Introduction

Hereditary medullary thyroid cancer (MTC) is inherited in an autosomal dominant fashion but may also arise de novo. Commonly, MTC is the first clinical manifestation of multiple endocrine neoplasia type 2 (MEN2). Intriguingly, the speed of progression from neoplastic C cell disease to node-negative and node-positive MTC is proportional to the transforming potential of the respective germline mutations in the RET (REarranged during Transfection) proto-oncogene [1,2,3,4]. Early diagnosis of MTC, facilitated by calcitonin and RET screening, is associated with improved oncologic outcome [5, 6].

The RET gene is located on chromosome 10q11.2 and codes for a transmembrane RET tyrosine kinase receptor that is essential for proliferation, differentiation, and survival of central and peripheral nerve lineages [7]. Constitutive activation of the RET tyrosine kinase receptor drives the malignant transformation of thyroid C cells in a genotype-specific manner, yielding close genotype-phenotype correlations [1, 2, 4] which open a surgical window of opportunity [8, 9].

The 2015 revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma [3], enhancing the 2001 international consensus guidelines for diagnosis and therapy of MEN type 1 and type 2 [10] and the 2012 European Thyroid Association guidelines for genetic testing and its clinical consequences in medullary thyroid cancer [11] and replacing the 2009 medullary thyroid cancer management guidelines of the American Thyroid Association [12], categorize RET mutations into highest risk (RET p.Met918Thr, producing MEN2B), high risk (RET p.Cys634, predominantly associated with MEN2A) and moderate risk mutations. Subsequent research [4] published in 2018 uncovered that moderate risk mutations are a fairly heterogeneous group of mutations that can be subdivided further into moderate-high (RET p.Cys609/611/618/620/630, associated with MEN2A from time to time) and low-moderate risk mutations (RET p.Glu768Asp, p.Leu790Phe, p.Val804Leu/Met and p.Ser891Ala, infrequently associated with MEN2A).

To deplete the pool of unrecognized RET carriers in the general population and enable early pre-emptive thyroidectomy in offspring, DNA-based screening would need to be targeted at sporadic-appearing MTC [13]. In accordance with this concept, systematic genetic screening would eventually (i) eliminate the need for neck reoperations for persistent disease by making the diagnosis earlier, (ii) decrease the number of index patients in the general population, and (iii) reduce the presence of MTC at preemptive thyroidectomy.

Thirty years into the genomic era, it remains unclear to what extent these goals have been achieved and which events helped transform the clinical landscape of hereditary MTC. Because not all mutations are created equal, the expected health benefits may not have developed in synchrony across all four RET risk categories. The present investigation, taking stock of temporal trends in the past four decades, was undertaken to address these research questions.

Patients and methods

Study population

This research was based on retrospective evaluation of prospectively collected data on all RET carriers who were referred to a tertiary surgical center for neck surgery that was carried out between 1986 and 2021.

All carriers of RET mutations (http://www.arup.utah.edu/database/MEN2/MEN2_display.php) [14] provided written informed consent before undergoing neck surgery at the authors’ institution between 1986 and 2021 [13, 15, 16]. All procedures complied with national and international clinical standards applicable at the time of surgery [3, 10,11,12] and were in accordance with the amended Declaration of Helsinki and relevant local rules (institutional review board approval reference 2020-237).

At the time of first referral to the authors’ institution, operative status was determined for each RET carrier: referral for thyroidectomy or referral for neck reoperation. Subsequent reoperations at the authors’ institution were disregarded so that each RET carrier was counted only once.

The diagnosis of MTC was based on histopathologic demonstration of cells immunoreactive with calcitonin extending beyond the basement membrane [17].

RET mutations were subdivided into highest, high, moderate-high or low-moderate risk mutations, as described above [4].

Statistical analysis

For statistical analysis, the software package SPSS® version 28 (IBM, Armonk, New York, USA) was used. Categorical and continuous data were tested with the two-tailed Fisher’s exact test and the two-tailed Kruskal-Wallis rank sum test, respectively, to determine differences among RET risk categories, referral intervals, and clinical variables.

Multiple testing was corrected for with the Bonferroni method. In doing so, observed (nominal) P values were adjusted by multiplying these by the number of statistical comparisons within each set of tests addressing the same research question [18]. Adjusted P values were compared to the significance level alpha = 0.05. To assess the impact of events on subsequent changes in referral patterns, Poisson regression analysis was performed using R version 4.1.2 [19].

