Thyroid Cancer in Childhood: A Retrospective Review of Childhood Course

Abstract

Background:

Thyroid cancer (TC) is an uncommon childhood malignancy, but the incidence may be increasing. Recent American Thyroid Association guidelines focus primarily on adult data. Natural history studies of TC in childhood are important to determine outcomes. The objectives of this study were to describe the demographics and outcomes in children with TC treated at The Hospital for Sick Children, Toronto, from 1983 to 2006. We hypothesized that childhood TC was increasing at our institution.

Methods:

Cases of papillary TC (PTC) (including follicular variant PTC) and follicular TC (FTC) were identified from pathology databases. Chart review was performed, and data were extracted on clinical, treatment, and outcome variables.

Results:

Sixty-one cases were identified, and complete data were available in 54, including 36 girls and 18 boys. There was no statistical change in numbers of cases diagnosed yearly during the study period. Younger children were more likely to have metastases at presentation or during follow-up. Pathological TC diagnosis included 40 PTC, 1 diffuse-sclerosing papillary, 7 follicular variant PTC, and 6 FTC. There was no difference in pathology findings between children less than or greater than 10 years old. Five patients had a history of previous malignancy, and five had a history of previous thyroid conditions. Three patients were born in areas of high TC endemnicity. Twenty-three patients had thyroiditis on pathology examination. All patients underwent total thyroidectomy, and 53/54 patients received therapeutic radioactive iodine ablation. Twenty-seven patients had metastases at presentation (19 lymph nodes only, 2 lung only, and 6 lymph node and distant) and 6 developed distant metastases during follow-up (3 lung, 2 thymus, and 1 paraspinal). Male sex was associated with development of metastases during follow-up. On multiple regression, tumor size was predicted positively by PTC but not by age, sex, or metastases at presentation or during follow-up.

Conclusion:

We did not find evidence of increasing numbers of cases of TC diagnosed yearly during the study period, or difference in tumor aggressiveness, or between outcomes in children aged less than or greater than 10 years. Children with metastases at presentation or during follow-up were likely to be younger than children without metastases. There is a need for prospective, collaborative multicenter studies of TC.

Introduction

T hyroid cancer (TC) is an uncommon childhood malignancy, with a reported age-standardized incidence rate of 2.0 per million per year in children aged 5–9 and 4.0 per million per year in children aged 10–14 in Canada from 2000 to 2004, accounting for 1.49% of all childhood malignancies under 14 years (1). The incidence of TC appears to be steadily increasing over the last two decades among adolescents and young adults, in part attributed to the treatment of head and neck malignancies with ionizing radiation from the 1930s to the 1960s (2) and also due to increased vigilance and early diagnosis (3). Some data support the impression that the incidence of TC is increasing (4,5). In Australia, two reports from the same institution describe 17 patients in whom the disease was diagnosed between 1966 and 1996 (6) and 14 between 1997 and 2004 (4).

Younger children appear to have more disease burden at presentation and decreased disease-free survival, but without an overall increase in mortality, compared to adults (7). Childhood TC has been associated with the following high-risk factors: younger age, male sex, large primary tumor size, extrathyroidal tumor extension, palpable lymph nodes, distant metastases at diagnosis, residual cervical disease after thyroidectomy, inadequate thyroid hormone suppression, diffuse sclerosing, or follicular thyroid histology (5).

Evidence-based guidelines for the management of TC published by the American Thyroid Association refer primarily to adult populations, and include limited reference to children (8). Natural history studies of TC in childhood are important given the paucity of literature relevant to this patient population. The aim of this study was to describe the outcome and demographics in children with TC treated at a large tertiary care children's hospital for 23 years.

Materials and Methods

Cases were identified by searching the Department of Pathology database, for confirmed histologic diagnoses of TC (papillary TC [PTC], follicular variant of PTC [F-PTC], follicular TC [FTC], and diffuse sclerosing) in patients aged less than 18 years at presentation. All tissue was obtained after surgical intervention (hemithyroidectomy, completion thyroidectomy [CT], or total thyroidectomy [TT]) at The Hospital for Sick Children (SickKids) from January 1984 to December 2006. Cases of medullary TC were excluded. Review of health records was conducted to determine patient demographics, risk factors, family history of thyroid disorders or malignancies, preoperative investigations, interventions (surgery and radioactive iodine [RAI]), and outcome until transfer to an adult institution at age 18 years. Data collection ended in December 2006. The study was approved by The Hospital for Sick Children Research Ethics Board.

