Introduction
Differentiated thyroid cancer (DTC), which includes papillary and follicular thyroid cancers, is the most common endocrine malignancy and accounts for 90% of all thyroid malignancies [
1‐
4]. Therapy modalities of DTC mainly include surgery, radioactive iodine (RAI) therapy and thyrotropin (TSH) suppression therapy [
5‐
7]. RAI therapy plays a crucial role in the purge of potential residual thyroid cancer and reducing risk of mortality [
7,
8]. DTC, with appropriate therapy, tend to have excellent prognosis [
7,
9,
10]. However, recurrence and persistence can occur several decades later [
11‐
13]. Patients having persistent radioiodine avid lesions usually require long-term surveillance and repeated high dose of RAI [
14,
15]. Therefore, it is very important to identify risk factors to predict disease persistence for DTC in order to help clinician to determine the most appropriate follow-up for these patients.
One of the risk assessment tools for persistence and recurrence is dynamic risk stratification (DRS) [
7,
16,
17]. Based on an optimal response to initial therapy, DRS is composed of four response groups: excellent, indeterminate, biochemical incomplete and structural incomplete. Biochemical incomplete response (BIR) classification was defined as having persistently abnormal suppressed and/or stimulated thyroglobulin (Tg) or rising anti-Tg antibodies (TgAb) without structural evidence of disease [
7]. Patients exhibiting BIR have shown to have a high risk of structural recurrent and persistent disease [
18,
19].
Data about BIR in Middle Eastern DTC is limited. Therefore, we conducted this retrospective study on a large cohort of Middle Eastern DTC to determine: 1. The prevalence of BIR in this cohort, 2. Clinico-pathological parameters to predict BIR after surgery and before RAI, and 3. Long-term clinical outcomes of patient with BIR. Furthermore, we tried to investigate the relationship between the timing of initiating RAI therapy and BIR in this cohort.
Materials and methods
Clinical cohort
One-thousand eight-hundred and twenty-two DTC patients diagnosed between 1988 and 2018 at King Faisal Specialist Hospital and Research Center (Riyadh, Saudi Arabia) were available to be included in the study. The patients were included in the study after considering the following inclusion and exclusion criteria:
Inclusion criteria
1.
Pathologically proven DTC post total thyroidectomy
2.
Received radioactive iodine ablation post surgery
3.
Adequate clinical follow-up data for at least 12 months is available
Exclusion criteria
1.
Less than total thyroidectomy (n = 63)
2.
No radioactive iodine ablation received (n = 183)
3.
Patients with elevated TgAb during follow-up (n = 121)
4.
Patients with persistent positive WBS (n = 118)
5.
Patients with follow-up duration less than 12 months (n = 51)
After applying the inclusion and exclusion criteria, 1286 DTC patients were included for the final analysis. The Institutional Review Board of the hospital approved this study and since only retrospective patient data were used, the Research Advisory Council (RAC) provided waiver of consent under project RAC # 221 1168 and # 2110 031. The study was conducted in accordance with the Declaration of Helsinki.
Clinico-pathological and follow-up data
Baseline clinico-pathological data were collected from case records and have been summarized in Table
1. Staging of DTC was performed using the eighth edition of American Joint Committee on Cancer (AJCC) staging system [
16]. Following initial surgery, all patients had pre-ablative stimulated Tg (presTg) evaluated before receiving RAI therapy. Patients were stratified into low, intermediate and high risk based on 2015 ATA guidelines [
7]. Low-risk DTC patients were followed up annually, intermediate risk patients were followed up at 6 months’ intervals and high risk patients were followed up at 3 months’ intervals. At each follow-up, neck ultrasound, thyroid function tests, thyroglobulin (Tg) levels and thyroglobulin antibodies were performed. In addition, for high risk patients, whole body scan (WBS) and/or PET CT scan were performed to identify tumor persistence/recurrence. Patients with stimulated Tg of less than 10.0 ng/ml with no clinical or imaging evidence of tumors were considered complete treatment responses. BIR was considered in patients with a raised stimulated serum Tg level (>10 ng/ml) and a negative WBS.
Table 1
Clinico-pathological details of the study cohort (n = 1286)
Age at diagnosis, years (mean ± SD) | 38.9 ± 16.5 |
Gender | |
Male | 322 (25.0) |
Female | 964 (75.0) |
Tumor laterality | |
Unilateral | 858 (66.7) |
Bilateral | 421 (32.7) |
Unknown | 7 (0.6) |
Tumor focality | |
Unifocal | 638 (49.6) |
Multifocal | 642 (49.9) |
Unknown | 6 (0.5) |
Extrathyroidal extension | |
Present | 552 (42.9) |
Absent | 724 (56.3) |
Unknown | 10 (0.8) |
pN | |
N0 | 499 (38.8) |
N1 | 664 (51.6) |
Nx | 123 (9.6) |
Distant metastasis at diagnosis | |
Present | 72 (5.6) |
Absent | 1214 (94.4) |
Stage | |
I | 1087 (84.5) |
II | 137 (10.6) |
III | 14 (1.1) |
IV | 47 (3.7) |
Unknown | 1 (0.1) |
BRAF mutation | |
Present | 672 (52.2) |
Absent | 540 (42.0) |
Unknown | 74 (5.8) |
TERT mutation | |
Present | 150 (11.7) |
Absent | 1010 (78.5) |
Unknown | 126 (9.8) |
Interval to RAI therapy | |
<3 months | 630 (49.0) |
≥3 months | 656 (51.0) |
ATA risk stratification | |
Low | 198 (15.4) |
Intermediate | 450 (35.0) |
High | 638 (49.6) |
BRAF and TERT mutation analysis
BRAF and
TERT mutation data for the DTC cohort was available from our previous studies [
20,
21].