As predictors were considered the number of years, taking 1986 as year 1, and the number of years elapsed since each event. For instance, when an event occurred in 1993, 1994 was taken as post-event year 1. These predictors were included in the Poisson regression model using multivariable fractional polynomials [20] to allow for non-linearity in the data. Backward selection with the Akaike information criterion (AIC) was applied to remove weak predictors and to select variables for inclusion in the final Poisson regression model. These analyses were repeated using forward selection to evaluate the sensitivity of the regression model.

Results

Baseline characteristics of the study population

Altogether, 496 RET carriers were referred for thyroidectomy (388 carriers) or neck reoperation (108 carriers) between 1986 and 2021 (Table 1). Of these, 44 carriers had highest risk mutations (p.Met918Thr), 164 carriers high risk mutations (p.Cys634Arg/Gly/Phe/Ser/Trp/Tyr/insHisGluLeuCys), 116 carriers moderate–high risk mutations (p.Cys609/611/618/620/630Arg/Gly/Phe/Ser/Tyr) and 172 carriers low–moderate risk mutations (p.Glu768Asp, p.Leu790Phe, p.Val804Leu/Met, or p.Ser891Ala).

Table 1 Baseline characteristics of the study population referred to the authors’ institution in 1986–2021
Full size table

Significant differences across RET categories (from highest to low–moderate risk) were observed in referrals regarding the proportion of index patients (84% in the highest RET category vs. 21–31% in the other three RET categories; P < 0.001), age at initial thyroidectomy (medians of 12, 10.5, 21.5 and 38 years; P < 0.001), and presence of MTC at thyroidectomy (95, 70, 58, and 59%; P < 0.001). No significant differences were noted for sex (279 female versus 217 male carriers) or in respect to whether thyroidectomy had been performed at the authors’ institution (388 carriers) or elsewhere (108 carriers).

Seminal events and referral of carriers to the authors’ institution by year of referral

Figure 1 chronicles the ups and downs of the annual number of referrals to the authors’ institution between 1986 and 2021 (no referrals took place in 1987 and 1992).

Fig. 1
figure 1

Seminal events and referral of RET carriers to the authors’ institution by year of referral. A Operative status. B Index status. C Medullary thyroid cancer. White circles denote first mutational reports, black circles other high-end RET publications, and arrows international management guidelines, with numbers indicating references. ATA American Thyroid Association, ETA European Thyroid Association

Full size image

Altogether, there were three event clusters that drove referral numbers upward (Fig. 1):

  1. 1.

    The first seminal event comprised a series of genomic discoveries: two first reports from competing research groups [21, 22] in 1993 demonstrating that high and moderate–high RET mutations caused MEN2A and familial MTC (p.Cys634 and p.Cys609/611/618/620/630)—at a time when referral numbers to the authors’ institution hit a record low (first nadir).

    This was followed by another first report in 1994 that found that highest risk RET mutations (p.Met918Thr) caused MEN2B [23] and a string of first reports from other institutions revealing that low–moderate risk RET mutations caused familial MTC in 1995 (p.Glu768Asp and p.Val804Leu) [24, 25], 1996 (p.Val804Met) [26], 1997 (p.Ser891Ala) [27], and 1998 (p.Leu790Phe) [28].

    This rise in referral numbers was buttressed by high-end publications that appeared in 1994 [29, 30] and 1996 [1].

  2. 2.

    The second seminal event was the publication of the international consensus guidelines for diagnosis and therapy of MEN type 1 and type 2 [10] in 2001 – at a time when referral numbers to the authors’ institution were down again, though on a higher level (second nadir).

    The subsequent surge in referrals to a record high at the authors’ institution in 2007 was fueled by two high-ranking publications in 2003 [2] and 2005 [31]. Interestingly, publication of the medullary thyroid cancer management guidelines of the American Thyroid Association [12] in 2009 may have stemmed the accelerating decline of referral numbers that began taking up pace in 2008. In this endeavor, the 2012 European Thyroid Association guidelines for genetic testing and its clinical consequences in medullary thyroid cancer [11] may have been less successful.

  3. 3.

    The third seminal event was the publication of the revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma [3] in 2015—at a time when referral numbers to the authors’ institution hit another record low (the third nadir).

These associations were explored further using Poisson regression analysis. The Akaike information criterion, striking a balance between the risks of overfitting and underfitting, did not permit inclusion of the 2009 (medullary thyroid cancer management guidelines of the American Thyroid Association [12]) and 2012 (European Thyroid Association guidelines for genetic testing and its clinical consequences in medullary thyroid cancer [11]) events, neither in the backward nor in the forward Poisson regression model.

