Non–Pure Intestinal Phenotype as an Indicator of Progression in Sporadic Nonampullary Duodenal Adenomas: A Multicenter Retrospective Cohort Study

Abstract

INTRODUCTION:

We aimed to evaluate the natural course of sporadic nonampullary duodenal adenomas (SNDAs) and determine the risk factors of progression.

METHODS:

We retrospectively analyzed the follow-up outcomes of patients with biopsy-diagnosed SNDA between April 2010 and March 2016 at 13 institutions. All initial biopsy specimens were centrally evaluated. Only those diagnosed with adenomas were included. Mucinous phenotypes were classified into pure intestinal and non–pure intestinal phenotypes. Cumulative incidence rates of carcinoma and tumor enlargement were evaluated. Tumor enlargement was defined as a ≥25% or 5-mm increase in tumor size.

RESULTS:

Overall, 121 lesions were analyzed. Within a median observation period of 32.7 months, 5 lesions were diagnosed as carcinomas; the cumulative 5-year incidence of carcinoma was 9.5%. Male sex (P = 0.046), initial lesion size ≥10 mm (P = 0.044), and non–pure intestinal phenotype (P = 0.019) were significantly associated with progression to carcinoma. Tumor enlargement was observed in 22 lesions, with a cumulative 5-year incidence of 33.9%. Initial lesion size ≥10 mm (P < 0.001), erythematous lesion (P = 0.002), high-grade adenoma (P = 0.002), Ki67 negative (P = 0.007), and non–pure intestinal phenotype (P = 0.001) were risk factors of tumor enlargement. In a multivariate analysis, an initial lesion size ≥10 mm (P = 0.010) and non–pure intestinal phenotype (P = 0.046) were independent and significant risk factors of tumor enlargement.

DISCUSSION:

Lesion size ≥10 mm and non–pure intestinal phenotype on initial biopsy are risk factors of cancer progression and tumor enlargement in cases with SNDA. Thus, management effectiveness may be improved by focusing on lesion size and the mucinous phenotype.

INTRODUCTION

Duodenal cancer is an extremely rare type of gastrointestinal cancer; however, its incidence has increased recently (1–4). Adenomas arising from the duodenum are considered precancerous lesions because carcinogenesis through the adenoma-carcinoma sequence pathway is presumed, as in the colon (5–7). Therefore, they are typically considered for treatment. Among duodenal adenomatous lesions, sporadic nonampullary duodenal adenoma (SNDA) is one of the most frequently encountered lesions (8). The prevalence of SNDA in subjects who underwent endoscopy has been reported to be less than 0.4%; however, with the advent of endoscopy, the detection of SNDA is gradually increasing (9,10). Because postresection complications of these lesions can be severe, the risks and benefits of resection or observation should be considered.

Understanding the natural course of SNDA is important for establishing its optimal management strategy. There are several reports focusing on the natural course of SNDA. Okada et al (11) reported that high-grade dysplasia and a tumor diameter >20 mm were associated with SNDA progression. Cassani et al also showed that tumor atypia was exacerbated in some SNDAs (12). Sakaguchi et al (13) reported that changes in the macroscopic size or shape of SNDAs indicate a high risk of progression to invasive cancer. In those reports, the progression rate from adenoma to cancer was reported to be 4.3%–5.6% (11–13). However, most cases were low-grade adenomas (LGAs), and no reports have followed up a large number of high-grade adenomas (HGAs). Therefore, the biological malignant potential of SNDAs remains unknown and their natural course is unclear.

In previous reports, red color, depression, mixed-type morphology, and preampullary location were reported as indicators for harboring high-grade dysplasia (14,15). Several studies have also reported that SNDA with a gastric-type mucin phenotype is highly malignant (16–20). However, these data were obtained from resected specimens and biopsy tissues, and it remains unclear whether SNDAs with such characteristics show aggressive behavior during follow-up. Thus, we aimed to examine the natural course of SNDA and determine the risk factors associated with tumor enlargement and progression to cancer.

METHODS

Study design

This multicenter retrospective cohort study was conducted at the Osaka University Hospital and its 12 affiliated institutions. The study was approved by the institutional review board of Osaka University Hospital (IRB No. 18059) and was conducted in accordance with the principles of the Declaration of Helsinki. Although the need for written informed consent was waived owing to the retrospective nature of the study, all participants were given the opportunity to refuse to participate in the study using an opt-out procedure available on the website of each participating institution.

Patients

We evaluated consecutive patients with SNDA diagnosed for the first time by endoscopy and biopsy at participating institutions between April 2010 and March 2016. The exclusion criteria were resection by endoscopic or surgical treatment within 3 months; polyposis, such as familial colorectal adenomatosis; an observation period <3 months; and unavailable initial biopsy specimens. All initial biopsy specimens were centrally evaluated by a board-certified pathologist at the Department of Pathology, Osaka University Graduate School of Medicine, as described further, and only those diagnosed as LGA (category 3) or HGA (category 4.1) based on the revised Vienna classification (17) were included. For eligible cases, endoscopic images, final biopsy specimens in unresected lesions, and resected specimens in resected lesions were also collected and centrally evaluated.

