8. Medullary thyroid carcinoma (MTC) – Thyroid Cancer

Sporadic MTC usually manifests itself in the fourth to fifth decade of life. Of the MTC diagnosed via a palpable thyroid nodule, 70% already have cervical lymph node metastases and 13% have distant metastases in the liver, lungs or bones [1063], [1064], [1065], [1066]. The 10-year survival is approx. 61-81% [1067], [1068], depending on the tumor stage at diagnosis, age, gender and pre- and postoperative Ctn levels. In the last decade, due to increasingly earlier diagnosis, more favorable tumor stages and improved surgical strategies, survival has increased overall, but also in the higher tumor stages [1067], [1069].

The most important laboratory test leading to the diagnosis of MTC is Ctn. The molecular genetic detection of an activating pathogenic germline variant in the RET gene proves the familial variant of MTC. If a hereditary form of MTC (MEN2) is suspected, molecular genetic germline diagnostics should be performed preoperatively (see Chapter Multiple endocrine neoplasia type 2 (MEN2)). If an activating pathogenic germline variant in the RET gene is present or there is a history of MEN2, a pheochromocytoma and primary hyperparathyroidism must first be excluded or confirmed biochemically before a thyroidectomy. If a pheochromocytoma is detected, it should be operated on first to avoid hypertensive crises during thyroidectomy (see Chapter Multiple endocrine neoplasia type 2 (MEN2)).

Desmoplasia is generally defined as a neoplastic-induced and thus tumor-associated new formation of collagenous connective tissue, which is mainly found in carcinomas. Desmoplasia must be distinguished from reactive new connective tissue formation in the context of physiological scarring as a result of tissue injury and/or inflammation. In the case of a histomorphological assessment of malignant neoplasia, it is not always possible to distinguish with certainty whether the intra- or peritumoral fibrosis is desmoplastic or regenerative connective tissue formation.

Fine needle aspiration, which is frequently performed when a suspicious nodule of the thyroid gland is clinically diagnosed, leads to histologically detectable regenerative effects in thyroid and tumor tissue. In everyday practice, the histopathologist recognizes such areas in the tumor on the basis of tumor capsule defects, focal fibrosis with residual bleeding, fibroblastic stromal reaction and sometimes radially arranged tumor epithelial strands with regenerative changes [1116], [1117], [1118]. Pandit et al. were able to demonstrate reparative effects in a study of 265 thyroid specimens after fine needle aspiration. In 38% of cases, suspicious changes such as regenerative tumor cell changes with nuclear atypia, capsular pseudoinvasion and metaplastic changes were also described [1117]. Bolat et al. compared 150 thyroid specimens with a condition after fine needle aspiration with 150 thyroid specimens without prior aspiration. As a result, significantly increased signs of hemorrhages, bleeding residuals, granulation tissue, pseudopapillary hyperplasia, fibrosis, calcification, capsular defects, cholesterol crystal clefts and focal microthrombus formation were detected [1118]. Against this background, the clinical information of a previous fine needle aspiration to the pathologist is necessary to allow a correct assessment of the intratumoral changes including fibrosis/desmoplasia.

The significance of stromal desmoplasia in MTC as a predictive marker of lymphonodal spread has been described several times and is also mentioned in the current version of the WHO classification for tumors of the endocrine organs [172]. In a retrospective study of a cohort of 50 hereditary and sporadic MTC, Kaserer et al. 2001 described for the first time the absence of tumor desmoplasia as a statistically significant histological marker against lymphonodal tumor spread [1119]. The same working group published further retrospective studies on 120 sporadic and 196 hereditary and sporadic MTC [175], [176]. Both studies supported the role of the lack of tumor desmoplasia as a strong positive prognostic marker and discussed the possibility of intraoperative frozen section diagnostics to avoid lateral lymph node dissection. In an extended and partly prospective study of this group, this thesis was further supported in a total of 360 hereditary and sporadic MTC. In the analysis, 17.8 % of the MTC showed no tumor desmoplasia. The patients who did not receive lateral lymph node dissection due to the diagnosis of no tumor desmoplasia made in the intraoperative frozen section nevertheless all achieved biochemical healing in the course of the procedure [178]. An independent study by another working group was able to collect similar data. Aubert et al. also showed the positive prognostic value of the absence of tumor desmoplasia in 54 sporadic MTC. A threshold value of desmoplasia of 20% of the tumor area was defined as predictive for the absence of lymphonodal spread [1120]. Only recently, a German working group was able to support these statements in a study of 139 retrospectively reviewed cases of MTC with a diameter of ≤ 25 mm. Tumors with a desmoplasia of < 10 % showed no lymph node metastases [179]. The intraoperative assessment of desmoplasia in MTC requires a high degree of experience with regard to the percentage classification of fibrosis based on the differentiation of true desmoplasia as opposed to perivascular intratumoral fibrosis in vascularized parts and amyoid deposits.

