NEJM Iod Radioactiv

The n e w e ng l a n d j o u r na l of m e dic i n e
Original Article
Thyroidectomy without Radioiodine

in Patients with Low-Risk Thyroid Cancer
S. Leboulleux, C. Bournaud, C.N. Chougnet, S. Zerdoud, A. Al Ghuzlan,
B. Catargi, C. Do Cao, A. Kelly, M.-L. Barge, L. Lacroix, I. Dygai, P. Vera, D. Rusu,
O. Schneegans, D. Benisvy, M. Klein, J. Roux, M.-C. Eberle, D. Bastie,
C. Nascimento, A.-L. Giraudet, N. Le Moullec, S. Bardet, D. Drui, N. Roudaut,
Y. Godbert, O. Morel, A. Drutel, L. Lamartina, C. Schvartz, F.-L. Velayoudom,
M.-J. Schlumberger, L. Leenhardt, and I. Borget
A BS T R AC T
BACKGROUND
In patients with low-risk differentiated thyroid cancer undergoing thyroidectomy, The authors’ full names, academic de-
the postoperative administration of radioiodine (iodine-131) is controversial in the grees, and affiliations are listed in the
Appendix. Dr. Leboulleux can be contact-
absence of demonstrated benefits. ed at sophie.leboulleux@gustaveroussy.fr
or at the Department of Nuclear Medi-
METHODS cine and Endocrine Oncology, Gustave
In this prospective, randomized, phase 3 trial, we assigned patients with low-risk Roussy, Université Paris Saclay, 114 Rue
differentiated thyroid cancer who were undergoing thyroidectomy to receive abla- Edouard Vaillant, 94805 Villejuif, France.
tion with postoperative administration of radioiodine (1.1 GBq) after injections of N Engl J Med 2022;386:923-32.
recombinant human thyrotropin (radioiodine group) or to receive no postoperative DOI: 10.1056/NEJMoa2111953
Copyright © 2022 Massachusetts Medical Society.
radioiodine (no-radioiodine group). The primary objective was to assess whether
no radioiodine therapy was noninferior to radioiodine therapy with respect to the CME
absence of a composite end point that included functional, structural, and bio- at NEJM.org
logic abnormalities at 3 years. Noninferiority was defined as a between-group
difference of less than 5 percentage points in the percentage of patients who did
not have events that included the presence of abnormal foci of radioiodine uptake
on whole-body scanning that required subsequent treatment (in the radioiodine
group only), abnormal findings on neck ultrasonography, or elevated levels of
thyroglobulin or thyroglobulin antibodies. Secondary end points included prog-
nostic factors for events and molecular characterization.
RESULTS
Among 730 patients who could be evaluated 3 years after randomization, the per-
centage of patients without an event was 95.6% (95% confidence interval [CI], 93.0
to 97.5) in the no-radioiodine group and 95.9% (95% CI, 93.3 to 97.7) in the radio-
iodine group, a difference of −0.3 percentage points (two-sided 90% CI, −2.7 to
2.2), a result that met the noninferiority criteria. Events consisted of structural or
functional abnormalities in 8 patients and biologic abnormalities in 23 patients
with 25 events. Events were more frequent in patients with a postoperative serum
thyroglobulin level of more than 1 ng per milliliter during thyroid hormone treat-
ment. Molecular alterations were similar in patients with or without an event. No
treatment-related adverse events were reported.
CONCLUSIONS
In patients with low-risk thyroid cancer undergoing thyroidectomy, a follow-up strat-
egy that did not involve the use of radioiodine was noninferior to an ablation strat-
egy with radioiodine regarding the occurrence of functional, structural, and bio-
logic events at 3 years. (Funded by the French National Cancer Institute; ESTIMABL2
ClinicalTrials.gov number, NCT01837745.)
n engl j med 386;10 nejm.org March 10, 2022 923
The New England Journal of Medicine
Downloaded from nejm.org by DAN NICULESCU on March 10, 2022. For personal use only. No other uses without permission.
Copyright © 2022 Massachusetts Medical Society. All rights reserved.
T
he majority of patients with thy- with support from the French Ministry of Health
roid cancer are at low risk for recurrence through a grant from the National Cancer Insti-
(<5%),1-3 and their risk of cancer-related tute. All the patients provided written informed
death is even lower.4 After thyroidectomy, radio- consent. The trial was conducted in accordance
iodine (iodine-131) is generally administered both with the protocol (available with the full text of
A Quick Take
is available at to ablate residual normal thyroid tissue and to treat this article at NEJM.org), which was approved by a
NEJM.org persistent disease. Two large, randomized trials central ethics committee and by national authori-
have shown that in patients with low-risk thyroid ties. Data were gathered by all the authors and
cancer, the postoperative administration of low- were analyzed by the first and last authors. The
activity radioiodine (1.1 GBq) after injections of manuscript was reviewed by all the authors be-
recombinant human thyrotropin was noninferior fore submission for publication. All the authors
to the administration of high-activity radioiodine assume responsibility for the accuracy and com-
(3.7 GBq) after withdrawal of thyroid hormone pleteness of the data and for the fidelity of the
treatment with respect to the ablation success trial to the protocol.
rate at 1 year and the recurrence rate at 5 years.5-8
There is a consensus to avoid radioiodine ad- Patients
ministration in patients with a unifocal microcar- Patients were adults (≥18 years of age) with a
cinoma (≤10 mm in diameter), but the benefits of differentiated thyroid carcinoma (papillary, follic-
radioiodine administration in other patients with ular, or oncocytic [Hürthle-cell cancer]), with a
low-risk thyroid cancer remain controversial.4,9 In multifocal pT1a tumor (a diameter of each lesion
patients with pathological tumor–node–metasta- of ≤1 cm and a sum of the longest diameters of
sis (pTNM) stage 1 disease, retrospective studies the lesions of ≤2 cm) or a pT1b tumor (>1 cm
have shown inconsistent results regarding the use- and ≤2 cm); with both tumor stages, patients
fulness of radioiodine administration.10,11 The pTNM had a nodal status of N0 (no evidence of re-
staging system predicts the risk of thyroid cancer– gional node involvement) or Nx (regional lymph
related death and was used until recently for the nodes cannot be assessed in the absence of neck
decision regarding postoperative administration dissection), in the absence of extrathyroidal ex-
of radioiodine.12,13 The risk of recurrence was de- tension (Table S1 in the Supplementary Appen-
fined by the risk stratification recommended by the dix, available at NEJM.org). Patients who had ag-
American Thyroid Association.4 On the basis of gressive histologic subtypes (tall-cell, clear-cell,
this stratification, retrospective studies have not columnar-cell, and diffuse sclerosing variants of
shown benefits of postoperative radioiodine ad- papillary thyroid cancer, poorly differentiated)
ministration in patients with low-risk disease.14,15 were excluded from the trial.16
Nevertheless, the absence of prospective studies Two to five months before randomization, eli-
that address this question has been used as an ar- gible patients had undergone total thyroidecto-
gument in favor of recommending the use of radio- my with or without dissection of cervical lymph
iodine in all patients with low-risk thyroid cancer. nodes (neck dissection), according to ongoing
We conducted the randomized, phase 3 Essai protocols, with complete tumor resection. In ad-
Stimulation Ablation 2 (ESTIMABL2) trial involv- dition, all the patients had undergone postop-
ing patients with low-risk thyroid cancer to assess erative neck ultrasonography without the detec-
the noninferiority of a follow-up strategy that does tion of suspicious abnormalities.
not include the postoperative administration of
radioiodine as compared with radioiodine admin- Procedures
istration (ablation with 1.1 GBq after stimula-
Patients were randomly assigned to undergo post-
tion with recombinant human thyrotropin) withoperative administration of radioiodine (radio-
respect to the percentage of patients without an
iodine group) or not to undergo such adminis-
event during 3 years after randomization. tration (no-radioiodine group), with stratification
according to the trial site and lymph-node status
(N0 or Nx). In the radioiodine group, while pa-
Me thods
tients were receiving thyroid hormone treatment,
Trial Oversight 1.1 GBq (30 mCi) of radioiodine was adminis-
The ESTIMABL2 trial was performed within the tered 24 hours after the second intramuscular
Tumeurs de la Thyroïde Refractaires Network injection of recombinant human thyrotropin
924 n engl j med 386;10 nejm.org March 10, 2022

