Development and Assessment of a Model for Predicting Individualized Outcomes in Patients With Oropharyngeal Cancer

Beesley, Lauren J.; Shuman, Andrew G.; Bellile, Emily L.; Rosen, Benjamin; Casper, Keith; Ibrahim, Mohannad; Dermody, Sarah M.; Wolf, Gregory T.; Chinn, Steven B.; Spector, Matthew E.; Jong, Robert J. Baatenburg de; Dronkers, Emilie A. C.; Taylor, Jeremy M. G.

doi:10.1001/jamanetworkopen.2021.20055

Cited by 10 publications

(17 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Current surveillance guidelines do not consider patient‐ and tumor‐specific risk factors, such as tumor subsite, stage, and HPV status, which greatly affect treatment response and patterns of failure 36 . Furthermore, most recurrences are reported by patient symptoms rather than identified by the provider, 37 and adherence to current surveillance guidelines fail to improve oncologic outcomes or quality of life 38,39 .…”

Section: Discussionmentioning

confidence: 99%

“…mOS for patients with oligometastatic disease was Current surveillance guidelines do not consider patient-and tumor-specific risk factors, such as tumor subsite, stage, and HPV status, which greatly affect treatment response and patterns of failure. 36 Furthermore, most recurrences are reported by patient symptoms rather than identified by the provider, 37 and adherence to current surveillance guidelines fail to improve oncologic outcomes or quality of life. 38,39 The development of evidence-based surveillance guidelines requires consideration of the clinical utility of early F I G U R E 2 Time to recurrence.…”

Section: Oligometastatic Diseasementioning

confidence: 99%

See 1 more Smart Citation

Patterns of recurrence in head and neck squamous cell carcinoma to inform personalized surveillance protocols

Haring

Kana

Dermody

et al. 2023

Cancer

Self Cite

View full text Add to dashboard Cite

BackgroundDevelopment of evidence‐based post‐treatment surveillance guidelines in recurrent/metastatic head and neck squamous cell carcinoma (R/M HNSCC) is limited by comprehensive documentation of patterns of recurrence and metastatic spread.MethodsA retrospective analysis of patients diagnosed with R/M HNSCC at a National Cancer Institute‐designated cancer center between 1998– 2019 was performed (n = 447). Univariate and multivariate analysis identified patterns of recurrence and predictors of survival.ResultsMedian overall survival (mOS) improved over time (6.7 months in 1998–2007 to 11.8 months in 2008–2019, p = .006). Predictors of worse mOS included human papillomavirus (HPV) negativity (hazard ratio [HR], 1.8; 95% confidence interval [CI], 1.2–2.6), high neutrophil/lymphocyte ratio (HR, 2.1 [1.4–3.0], disease‐free interval (DFI) ≤6 months (HR, 1.4 [1.02–2.0]), and poor performance status (Eastern Cooperative Oncology Group, ≥2; HR, 1.91.1–3.4). In this cohort, 50.6% of recurrences occurred within 6 months of treatment completion, 72.5% occurred within 1 year, and 88.6% occurred within 2 years. Metachronous distant metastases were more likely to occur in patients with HPV‐positive disease (odds ratio [OR], 2.3 [1.4–4.0]), DFI >6 months (OR, 2.4 [1.5–4.0]), and body mass index ≥30 (OR, 2.3 [1.1–4.8]). Oligometastatic disease treated with local ablative therapy was associated with improved survival over polymetastatic disease (HR, 0.36; 95% CI, 0.24–0.55).ConclusionThese data regarding patterns of distant metastasis in HNSCC support the clinical utility of early detection of recurrence. Patterns of recurrence in this population can be used to inform individualized surveillance programs as well as to risk‐stratify eligible patients for clinical trials.Plain Language Summary After treatment for head and neck cancer (HNC), patients are at risk of recurrence at prior sites of disease or at distant sites in the body. This study includes a large group of patients with recurrent or metastatic HNC and examines factors associated with survival outcomes and recurrence patterns. Patients with human papillomavirus (HPV)‐positive HNC have good survival outcomes, but if they recur, this may be in distant regions of the body and may occur later than HPV‐negative patients. These data argue for personalized follow‐up schedules for patients with HNC, perhaps incorporating imaging studies or novel blood tests.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Oligometastatic Diseasementioning

confidence: 99%

Patterns of recurrence in head and neck squamous cell carcinoma to inform personalized surveillance protocols

Haring

Kana

Dermody

et al. 2023

Cancer

Self Cite

View full text Add to dashboard Cite

show abstract

“…The specialized data file includes a clinically validated variable indicating the HPV etiology (HPV in situ hybridization, tissue polymerase chain reaction, in situ hybridization for E6/7 RNA, real‐time polymerase chain reaction for E6/7 RNA) of each HNC tumor within the following HNC sites: Lip, Hypopharynx, Nasopharynx, Oropharynx, Pharyngeal Tonsil, Pharynx Other, Palate Soft, Tongue Base, Gum and Other Mouth, and Other Oral Cavity 30 . Following established practice, missing HPV‐status was imputed with data on tumor grade, histology, site, behavior, and encrypted census tract‐level socioeconomic indicators via a logistic regression model through Multiple Imputation Chained Equations 31–38 . This specialized data includes cancer diagnoses for year 2010–2016.…”