Tumor size was determined by direct measurement at pathological examination after excision. Lymph node metastases were defined as pathological evidence of tumor metastases to lymph nodes. Distant metastases were defined as either pathological evidence of distant metastases or radiological evidence of metastases on either stimulated or unstimulated diagnostic RAI scan. Metastases that developed during follow-up were defined as pathological or diagnostic RAI scan evidence of metastases that were not present on evaluation at initial presentation, but were detected during the follow-up, after the initial therapy. Thymic uptake of RAI was deemed to be likely thymic metastases (or thyroid bed residual disease) if the patient was subsequently treated with further RAI or other therapy for this thymic uptake. Disease-free status was defined as the absence of detectable residual or metastatic disease. During the study, this definition changed, but the definitions over time included normal stimulated or unstimulated RAI scan, as well as normal stimulated or unstimulated thyroglobulin, and, on an individual patient basis, sometimes normal thyroid ultrasonography or other imaging modalities. Tumor staging was performed as recommended by the National Thyroid Cancer Treatment Cooperative Study Group, based on the TNM tumor classification system (9). All TC patients aged less than 45 years—thus, all patients in this study—are classified as stage 1 (any T, any N, M0) or stage 2 (any T, any N, M1). Staging was assessed using data collected after thyroidectomy, and included data from initial diagnostic RAI scan, before administration of the first therapeutic RAI dose.

Statistical analyses were performed using SAS (9.1.2). T-tests, χ ²-tests, and analysis of variance were used as appropriate depending on the continuous or discrete nature of the data. Pearson correlations were performed to assess the relationship between continuous variables. A multiple regression was performed on selected clinically important variables. Significance was set at p < 0.05.

Results

Demographics

We identified 61 eligible cases: 7 were excluded because of postsurgical transfer for follow-up to another institution. Data of 54 patients were analyzed, including 36 girls and 18 boys (p = 0.01). Mean age at presentation was 13.0 ± 3.5 years and 13.4 ± 2.8 years, in boys and girls, respectively (p = 0.66). Median duration of follow-up was 2.9 years (range 0.1–11.5), and 35/54 (64.8%) patients were transitioned to adult care by the time of study end.

The number of patients in whom the disease was newly diagnosed remained constant from 1984 to 2003, but increased nonsignificantly (p = 0.38) from 2004 to 2006 (Fig. 1). Three additional patients were found to have the disease in each of 2007 and 2008, and five patients from January to June 2009. There was no statistically significant increase in diagnoses of TC during the period of observation.

FIG. 1.

Number of patients presenting with differentiated thyroid cancer per year of study: 1984–2006. Note: For comparison, three patients were found to have the disease during 2007 and 2008, and five patients during the first 6 months of 2009 (data not shown). There was no significant increase in the number of cases of thyroid cancer diagnosed annually (p = 0.38).

Clinical presentation

Forty-four (78%) patients presented with an asymptomatic thyroid nodule, and 12 (22%) had cervical lymphadenopathy, including 1 who presented with a goiter on clinical examination. Mean tumor size was 2.3 ± 1.5 cm (range 0.23–9.4 cm). Eleven tumors were less than 1.5 cm, and these included four with lymph node and one with distant metastases at presentation, and one with the development of possible thymic metastases during follow-up.

The age at presentation did not change during the study period (data not shown). Children with lymph node or distant metastases at presentation were younger than those without metastases at presentation (12.4 ± 3.1 [n = 27] vs. 14.2 ± 2.6 years [n = 27], respectively; p = 0.02), despite similar tumor size (2.1 ± 1.0 cm vs. 2.5 ± 1.9 cm, respectively; p = 0.34). Similarly, children who developed metastases during follow-up were younger than those without metastases (12.4 ± 3.1 years [n = 6] vs. 14.2 ± 2.6 years [n = 48]; p = 0.02), with no difference in initial tumor size (3.4 ± 3.1 vs. 2.3 ± 1.4 cm, respectively; p = 0.11). Pubertal data were incomplete in medical records, and hence could not be evaluated. However, there was no difference in pathological evidence of tumor aggressiveness (defined by extrathyroid extension, lymph node involvement, or lung metastases at presentation) between children aged less than (n = 9) or greater than (n = 45) 10 years (Table 1). More children aged greater than 10 years at presentation had tumor at the resection margin. However, all children had TT at either initial procedure or at CT.