Thyroidectomy without R adioiodine in Low-Risk Cancer
(Thyrogen, Sanofi), which was given at a dose of lin level was defined as a value of more than 5 ng
0.9 mg on 2 consecutive days. Whole-body scan- per milliliter during thyroid hormone treatment
ning and single-photon-emission computed tomog- or a value of more than 2 ng per milliliter during
raphy (SPECT) of the neck were performed 2 to thyroid hormone treatment on two consecutive
5 days after radioiodine administration. measurements obtained 6 months apart. In the
The follow-up protocol was consistent with two groups, the detection of a thyroglobulin
the standard of care in France and consisted of antibody titer above the upper limit of the nor-
the measurement of thyroglobulin and thyro- mal range or an increase in the thyroglobulin
globulin antibodies in all patients at 10 months antibody titer by more than 50% between two
and yearly thereafter. Thyroglobulin was mea- measurements performed 6 months apart was
sured while the patient was receiving thyroid also considered to be part of the composite cri-
hormone treatment, except for the measurement teria. Serum samples for the measurement of
at 10 months after randomization in the radio- thyroglobulin and thyroglobulin antibodies that
iodine group, in whom the measurement was were obtained at randomization and at the
performed after stimulation with recombinant hu- 10-month and 3-year follow-up visits were as-
man thyrotropin. Ultrasonography of the neck was sessed both locally and in a central laboratory
performed in all patients 10 months and 3 years (Table S2).17
after randomization. No diagnostic radioiodine Secondary end points were quality of life,
scanning was performed after the whole-body anxiety, fear of recurrence, and dysfunction of
scanning that was performed after therapy. lacrimal and salivary glands. Questionnaires were
administered to all the patients at the time of
Primary and Secondary End Points randomization, after radioiodine administration
The primary objective was to assess noninferiority during hospitalization in the radioiodine group
in the no-radioiodine group as compared with or 2 months after randomization in the no-radio-
the radioiodine group with respect to the per- iodine group, and at 10 months and 3 years after
centage of patients without a functional, struc- randomization in the two groups.
tural, or biologic event during 3 years after ran-
domization. An event was a composite end point Prognostic Factors
that consisted of several criteria. Prognostic factors for an event were evaluated
In the radioiodine group only, events included with the use of univariate logistic regression. A
the presence of foci of radioiodine uptake out- post hoc analysis comparing the percentages of
side the thyroid bed on postablation whole-body patients with no evidence of disease (as defined
scanning or on SPECT (functional event) that by the 2015 guidelines of the American Thyroid
resulted in additional treatment (radioiodine ad- Association as an “excellent response” to treat-
ministration or surgery). In both groups, events ment) was performed in the two groups at 10
included abnormal findings on ultrasonography months and at 3 years. An excellent response at
of the neck (structural events), which were de- 10 months was defined as normal findings on
fined as a suspicious lymph node or thyroid neck ultrasonography with a thyroglobulin level
mass associated with abnormal cytologic find- of less than 1 ng per milliliter after the admin-
ings or with a serum thyroglobulin level in the istration of recombinant human thyrotropin (in
aspirate fluid of more than 10 ng per milliliter the radioiodine group) or during thyroid hor-
or an elevated level of thyroglobulin or thyro- mone treatment (in the no-radioiodine group) in
globulin antibodies (biologic events). In the ra- the absence of an elevated level of thyroglobulin
dioiodine group, in the absence of thyroglobulin antibodies. An excellent response at 3 years was
antibodies, an elevated thyroglobulin level was defined as normal findings on neck ultrasonog-
defined as a value of more than 5 ng per milli- raphy with a thyroglobulin level of less than 0.2 ng
liter (after the receipt of recombinant human per milliliter (in the radioiodine group) and a
thyrotropin or during thyroid hormone treat- level of less than 1 ng per milliliter (in the no-
ment) or a value of more than 1 ng per milliliter radioiodine group) during thyroid hormone treat-
during thyroid hormone treatment on two con- ment in the absence of an elevated level of thy-
secutive measurements obtained 6 months apart. roglobulin antibodies. Results were assessed after
In the no-radioiodine group, in the absence of review of findings from both local and central
thyroglobulin antibodies, an elevated thyroglobu- laboratories.4

