Lisa Caulley

The introductory chapter provides an overview of the rationale for evidence-based surgery in the management of endocrine tumors in the head and neck and outlines the objectives for this thesis.

Evidence-based medicine is defined as the conscientious, explicit and judicious use of the best scientific evidence to make decisions about clinical care. The use of evidence-based medicine in the appropriate setting leads to an improvement in patient care and, in most cases, a decrease in resource waste. There is increasing pressure for evidence-based surgery and for promotion of best practices in surgical research and decision making. There are scientific approaches that do not require the highly complex and costly scientific techniques of clinical trials to evaluate contemporary practice of otolaryngology-head and neck surgery using the best-available evidence, including systematic reviews, analyses of real-world data in databases and creation of decision models that account for uncertainty.

This thesis explores the use of evidence-based techniques in evaluating the diagnostic and treatment strategies of endocrine tumors. Diagnosis and treatment of endocrine tumors requires a multidisciplinary approach, drawing frequently on the expertise of surgeons, endocrinologists, and oncologists for the diagnosis, surgical and medical management of these tumors and appropriate surveillance for disease recurrence.

PART 1: REVIEW OF DECISION MODELING

Chapter 2 reviews health care evaluation as it relates to otolaryngology-head and neck surgery. Decisions about resource allocation are increasingly based on value trade-offs between health outcomes and cost. This process relies on comprehensive and standardized definitions of health outcomes which accurately measure the physical, mental, and social well-being of patients across disease states. These health-related quality of life metrics can be obtained through clinical trials, observational studies, and health surveys, and facilitate the integration of patient preferences into clinical practice. This chapter provides a narrative review of health outcome valuation, as well as the integration of both quality and quantity of life into standardized metrics for health research, program planning, and resource allocation.

Chapter 3 addresses the role of decision analysis in otolaryngology-head and neck surgery. Methods have been developed to support decisions based on data. Understanding the related techniques can help promote an evidence-based approach to clinical management and policy. This chapter provides a narrative review and step-by-step introduction to decision models, their typical framework, and favored approaches to inform data-driven practice for patient-level decisions, as well as comparative assessments of proposed health interventions for larger populations.

Chapter 4 provides a comprehensive review of cost-effectiveness analyses in otolaryngology-head and neck surgery. Clinicians seek to pursue the most clinically effective treatment strategies, but costs have also become a key determinant in contemporary health care. Economic analyses have thus emerged as a valuable resource to both quantify and qualify the value of existing and emerging interventions and programs. Cost-effectiveness analyses estimate the benefits gained per monetary unit, providing insights to guide resource allocation. This chapter discusses the related concepts and considerations to facilitate understanding and appraisal of these analyses so as to better inform the stakeholders in our otolaryngology community and it provides a review of available cost-effectiveness analyses in otolaryngology-head and neck surgery.

PART II: PITUITARY ADENOMAS

In Chapter 5, we present a protocol for a systematic review and meta-analysis of outcomes from surgical and non-surgical randomized clinical trials (RCTs) for the management of growth-hormone secreting pituitary adenoma tumors. The protocol outlines the rationale, hypothesis, and planned methods of the review.

Chapter 6 provides a synthesis of evidence on outcomes from RCTs to evaluate the effects of surgical and non-surgical interventions for primary and salvage treatment of growth-hormone-secreting pituitary adenoma tumors in adults. Our analysis of 8 RCTs evaluates 445 adults with a growth hormone-secreting pituitary adenoma tumor that received surgical therapy alone, pharmacological therapy alone, or combination surgical and pharmacological therapy. Comparisons were categorized into 2 groups: surgical therapy alone versus pharmacological therapy alone; and surgical therapy alone versus preoperative pharmacological therapy and surgery. The evidence suggested no benefit or detriment of pharmacological therapy alone as compared to surgery alone in reducing disease-related diabetes, disease-related impaired glucose tolerance, or treatment-related biliary problems. We did not find evidence of improvement in health-related quality of life with pharmacological therapy alone compared to surgery alone. While evidence suggests possible benefit from preoperative pharmacological therapy and surgery intervention in reducing postoperative cerebrospinal fluid leak, the evidence did not suggest benefit or detriment of preoperative pharmacological therapy and surgery compared to surgery alone in reducing overall surgical complications, postoperative diabetes insipidus, surgery for recurrent or persistent disease, or nonsurgical therapy for recurrent or persistent disease. The current evidence is limited by the small number of included RCTs with small sample populations, as well as the unclear risk of bias in most studies.

Chapter 7 builds on the findings of Chapter 2-6 to perform a cost-effectiveness analysis of preoperative treatment of GH-secreting pituitary adenomas as compared to direct surgery using a decision analysis model. Under base case assumptions, direct surgery was found to be the dominant strategy as it yielded lower costs and greater health effects (QALYs) compared to the preoperative treatment strategy in the second-order Monte Carlo microsimulation. The incremental cost and effects were most sensitive to probability of remission following primary therapy and duration of preoperative octreotide therapy. Accounting for joint parameter uncertainty, direct surgery had a higher probability of demonstrating a cost-effective profile compared to preoperative octreotide treatment at 77% compared to 23%, respectively.

