Characteristics of Adult Intussusception due to Malignancy in Japanese Patients
Funding: The authors received no specific funding for this work.
ABSTRACT
Aims
Adult intussusception (AI) is often associated with organic diseases. However, few studies have examined the causes of AI in Japanese patients. This study aimed to elucidate the clinical characteristics of AI due to malignancy in Japanese patients.
Methods and Results
From 2013 to 2021, 54 Japanese patients with AI (≥ 20 years) diagnosed at our hospital were enrolled and divided into two groups according to the cause of AI (malignancy group, n = 26; other diseases group, n = 28). The patients' clinical characteristics were retrospectively evaluated. Patients in the malignancy group were significantly older than those in the other diseases group (p < 0.001). The cutoff value for age as a factor associated with AI due to malignancy was 64 years (area under the curve: 0.78, 95% confidence interval [CI]: 0.65–0.90). The frequency of chronic symptoms (> 14 days) in AI due to malignancy was significantly higher than the frequencies of acute (≤ 4 days) and subacute (4–14 days) symptoms (p = 0.010 and p = 0.027, respectively). The colonic type of AI was significantly more common than the small intestinal and ileocecal types in the malignancy group (both p < 0.001). Multivariate analysis showed that age of ≥ 64 years, chronic symptoms, and the colonic type were independently associated with AI due to malignancy (adjusted odds ratio [OR] 16.00, 95% CI 1.23–208.00; adjusted OR 32.70, 95% CI 1.50–712.00; adjusted OR 31.20, 95% CI 2.68–363.00, respectively).
Conclusion
Advanced age (≥ 64 years), chronic symptoms, and AI in the colon are characteristics of AI due to malignancy.
1 Introduction
Adult intussusception (AI) differs from pediatric intussusception in that many cases are associated with organic diseases [1]. AI due to malignancy is particularly important because it significantly impacts the patient's prognosis. However, AI itself is difficult to diagnose through interviews and physical examinations only [2], and a preoperative diagnosis is generally established on the basis of abdominal ultrasonography [3, 4] and computed tomography [5-7] findings. Moreover, it is difficult to determine whether AI is caused by malignancy or other diseases [8]. Preoperative reduction is generally performed depending on the cause of AI [9, 10]. Elucidating the clinical features of AI due to malignancy is very important to determine the choice of treatment after diagnosis of AI, because AI is rare with few reported clinical features. Studies involving more than 20 cases of AI have been reported from several countries [2, 6, 11-18]; however, the clinical features of AI in Japanese patients have been reported in only one study [19]. In addition, the characteristics of AI due to malignancy have not been well described [14, 15, 20]. Accordingly, the aim of this study was to elucidate the clinical characteristics of malignancy-related AI in Japanese patients.
2 Methods
2.1 Study Population
Between 2013 and 2021, 58 Japanese patients aged ≥ 20 years were diagnosed with AI on the basis of clinical, ultrasound, and computed tomography examinations at Ehime Central Hospital. One patient with no identifiable cause was excluded. Two patients with malignant lymphoma and one patient with metastasis to other organs were excluded because the number of malignant diseases other than primary carcinoma was small, and the treatment methods for primary carcinoma and malignant lymphoma/multiorgan cancer metastasis were different. Finally, 54 patients were enrolled in this study. To identify the factors associated with AI due to malignancy, we divided the patients into two groups: AI due to malignancy (n = 26) and AI due to other diseases (n = 28). Clinical characteristics, including age, sex, time from onset to diagnosis of AI, disease type, anemia, and hypoalbuminemia, were retrospectively evaluated and compared between the two groups.
2.2 Classification of Symptoms
Depending on the overall time from onset to diagnosis of AI, the symptoms were classified as acute (≤ 4 days), subacute (4–14 days), and chronic (> 14 days) according to the criteria provided by Wang et al. [16].
2.3 Measurements of Other Clinical Data
Body mass index (BMI) and blood test results obtained on the day of admission were evaluated.
2.4 Definition of Anemia and Hypoalbuminemia
Anemia was defined as a hemoglobin level of < 14 g/dL in men and < 12 g/dL in women. Hypoalbuminemia was defined as an albumin level of ≤ 3.5 g/dL.
