Surg Today DOI 10.1007/s00595-016-1417-2

ORIGINAL ARTICLE

Is body mass index relevant to prognosis of papillary thyroid carcinoma? A clinicopathological cohort study Yoo Seung Chung1 · Joon‑Hyop Lee1 · Young Don Lee1 

Received: 24 May 2016 / Accepted: 22 July 2016 © Springer Japan 2016

Abstract  Purpose  Obesity appears to be related to papillary thyroid carcinoma (PTC) in the observational studies, although its relationship concerning the PTC prognosis has not been established. We investigated the association between body mass index (BMI) and the prognosis of PTC. Methods  The WHO BMI classification was used to stratify the degree of obesity. The final outcome was disease status, including recurrence and persistence, of 783 PTC patients. We reviewed patients’ BMI, disease status, and other prognostic factors retrospectively. Results The mean BMI was 24.2 kg/m2. When stratified according to the WHO BMI classification, 21 were Underweight, 482 were Normal, 232 were Overweight, and 48 were Obese. We divided patients into two groups: <25.0 kg/m2 (n  = 503) vs. ≥25.0 kg/m2 (n  = 280). The BMI ≥25.0 group was older and more likely to be male in a multivariate analysis (p < 0.001). For those with BMI <25.0 and ≥25.0, recurrence occurred in 3.0 and 2.1 % (p = 0.486), persistence in 7.2 and 5.1 % (p = 0.265), and either recurrence or persistence in 9.9 and 7.1 %, respectively (p  = 0.189). A multivariate analysis revealed that older age and male gender in Overweight vs. Normal, older * Young Don Lee [email protected] Yoo Seung Chung [email protected] Joon‑Hyop Lee [email protected] 1



Thyroid and Endocrine Surgery Section, Department of Surgery, Gachon University Gil Medical Center, Gachon University School of Medicine, Gachon University, 1198 Guwol‑dong, Namdong‑gu, Incheon 405‑760, Republic of Korea

age in Obese vs. Normal, and advanced T stage in Normal vs. Underweight were statistically significant prognostic factors. Conclusions There was no significant difference in the prognosis according to BMI in PTC patients. However, old age, male gender, and advanced T-stage patients were found more frequently in the higher BMI group than in the lower BMI group. Keywords  Papillary thyroid carcinoma · Obesity · BMI · Prognosis

Introduction Papillary thyroid carcinoma (PTC) is the most common endocrine-related carcinoma, and its incidence has been increasing recently, a phenomenon mainly caused by the increase in the incidence of small PTC. This may be due to improvements in diagnostic tools, such as thyroid ultrasonography (USG) and fine needle aspiration cytology (FNAC), or incidental findings during a cervical vascular imaging study or computed tomography (CT) for chest or spinal evaluation [1–3]. However, the incidence of advanced stage thyroid cancer has also increased with the increase in the incidence of small-sized thyroid cancer [4], which cannot be explained by increased medical surveillance and the usage of diagnostic tools alone. Environmental radiation, CT using iodine contrast media, and the extension of the human lifespan may all contribute to the increasing incidence of PTC [5]. Obesity is a serious health problem causing several chronic diseases and is known to contribute to the cancer development of the colon, breast, kidney, and endometrium. In addition, obesity is reported to have an adverse

13



Surg Today

effect on the prognosis of various types of carcinomas. Likewise, BMI is significantly associated with higher rates of mortality due to esophagus, colorectal, liver, gallbladder, pancreas, endometrium, ovary, breast, and kidney cancers [6, 7]. A positive correlation between the risk of thyroid carcinoma and obesity has been noted in many observational studies [8–11]. Although the relationship between obesity and thyroid carcinoma has been epidemiologically evaluated, the underlying mechanism has not been confirmed. Insulin resistance, IGF-1, cytokines, inflammation, TSH, estrogens, adiponectin, and leptin have received attention as associated factors [5, 12, 13]. Many researchers have also made efforts to establish a connection between obesity and the prognosis of PTC [14–21]. However, these studies have not shown consistent results, and whether or not the severity of obesity influences the prognosis of PTC remains unclear. A large cohort of thyroid cancer patients will need to be analyzed with long-term follow-up to answer these questions. Therefore, we evaluated the relevance of obesity on the prognosis and clinicopathologic prognostic factors of PTC, with a particular focus on the feasibility of obesity as a prognostic factor of PTC.

