Radioactive iodine in the management of medullary carcinoma of the thyroid

Abstract

Objectives:

Although it is generally accepted that medullary thyroid cancer (MTC) cells do not take up iodine, there are reports indicating that this can occur. Additionally, the potential for radioactive iodine (RAI) to reduce the risk of recurrence within the thyroid bed following thyroid remnant ablation in MTC is uncertain. A systematic review was therefore undertaken.

Methods:

Studies of patients with MTC of any age or stage receiving RAI, either as adjuvant postoperative treatment or primary treatment for unresectable disease, or as treatment for recurrent or metastatic disease were eligible for inclusion.

Randomised and non-randomised studies were identified by electronic searching of Medline and Embase databases. A risk of bias assessment (ROBINS-I) was carried out for each study.

Outcome measures sought included overall survival, locoregional relapse-free survival, rates of locoregional recurrence, and changes in serum calcitonin.

A protocol was registered with PROSPERO before the systematic review was undertaken.

Results:

There were no randomised studies. Ten non-randomised studies (525 patients) and ten case reports (21 patients) met the inclusion criteria, with all studies containing a high risk of bias. There were case reports reporting responses to RAI, both as adjuvant treatment and for recurrent/metastatic disease.

Conclusions:

The proportion of metastatic or recurrent MTC which take up iodine remains unknown. A possible role of RAI ablation for patients with localised MTC and raised calcitonin post-thyroidectomy should be explored.

Advances in knowledge:

Although there is insufficient data to recommend changes to current treatment policies, this review suggests avenues for further research.

Introduction

Whereas radioactive iodine ablation (RAI) is standard therapy following thyroidectomy in selected cases of differentiated thyroid cancer (DTC, i.e. papillary or follicular thyroid cancer), RAI is regarded as being futile in medullary thyroid cancer (MTC), the rationale being that as MTC arises from parafollicular C-cells, and not thyroid follicular cells, cancer cells have no capacity to take up radioactive iodine and are not susceptible to its tumoricidal effects. National guidelines therefore do not support the use of radioactive iodine in MTC. ^1,2

The tumoricidal effects of RAI are related to absorbed radiation dose. This dose is related both to the avidity of uptake and size of the focus taking up RAI. As the range of the β-particles emitted by ¹³¹I is approximately 0.8 mm, some of the dose from ¹³¹I concentrated in very small metastases may be deposited in immediately adjacent normal tissues rather than the deposit itself. Thyroid remnants being larger, and uptake into remaining thyroid follicular cells being more avid, absorbed doses within the thyroid remnant are much higher (up to 300 Gy) ³ and effective in not just ablating the thyroid remnant but also killing any cancer cells in the immediate vicinity regardless of their potential to take up iodine – the so-called bystander effect. Following optimal surgery for MTC, there is a high risk of residual cancer within the thyroid bed when there is extrathyroidal extension, as seen in 26% of excised cancers. ⁴ While postoperative RAI would not be expected to have any impact on residual MTC in nodal areas or areas away from the thyroid bed, cells within the thyroid bed would receive a high radiation dose by the bystander effect. In a systematic review of the effectiveness of post-operative external beam radiotherapy, there was a reduction in risk of locoregional recurrence of at least 37% ⁵ suggesting that MTC is a relatively radiosensitive cancer, probably to a similar degree to DTC. This suggests that if an adequate radiation dose can be delivered, RAI might be of clinical benefit. To benefit from this, foci of residual MTC would need to be located in or very close to residual thyroid tissue (i.e., within the thyroid bed).

However, when RAI is given as treatment for recurrent or metastatic disease, cell kill is dependent on uptake by cancer cells themselves, although this is somewhat variable in DTC with up approximately 75% of follicular or follicular-variant papillary cancers taking up RAI compared to approximately 20% of classical variant papillary or Hurthle cell cancers. ⁶ Iodine uptake (as an iodide ion) occurs via the sodium-iodide symporter (NIS). It is now known that many other normal tissues express NIS (most notably salivary gland and breast) and many tumours express NIS at higher levels than their host tissues (for example tumours arising in lung, breast and cervix). ⁷ However, to be effective in iodine transport, NIS needs to be located on the plasma membrane (as opposed to within the cytoplasm) and this process of subcellular trafficking is largely regulated by TSH. ⁷ Loss of NIS activity, either by reduced expression of NIS or location away from the cell membrane, can contribute to iodine-refractoriness. ⁸ NIS activity and associated iodine uptake have been observed in MTC stem cells ⁹ and following virally-mediated NIS gene transfer in both MTC cell xenografts ¹⁰ and in human subjects within ongoing clinical trials. ¹¹ NIS activity in MTC cells can be upregulated by all-trans retinoic acid further increasing the potential for iodine uptake. ⁹

