaluating Adjuvant Therapy With Chemoradiation vs Radiation Alone for Patients With HPV-Negative N2a Head and Neck Cancer

Importance

The American Joint Committee on Cancer staging system (Cancer Staging Manual, 8th Edition) for head and neck squamous cell carcinoma (HNSCC) now categorizes human papillomavirus (HPV)-negative HNSCC in a single positive lymph node smaller than 3 cm with pathologic extranodal extension (ENE) as N2a. The standard of care for pathologic ENE is adjuvant chemoradiation therapy (CRT). Whether adding chemotherapy concurrent with adjuvant radiation therapy improves survival in this clinical scenario is unknown.

Objective

To assess whether adjuvant CRT relative to radiation therapy alone is associated with improved survival among patients with pN2a HPV-negative HNSCC with ENE.

Design, Setting, and Participants

This retrospective cohort study included 504 patients with pN2a HPV-negative HNSCC with ENE who had undergone margin-negative surgery and adjuvant therapy. The patients were identified from the National Cancer Database from January 1, 2004, to December 31, 2015. Statistical analyses were conducted from September 1, 2019, to April 16, 2020.

Main Outcomes and Measures

The primary end point was overall survival. The association of adjuvant CRT with overall survival was analyzed using univariate and multivariate Cox proportional hazards regression analyses. Planned subset analyses were conducted in patients younger than 70 years with no comorbidities (the subset most likely to be eligible for a clinical trial of cisplatin-based chemoradiation) and in patients with pT3/T4 disease classification.

Results

Of 504 patients (mean [SD] age, 60.5 [12.7] years; 319 [63.3%] men; 434 [86.1%] White) with pN2a HPV-negative HNSCC with ENE who had undergone margin-negative surgery and adjuvant therapy, 298 patients (59.1%) received adjuvant CRT. For the overall cohort of patients with pN2a ENE, adjuvant CRT was not associated with improved overall survival relative to adjuvant radiation therapy alone in a multivariate analysis (adjusted hazard ratio, 0.98; 95% CI, 0.74-1.30). Adjuvant CRT was still not associated with improved overall survival in a subset analysis of 304 patients younger than 70 years with no comorbidities (adjusted hazard ratio, 0.98; 95% CI, 0.66-1.45) nor in a subset of 220 patients with pT3/T4 disease classification (adjusted hazard ratio, 1.03; 95% CI, 0.70-1.54).

Conclusions and Relevance

This study found that for patients with pN2a HPV-negative HNSCC with ENE who underwent margin-negative surgery and adjuvant therapy, adding chemotherapy concurrent with adjuvant radiation therapy was not associated with improved overall survival. Additional research is necessary to identify the optimal treatment paradigm for this clinical scenario.

Impact of Tumor Site and Adjuvant Radiotherapy on Survival of Patients with Adenoid Cystic Carcinoma: A SEER Database Analysis

Adenoid cystic carcinoma (ACC) is a rare salivary gland tumor, displaying aggressive behavior with frequent recurrence and metastasis. Little information exists regarding the impact of clinicopathological parameters and adjuvant radiotherapy (aRT) on ACC disease specific (DSS) and overall survival (OS). We extracted demographic, treatment, and survival information of 1439 patients with major or minor intraoral salivary gland ACC from the Surveillance, Epidemiology, and End Results (SEER) database. The associations between tumor characteristics and aRT with OS and DSS were estimated using hazard ratios (HR) and 95% confidence intervals (CI). Submandibular gland ACCs had the worst prognosis (adjusted DSS HR = 1.48; 95% CI = 0.99-2.20, compared to parotid), and this difference was more pronounced among patients with advanced-stage tumors (adjusted DSS HR = 1.93; 95% CI = 1.13-3.30). aRT was associated with increased overall survival only among stage III submandibular ACC patients (HR = 0.64; 95% CI = 0.42-0.98) and had no benefit in any other group. In conclusion, submandibular gland ACC carries a worse prognosis than other gland subsites and may benefit from aRT. The different outcomes between submandibular gland and other major or minor gland ACCs warrant further mechanistic investigation.

Prognostic risk factors for initially diagnosed T2N0M0 glottic cancer: competing risk analysis and propensity-score matched cohort analysis

BACKGROUND

Prognostic risk factors of patients with initially diagnosed T2N0M0 glottic cancer remain unclear. This study was aimed to conduct a comprehensive analysis to identify valuable prognostic risk factors for initially diagnosed T2N0M0 glottic cancer.