ipsilateral central LN dissection were performed for unilateral thyroid tumors (<1 cm in maximum diameter) suggestive of malignancy by FNAC when there was no suspicion of a nodule in the contralateral lobe. We also performed total thyroidectomy with bilateral central LN dissection. Central LN dissection was carried out to harvest pretracheal, paratracheal, and Delphian LNs during thyroidectomy. Lateral neck LN dissection was performed in patients suspected of having LNM by USG and nodal FNAC or Tg washout measurement. Levels II–V were routinely dissected during lateral neck LN dissection in thyroid cancer patients. Follow-up studies included routine cervical USG, thyroid function test (TFT), CT, and 131I whole-body scan (WBS). Starting with TFT 2 months after and neck USG 1 year after operation, at least a semiannual TFT and annual cervical USG were performed. If needed, chest CT, serum off-Tg, and 131I WBS were performed. Routine serum offTg with thyroid hormone withdrawal in all 783 patients was not performed during the follow-up period. Distant metastasis was confirmed based on the first postoperative 131 I WBS or postoperative chest CT. Our institutional review board approved this retrospective study before the patient list was retrieved from the hospital database.

Materials and methods

BMI calculation

Data source and study design

The World Health Organization (WHO) BMI classification was used to stratify the severity of obesity. BMI was calculated as the weight in kilograms divided by the square of the height in meters (kg/m2). The height and weight of patients at the time of operation were measured. We divided the 783 patients into 2 BMI groups (<25.0 kg/m2 and ≥25.0 kg/m2) and analyzed the results between the two groups. BMI was further classified into four groups as follows: Underweight, BMI <18.5 kg/m2; Normal, 18.5–24.9 kg/m2; Overweight, 25.0–29.9 kg/m2; and Obese, ≥30.0 kg/m2. We compared patients’ clinicopathologic results between the Normal group vs. the Underweight, Overweight, and Obese groups.

A total of 783 primary PTC patients who underwent an operation between October 2006 and December 2009 were enrolled in this cohort. During this period, we performed 1225 thyroid operations. We chose 783 of the primary PTCs among 880 malignancies, excluding recurred thyroid malignancies and other malignant histologies. Patients’ medical charts were reviewed retrospectively. We reviewed patients’ BMIs at the time of their operation, disease status during follow-up, and other prognostic factors at the time of diagnosis. These factors included age, sex, tumor size, extrathyroidal extension (ETE), multiplicity, bilaterality, lymph node metastasis (LNM), distant metastasis, and AJCC/UICC TNM stage documented in the medical records, pathologic reports, and operation records. Age was managed as a continuous variable or divided into two groups at a cutoff of 45 years. As T2 and T4 comprised such a small portion, T was dichotomized into two groups as T1 and T2/3/4. TNM stage was considered as two groups of Stage I/II and Stage III/IV, because the portion of stage II and IV was too small. In patients who had a lobectomy with isthmectomy, we determined that there was no bilaterality based on preoperative USG. Thyroid lobectomy and isthmectomy with

13

Primary endpoints The final outcome was disease status, such as recurrence, persistent disease, and no disease events. Recurrence was defined as situations, in which patients were pathologically confirmed to have recurrent disease by cytology using fine needle aspiration or specimens from surgical excision 12 months after the first thyroid cancer operation. When the serum on- or off-Tg increased without cervical pathologic lesions, 131I WBS and chest CT were used to check for distant metastasis. Persistent disease was defined as patients not satisfying the following criteria of no evidence

Surg Today

of disease: (1) no clinical evidence of tumor; (2) no imaging evidence of tumor (no uptake outside the thyroid bed on the initial post-treatment WBS, or if uptake outside the thyroid bed was present, no imaging evidence of tumor on a recent diagnostic scan and neck US); and (3) undetectable serum Tg levels during TSH suppression and stimulation in the absence of interfering antibodies. Naturally, patients with no disease events were those who had not experienced recurrence or persistent disease during follow-up. Patients with no disease events were the favorable prognosis group that we sought. Statistical analyses Statistical analysis was performed using the SPSS version 20.0. Pearson’s χ2 test, Fisher’s exact test, and the independent t test were used to evaluate the differences between the WHO BMI classifications. In addition, a multivariate logistic regression analysis and Kaplan–Meier survival analysis with log-rank tests were performed. Statistical significance was accepted for p values <0.05.

