Type 1 Gastric Carcinoid: Is the Conservative Approach Always Indicated?

We read with interest the study by Ichikawa et al. published in the March 2003 issue of Endoscopy [1], focusing on the current indications for clinical management of type 1 gastric neuroendocrine tumors in the Rindi classification [2].

Such lesions, which represent the result of hypergastrinemia that is unrelated to gastrinoma, account for almost 80 % of enterochromaffin-like (ECL) cell gastric tumors. They are generally located in the gastric body and fundus as small-sized single, or more usually multiple, polyps, and they are often associated with chronic atrophic gastritis. Type 1 tumors are considered to be neoplasms that are benign, slow-growing, rarely infiltrative, and very rarely metastatic. In some cases, sudden disappearance of the lesion may follow the removal of the hypergastrinemic stimulus. It has been suggested that a conservative approach, either endoscopic or surgical - e. g., laparoscopic enucleation or gastric antrum resection - is the best method [2] [3], but simple observation [4] and chronic long-acting octreotide therapy [5] have also recently been described in patients who are unfit for surgery.

However, the preoperative distinction between less aggressive (type 1) and more aggressive (type 3) carcinoids, based on gastrin dependency and background gastric mucosa characteristics, may prove difficult in some cases, as there are histomorphological features common to the different types. In addition, it has been reported that 2 - 5 % of type 1 carcinoids are capable of developing an aggressive clinical course, with lymph-node and hematogenous metastases [2] [3]. The following case illustrates such controversies.

A 67-year-old woman complaining of epigastric pain, diarrhea, and flushing underwent upper digestive endoscopy at a primary-care medical center. A single polyp was found at the junction between the gastric body and antrum, about 2 cm in size; endoscopic polypectomy was performed. The histological examination showed a low-grade carcinoid, with two mitoses per 10 high-powered fields, infiltrating the submucosa but not extending to the muscularis propria. A full histological assessment of lymphatic invasion, vascular invasion, grading and Ki-67 was not carried out at that time, and the endoscopic therapy provided was considered definitive. The serum gastrin level was 212 pg/ml. The patient’s epigastric pain disappeared, but she complained of a recurrent flushing syndrome and was referred to our center. A repeat upper endoscopic examination with stepwise biopsies showed no signs of recurrence, but there was evident ECL cell hyperplasia. Computed tomography excluded the presence of gross metastases. However, a reevaluation of the slides from the previous polypectomy showed that the polyp was not correctly oriented, and consequently that there were no inferior resection margins that were free of neoplastic cells. On the basis of these findings, the size of the polyps, the persistence of the symptoms, and the patient’s own preference, a total gastrectomy was carried out. A macroscopic suspicion of nodal involvement at the level of the lesser curvature prompted us to perform a lymphadenectomy extending to nodal stations 7, 8, 9, 11, and 12a. A complete histological examination confirmed the presence of chronic atrophic gastritis, diffuse hyperplasia of the ECL cells of the gastric body and fundus, positively staining for chromogranin A, and the absence of any residual carcinoid at the endoscopic resection site. However, the single first-level node that had been considered suspect during surgery was completely replaced by a metastatic carcinoid cell tumor. The postoperative clinical course was uneventful, and the patient was discharged on the 11th postoperative day. She is receiving chronic long-acting octreotide therapy, with a planned follow-up schedule including octreo-scanning and endoscopy. After 15 months of follow-up, the patient was well and free of disease.

Several clinicopathological findings have been considered to be of value for decision-making in connection with type 1 gastric ECL cell tumors [6]. The number of lesions, size, deep invasion of the gastric wall, vascular invasion, lymphatic invasion, mitotic index, Ki-67 index, and histological grade are all thought to be capable of identifying the very small group of patients who have a substantial risk of aggressive disease, to whom surgical therapy can be offered instead of endoscopic resection. The number and diameter of the polyps are the only parameters available before endoscopic removal, as endoscopic ultrasonography was recently reported to be unreliable in evaluating the depth of invasion through the gastric wall [7]. There is a broad consensus in the literature on the proposed rule that single polyps larger than 1 cm in size require more than endoscopic treatment. However, further good experience and additional reports have given rise to the enthusiasm for a nonaggressive treatment policy for this type of tumor, and larger thresholds have been proposed for the size at which to change the therapeutic approach [3]. On the other hand, in the broad retrospective analysis presented by Rindi et al. [7], the only two patients with extragastric metastases (one nodal and one hepatic) had single polyps 1 cm and 1.5 cm in size, limited to the submucosa and mucosa, respectively. Thus, size criteria may in rare cases prove unreliable for predicting the clinical course of the neoplasia.