Statistical analysis
The associations between clinico-pathological variables and BIR was performed using contingency table analysis and Chi square tests or Mann-Whitney U test for categorical and continuous variables, respectively. Disease-free survival (DFS) was determined using Kaplan-Meier estimates. DFS was defined as the time from diagnosis to the occurrence of recurrent disease or death. Logistic regression analysis was used for analyzing the prognostic factors that could predict BIR and structural persistent disease, in univariate and multivariate manner. Two-sided tests were used for statistical analyses with a limit of significance defined as p value < 0.05. Data analyses were performed using the JMP14.0 (SAS Institute, Inc., Cary, NC) software package.
Receiver operating characterisitcs (ROC) curve analysis was performed using MedCalc software, version 10.4.7.0 for Windows (MedCalc, Ostend, Belgium).
Discussion
RAI therapy is a well-established therapeutic modality for DTC patients. Despite excellent prognosis, a significant percentage of DTCs develop biochemical incomplete response (BIR) [
18,
22,
23]. Data on prevalence of BIR, clinico-pathological and biochemical risk factors to predict BIR after surgery and before radioiodine ablation in DTC patients from Middle Eastern ethnicity is not fully explored. Therefore, we conducted this retrospective study to evaluate BIR prevalence as well as predictive markers after surgery and before RAI therapy. In addition, we also evaluated long-term clinical outcomes of DTC patients showing BIR following initial therapy.
In this retrospective study of 1286 DTC patients, we noted that 20.7% (266/1286) of the patients show BIR. This incidence is consistent with a previous report [
24]. However, a recent study has identified higher incidence (~36%) of BIR and have attributed this to the fact that most of their patients belong to intermediate or high risk categories [
18]. In our study as well, ~90% of BIR patients are in the intermediate or high-risk category.
We further evaluated multiple clinico-pathological, biochemical and molecular markers to predict the risk of BIR before radioiodine ablation. In our study, we noted four risk factors as significant predictors of BIR on multivariant analysis: presTg, the time interval between thyroidectomy and first dose of RAI therapy (≥ 3 months), male gender, and LNM. Several previous reports have identified the impact of post-operative TSH stimulated Tg in predicting persistent and/or recurrent disease in DTC [
25‐
28]. Our result is consistent with a recent study that found BIR risk to be very high in patients who had high presTg [
18].
In our study, we found that presTg ≥6.9 ng/ml had 96.2% sensitivity to predict BIR prior to RAI ablation. We further noted that low presTg (<6.9 ng/ml) had excellent NPV (99.2%) to rule out BIR. We found reasonable specificity for presTg cutoff ≥10 ng/dl (specificity = 80.6%). As opposed to high NPV, the PPV of presTg over 10 ng/ml was quite low (57%). However, the main value of presTg is as a negative predictor of BIR when presTg values are low. Although the predictive value of presTg below 10 ng/ml (which is routinely used) has been demonstrated in our analysis, it is likely that lower cutoffs such as 6.9 ng/ml suggested by ROC curve would demonstrate an even higher NPV, albeit for a smaller group of patients.
Interestingly, our study has identified that the timing of initiating radioiodine adjuvant therapy could be a significant predictor of BIR in patients from this ethnicity. Although, the current DTC guidelines have no recommendation for timing of RAI [
7], several previous studies have explored the relationship between RAI initiating time and DTC clinical outcome with conflicting conclusions [
23,
29‐
31]. Our study showed that delayed initial RAI (≥ 3 months after thyroidectomy) was an independent predictor of BIR in this cohort.
Another interesting finding in this study was the association between BIR and male gender. Sex disparity in the incidence of DTC and its effect have been well documented [
32‐
34]. However, the impact of gender on DTC from Middle Eastern ethnicity appears to be more pronounced as we have identified in our recent study [
35] that male sex was an independent prognostic factor for recurrence-free survival in PTC.
We found that presence of LNM was an independent predictor of BIR in this cohort. This is in concordance with several previous studies, where LNM was shown to be associated with unfavorable outcome in DTC patients [
36‐
38].
We further sought to evaluate the long-term outcome of BIR patients in this cohort. With a median follow-up of 9.8 years, 36% of patients had structural disease, which was found to be significantly associated with male gender and increasing Tg after initial therapy even in multivariant analysis. According to the number of risk factors, risk stratification related to poor DFS was attempted: low-risk group (having no risk factor), intermediate-risk-group (having any one risk factor), and high-risk group (having both risk factors). 10-year DFS rates in the low-, intermediate- and high-risk groups were 82.3%, 50.6%, and 21.4%, respectively. DFS was significantly different among the three risk groups, being worst in the high-risk group. Our findings suggest that a risk adaptive management in patients with BIR could be beneficial.
Our research has a few limitations. First, it was a single-center, retrospective study due to which bias cannot be excluded. Second, the study involved a specific ethnicity, which could prevent generalizing the findings on other patient populations.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.