Results Table  1 shows the clinicopathologic characteristics of the 783 PTC patients. The follow-up duration was 80.86  ± 22.80 months. The mean height of the patients was 159.5 cm (140.0–189.0), the mean weight was 61.7 kg (36.0–108.0), and the mean BMI was 24.2 kg/m2 (13.7– 40.7). The 783 patients were classified based on the WHO BMI classification as follows: 21 (2.7 %) as Underweight (<18.5), 482 (61.6 %) as Normal (18.5–24.9), 232 (29.6 %) as Overweight (25.0–29.9), and 48 (6.1 %) as Obese (≥30.0). Total thyroidectomy was performed in 593 patients (75.7 %) and lobectomy with isthmectomy in 148 (18.9 %). Lateral neck dissection was performed in 42 patients (5.4 %) accompanied by total thyroidectomy. T2 and T4 were 15 (1.9 %) and 1 (0.1 %), respectively. There were 10 patients (1.3 %) in stage II and 23 (2.9 %) in stage IV. Distant metastasis was found in 4 patients (0.5 %) at the lungs. Recurrence occurred in 21 patients (2.7 %), 19 of whom were women. The mean age was 42.86 ± 11.05 years, and the follow-up duration was 89.10 ± 9.26 months. The tumor size was 1.72 ± 1.09 cm. The mean time until recurrence was 33.14 ± 15.88 months, and the BMI was 23.6 kg/m2 (19.4–27.3). The site of recurrence was in the regional cervical area for all patients. The lateral neck LN was the most common site (18/21), and there was one case of contralateral lobe recurrence, one central LN, and one

Table 1  Characteristics of PTC patients Variables

Results (n = 783)

Age (years), mean (SD)  <45  ≥45

47.59 (11.07) 297 (37.9 %) 486 (62.1 %)

Sex, female:male, n (%)

689 (88.0 %):94 (12.0 %)

Tumor size (cm), mean (SD) PTMC  ≤1 cm  >1 cm ETE (+) Multifocality (+) Bilaterality (+) T  T1  T2/3/4 N  N1 M  M1 Stage  I/II  III/IV Recurrence

1.22 (0.90)

Recurrence + persistent disease

70 (8.9 %)

399 (51.0 %) 384 (49.0 %) 480 (61.3 %) 233 (29.8 %) 143 (18.3 %) 283 (36.1 %) 500 (63.9 %) 390 (49.8 %) 4 (0.5 %) 431 (55.0 %) 352 (45.0 %) 21 (2.7 %)

SD standard deviation, PTC papillary thyroid carcinoma, PTMC papillary thyroid microcarcinoma, ETE extrathyroidal extension

lateral neck soft tissue. Persistent disease was found in 49 patients (6.3 %). The mean age was 45.92 ± 13.06 years, and the follow-up duration was 81.49 ± 26.52 months. The tumor size was 1.84 ± 1.26 cm, and the BMI was 23.7 kg/ m2 (18.1–31.3). The BMI in patients with no disease events was 24.2 kg/m2 (13.7–40.7), and there was no significant difference in the BMI between the patients with recurrence, persistent disease, and no disease events (p = 0.396). As obesity might prevent patients from detecting neck lumps due to excess flesh at their necks, thereby leading to a delayed diagnosis, we evaluated the chance of patients recognizing signs of thyroid tumor according to the severity of obesity. About 18.4 % (144/783) of patients came to the clinic after detecting a neck mass either through self-palpation or others noticing it. There was no marked difference in the rate of detection among the 4 BMI classifications (14.3, 20.5, 13.8, and 20.8 % in the Underweight, Normal, Overweight, and Obese groups, p = 0.159). After dividing all cases into 2 groups of BMI (<25.0 vs. ≥25.0 kg/m2), we still could not find detect any significant difference in the rate of detecting a thyroid tumor between the two groups (20.3 vs. 15.0 %, p = 0.083).