If the RET analysis is not available, a pheochromocytoma and primary hyperparathyroidism must be ruled out before surgery for MTC. Since a familial MTC cannot be ruled out or proven on the basis of family history alone, molecular genetic diagnostics are recommended for all MTCs. In approx. 3-7% of apparently "sporadic" MTC with an unremarkable family history, a pathogenic germline variant is surprisingly detectable, either as a de novo variant or as a pathogenic variant in the RET gene with low penetrance of the MTC [1121]. Molecular diagnostics for the presence of an activating pathogenic germline variant in the RET gene must be initiated at the latest after histological detection of an MTC (see [8.15] / Chapter Classification according to risk groups - follow-up strategy).

The aim of MTC surgery is to cure the tumor by completely removing the primary tumor and any affected lymph nodes.
Clinically detectable MTCs almost always already have more than 10 lymph nodes and often distant metastases and cannot usually be cured by surgery.

In the case of grossly invasive tumors, highly metastatic tumors or distant metastases, the goal is local tumor control to avoid local complications from the tumor.

For early diagnosis of MTC, basal Ctn values shall be determined in patients with thyroid nodules [1122]. MTC is considered highly probable with Ctn values above the limit of 30 pg/ml in women and above 60 pg/ml in men (see Chapter Calcitonin).

For Ctn values above the reference value of up to 30 pg/ml in women and up to 60 pg/ml in men (gray zone), there is only a relative indication for surgery, as a micro-MTC (pT1a) is only rarely found, but C-cell hyperplasia with questionable precancerous significance is found far more frequently. Lymph node involvement is rarely observed here. If surgery is not decided upon, Ctn checks are recommended every 3-6 months and later annually. An increase of more than 30 pg/ml in women and more than 60 pg/ml in men is an indication for surgery. With basal Ctn values of up to 100 pg/ml, a cure is possible in almost all cases [1073].

In the case of biochemically proven MTC, thyroidectomy with bilateral central LN dissection is recommended, as 10% of sporadic MTCs also occur multifocally (bilaterally) and micro-MTCs are also associated with lymph node metastases (exception: lack of desmoplasia detection in frozen section) [1123], [1124]. It has been shown that in women with basal Ctn ≥ 85 pg/ml and men ≥ 100 pg/ml, at least 40% of women and 53% of men have lateral lymph node metastases [1093]. The authors therefore recommend performing a (unilateral) lateral LN dissection on the tumor side in addition to thyroidectomy and bilateral central LN dissection in the initial procedure. If the basal Ctn is >200pg/ml at the latest, a contralateral lateral LN dissection shall also be performed, as contralateral LN involvement has been described [1073].

Systematic lateral (unilateral) LN dissection as a compartment dissection can also be performed in individual cases as part of a second operation [1125]. Due to the early tendency to micrometastasize, central LN dissection is an important part of the initial procedure. The associated morbidity risk is estimated to be lower than the deterioration in prognosis or local complications arising during the course of the disease due to potentially affected but retained lymph nodes. In addition, the risk of morbidity is increased in the case of central re-intervention.

A systematic, en bloc compartment dissection with quantitative lymph node analysis (ratio of removed to affected lymph nodes) should always be performed. Selective, i.e. non-systematic lymph node removal ("berry picking" of enlarged lymph nodes) should be avoided. It reduces the chance of a cure, as the surgical conditions are more difficult in the second or completion operation that is then often necessary due to the scarring, which significantly increases morbidity [1126]. Re-surgery may be necessary in special individual cases after previous adequate compartment resection.

A transsternal mediastinal lymph node dissection is only indicated if there is suspected involvement of the corresponding lymph nodes and no distant metastases, or in individual cases to avoid local complications.

The tissue reaction caused by the fine needle biopsy is microscopically similar to a desmoplastic stromal reaction and should therefore be avoided. The intraoperative frozen section is necessary, as this is the only way to prove the absence of a desmoplastic stromal reaction, which means that lateral LN dissection can be avoided in approx. 27% of patients [175].

The desmoplastic stromal reaction is a characteristic of invasive tumor growth. MTC without desmoplastic stromal reaction is a special form. Tumors without desmoplastic stromal reaction do not form LN metastases. Patients with MTC without desmoplastic stromal reaction were cured in the long-term even without LN dissection [175]. Thus, in the absence of a desmoplastic reaction in the intraoperative frozen section, no further LN dissection is necessary regardless of the preoperative basal Ctn level.

However, since the absence of a desmoplastic stromal reaction in the frozen section can only be demonstrated after organ removal in the frozen section of the primary tumor, but central bilateral LN dissection is technically easier as long as the thyroid gland is still in situ, some centers always perform a central bilateral LN dissection first [1127].