The final Poisson regression hence focused on the 1993 (first description of the RET proto-oncogene as causative of hereditary MTC and MEN2) [21, 22], 2001 (international consensus guidelines for diagnosis and therapy of MEN type 1 and type 2 [10]) and 2015 (revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma) [3] events. Each of these three events, showing statistically significant effects (P < 0.001) when formally explored, was followed by an abrupt surge in referrals of RET carriers to the authors’ institution.

Figure 2, plotting the number of observed against the number of expected referrals of RET carriers to the authors’ institution per year, visualizes these boosts in referrals setting in immediately after the 1993, 2001 and 2015 events.

Fig. 2
figure 2

Number of observed annual referrals of RET carriers versus number of expected annual referrals of RET carriers (final Poisson regression model). Events not included in the final Poisson regression model are represented by dashed vertical lines. ATA American Thyroid Association, ETA European Thyroid Association

Full size image

Referral of carriers with certain clinical features to the authors’ institution by year of referral

Figure 1 depicts time trends of referral numbers for each year of referral to the authors’ institution from 1986 to 2021 (no referrals took place in 1987 and 1992).

By and large, referral numbers for neck reoperations paralleled total referral numbers over the years, but in recent years were on the wane (Fig. 1A). After 30 years of sluggish decline (1988–2018), referrals for reoperation came to an end, as illustrated by a complete lack of referrals for neck reoperation in 2019, 2020, and 2021.

Congruent with this, index patients (Fig. 1B) and carriers with MTC (Fig. 1C) displayed fairly similar referral patterns, although more carriers were affected by these conditions. Index patients, most recently a 3-year-old girl with C cell hyperplasia alone who was found to carry the highest risk p.Met918Thr RET mutation during work-up for stunted growth, continued to be referred to the authors’ institution into 2021 (Fig. 1B).

Referral of carriers to the authors’ institution by RET category and referral period

When referrals were grouped in 5-year increments (Table 2), the decline of referrals for reoperation was statistically significant for carriers of high, moderate–high and low–moderate risk mutations (P < 0.001), but missed statistical significance for carriers of highest risk mutations.

Table 2 Referrals of all 496 RET carriers for thyroidectomy or reoperation to the authors’ institution in 1986–2021
Full size table

Referral of carriers to the authors’ institution by reoperation, RET category and referral period

Evaluating all annual referrals together can provide valuable clues but may obscure differential referral patterns for carriers who differ in operative status, index status, and thyroid histopathology and harbor RET mutations that confer disparate risks.

To control for such confounders, Fig. 3 stratifies 5-year referral patterns of index patients (Fig. 3A), carriers with MTC (Fig. 3B), and age at thyroidectomy (Fig. 3C) by operative status and RET risk category.

Fig. 3
figure 3

Referral of carriers to the authors’ institution by reoperation, RET category and referral period. A Index status. B Medullary thyroid cancer. C Age at thyroidectomy

Full size image

As pictured in Fig. 3, increases in referrals for thyroidectomy peaked in 1996–2000 for carriers of highest and high risk mutations, and in 2006–2010 for carriers of moderate–high and low–moderate risk mutations, some 10 years later.

Corresponding trends, although attenuated, were present also in referrals for reoperations, involving mainly index patients (Fig. 3A) and almost always carriers with MTC (Fig. 3B).

The 108 carriers referred for reoperation were considerably older at outside thyroidectomy than the 388 carriers referred for thyroidectomy to the authors’ institution (Fig. 3C). Likewise, the spread in age at thyroidectomy across RET risk categories was greater in the former group who were older than in the latter group of younger carriers who were diagnosed earlier.

When formally tested, the above referral trends were consistently significant for referrals of carriers of high risk mutations for thyroidectomy (Table 3), and consistently negative for referrals for reoperation (Table 4).

Table 3 Referral of all 388 RET carriers to the authors’ institution for thyroidectomy in 1986–2021
Full size table
Table 4 Referral of all 108 RET carriers to the authors’ institution for neck reoperation in 1986–2021
Full size table

Discussion

This is the first investigation to explore the impact of international consensus and professional management guidelines on referrals of RET carriers to a tertiary surgical center before and after genomic testing became available in the 1990s. Following a string of first reports linking RET missense mutations to MEN2 and MTC, international management guidelines on MEN2 and MTC highlighted and emphasized the importance of genetic screening of patients with sporadic-appearing MTC and offspring of known RET families, reaching a wider, more diverse clinical audience and sparking off two further waves of referrals (Fig. 2).