Table 1.

Comparison of Features of Disease Aggressiveness Between Children Aged Less Than 10 Years (n = 9) and Children Aged Greater Than 10 Years

	Whole group, n = 54	Age < 10 years, n = 9	Age > 10 years, n = 45	p-Value
Background thyroid inflammation	23	4	19	1.00
Multiple tumor foci in one lobe	26	4	22	1.00
Tumor foci in both lobes	15	5	10	0.10
Vascular invasion	18	4	14	0.46
Capsular invasion	35	8	27	0.14
Invasion of soft tissues	15	5	10	0.10
Lymphatic invasion	2	0	2	1.00
Tumor at the surgical resection margins	12	5	7	0.02
Tumor metastases to lymph nodes	25	5	17	0.46

Lymph node metastases at presentation were associated with PTC histology. Younger age was associated with a trend to distant metastases at presentation (p = 0.06) and development of subsequent distant metastases during follow-up (p = 0.06). Male sex was associated with the development of distant metastases during follow-up (p = 0.015), but not with histological subtype (p = 0.6).

Histology

Forty (74%) patients had PTC, 1 (2%) had diffuse sclerosing-PTC, 7 (13%) had F-PTC, and 6 (11%) had FTC. PTC tumors were more likely to be associated with lymph node metastases (p < 0.05) and vascular invasion (p = 0.047). Tumors were bilateral in 15/54 (27.8%) cases, and there were multiple foci of tumor in one lobe in 26/54 (48.1%) cases. Multiple foci of tumor and capsular invasion were more likely to be associated with lymph node or distant metastases at presentation (p = 0.044 and p = 0.019, respectively). Pathological evidence of background thyroiditis was evident in 23/54 (43%) thyroid specimens, including all 5 with clinical thyroid disease. Further details of the patients with FTC are given in Table 2.

Table 2.

Details of Patients with Follicular Thyroid Cancer (n = 6)

Descriptor	Number of patients (total n = 6)
Age, years (mean ± SD, range)	11.3 ± 3.0, 7–15
Sex	3 boys, 3 girls
Tumor size, cm (mean ± SD, range)^a	3.1 ± 1.9, 0.25–5
Positive lymph nodes at presentation^a	0
Distant metastases at presentation^b	0
Distant metastases during follow-up^b	0
Surgical procedure: total thyroidectomy	6
Received RAI	6
Bilateral disease^a	0
Multifocal disease^a	0
Capsular and/or vascular invasion^a	6
Soft tissue invasion^a	0
Lymphatic invasion^a	0
Background thyroiditis^a	2
TNM staging	T1N0M0 (n = 1)
	T2N0M0 (n = 2)
	T3N0M0 (n = 3)

Defined by pathological examination.

Defined by either pathological examination or radiological/biochemical investigations.

RAI, radioactive iodine; SD, standard deviation.

Stage

Forty-six (85%) patients had stage 1 disease and 8 (15%) had stage 2 disease. TNM staging details of all patients are summarized in Table 3.

Table 3.

TNM Staging Details of Patients (n = 54)

TNM staging
T	N M	No. of patients (%)
T1	N0 M0	9 (17)
	N0 M1	1 (2)
	N1 M0	8 (15)
T2	N0 M0	13 (24)
	N1 M0	4 (7)
	N1 M1	1 (2)
T3	N0 M0	3 (5)
	N0 M1	1 (2)
	N1 M0	1 (2)
T4	N0 M0	3 (6)
	N0 M1	2 (4)
	N1 M0	5 (9)
	N1 M1	2 (4)
TX	N1 M1	1 (2)