Lifelong follow-up is indicated for all MTC. A prerequisite for adequate follow-up care after surgical removal of the MTC is histological confirmation of the diagnosis, if necessary with the help of immunohistology, classification according to the pTNM scheme and thus staging, a postoperative basal Ctn and CEA level and the result of molecular genetic diagnostics for the presence of a pathogenic germline variant in the RET gene for classification into sporadic or hereditary variant of the MTC. These parameters determine the intensity and frequency of biochemical and imaging examinations as part of risk-adapted follow-up care. In the further course, risk-adapted follow-up care can be adapted based on the recording of progression by means of Ctn/CEA doubling times and imaging examinations with recording of growth according to RECIST criteria (dynamic risk stratification). Depending on the response to primary therapy, 3 groups of patients can be defined: biochemically cured, biochemically incomplete and structurally incomplete [1136], which also differ in terms of their prognosis and follow-up examination intervals [1137], [1138], [1139]:

Biochemically cured, very good result of primary therapy: the Ctn level is not measurably low postoperatively, CEA in the normal range, the tumor has been completely removed microscopically (R0 resection, usually stage I and II): then the patient can be assumed to be biochemically cured provided the histology is correct (Ctn immunohistology obligatory). In the first 2 years, six-monthly follow-up examinations with sonography of the neck, Ctn and CEA determination shall be carried out [1140]. Thereafter, if there has been no increase in Ctn, annual checks or even longer intervals can be carried out. The prognosis is very good: 87% survival after 20 years [1136].

Biochemically incomplete response: The Ctn level is usually persistently low to moderately elevated (usually ≤ 1,000 pg/ml) (mostly stage II and III). It can be assumed that MTC remains but is usually undetectable on imaging (clinically occult persistent). If the previous operation was not "adequate", a completion surgery should be performed if the postoperative Ctn level is elevated, even if the cervical imaging "staging" is negative and distant metastases have been excluded (see Chapter Indication for completion surgery) [1141], [1142]. In the further course, six-monthly follow-up checks are usually sufficient. The frequency of diagnostic imaging can be planned on the basis of the primary findings and the tumor marker doubling time; close-meshed maximum diagnostics without therapeutic consequences is not advisable; the diagnostic imaging serves to determine the tumor burden or for "staging" in connection with a new therapeutic option. In most cases, active surveillance is sufficient, as the survival rate in this group can be up to 82% after 20 years [1136].

Structurally incomplete response: The Ctn level is significantly elevated (usually ≥ 1,000 pg/ml): Here, locally infiltrating tumor tissue (pT4a-b) or distant metastasis (stage IV) mostly in the lungs, liver and/or bones (clinically manifest MTC) can be assumed. A curative approach is no longer possible. Despite distant metastases, many patients have a good quality of life for years, as tumor growth is often very slow. An "active surveillance" strategy is then often possible. In the case of progressive tumor growth, the indication for intervention can be assessed with the help of the Ctn doubling time and imaging during the course (e.g. RECIST criteria) (see Chapter Treatment of metastatic MTC (including persistence and MTC recurrence)) [1143]. All therapies in this situation are non-curative, so the indication must also take particular account of the patient's quality of life and wishes. Even with distant metastasis at the time of diagnosis, the 10-year survival rate is 44% [1144].

After thyroidectomy, levothyroxine substitution should be checked at every outpatient presentation. As MTC does not grow in a TSH-dependent manner, TSH in the reference range is sufficient.

Depending on the level of the preoperative basal Ctn value, the drop in the (high) Ctn value can take weeks. The same applies even more to CEA, which has a longer half-life per se. The nadir is reached 6-8 weeks postoperatively [1145]. If the basal Ctn cannot be measured or is in the low reference range, this indicates a biochemical cure with an excellent therapeutic outcome. The six-monthly checks with non-measurable Ctn confirm biochemical healing. If the Ctn level is persistently elevated during controls, this indicates that tumor tissue is still present; levels up to 150 pg/ml usually indicate locoregional persistence, ≥ 1,000 pg/ml often already indicate distant metastases. Imaging at Ctn ≤ 150 pg/ml is usually unsuccessful due to insufficient tumor mass; in terms of primary therapy, these patients belong to the group with a biochemically incomplete response.

The progression of the tumor disease in the further course can vary greatly and can be estimated relatively well using the Ctn and CEA doubling time (a calculation aid can be found at www.thyroid.org/professionals/calculators/CDTC.php [1146]. At least 4 Ctn/CEA values over 2 years are a prerequisite for a good statement. Patients with tumor marker doubling times ≤ 24 months also had image morphological progression in 94% of cases; if the tumor marker doubling times were ≥ 24 months, 86% of patients had no detectable tumor growth [1147]. In favorable cases (no or long tumor marker doubling time > 2 years), further checks every 12 months are sufficient, treatment of asymptomatic metastases is not necessary, "active surveillance" is appropriate. The same applies to higher Ctn values and slow progression in the group with a structurally incomplete response to primary therapy. On the other hand, if Ctn and CEA doubling times are ≤ 1 year, monitoring of Ctn, CEA and also imaging at 3-month intervals is appropriate for treatment decisions (e.g. tyrosine kinase inhibitors).

