2017 Section 7 Green Book

Surgery June 2012

Lang et al

calculated by dividing observed incidence rates in our cohort by expected incidence rates in the gen- eral population. The observed number of each NSPM was calculated by compiling the person- years of observation according to 5-year age groups by sex from the diagnosis of DTC to the date of death, date of last follow-up, or date of diagnosis of NSPM, whichever occurred first. The expected number of each NSPM in 5-year age groups by sex were derived from cancer incidence statistics in 2008 reported by the Hong Kong Cancer Regis- try ( http://www3.ha.org.hk/cancereg/ ) and were multiplied by the accumulated person-years of ob- servation at risk to obtain the expected number of NSPM. For the comparison for dichotomous variables between the RAI + and RAI groups, v 2 and Fisher exact tests were used. The Mann–Whitney U test was used for the comparison of continuous varia- bles between RAI + and RAI groups. All statistical analyses were conducted using SPSS software (ver- sion 18.0; SPSS, Inc., Chicago, IL). RESULTS Table I shows the baseline patient characteris- tics. Six hundred ninety-five (77.7%) patients had papillary thyroid carcinoma and 200 (22.3%) had follicular thyroid carcinoma. The majority were fe- male (80.6%) and ethnic Chinese (94.1%). The median age of DTC diagnosis was 44.0 years (range, 7.1–90.6), and the median follow-up pe- riod was 93.5 months (range, 23.1–570.8). Seven hundred sixty-three (85.3%) patients underwent bilateral thyroid resection, and of these, 643 (84.3%) patients received at least 1 course of RAI therapy, whereas of the 132 (14.7%) patients who underwent unilateral thyroid resection, no patient received RAI therapy. There were 55 (6.1%) pa- tients who received $ 2 doses of RAI therapy. Table II shows a comparison of demographics, pe- riod of DTC diagnosis, number and type of NSPM, major histologic types of DTC, and stage of DTC between the RAI + and RAI groups. Patients in the RAI + group were significantly older at the time of DTC diagnosis (47.5 vs 44.0; P < .001) and there was a significantly greater proportion of patients belonging to the $ 50-year-old age group (39.9% vs 28.2%; P < .001). There were a similar proportion of males in the 2 groups. When the periods of DTC diagnosis were com- pared, there were a significantly greater propor- tion of patients not receiving RAI treatment towards the later period. Tumor size was signifi- cantly larger in the RAI + group because size was an important criterion for RAI ablation. Similarly,

Table I. Baseline patient demographics and char- acteristics ( n = 895) Median Range or % Age at diagnosis of DTC, y 44.0 7.1–90.6 Sex Male 174 19.4 Female 721 80.6 Major histologic types of DTC Papillary 695 77.7 Follicular 200 22.3 Tumor stage of DTC by TNM I 586 65.5 II 59 6.6 III 136 15.2 IV 114 12.7 Follow-up time, months 93.5 23.4–570.8

No. of patients with NSPM * detected during follow-up Latency period to NSPM * from time of DTC, months

64

7.2

189.5 22.8–531.1

Radiation exposure

No RAI therapy given

252 643 588

28.2 71.8 65.7

RAI therapy given

1 course 2 courses

26 29

2.9 3.2

$ 3 courses

*Only second primary malignancy which occurred > 12 months after the diagnosis of DTC was included. DTC , Differentiated thyroid carcinoma; NSPM , nonsynchronous second primary malignancy; RAI , radioactive iodine; TNM , American Joint Can- cer Committee/Union Internationale Contre le Cancer tumor-nodes- metastasis staging system, 6th edition.

the stages of DTC were more advanced in the RAI + group as reflected by the higher proportion of stage III and IV patients (18.8% vs 6.0% and 15.6% vs 5.6%, respectively; P < .001). In terms of the type of NSPM, primary breast, colon, and lung cancers were the 3 most common NSPMs in the RAI + and RAI groups. The Figure shows the cumulative risk of devel- oping NSPM after the diagnosis of DTC in the RAI + and RAI groups. The 20-year cumulative risk of NSPM in the RAI + group was significantly higher than that in the RAI group (13.5% vs 3.1%; P = .015). The mean times to development of NSPM in the RAI + and RAI groups were 34.37 years (95% CI, 32.6–36.1) and 43.05 years (95% CI, 40.3–45.8). Table III shows the Cox proportional hazards analysis of factors influencing the development of NSPM in patients with DTC. The analysis included variables that were significant in the comparison between RAI + and RAI groups and factors that might be linked to development of NSPM. Tumor size was not entered because the stages of DTC by

93