Thyroid Cancer Active Surveillance: The Devil You Know or The Devil You Don't

Perhaps one of the most controversial topics in thyroid cancer is the extent of surgery. For decades the pendulum has swung between the purported merits of hemithyroidectomy or total thyroidectomy, with advocates fiercely debating oncological considerations and long-term prognosis. Clinicians have wrestled with balancing the comparative risk of complications, the need for thyroid hormone replacement, and the quality-of-life changes (e.g., asthenia) that to this day remain difficult to forecast.

The recognition that many thyroid cancers are indolent has lent credence to more minimalist approaches. Guidelines now support either approach for 1–4 cm papillary thyroid carcinomas (PTCs) or hemithyroidectomy alone for ≤1 cm cancers (microPTCs). ¹ Yet the high incidence and superb outcomes for small thyroid cancers betray the initial debate as simplistic, and perhaps spurious. It begs the question that physicians and patients increasingly face for many of their low-risk patients: is surgery needed at all?

In this issue, two groups greatly contribute to answering this question, building upon initial Japanese studies that pioneered the concept of thyroid cancer active surveillance. ^2,3 In South Korea, Lee and colleagues in the multicenter prospective MAeSTro trial followed 1177 microPTC patients who self-assigned to surgery (36%) or surveillance (64%). ⁴ Progression was defined as 3 mm growth (single diameter), 2 × 2 mm growth, extrathyroidal extension, or lymph node metastasis (LNM). Of interest, they found that younger age, male sex, and tumor size >6 mm were associated with progression.

Concurrently, Tuttle and colleagues in the United States conducted active surveillance for 483 PTCs up to 1.5 cm in size. ⁵ Progression was defined as >72% increase in tumor volume or new LNM. Based on observations, they proposed six tumor volume kinetic patterns that categorize how cancers change over time, to guide decision making. By far the most common pattern was stability (79%); reassuringly, the rarest pattern was stability followed by sudden growth (0.6%).

While the studies tackled active surveillance differently, it is their similarities that are most striking: 5-year progression was 14.2–15.9%, 77–82% remained on active surveillance, 1.3–1.4% developed LNM, and 5.5–5.6% showed tumor shrinkage between the Korean and U.S. studies, respectively. Such parallels over large cohorts, together with other recent studies, ^{6 –9} indicate high generalizability: active surveillance, like prostate cancer, ¹⁰ seems effective and not just limited to certain populations, particular countries, or select centers.

Perhaps most encouragingly, it is important to acknowledge that progression is not a surrogate indicator for death, but merely a proxy for surgery: all patients who went on to delayed surgery did well. The relatively short follow-up (3–4 years), however, warrants longer monitoring to validate their findings in a slow-growing cancer.

While these studies are prospective, the nonrandomized nature is a limitation. It is true that a randomized controlled trial, using ultrasound to measure progression, might solidify acceptance for active surveillance. Yet perhaps the bigger limitation of these studies is what ultrasounds cannot objectively capture: patient preference. In the Korean study, 36% elected for surgery, while 11% in the U.S. study declined active surveillance—some may consider this encouraging, others unrealistic in a community setting.

Increasingly, factors such as patient worry and individual priorities are considered alongside cancer eradication. These qualitative factors may eclipse tumor volume doubling time or sonographic features. ^11,12 Physician reluctance remains another parameter that has proven difficult to overcome, ¹³ as is the fact that thyroidectomy could be considered a cleaner more definitive approach compared with years to decades of monitoring. Other interventions that straddle surgery and surveillance, such as radiofrequency ablation, are also emerging, ¹⁴ though are not without flaws.

In the end, active surveillance will likely take its place alongside surgery, as one of numerous pillars of treatment, rather than simply replace it. There will always be a spectrum of disease whereby surveillance is more appropriate; there will always be a spectrum of patient anxiety whereby surgery is best. Other recent trial data validate the concept that psychosocial distress may be strong determinants for decision making. The investigators should be congratulated for further illuminating the uncertainties underlying active surveillance. As an approach that is now no longer theoretical, it is now up to patients and physicians to decide whether the devil they know is better than the devil they don't.