Society: ASGE

Background and Aims

Recently macroscopic-onsite evaluation (MOSE) technique for Endoscopic Ultrasound guided Tissue acquisition (EUS-TA) for solid gastrointestinal lesions has been advocated to provide high diagnostic accuracy in settings where rapid- onsite evaluation is not available. However, MOSE technique has not yet been standardized. Our aim was to compare MOSE guided EUS-TA with conventional techniques and to find out the optimum parameters required for MOSE technique.

Materials and Methods

In this prospective randomized study, 96 consecutive adult patients undergoing EUS-TA were randomized to MOSE and conventional groups. MOSE technique was performed by segregating material obtained in each pass into Macroscopically visible Core (MVC), and red paste. Obtaining total MVC length greater than 4 mm was considered complete, and no further passes were taken. Sensitivity, specificity, positive predictive value, negative predictive value, diagnostic yield, diagnostic accuracy was compared between two arms. Optimum length and number of MVC required was also evaluated.

Results

When comparing MOSE group (n=48) with conventional group (n=48), diagnostic accuracy, sensitivity ,specificity, positive predictive value and negative predictive value were 97.9% vs 95.8%,93.3 % vs 79.4%, 100 % vs 100%, 100 % vs 100% and 90% vs 87.8% respectively (Table 1).Diagnostic yield was similar in both the groups (95.8% vs 91.6%, p=0.39).The mean length of MVC obtained was 15.46 ± 3.22 mm and number of MVC obtained was 3.17 ± 0.83 in MOSE group. No significant difference in duration of procedure and adverse events were noted in two groups. Number of passes needed in MOSE group was significantly less then needed in conventional arm ( median 2 vs 3 , mean 2.08 ± 0.28 vs 2.65 ± 0.53, p <0.001). Obtaining a 11.5 mm length of MVC was 93.3% sensitive whereas obtaining total 2.5 number of MVC was 86.7% sensitive to diagnose malignancy (Figure 1).

Conclusion

MOSE technique guided EUS-TA results in fewer number of passes needed to achieve high diagnostic accuracy and diagnostic yield without increasing the procedure duration and adverse events as compared to conventional technique. Obtaining longer length and more number of MVC increases the sensitivity to diagnose malignancy.

<b><u>TABLE 1: Comparing the clinical details, technical parameters, diagnostic accuracy and diagnostic yield between MOSE group and Conventional group. </u></b>

TABLE 1: Comparing the clinical details, technical parameters, diagnostic accuracy and diagnostic yield between MOSE group and Conventional group.

FIGURE 1 : Area under the receiver operating characteristic graph shows that in in patients where Macroscopic onsite evaluation (MOSE) was used, obtaining a total Macroscopic visible core (MVC) length of 11.5 mm had 93.3% sensitivity and obtaining 2.5 MVC cores (each 4 mm) had 86.7% sensitivity in diagnosing malignancy.

FIGURE 1 : Area under the receiver operating characteristic graph shows that in in patients where Macroscopic onsite evaluation (MOSE) was used, obtaining a total Macroscopic visible core (MVC) length of 11.5 mm had 93.3% sensitivity and obtaining 2.5 MVC cores (each 4 mm) had 86.7% sensitivity in diagnosing malignancy.

Background:
Endoscopic ultrasound-guided fine-needle aspiration/biopsy (EUS-FNA/B) is the gold standard for diagnosing gastric subepithelial lesions (SELs). However, especially in SELs <20 mm, collecting sufficient tissue for pathological diagnosis can sometimes be challenging. To improve the diagnostic accuracy of the small SEL, we have developed a new method, termed traction-assisted EUS-FNB using the clip-with-thread (TA-EUS-FNB). That can fix the lesion to prevent respiratory fluctuation, orient the lesion perpendicular to the endoscope, and obtain counter traction to facilitate puncture. This study aimed to evaluate the diagnostic accuracy of TA-EUS-FNB.
Method:
This is a prospective randomized controlled crossover study conducted from July 2019 to November 2022. Consecutive 30 patients with gastric SELs <20 mm (14 men; mean age 57.7 [± 13.9] years) were enrolled in this study. These patients were randomly assigned to either TA-EUS-FNB first group or the conventional EUS-FNB first group. Four punctures in each patient were performed, namely twice by the designated first method and then twice by the other method. The final diagnosis was determined by surgical pathology (n=13) or, in patients who did not undergo surgery, by biopsy samples such as EUS-FNB (n=17). The primary outcome was set as the diagnostic accuracy of the two methods mentioned above. The specimen adequacy and diagnostic capability of sensitivity, specificity, and positive predictive value to differentiate gastrointestinal stromal tumor (GIST) from non-GIST were analyzed as the secondary outcome.
Result:
The mean tumor size was 15.1 (± 4.0), and the final diagnosis was GIST (n=15, 50%), leiomyoma (n=8, 26.7%), schwannoma (n=2, 6.7%), aberrant pancreas (n=3, 10%), and inflammation (n=2, 6.7%). Compared to conventional EUS-FNB, TA-EUS-FNB has significantly higher diagnostic accuracy (86.7% vs. 63.3%, p=0.037). The specimen adequacy of TA-EUS-FNB was also significantly higher than that of conventional EUS-FNB (90% vs. 66.7%, p=0.028), while the sensitivity and specificity, and PPV of TA-EUS-FNB were comparable to that of conventional EUS-FNB (sensitivity [86.7% vs. 66.7%, p=0.195], specificity [100% vs. 100%], positive predictive value [100% vs. 100%]) (Table).
Conclusion:
The TA-EUS-FNB is superior to the conventional EUS-FNB in specimen adequacy and diagnostic accuracy for SELs <20mm, which should be a good option.

Study design and method

table of outcome

(Introduction) Although the gold standard for diagnosing subepithelial lesions (SELs) is endoscopic ultrasound guided fine-needle aspiration/biopsy (EUS-FNB), its accuracy for SELs <20 mm is low. To overcome this situation, we developed a new diagnostic method to differentiate between gastrointestinal stromal tumor (GIST) and non-GIST by measuring impedance value using an EUS-FNB needle.
(Methods) The impedance of 16 gastric epithelial neoplasms was measured with a conventional impedance probe to determine if the impedance is clinically useful in assessing cell density in Study 1. The impedance values of non-exposed/exposed SELs from 25 cases using the conventional probe in Study 2, and non-exposed SELs from 20 cases using the EUS-FNB needle probe in Study 3 were measured to determine the diagnostic yield of the impedance to differentiate GISTs from non-GISTs.
(Results) There was a significant positive correlation between the impedance and cell density (P=0.030) in Study 1. In Study 2, the impedance value of 99.5 for GIST measured with a conventional probe was significantly higher (P<0.01) than for muscle layer (82.4) and leiomyoma (89.2). The impedance of GIST measured with the EUS-FNB needle was also significantly higher than that of leiomyoma (GIST: 80.2 vs leiomyoma: 71.8, P=0.015). The diagnostic yield of the impedance method to differentiate GIST from non-GIST was 94.4% accuracy, 88.9% sensitivity, 100% specificity, and area under the curve was 0.95. Diagnostic performance was not affected by lesion size (P=0.86) (Study 3).
(Conclusion) Impedance measurement of EUS-FNB to differentiate gastric GISTs from non-GISTs would be a good option, especially if the lesion is less than 20 mm.

Yosuke Minoda

Appearances