In sporadic MTC, a somatic RET variant can be detected in up to 90% of tumor tissue, most frequently the pathogenic variant M918T with high inter- and intra-tumoral heterogeneity [1148]. Somatic RET variants are rarely found in small primary tumors, whereas the prevalence increases in large primary tumors and lymph node metastases, and somatic RET variants are most frequently found in distant metastases [1149], [1150]. This has prognostic and therapeutic significance. Patients with a somatic RET M918T variant have a more aggressive course of disease and a poorer prognosis. The detection of a pathogenic RET variant is necessary for the treatment decision, especially for the selective RET kinase inhibitors that will be available in the future.

Depending on the location of the suspected metastases or the locoregional recurrence, imaging procedures that may be considered include sonography of the neck and abdomen, CT with contrast agent, magnetic resonance imaging (MRI) with liver-specific contrast agent if necessary, bone scintigram, DOPA-PET/CT, [¹⁸F]FDG-PET/CT or SSTR-PET [1151]. The intervals of the examination, every 6 months to once a year, depend on the Ctn and CEA values and their doubling time, the size and location of the residual tumor/metastases and the extent of the progression of the tumor, the patient's symptoms and, if applicable, the planned palliative therapy [1152]. Imaging only makes sense from a Ctn level of >150 pg/ml [335]. In patients who have only undergone lobectomy (Ctn not measurably low), examinations at yearly intervals and with sonography examination of the neck if Ctn is elevated up to 150 pg/ml are sufficient for long-term observation. This also applies to biochemically incomplete response to primary therapy including recurrence/completion surgery and long Ctn/CEA doubling time (> 2 years).

If Ctn levels are persistently elevated postoperatively, residual tumor tissue in the neck or distant metastases should be assumed. If primary surgery is inadequate, completion surgery (see Chapter Indication for completion surgery) shall be performed [1141], [1142]. Completion surgery should be used to attempt curative surgical treatment. However, if the following limits (postoperative persistently elevated Ctn levels up to max. 1,000 pg/ml without evidence of distant metastases, less than 5-10 affected lymph nodes, only two affected lymph node compartments and absence of lymphangiosis carcinomatosa) are exceeded, a curative approach is no longer possible. Therefore, all further therapeutic measures should be considered while weighing up the risk of morbidity and quality of life in the case of a remaining tumor with a possibly favourable prognosis. If the primary operation was "adequate" (total thyroidectomy with central and lateral lymph node dissection), a re-surgery may be indicated after imaging "staging" with locoregional MTC findings.

Re-surgeries using a palliative approach are particularly useful in cases of progressive local recurrence or painful lymph node metastases in the central neck area (near the trachea or esophagus/infiltration) in order to reduce local complications. Automatic surgery of newly occurring cervical metastases ("berry picking") should not be performed, as this leads to numerous interventions that impair the quality of life and do not affect the prognosis, but can often lead to complications such as recurrent paresis, hypoparathyroidism and arm muscle paralysis [450].

Local radiotherapy (tumor bed irradiation) may be useful for symptomatic tumor manifestations after surgical procedures have been exhausted [1160].

In the case of locally infiltrating tumor growth in the aerodigestive tract, resection of the larynx, trachea or esophagus is possible if the local situation requires resection. In most cases, advanced tumor disease with disseminated distant metastasis is present [458]. Accordingly, the indication for extended, organ-spanning resections should be viewed critically, as they often significantly impair quality of life.

In the case of a locally inoperable MTC, radiotherapy may be indicated if local symptoms are present at the same time or to avoid complications.

Bone metastases of MTC usually occur together with other distant metastases. They are usually multifocal, most frequently localized in the spine and pelvis, and are osteolytic, osteoplastic or mixed [1161], [1162]. Their treatment poses a particular challenge. Smaller, clinically unremarkable bone metastases can be observed initially. External radiotherapy or surgical intervention is indicated for painful bone metastases and/or if there is a risk of fracture, particularly in the case of vertebral metastases at risk of fracture. Systemic therapy alone, e.g. with tyrosine kinase inhibitors in advanced MTC, is often not sufficient to control bone metastasis. The use of antiresorptive therapy (bisphosphonates/denosumab) is therefore recommended, particularly in the case of disseminated bone metastasis. There are no prospective studies on MTC itself [1089], [335]. The recommendation is therefore primarily based on studies in other tumor entities that have shown a delay in the occurrence of skeletal complications [1161], [1058], [1059], [1163], [1060], [1061]. A recent study provides evidence that the complication rate and prognosis are less favorable in MTC with osteolytic metastases and that antiresorptive therapy may be appropriate [1162].

Liver metastases of MTC are often small and disseminated and usually cause few symptoms. Clearly progressive and painful liver metastases should be treated, especially if there is pronounced diarrhea that cannot be controlled with loperamide or tinctura opii. Single or large metastases should, if possible, be surgically removed, but in most cases the metastasis is extensive and treatment with a local ablative procedure is therefore often of limited efficacy. Systemic therapy with a tyrosine kinase inhibitor would be useful in the case of aggressive MTC (see Chapter Systemic drug therapy).

Lung metastases of MTC are often miliary, multiple and frequently associated with mediastinal LN. Local surgical or radiotherapeutic measures can be taken if the airways are obstructed with respiratory problems, but they also represent an indication for systemic therapy.