Separate from the process of potential iodine uptake into MTC cells, are reports of iodine uptake by cancers of mixed histology (i.e., mixed medullary-papillary and mixed medullary-follicular). In these studies, it is generally believed that the iodine uptake into sites of metastasis occurs via the papillary/follicular elements of these tumours, ¹² although in some mixed tumours, cells have been seen to stain for both thyroglobulin and calcitonin. ¹³ For this reason, mixed tumours have been specifically excluded from this review.

The purpose of this review is to examine the evidence for clinical benefit from RAI in MTC, whether in the context of adjuvant treatment or inoperable, recurrent or metastatic disease, and identify areas where further research might be worthwhile.

Methods

A protocol for this review was registered with PROSPERO ¹⁴ prior to the commencement of the review and the review was carried out in line with the PRISMA 2020 statement ¹⁵ and the SWiM reporting guideline. ¹⁶ The PRISMA 2020 checklist is shown in Supplementary Table 1 and the PRISMA 2020 Checklist for Abstracts shown in Supplementary Table 2.

Criteria for considering studies for this review

Types of studies, participants and interventions

Randomised or non-randomised studies of males or females with medullary thyroid carcinoma (familial or sporadic) of any age or stage receiving RAI, either postoperatively, as primary treatment for unresectable disease or as treatment for recurrent or metastatic disease.

Outcome measures

Iodine uptake into thyroid tissue or foci of residual or recurrent disease.

Effects of RAI on serum calcitonin.

Incidence of local, locoregional or distant metastasis; overall and locoregional progression-free survival at five and ten years.

Search strategy for identification of studies

Electronic searches of Medline and Embase databases were carried out and updated in January 2023, with identification of further studies from references cited in the papers identified by electronic searching. The search strategy is shown in Supplementary Figure 1. There was no restriction by date of publication, language or duration of follow-up. Studies reporting cases with mixed histology (i.e., mixed medullary-follicular or medullary-papillary) were excluded.

Data extraction and analysis

Data were extracted from each paper and summarised. Studies were separated into cohort studies, retrospective studies and case reports.

Risk of bias (quality) assessment

Formal risk of bias assessment was undertaken for each study using the ROBINS-I assessment tool. ¹⁷ This assesses risk of bias in six domains (patient selection, treatment allocation, measurement of outcomes, reporting of results, missing data and risk of bias due to confounding).

Results

No randomised studies were identified. There were two cohort studies, eight retrospective studies and ten case reports as shown in the PRISMA flow diagram (Supplementary Figure 2).

One population-based cohort study using data from the Surveillance Epidemiology and End Results (SEER) program reported that 51/1344 patients (4.0%) with MTC received postoperative RAI between 1973–2006. ¹⁸ A second study using data from the US National Cancer Database found that 217/6375 MTC patients (3.4%) received postoperative RAI between 1995–2011. ¹⁹ Neither study provided outcome data with respect to RAI use.

Retrospective case series

The eight retrospective case series, including 525 patients with MTC (240 men, 283 women, two not recorded; mean age 41) ^20–27 differed in the proportion of patients receiving RAI (Table 1). Family history was reported in six studies (211/465 patients; 45%). RAI was given postoperatively in all but one study where the timing is uncertain. ²⁴ Follow-up appears to have been predominantly clinical with routine calcitonin estimation with additional ¹³¹I-MIBG ²⁵ or ^99mTc-V-DMSA imaging ²⁷ in selected patients. Chemotherapy was administered to some patients. ^24,27 No patient received targeted therapy. Only two studies reported outcomes with and without RAI. The MD Anderson study ²³ found an almost identical 10 year survival in 15 patients receiving postoperative RAI and in 84 patients not receiving RAI (75% vs  74%), although there were fewer patients with nodal involvement in the RAI group (27% vs  43% respectively; p = 0.36). In the same study, relapse at any site was unaffected by RAI use in those without nodal involvement (36% vs  35%, with and without RAI). In two further studies, RAI had no significant effect on overall survival in multivariate analysis. ^26,27

Table 1.