13



Surg Today

BMI <25.0 kg/m2 vs. BMI ≥25.0 kg/m2 We divided the patients into two groups based on BMI: <25.0 kg/m2 (n = 503, 64.2 %) vs. ≥25.0 kg/m2 (n = 280, 35.8 %). There was no significant difference in disease status between the groups. Recurrence was detected in 3.0 % of BMI <25.0 (15/503) and 2.1 % of BMI ≥25.0 patients (6/280) (p  = 0.486). In a Kaplan–Meier analysis, there was no significant difference in the rate of recurrence (p = 0.510). There were no disease events in 90.1 % of the BMI <25.0 (453/503) group and in 92.9 % of the BMI ≥25.0 (260/280) group (p = 0.189). Persistent disease was detected in 6.4 % (49/762, excluding patients with recurrence). There was no significant difference in persistent disease rate, at 7.2 % (35/488, BMI <25.0) and 5.1 % (14/274, BMI ≥25.0) (p = 0.265). The follow-up duration was also not markedly different between the two groups (81.51 ± 22.42 months in BMI <25.0 vs. 79.70 ± 23.47 in BMI ≥25.0, p = 0.285).

Table  2 shows the comparison of prognostic factors between BMI <25.0 and BMI ≥25.0 patients. The BMI ≥25.0 group tended to be older, had a higher ratio of males, and was more advanced in stage than the BMI <25.0 group. According to a multivariate analysis, significant differences between the groups in age and gender were observed (p < 0.001). We evaluated the difference in the metastatic LN ratio as the number of lymph node metastases between BMI <25.0 and BMI ≥25.0 groups. The number of LNs obtained during surgery was not markedly different between the two groups (16.25 ± 14.92 in BMI <25.0 and 15.49 ± 15.41 in BMI ≥25.0, p = 0.500). The metastatic LN ratio was evaluated when the number of obtained LNs was more than 6 (n = 656). The LN ratio was 0.16 ± 0.21 in the BMI <25.0 group and 0.16 ± 0.20 in the BMI ≥25.0 group. There was no significant difference in the metastatic LN ratio according to BMI (p = 0.960).

Table 2  Comparison of patients’ characteristics between BMI <25.0 kg/m2 and BMI ≥25.0 kg/m2 Variables

<25 kg/m2, n = 503 (64.2 %)

≥25 kg/m2, n = 280 (35.8 %)

Univariate

Multivariate

p

p

OR

95 % CI

Age (years), mean ± SD  <45  ≥45 Sex  Female  Male Size (cm), mean ± SD PTMC  ≤1 cm  >1 cm ETE (+) Multifocality (+) Bilaterality (+) T stage  T1  T2/3/4 N stage  N1 M stage  M1 Stage  I/II  III/IV

46.00 ± 11.20 221 (43.9 %) 282 (56.1 %)

50.45 ± 10.23 76 (27.1 %) 204 (72.9 %)

<0.001 <0.001

<0.001 na

1.039 na

1.025–1.054 na

460 (91.5 %) 43 (8.5 %) 1.22 ± 0.90

229 (81.8 %) 51 (18.2 %) 1.22 ± 0.89

<0.001

<0.001

2.491

1.596–3.889

0.975

0.335

0.873

0.663–1.150

262 (52.1 %) 241 (47.9 %) 308 (61.2 %) 145 (28.8 %) 89 (17.7 %)

137 (48.9 %) 143 (51.1 %) 172 (61.4 %) 88 (31.4 %) 54 (19.3 %)

0.397

0.305

1.259

0.811–1.955

0.957 0.445 0.581

0.463 0.698 0.923

0.669 1.100 0.972

0.229–1.957 0.679–1.784 0.548–1.725

183 (36.4 %) 320 (63.6 %)