Table 1. Characteristics of the Patients at Baseline.* iodine group would be noninferior to that in the
radioiodine group, with a between-group differ-
Radioiodine No Radioiodine ence of less than 5 percentage points at 3 years.
Characteristic (N = 389) (N = 387)
We calculated that the enrollment of 750 pa-
Age — yr 52.2±13.4 52.6±13.5 tients would provide the trial with 90% power to
Female sex — no. (%) 319 (82.0) 323 (83.5) determine the noninferiority margin, assuming
Interval between thyroidectomy 92.1±23.2 91.2±23.6 that 95% of the patients in the radioiodine group
and randomization — days would not have an event at 3 years and including
Histologic analysis — no. (%) the potential loss to follow-up; a P value of less
Papillary 372 (95.6) 372 (96.1) than 0.05 was considered to indicate statistical
Follicular 13 (3.3) 11 (2.8)
significance.
Descriptive quantitative data were expressed
Oncocytic: Hürthle cell 4 (1.0) 4 (1.0)
as means and standard deviations; qualitative
Largest tumor dimension — mm 13.4±3.6 13.7±3.9 data were expressed as percentages and 95%
Primary tumor–node–metastasis confidence intervals. We calculated the differ-
stage — no. (%)
ence in the observed percentages of patients
pT1aN0 26 (6.7) 23 (5.9) without an event and its 95% one-sided confi-
pT1aNx 56 (14.4) 42 (10.9) dence interval, which was equivalent to a two-
pT1bN0 143 (36.8) 148 (38.2) sided 90% confidence interval, because the trial
pT1bNx 164 (42.2) 174 (45.0) was designed to use one-sided hypothesis test-
Multifocality — no. (%) 178 (45.8) 156 (40.3)
ing at an alpha level of 0.05. We used logistic
regression to perform an analysis of the primary
Neck dissection performed 171 (44.0) 171 (44.2)
— no. (%)† end point after adjustment for stratification fac-
tors as a test of robustness.
Central only 69 (17.7) 77 (19.9)
The primary analysis included all the patients
Central and lateral 74 (19.0) 65 (16.8)
who could be evaluated in the per-protocol
Lateral only 26 (6.7) 27 (7.0) population (i.e., all the patients whose treatment
Unknown 2 (0.5) 2 (0.5) and 3-year follow-up had adhered to the study
protocol); patients who had been followed for
* Plus–minus values are means ±SD. The percentages may not total 100 because
of rounding. fewer than 3 years were excluded from the pri-
† Neck dissection was performed according to ongoing protocols at each trial mary population. The percentage of patients
center. without an event during the trial period was
calculated without consideration for the timing
Molecular Analysis of the event during that period. A sensitivity
The objective was to describe the type and num- analysis was performed in the intention-to-treat
ber of molecular alterations, according to the population that included the results for all the
occurrence of an event (Table S2). For this pur- patients until their last participation in the trial
pose, we designed a nested case–control study in in order to consider those who could not be
which cases were patients with an event and evaluated at 3 years. We report the between-
controls were patients without an event, regard- group difference in event-free survival at 3 years
less of which randomized treatment had been using a time-to-event analysis and its two-sided
received. Each case patient was paired with two 90% confidence interval.
controls on the basis of five criteria: pT classifi- For secondary end points, results are reported
cation (pT1a or pT1b), histologic results (papil- as point estimates and 95% confidence intervals,
lary or follicular), sex, age group (≤55 years or since the statistical analysis plan did not include
>55 years), neck dissection (yes or no), and treat- a provision for correcting for multiple compari-
ment group (radioiodine or no radioiodine). sons. Thus, the widths of the confidence inter-
vals have not been adjusted for multiplicity, so
Statistical Analysis they should not be used to infer definitive treat-
We designed our trial as a noninferiority study ment effects. All the analyses were performed
to answer the question of whether the percent- with the use of SAS software, version 9.4 (SAS
age of patients without an event in the no-radio- Institute).

776 Patients underwent randomization
389 Were assigned to receive 387 Were assigned to follow-up only,

postoperative radioiodine without postoperative radioiodine
20 Were excluded
26 Were excluded
1 Was included in error
11 Were lost to follow-up
19 Could not be evaluated 3 yr
3 Died (unrelated to
after randomization
thyroid cancer)
11 Were lost to follow-up
12 Withdrew consent
2 Died (unrelated to
thyroid cancer)
6 Withdrew consent
363 Were included in the per-protocol 367 Were included in the per-protocol
analysis analysis
Figure 1. Enrollment and Follow-up.