Risk factors for TC

Five of 54 (9%) patients had received external therapeutic radiation for treatment of previous malignancies, including acute lymphoblastic leukemia (n = 2), Burkitt's lymphoma (n = 1), Ewings Sarcoma (n = 1), and Wilms tumor (n = 1). Three patients had received external therapeutic radiation as part of their initial therapy, and two patients for treatment of relapsed disease. Data on the radiation field were not collected, so it is unclear which patients had direct exposure of the thyroid during their treatment. Further details are given in Table 4. The mean age at diagnosis was 15.7 ± 2.0 years. Median tumor diameter was 1.5 ± 0.6 cm. Median time from diagnosis of primary malignancy to diagnosis of secondary TC was 5.1 ± 3.4 years (range 5–12 years). TC therapy consisted of complete thyroidectomy and RAI in all patients. One patient developed lung metastases 15 months after initial presentation, and required further RAI therapy.

Table 4.

Details of Patients with History of Radiotherapy Before Diagnosis of Thyroid Cancer (n = 5)

Descriptor	No. of patients (total n = 5)
Age, years (mean ± SD, range)	15.7 ± 2.0, 9–18
Sex	2 boys, 3 girls
Histological tumor type	5 papillary thyroid cancer
Tumor size, cm (mean ± SD, range)^a	1.7 ± 0.5, 1–2.2
Positive lymph nodes at presentation^a	3
Distant metastases at presentation^b	0
Distant metastases during follow-up^b	1 (lung metastases)
Surgical procedure: total thyroidectomy	5
Received RAI	5
Bilateral disease^a	2
Multifocal disease^a	4
Capsular invasion^a	4
Vascular invasion^a	0
Soft tissue invasion^a	0
Lymphatic invasion^a	0
Background thyroiditis^a	0
TNM staging	T1N0M0 (n = 1)
	T2N0M0 (n = 1)
	T2N1M0 (n = 1)
	T4N1M0 (n = 1)
	TXN1M0 (n = 1)

Defined by pathological examination.

Defined by either pathological examination or radiological/biochemical investigations.

Six of 54 (11%) patients were not born in Canada, including 3 born in Eastern European regions with high TC endemnicity. For these three patients, there was no family history of thyroid condition recorded in medical notes.

Four (7%) patients, all with PTC, had a family history of PTC, affecting a grandparent (n = 2) or a mother (n = 2). Ten of the remaining 50 patients (20%) had a family history of other thyroid disorders. No patient had a family history of diagnosed genetic syndromes predisposing to PTC. However, one patient had a strong family history of cancer (breast, colon, and lung) and was referred for further cancer genetics studies (results pending).

Past thyroid medical history

Five (9%) patients were followed up for preexisting thyroid disease, including Grave's disease requiring methimazole therapy (n = 2), Hashimoto's thyroiditis requiring levo-thyroxine (n = 1), and asymptomatic goiter with no treatment (n = 1). One patient had a previous thyroid nodule biopsy, which was negative for malignancy but was followed up in endocrinology clinic thereafter.

Treatment

Surgery

All patients underwent TT, including 41 (76%) patients who had TT as the initial procedure and 13 (24%) who underwent CT as a second surgical procedure, after review of the final pathology result of the initial hemithyroidectomy specimen.

Eleven (20%) patients had postoperative complications, all with hypocalcemia (transient in 4 and persistent in 7) and 1 who also had transient hoarseness, which recovered spontaneously. There was no difference in the occurrence of complications after TT or CT.

Radioactive iodine

Fifty-three (98%) patients underwent therapeutic RAI with a range of 1–4 doses, and a mean cumulative I-131 dose of 133 ± 72 mCi. The RAI-untreated patient presented with 1.5-cm-diameter PTC at age 15.8 years, and had no lymph nodes or distant metastases at either initial presentation or during subsequent follow-up investigations. Eight (15%) demonstrated nonthyroidal postoperative uptake of RAI, of whom six had lymph node metastases at the time of surgery. Of eight patients with evidence of distant metastases, six had PTC and two had F-PTC (including one with lung metastases and one with bone metastases).