Retrospective series of MTC patients treated with and without radioactive iodine (¹³¹I)

Series (Reference)	Years Studied	Number with MTC	Number (%) Receiving Radioactive Iodine	Overall Survival at 10 Years (%) without Iodine	Overall Survival at 10 Years (%) with Iodine	Number of Locoregional Recurrences without Iodine (%)	Number of Locoregional Recurrences with Iodine (%)
Dordrecht 20	1980–2007	230	51 (22)			55/179 (31)	18/51 (35)
Ankara 21	ns	7	7 (100)				3/7 (43)
Leiden 22	prior to 1984	20	20 (100)				0/20 (0)
MDAnderson 23	1951–1982	99	15 (15)	74	75		4/15 (27)
Hammersmith/Kings 24	1950–1966	17	10 (59)
Vellore 25	1982–2002	40	15 (38)
Vilnius 26	1977–2006	59	33 (58)
Busto Arsizio 27	1970–1992	53	9 (17)			4/53 (8)
Total		525	160 (31)

ns, not stated.

In a multicentre study of 230 MTC patients from the Netherlands, there was no significant difference in the number experiencing locoregional recurrence (35% vs  31%, with and without RAI; p = 0.65) although details of prognostic factors between the two groups were not reported. ²⁰

In a study of seven patients receiving postoperative RAI, significant reductions in basal and pentagastrin-stimulated calcitonin following RAI were observed in all three with disease confined to the thyroid but in only one of four who had lymph node involvement. ²¹

In a study of 20 patients, eight of nine with localised disease had normal serum calcitonin postoperatively and the ninth normalised after RAI. ²² However, of 11 with lymph node involvement, six had raised levels postoperatively, with one of these normalising after RAI.

RAI uptake within the thyroid bed was reported in all 35 patients in two studies. ^22,23 An earlier third study reported “no uptake in tumour or its secondaries”. ²⁴

Case reports

The ten case reports (21 patients; median age 50 years, range 16–78; Table 2) were divided into those reporting on adjuvant postoperative RAI (similar to that reported in the retrospective series), ^28–33 and those reporting on RAI treatment for inoperable, recurrent or metastatic MTC. ^32–37 Three case reports specifically confirmed the presence of “pure” MTC (i.e., the absence of any mixed histology) ^34,36,37 and three reported negative immunostaining for thyroglobulin. ^29,33,34 Imaging studies undertaken prior to the administration of RAI are shown in Supplementary Table 3.

Table 2.

Summary of clinical details of the 21 patients included in the ten case-reports. One patient received treatment both in the adjuvant setting and for recurrence in the neck seven years later

Setting	Total Evaluable	Male / Female	Familial / Sporadic	Total Thyroidectomy	Neck Dissection	Adjuvant Radiotherapy
Adjuvant	10	3/7	2/6	9/9 a	3/5	4/10
Inoperable / Recurrent / Metastatic	11 b	6/5	1/2	2/5	2/3	4/8
Total	21	9/12	3/8	11/14	5/8	7/18

^aextent of surgery not specified in remaining patient

^balso includes a twelfth patient previously treated in the adjuvant setting

Details of those treated in the adjuvant setting are shown in Table 3. Four patients with elevated calcitonin following surgery reported reduced calcitonin levels after RAI, two with localised disease and two with nodal involvement. Uptake into the thyroid remnant was a common feature of these studies. Subsequent treatment in this group included chemotherapy in two patients and further surgery in one.

Table 3.

Details of case reports of patients treated with radioactive iodine (¹³¹I) as adjuvant (postoperative) treatment

Reference	Nodal Staging (N)	Iodine Activity Iodine Administered (GBq)	Sites of Iodine Uptake	Uptake on Subsequent Iodine Scans	Calcitonin before Iodine (pg/ml)	Calcitonin after Iodine (pg/ml)
29	1	5.5	thyroid bed	none 1 year later	890	786
30	0	1.1	thyroglossal remnant			reduced a
31	0	5.5	thyroid bed		680	50
28	1	ns	“no distant”		162	40.5
32	ns	2.7	thyroid bed
32	ns	3.4	thyroid bed
32	1	4.1	thyroid bed & superior mediastinum
32	1	3.5	thyroid bed
33 b	1	3.0, 5.5	ns	“none on later scan”
33	ns	ns	neck

ns, not stated.

