Author ORCID Identifier

https://orcid.org/0000-0002-6642-1073

Semester

Spring

Date of Graduation

2024

Document Type

Dissertation

Degree Type

PhD

College

School of Pharmacy

Department

Pharmaceutical Systems and Policy

Committee Chair

Kamal Khalid

Committee Member

Usha Sambamoorthi

Committee Member

George A. Kelley

Committee Member

Murtuza Bharmal

Committee Member

Traci LeMasters

Committee Member

Joanna Kolodney

Abstract

Although considered a rare form of nonmelanoma skin cancer, Merkel Cell Carcinoma incidence rate has increased to 0.74 per 100,000 and projections expect this number to steadily rise over the next decades and to reach an average of 3,285 cases per year in 2025. MCC incidence increases significantly with age. Among those aged 40-44, it is 0.44, it rises to 1.0/100,000 in the 60-64 age group and soars to 9.8/100,000 among those ≥85 years and above. Due to the small number of cases, MCC is often managed using institutional practices or provider preferences. Formal clinical trials that enroll early stage MCC are ongoing and therefore, robust estimates of overall survival by stage and clinical management are not yet available. However, case series and retrospective studies and single-site studies suggest better overall survival with early stages. Thus, the real benefit of current standard of care needs to be assessed in early-stages to obtain a comprehensive overview of MCC treatment. Further, chronic conditions are highly prevalent in older adults with MCC. Although, evidence speaks of their prevalence, less is known about their temporal origin and their influence on cancer treatment choice. As older adults with cancer have on average three or more chronic conditions, these comorbidities typically precede a cancer diagnosis. Patients with comorbidities have poor prognosis and survival, are less likely to receive curative treatment choices and also are more likely to be offered less aggressive care. Thus, the presence and type of chronic conditions in MCC requires serious investigation as well as their impact on the choice of MCC treatment type. Cancer comorbidities not only influence treatment but can also affect economic burden among older adults with MCC. A diagnosis of cancer often leads to the neglect of any underlying chronic condition. an acute cancer diagnosis with an underlying chronic condition. However, the impact of chronic conditions among MCC patients over time remains unknown. To fill this knowledge gap, the three related aims of this dissertation were: 1) To conduct a systematic review with meta-analysis, wherever possible, of the effectiveness and safety of SRx, RTx, CTx or combination therapies in MCC patients; 2) To investigate the association of presence and type of pre-existing chronic conditions with treatment patterns among patients with incident primary MCC using interpretable machine learning methods; 3) To build predictive and interpretable machine learning of economic burden (total healthcare and out-of-pocket expenditures) and their association with chronic conditions among patients with incident primary MCC. In the first aim, among early-stages MCC patients, a statistically significant improvement in overall survival (HR=0.78, 95% CI, 0.62 to 0.99), and disease-free survival (HR=0.35, 95% CI, 0.13 to 0.93) was observed for adjuvant RTx compared to SRx. Although not statistically significant, adjuvant RTx also improved control of regional recurrence (HR=0.41, 95% CI, 0.09 to 1.78), as well as disease specific survival (HR=0.58, 95% CI, 0.24 to 1.40). However, it decreased control of local recurrence LR (HR=1.52, 95% CI, 0.37 to 6.19) compared to SRx. Subgroup analyses revealed that adjuvant RTx was more effective among local compared to regional stages. The E-value informed on the likelihood of total radiotherapy dose as a confounder, but also suggested that chemotherapy provided to those that had already received adjuvant radiotherapy did not impact the strength of evidence of the aforementioned outcomes. In the second aim, among older (> 65 years) adults with primary incident MCC (N=1,668), high cholesterol (75.5%), HIV (71.5%), hypertension (67.7%), arthritis (54.9%), coronary artery disease (47.1%), diabetes (43.5%), and hepatitis (37.1%) were some of the highly prevalent pre-existing chronic conditions. MCC treatment varied by type of chronic conditions and treatment modality. For example, a lower percentage of those with hypertension received ITx compared to those without hypertension (5.7% vs. 17.1%). XGBoost predictions revealed high predictive accuracy (area under the curve ranged from 0.72 (CTx) to 0.99 (ITx)). Hypertension (ITx), diabetes and thyroid disorders (HTx), thyroid diseases (RTx), and high cholesterol (CTx) were among the top ten predictors of specific MCC treatment. In the third aim, among older Medicare beneficiaries with incident MCC, congestive heart failure (CHF), chronic kidney disease (CKD) and depression had the highest average incremental total expenditures during pre-diagnosis, treatment, and post-treatment phases respectively ($25,004, $24,221, and $16,277 (CHF); $22,524, $19,350, $20,556 (CKD); and $21,645, $22,055, $18,350 (depression)). Whereas the average incremental out-of-pocket expenditures during the same periods were: $3703, $3,013, $2,442 (CHF); $2,457, $2,518, $2,914 (CKF); and $3,278, $2,322, $2,783 (depression). Except for hypertension, and HIV all chronic conditions had higher expenditures compared to those without the chronic conditions. Predictive models in each of the three phases indicated that CHF, CKD, and heart diseases were among the top 10 leading predictors. The root mean square error (RMSE) ranged from as low as 0.63 (during treatment) to as high as 1.05 (pre-diagnosis) for total expenditures, and from 0.52 (during treatment ) to 0.8 (pre-diagnosis) for out-of-pocket suggesting modest predictive performance. Although CHF, CKD, heart diseases was among the top common 10 leading predictors of total and out-of-pocket expenditures, their feature importance ranking declined over time. The findings from these three aims have implications for programs and policies, treatment, and practice. Our systematic review suggested that research with robust study designs are needed to compare the effectiveness of cancer therapies for early-stages MCC. Findings from our retrospective studies with real world data, comprehensive list of biological, clinical, lifestyle factors and social determinants of health applications of novel predictive and interpretable machine learning methods suggest the heterogeneous associations of chronic conditions with treatment and economic burden among older adults with MCC. These findings call for a personalized approach to treatment and cost containment efforts.

Embargo Reason

Publication Pending

Download

Available for download on Friday, January 03, 2025

Share

COinS