Brain metastases of MTC are rare and usually only become manifest with clinical neurological symptoms. Surgical resection, stereotactic radiosurgery or conventional radiotherapy may be considered.

Soft tissue metastases of MTC in the skin and breast are rare and only occur at an advanced stage. If possible, they can be treated locally (resection or radiotherapy); in the case of aggressive MTC, systemic therapy is advisable.

Cytotoxic chemotherapy is only partially effective in metastatic MTC, which is why new molecular-targeted therapeutic approaches are preferred [335]. Tyrosine kinase inhibitors (TKIs) have been investigated in trials for advanced thyroid cancer for more than 10 years. The tyrosine kinase inhibitor vandetanib has been available in Germany for the treatment of advanced, rapidly progressive and symptomatic MTC since February 2012, and cabozantinib since October 2014.

Vandetanib is an orally administered TKI that acts on the RET tyrosine kinase receptor, which is frequently mutated in MTC, but also on the EGF receptor and VEGF receptor. A randomized phase III study (n = 331 patients) with 300 mg vandetanib/day by Wells [1165] showed that progression-free survival was significantly extended by 11.2 months to 30.5 months with vandetanib vs. 19.3 months with placebo. The objective response rate was 45% according to RECIST criteria. Adverse events were more common in the vandetanib arm than in the placebo arm, the most common being diarrhea (56% vs. 26%), rash (45% vs. 11%), nausea (33% vs. 16%), hypertension (32% vs. 5%) and headache (26% vs. 9%). The most common grade 3–4 adverse events were diarrhea (11% vs. 2%), hypertension (9% vs. 0%), QT time prologation (8% vs. 1%), fatigue (6% vs. 1%). Treatment discontinuation due to side effects occurred in 28% of patients vs. 3% in the placebo arm.

A recent re-analysis of patients from the vandetanib pivotal trial who were originally included with unknown mutation status, published only in the form of a red-handed letter, showed an inadequate objective response rate of 18.2% in RET mutation-negative patients compared with an objective response rate of 51.7% in patients with a RET mutation. In view of this change in data, the efficacy of vandetanib is considered insufficient to justify the risk of side effects in RET-negative patients. Therefore, the indication for vandetanib was limited in January 2023 to the treatment of aggressive and symptomatic RET-mutated medullary thyroid cancer in patients with unresectable, locally advanced or metastatic disease.

Cabozantinib is also administered orally and acts on the RET tyrosine kinase receptor, the c-Met receptor and the EGF and VEGF receptors. In the pivotal phase III trial, 330 patients with image morphologically confirmed MTC progression after RECIST were shown to have significantly longer progression-free survival (11.2 months) compared to placebo (4.0 months) with an initial dose of 140 mg cabozantinib/day [1166]. Subgroup analyses show that patients with RET mutations have the longest extension of progression-free survival and that there are also indications of an extension of overall survival in this subgroup - even if these analyses must be interpreted with caution, as they were only collected post-hoc and have not been prospectively validated. For patients with unknown RET status and RET-negative status, efficacy in terms of prolongation of progression-free survival was also demonstrated [1166]. According to RECIST, 28% of patients treated with cabozantinib showed a partial remission. Grade 3 or 4 adverse events were reported in 69% of patients in the cabozantinib arm (most frequently diarrhea (15.9%), hand-foot syndrome 12.6%, fatigue 9.3%) vs. 33% in the placebo arm. The dose had to be reduced in 79% of patients due to significant toxicity (vs. 9% in the placebo arm), treatment discontinuation due to side effects occurred in 16% of patients on cabozantinib (vs. 8% in the placebo arm).

The results of both studies are not comparable due to conceptual differences in the study design. While only patients with radiologically documented progression within a period of 14 months were included in the cabozantinib study, patients without progression were also included in the vandetanib study. In addition, in the vandetanib study, a switch to the verum arm was possible in the event of radiologically proven progression during treatment. In the cabozantinib study, cross-over was not planned. These conceptual differences, which make it impossible to compare the results obtained, are also reflected in the shorter time of progression-free survival in the verum and placebo arm of the cabozantinib study. There are no studies that directly compare the effectiveness of vandetanib and cabozantinib. Therefore, a decision must be made on the basis of the patient's medical history, physical examination, ECG and laboratory results, previous illnesses, concomitant medication and tumor characteristics as to which drug is preferred in each individual case.

To date, no TKI has been convincingly proven to prolong survival in metastatic MTC. Complete remissions have also not been described. Furthermore, there is no evidence that an early start of therapy increases its effectiveness.