100 (35.7 %) 180 (64.3 %)

0.852

0.491

1.479

0.486–4.501

244 (48.5 %)

146 (52.1 %)

0.330

0.131

1.315

0.922–1.876

3 (0.6 %)

1 (0.4 %)

0.549

0.879

0.819

0.063–10.700

299 (59.4 %) 204 (40.6 %)

132 (47.1 %) 148 (52.9 %)

0.001

0.983

1.005

0.647–1.561

No disease events

453 (90.1 %)

260 (92.9 %)

0.189

0.213

0.685

0.378–1.242

OR odds ratio, CI confidence interval, SD standard deviation, na not assessed, PTMC papillary thyroid microcarcinoma, ETE extrathyroidal extension

13

Surg Today

Normal (18.5 kg/m2 ≤ BMI < 25.0 kg/m2) vs. Overweight (25.0 kg/m2 ≤ BMI < 30.0 kg/m2) The patients in the Overweight group tended to be older (46.05  ± 11.03 vs 50.66 ± 10.04, p < 0.001) and had a higher proportion of males (42/482, 8.7 % vs 45/232, 19.4 %, p < 0.001) than the Normal group. There were more advanced stage cases in the Overweight (200/482, 41.5 % vs. 126/232, 54.3 %, p = 0.001) than in the Normal group. Overall, significant differences in the rates of old age and male gender were seen between these two groups in a multivariate analysis (Table 3). Disease status, such as recurrence (p = 0.697), persistent disease (p = 0.227), and no disease events (p = 0.220), did not show any significant differences between these groups (Table 4). In a Kaplan– Meier analysis, there was no significant difference in the rate of recurrence between the groups (p = 0.720). Normal (18.5 kg/m2 ≤ BMI < 25.0 kg/m2) vs. Obese (BMI ≥30 kg/m2) As confirmed in a multivariate analysis (Table 3), patients in the Obese group tended to be older (46.05 ± 11.03 vs

49.46 ± 11.18, p = 0.042) than in the Normal group. Disease status, such as recurrence (p = 0.382), persistent disease (p  = 0.540), and no disease events (p  = 0.609), did not show any significant differences between these groups (Table 4). In addition, a Kaplan–Meier analysis showed no significant difference in the rate of recurrence between the groups (p = 0.227). Normal (18.5 kg/m2 ≤ BMI < 25.0 kg/m2) vs. Underweight (BMI <18.5 kg/m2) The patients with Normal BMI had more ETE, advanced T/ TNM stage, and larger tumors than Underweight patients. The incidence of ETE was 33.3 % (7/21) in Underweight and 62.4 % (301/482) in Normal (p  = 0.007), that of T2/3/4 was 33.3 % (7/21) in Underweight and 64.9 % (313/482) in Normal (p = 0.003), and that of Stage III/IV was 19.0 % (4/21) in Underweight and 41.5 % (200/482) in Normal (p = 0.040). The mean tumor size in the Underweight group was 0.87 ± 0.50 cm, and that in the Normal group was 1.24 ± 0.91 cm (p  = 0.003). In a multivariate analysis, there was a statistically significant difference in T stage between the groups (Table 3). Disease status, such as

Table 3  Statistically significant variables between WHO BMI classification groups Variables

Underweight (n = 21)

Overweight (n = 232)

Obese (n = 48)

ETE T stage Stage Tumor size Age ≥45 years Male gender Stage

Normal (n = 482) Univariate

Multivariate

p

p

OR

95 % CI

ns 0.008a ns ns <0.001b na <0.001b ns

ns 3.548a ns n.s 1.042b na 2.584b ns

ns 1.402–8.983a ns ns 1.026–1.058b na 1.624–4.110b ns

0.027c

1.031c

1.004–1.059c

a

0.007 0.003a 0.040a 0.003a <0.001b <0.001b <0.001b 0.001b 0.042c

Age

OR odds ratio, CI confidence interval, ETE extrathyroidal extension, ns not significant, na not assessed a