Outcome

Two patients with initially localized disease (both with 3-cm-diameter PTC primary tumors, with negative lymph nodes and no detected distant metastases) developed lung metastases 3 and 7 years after their initial therapy. There were no other high-risk features in these patients. Of the 22 patients who presented with lymph node metastases, 5 (23%) developed distant lesions, including 3 with lung and 2 with thymic metastases. Of the eight patients who presented with distant metastases, two (25%) went on to develop new distant lesions (thymus and lungs, respectively). Mean time to development of metastases was 1.3 ± 1.1 years from initial diagnosis. Mean RAI dose delivered at initial presentation was 90.1 ± 20.0 mCi. Thirty-three (61%) patients transitioned to adult care during the study period, two of whom developed lung metastases before transition.

Patients found to have TC in the earlier years of the study were more likely to present with advanced disease including lymph node or distant metastases, compared with patients who presented in later years (χ ² = 15.9, p = 0.0003). However, there was no difference between the year of diagnosis and development of subsequent metastases during follow-up (χ ² = 1.27, p = 0.53).

Multiple regression

On multiple logistic regression, PTC, as compared to other histological subtypes, trended toward association with larger tumor diameter (p < 0.05). Tumor diameter did not correlate with age at presentation, sex, lymph nodes at presentation, and development of lymph node or distant metastases during follow-up.

Discussion

The annual number of children found to have TC at our institution did not increase during the study period, but we did identify a nonsignificant trend toward larger tumors in recent years, which may be a chance finding. Although it is possible that the regional incidence of TC in childhood is increasing, our hospital is the only tertiary referral centre for thyroid conditions and thyroid surgery in children in the Greater Toronto Area, a catchment area of approximately 1,035,000 children aged less than 18 years (10). Our data likely reflect accurately TC incidence in the Greater Toronto Area and are consistent with the findings of others (4,6). Nonetheless, with increasing numbers of childhood cancer survivors who were treated with radiotherapy, TC as a second malignancy is increasing (11,12). The incidence of TC in our study is affected by an ascertainment bias, as there was not a uniform approach to screening and diagnosing TC in this population during the study period.

The current study represents one of the largest single-institution childhood TC cohorts reported to date. Our study is limited by the inherent biases of a retrospective review, and lacks details on specific management practices, which may have changed during the study period. Also, descriptions of metastases and primary disease have been given, based in part on the decision of the original treating physician, supplemented with historic evidence interpreted by the authors, when possible. Further, our definition of metastases that developed during follow-up does not take into account iodine nonavid disease, and the possibility of false-positive (physiologic) thymic uptake could not be accurately determined in this retrospective study. However, given the changes in clinical practice and improvements in investigations over the study period, it is likely that clinical decision making was not uniform throughout the study period. Finally, long-term survival data are not available because data were not available once patients were transitioned to adult care. Nonetheless, some important features of the presentation and early course of pediatric TC were identified.

Previous studies have suggested that pediatric TC presents with more advanced disease compared with adults (5). A recent retrospective study suggested that TC is more aggressive in prepubertal children (13), with an increased incidence of extrathyroidal extension, lymph node involvement, and lung metastases at presentation compared with postpubertal children. In our retrospective study, we were unable to address the issue of puberty due to incomplete data. However, we did identify that younger children, less than 10 years old, were more likely to have metastatic lymph nodes at presentation compared with older children, greater than 10 years old. Further, there was a trend toward younger children being more likely to present with distant metastases and to develop metastases during follow-up. However, there were no deaths during the study period. Our data support previous suggestions of more metastases in younger children, but without consequent increased mortality. Nonetheless, consideration of TT followed by remnant radioactive ablation in young patients may be appropriate, regardless of initial tumor size (5), although there is no evidence of inferior overall survival of pediatric patients compared with adults. Future studies will be required to determine whether variations in treatment strategy change the chance of recurrent disease. Our data also indicate that boys were more likely to develop subsequent metastases. This is in keeping with previous published studies that have identified decreased disease-free survival associated with male sex (14). Genetic conditions associated with differentiated TC include familial adenomatous polyposis and PTC, and Cowden's syndrome and FTC. There was no genetic condition diagnosed in our cohort, although genetic studies from one family are still pending. More than 70% of PTC are associated with genetic alterations in the RET/PTC, BRAF, and RAS genes, leading to activation of the MAPK pathway (15). Some studies have examined genetic differences in pediatric and adult TC to date. These and additional future studies would be useful in correlating biological features with clinical behavior of the tumor.