As mentioned above, therapy with TKIs is often associated with side effects: Diarrhea, skin rash, nausea, hypertension, headache and also severe side effects such as ECG changes. These side effects sometimes significantly impair quality of life, but can be alleviated by symptomatic therapies. Nevertheless, it is not uncommon for a dose reduction, therapy break or discontinuation of therapy to be necessary. As there is no clear evidence of benefit in the early stages of the disease, the indication for treatment with TKIs is currently only seen with the following criteria under individual risk-benefit assessment, also based on international guidelines [1089], [1146], [335]: high tumor burden and progressive (RECIST criteria), short Ctn/CEA doubling time < 1 year and pronounced symptomatic disease when local therapy measures have been exhausted. Uncontrollable pain and treatment-resistant diarrhea can also support the indication for TKI therapy in individual cases. Since the majority of patients are only slowly progressive even in the advanced metastatic stage and have a barely impaired quality of life, the sole evidence of metastasis and progression is not an indication for therapy, nor is the sole evidence of increased or rising tumor markers Ctn and/or CEA.

The recommendation is based on the high evidence quality of the studies for the response to therapy and is aimed at patients in need of therapy.

If a patient is not a good candidate for vandetanib or cabozantinib or if there is significant progression during the course of treatment, inclusion in a clinical trial should be considered. Alternatively, other TKIs can be used that have shown positive effects in phase II trials in MTC and have been used in other tumor entities as part of an individual treatment trial due to their approval (e.g. axitinib, pazopanib, sorafenib, sunitinib, lenvatinib) [1167], [1168], [1028], [1169][1169], [1030], [1170], [1171].

The phase I/II study [1034] and also the phase III study [1173] on the selective RET inhibitor selpercatinib show very good response rates in progressive advanced MTC in first-line therapy and after TKI pre-treatment with a significantly lower rate of side effects compared to the TKI inhibitors cabozantinib and vandetanib.

In the completed Phase I/II study of selpercatinib [1034], 55 patients with RET-positive medullary thyroid cancer who had been pretreated with vandetanib and/or cabozantinib showed an objective response rate of 69% (complete remission 9%, partial remission 60%) and a 1-year PFS of 82%. In 88 patients with RET-positive medullary thyroid carcinoma without prior therapy, the objective response rate was 73% (complete remission 11%, partial remission 61%) and 1-year PFS 92%. The side effects were predominantly mild (grade 1-2), the most common grade 3, 4 side effects were hypertension (21%), transaminase elevation (GPT 11%, GOT 9%), hyponatremia (8%), and diarrhea (6%). Based on these study results, selpercatinib was approved in the USA in May 2020 for adult and pediatric patients ≥ 12 years of age with advanced or metastatic RET positive medullary thyroid carcinoma who require systemic therapy. In Europe, approval was granted in March 2021 for second-line therapy after treatment with vandetanib and/or cabozantinib in patients ≥ 12 years of age with advanced RET-positive medullary thyroid cancer.

The phase III trial of selpercatinib in advanced RET-mutated medullary thyroid cancer has now been published [1173]. A total of 291 patients were included in the study. It is a first-line therapy with selpercatinib compared to vandetanib or cabozantinib. The median follow-up was 12 months. The median progression-free survival was not achieved in the selpercatinib group; in the control group it was 16.8 months. Progression-free survival after 12 months was 86.8% (95% CI, 79.8 - 91.6) in the selpercatinib group and 65.7% (95% CI, 51.9 - 76.4) in the control group. The median treatment failure-free survival was not reached in the selpercatinib group and amounted to 13.9 months in the control group. The objective response rate was 69.4% in the selpercatinib group and 38.8% in the control group. A dose reduction due to side effects was necessary in 38.9% of patients in the selpercatinib group compared to 77.3% in the control group. Treatment discontinuation was necessary in 4.7% of patients in the selpercatinib group and 26.8% in the control group. Important side effects in the selpercatinib group (grade ≥ 3) were arterial hypertension (18.7%), diarrhea (3.1%), increased liver enzymes (10.4%), etc. Based on these data, the EMA extended the approval for selpercatinib in July 2022 to include first-line treatment of RET-mutated medullary thyroid carcinoma.

The selective RET inhibitor pralsetinib is currently only approved in Germany for RET fusion-positive non-small cell lung cancer. Approval for the treatment of medullary thyroid carcinoma is not being sought. It was investigated in a Phase I/II trial in patients with RET-positive medullary thyroid carcinoma [1036]. The study included 55 patients with RET-positive medullary thyroid carcinoma who had been pretreated with vandetanib and/or cabozantinib. There was an objective response rate of 60% (complete remission 2%, partial remission 58%) and stabilization in 33% of patients, and in patients without pretreatment (n=21) an objective response rate of 71% (complete remission 5%, partial remission 67%) and stabilization in 29% of patients. The median PFS had not yet been reached. The side effects were relatively rare, predominantly grade 1-3. The most common grade 3 or higher side effects were hypertension (17%), neutropenia (13%), lymphopenia (12%) and anemia (10%) [1172]. Based on these data, approval was granted in the USA in December 2020 for adult and pediatric patients ≥12 years of age with advanced or metastatic RET-positive medullary thyroid carcinoma who require systemic therapy. Due to the lack of phase III studies, the significance of the already approved substances vandetanib and cabozantinib as well as in comparison to the selective RET inhibitor selpercatinib cannot be conclusively assessed. The positive recommendation takes into account the patients' need for treatment and the proof of efficacy of the drugs with good quality of evidence.