2.5 Statistical Analysis
All statistical analyses were performed using Easy-R (EZR) version 1.37 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). The paired Wilcoxon rank-sum test was used to compare continuous variables, while Fisher's exact test was used to compare nominal variables. The familywise error rate (alpha inflation) was adjusted using the Holm method [21]. A p value of < 0.05 was considered statistically significant. To evaluate the discriminatory ability of age for AI due to malignancy, the area under the curve (AUC) was calculated using receiver operating characteristic (ROC) curve analysis, and the cutoff value was obtained using the Youden index. Multivariate analysis was performed to identify the factors associated with AI due to malignancy.
2.6 Ethical Considerations
The study protocol was established in accordance with the 1964 Declaration of Helsinki and its subsequent versions, and the study was approved by the Institutional Ethics Committee of Ehime Prefectural Central Hospital (approval no. 03-75). Informed consent was obtained from all participants in the form of opt-out on the website. Patients who refused the use of their data for this study were excluded from the analysis.
3 Results
3.1 Clinical Findings
Table 1 shows the baseline clinical characteristics of patients with AI. The median age of the 54 patients was 71 years, and 30 patients (55.6%) were female. The most common complaint was abdominal pain (57.4%); other symptoms included vomiting, diarrhea, bloody stools, decreased appetite, black stools, and weight loss. Eight (14.8%) patients were asymptomatic, and AI was incidentally detected in these patients. Acute (≤ 4 days) symptoms were the most common (n = 24; 44.4%), followed by subacute (4–14 days) (n = 7; 13.0%) and chronic (> 14 days) (n = 15; 27.8%) symptoms. The eight asymptomatic patients were considered unclassifiable because of the lack of information. AI was classified into three types according to the anatomical site: small intestine (small intestinal type), ileocecal tract (ileocecal type), and colon (colonic type); these were observed in 15 (27.8%), 10 (18.5%), and 29 (53.7%) patients, respectively.
| AI (N = 54) | |
|---|---|
| Age, yearsa | 71 (57–78) |
| Sex, male/female | 24/30 |
| Body mass index, kg/m2a | 19.5 (18.2–21.8) |
| Symptoms, N [%] | |
| Abdominal pain | 31 (57.4%) |
| Vomiting | 13 (24.1%) |
| Diarrhea | 9 (16.7%) |
| Bloody stools | 9 (16.7%) |
| Decreased appetite | 6 (11.1%) |
| Black stools | 4 (7.4%) |
| Weight loss | 3 (5.6%) |
| Others | 9 (16.7%) |
| Asymptomatic | 8 (14.8%) |
| Time from onset to diagnosis of AI, N [%] | |
| Acute (≤ 4 days) | 24 (44.4%) |
| Subacute (4–14 days) | 7 (13.0%) |
| Chronic (> 14 days) | 15 (27.8%) |
| Unclassifiable | 8 (14.8%) |
| Disease type, N [%] | |
| Small intestine | 15 (27.8%) |
| Ileocecal | 10 (18.5%) |
| Colonic | 29 (53.7%) |
| White blood cells, /μLa | 6820 (5300–9110) |
| Hemoglobin, g/dLa | 11.9 (9.1–13.4) |
| Albumin, g/dLa | 3.6 (2.7–4.0) |
| C-reactive protein, mg/dLa | 0.6 (0.1–2.6) |
- Abbreviation: AI, adult intussusception.
- a Median (interquartile range).
3.2 Etiology
The etiology of AI was benign disease in 13 (24.1%) patients, malignant disease in 26 (48.1%), and idiopathic disease in 15 (27.8%) (Table 2). Lipoma was the most common benign disease (n = 7; 13.0%), followed by inflammatory fibroid polyp (n = 2; 3.7%) and other diseases such as Cronkhite-Canada syndrome, appendiceal mucous cyst, colonic diverticulum, and colorectal adenoma. Malignant diseases included colorectal carcinoma in 25 patients and small intestine carcinoma in one. Malignant diseases were most frequently located in the cecum, followed by the transverse colon. The most common macroscopic morphology was Type 1, and the most common depth of invasion was T3 (see table in Supporting Information S1).