Evaluation of histological features

All specimens were collected as paraffin blocks or unstained slides and subjected to hematoxylin and eosin (HE) staining and immunohistochemistry (IHC) at the Department of Pathology, Osaka University Graduate School of Medicine. IHC staining was conducted on the Dako Autostainer Link 48 platform (Agilent, Santa Clara, CA) using an automated staining protocol. The following primary antibodies were used for IHC: TP53, Ki67, MUC5AC, MUC6, and CD10 (see Supplementary Table 1, http://links.lww.com/CTG/B25, which demonstrates details of the primary antibodies). All HE and IHC specimens were centrally evaluated by a board-certified pathologist blinded to clinical information. For cases that were difficult to evaluate, 2 or more board-certified pathologists discussed the evaluation of the specimens. All specimens were classified according to the revised Vienna classification of gastrointestinal epithelial neoplasia (21). TP53 positivity was defined as diffuse or focal staining, while negativity was defined as no staining. For Ki67, total or near-superficial staining was considered positive, whereas staining limited to the proliferative zone was considered negative.

Mucinous phenotypes were classified into 4 categories based on IHC findings: pure intestinal (MUC5AC and/or MUC6 negative, and CD10 positive), intestinal mucin–dominant gastrointestinal (MUC5AC and/or MUC6 partially positive, and CD10 positive), gastric mucin–dominant gastrointestinal (MUC5AC and/or MUC6 positive, and CD10 partially positive), and pure gastric phenotypes (MUC5AC and/or MUC6 positive, and CD10 negative) (Figure 1). CD10, MUC5AC, and MUC6 were evaluated as positive if >5% of the tumor cells were stained. Moreover, because only a small proportion of cases had a predominantly gastric phenotype, we classified mucinous phenotypes into the pure intestinal (PI) and non-pure intestinal (NPI) phenotype (i.e., without and with a gastric phenotype component) groups.

[Figure omitted. See PDF]

Procedure of endoscopic examination and biopsy during follow-up

For patients who did not undergo immediate resection, upper gastrointestinal endoscopy was performed for follow-up observation every 6–12 months. As a rule, a standard diameter endoscope was used, and biopsies were performed at the discretion of the physicians. One of the following biopsy forceps was used depending on the facility: Radial Jaw 4P (Boston Scientific, Boston, MA), Radial Jaw 4 Standard Capacity (Boston Scientific, Boston, MA), EndoJaw FB-230K (Olympus, Tokyo, Japan), or EndoJaw FB-231K (Olympus). The choice between a normal biopsy forceps and a thin biopsy forceps was made at the discretion of the physicians. During the surveillance phase, resection was contemplated based on morphological alterations, biopsy-proven pathological advancement, or patient preference. The lesions that disappeared following the initial biopsy or subsequent additional biopsies, and those that could not be identified, were defined as disappeared lesions.

Evaluation of endoscopic images

All endoscopic images were collected and centrally evaluated by 3 board-certified gastrointestinal endoscopists blinded to clinicopathological information. We assessed lesion diameter, macroscopic type, color, and presence or absence of whitening (described as milk-white mucosa in a previous report) (22). Each evaluator first received a brief lecture on the key findings of each observation before independently assessing all lesions. The lesion images were independently viewed, and the diameters were assessed by the 3 evaluators, and their average value was taken. To assess changes in lesion size, each evaluator compared endoscopic images at initial examination and at follow-up examination. Other findings were determined by a majority vote or by discussion if a consensus could not be reached. Macroscopic types were classified according to the Paris Classification (23). For analysis, 0-I types were designated as protruded and 0-IIa/IIb/IIc as flat or depressed based on the primary macroscopic type of each lesion.

Outcome measures

Two outcome measures were assessed. The first was the cumulative incidence rate of carcinoma development among received SNDAs that were followed up. For resected lesions, the diagnosis derived from the resected specimen was considered as the final diagnosis. For lesions that were not resected, the most malignant diagnosis among all biopsies, including follow-up biopsies, was deemed the final diagnosis of the lesion. Furthermore, if only 1 biopsy was performed, the pathological result of this initial biopsy was considered as the final diagnosis. The time to carcinoma development was defined as the time from the date of endoscopy when the first biopsy was performed to the date of endoscopy when the carcinoma was detected by resection or biopsy. The second outcome measure was the cumulative incidence rate of tumor enlargement at follow-up. We discussed before study and defined tumor enlargement as ≥25% or 5-mm increase in tumor size, considering these as clinically impactful cutoff values. The time to tumor enlargement was defined as the time from the date of endoscopy at first biopsy to the date of endoscopy when the tumor enlargement was observed.

Statistical analysis

Cumulative incidence rates of carcinoma development and tumor enlargement were plotted using the Kaplan–Meier method, and differences among the groups were analyzed using the log-rank test. Univariate and multivariate Cox regression analyses were performed to identify risk factors of progression to cancer and tumor enlargement. All statistical analyses were performed using JMP Pro version 16 (SAS Institute, Cary, NC) software and PRISM version 8 (GraphPad Software, San Diego, CA). Statistical significance was set at P < 0.05.