Because they are more likely to be the first family member to present for genetic counseling, carriers with early-onset disease typically become the index patient who brings the family under study. As a rule, carriers of highest and high risk RET mutations meet the criteria for family cascade testing more frequently than carriers of moderate–high and low–moderate risk RET mutations. Because of this, the former group of patients was referred more often with the first wave, whereas the latter group of patients was more commonly referred with the second and third wave (Fig. 3A, B).

A disproportionate number of carriers of highest risk mutations, 37 of 44 carriers or 84% (Table 1), were index patients carrying de novo RET p.Met918Thr mutations. Owing to the burden of disease, many index patients, developing MEN2B in infancy and early childhood, will not pass on the trait to future generations. This limits the pool of nonindex patients inheriting highest risk RET mutations, thwarting efforts by family cascade testing to spot these mutations earlier.

In spite of this, RET screening is crucial to make the diagnosis early, both at the level of the individual and the level of the population. The clinical benefits of systematic calcitonin screening before (pre-1993) and systematic molecular screening in the genomic era (starting in 1993), notably reductions in the number of index patients, in the number of patients with MTC and in age at thyroidectomy, were greater among carriers of high risk mutations than among carriers of moderate risk mutations (Table 3). This finding is explained by the fact that carriers of highest risk mutations, which are more penetrant, acquired the trait frequently from a clinically affected MEN2A parent, whereas carriers of moderate risk mutations, which are less penetrant, inherited the trait often from a mildly affected parent.

For many RET carriers, timely thyroidectomy alone will be adequate treatment, sparing them additional node dissection with incremental morbidity coming with the procedure [8]. Early preemptive thyroidectomy also helps diminish neck reoperations, which carry even greater operative risks. In this respect, it may be interesting to note that referrals to the authors’ institution for neck reoperations petered out recently (Fig. 1A).

The present work, breaking new ground, has strengths and weaknesses. A key asset were the large number of almost 500 RET carriers referred to, and carefully documented at the authors’ institution over a 36-year period, which resulted in complete data sets. Of note, all key performance indicators – operative status at referral, index or nonindex patient status, presence of MTC at thyroidectomy, year of and age at thyroidectomy, and year of referral – were prospectively documented and are robust to temporal change.

In view of the long observation period, it cannot be ruled out that referrals to the authors’ institution may have been subject to change in the course of time. Remarkably, the 2020 climax of the Covid-19 pandemic in Germany increased, rather than decreased, the number of referrals to the authors’ institution where patients in need of prophylactic or therapeutic cancer surgery continued to be prioritized. Given the senior author’s longstanding expertise in prophylactic thyroidectomy [32] and compartment-oriented node dissection [33], intervening changes in referrals of RET carriers, if any, may have skewed the disease spectrum at the authors’ institution towards advanced MTC. If so, the present research may underestimate the size of the impact of seminal events on the presentation of hereditary MTC and the magnitude of health benefits associated with routine RET screening.

Having said that, cause (publication) and effect (promptly being acted on by the medical community) can be implied, but not proven without a shadow of doubt. This is a challenge inherent in any public health care research. In the era of the internet, though, it is reasonable to presuppose expeditious uptake of seminal publications that appear in computerized biomedical bibliographic retrieval systems upon acceptance.

Notoriously hard, if not impossible, to disentangle are additive effects that may be reinforcing referrals to specialist surgical centers. For instance, two key publications, appearing in the New England Journal of Medicine in 2003 [2] and 2005 [31], arguably may have helped sustain the second wave of referrals to the authors’ institution, which set off with the 2001 publication of the international consensus guidelines for diagnosis and therapy of MEN type 1 and type 2 [10].

The present study of 496 MEN2 gene carriers holds useful learning lessons for other monogenic disease. Once genomic screening becomes available, carriers presenting with early onset and more severe disease will be targeted – and referred for pre-emptive surgery as appropriate – in the first place, whereas carriers presenting with late onset and indolent disease stand to be lost sight of at the outset. The international community is called on to quickly develop and promote international management guidelines encompassing the complete spectrum of genetic disease to build and increase the pressure throughout the entire health care system towards screening of sporadic-appearing disease and offspring of known gene families early on.