  Significant results between Underweight BMI and Normal BMI classification

b

  Significant results between Normal BMI and Overweight BMI classification

c

  Significant results between Normal BMI and Obese BMI classification

Table 4  Disease status of PTC patients based on BMI classifications

Persistent diseasea

Recurrence

No disease events

Underweight, n = 21

0/21

0.0 %

1/21

4.8 %

20/21

95.2 %

Normal, n = 482

15/482

3.1 %

34/467

7.3 %

433/482

89.8 %

Overweight, n = 232

6/232

2.6 %

11/226

4.9 %

215/232

92.7 %

Obese, n = 48

0/48

0.0 %

3/48

6.3 %

45/48

93.8 %

PTC papillary thyroid carcinoma, BMI body mass index a

  Persistent disease was calculated excluding recurrent patients

13



recurrence (p = 0.523), persistent disease (p = 0.546), and no disease events (p = 0.710), did not show any significant differences between the groups (Table 4). In a Kaplan– Meier analysis, the two groups showed no significant difference in the recurrence rate (p = 0.420). Difference in complication rates: BMI <25.0 kg/m2 vs. BMI ≥25.0 kg/m2 Complications, such as nerve injury, postoperative bleeding, hypocalcemia, and chyle leakage, were evaluated between the BMI <25.0 and BMI ≥25.0 groups. Nerve injury during operation occurred in 5 patients (5/783, 0.6 %), all of whom were BMI <25.0 (5/503, 1.0 %). There was no significant difference in the rate of nerve injury between the two groups (p = 0.166). There were four cases of postoperative bleeding. Two occurred in the BMI <25.0 group (2/503, 0.4 %), while the other two were in the BMI ≥25.0 group (2/280, 0.7 %), a non-significant difference (p = 0.620). Postoperative hypocalcemia was analyzed in patients who underwent total thyroidectomy (n = 635). Transient hypocalcemia occurred in 10.6 % (67/635), which translated to 10.5 % (43/409) in the BMI <25.0 group and 10.6 % (24/226) in the BMI ≥25.0 group (p  = 0.967). Permanent hypocalcemia occurred in 1.1 % (7/635), which translated to 1.5 % (6/409) in the BMI <25.0 group and 0.4 % (1/226) in the BMI ≥25.0 group (p = 0.431). There were 3 cases of chyle leakage: 1 in the BMI <25.0 group (1/503, 0.2 %) and 2 in the BMI ≥25.0 group (2/280, 0.7 %), a non-significant difference (p = 0.292).

Discussion In our series, higher BMI was significantly related to older age and male gender when comparing patients with BMI <25.0 kg/m2 and those with BMI ≥25.0 kg/m2. The same trend was noted between the Normal and Overweight groups. Patients tended to be older in the Obese group than in the Normal group. In addition, T stage tended to be more advanced in the Normal group than in the Underweight group. Despite not showing statistical significance in a multivariate analysis, advanced stage, ETE, and larger tumor size were more frequent in the higher BMI group than in the lower BMI group. There were no significant differences between the BMI groups in recurrence, persistent disease, and no disease events. Trésallet et al. reported that obese patients had an increased risk of disease persistence in macro-PTC [14]. When macrocarcinoma and microcarcinoma were combined, there were more older and male patients in the higher BMI group, which was the same as in our study. We analyzed our macro-PTC patients separately, although there was no significant difference in the rates of disease status, such as recurrence, persistent disease, and no disease