RESULTS

Patient and lesion characteristics

Among the 285 duodenal tumor lesions diagnosed as adenomas by biopsy, 157 lesions were excluded because of resection within 3 months (n = 79), follow-up <3 months (n = 56), familial adenomatous polyposis (n = 21), or an unavailable specimen (n = 1). Finally, 128 lesions underwent central pathological evaluation. The evaluation revealed 1 carcinoma and 6 non-neoplastic lesions. The remaining 121 lesions were subjected to subsequent analysis (see Supplementary Figure 1, http://links.lww.com/CTG/B25, which demonstrates the flowchart of patient selection). Background characteristics of lesions are summarized in Table 1. To examine the validity of the lesion diameter and endoscopic findings, we examined the degree of agreement among the 3 endoscopists, which proved to be acceptable (see Supplementary Table 2, http://links.lww.com/CTG/B25, which demonstrates the agreement of the central evaluation of endoscopic findings, and Supplementary Figure 2, http://links.lww.com/CTG/B25, which demonstrates details of the central evaluation of lesion diameter). The median lesion diameter was 8.0 mm. Of the 121 lesions considered, 71 (58.7%) and 50 (41.3%) were diagnosed as LGAs and HGAs, respectively, on initial biopsy. Concerning the mucinous phenotype, 64 (52.9%) and 57 (47.1%) lesions had PI and NPI phenotypes, respectively. The median observation period was 32.7 months.

Table 1.

Clinical and histological data of 121 SNDA lesions of 115 patients

Initial diagnosis	Total (n = 121)
Patients' characteristics
Age, yr (mean ± SD)	68.4 ± 9.5
Sex, n
Male	83
Female	38
Lesions characteristics
Location, n
Bulb	21
Second portion (preampullary)	35
Second portion (postampullary)	61
Third portion	4
Initial lesion size, mm
Median (range)	8.0 (2–60)
Macroscopic type, n
0-I	18
0-IIa	70
0-IIb	1
0-IIc	5
0-I + IIa	4
0-IIa + IIc	27
Whitening, n
Present	94
Absent	27
Color, n
Erythematous	28
Isochromatic or white	93
Gastric atrophy, n
None	34
Closed type	45
Opened type	40
Follow-up period, mo (median, range)	32.7 (3.0–111.4)
No. of endoscopic examinations during follow-up (median, range)	3 (1–14)
No. of biopsies during follow-up (median, range)	1 (0–12)
Histological findings of initial biopsy
Vienna classification
LGA (category 3)	71
HGA (category 4.1)	50
Mucinous phenotype
PI phenotype
NPI phenotype	64
Intestinal mucin–dominant gastrointestinal phenotype	52
Gastric mucin–dominant gastrointestinal phenotype	2
Pure gastric phenotype	3

HGA, high-grade adenoma; LGA, low-grade adenoma; NPI, non–pure intestinal; PI, pure intestinal; SNDA, sporadic nonampullary duodenal adenoma.

The natural course of SNDA is shown in Figure 2. Of the 71 lesions diagnosed as LGA on initial biopsy, 18, 46, and 7 lesions disappeared by biopsy, showed no change, and enlarged, respectively. Overall, 14 lesions were eventually resected, and 39 lesions were followed up. The final diagnoses were carcinoma, HGA, and LGA in 2, 7, and 64 lesions, respectively. Of the 50 lesions diagnosed as HGA on the initial biopsy, 5, 32, and 13 lesions disappeared by biopsy, showed no change, and enlarged, respectively. In total, 20 lesions were eventually resected, while 25 lesions were followed up. The final diagnoses were carcinoma and HGA for 3 and 47 lesions, respectively.

[Figure omitted. See PDF]

Risk factors of progression to carcinoma

Five lesions were finally diagnosed as carcinomas by biopsy or resection (see Supplementary Table 2, http://links.lww.com/CTG/B25, which summarizes details of the 5 lesions finally diagnosed as carcinoma). The cumulative 3-year and 5-year incidence rates of carcinoma were 2.1% and 9.5%, respectively (Figure 3a). Focusing only on carcinomas confirmed by resected specimens, the cumulative 3-year and 5-year incidence rates were 1.2% and 6.5%, respectively. Table 2 summarizes the results of a univariate analysis of factors associated with progression to carcinoma. Male sex (P = 0.046), initial lesion size ≥10 mm (P = 0.044), and an NPI phenotype (P = 0.019) were significantly associated with progression to carcinoma. Conversely, dysplasia grade, which was a predictor of cancer in a previous study (9), was not a significant factor (P = 0.270). Given the small number of developing carcinomas, a multivariate analysis was not performed. Kaplan–Meier curves stratified by initial lesion size, dysplasia grade, and mucinous phenotype (Figure 3b–d, respectively) showed that the mucinous phenotype was the strongest predictor of progression to cancer.

