The rise and fall of surgical aggressiveness for known or suspected differentiated thyroid cancer

Abstract

“The scientist explains the world by successive approximations.” —Edwin Hubble Sonographically detectable thyroid nodules affect up to one-half of the adult population. Because the underlying malignancy risk is approximately 5%, the first principle in managing thyroid nodules is to avoid unnecessary intervention in the 95% of benign cases (for the moment, let us set aside symptomatic benign enlargement and benign hyperfunction). At the same time, appropriate intervention is of course required in the small fraction of patients who harbor malignant nodules likely to cause morbidity and (rarely) mortality. Many diagnostic interventions, such as fine-needle aspiration biopsies and thyroid resections for indeterminate nodules, arise because of clinical uncertainty: although the risk of an adverse outcome related to a given nodule may be small, this nonzero risk drives clinicians and patients to achieve a state of reassurance. Depending on the balance of the scientific literature, the state of reassurance may be achieved at various average levels of intervention per capita. Over the past several decades, surgical aggressiveness for known or suspected differentiated thyroid cancer (DTC) has risen and then fallen in response to a series of important scientific developments. Sir William Osler stated that, “Medicine is a science of uncertainty and an art of probability.” Indeed, the management of thyroid nodules is a distillation of this maxim. Like all decisions, medical decisions are most straightforward at the extremes of probability. The decision to proceed with surgery is easy to make when a malignant diagnosis is close to certain (>95%), and the decision to observe is easy when the risk of malignancy approaches zero (<5%). Decision-making is difficult in the broad realm of intermediate probabilities.1 Since its inception, diagnostic thyroid cytopathology has been characterized by 3 basic categories: malignant, benign, and indeterminate. The indeterminate category is often considered troublesome because of the clinical uncertainty it carries. However, the indeterminate category is the precise reason why thyroid cytopathology is useful at all: its existence permits the diagnostic probabilities associated with the malignant and benign categories to be pushed to the extremes, where patients can confidently undergo surgery or comfortably be observed. In the absence of the indeterminate category, both the malignant and benign categories would be contaminated to an extent that would bring them into the range of intermediate probabilities, thus greatly diminishing patient value. Debates regarding the appropriate extent of initial surgery for low-risk DTC date back several decades.2 A series of events occurring between 1980 and 2012 caused a progressive increase in surgical aggressiveness for known or suspected thyroid cancer (Table 1).3-24 The development of thyroglobulin as a tumor marker provided a useful quantitative indicator of disease status during thyroid cancer surveillance. At the same time, because death from thyroid cancer is rare, thyroglobulin levels and structural recurrences detected on imaging emerged as secondary outcome measures of treatment efficacy. Put another way, the achievement of low thyroglobulin levels became a goal in and of itself. As thyroglobulin assays and ultrasound imaging grew progressively more sensitive, so did the intensity of postoperative scrutiny, driving surgeons toward increasingly aggressive operations. These operations aimed not only to prevent clinical recurrences but also to eliminate subclinical recurrences (even “biochemical-only” recurrences that manifest as isolated, nonzero thyroglobulin levels). A landmark 1994 study by Mazzaferri and Jhiang established total thyroidectomy, radioactive iodine ablation, and the administration of levothyroxine to suppress plasma thyroid-stimulating hormone levels as the standard of care for the initial treatment of DTC.5 This treatment strategy was reinfo