Natural Course of the American Thyroid Association Response to Therapy Statuses (Dynamic Risk Stratification) in Differentiated Thyroid Cancer

Introduction

Differentiated thyroid cancer (DTC) is the most common endocrine malignancy and its incidence has been increasing over the last 4 decades [1-4]. Although some data suggest a real increase in incidence [3], it is generally attributed to better detection methods [2, 4, 5]. The widespread use of neck ultrasonography and other imaging modalities has resulted in high detection rates of thyroid nodules; a small but significant proportion of them are malignant [3-5]. The vast majority of those tumors are papillary thyroid cancer, and more than 70% are small in size and carry minimal risk for mortality [1]. Realizing these facts, many experts in the field proposed a change in the diagnostic and therapeutic approaches to these low-risk DTC [6-9]. Over the last decade, this better understanding of the pattern and clinical course of DTC has resulted in a paradigm shift in its management with the adoption of risk-tailored strategies based on the estimation of the risk of recurrence and mortality [8]. Tumors that are low risk are usually treated more conservatively with lobectomy and do not routinely receive radioactive iodine (I-131) while high-risk tumors are treated more proactively with total or near-total thyroidectomy with or without lymph node dissection and I-131 adjuvant therapy [6, 8]. The risk stratification concepts were further extended to include follow-up evaluations after the initial management based on the available information at the time of evaluation, so-called dynamic risk stratification [6, 10, 11]. Dynamic risk stratification indicates that risk changes due to the institution of therapeutic measures and the natural course of the disease [8-10]. For example, a high-risk tumor at the time of initial management may turn to be of low risk at the time of follow-up. The 2015 American Thyroid Association (ATA) guidelines embraced and emphasized these concepts, which have become part of the DTC standard of care [6].

While conservative approaches to the management of low-risk patients are gaining more acceptance, there are significant variations in practice between different institutions and different countries. Total or near-total thyroidectomy with or without lymph node dissection followed by I-131 remnant ablation/adjuvant therapy remains the most commonly practiced management strategy for intermediate and high-risk patients and even for a good proportion of low-risk patients [12, 13]. Patients are usually evaluated at 6–12 months after I-131 ablation/therapy for the response to therapy of the disease. The ATA guidelines classify the response to therapy into four categories [6]. An excellent response indicates no clinical, biochemical, or radiological evidence of the disease with suppressed serum Tg level <0.2 ng/mL or stimulated thyroglobulin (Tg) level of <1 ng/mL in the absence of anti-Tg antibodies (Abs) and negative imaging studies. An indeterminate response refers to a situation when residual or recurrent disease cannot be confidently excluded with mild elevation of Tg (suppressed Tg 0.2 to <1 ng/mL or stimulated Tg 1 to <10 ng/mL), positive anti-Tg Abs, and/or some nonspecific findings on imaging studies. A biochemically incomplete status refers to a significant elevation of serum Tg (suppressed ≥1 ng/mL or stimulated ≥10 ng/mL) in the absence of radiological or cytological evidence of the disease. A structurally incomplete status refers to the presence of disease on radiological, cytological, and/or histopathological investigations, usually accompanied by a significant elevation of serum Tg or anti-Tg Abs. The natural history of these different statuses over time is important for predicting what will happen to the patients and for implementing timely and cost-effective monitoring and intervention strategies. In this study, our primary objective was to study the natural course of the response to therapy statuses after the initial management over time. Our hypothesis is that the response to therapy at the first evaluation after the initial management is highly predictive of the course of the disease over time and those patients in an excellent response are expected to do well over many years, need less intense follow-up and can be considered for discharge after a few years of follow-up. By contrast, we hypothesized that patients with an incomplete response to therapy are destined for recurrence, progression, and death and this group needs close follow-up and interventions when appropriate. For that reason, we analyzed the long-term course of patients in any of these statuses at the first evaluation after I-131 ablation/therapy over many years of follow-up. In addition, we assessed the performance of the ATA and the American Joint Committee on Cancer Tumor Node Metastasis (AJCC TNM) staging system in our patients.