[Figure omitted. See PDF]

Table 2.

Univariate analysis of the factors associated with progression to carcinoma

Risk factors	No. of lesions	No. of events	Hazard ratio (95% CI)	P value
Patient characteristics
Age
<70 y	60	2	1.00
≥70 y	61	3	1.937 (0.325–11.56)	0.468
Sex
Female	38	0	1.00
Male	83	5	6.188 (1.034–37.04)	0.046
Lesions characteristics
Location
Postampullary	65	2	1.00
Preampullary	56	3	1.479 (0.254–8.610)	0.663
Initial lesion size
<10 mm	70	2	1.00
≥10 mm	51	3	6.537 (1.053–40.59)	0.044
Macroscopic type
Flat or depressed	103	4	1.00
Protruded	18	1	1.692 (0.132–21.64)	0.686
Whitening
Present	94	2	1.00
Absent	27	3	6.311 (0.838–47.53)	0.074
Color
Isochromatic or white	93	3	1.00
Erythematous	28	2	5.117 (0.494–52.99)	0.171
Histological findings of initial biopsy
Vienna classification
LGA (category 3)	71	2	1.00
HGA (category 4.1)	50	3	2.805 (0.449–17.52)	0.270
p53
Positive	80	3	1.00
Negative	41	2	1.323 (0.207–8.452)	0.768
Ki67
Positive	68	2	1.00
Negative	53	3	2.438 (0.402–14.79)	0.333
Mucinous phenotype
PI phenotype	64	0	1.00
NPI phenotype	57	5	8.168 (1.407–47.41)	0.019

CI, confidence interval; HGA, high-grade adenoma; LGA, low-grade adenoma; NPI, non–pure intestinal; PI, pure intestinal.

Risk factors of tumor enlargement

Twenty-two lesions increased in size by ≥ 5 mm or 25%, and the cumulative 3-year and 5-year incidence rates of tumor enlargement were 12.1% and 33.9%, respectively (Figure 4a). Table 3 summarizes results of univariate and multivariate analyses of factors associated with tumor enlargement. In the univariate analysis, an initial lesion size ≥10 mm (P < 0.001), erythematous lesion (P = 0.002), HGA (P = 0.002), Ki-67 negativity (P = 0.007), and NPI phenotype (P = 0.001) were significantly associated with tumor enlargement. However, multivariate analysis of the 3 clinically important factors (initial lesion size, dysplasia grade, and mucinous phenotype) showed that only an initial lesion size ≥10 mm (P = 0.010) and NPI phenotype (P = 0.046) were independent predictors of tumor enlargement. Kaplan–Meier curves stratified by initial lesion size, dysplasia grade, and mucinous phenotype are shown in Figure 4b–d, respectively.

[Figure omitted. See PDF]

Table 3.

Univariate and multivariate analyses of the factors associated with tumor enlargement

Risk factors	No. of lesions	No. of events	Univariate analysis		Multivariate analysis
			Hazard ratio (95% CI)	P value	Hazard ratio (95% CI)	P value
Patient characteristics
Age
<70 y	60	8	1.00
≥70 y	61	14	2.322 (0.963–5.597)	0.061
Sex
Male	83	12	1.00
Female	38	10	1.103 (0.472–2.582)	0.820
Lesions characteristics
Location
Preampullary	56	11	1.00
Postampullary	65	11	1.104 (0.474–2.571)	0.819
Initial lesion size
<10 mm	69	5	1.00		1.00
≥10 mm	52	17	6.838 (2.506–18.66)	<0.001	3.940 (1.385–11.21)	0.010
Macroscopic type
Flat or depressed	103	18	1.00
Protruded	18	4	1.727 (0.582–5.123)	0.325
Whitening
Absent	27	3	1.00
Present	94	19	2.080 (0.615–7.036)	0.239
Color
Isochromatic or white	93	13	1.00
Erythematous	28	9	4.057 (1.710–9.624)	0.002
Histological findings of initial biopsy
Vienna classification
LGA (category 3)	71	7	1.00		1.00
HGA (category 4.1)	50	15	4.314 (1.725–10.79)	0.002	1.708 (0.646–4.515)	0.281
p53
Positive	80	13	1.00
Negative	41	9	1.344 (0.573–3.150)	0.496
Ki67
Positive	68	7	1.00
Negative	53	15	3.448 (1.394–8.529)	0.007
Mucin phenotype
PI phenotype	64	3	1.00		1.00
NPI phenotype	57	19	7.456 (2.194–25.33)	0.001	3.790 (1.027–13.98)	0.046

CI, confidence interval; HGA, high-grade adenoma; LGA, low-grade adenoma; NPI, non–pure intestinal; PI, pure intestinal.

