Are Patients With Autoimmune Thyroid Disease and Autoimmune Gastritis at Risk of Gastric Neuroendocrine Neoplasms Type 1?

Krystallenia I. Alexandraki; Argiro Nikolaou; Dimitrios Thomas; Vassiliki Syriou; Penelope Korkolopoulou; Stavros Sougioultzis; Gregory Kaltsas


Clin Endocrinol. 2014;80(5):685-690. 

In This Article


In this prospective study, we found that 33% of patients with AITD have histologically and serologically confirmed AG with gastrin being the only factor predicting its presence. In up to 17·5% of these patients, ECL-cell hyperplasia may develop, progressing to a GNEN1 in a further 14% of those with ECL-cell hyperplasia. Taking into consideration the young age of our studied group and the fact that only a minority of them (22·5%) had normal histological findings, the present study suggests that UGE might be of clinical value in patients with AITD irrespective of the presence of relevant symptoms.

A similarly high rate of AG in patients with AITD has previously been reported.[1] It has been suggested that the co-existence of both these entities may be related to a common antigenic epitope.[15] This view is also supported by the positive correlation observed between APCA and anti-TPO titres in our studied population. Taking into account that APCA can be detected in young persons with juvenile AITD leading to hypergastrinaemia and ECL-cell hyperplasia,[16] it is prudent to speculate that this population may be at increased risk of GNEN1 development. This is also supported by our 32-year-old female patient in whom a GNEN1 developed from nodular ECL-cell hyperplasia after the relatively short follow-up period of 39 months. Furthermore, HP infection can also affect ECL cells via mucosal pro-inflammatory cytokine production resulting in inhibition of histamine secretion, thus predisposing to hypochlorhydria, hypergastrinaemia[6,7,8] and consequent AG and ECL-cell hyperplasia. Although the role of HP infection in susceptibility to AG has not been precisely defined, we examined its presence and potential role in our patients with AG; nonetheless, we did not find any impact of HP infection in ECL-cell hyperplasia development. In support of this are studies showing that APCA positivity and hypergastrinaemia, but not HP infection per se, are related to atrophic gastritis.[8] It has also been suggested that among patients with AG due to HP infection, a proportion, approximately 20%, will normalize their gastric mucosa after HP eradication.[8] Interestingly, our patient who developed a GNEN1 had unsuspected HP infection with an indolent course. However, despite eradication of HP infection following appropriate treatment, AG persisted and GNEN1 developed later on.

Serum gastrin and CgA levels were increased in patients with AG, ECL-cell hyperplasia and GNEN1 as expected.[17] In addition, a positive correlation of APCA titre with CgA and gastrin concentrations was also observed supporting their interplay. However, only gastrin reliably predicted ECL-cell hyperplasia on our logistic regression analysis model. Although no specific gastrin cut-off value predicting the presence of ECL-cell hyperplasia in patients with AG was found, gastrin measurements could be of potential use as a tumour marker. Larger studies including more patients with AITD and AG who have developed GNEN1 could allow ROC analysis to be performed to potentially identify a meaningful biochemical marker putting such patients with AG at risk of GNEN1 development. Furthermore, PPIs should be avoided in these patients as they result in a rise of serum gastrin levels confounding the interpretation of gastrin measurements, whereas they may also promote ECL-cell hyperplasia.[6]

It has recently been suggested that APCA should be considered in any patient (adult or young) known to have AITD[1,16] along with yearly gastrin measurements. As normal findings in UGE were only seen in a minority of patients studied (22·5%), the present study provides information in favour of performing an UGE in patients with AITD. Although the period to repeat this procedure cannot be determined, it may need to be shortened compared with the 5-year time period previously proposed.[16,18,19] The identification of AG also helps explaining the need for increased L-thyroxine requirements in patients with AITD in whom the required L-thyroxine dose is disproportional to their BMI.[20]

The present study has several limitations, namely its short-term follow-up and the small number of patients studied. However, its prospective design provides some important information regarding the development of GNEN1 in patients with AITD and its potential implications. This is particularly important as the population studied was of a relatively young age and GNEN1 constitutes the commonest GEP-NETs, of which a subset may follow not such an indolent course.

In conclusion, our study demonstrates that patients with AITD are at risk of ECL-cell hyperplasia and consequently GNEN1 and that they may benefit from early assessment of the presence of AG. The exact protocol of follow-up of these patients and potential clinical implications need longer-term studies.