Patients and Methods

This study was undertaken at the King Faisal Specialist Hospital & Research Center (KFSHRC), Riyadh, Saudi Arabia. An institutional review board approval was obtained from the Office of Research Affairs of KFSHRC. KFSHRC is the premier tertiary care institution and the main cancer care center in Saudi Arabia. Each year, 200–300 patients with newly diagnosed thyroid cancer are referred to KFSHRC for management and long-term follow-up. The management of DTC has been following international standards and guidelines since 1998. Except for very low-risk patients (T1N0M0), our practice had been in the form of near-total or total thyroidectomy and therapeutic lymph node dissection if pathological lymph nodes are found on preoperative evaluation or intraoperative examination. Radioactive iodine (I-131) ablation had been given to most cases except for very low-risk patients (T1aN0M0). Thyroid hormone suppression was practiced aiming at serum TSH level of 0.1–0.5 mU/L for low to intermediate risk and <0.1 mU/L for high-risk patients. Patients were usually evaluated 6–18 months after I-131 ablation with measurement of serum TSH, FT4, Tg, and anti-Tg Abs, and by neck ultrasonography. I-123 diagnostic whole-body scan and stimulated Tg levels were usually done once after I-131 ablation/therapy for patients with high and intermediate risk features without evidence of distant metastasis and more frequently as a preparation for more I-131 therapies for those with distant metastasis. Patients in an excellent response or an indeterminate response were usually seen on a yearly basis for follow-up, mostly with suppressed Tg, anti-Tg Abs, TSH, FT4, and neck ultrasonography. Other imaging studies (including computed tomography (CT) of the lungs and neck, combined CT/F-18 flurodeoxyglucose positron emission tomography and other imaging as appropriate) and more frequent monitoring were followed in patients with evidence of local disease or distant metastasis.

To ensure that we have an adequate long-term follow-up period for this study, we studied all patients diagnosed during the period January 2004–December 2006. A total of 814 patients with thyroid cancer were diagnosed at our institution during this period (Fig. 1). Considering the aim of the study of assessing the natural course of different ATA statuses and in order to ensure a homogeneous population that represents the majority of patients in the usual clinical practice, we excluded patients with anaplastic (18 patients), poorly differentiated (19 patients), and medullary thyroid cancers (19 patients). For the same reasons, we also excluded patients who did not receive I-131 (146 patients) since the definitions of different ATA response to therapy statuses depend to a great extent on serum Tg with different cutoff values for patients who received I-131 compared to those who did not. Finally, we excluded patients who did not have adequate follow-up data or were lost to follow-up (111 patients). A total of 501 cases of DTC remained and were analyzed for the objectives of this study (Fig. 1).

View Full Size

Overall design and the results of the study showing the initial number of patients, the excluded patients, the initial management, response to therapy at the first evaluation after thyroid surgery and I-131 ablation, the additional interventions for a subgroup of 114 patients, and the final response to therapy statuses at the last follow-up.

Citation: European Thyroid Journal 10, 3; 10.1159/000511708

• Download Figure
• Download figure as PowerPoint slide

Results

Characteristics and Initial Management

We studied 501 patients (102 males and 399 females) with a median age of 37 years (IQR 29–48). The clinical and pathological characteristics are summarized in Table 1. The overall study design and the main results are summarized in Figure 1.

Table 1.

Clinical and pathological features of 501 patients with DTC

View Table

ATA and AJCC TNM Staging Systems Predict the Response to Therapy at First Evaluation and Last Follow-Up

The majority of patients were in low (246 patients) to intermediate (199 patients) ATA stages (Table 1). At the first evaluation after the initial management, 64.2 and 25.6% of patients in ATA low stage and 48.7 and 27.6% of the ATA intermediate stage achieved an excellent response and an indeterminate response, respectively (Table 2). By contrast, the majority of patients in the ATA high-stage DTC had either structurally incomplete (58.9%) or biochemically incomplete (14.3%) statuses (Table 2).

Table 2.