| No. | ||
|---|---|---|
| Idiopathic (15) | ||
| Benign (13) | Lipoma | 7 |
| Inflammatory fibroid polyp | 2 | |
| Colorectal adenoma | 1 | |
| Cronkhite-Canada syndrome | 1 | |
| Appendiceal mucous cyst | 1 | |
| Colonic Diverticulum | 1 | |
| Malignant (26) | Colorectal carcinoma | 25 |
| Small intestine carcinoma | 1 |
3.3 Treatment
Surgical, endoscopic, and conservative treatments were implemented for 36 (66.7%), five (9.2%), and 13 (24.1%) patients, respectively (see table in Supporting Information S2). Among surgically treated patients, preoperative reduction was performed for nine (25.0%) patients.
3.4 Analysis of Factors Associated With AI due to Malignancy
To identify the factors associated with AI due to malignancy, we divided the patients into two groups: AI due to malignancy (n = 26) and AI due to other diseases (n = 28) (Table 3). There were no between-group differences in sex distribution and the frequency of anemia or hypoalbuminemia; however, patients in the malignancy group were significantly older than those in the other diseases group (p < 0.001). Moreover, patients with chronic symptoms were significantly more likely to have AI due to malignancy than those with acute and subacute symptoms (p = 0.010 and p = 0.027, respectively, after Holm correction). The colonic type of AI was significantly more common than the small intestinal and ileocecal types in the malignancy group (both p < 0.001, after Holm correction) (Table 3). To determine whether age could determine AI due to malignancy, the AUC and cutoff value for age were calculated using ROC analysis. The AUC was 0.78 (95% confidence interval [CI]: 0.65–0.90, see figure in Supporting Information S3), with a cutoff age of 64 years. Accordingly, patients with AI were divided into those aged ≥ 64 years and those aged < 64 years. Multivariate analysis including advanced age (≥ 64 years), sex, time from onset to diagnosis of AI, and disease type as variables revealed that an age of ≥ 64 years, chronic symptoms, and the colonic type were independently associated with AI due to malignancy (adjusted odds ratio [OR] 16.00, 95% CI 1.23–208.00; adjusted OR 32.70, 95% CI 1.50–712.00; adjusted OR 31.20, 95% CI 2.68–363.00, respectively; Table 4).
| AI due to malignancy (n = 26) | AI due to other etiologies (n = 28) | p b | |
|---|---|---|---|
| Age, yearsa | 77 (69–84) | 60 (52–71) | < 0.001 |
| Advanced age (≥ 64 years), N [%] | 23 (88.5%) | 10 (35.7%) | < 0.001 |
| Sex, male/female | 9/17 | 15/13 | 0.183 |
| Time from onset to diagnosis of AI, N | |||
| Acute (A) (≤ 4 days) | 9 | 15 | 1.000 (A vs. S) |
| Subacute (S) (4–14 days) | 2 | 5 | 0.027 (S vs. C) |
| Chronic (C) (> 14 days) | 13 | 2 | 0.010 (C vs. A) |
| Disease type, N | |||
| Small intestine (SI) | 1 | 14 | 1.000 (SI vs. I) |
| Ileocecal (I) | 1 | 9 | < 0.001 (I vs. C) |
| Colonic (C) | 24 | 5 | < 0.001 (C vs. SI) |
| Anemia (male < 14 g/dL, female < 12 g/dL) | 19 (73.1%) | 17 (60.7%) | 0.257 |
| Hypoalbuminemia (≤ 3.5 g/dL) | 15 (57.7%) | 10 (35.7%) | 0.156 |
- Abbreviation: AI, adult intussusception.
- a Median (interquartile range).