The five listed deaths that occurred during the trial period were from pulmonary embolism, sarcomatoid lung cancer,
and aneurysm rupture in the radioiodine group and heart failure and peritoneal cancer in the no-radioiodine group.
R e sult s Events consisted of functional or structural ab-

normalities in 8 patients and biologic abnor-
Trial Population malities with respect to levels of thyroglobulin
From May 2013 through March 2017, a total of or thyroglobulin antibodies in 23 patients with
776 patients underwent randomization at 35 cen- 25 events (Table 2).
ters in France (Table 1 and Table S3). The mean Overall, subsequent treatments (surgery, radio-
age of the patients was 52 years, and 83% were iodine administration, or both) were performed
women. Histologic analyses revealed mostly in 4 patients in the no-radioiodine group and in
papillary tumors (95.9%), and pTN stages were 10 in the radioiodine group. Other patients were
mostly pT1b N0 or Nx (81.1%). One patient did followed without additional treatment. Adverse
not meet the inclusion criteria, and 45 patients events occurred in 30 patients, and no adverse
(5.8%) could not be evaluated at 3 years after events were determined by the investigators to
randomization (Fig. 1). be related to treatment (Table S6). There were no
thyroid-related deaths.
Primary End Point
Among the 730 patients who could be evaluated Quality-of-Life Scores
at 3 years, the percentage without an event was In the two groups, the patients had similar scores
95.6% (95% confidence interval [CI], 93.0 to 97.5) regarding quality of life, anxiety, distress, and
in the no-radioiodine group and 95.9% (95% CI, fear of recurrence (Table S7 and Fig. S1). The
93.3 to 97.7) in the radioiodine group, for a frequency of salivary or lacrimal dysfunction
between-group difference of −0.3 percentage was also similar in the two groups at all time
points (two-sided 90% CI, −2.7 to 2.2), which points, except for lacrimal discomfort, which
met the noninferiority cutoff for the nonuse of was more common in the radioiodine group at
radioiodine. Similar results were obtained in the 2 months after randomization (Table S8).
intention-to-treat population and after adjustment
for stratification factors (Tables S4 and S5). Prognostic Factors for Event Occurrence
Primary events occurred in 16 of 367 patients A higher risk of events was observed in patients
(4.4%) in the no-radioiodine group and in 15 of with a tumor size of less than 14 mm and in pa-
363 patients (4.1%) in the radioiodine group. tients with a postoperative serum thyroglobulin

Table 2. Number and Type of Events Occurring during 3 Years (Per-Protocol Population).*
Radioiodine No Radioiodine Between-Group

Variable (N = 363) (N = 367) Difference†
percentage points
no. % (95% CI) no. % (95% CI) (90% CI)
Primary composite end point
No primary event during 3 yr 348 95.9 351 95.6 −0.3
(93.3 to 97.7) (93.0 to 97.5) (−2.7 to 2.2)
Occurrence of event during 3 yr 15 4.1 16 4.4
(2.3 to 6.7) (2.5 to 7.0)
Functional event
Foci of radioiodine uptake outside thyroid bed 3‡ NA
resulting in further treatment
Structural event
Abnormal results on neck ultrasonography with
abnormal cytologic or elevated thyro
globulin findings in aspirate fluid
Abnormal lymph node 2§ 3¶
Abnormal thyroid mass 0 0
Biologic event
Thyroglobulin level after recombinant human
thyrotropin
>5 ng per ml 3‖ NA
>1 ng to ≤5 ng per ml on 2 consecutive 6‖ NA
measurements
Thyroglobulin level during THT
>5 ng per ml 0 3**
>1 ng to ≤5 ng per ml on 2 consecutive 0 NA
measurements
>2 ng to ≤5 ng per ml on 2 consecutive NA 4**
measurements
Thyroglobulin antibodies
Appearance of elevated levels 2 5
Increase in titer during trial 0 2
* The primary objective was to assess noninferiority in the no-radioiodine group as compared with the radioiodine
group with respect to the absence of a 3-year composite end point, which was defined as the presence of abnormal
foci of radioiodine uptake on whole-body scanning that required subsequent treatment (in the radioiodine group), ab-
normal findings on neck ultrasonography, or elevated levels of thyroglobulin or thyroglobulin antibodies. The widths
of the confidence intervals have not been adjusted for multiplicity, so the intervals should not be used to infer defini-
tive treatment effects for secondary outcomes. NA denotes not applicable, and THT thyroid hormone treatment.
† A two-sided 90% confidence interval (equivalent to a 95% one-sided confidence interval) is reported for the between-
group difference in the primary end point because the trial was designed to use one-sided hypothesis testing at an
alpha level of 0.05.
‡ The 3 patients in this category had postoperative thyroglobulin levels of 0.2 ng, 0.3 ng, and 1.33 ng per milliliter while
receiving THT.
§ The 2 patients in this category had postoperative thyroglobulin levels of 0.15 ng and 0.04 ng per milliliter while receiv-
ing THT.
¶ The 3 patients in this category had postoperative thyroglobulin levels of 0.7 ng, 0.9 ng, and 1.7 ng per milliliter while
receiving THT.
‖ This category includes 1 patient who had two events at different time points: an elevated thyroglobulin level after the
administration of human recombinant thyrotropin (>1 ng to ≤5 ng per milliliter) on two consecutive measurements,
which was followed by an increased level (>5 ng per milliliter).
** This category includes 1 patient who had two events at different time points: an elevated thyroglobulin level during THT
(>2 ng to ≤5 ng per milliliter) on two consecutive measurements, which was followed by an increased level (>5 ng per
milliliter).