Response to therapy status at the first evaluation after initial management (thyroidectomy and I-131 ablation) and at the last evaluation against the ATA and AJCC TNM8 staging systems

View Table

Similarly, 95.1 and 83.4% of the patients in the ATA low and intermediate stages achieved an excellent response at the last follow-up evaluation. In comparison, only 30.4% of patients in the ATA high stage achieved an excellent response, and 66.1% continued to have a structurally incomplete response, including 17.9% of them who died secondary to structural disease (Table 2). The ATA staging was highly predictive of the long-term outcome (Fig. 2a).

View Full Size

Kaplan-Meier analyses showing the disease-free probabilities based on the ATA risk stratification (a), the AJCC/TNM8 stages (b) and response to therapy statuses at the first evaluation after initial management (c). All of these risk stratification and staging systems correlated well with the final outcome (median duration of 105 months [IQR 71–132] after I-131 ablation/therapy). ATA, American Thyroid Association; AJCC TNM, American Joint Committee on Cancer Tumor Node Metastasis.

Citation: European Thyroid Journal 10, 3; 10.1159/000511708

• Download Figure
• Download figure as PowerPoint slide

AJCC TNM staging also was predictive of the outcome at first follow-up and last evaluation after the initial management (Table 2). At the first evaluation, around 60% of stage I were in an excellent response while 70.8% of stage IV had a structurally incomplete disease. At the last evaluation, 93.3% and 49.1% of stage I and II were in an excellent response while 100% and 91% of stage III and IV had a structural disease (Table 2). AJCC TNM staging was also predictive of the long-term outcome although less distinctly than ATA stages as some overlap between AJCC TNM stages occurred (Fig. 2b).

Overall Course of the Response to Therapy Statuses

At a median duration of 105 months (IQR 71–132) after I-131 ablation/therapy, the change of the response to therapy status from the time of the first evaluation after the initial management to the last visit of the 501 patients including those who had no additional intervention (387 patients) and those who had 1 or more additional interventions (114 patients) was as follows (Fig. 3): excellent status, 263 patients (52.5%) at initial evaluation increased to 417 (83.2%) at last visit; indeterminate status, 125 patients (25%) decreased to 5 patients (1%); biochemically incomplete, 32 patients (6.4%) decreased to 18 patients (3.6%); and structurally incomplete, 81 patients (16.2%) decreased to 61 patients (12.2%) including 11 patients (2.2%) who died with structurally incomplete disease. The response to therapy status at the first evaluation after the initial management was predictive of the response to therapy statuses at the last follow-up (Fig. 2c). A major change occurred in the indeterminate status group. However, there was no significant difference in this change between those who were labeled to have an indeterminate status based on positive anti-Tg antibodies (36/38 patients achieved excellent response at the last follow-up) and those who had mildly elevated Tg or nonspecific changes on imaging studies (79/87 patients achieved an excellent response at the last follow-up) with p value of 0.72.

View Full Size

A summary diagram showing the changes in the numbers and percentages of patients in different response to therapy statuses at the first evaluation after the initial management at a median time of 17 months (IQR 14–22) after I-131 ablation and at the last follow-up visit (median duration 105 months [IQR 71–132] after I-131 ablation). IQR, interquartile range. *Including 11 patients who died with a structurally incomplete status.

Citation: European Thyroid Journal 10, 3; 10.1159/000511708

• Download Figure
• Download figure as PowerPoint slide

Discussion

In this study, we have assessed the natural course of patients with different ATA response to therapy statuses after the initial management. We have demonstrated that (1) there is a good correlation between the ATA and AJCC TNM8 staging systems and the outcome at the first evaluation after initial management and at the final evaluation, (2) the ATA response to therapy status at the first evaluation is predictive of the long-term outcome, (3) the majority of patients with an indeterminate status and a significant proportion of those in a biochemically incomplete status achieved an excellent response status over time without further intervention, and (4) disease progression and mortality occurred predominantly in high ATA stage and the structurally incomplete response group at the first follow-up evaluation visit after initial management.