^acalcitonin response to calcium infusion and pentagastrin stimulation both significantly reduced (actual value not given).

^balso received ¹³¹I as treatment for neck recurrence (Table 4).

Table 4 shows the results of RAI treatment in those with inoperable disease (two patients), metastatic disease (four patients) or locoregional recurrence (six patients). One with locoregional recurrence had also received RAI in the adjuvant setting seven years previously. Serum calcitonin fell significantly after RAI in both patients with inoperable disease, one also demonstrating iodine uptake in a metastasis in the femur. Of four patients treated for known metastases, three demonstrated iodine uptake into metastatic sites. Of those with recurrent disease in the neck, three of six demonstrated iodine uptake. Two of these received only a tracer dose, and four received one or two treatment doses. In two of these four, there was a significant response on subsequent studies with disappearance of iodine uptake accompanied by significant falls in serum calcitonin. A third patient with raised calcitonin but no demonstrable sites of uptake experienced a significant fall in calcitonin following two doses of RAI. Symptomatic improvement was noted in five of ten (excluding the two who received only a tracer dose). Diarrhoea, a frequent accompaniment to advanced disease, improved in the four patients reporting it. Other symptomatic improvements included reduction in pain or breathlessness, and improvement in vision. Subsequent treatment included chemotherapy in four, surgery in two and radiotherapy in one patient. No patient received targeted therapy either before or after RAI.

Table 4.

Details of case reports of inoperable, metastatic or recurrent MTC treated with radioactive iodine (¹³¹I)

Reference	Sites of Disease	Activity of Iodine Administered (GBq)	Sites of Iodine Uptake after Therapy Dose or Prior Tracer Dose	Response on Subsequent Iodine Scans	Calcitonin before Iodine (pg/ml)	Calcitonin after Iodine (pg/ml)	Symptomatic Response
35	inoperable (thyroid and bilateral neck)	1.85, 3.7, 3.7	ns	ns		reduced by 63% (actual values not given)	diarrhoea resolved after second dose
35	inoperable (thyroid, neck and mediastinum); metastasis in femur	3.7, 4.6, 4.6, 3.7, 3.7	thyroid left femur		19,420	8,650	pain and dyspnoea resolved; diarrhoea resolved after first dose
34	neck, bones	3.7, 4.14, 7.4	neck, bones and “internal”	regression in size of lesions	ns	ns	less pain
36	lungs, choroidal metastases	11.9	lungs		250,000	235,000	vision improved after 2 days; dyspnoea and diarrhoea improved
35	neck, lungs	ns	ns			no response	none
32	lungs	two doses (activity not stated)	ns			no response	none
37	neck and anterior mediastinum	tracer only	neck and anterior mediastinum			not applicable: tracer only	not applicable: tracer only
35	none identified	3.7, 3.7	none		141,000	67,800	no diarrhoea after 4 days
35	neck	1.85, 3.7	neck	disappearance of uptake in neck	31,900	16,380
35	neck	1.85, 2.8	neck	disappearance of uptake in neck	220,400	31,500
33	neck	tracer	none			not applicable: tracer only	not applicable: tracer only
33 *	neck	single dose (activity not stated)	none			raised; no response
summary of responses			6/12			reduced by > 48% in 5/9	5/10

ns, not stated.

^aalso received ¹³¹I as adjuvant therapy (Table 3).

Risk of bias assessment was carried out for all included studies (Supplementary Table 4). The cohort studies and retrospective studies were considered at critical risk of bias because insufficient allowance was made for the effects of nodal involvement. It is considered insufficient just to stratify nodal involvement as present or absent, as there is a continuous gradation of outcome by number of nodes involved ^38,39 so that the likelihood of higher risk patients with a greater number of involved nodes being referred for RAI cannot be excluded. All case reports were considered at critical risk of bias based on the criteria for selection of participants.

Discussion

The presence of a high risk of bias affecting all included data means that no firm conclusions can be drawn. In these studies, the overall number of patients is low and outcome data often incomplete, yet there is evidence of objective responses in a number of patients. Much of the data is from a time period when RAI was used in the treatment of MTC and when cross-sectional imaging (i.e., CT and MRI) was not routinely available. A key question therefore is whether the data reported can be considered sufficiently robust to form the basis for further investigation.