level that was higher than the two cutoff values Table 3. Prognostic Factors Associated with a Primary Event in 730 Patients
we evaluated (0.5 ng and 1 ng per milliliter) Who Could Be Evaluated during 3 Years (Univariate Analysis).*
while receiving thyroid hormone therapy in the
absence of an elevated level of thyroglobulin Variable Patients with Event Odds Ratio (95% CI)†
antibodies (Table 3 and Table S9). no./total no. (%)
Age
Molecular Data
≤55 yr 14/402 (3.5) Reference
In the nested case–control study (in which cases
were patients with an event and controls were >55 yr 17/328 (5.2) 1.5 (0.7–3.1)
patients without an event), we performed mo- Sex
lecular analysis on 100 tumor samples obtained Female 27/604 (4.5) Reference
from 96 patients, among which 90 samples Male 4/126 (3.2) 0.7 (0.2–2.0)
could be analyzed (Table S10). Overall, 50 tu- Histologic analysis
mors had a BRAF mutation, 14 had a RAS muta-
Papillary 30/698 (4.3) Reference
tion, and 6 had an oncogenic fusion (Table S11).
The frequency of BRAF mutations was not mate- Follicular 1/24 (4.2) 1.0 (0.1–7.4)
rially different between cases (61.5%) and con- Oncocytic: Hürthle cell 0/8 NE
trols (53.1%). A total of 17 tumors did not have Largest median tumor size
any genetic abnormalities, including 4 samples ≤14 mm 23/407 (5.7) Reference
with noncontributive ribonucleic acid sequenc- >14 mm 8/322 (2.5) 0.4 (0.2–0.96)
ing. No TERT promoter mutations were detected.
Nodal status
Measures of Excellent Response N0 9/320 (2.8) Reference
Patients in the two groups had similar rates of Nx 22/410 (5.4) 2.0 (0.9–2.8)
excellent response, as defined by the American Multifocality
Thyroid Association, with respect to thyroglobu- No 15/411 (3.6) Reference
lin levels that were measured at a central labora- Yes 16/319 (5.0) 1.4 (0.7–2.8)
tory at both 10 months and 3 years (Table 4). Thyroglobulin level‡
Results regarding thyroglobublin levels as deter-
Higher cutoff value
mined by local laboratories are provided in Table
S12. ≤1 ng/ml 16/507 (3.2) Reference
>1 ng/ml 7/48 (14.6) 5.2 (2.0–13.5)
Lower cutoff value
Discussion
≤0.5 ng/ml 12/426 (2.8) Reference
Our randomized trial addressed the usefulness >0.5 ng/ml 11/129 (8.5) 3.2 (1.4–7.5)
of follow-up without postoperative administra-
tion of radioiodine in patients with low-risk * NE denotes not evaluated.
thyroid cancer undergoing thyroidectomy. We † The widths of the confidence intervals have not been adjusted for multiplicity,
so the intervals should not be used to infer definitive treatment effects within
observed that less than 5% of the patients in the subgroups. No multivariate analysis was performed.
two groups had events that included abnormal ‡ The listed thyroglobulin level was evaluated during THT in 555 patients without
findings on whole-body scanning or neck ultra- an elevated level of thyroglobulin antibodies at randomization. Two thresholds
of thyroglobulin level were tested.
sonography or elevated levels of thyroglobulin or
thyroglobulin antibodies during the first 3 years
of follow-up. This rate is concordant with the and 65 to 75% of those with a nodal status of
definition of low-risk thyroid cancer, and our N0 or Nx.3,18-20
trial showed that the risk of events was not Radioiodine is administered postoperatively
higher in the absence of postoperative adminis- to destroy remaining normal thyroid tissue, to
tration of radioiodine.4 The patients in our trial facilitate follow-up, and to diagnose and eventu-
were representative of patients with a pT1 thy- ally treat microscopic disease and therefore in-
roid tumor measuring 20 mm or less, a category crease recurrence-free survival. In accordance
that constitutes 50 to 70% of all thyroid cancers with these goals, we defined events as structural

Table 4. Response to Treatment at 10 Months and 3 Years.
Response Radioiodine No Radioiodine

Evaluation at 10 mo
No. of patients 325 342
Distribution of response — no. (%)
Excellent* 282 (86.8) 295 (86.3)
Structurally incomplete: abnormal imaging 4 (1.2) 1 (0.3)
Biochemically incomplete† 8 (2.5) 13 (3.8)
Indeterminate‡ 31 (9.5) 33 (9.6)
Evaluation at 3 yr
No. of patients 363 367
Distribution of response — no. (%)
Excellent§ 265 (73.0) 272 (74.1)
Structurally incomplete: abnormal imaging 4 (1.1) 3 (0.8)
Biochemically incomplete¶ 2 (0.6) 12 (3.3)
Indeterminate‖ 40 (11.0) 37 (10.1)
Missing data for central laboratory assessment 52 (14.6) 43 (11.7)
* Listed are the results of a post hoc analysis comparing the percentages of patients with an excellent response (as
defined by the 2015 guidelines of the American Thyroid Association) according to central laboratory measurements.
The definition of an excellent response to treatment at 10 months was normal findings on neck ultrasonography and
a thyroglobulin level of less than 1 ng per milliliter after the administration of recombinant human thyrotropin in the
radioiodine group or while patients were receiving THT in the no-radioiodine group in the absence of an elevated level
of thyroglobulin antibodies.
† A biochemically incomplete response at 10 months was defined as normal results on neck ultrasonography and a
thyroglobulin level of 10 ng per milliliter or more after the administration of recombinant human thyrotropin in the
radioiodine group or a level of 2 ng per milliliter or more during THT in the no-radioiodine group in the absence of
an elevated level of thyroglobulin antibodies or an increasing level over time.
‡ The criteria for an indeterminate response at 10 months were the same as those for a biochemically incomplete re-
sponse except that the thyroglobulin cutoff was 1 ng to less than 10 ng per milliliter in the radioiodine group and 1 ng
to less than 2 ng per milliliter in the no-radiodine group; or the presence of elevated thyroglobulin antibodies with a
decreasing level over time.
§ An excellent response at 3 years was defined as normal findings on neck ultrasonography and a thyroglobulin level of
less than 0.2 ng per milliliter in the radioiodine group and a level of less than 1 ng per milliliter in the no-radioiodine
group during THT in the absence of an elevated level of thyroglobulin antibodies.
¶ A biochemically incomplete response at 3 years was defined as normal findings on neck ultrasonography and a thyro-
globulin level of 1 ng per milliliter in the radioiodine group and a level of 2 ng per milliliter in the no-radioiodine group
during THT in the absence of an elevated level of thyroglobulin antibodies; or the presence of an elevated level of thyro-
globulin antibodies with an increasing level over time.
‖ The criteria for an indeterminate response at 3 years were the same as those for an indeterminate response at 10 months
except that the thyroglobulin cutoff was 0.2 ng to less than 1 ng per milliliter in the radioiodine group and a level of 1 ng
to less than 2 ng per milliliter in the no-radioiodine group.
or functional, with the need for further treat- Thresholds for abnormal thyroglobulin levels
ments (surgery or radioiodine administration), were more than 5 ng per milliliter after the ad-
or biologic, defined as an elevated level of thyro- ministration of recombinant human thyrotropin
globulin or thyroglobulin antibodies. The mea- or more than 1 ng per milliliter in the radioio-
surement of thyroglobulin antibodies was includ- dine group and more than 2 ng per milliliter in
ed in this composite definition of events, even the no-radioiodine group while the patients were
though the clinical relevance is questionable. receiving thyroid hormone therapy. These levels
The appearance of elevated levels of thyroglobu- are consonant with the standard of care at the
lin antibodies or an increase in the titer over initiation of the trial, with 5% of patients having
time is not always related to structural recur- a thyroglobulin level between 1 ng and 2 ng per
rence but frequently leads to further diagnostic milliliter while receiving thyroid hormone ther-
examination or administration of radioiodine. apy in the absence of recurrence after total thy-