The ATA and AJCC TNM staging systems are excellent tools to assess the risk of recurrence and mortality and their predictive power has been demonstrated in many studies from around the world [8, 9, 14-18]. This is again demonstrated in this study from a Middle Eastern population (Table 2). However, these staging systems are static systems that utilize clinicopathological features and radiological information available at the time of initial management and do not consider the possible changes in the clinical course over time and the additional biochemical and radiological data that become available during follow-up. This is where the concept of dynamic risk stratification was introduced which is based on the fact that the disease status may change over time, and the risk is a dynamic process that should be assessed at each visit [6, 8, 10, 17]. Initially, dynamic risk stratification was suggested to be used during the first 2 years after the initial management [10]. However, studies have shown that this system is valid and robust at all stages of surveillance as it utilizes the available clinical, biochemical, and radiological data at any time of assessment during the follow-up [9, 14, 19]. Our study shows that this system is highly reliable in predicting the long-term outcome. More importantly, it shows also that the dynamic response to therapy risk stratification during the first 1–2 years highly correlates with the dynamic risk stratification at the last follow-up evaluation (online suppl. Table 2, 3; Fig. 3). In fact, the overall outcome of patients at the last follow-up was even better than at the first 1–2 years as many patients achieve excellent response either spontaneously (the majority of patients) or after additional therapies (Fig. 3; online suppl. Table 2, 3).

An indeterminate status is a situation where the presence or absence of the disease cannot be confidently confirmed [6, 8]. It is characterized by mildly detectable Tg (<1 ng/mL on thyroxine or 1–<10 ng/mL off thyroxine), elevated anti-Tg Abs that make Tg level unreliable as a tumor marker, and/or nonspecific imaging findings [6]. The ATA guidelines recommend expectant management of patients with an indeterminate response to therapy [6]. In our study, at the first evaluation after the initial management, over a quarter (26.1%) of the patients who did not receive any further intervention were in an indeterminate response (online suppl. Table 2). Without any further intervention, only 1% of them continued to have an indeterminate status. The vast majority of them spontaneously achieved an excellent response (online suppl. Table 2; Fig. 1). This supports following a more expectant approach with observation and follow-up of this group of patients. Although some studies suggested a difference in the outcome between those labeled to have an indeterminate status based on positive anti-Tg Abs and those due to mildly elevated Tg or nonspecific changes on imaging studies [20, 21], we did not find a significant difference between these 2 subgroups of indeterminate response to therapy. This might be due to the small numbers of patients with indeterminate response to therapy at the last follow-up (Fig. 3).

Similarly, patients who achieved an excellent response status at the first evaluation after the initial management continued to enjoy excellent long-term outcome (Fig. 1; Fig. 3). This indicates that recurrence of the disease is rare and most cases that develop structural disease during the follow-up had never been free of the disease (persistent disease). This also suggests that this large group of patients who are in an excellent response can be infrequently monitored with serum Tg and anti-Tg Ab and can be considered for discharge after a few years of follow-up. Patients with a biochemically incomplete status are presumed to have small volume disease that escapes the limits of detection of the commonly used imaging techniques, or their disease is in unusual sites that were not scanned or seen. However, 23% and 11% of them spontaneously regressed to excellent and indeterminate responses, respectively, while 29% of them progressed to structural disease (Fig. 1). The rest continued to have biochemical evidence of the disease without clear structural lesions.

Our study showed clearly that the group of patients that need more proactive monitoring and interventions are those with a structural disease since a significant percentage of them developed more progressive disease, and the vast majority of patients who died were in this category. This was also shown in previous studies. Therefore, close monitoring and appropriate interventions should be implemented as necessary in this group of patients. This is also supported by the fact that our study also shows that additional therapeutic interventions, when indicated, are effective in a good percentage of patients, as demonstrated in the 114 patients who received additional interventions in this study (online suppl. Table 3). In this subgroup, the percentage of patients who achieved an excellent response increased from zero before additional interventions to 60 patients (52.6%) while the rates of biochemically incomplete and structurally incomplete statuses decreased from 21.9% and 60.5% before interventions to 9% and 37.7% at the last follow-up (online suppl. Table 3).