Tumor growth rate and a new risk stratification system

We subsequently evaluated whether growth rate trends of tumors of different phenotypes and sizes varied. Figure 5a shows the observation period and degree of tumor enlargement of each lesion, stratified into 4 groups according to mucinous phenotype (PI or NPI) and lesion size (≥10 or <10 mm). The gradient of the approximate line is presented on each graph. The tumor growth rate of the participants with lesion size ≥10 mm and NPI phenotype was the greatest. In that group, 2 lesions showed significant enlargement. One lesion had a PI phenotype on final biopsy; however, the other displayed a gastric-type component on resection, prompting a final diagnosis of NPI phenotype.

[Figure omitted. See PDF]

We developed a new classification system to stratify risk of tumor enlargement and cancer progression during SNDA follow-up (Figure 5b). Three groups were classified based on initial lesion size and the presence of a gastric phenotype. Lesions with an initial size ≥10 mm and gastric phenotype, either initial size ≥10 mm or gastric phenotype, and without either factor were classified to high-risk, moderate-risk, and low-risk groups, respectively. This classification significantly stratified both tumor enlargement (P < 0.001) (Figure 5c) and carcinoma development (P = 0.010) risk (Figure 5d).

DISCUSSION

We have demonstrated the possibility that multiple factors were involved in progression to carcinoma and enlargement of SNDA, particularly showing that initial lesion size ≥10 mm and NPI phenotype were strong risk factors. Furthermore, we proposed a new strategy based on these factors. To our knowledge, this is the first report demonstrating mucinous phenotype as a significant risk factor in the natural course of SNDA.

In recent years, several studies using resected specimens have reported that SNDAs with a gastric phenotype were highly malignant, and classification based on mucinous phenotype is becoming important (16–20). Mucinous phenotypes have been reported to be associated with genetic mutations in other gastrointestinal tumors. In gastric cancer, intestinal-type tumors have a higher frequency of APC and TP53 mutations, whereas gastric-type tumors are more likely to be microsatellite instability high (MSI-H) (24–26). In colorectal cancer, the gastric phenotype is associated with MSI-H (27). MSI-H is associated with aggressive tumor behavior in gastric and colorectal cancers and is a poor prognostic factor of colorectal cancer (28,29). Concerning duodenal tumors, several studies showed that genetic mutations (e.g., GNAS, KRAS) in gastric-type tumors of the stomach were also found in gastric-type tumors of the duodenum (30,31). Particularly, Ishizu et al (31) reported a low mutation rate of the APC gene in nonpapillary duodenal adenocarcinomas, whereas the mutation rate was high in intestinal-type adenomas, suggesting that common intestinal-type adenomas are not likely major precursors of adenocarcinoma. These results suggest that a gastric phenotype likely reflects genetic mutations associated with increased carcinogenesis.

We also found that a lesion size ≥10 mm was associated with subsequent cancer progression and tumor enlargement, which is consistent with previous reports (11,15). Large lesions (≥20 mm) are likely to be treated with piecemeal resection, and treatment itself is associated with a high risk of adverse events, such as perforation and bleeding; thus, it is safer to resect the lesion while it is small (10). On the contrary, no clear relationship between dysplasia grade and tumor progression could be demonstrated in this study, in contrast to the previous report (11). This discrepancy could stem from 2 factors. First, small number of lesions were finally diagnosed as carcinoma. A larger sample size might yield statistically significant results. Second, there was considerable difficulty in assessing the degree of atypia within biopsy specimens, which depended largely on the subjective judgment of the pathologist. Several studies reported that the accuracy of preoperative biopsies was low, which may reflect tumor heterogeneity and difficulty regarding pathologic evaluation (32,33). Contrarily, in this study, mucinous phenotype was evaluated based on whether more than 5% staining was observed through IHC corresponding to either gastric or intestinal type. This method, compared with assessing the dysplasia grade by HE staining, needed less specialized training. Therefore, the mucinous phenotype may be a more objective indicator.

Based on our findings, we developed a new risk classification system for SNDA according to lesion size and the presence of a gastric phenotype (Figure 5b). Considering the risk of tumor enlargement and cancer, we suggest a strong consideration for resection in the high-risk group, and resection or strict follow-up in the moderate-risk group. In the low-risk group, the risk of tumor enlargement and cancer is small; therefore, resection is necessarily not needed. However, given the emergence of treatments with low complication risks such as underwater endoscopic mucosal resection (34) and cold snare polypectomy (35), resection is worth considering when taking into account the medical costs and psychological anxiety associated with observation. Taken together, considering the individual patient risk, such as comorbidities and the administration of anticoagulation therapy, observation can be considered as one option for low-risk group. Moreover, due to the significant difficulty and risk of treatment in the horizontal and ascending parts of duodenum, this strategy could serve as an indicator for choosing observation in such cases. If the course of observation is chosen, it might be suggested to conduct an endoscopic examination every 6–12 months for the moderate-risk group and every 12–24 months for the low-risk group, taking into account the risks of progression to carcinoma, tumor enlargement, and the tumor growth rate. Nevertheless, further research is required to confirm the appropriateness of these surveillance intervals.