roidectomy without radioiodine.17,21 A post hoc In our trial, a postoperative thyroglobulin

analysis that we performed using thyroglobulin level that was higher than the two cutoff values
thresholds recommended by the 2015 American we evaluated (0.5 ng and 1 ng per milliliter) in
Thyroid Association guidelines (0.2 ng per milli- patients who were undergoing thyroid hormone
liter in patients treated with radioiodine and 1 ng therapy was associated with an increased risk of
per milliliter in the absence of radioiodine) an event, which suggests that elevated postop-
showed a similar percentage of patients with an erative levels should be an indication for closer
excellent response in the two trial groups. follow-up. In our trial, a tumor size of less than
In addition, the patients in the two groups 14 mm was predictive of an event; this unex-
had similar scores with respect to quality of life, pected finding could be related to multifocality,
anxiety, distress, and fear of recurrence, find- which was present at a high rate in the trial
ings that reflect the absence of consequences in patients.
patients who did not receive postoperative radio- Our trial has certain limitations. Our results
iodine. Also, lacrimal discomfort was more fre- should be confirmed with longer follow-up. How-
quent at 2 months in the radioiodine group.6 ever, in previous retrospective studies involving
Among the trial patients, genomic abnormali- patients with low-risk thyroid cancer, most re-
ties were similar to those listed in the thyroid currences occurred during the first 5 years of
data set of the Cancer Genome Atlas, with the follow-up.27,28 The 3-year period in our trial al-
BRAF V600E mutation present in more than half lowed for control of indeterminate findings over
the patients.22 Although the presence of the tall- time while limiting loss to follow-up.
cell variant of papillary thyroid cancer was an In this randomized trial involving patients with
exclusion criterion, this variant was detected in low-risk differentiated thyroid cancer, we found
25% of the samples that underwent central labo- that follow-up without the use of radioiodine after
ratory review, which explains the high incidence thyroidectomy was noninferior to the administra-
of BRAF mutations and underlines the impor- tion of 1.1 GBq of radioiodine after the adminis-
tance of central histologic review to detect these tration of recombinant human thyrotropin.
pathological subtypes.23 We did not find any Supported by the French Ministry of Health through a grant
from the National Cancer Institute.
TERT promoter mutations.24 The presence of the Disclosure forms provided by the authors are available with
BRAF V600E mutation is correlated with a more the full text of this article at NEJM.org.
advanced pTNM stage.25,26 In retrospective stud- A data sharing statement provided by the authors is available
with the full text of this article at NEJM.org.
ies, the presence of this mutation has been asso- We thank the patients for their participation in the trial;
ciated with a poor outcome, but this was not the Sylviane Iacobelli, Anne-Marine Lenzotti, and Nathalie Bouvet-
Forteau for data management; Thibaud Motreff for serving as
case in a study involving patients with N1 micro- managing sponsor of the clinical research assistants in the
carcinoma.26 On the basis of our findings, it trial; Muriel Wartelle and Tasnim Hedoui for informatics;
appears that in patients with low-risk disease, Caroline Pradon and Lea Ciniello for centralized thyroglobu-
lin measurements; Nathalie Bourcigaux, Lise Criniere, Ciprian
the presence of a BRAF mutation would not be Draganescu, and Patrice Rodien for their participation; and all
an indication for radioiodine administration. the clinical research associates for data collection.
Appendix
The authors’ full names and academic degrees are as follows: Sophie Leboulleux, M.D., Ph.D., Claire Bournaud, M.D., Cecile N. Chougnet,
M.D., Slimane Zerdoud, M.D., Abir Al Ghuzlan, M.D., Bogdan Catargi, M.D., Christine Do Cao, M.D., Antony Kelly, M.D., Marie‑Luce
Barge, M.D., Ludovic Lacroix, Pharm.D., Ph.D., Inna Dygai, M.D., Pierre Vera, M.D., Ph.D., Daniela Rusu, M.D., Olivier Schneegans,
M.D., Danielle Benisvy, M.D., Marc Klein, M.D., Julie Roux, M.D., Marie‑Claude Eberle, M.D., Delphine Bastie, M.D., Camila Nasci-
mento, M.D., Anne‑Laure Giraudet, M.D., Nathalie Le Moullec, M.D., Stéphane Bardet, M.D., Delphine Drui, M.D., Nathalie Roudaut,
M.D., Yann Godbert, M.D., Olivier Morel, M.D., Anne Drutel, M.D., Livia Lamartina, M.D., Ph.D., Claire Schvartz, M.D., Fritz‑Line
Velayoudom, M.D., Martin‑Jean Schlumberger, M.D., Laurence Leenhardt, M.D., Ph.D., and Isabelle Borget, Pharm.D., Ph.D.
The authors’ affiliations are as follows: the Department of Nuclear Medicine and Endocrine Oncology (S.L., L. Lamartina, M.-J.S.),
the Department of Medical Biology and Pathology (A.A.G., L. Lacroix), and the Biostatistics and Epidemiology Office, Oncostat,
INSERM Unité 1018 (I.B.), Gustave Roussy and Université Paris-Saclay, Villejuif, the Nuclear Medicine Department, Hospices Civils de
Lyon, Groupement Hospitalier Est, Bron (C.B.), the Endocrine Oncology Department, Assistance Publique–Hôpitaux de Paris (AP–HP)
Hôpital Saint-Louis (C.N.C.), and the Thyroid and Endocrine Tumors Unit, Pitié–Salpétrière Hospital AP–HP, Institute of Cancer IUC
Sorbonne University (L. Leenhardt), Paris, the Department of Medical Imaging, Nuclear Medicine, IUCT Oncopole Toulouse–Institut
Claudius Regaud (S.Z.), and the Nuclear Medicine Department, CHU Rangueil (D. Bastie), Toulouse, the Department of Endocrinology–
Metabolic Diseases, Hôpital Saint-André, Centre Hospitalier Universitaire (CHU) de Bordeaux (B.C.), and the Thyroid Oncology and
Nuclear Medicine Department, Institut Bergonié (Y.G.), Bordeaux, the Endocrine Department, CHRU de Lille–Hôpital Claude Huriez,
Lille (C.D.C.), the Nuclear Medicine Department, Centre Jean Perrin, Clermont-Ferrand (A.K.), the Nuclear Medicine Department,
Centre Eugene Marquis, Rennes (M.-L.B.), the Nuclear Medicine Department, Centre Georges François Leclerc, Dijon (I.D.), the Nuclear