METHODS

Data of patients with initially diagnosed T2N0M0 glottic cancer were extracted from the Surveillance, Epidemiology, and End Results database. Survival analyses and Cox regression analyses were conducted to evaluate overall survival (OS) and cancer-specific survival (CSS). In consideration of competing events, the competing risk (CR) analysis was applied. Furthermore, propensity-score matching (PSM) was applied to mimic randomized-controlled trials and reduce selection bias.

RESULTS

A total of 923 eligible patients met the inclusion criteria. Survival analyses showed that age, marital status, primary site surgery, and radiation were independent predictors of OS. Besides, age, marital status, primary site surgery, radiation, and chemotherapy were independent predictors of CSS. Cox regression analyses and the CR analysis were basically consistent with this result. In addition, an internal validation and PSM were performed to explore the role of chemotherapy.

CONCLUSION

We conducted a comprehensive analysis to prove that age, marital status, primary site surgery, radiation, and chemotherapy may be valuable prognostic risk factors for initially diagnosed T2N0M0 glottic cancer. Primary site surgery and radiation should be recommended, whereas chemotherapy was likely not suitable so far. Furthermore, we constructed a CR nomogram to predict survival rates.

Prognostic factors of patients with initially diagnosed T1a glottic cancer: Novel nomograms and a propensity-score matched cohort analysis

The option of T1a glottic cancer treatments remarkably varied in different countries. This study aimed to construct predictive models to predict overall survival (OS) and cancer-specific survival (CSS) of patients with initially diagnosed T1a glottic cancer. And we used propensity score matching (PSM) to reassess the effect of treatments.Data of patients with initially diagnosed T1a glottic cancer were extracted from the Surveillance, Epidemiology, and End Results database. Patients with complete information were randomly divided into the training and the validation cohorts (7:3). Cox regression was conducted to screen significant predictors of the OS and the CSS. PSM was performed to mimic randomized controlled trials. Survival analyses were performed by Kaplan-Meier survival methods, and log-rank tests were utilized.A total of 2342 patients met the inclusion criteria. Survival analyses showed that patients who underwent primary site surgery would have better OS and CSS. Univariate analyses and multivariate analyses proved that stage, N stage, primary site surgery, and chemotherapy significantly affected both the OS and the CSS. Predictive nomograms were established to predict patients' prognosis. Finally, the OS and the CSS for patients who underwent primary site surgery alone were significantly longer than those who underwent radiation alone before and after PSM.We constructed nomograms predicting the OS and the CSS of patients with initially diagnosed T1a glottic cancer. Compared to our previous studies, this study indicated that primary site surgery may be superior to radiation and chemotherapy. At present, chemotherapy should be not recommended for T1a glottic cancer patients.

The primary tumor resection in patients with distant metastatic laryngeal carcinoma

BACKGROUND

The role of primary tumor resection in patients with distant metastatic laryngeal carcinoma (DMLC) has not been clarified completely. Thus, we used propensity score matching (PSM) and survival analysis to address this issue.

METHODS

The PSM was utilized to avoid selection bias and disproportionate distributions of the confounding factors. Kaplan-Meier estimates and Cox proportional hazard analysis were utilized to evaluate overall survival (OS) and cancer-specific survival (CSS).

RESULTS

From the Surveillance, Epidemiology, and End Results Program database, a cohort of 480 patients with DMLC were included. After PSM, the OS and CSS for patients who underwent resection were significantly longer than those without resection (median OS: 19 months vs. 8 months, P < 0.001; median CSS: 19 months vs. 9 months, P = 0.002). Tumor resection significantly prolonged survival of DMLC patients with appropriate demographic and clinical characteristics. In the multivariate analysis, age at diagnosis, race, pathologic subtype, and marital status were found significantly affecting both OS and CSS of patients who underwent surgical resection. Predictive nomograms were developed to help distinguish patients with early mortality potential after surgical resection.

CONCLUSIONS

This study is the first one using PSM to assess the role played by surgical resection in DMLC and evaluate the prognostic factor of resected patients. Premised on well controlled postoperative complications, resection could significantly prolong OS and CSS of certain patients.