One limitation of this strategy is that not all institutes are capable of performing and evaluating IHC for mucinous phenotype. Although the number of capable institutes is expected to increase as the importance of evaluating mucinous phenotype becomes more widely recognized, it is currently not practical to evaluate the mucinous phenotype in all SNDA. In this strategy, cases of lesions >10 mm would fall into the moderate-risk or high-risk group, regardless of the presence or absence of gastric phenotype. Duodenal biopsy reportedly causes lesion fibrosis and hinders treatment (32,36,37). Hence, in cases where the tumor is ≥ 10 mm and treatment is well tolerated, resection without biopsy may be an option. In recent years, attempts have been made to diagnose the mucinous phenotype of SNDA by endoscopy without biopsy. Multiple approaches have been reported, including magnifying endoscopy, crystal violet, and acetic acid, but the process is still in the developing stage (18,19,38,39). If further research makes it possible to diagnose the mucinous phenotype without biopsy, it may be possible to adapt this strategy without biopsy. Although further prospective studies are needed to determine the effectiveness of this strategy, our findings provide a basis for determining the indication for treatment.

This study has some limitations. First, this was a retrospective multicenter study, with treatment strategies and follow-up methods varying among institutions. Ideally, a large-scale prospective study should be conducted; however, given the rarity of SNDA, conducting such a study presents considerable challenges. Particularly, the number of patients with SNDA who choose observation is expected to gradually decrease as treatment options have expanded recently. Therefore, future studies on the natural course of SNDA will become more difficult. Second, the number of lesions eventually diagnosed as carcinoma was small, and the assessment of cumulative incidence of carcinoma is limited. This would reflect differences in the diagnostic criteria of pathologists. However, because there are differences in the diagnosis of SNDA among pathologists, the diagnosis of carcinoma itself is less important, and biological behaviors, such as invasion and tumor enlargement, are more important. Accordingly, we mitigated this limitation by evaluating outcomes of both carcinoma development and tumor enlargement. Third, because not all lesions were eventually resected, pathological evaluation of biopsied and resected lesions differed. In SNDA, several degrees of dysplasia may coexist in one lesion, and making a final diagnosis based solely on biopsy can result in underestimation. Ideally, the final diagnosis should be confirmed by resection. However, because this was a multicenter retrospective study focusing on follow-up cases, all lesions were not resected. Previous reports on the natural course of SNDA also based their final diagnoses on biopsies in nonresected lesions (11–13), which can be regarded as a limitation in studies observing the natural course. To mitigate this limitation, we have also shown the cumulative incidence rate of carcinoma confirmed by resection (Figure 3). Fourth, we did not evaluate the lesion diameter using objective measures. Accurate evaluation of lesion diameter is critical in evaluating tumor enlargement. However, because this study was retrospective and focused on observed lesions, objective evaluation was difficult. Nevertheless, the lesion diameters evaluated by 3 evaluators matched very accurately, and it was acceptable (see Supplementary Figure 2, http://links.lww.com/CTG/B25). Finally, we included only follow-up cases, and selection bias could not be eliminated. It was presumed that malignant-looking lesions were excluded because immediate treatment and that there may be a difference in the distribution of lesions observed in the current study vs the general population. Therefore, our findings may not be generalizable to all SNDAs. Although a new large-scale prospective cohort study is needed to overcome these limitations, such a study is difficult to conduct for the aforementioned reasons. Thus, this study is still valuable.

In conclusion, a lesion size ≥10 mm and the presence of a gastric phenotype on initial biopsy are possible risk factors of subsequent cancer progression and tumor enlargement in SNDA. Effectiveness of the management may be improved by focusing on lesion size and the mucinous phenotype.

CONFLICTS OF INTEREST

Guarantor of the article: Tetsuo Takehara, MD, PhD.

Specific author contributions: R.U., M.K., Y.S., T.Y., Y.H., S.S., and I.H.: designed the study. R.U., M.K., N.S., N.H., M.H., T.A., M.Y., S.H., K.K., H.O., S.Y., S.E., and T.K.: acquired the data. H.K. and E.M.: performed histological assessment. R.U., M.K., and S.Y.: performed endoscopic assessment. R.U.: performed statistical analysis. R.U.: drafted the manuscript. Y.H. and T.T.: performed critical revision. All authors gave final approval of the submitted manuscript.

Financial support: None to report.

Potential competing interests: None to report.

IRB approval statement: This study was approved by the institutional review board of Osaka University Hospital (IRB No. 18059).

Study Highlights

WHAT IS KNOWN

• ✓ Large tumor diameter and high-grade dysplasia are associated with tumor progression in sporadic nonampullary duodenal adenomas (SNDA).

• ✓ Postresection complications of SNDA can be severe and risk factors must be considered.

WHAT IS NEW HERE

• ✓ SNDA with a gastric phenotype is associated with a high malignancy risk.

• ✓ Cases with a gastric phenotype and/or a lesion >10 mm should be considered for resection.