Section: Methodsmentioning

confidence: 99%

Examining the effect of Medicaid expansion on early detection of head and neck cancer of the oral cavity and pharynx by HPV‐type and generosity of dental benefits

Semprini

2023

Cancer Reports

View full text Add to dashboard Cite

BackgroundOver a decade of evidence supports the claim that increased access to insurance through Medicaid expansions improves early detection of cancer. Yet, evidence linking Medicaid expansions to early detection of head and neck cancers (HNC) of the oral cavity and pharynx, specifically, may be limited by the lack of attention to Human Papillomavirus (HPV) etiology, generosity of dental coverage, and valid inference analyzing state cancer registry data.AimsThis study reexamined the effect of Medicaid expansion on early detection of HPV+/− HNC in states offering extensive dental benefits.Materials and MethodsSpecialized data from the Surveillance, Epidemiology, and End Results (SEER) program was analyzed to account for, previously unmeasurable, differential detection patterns of HNCs associated with HPV. Then, to identify the effect of increasing Medicaid eligibility on staging patterns in states offering extensive benefits amidst potentially non‐common trends between states, a “Triple Differences” design identifies the differential effect of Medicaid Expansion (with dental coverage) on HPV‐negative HNCs relative to the change in HPV‐positive HNCs. For valid inference analyzing a small number of state clusters (12) in cancer registry data, each regression model applies a Wild Cluster Bootstrap.ResultsExpanding Medicaid eligibility was found to be associated with a decrease in the proportion of distant‐stage diagnoses of HPV(−) HNCs, but only among states which increased Medicaid dental generosity at the time of Medicaid expansion.ConclusionsThese results suggest that adding extensive Medicaid dental benefits was the primary mechanism impacting HNC detection. This study highlights the potential positive spillover effects of policies which increase access to public dental coverage for low‐income adults, while also showing the limitation of access to dental services for improving early detection of HPV+ HNCs.

show abstract

“…Note that for transitions after recurrence, the clock is reset to zero at the time of recurrence. The full details of this MSM are provided in Beesley et al 20 but a brief summary-including descriptions of additional component models and order restrictions on some parameters-is presented in the Appendix (Section A.1).…”

Section: Bayesian Multistate Modelmentioning

confidence: 99%

“…Transitions from states 1 or 2 directly to state 5 correspond to other-cause death, while transitions through state 3, 4, or 6 to state 5 correspond to death after recurrence; these deaths correspond largely to cancer-specific death although death could be due to any cause. Figure is adapted from Beesley et al 20 missingness. Along with missing covariate values, unobserved outcomes (e.g., type of recurrence), and cure status are also imputed within each MCMC iteration.…”

Section: Bayesian Multistate Modelmentioning

confidence: 99%

Comparing Individualized Survival Predictions From Random Survival Forests and Multistate Models in the Presence of Missing Data: A Case Study of Patients With Oropharyngeal Cancer

et al. 2023

View full text Add to dashboard Cite

Background: In recent years, interest in prognostic calculators for predicting patient health outcomes has grown with the popularity of personalized medicine. These calculators, which can inform treatment decisions, employ many different methods, each of which has advantages and disadvantages. Methods: We present a comparison of a multistate model (MSM) and a random survival forest (RSF) through a case study of prognostic predictions for patients with oropharyngeal squamous cell carcinoma. The MSM is highly structured and takes into account some aspects of the clinical context and knowledge about oropharyngeal cancer, while the RSF can be thought of as a black-box non-parametric approach. Key in this comparison are the high rate of missing values within these data and the different approaches used by the MSM and RSF to handle missingness. Results: We compare the accuracy (discrimination and calibration) of survival probabilities predicted by both approaches and use simulation studies to better understand how predictive accuracy is influenced by the approach to (1) handling missing data and (2) modeling structural/disease progression information present in the data. We conclude that both approaches have similar predictive accuracy, with a slight advantage going to the MSM. Conclusions: Although the MSM shows slightly better predictive ability than the RSF, consideration of other differences are key when selecting the best approach for addressing a specific research question. These key differences include the methods’ ability to incorporate domain knowledge, and their ability to handle missing data as well as their interpretability, and ease of implementation. Ultimately, selecting the statistical method that has the most potential to aid in clinical decisions requires thoughtful consideration of the specific goals.

show abstract

Development and Assessment of a Model for Predicting Individualized Outcomes in Patients With Oropharyngeal Cancer

Cited by 10 publications

References 55 publications

Patterns of recurrence in head and neck squamous cell carcinoma to inform personalized surveillance protocols

Patterns of recurrence in head and neck squamous cell carcinoma to inform personalized surveillance protocols

Examining the effect of Medicaid expansion on early detection of head and neck cancer of the oral cavity and pharynx by HPV‐type and generosity of dental benefits

Comparing Individualized Survival Predictions From Random Survival Forests and Multistate Models in the Presence of Missing Data: A Case Study of Patients With Oropharyngeal Cancer

Contact Info

Product

Resources

About