13

Surg Today

events (data not shown). Only older age and male gender showed a significant difference in macro-PTC patients. An advanced stage of III/IV was more frequent in the higher BMI group than in the lower BMI group [15–17]. Although it was not statistically significant in a multivariate analysis, stage III/IV was associated with a higher BMI in our results. In a univariate analysis comparing the BMI <25.0 and ≥25.0 kg/m2 groups, the Underweight and Normal groups, and the Normal and Overweight groups, all higher BMI groups showed a greater proportion of advanced stage patients than the lower BMI groups. Regarding tumor size, larger tumors were more likely to be found in patients with higher BMI than in those with lower BMI [16–18]. In our series, tumor size tended to be larger in the Normal group than in the Underweight group. ETE [16], LNM, and tumor multiplicity [19] were suggested to be related to BMI. Despite lacking statistical significance in a multivariate analysis, ETE was more frequent in the Normal BMI group than in the Underweight groups in our series. LNM and tumor multiplicity were only related to BMI in patients ≥45 years [19]. When we analyzed ≥45-year-old patients separately, there was no significant difference in the LNM, tumor multiplicity, and disease status (data not shown). Only older age and male gender showed a significant difference in ≥45-year-old patients. However, several studies have reported that BMI was not associated with the clinicopathologic features or the prognosis of PTC. Kim et al. stated that there was no independent association between BMI and T/N stage [20]. Paes et al. reported that no association was detected between BMI and T/M stage, recurrence, or persistent disease. Interestingly, there was an inverse relationship between BMI and LNM and tumor invasion. Patients with lower BMI showed more LNM and tumor invasion after a multivariate analysis [21]. Similarly, more LNM was detected in the lower BMI group than in the higher BMI group [14]. We evaluated the Korean researchers’ studies collectively for the consideration of genetic characteristics. However, no consistent results were obtained [16, 18–20]. Therefore, we were unable to confirm a relationship between poor prognostic factors and BMI. However, the disease status, such as recurrence, persistent disease, and no disease events, was definitely not associated with BMI. Our study has several strengths. First, the number of subjects was relatively large. Although there have been a few studies, including over 1000 subjects [14, 16, 17], most researchers deal with several hundred cases in their studies [15, 18–21]. Therefore, our subject number (783 patients) was not small compared with that of other studies. Second, our follow-up period (about 6.7 years) was not short. This was longer than that of other studies dealing with prognosis (6.2 years) [14, 21]. Furthermore, 89 % (698/783) of our patients had a follow-up period of over 60 months. Third, we

Surg Today

dealt with three types of disease status: recurrence, persistent disease, and no disease events. There have been a few studies investigating the prognosis of PTC and BMI [14, 21]. However, they did not distinguish between recurrence and persistent disease. When we analyzed persistent disease, we excluded recurrent cases for a more detailed categorization of disease status. The patients with no disease events had experienced neither recurrence nor persistent disease. We believe that our exclusion of recurrent cases is the most accurate method of reflecting the patients’ status. No study has analyzed the status of recurrent, persistent, and no disease events. However, several limitations associated with this study also warrant mention. There are various indexes indicating the extent of obesity, and although BMI measurement is the most universal method, waist-to-hip ratio or waist circumference is also proposed indicator of abdominal adiposity. It has been observed that abdominal adiposity is more closely associated with cancer than BMI [8]. Because we did not measure the waist and hip circumference before the operations, abdominal adiposity could not be evaluated. We also did not have detailed information of the patients’ medical history, such as diabetes, hypercholesterolemia, cardiovascular disease, smoking, alcohol intake, and degree of physical activity. These may be confounding factors influencing the relationship between BMI and the prognosis of PTC. We were unable to evaluate the biological factors related to obesity. Several biological factors, such as insulin resistance, IGF-1, and cytokines, might be potential underlying mechanisms [5, 12, 13]. In future prospective studies, the evaluation of a detailed medical history, abdominal adiposity, and biological parameters should be incorporated to address these factors. In conclusion, there was no significant difference in the prognosis according to BMI in PTC patients. BMI cannot be used as a prognostic factor for PTC at present. However, prognostic factors, such as old age, male gender, and advanced T stage were related to higher BMI. A prospective study in a large-scale cohort with long-term follow-up might obtain a more conclusive relationship between BMI and the prognosis of PTC. Compliance with ethical standards  Conflict of interest The authors declare no conflicts of interest in association with this study.

References 1. Davies L, Welch HG. Increasing incidence of thyroid cancer in the United States, 1973–2002. JAMA. 2006;295:2164–7. 2. Davies L, Ouellette M, Hunter M, Welch HG. The increasing incidence of small thyroid cancers: where are the cases coming from? Laryngoscope. 2010;120:2446–51.