ACKNOWLEDGEMENTS

We thank Drs. Akira Kaneko, Yuhei Wakahara, Syouhei Ouchi, Shuji Wakamatsu (Department of Gastroenterology, Daini Osaka Police Hospital), Mai Horie, Akinori Shimayoshi (Department of Internal Medicine, Osaka Police Hospital), Ryu Ishihara (Department of Gastroenterology Oncology, Osaka International Cancer Institute), Hideki Hagiwara (Department of Gastroenterology, Kansai Rosai Hospital), Syusaku Tsutsui (Department of Gastroenterology, Itami City Hospital), Tsutomu Nishida (Department of Gastroenterology, Toyonaka Municipal Hospital), Hiroyuki Ogawa (Department of Gastroenterology, Nishinomiya Municipal Central Hospital), Kentaro Nakagawa, Hirotsugu Saiki, Keiichi Kimura, Takanori Inoue, and Akihiko Sakatani (Department of Gastroenterology and Hepatology, Osaka University Graduate School of Medicine) for significant assistance with study preparation. We also thank Masaharu Kohara, Takako Sawamura, Megumi Nihei, and Etsuko Maeno (Department of Pathology, Osaka University Graduate School of Medicine) for conducting immunohistochemistry. We thank Editage (www.editage.com) for English language editing.

Author Notes

Correspondence: Tetsuo Takehara, MD, PhD. E-mail: [email protected].

SUPPLEMENTARY MATERIAL accompanies this paper at http://links.lww.com/CTG/B25

REFERENCES

1.Lu, Y, Fröbom, R, Lagergren, J. Incidence patterns of small bowel cancer in a population-based study in Sweden: Increase in duodenal adenocarcinoma. Cancer Epidemiol 2012;36(3):e158–63.

2.Bojesen, RD, Andersson, M, Riis, LB, et al. Incidence of, phenotypes of and survival from small bowel cancer in Denmark, 1994-2010: A population-based study. J Gastroenterol 2016;51(9):891–9.

3.Legué, LM, Bernards, N, Gerritse, SL, et al. Trends in incidence, treatment and survival of small bowel adenocarcinomas between 1999 and 2013: A population-based study in The Netherlands. Acta Oncol 2016;55(9–10):1183–9.

4.Siegel, RL, Miller, KD, Fuchs, HE, et al. Cancer statistics, 2022. CA Cancer J Clin 2022;72(1):7–33.

5.Spigelman, AD, Talbot, IC, Penna, C, et al. Evidence for adenoma-carcinoma sequence in the duodenum of patients with familial adenomatous polyposis. The leeds castle polyposis group (upper gastrointestinal committee). J Clin Pathol 1994;47(8):709–10.

6.Wagner, PL, Chen, YT, Yantiss, RK. Immunohistochemical and molecular features of sporadic and FAP-associated duodenal adenomas of the ampullary and nonampullary mucosa. Am J Surg Pathol 2008;32(9):1388–95.

7.Sakaguchi, Y, Yamamichi, N, Tomida, S, et al. Identification of marker genes and pathways specific to precancerous duodenal adenomas and early stage adenocarcinomas. J Gastroenterol 2019;54(2):131–40.

8.Culver, EL, McIntyre, AS. Sporadic duodenal polyps: Classification, investigation, and management. Endoscopy 2011;43(2):144–55.

9.Jepsen, JM, Persson, M, Jakobsen, NO, et al. Prospective study of prevalence and endoscopic and histopathologic characteristics of duodenal polyps in patients submitted to upper endoscopy. Scand J Gastroenterol 1994;29(6):483–7.

10.Yamasaki, Y, Uedo, N, Takeuchi, Y, et al.Current status of endoscopic resection for superficial nonampullary duodenal epithelial tumors. Digestion 2018;97(1):45–51.

11.Okada, K, Fujisaki, J, Kasuga, A, et al. Sporadic nonampullary duodenal adenoma in the natural history of duodenal cancer: A study of follow-up surveillance. Am J Gastroenterol 2011;106(2):357–64.

12.Cassani, LS, Lanke, G, Chen, HC, et al.Comparison of nonampullary duodenal adenomas in patients with familial adenomatous polyposis versus patients with sporadic adenomas. Gastrointest Endosc 2017;85(4):803–12.

13.Sakaguchi, Y, Tsuji, Y, Ushiku, T, et al. The natural history of sporadic non-ampullary duodenal epithelial tumors: Can we wait and see? DEN Open 2021;1(1):e9.

14.Kakushima, N, Yoshida, M, Iwai, T, et al. A simple endoscopic scoring system to differentiate between duodenal adenoma and carcinoma. Endosc Int Open 2017;5(8):E763–8.

15.Matsueda, K, Uedo, N, Kitamura, M, et al. Pre-ampullary location and size ≥10 mm are independent predictors for high-grade superficial non-ampullary duodenal epithelial tumors. J Gastroenterol Hepatol 2021;36(6):1605–13.

16.Toba, T, Inoshita, N, Kaise, M, et al. Clinicopathological features of superficial non-ampurally duodenal epithelial tumor; gastric phenotype of histology correlates to higher malignant potency. J Gastroenterol 2018;53(1):64–70.

17.Minatsuki, C, Yamamichi, N, Inada, KI, et al. Expression of gastric markers is associated with malignant potential of nonampullary duodenal adenocarcinoma. Dig Dis Sci 2018;63(10):2617–25.