There was a background of clinical thyroid disease in 9% of our patients. In three of these five patients, methimazole or levo-thyroxine was used to treat the thyroid disorder before development of a dominant malignant nodule. Further, 23 (43%) pathology specimens had evidence of background thyroiditis, including all 5 with clinical thyroid disease. These data suggest that there is a high incidence of thyroiditis in children with TC in our cohort. A recent multicenter retrospective study (11) reported 365 children followed up for autoimmune thyroid disease, of whom 31.5% (115/365) had nodular thyroid disease and 3% (11/365) had TC. Thus, autoimmune inflammation of the thyroid gland is common in TC, and, conversely, it is possible that TC may evolve from preexisting autoimmune thyroid disease.

Five (10%) patients in the current cohort developed TC as a secondary malignancy after treatment of a primary malignancy, of whom three had received irradiation as part of their initial therapy. There is a high risk of TC as a second malignancy in children who have received radiotherapy to the head, neck, or upper thorax (12). Some clinicians advocate following up survivors of pediatric cancer with regular ultrasonography to screen for thyroid nodules, and biopsy for suspicious nodules. One report identified five thyroid malignancies (all PTC) in 129 individuals screened following a previous malignancy (11). The median age at diagnosis of TC was 25.1 years (range 17.5–43.8 years) and the median duration from first malignancy to diagnosis of TC was 15.8 years (range 6.1–34.8 years). This contrasts with patients in our study, who presented with TC at a median age of 13 years (range 13–17.8 years) and at median of 8.8 years after their first malignancy (range 5–12 years). This might reflect a referral bias for thyroid ultrasonography in the endocrinology late-effects clinic at our institution in the recent years of this study. However, in our study, all those with an initial lesion identified on ultrasonography had a palpable thyroid nodule or lymph node on examination. Further, 11% of patients had immigrated from areas of high TC endemnicity in Eastern Europe. Although ethnicity does not indicate causality of TC tumors, clinicians should be aware of increased TC risk among individuals with therapeutic or environmental radiation exposure (16). TC tumors less than 1 cm in diameter are increasingly difficult to palpate. Thus, although screening ultrasonography may facilitate earlier detection of smaller tumors, there is insufficient evidence to recommend their routine use. If they are routinely performed, then future research and institutional reviews should evaluate their contribution to both earlier diagnosis and altered outcomes.

The risk of operative complications such as recurrent laryngeal nerve damage or hypoparathyroidism after TT or CT was low in the current study. However, our institution is a large tertiary-care referral center for pediatric surgery. Thyroid surgery should be performed by skilled surgeons with experience in pediatric thyroid surgery. The frequency with which an individual surgeon performs thyroidectomies is associated with a lower likelihood of operative complications (17).

A recent retrospective study suggested that TC is more aggressive in prepubertal children (13), with prepubertal children having more extrathyroid extension, more lymph node involvement, and more lung metastases at presentation than postpubertal children. Our retrospective study did not allow us to address puberty as a variable, owing to incomplete medical records. However, using age 10 years as a surrogate marker of puberty, there were no differences in tumor size or aggressiveness between patients aged less than or greater than 10 years in our study. Future prospective studies should re-address this issue.

Conclusion

We have described a large cohort of children treated for TC at a single institution over 23 years. Our data suggest that younger children are more likely to have lymph node disease at presentation, and possibly more likely to develop subsequent metastases. We did not identify a statistically significant increase in incidence in recent years of the study. Treatment for a prior malignancy and a history of thyroiditis were identified as common features among pediatric TC.

To facilitate improved understanding of the biological differences of TC related to age, there is need to collect and store tissue from thyroidectomies in tissue bank repositories. Close collaboration between centers for pediatric and adult TC care will allow long-term follow-up of patients after transition to adulthood. This is necessary to gain information on the long-term outcome of pediatric TC and to determine whether age-specific approaches to therapy are justified. Future prospective multicenter pediatric studies are required to answer questions regarding the natural history of this disease in pediatric patients.

Footnotes

Disclosure Statement

This study was not funded. The authors declare that no competing financial interests exist.