In summary, an RET mutation analysis (somatic mutation) of the tumor tissue is therefore recommended if medullary thyroid carcinoma is detected (with the exception of patients in whom a pathogenic RET germline variant is known). In the case of a RET mutation, a selective RET inhibitor (e.g. selpercatinib) shall be used in advanced MTC with significant tumor burden and symptomatic or progressive (according to RECIST) metastatic disease. This is also preferable to treatment with a non-selective multikinase inhibitor.
In order to ensure optimal individual treatment planning while exhausting all possible treatment options, including inclusion in controlled clinical trials, it is strongly recommended that these patients be cared for by an interdisciplinary team of experts at specialized centers.

MTC occurs hereditarily in approx. 25% of cases in the context of multiple endocrine neoplasia type 2 (MEN2). MEN2 is one of the hereditary tumor syndromes in which MTC can occur in association with pheochromocytomas in up to 50% (unilateral or bilateral) and with primary hyperparathyroidism (parathyroid adenomas/hyperplasia) in up to 20% depending on the type of pathogenic RET variant [335], [1174]. MEN2 syndrome has an estimated frequency of 1:30,000 and is inherited in an autosomal dominant manner. Hereditary MTC occurs in two different phenotypic forms (MEN2A approx. 95% and MEN2B approx. 5%), which differ in terms of frequency, genetics, age at clinical manifestation, association with other tumors, aggressiveness and prognosis [335], [1175]:

MEN2A (Sipple syndrome) - characterized by an MTC, with unilateral or bilateral pheochromocytomas in up to 50%, primary hyperparathyroidism on the basis of multi-glandular disease in up to 10-20%, and rarely with a skin change on the back (cutaneous lichen amyloidosis) or, extremely rarely, Hirschsprung's disease. A rare variant must be distinguished from this: FMTC - familial MTC with the sole occurrence of MTC without pheochromocytoma or primary hyperparathyroidism.

MEN2B - associated with MTC, unilateral or bilateral pheochromocytomas and a childhood-onset phenotype with marfanoid habitus and centrofacial mucosal neuromas of the lips, tongue and eyelids and intestinal ganglioneuromatosis that can lead to episodes of constipation and diarrhea. Primary hyperparathyroidism is not observed. In most cases, the typical phenotype (bulging lips, multiple ganglioneuromas, etc.) allows a visual diagnosis of MEN2B [1176], [1177].

Hereditary MTC is the leading tumor in MEN2 and manifests earlier than in the sporadic form. In MEN2B, MTC already develops in the first year of life, in classic MEN2A usually in the first/second decade of life and in the mild MEN2A forms in the third/fourth decade of life or later, depending on the causative pathogenic variant in the RET gene. Histopathologically, hereditary MTC is often associated with "neoplastic" C-cell hyperplasia, multifocality, with occurrence in one or both thyroid lobes [1178]. Clinically, the index case presents like a sporadic MTC. The MEN2-associated tumors (pheochromocytoma or primary hyperparathyroidism) often occur after the manifestation of MTC and can be bilateral or multiglandular. Primary hyperparathyroidism often has an "asymptomatic" (mild) course.

The cause of autosomal dominant hereditary MEN2 is an activating germline variant in the RET protooncogene on chromosome 10q11.2, a gene that codes for a tyrosine kinase receptor. Certain variants in the RET protooncogene lead to a permanent activation of the RET receptor and thus activation of the postreceptor signaling cascade, which is responsible for the proliferation of neuroendocrine cells, among other things. C-cell hyperplasia, defined as multifocal proliferation of C-cells within the thyroid gland, can be considered a precursor or precancerous condition of hereditary MTC.

To date, over 100 different pathogenic RET variants are known in MEN2. A list of classified RET variants can be found, for example, in the ClinVar database: https://www.ncbi.nlm.nih.gov/clinvar/?term=MEN2. In almost all families that fulfill the clinical criteria of MEN2, activating, pathogenic germline variants in the RET gene from the DNA of peripheral lymphocytes could be detected and serve as genetic markers. There is a good correlation between the genotype (type and location of the pathogenic variant in the RET protooncogene) and the phenotype (clinical appearance). The expression and penetrance of MTC and other MEN2 tumors is highly variable, from early metastatic MTC in infancy with an activating variant in codon 918 (MEN2B), to MTC occurring in adolescence with variants in codon 634 (MEN2A), to MTC occurring in adulthood associated with variants in exons 13-15 (MEN2A). Three risk groups are formed according to the age at manifestation of the MTC and the clinical course: with "highest" risk (MEN2B, RET variant M918T), with "high" risk (MEN2A, e.g. RET variants affecting codon 634) and with "moderate" risk (MEN2A, e.g. RET variants in exons 13-15) [335] (Table "Risk stratification and management of patients with MEN2 depending on the pathogenic RET variant).