18.Yoshida, M, Shimoda, T, Abe, M, et al. Clinicopathological characteristics of non-ampullary duodenal tumors and their phenotypic classification. Pathol Int 2019;69(7):398–406.

19.Akazawa, Y, Ueyama, H, Tsuyama, S, et al. Endoscopic and clinicopathological features of superficial non-ampullary duodenal tumor based on the mucin phenotypes. Digestion 2021;102(5):663–70.

20.Fukusada, S, Shimura, T, Iwasaki, H, et al. Relationship between immunophenotype and clinicopathological findings for superficial nonampullary duodenal epithelial tumor. Digestion 2021;102(6):870–7.

21.Dixon, MF. Gastrointestinal epithelial neoplasia: Vienna revisited. Gut 2002;51(1):130–1.

22.Yoshimura, N, Goda, K, Tajiri, H, et al. Endoscopic features of nonampullary duodenal tumors with narrow-band imaging. Hepatogastroenterology 2010;57(99–100):462–7.

23.The Paris endoscopic classification of superficial neoplastic lesions: Esophagus, stomach, and colon: November 30 to December 1, 2002. Gastrointest Endosc 2003;58(6 Suppl):S3–43.

24.Yamazaki, K, Tajima, Y, Makino, R, et al. Tumor differentiation phenotype in gastric differentiated-type tumors and its relation to tumor invasion and genetic alterations. World J Gastroenterol 2006;12(24):3803–9.

25.Oue, N, Sentani, K, Sakamoto, N, et al. Clinicopathologic and molecular characteristics of gastric cancer showing gastric and intestinal mucin phenotype. Cancer Sci 2015;106(8):951–8.

26.Endoh, Y, Sakata, K, Tamura, G, et al. Cellular phenotypes of differentiated-type adenocarcinomas and precancerous lesions of the stomach are dependent on the genetic pathways. J Pathol 2000;191(3):257–63.

27.Biemer-Hüttmann, AE, Walsh, MD, McGuckin, MA, et al. Mucin core protein expression in colorectal cancers with high levels of microsatellite instability indicates a novel pathway of morphogenesis. Clin Cancer Res 2000;6(5):1909–16.

28.Kim, JY, Shin, NR, Kim, A, et al. Microsatellite instability status in gastric cancer: A reappraisal of its clinical significance and relationship with mucin phenotypes. Korean J Pathol 2013;47(1):28–35.

29.Lenz, HJ, Ou, FS, Venook, AP, et al. Impact of consensus molecular subtype on survival in patients with metastatic colorectal cancer: Results from CALGB/SWOG 80405 (alliance). J Clin Orthod 2019;37(22):1876–85.

30.Hida, R, Yamamoto, H, Hirahashi, M, et al. Duodenal neoplasms of gastric phenotype: An immunohistochemical and genetic study with a practical approach to the classification. Am J Surg Pathol 2017;41(3):343–53.

31.Ishizu, K, Hashimoto, T, Naka, T, et al. APC mutations are common in adenomas but infrequent in adenocarcinomas of the non-ampullary duodenum. J Gastroenterol 2021;56(11):988–98.

32.Kinoshita, S, Nishizawa, T, Ochiai, Y, et al. Accuracy of biopsy for the preoperative diagnosis of superficial nonampullary duodenal adenocarcinoma. Gastrointest Endosc 2017;86(2):329–32.

33.Kakushima, N, Kanemoto, H, Sasaki, K, et al. Endoscopic and biopsy diagnoses of superficial, nonampullary, duodenal adenocarcinomas. World J Gastroenterol 2015;21(18):5560–7.

34.Yamasaki, Y, Uedo, N, Takeuchi, Y, et al. Underwater endoscopic mucosal resection for superficial nonampullary duodenal adenomas. Endoscopy 2018;50(2):154–8.

35.Maruoka, D, Matsumura, T, Kasamatsu, S, et al. Cold polypectomy for duodenal adenomas: A prospective clinical trial. Endoscopy 2017;49(8):776–83.

36.Kakushima, N, Ono, H, Takao, T, et al. Method and timing of resection of superficial non-ampullary duodenal epithelial tumors. Dig Endosc 2014;26(Suppl 2):35–40.

37.Tsuji, S, Doyama, H, Tsuji, K, et al. Preoperative endoscopic diagnosis of superficial non-ampullary duodenal epithelial tumors, including magnifying endoscopy. World J Gastroenterol 2015;21(41):11832–41.

38.Toya, Y, Endo, M, Oizumi, T, et al. Diagnostic algorithm of magnifying endoscopy with crystal violet staining for non-ampullary duodenal epithelial tumors. Dig Endosc 2020;32(7):1066–73.

39.Tanaka, Y, Fujii, S, Okada, K, et al. Magnifying narrow-band imaging with acetic acid spray for superficial non-ampullary duodenal epithelial tumors based on the mucin phenotype. Digestion 2022;103(6):421–7.