Although dynamic risk stratification has become part of the routine care of patients with DTC, certain issues remained of concern including the impracticality and non-convenience of doing stimulated Tg and I-123 whole body scans. However, these are not needed in most patients as definitions of the different statuses in non-stimulated situations exist and correlate well with stimulated studies [6]. Definitions of the different statuses were initially established for patients who undergo total or near-total thyroidectomy and I-131 ablation [6]. However, currently more patients in the low and intermediate risk groups undergo lobectomy for DTC and do not receive I-131 ablation. For that reason, studies also have been performed to validate dynamic risk stratifications in patients undergoing lobectomy alone or total thyroidectomy without I-131 ablation [22-24].

Our study has some weaknesses and strengths. The study is retrospective and included patients at a time when the general approach for management of DTC was more in the form of “one size fits all.” However, the patients included were consecutive nonselected and were managed in the usual standard approach at the time of the study. All of them had near-total or total thyroidectomy ± lymph node dissection and I-131 ablation, and this minimized variation due to different management approaches. A large proportion of them (387 patients, 77%) did not receive any additional therapy after the initial management, and this enabled us to study the natural history of different ATA statuses over a long period of follow-up without intervention. It also confirms the usability and generalizability of the dynamic risk stratification across different ethnic populations. To further assess the robustness of the dynamic risk stratification, we analyzed a large subgroup of our patients (291 patients) who had an intermediate to high risk disease (T3-T4 disease and/or those with lymph node or distant metastases) and would have been treated with total or near-total thyroidectomy and I-131 ablation even in this era of the “minimalistic approach” to the management of DTC [6]. The results showed the ability of dynamic risk stratification at the first evaluation to predict the long-term outcome. It also shows that this is the group of patients in whom the risk of structural disease and mortality is high despite instituting aggressive management measures. Although the follow-up period in our study is long [median 105 months (IQR 71–132)], it is less than expected since the study included patients diagnosed in 2004–2006. The reason for this apparently shorter than expected follow-up time is that we have discharged a significant number of patients who have been in excellent response for years to decongest our practice in accordance with the ATA guidelines 2015 [6].

In summary, this study from a previously non-included ethnic group (the Middle East), shows that the “static” ATA and AJCC TNM staging systems perform very well in predicting the short- and long-term course of DTC and correlate well with the ATA response to therapy dynamic risk stratification system. The ATA response to therapy statuses in the first 1–2 years after the initial management predicts the long-term outcome and correlates well with ATA response to therapy statuses at the last evaluation. Most patients with an indeterminate response and a good percentage of those in the biochemically incomplete statuses spontaneously achieve excellent status over time without intervention, and an expectant management for these groups, especially the indeterminate status, is warranted. On the other hand, patients with a structurally incomplete status are at the highest risk of progressive disease and mortality and should be carefully monitored and managed.

Acknowledgements

The authors would like to thank colleagues at the Endocrine Section and Department of Medicine of the KFSHRC for their support.

Statement of Ethics

This study was undertaken at the King Faisal Specialist Hospital & Research Center (KFSHRC), Riyadh, Saudi Arabia. An institutional review board approval was obtained from the Office of Research Affairs of KFSHRC (RAC# 2130–015). Since the study was retrospective and involved only confidential medical record review without direct interaction with patients, a waiver of consent was granted by the Ethics Committee. All research included in this manuscript followed the guidelines for human studies and should include evidence that the research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

No specific funding was obtained for this study.

Author Contributions

Noha Mukhtar: conception of the research idea, data collection, tabulation, and analysis. Hadeel Aljamei: data collection, tabulation, and analysis. Abeer Aljomaiah: data collection, editing, and analysis. Yosra Moria: conception of the idea, analysis of the data, and review of the manuscript. Ali Alzahrani: conception of the research idea, supervising the whole research process, data analysis, and writing the manuscript.