If a pathogenic germline variant is found in the RET protooncogene in an index patient, the familial RET variant can be identified in other family members. In these individuals with a pathogenic variant in the RET gene, further clinical, biochemical and morphological examinations are then necessary to clarify the extent and confirm the diagnosis of MTC, pheochromocytoma and pHPT. In MEN2B, "de novo" variants are often found. The predictive exclusion of a known familial RET variant exonerates these individuals, who do not require further examination for manifestations of MEN2.

Ctn determination in carriers of a pathogenic RET variant in families with MEN2 is used to determine the time for prophylactic thyroidectomy. In these cases, a neoplastic precursor, C-cell hyperplasia (then prophylactic thyroidectomy), or a micro-MTC (then early thyroidectomy) is to be expected due to the significantly higher pre-test probability if the Ctn rises above the upper reference range. The optimal timing of prophylactic thyroidectomy is recommended taking into account the RET variant, the typical age at manifestation (3 risk groups: "moderate, high, highest") and the Ctn level (Table Risk stratification and management of patients with MEN2 depending on the pathogenic RET variant). It should be noted that children have higher Ctn reference values depending on their age.

Since the lifetime risk of carriers of a pathogenic germline variant in the RET protooncogene developing MTC is over 90%, prophylactic thyroidectomy is recommended before the manifestation of a thyroid-transcending MTC. The optimal timing of prophylactic thyroidectomy of asymptomatic carriers of a pathogenic germline variant of the RET gene is recommended based on the risk classification of the specific RET variant into "moderate, high and highest" risk for early development of MTC (early or late penetrance) [335]. The Ctn level complements and individually defines the timing of prophylactic thyroidectomy. As soon as the Ctn level rises above the age- and sex-specific normal range, (mostly) neoplastic C-cell hyperplasia as a precursor of MTC can be assumed. At this stage, the MTC is unlikely to spread beyond the thyroid gland, central (prophylactic) lymph node removal is then not necessary and a thyroidectomy is sufficient. With increasing Ctn levels, a micro-MTC is to be expected, which involves a risk of LN metastases; depending on the Ctn level, a central LN dissection may then have to be discussed [1179]. The time of MTC development in the "highest" risk group is to be expected in the first year of life, possibly in the first month of life, in the "high" risk group at 3-5 years and in the "moderate" risk group after the age of 5 in early adolescence. If the Ctn level measured at this time is within the reference range, thyroidectomy can be postponed by measuring Ctn every 6 months until Ctn rises to a maximum of 30 pg/ml; this applies in particular to patients in the "moderate" risk group [1180], [1181]. With this approach, almost all children can be cured with minimal morbidity.

Index patients and later diagnosed carriers of a pathogenic RET variant with clinically manifest MTC and basally elevated Ctn receive a thyroidectomy with bilateral central and possibly lateral lymph node dissection (compartments I-III) in the same way as patients with sporadic MTC (see Chapter Extent of surgery according to calcitonin level and desmoplasia).

Depending on the genotype, MEN2 patients develop a pheochromocytoma in up to 50 % of cases in the course of their lives , often bilaterally. In most cases, MTC precedes pheochromocytoma by an average of 10 years; it is usually benign, multicentric and restricted to the adrenal gland [1182]. It is important to diagnose pheochromocytoma biochemically (plasma metanephrines, plasma normetanephrines) [1183] and, if necessary, by imaging (CT/MRI/MIBG scintigraphy or F-18-DOPA-PET/CT) before thyroidectomy in order to first operate on the pheochromocytoma and then the MTC. In early thyroidectomized MEN2 patients, annual biochemical screening for the presence of pheochromocytoma is required even in asymptomatic patients. Screening begins in the "highest" and "high" risk groups at the age of 11, and in the "moderate" group at the age of 16. For smaller pheochromocytomas and depending on the location of the tumor in the adrenal gland, subtotal adrenalectomy to preserve the adrenal cortex can be performed laparoscopically or endoscopically retroperitoneally [1184]. A further pheochromocytoma often develops in the residual tissue or on the opposite side over the next 10 years [1185]. After total adrenalectomy, lifelong glucocorticoid and mineral corticoid substitution as well as regular patient training and provision of an emergency card are necessary to prevent adrenal insufficiency.

Primary hyperparathyroidism also shows a genotype-phenotype correlation; it is not observed in MEN2B patients, but typically in patients with the pathogenic variants in RET codon 634 in up to 20 %, usually also only after the manifestation of MTC. Primary hyperparathyroidism is usually asymptomatic with often only slightly elevated serum calcium [1186], [1187]. It is often a single-gland disease, more rarely a multi-gland disease. Depending on the macroscopic findings, surgery may involve selective removal of the enlarged parathyroid gland(s), subtotal (3 ½) resection or, alternatively, total parathyroidectomy with synchronous autotransplantation of a parathyroid gland. A higher probability of recurrence with selective limited surgery must be weighed against the higher risk of postoperative hypoparathyroidism with total thyroidectomy with autotransplantation [1188], [1189]. Examinations such as sestamibi SPECT/CT or choline PET/CT to localize the leading parathyroid gland or prior to recurrent surgery to localize the persistent/recurrent hyperactive parathyroid gland are helpful [1190].