Medicine Department, Centre Henri Becquerel and Laboratoire QUANTif, Rouen (P.V.), the Nuclear Medicine Department, Centre René
Gauducheau, Saint Herblain (D.R.), the Nuclear Medicine Department, Centre Paul Strauss, Strasbourg (O.S.), the Nuclear Medicine
Department, Antoine Lacassagne, Nice (D. Benisvy), the Endocrine Department, Centre Hospitalier Régional Universitaire (CHRU) de
Nancy, Hôpitaux de Brabois, Vandoeuvre Les Nancy (M.K.), the Nuclear Medicine Department, CHU Grenoble–Alpes, Grenoble (J.R.),
the Nuclear Medicine Department, Institut du Cancer de Montpellier, Institut Régional du Cancer Val d’Aurelle, Montpellier (M.-C.E.),
the Nuclear Medicine Department, Institut Curie Site Saint-Cloud, Saint-Cloud (C.N.), the Nuclear Medicine Department, Centre Léon
Bérard, Lyon (A.-L.G.), the Endocrine Department, CHU Saint Pierre, Saint Pierre (N.L.M.), the Nuclear Medicine Department and
Thyroid Unit, Centre François Baclesse, Caen (S.B.), the Endocrine Department, Institut du Thorax, CHU de Nantes–Hopital Laennec
Saint-Herblain, Nantes (D.D.), the Endocrine Department, CHU La Cavale Blanche, Brest (N.R.), the Nuclear Medicine Department,
Institut de Cancérologie de l’Ouest, Angers (O.M.), the Endocrine Department, CHU Dupuytren, Limoges (A.D.), the Thyroid Unit,
Institut Jean Godinot, Reims (C.S.), and the Endocrine Department, CHU de Guadeloupe, Hôpital Ricou, Les Abymes (F.-L.V.) — all in
France.
References
1. Vaccarella S, Franceschi S, Bray F, decade. Eur J Nucl Med Mol Imaging 2020; work. Progression of incidence and esti-
Wild CP, Plummer M, Dal Maso L. World- 47:78-83. mate of net survival from papillary thy-
wide thyroid-cancer epidemic? The in- 10. Sawka AM, Brierley JD, Tsang RW, roid cancers diagnosed between 2008 and
creasing impact of overdiagnosis. N Engl et al. An updated systematic review and 2016 in France. Ann Endocrinol (Paris)
J Med 2016;375:614-7. commentary examining the effectiveness 2020;81:530-8.
2. Lim H, Devesa SS, Sosa JA, Check D, of radioactive iodine remnant ablation in 19. Ganly I, Nixon IJ, Wang LY, et al. Sur-
Kitahara CM. Trends in thyroid cancer well-differentiated thyroid cancer. Endo- vival from differentiated thyroid cancer:
incidence and mortality in the United crinol Metab Clin North Am 2008;37:457- what has age got to do with it? Thyroid
States, 1974-2013. JAMA 2017;317:1338- 80. 2015;25:1106-14.
48. 11. Carhill AA, Litofsky DR, Ross DS, et al. 20. Davies L, Welch HG. Current thyroid
3. Lamartina L, Durante C, Lucisano G, Long-term outcomes following therapy in cancer trends in the United States. JAMA
et al. Are evidence-based guidelines re- differentiated thyroid carcinoma: NTCTCS Otolaryngol Head Neck Surg 2014;140:
flected in clinical practice? An analysis of registry analysis 1987-2012. J Clin Endo- 317-22.
prospectively collected data of the Italian crinol Metab 2015;100:3270-9. 21. Durante C, Montesano T, Attard M,
Thyroid Cancer Observatory. Thyroid 2017; 12. Pacini F, Schlumberger M, Dralle H, et al. Long-term surveillance of papillary
27:1490-7. Elisei R, Smit JW, Wiersinga W. European thyroid cancer patients who do not un-
4. Haugen BR, Alexander EK, Bible KC, consensus for the management of pa- dergo postoperative radioiodine remnant
et al. 2015 American Thyroid Association tients with differentiated thyroid carci- ablation: is there a role for serum thyro-
management guidelines for adult patients noma of the follicular epithelium. Eur J globulin measurement? J Clin Endocrinol
with thyroid nodules and differentiated Endocrinol 2006;154:787-803. Metab 2012;97:2748-53.
thyroid cancer: the American Thyroid As- 13. The American Thyroid Association 22. Giordano TJ. The cancer genome atlas
sociation Guidelines Task Force on Thy- (ATA) Guidelines Taskforce on Thyroid research network: a sight to behold. En-
roid Nodules and Differentiated Thyroid Nodules and Differentiated Thyroid Can- docr Pathol 2014;25:362-5.
Cancer. Thyroid 2016;26:1-133. cer. Revised American Thyroid Associa- 23. Hescot S, Sheikh-Alard H, Kordahi M,
5. Mallick U, Harmer C, Yap B, et al. Ab- tion management guidelines for patients et al. Impact of expert review of histologi-
lation with low-dose radioiodine and thy- with thyroid nodules and differentiated cal diagnosis of papillary and follicular
rotropin alfa in thyroid cancer. N Engl J thyroid cancer. Thyroid 2009;19:1167-214. thyroid cancer. Endocrine 2021;72:791-7.
Med 2012;366:1674-85. 14. Hay ID, Kaggal S, Iniguez-Ariza NM, 24. Xu B, Tuttle RM, Sabra MM, Ganly I,
6. Schlumberger M, Catargi B, Borget I, Reinalda MS, Wiseman GA, Thompson GB. Ghossein R. Primary thyroid carcinoma
et al. Strategies of radioiodine ablation Inability of radioiodine remnant ablation with low-risk histology and distant metas-
in patients with low-risk thyroid cancer. to improve postoperative outcome in adult tases: clinicopathologic and molecular
N Engl J Med 2012;366:1663-73. patients with low-risk papillary thyroid characteristics. Thyroid 2017;27:632-40.
7. Schlumberger M, Leboulleux S, Cat- carcinoma. Mayo Clin Proc 2021;96:1727- 25. Xing M, Alzahrani AS, Carson KA, et al.
argi B, et al. Outcome after ablation in 45. Association between BRAF V600E muta-
patients with low-risk thyroid cancer 15. Lamartina L, Durante C, Filetti S, tion and recurrence of papillary thyroid
(ESTIMABL1): 5-year follow-up results of Cooper DS. Low-risk differentiated thy- cancer. J Clin Oncol 2015;33:42-50.
a randomised, phase 3, equivalence trial. roid cancer and radioiodine remnant ab- 26. Perera D, Ghossein R, Camacho N, et al.
Lancet Diabetes Endocrinol 2018;6:618-26. lation: a systematic review of the litera- Genomic and transcriptomic characteriza-
8. Dehbi H-M, Mallick U, Wadsley J, ture. J Clin Endocrinol Metab 2015;100: tion of papillary microcarcinomas with
Newbold K, Harmer C, Hackshaw A. Re- 1748-61. lateral neck lymph node metastases. J Clin
currence after low-dose radioiodine abla- 16. Brierley JD, Gospodarowicz MK, Witte Endocrinol Metab 2019;104:4889-99.
tion and recombinant human thyroid- kind C, eds. TNM classification of malig- 27. Durante C, Montesano T, Torlontano
stimulating hormone for differentiated nant tumours. 8th ed. Oxford, United M, et al. Papillary thyroid cancer: time
thyroid cancer (HiLo): long-term results Kingdom:Wiley Blackwell, 2017. course of recurrences during postsurgery
of an open-label, non-inferiority random 17. Nascimento C, Borget I, Troalen F, et al. surveillance. J Clin Endocrinol Metab 2013;
ised controlled trial. Lancet Diabetes En- Ultrasensitive serum thyroglobulin mea- 98:636-42.
docrinol 2019;7:44-51. surement is useful for the follow-up of 28. Dong W, Horiuchi K, Tokumitsu H,
9. Verburg FA, Flux G, Giovanella L, van patients treated with total thyroidectomy et al. Time-varying pattern of mortality
Nostrand D, Muylle K, Luster M. Differen- without radioactive iodine ablation. Eur J and recurrence from papillary thyroid
tiated thyroid cancer patients potentially Endocrinol 2013;169:689-93. cancer: lessons from a long-term follow-
benefitting from postoperative I-131 ther- 18. Colonna M, Borson-Chazot F, Delafosse up. Thyroid 2019;29:802-8.
apy: a review of the literature of the past P, Schvartz C, Guizard AV, FRANCIM net- Copyright © 2022 Massachusetts Medical Society.


