2017 Section 7 Green Book

Surgery Volume 151, Number 6

Lang et al

for the association between DTC and breast cancer included the presence of sodium iodide symporters in mammary tissue leading to cumulative doses of I 131 , the female dominance in DTC survivors, ge- netic predisposition, and surveillance bias. 10,19 Nevertheless, when all types/sites of NSPM were considered, our data are consistent with the hy- pothesis that RAI therapy increased the overall risk of developing NSPM in DTC survivors. In the multivariable analysis, after adjusting for other possible factors linked to the risk of NSPM, a cumulative RAI activity of 3–8.9 GBq was found to be an independent factor for the development of NSPM in DTC survivors, and the relative increased risk was approximately 2–3 times higher than survivors who received no RAI. However, because the RR was not significant in the group who had cumulative RAI activity > 9.0, the present study was not able to establish a dose–effect relationship between cumulative RAI activity and risk of NSPM. This might have been because only 29 patients had cumulative RAI activity > 9.0 GBq, and our study was therefore underpowered to identify an effect. Another explanation might have been related to the disease threshold phenomenon, where the risk of RAI activity $ 9.0 GBq far exceeded the dose threshold and so imparted the same risk as RAI activity of 3.0–8.9 GBq. To further confirm that RAI therapy was the factor responsible for the increased risk of NSPM and not factors that influ- enced the decision to prescribe RAI therapy in the Table IV. Observed and expected number of cases and standardized incidence ratio with the corre- sponding 95% CIs of nonsynchronous second pri- mary malignancy for those who did and did not receive radioiodine in males, females, and both sexes * Observed no. of NSPMs Expected number of NSPMs Standardized incidence ratio 95% CI RAI + group Males 14 9.90 1.41 0.77–2.37 Females 42 27.28 1.54 1.11–2.08 Both sexes 56 37.18 1.51 1.14–1.96 RAI group Male 1 1.88 0.53 0.01–2.96 Female 7 7.59 0.92 0.37–1.90 Both sexes 8 9.47 0.84 0.36–1.66 *Expected numbers of NSPM were based on population incidence of all sites in 2008 after adjusting for age. CI , Confidence interval; NSPM , nonsynchronous second primary malig- nancy; RAI , radioactive iodine (I 131 ).

first place, the present analysis compared factors that might have influenced the decision to pre- scribe RAI (such as the stage of DTC, period of DTC diagnosis, and age of DTC), and these significant factors were entered into the multivar- iable analysis together with other well-reported risk factors, such as sex. 17,20 Although the male sex was not an independent factor for NSPM, 2 previous studies found that male survivors were at increased risk of NSPM. 17,20 Therefore, male DTC survivors treated with RAI might be at even greater risk of NSPM. Nevertheless, our SIR analysis in NSPM did not support this finding with the males in the RAI + group having a slightly lower SIR value than females in the RAI + group after adjusting for age (1.41 vs 1.54). However, there appeared to be a strong association between breast cancer and DTC; the higher risk of NSPM observed in RAI + female patients might have been a result of this association. 4-6 Unlike other studies, our data showed that the risk of NSPM in the RAI group was similar to that of the general popula- tion. 6,8,11,17 Perhaps this further strengthened the association between RAI therapy and risk of NSPM in DTC survivors. However, because < 30% of DTC patients received no RAI, our study might have been underpowered in this aspect. One of the strengths of the present study was the complete patient follow-up status and data collection, and this was not possible without the establishment of the territory-wide CMS in 1995. 21 Unlike previous larger-scale analyses, all patients in the present study were managed under a standard- ized treatment protocol, and the histology of DTC was confirmed by the same group of pathologists at our institution. Given the completeness of the clin- ical data, the chance of NSPM misclassification would have been minimal, and our SIR results were already adjusted for both patient age and sex. However, similar to previous single-center analysis, the total number of NSPMs remained rel- atively small, which limited the power of our study to identify smaller effects. Also, the present analysis was potentially subject to a degree of institution and referral biases, because our center is an aca- demic tertiary care institute to which more com- plex cases are often referred from other hospitals. Because there were a number of signifi- cant differences in baseline clinical and demo- graphics between RAI + and RAI groups, they might not have been fully adjusted by the multivar- iate analysis. A future multicenter study involving other institutions in our territory would be desir- able to further confirm our findings. Although none of the 64 patients with NSPM had known

96