It was widely held for many years that while thyroid follicular cells originate, embryologically, from endodermal tissue in the tongue, parafollicular C-cells arose from neuroectodermal tissue. However, evidence now suggests that, unlike other neuroendocrine cells, C-cells arise from pharyngeal endoderm, and that, in mammals, thyroid follicular cells and C-cells may have a common origin in the ultimobranchial body. ⁴⁰ Historically, such apparently diverse origins would seem to make iodine uptake in MTC unlikely but the discovery of NIS in MTC cells ⁹ shows that this is a more complex issue. Cases with mixed histology (i.e., mixed medullary-papillary or mixed medullary-follicular) as distinct from the simultaneous occurrence of two separate cancers which can occur in up to 15% of MTC thyroidectomy specimens, ⁴¹ might suggest a common cell of origin. However, a study of mixed medullary-follicular cancers showed different genetic abnormalities in the two components of these tumours, more consistent with the “hostage theory” than a common cell of origin. ⁴² On the other hand, in some mixed tumours, individual tumour cells stain for both calcitonin and thyroglobulin ⁴³ consistent with an origin in pluripotent stem cells of ultimobranchial origin. It is therefore possible that both mechanisms exist in different subsets of tumours.

Although both DTC and MTC cells possess NIS activity, DTC cells differ in being able, in varying degrees, to trap iodine as part of the process of organification (as steps toward thyroid hormone synthesis). In two of the case reports, perchlorate washout was used to demonstrate that iodine was not trapped within MTC cells, ^36,37 but conversely an increase in iodine half-life in another case report ³⁴ after administration of lithium carbonate (which slows the release of organified iodine products from the thyroid ⁴⁴ ) would suggest that some organification had taken place. Without organification, iodine efflux from thyroid cells follows zero-order kinetics but with the potential for maintenance of intracellular levels by re-uptake. ⁴⁵ In vitro, NIS-transfected MTC cells accumulated iodine with organification of 4% of accumulated iodine, a somewhat surprising finding suggesting that in these cells also the organification pathway had somehow been preserved. ⁴⁶ The finding of TSH-R (TSH receptor) and TTF-1 gene expression in surgical samples and fine-needle aspirate biopsies from MTC further blurs the lines between DTC and MTC. ⁴⁷ Also, in MTC cells in vitro, overexpression of NDRG2 was related to enhanced NIS activity and increased iodine uptake. ⁴⁸ In clinical terms, the extent to which the lack of organification would reduce the absorbed radiation dose (i.e., in MTC compared to DTC) is uncertain.

The case reports describe iodine uptake into the thyroid remnants as routinely seen in in DTC, with some showing a response in serum calcitonin following remnant ablation, particularly in those without nodal involvement. Whether RAI might be appropriate in node-negative cases with elevated calcitonin postoperatively requires further exploration. Since in DTC, RAI appears effective regardless of cancer cell-type as the effectiveness of RAI depends on a bystander effect and not on any iodine-concentrating properties of the tumour cells, it would seem logical to include MTC in this category even if uptake into MTC cells is much lower than for DTC sub types (or even zero). Such use of RAI for MTC would need to be considered in conjunction with any indication for external beam radiotherapy. ^1,5

The major limitations of this review are that much of the evidence comes from a small number of case-reports (rather than prospective studies or larger institutional case series), so that the actual incidence of iodine uptake in MTC remains unknown. Given the systematic nature of this review, it is considered unlikely that studies with significant numbers of participants were missed. The presence of missing outcome data (particularly the lack of cross-sectional imaging or routine calcitonin measurements in all cases) contributes to the lack of precision about response assessment. The possibility that some responses might have occurred in association with unrecognised mixed histology cannot be excluded.

The findings of this review, while not sufficient to change current practice, do introduce an element of uncertainty into our commonly held views on MTC. On this basis, it would seem logical to explore, in a prospective and controlled manner, the extent of any RAI uptake into sites of recurrent or metastatic disease. If iodine uptake were confirmed in a subset of MTC patients, this could then be seen as an additional treatment option prior to or subsequent to currently available targeted therapies.

Conclusions

There is no evidence to suggest any change to current treatment guidelines.

Further research is needed into the ability of MTC to take up RAI.

Exploration of RAI for thyroid remnant ablation is suggested where MTC is localised to the thyroid bed and calcitonin remains elevated postoperatively.