3. Oh CM, Jung KW, Won YJ, Shin A, Kong HJ, Lee JS. Ageperiod-cohort analysis of thyroid cancer incidence in Korea. Cancer Res Treat. 2015;47:362–9. 4. Enewold L, Zhu K, Ron E, Marrogi AJ, Stojadinovic A, Peoples GE, et al. Rising thyroid cancer incidence in the United States by demographic and tumor characteristics, 1980–2005. Cancer Epidemiol Biomark Prev. 2009;18:784–91. 5. Pazaitou-Panayiotou K, Polyzos SA, Mantzoros CS. Obesity and thyroid cancer: epidemiologic associations and underlying mechanisms. Obes Rev. 2013;14:1006–22. 6. Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. N Engl J Med. 2003;348:1625–38. 7. Reeves GK, Pirie K, Beral V, Green J, Spencer E, Bull D, Million Women Study Collaboration. Cancer incidence and mortality in relation to body mass index in the Million Women Study: cohort study. BMJ. 2007;335:1134. 8. De Pergola G, Silvestris F. Obesity as a major risk factor for cancer. J Obes. 2013;291546. 9. Kitahara CM, Platz EA, Freeman LE, Hsing AW, Linet MS, Park Y, et al. Obesity and thyroid cancer risk among U.S. men and women: a pooled analysis of five prospective studies. Cancer Epidemiol Biomark Prev. 2011;20:464–72. 10. Rinaldi S, Lise M, Clavel-Chapelon F, Boutron-Ruault MC, Guillas G, Overvad K, et al. Body size and risk of differentiated thyroid carcinomas: findings from the EPIC study. Int J Cancer. 2012;131:E1004–14. 11. Xu L, Port M, Landi S, Gemignani F, Cipollini M, Elisei R, et al. Obesity and the risk of papillary thyroid cancer: a pooled analysis of three case–control studies. Thyroid. 2014;24:966–74. 12. Pappa T, Alevizaki M. Obesity and thyroid cancer: a clinical update. Thyroid. 2014;24:190–9. 13. Marcello MA, Cunha LL, Batista FA, Ward LS. Obesity and thyroid cancer. Endocr Relat Cancer. 2014;21:T255–71. 14. Trésallet C, Seman M, Tissier F, Buffet C, Lupinacci RM, Vuarnesson H, et al. The incidence of papillary thyroid carcinoma and outcomes in operative patients according to their body mass indices. Surgery. 2014;156:1145–52. 15. Harari A, Endo B, Nishimoto S, Ituarte PH, Yeh MW. Risk of advanced papillary thyroid cancer in obese patients. Arch Surg. 2012;147:805–11. 16. Kim HJ, Kim NK, Choi JH, Sohn SY, Kim SW, Jin SM, et al. Associations between body mass index and clinico-pathological characteristics of papillary thyroid cancer. Clin Endocrinol (Oxf). 2013;78:134–40. 17. Dieringer P, Klass EM, Caine B, Smith-Gagen J. Associations between body mass and papillary thyroid cancer stage and tumor size: a population-based study. J Cancer Res Clin Oncol. 2015;141:93–8. 18. Na AS, Kang SY, Kim SK, Youn HJ, Jung SH. Clinicopathological relevance between body mass index and papillary thyroid carcinoma. Korean J Endocr Surg. 2014;14:171–6. 19. Kim SH, Park HS, Kim KH, Yoo H, Chae BJ, Bae JS, et al. Correlation between obesity and clinicopathological factors in patients with papillary thyroid cancer. Surg Today. 2015;45:723–9. 20. Kim JY, Jung EJ, Jeong SH, Jeong CY, Ju YT, Lee YJ, et al. The indices of body size and aggressiveness of papillary thyroid carcinoma. J Korean Surg Soc. 2011;80:241–4. 21. Paes JE, Hua K, Nagy R, Kloos RT, Jarjoura D, Ringel MD. The relationship between body mass index and thyroid cancer pathology features and outcomes: a clinicopathological cohort study. J Clin Endocrinol Metab. 2010;95:4244–50.

13