Chapter 8 integrates the current available literature on surveillance of pituitary adenomas following surgical resection with curative intent. An analysis of the literature revealed 77 articles that met the study inclusion criteria for data extraction. Surveillance strategies and timelines were widely heterogenous across studies. A pooled analysis demonstrated recurrence rates at 1 year, 5 years and 10 years for a non-functioning pituitary adenoma were 1%, 17%, and 33%; for prolactin-secreting adenoma were 6%, 21%, and 28%; and for growth-hormone pituitary adenoma they were 3%, 8% and 13%, respectively. Rates of recurrence prior to 1 year were less than 2% for non-functioning pituitary adenoma, prolactin-secreting pituitary adenoma and growth-hormone secreting pituitary adenoma. The mean time to disease recurrence for non-functioning pituitary adenoma, prolactin-secreting pituitary adenoma and growth-hormone secreting pituitary adenoma were 4.25, 2.52 and 4.18 years, respectively. Standardized reporting for duration and outcomes for post-operative surveillance were lacking and thus prevented a direct comparison of surveillance schedules. The paucity of clinical trials and cohort studies investigating cost-effectiveness of surveillance schedules and impact on quality of life of patients under surveillance suggest that further studies are needed to optimize follow-up.

In Chapter 9, we developed an individual-level state-transition microsimulation model to simulate costs and outcomes associated with 3 postoperative imaging strategies following surgery for non-functioning pituitary adenoma. The strategies were: 1) annual magnetic resonance imaging (MRI) surveillance; 2) tapered MRI surveillance which involved annual surveillance for 5 years followed by surveillance every 2 years thereafter; and 3) personalized surveillance described as annual surveillance for 5 years followed by surveillance every 2 years for equivocal MRI or surveillance at 7, 10 and 15 years for disease-free MRI. Under base case assumptions, personalized surveillance was considered cost-effective at a WTP threshold of $100,000 USD compared to alternative surveillance strategies. Tapered surveillance, from annual surveillance to biannual surveillance, was considered a low-cost surveillance strategy but provided less health benefit to patients. Annual surveillance for patients was found to be more costly and offered less health benefits for patients. When we included the impact of parameter uncertainty, the probability that personalised-surveillance was cost-effective was 79% and tapered-surveillance was 14% given a WTP threshold of $100,000 USD per QALY gained.

PART III: THYROID TUMORS

Chapter 10 provides a contemporary analysis of the mode of detection of thyroid tumors that resulted in surgery using real-world data. Sixteen hospitals in 4 countries performed a retrospective analysis of 1328 thyroid tumors and classified the mode of detection of thyroid tumors that led to surgery using a validated tool. Most thyroid cancers were discovered in patients without symptoms referable to the thyroid (40.8%), while 34% of surgeries were performed for tumors discovered due to thyroid-related symptoms, 14% for endocrinopathies and 12% for tumors under surveillance. Cancers detected in patients without thyroid-referable symptoms were on average smaller than symptomatic thyroid cancers (3.2 cm). However, some asymptomatic cancers were still large (range 0.05 to 8.8cm), consistent with findings from previous data.

Recently, the increased detection of early-stage papillary thyroid neoplasms without improvements in mortality has prompted development of strategies to prevent or mitigate overtreatment. Chapter 11 explored the uptake of two strategies developed to limit overtreatment of low-risk thyroid cancers: 1) a new classification - non-invasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP); and 2) hemithyroidectomy for selected papillary thyroid cancers up to 4 cm in size. A cross-sectional analysis of 3,368 pathology records from 18 hospitals in 6 countries during two time periods (2015 and 2019) demonstrated that only 38 (4.8%) were diagnosed as NIFTP. The proportion of eligible papillary thyroid cancers treated with total thyroidectomy in the two cohorts demonstrated a decreasing trend from 2015 to 2019: 75.3% vs 58.3%.

Chapter 12 provides a detailed assessment of the findings, limitations and future directions presented in this thesis. New trials in pituitary adenomas are needed to address the knowledge gap observed in these systematic reviews, including lack of long-term follow up, comprehensive reporting and analysis of treatment response by tumor subtype and direct comparison of postoperative surveillance strategies. Our findings suggest that most thyroid cancers continue to be discovered in asymptomatic patients. Further research into patients with thyroid cancer and locoregional or distant metastasis would provide additional insight into detection of high-risk disease. De-escalation strategies to prevent or mitigate overtreatment of early-stage papillary thyroid cancer have shown mixed success thus far. Our studies using real-world evidence can be optimized in future through use of advanced statistical techniques. Future decision models should be informed with the highest-quality trial data and adapted to reflect international perspectives and optimize external validity. Clinical trials that will provide the highest value in improving surgical care can be selected through value of information analyses to assess whether the current evidence is sufficient to determine the best intervention or if further trials are required. Future research efforts should be prioritized to identify direct evidence linking the use of these evidence-based studies to influence clinical guidance and improve health.