References

Canadian Cancer Statistics. www.cancer.ca/Canada-wide/About%20cancer/Cancer%20statistics/Canadian%20Cancer%20Statistics.aspx?sc_lang = en. 2009 November 25.

Liu

, Semenciw

, Ugnat

, Mao

. 2001. Increasing thyroid cancer incidence in Canada, 1970–1996: time trends and age-period-cohort effects. Br J Cancer, 85:1335–1339.

Colonna

, Grosclaude

, Remontet

, Schvartz

, Mace-Lesech

, Velten

, Guizard

, Tretarre

, Buemi

, Arveux

, Esteve

. 2002. Incidence of thyroid cancer in adults recorded by French cancer registries (1978–1997) Eur J Cancer, 38:1762–1768.

Hameed

, Zacharin

. 2005. Changing face of paediatric and adolescent thyroid cancer. J Paediatr Child Health, 41:572–574.

Rachmiel

, Charron

, Gupta

, Hamilton

, Wherrett

, Forte

, Daneman

. 2006. Evidence-based review of treatment and follow up of pediatric patients with differentiated thyroid carcinoma. J Pediatr Endocrinol Metab, 19:1377–1393.

Lafferty

, Batch

. 1997. Thyroid nodules in childhood and adolescence—thirty years of experience. J Pediatr Endocrinol Metab, 10:479–489.

Zimmerman

, Hay

, Gough

, Goellner

, Ryan

, Grant

, McConahey

. 1988. Papillary thyroid carcinoma in children and adults: long-term follow-up of 1039 patients conservatively treated at one institution during three decades. Surgery, 104:1157–1166.

Cooper

, Doherty

, Haugen

, Kloos

, Lee

, Mandel

, Mazzaferri

, McIver

, Pacini

, Schlumberger

, Sherman

, Steward

, Tuttle

. 2009. Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid, 19:1167–1214.

Jonklaas

, Sarlis

, Litofsky

, Ain

, Bigos

, Brierley

, Cooper

, Haugen

, Ladenson

, Magner

, Robbins

, Ross

, Skarulis

, Maxon

, Sherman

. 2006. Outcomes of patients with differentiated thyroid carcinoma following initial therapy. Thyroid, 16:1229–1242.

10.

www.toronto.ca/reportcardonchildren/pdf/factsheet1.pdfdownloaded on October 222009.

11.

Brignardello

, Corrias

, Isolato

, Palestini

, Cordero di Montezemolo

, Fagioli

, Boccuzzi

. 2008. Ultrasound screening for thyroid carcinoma in childhood cancer survivors: a case series. J Clin Endocrinol Metab, 93:4840–4843.

12.

Sigurdson AJ

, Mertens

, Stovall

, Smith

, Liu

, Berkow

, Hammond

, Neglia

, Meadows

, Sklar

, Robison

, Inskip

. 2005. Primary thyroid cancer after a first tumour in childhood (the Childhood Cancer Survivor Study): a nested case-control study. Lancet, 365:2012–2023.

13.

Lazar

, Lebenthal

, Steinmetz

, Yackobovitch-Gavan

, Phillip

. 2009. Differentiated thyroid carcinoma in pediatric patients: comparison of presentation and course between pre-pubertal children and adolescents. J Pediatr, 154:708–714.

14.

Shapiro

, Bhattacharyya

. 2005. Population-based outcomes for pediatric thyroid carcinoma. Laryngoscope, 115:337–340.

15.

Nikiforova

, Nikiforov

. 2008. Molecular genetics of thyroid cancer: implications for diagnosis, treatment and prognosis. Expert Rev Mol Diagn, 8:83–95.

16.

Cardis

, Krewski

, Boniol

, Drozdovitch

, Darby

, Gilbert

, Akiba

, Benichou

, Ferlay

, Gandini

, Hill

, Howe

, Kesminiene

, Moser

, Sanchez

, Storm

, Voisin

, Boyle

. 2006. Estimates of the cancer burden in Europe from radioactive fallout from the Chernobyl accident. Int J Cancer, 119:1224–1235.

17.

Tuggle

, Roman

, Wang

, Boudourakis

, Thomas

, Udelsman

, Ann Sosa

. 2008. Pediatric endocrine surgery: who is operating on our children? Surgery, 144:869–877discussion 877.