The primary endpoint is newly identified structural disease,1 which may include persistent, recurrent, or distant metastatic disease defined as follows.
1) Persistent disease would be defined as cervical/superior mediastinal disease identified by an imaging study (US, radioisotope scan, computed tomography [CT], magnetic resonance imaging [MRI, positron emission tomography [PET], etc.), which would ideally be biopsy proven. Direct measurements of recombinant human thyrotropin (rhTSH)-stimulated (Thyrogen®, Genzyme, Cambridge, MA) pre–radioactive iodine treatment Tg levels (and antibodies) would aid in identifying persistent structural disease. Persistent disease is defined as being diagnosed ≤6 months after total thyroidectomy and radioactive iodine treatment.
2) Recurrent/distant metastatic disease would be defined as occurring >6 months after total thyroidectomy and radioactive iodine treatment.
a. Locoregional recurrence is defined as cervical/superior mediastinal disease identified by an imaging study (US, radioisotope scan, CT, MRI, PET, etc.) and ideally biopsy proven. Direct measurements of rhTSH-stimulated or unstimulated Tg levels (and antibodies) would aid in identifying recurrent structural disease.
b. Distant metastatic disease is defined as disease identified outside of the neck by an imaging study (US, radioisotope scan, CT, MRI, PET, etc) or biopsy proven. Direct measurements of rhTSH-stimulated or unstimulated Tg levels (and antibodies) would aid in identifying distant metastatic disease.
The secondary endpoints are
1) Disease-specific mortality
2) Surgical mortality and morbidity
a. Operative mortality (defined as within 30 days postoperatively). General operative morbidity such as pulmonary embolism, pneumonia, cardiac complications, and deep venous thrombosis would be recorded.
b. Specific morbidities
i. Permanent hypoparathyroidism (transient hypoparathyroidism would be monitored but not considered a morbidity endpoint)
ii. Transient recurrent laryngeal nerve paresis (documented paresis that recovers within 6 months postoperatively)
iii. Permanent recurrent laryngeal nerve paresis (documented paresis that extends beyond 6 months postoperatively)
iv. Cervical neck hematoma
3) Other outcomes
a. Presence of undetectable suppressed Tg at 1-year follow-up.
b. Presence of undetectable stimulated Tg at 1-year follow-up.
c. Biochemical persistent disease defined based on elevated Tg levels is more likely to represent persistent disease rather than residual postoperative normal thyroid tissue. Persistent disease is defined as being diagnosed ≤6 months after total thyroidectomy and radioactive iodine treatment.1
d. Biochemical recurrent disease which is defined based on elevated Tg levels in a patient with a previously undetectable Tg diagnosed >6 months after total thyroidectomy and radioactive iodine treatment.1
e. Requirement of cervical reoperation for thyroid cancer after initial total thyroidectomy and radioactive iodine treatment.
f. Requirement of radioactive iodine treatment for thyroid cancer after initial total thyroidectomy and radioactive iodine treatment.
g. Operative time
h. Duration of hospital stay
i. Number of frozen section analyses during surgery
j. Molecular marker status
k. Quality of life assessment
Radioactive iodine treatment
All patients would undergo postoperative radioactive iodine treatment. The dose would be standardized to approximately 50 mCi for all patients in both group A and B, independent of the underlying stage of PTC. A pretherapy whole-body iodine scan would not be obtained since it would not affect the decision to treat or the activity of radioactive iodine that is administered (2). The treatment would ideally be performed at the participating institutions, but treatment at another institution/hospital would not be in violation of the protocol. All patients would be on a pretherapy low-iodine diet and the treatment would be performed after rhTSH stimulation. Posttherapy whole-body iodine scanning would be performed in all patients in accordance with ATA guidelines (2).
Surveillance and identification of newly identified structural disease, and biochemical persistent or recurrent disease
In this protocol, all patients would undergo surveillance for recurrence as well as TSH suppression, in accordance with ATA guidelines (2). The diagnosis of newly identified structural disease and biochemical persistent or recurrent disease is as outlined in the endpoints section. The time points for surveillance are as outlined in Table 2.
Definition of surgical complications
Permanent hypoparathyroidism would be defined as requirement of therapeutic vitamin D and/or calcium replacement at 6 months or a fasting albumin-corrected serum calcium below 8.0 mg/dL. In this study, unintentional recurrent laryngeal nerve paresis would be defined as new vocal cord paralysis diagnosed within 6 months postoperatively in a patient who had normal vocal cord movement preoperatively. Permanent recurrent laryngeal nerve injury is by definition persistent beyond the 6-month postoperative period. Routine laryngoscopy would be performed in all patients at 6 months following thyroid surgery unless demonstrated to be normal prior to 6 months postoperatively. In addition, durations of surgery and hospital stay would be recorded. Postoperative cervical hematoma is defined as postsurgical bleeding requiring reoperative intervention.
Sample size estimation
Sample size was estimated for the primary outcome, newly identified structural disease (i.e., persistent, recurrent or distant metastatic disease), using the method of Lakatos (23,24) to approximate a discrete time hazard model. In the calculations we assumed a 7-year study with 4 years of recruitment and 5 years of average follow-up (minimum=3 years, maximum=7 years), a type I error of 5% (two-sided), 80% power, annual dropout rates from 1% to 5% in both treatment groups, and a range of reductions in the hazard rate with prophylactic central lymph node dissection compared with no central lymph node dissection (20%, 25%, 33%, and 40%). The estimates of the newly identified structural disease rate for the no prophylactic central lymph node dissection group were obtained from contemporary series and author experiences and were assumed to not be constant over the study period. The sample sizes were not inflated for interim monitoring for efficacy and futility, which would be approximately 4%–6%.
Sample sizes to detect differences in complication rates between the two treatment groups were estimated for binomial proportions assuming a Type I error of 5% (two-sided) and 80% power. These sample size determinations were made using PASS 2008 (NCSS, Kaysville, UT).
1It is anticipated that occasionally the diagnosis of newly identified structural disease and biochemical persistent or recurrent disease may be debatable in individual cases (e.g., elevated Tg without structural disease on imaging or radioisotope scan uptake that is not biopsy proven). Such cases would be reviewed by a multidisciplinary committee of investigators to determine the diagnosis of newly identified structural disease or biochemical persistent and recurrent disease.