NEJM Iod Radioactiv

Uploaded by

Copyright:

Available Formats

NEJM Iod Radioactiv

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NEJM Iod Radioactiv

Uploaded by

Copyright:

Available Formats

The n e w e ng l a n d j o u r na l of m e dic i n e

Thyroidectomy without Radioiodine

924 n engl j med 386;10 nejm.org March 10, 2022

The New England Journal of Medicine

n engl j med 386;10 nejm.org March 10, 2022 925

926 n engl j med 386;10 nejm.org March 10, 2022

The New England Journal of Medicine

776 Patients underwent randomization

389 Were assigned to receive 387 Were assigned to follow-up only,

Figure 1. Enrollment and Follow-up.

R e sult s Events consisted of functional or structural ab-

n engl j med 386;10 nejm.org March 10, 2022 927

Radioiodine No Radioiodine Between-Group

928 n engl j med 386;10 nejm.org March 10, 2022

The New England Journal of Medicine

Measures of Excellent Response N0 9/320 (2.8) Reference

n engl j med 386;10 nejm.org March 10, 2022 929

Table 4. Response to Treatment at 10 Months and 3 Years.

Response Radioiodine No Radioiodine

930 n engl j med 386;10 nejm.org March 10, 2022

The New England Journal of Medicine

roidectomy without radioiodine.17,21 A post hoc In our trial, a postoperative thyroglobulin

n engl j med 386;10 nejm.org March 10, 2022 931

932 n engl j med 386;10 nejm.org March 10, 2022

The New England Journal of Medicine

You might also like