Once endoscopic resection has been carried out, histological analysis of the specimen is very important for correctly selecting the treatment and follow-up program. However, complete reconstruction of the resected tissue may prove difficult, as was the case in patient number 2 in the series reported by Ichikawa et al., due to heat-induced tissue degeneration and the multiple fragments sent to the pathologist [1]. Vascular invasion, the most important parameter to be evaluated in this phase, may be unclear in such conditions, while the mitotic index may not reflect the biological behavior of the tumor - as was the case in the patient described above, where it was below the Rindi classification threshold for aggressive therapy.

A third issue concerns recurrent symptoms after endoscopic removal of the tumor. In the case reported above, a complete biochemical examination, including the serum 5-hydroxyindolacetic acid (5-HIAA) level, was not carried out. However, on the basis of the patient’s typical description of the symptoms, it was possible to postulate residual ECL cell neoplasia. The lymph-node involvement was in fact massive, explaining the persistence of the flushing syndrome. In the rare cases of secreting gastric carcinoid, symptoms can be used to guide treatment.

Thus, the diameter of the lesion (more than 1.5 cm), an incomplete resection, unclear or impossible evaluation of the histological margin, and the persistence of symptoms after endoscopic resection may indicate a need for surgical intervention. Ichikawa et al. suggest that these patients should undergo antrectomy along with local resection and subsequent close endoscopic follow-up. However, in the case reported above, only aggressive surgery was able to achieve the goal of a radical cure. In addition, the very low figure given in the Ichikawa et al. report of only one case of nodal involvement in type 1 gastric carcinoids among 200 patients in the literature may be an underestimate. In the series reported by Schindl et al. [3], four total gastrectomies were carried out in patients with type 1 carcinoids. It is noteworthy that these patients were the only ones in whom the problem was definitively treated by eliminating all gastrin-producing cells, as one patient with endoscopic treatment had a recurrence of the disease after 8 months follow-up, and seven of eight patients who underwent local excision and one who had a partial gastric body resection had persistent gastric disease (chronic atrophic gastritis, intestinal metaplasia, ECL cell hyperplasia). Antrectomy was considered to be potentially ineffective due to autocrine gastrin secretion by hyperplastic ECL cells [3].

On the basis of the case reported above, we would recommend considering surgery in patients who have received conservative treatment when the tumor is larger than 1.5 cm in size and when the symptoms do not resolve. We believe that when an indication for surgery has been established, complete treatment of ECL cell disease by total gastrectomy with associated lymphadenectomy should be offered to patients who are fit to undergo surgery.

References

• 1 Ichikawa J, Tanabe S, Koizumi W. et al . Endoscopic mucosal resection in the management of gastric carcinoid tumors.  Endoscopy. 2003;  35 203-206
• 2 Rindi G, Luinetti O, Cornaggia M. et al . Three subtypes of gastric argyrophil carcinoid and the gastric neuroendocrine carcinoma: a clinicopathologic study.  Gastroenterology. 1993;  104 994-1006
• 3 Schindl M, Kaserer K, Niederle B. Treatment of gastric neuroendocrine tumors: the necessity of a type-adapted treatment.  Arch Surg. 2001;  136 49-54
• 4 Johnson F E, Smith J B, Janneycm . et al . Multiple gastric carcinoids.  J Am Coll Surg. 2003;  196 984-985
• 5 Prommegger R, Bale R, Ensinger C. et al . Gastric carcinoid type I tumor: new diagnostic and therapeutic method.  Eur J Gastroenterol Hepatol. 2003;  15 705-707
• 7 Lachter J, Chemtob J. EUS may have limited impact on the endoscopic management of gastric carcinoids.  Int J Gastrointest Cancer. 2002;  31 181-183