De inleiding van het proefschrift geeft een overzicht van de grondgedachte van evidence-based chirurgie binnen de behandeling van endocriene tumoren in het hoofd-halsgebied en schetst de doelstellingen voor dit proefschrift.

Evidence-based geneeskunde wordt gedefinieerd als het gewetensvol, expliciet en oordeelkundig gebruik van het best beschikbare wetenschappelijke bewijs om beslissingen te nemen over klinische zorg. Het juiste gebruik van evidence-based geneeskunde leidt tot een verbetering van de patiëntenzorg en, in de meeste gevallen, minder verspilling. De noodzaak voor evidence-based chirurgie en het bevorderen van optimale praktijken in de chirurgische besliskunde neemt toe. Er zijn verschillende wetenschappelijke benaderingen ontwikkeld waarvoor zeer complexe en kostbare wetenschappelijke klinische studies niet noodzakelijk zijn om de hedendaagse praktijk van de hoofd-hals chirurgie te evalueren, waaronder systematische reviews, analyses van geobserveerde gegevens in databanken en besliskundige modellen die onzekerheid in beschouwing nemen.

Dit proefschrift onderzoekt het gebruik van evidence-based technieken bij het evalueren van de diagnostische en behandelingsstrategieën van endocriene tumoren. De diagnostisering en behandeling van endocriene tumoren vereist een multidisciplinaire aanpak. Voor de diagnose is vaak specialistische kennis nodig van chirurgen, endocrinologen en oncologen over de diagnose, chirurgische en medische behandeling van deze tumoren en voor het monitoren van de patiënt.

DEEL 1: EVALUATIE VAN BESLISKUNDIGE MODELLEN

Hoofdstuk 2 geeft een overzicht van evaluaties met betrekking tot de otorinolaryngologie, beter bekend als keel-neus-oor-(KNO)-heelkunde. Beslissingen over het toewijzen van middelen zijn steeds vaker gebaseerd op een afweging tussen gezondheidsresultaten en kosten. Dit proces is gebaseerd op uitgebreide en gestandaardiseerde definities van gezondheidsuitkomsten die het fysieke, mentale en sociale welzijn van patiënten in verschillende ziektetoestanden nauwkeurig meten. Deze maatstaven voor kwaliteit van leven kunnen worden verkregen via klinische onderzoeken, observationele studies en gezondheidsenquêtes, en vergemakkelijken de integratie van patiëntvoorkeuren in de klinische praktijk. Dit hoofdstuk geeft een narratief overzicht van de waardering van gezondheidsuitkomsten, evenals de integratie van zowel kwaliteit van leven als het aantal levensjaren voor gezondheidsonderzoek, projectplanning en de toewijzing van middelen.

Hoofdstuk 3 behandelt de rol van besliskundige analyses in de KNO-heelkunde. Er zijn verschillende methoden ontwikkeld om beslissingen te ondersteunen met data. Deze technieken kunnen helpen bij het bevorderen van een evidence-based benadering van klinisch management en beleid. Daarom biedt dit hoofdstuk een narratief overzicht en een stapsgewijze introductie van besliskundige modellen, inclusief de gebruikelijke opzet van deze modellen. Ook behandelt dit hoofdstuk de gebruikelijke benaderingen om de praktijk te informeren over beslissingen op patiëntniveau en hoe gezondheidsinterventies vergeleken kunnen worden voor grotere populaties.

Hoofdstuk 4 geeft een uitgebreid overzicht van kosteneffectiviteitsanalyses binnen de KNO-heelkunde. Clinici streven er doorgaans naar behandelingsstrategieën te gebruiken die het grootste klinische effect opleveren, maar in de hedendaagse gezondheidszorg zijn ook de kosten van belang. Economische analyses zijn een waardevol hulpmiddel om de waarde van bestaande en nieuwe interventies zowel te kwantificeren als te kwalificeren voor uitvoering. Kosteneffectiviteitsanalyses schatten de verwachte gezondheidsvoordelen per kosteneenheid en bieden belangrijke inzichten om de toewijzing van middelen aan te sturen. Dit hoofdstuk bespreekt de gerelateerde concepten en overwegingen om het begrip en de beoordeling van deze analyses te vergemakkelijken. In dit hoofdstuk streven we er ook naar om de belanghebbenden in de KNO-gemeenschap beter te informeren middels een overzicht van beschikbare kosteneffectiviteitsanalyses in de KNO-heelkunde.

DEEL II: HYPOFYSEADENOMEN

In Hoofdstuk 5 presenteren we een protocol voor een systematisch review en meta-analyse dat kijkt naar de uitkomsten van chirurgische en niet-chirurgische gerandomiseerde klinische studies (RCT’s) voor de behandeling van groeihormoon afscheidende hypofyseadenomen. Het protocol beschrijft de grondgedachte, hypothese en geplande methoden om deze uitkomsten samen te vatten.

Hoofdstuk 6 geeft een