- b The Wilcoxon rank-sum test was used to compare age between patients with AI due to malignancy and those with AI due to other etiologies. Fisher's direct probability test was used to compare older age, sex, time from onset to diagnosis of AI, disease type, anemia, and hypoalbuminemia between AI due to malignancy and AI due to other etiologies. The p values for time from onset to diagnosis of AI and disease type were corrected using the Holm method.
| Variable | Prevalence (%) | Crude OR (95% CI) | Adjusted ORa (95% CI) |
|---|---|---|---|
| AI due to malignancy | |||
| Advanced age (≥ 64 years) | |||
| Yes | 23/33 (69.7) | 13.80 (3.30–57.70) | 16.00 (1.23–208.00) |
| No | 3/21 (14.3) | Reference | Reference |
| Sex | |||
| Male | 9/24 (37.5) | 0.46 (0.15–1.37) | 0.22 (0.02–2.23) |
| Female | 17/30 (56.7) | Reference | Reference |
| Time from onset to diagnosis of AI | |||
| Chronic (> 14 days) | 13/15 (86.7) | 11.80 (2.25–62.20) | 32.70 (1.50–712.00) |
| Acute (≤ 4 days) + subacute (4–14 days) | 11/31 (35.5) | Reference | Reference |
| Disease type | |||
| Colonic | 24/29 (82.8) | 55.20 (9.72–313.00) | 31.20 (2.68–363.00) |
| Ileocecal + small intestine | 2/25 (8.0) | Reference | Reference |
- Abbreviations: AI, adult intussusception; CI, confidence interval; OR, odds ratio.
- a Multivariate analysis was performed with advanced age (≥ 64 years), sex, time from onset to diagnosis of AI, and disease type as variables.
4 Discussion
The present study revealed that AI due to malignancy was significantly more frequent in elderly patients (≥ 64 years old) and patients with chronic symptoms; moreover, patients with AI in the colon were more likely to have AI due to malignancy. To the best of our knowledge, this is the first report on the clinical characteristics of Japanese patients with AI due to malignancy.
Several studies have reported the rates of AI due to malignancy; these were 27% (21/77), 38% (9/24), 20% (4/20), 34% (16/47), 57% (25/44), and 39% (13/33) in a Korean study of 77 patients [15], a Taiwanese study of 24 patients [16], a Turkish study of 20 patients [17], another study from Turkey of 47 patients [18], a Japanese study of 44 patients [19], and another Korean study of 33 patients [20], respectively.
However, evidence regarding the characteristics of AI due to malignancy is limited; only three studies have reported these characteristics. A study of 60 patients with AI conducted in Singapore reported that the site of intussusception and anemia were independent factors associated with AI due to malignancy [14]. The remaining two studies were conducted in Korea. In one study, chronic symptoms and the colonic type were associated with AI due to malignancy in univariate analysis [20]. The other Korean study involved 77 patients with AI and showed that chronic symptoms and the colonic type were independently associated with AI due to malignancy in multivariate analysis [15]. In the present study as well, chronic symptoms and the colonic type of AI were identified as independent factors associated with AI due to malignancy. Symptoms were classified depending on the overall time from onset to diagnosis of AI. In patients with AI due to malignancy, symptoms determining the onset date may not have been symptoms of AI; rather, they may have been symptoms of the carcinoma itself. This may have prolonged the time between onset and diagnosis of AI due to malignancy, and the symptoms would have been considered chronic. In addition, the colonic type may have been identified as a related factor because colorectal carcinoma accounted for the majority of malignant diseases in this study.
In this study, advanced age (≥ 64 years) was independently associated with AI due to malignancy. The median age in previous studies reporting the characteristics of AI due to malignancy was 50.5 [14], 58.5 [15], and 50.5 [20] years, whereas, the median age in our study was 71 years. The incidence of colorectal cancer has been reported to increase rapidly after the mid-60s in Japanese individuals [22]. This might explain the higher frequency of AI due to malignancy in patients aged ≥ 64 years in the present study.
Reduction is performed before bowel resection to prevent short bowel syndrome in cases of AI due to idiopathic or benign disease [9]. In contrast, en bloc resection without reduction is recommended for AI caused by malignancy because of the risk of perforation and spillage of microorganisms and malignant cells [10]. Among the 36 patients who underwent surgery in this study, all eight patients who met the three criteria had malignancy, whereas, the seven patients who did not meet all three criteria did not have malignancy. Therefore, the clinical features of AI due to malignancy identified in this study may help to decide the treatment for AI.
This study has some limitations. First, it was a retrospective study, and there may be potential inherent biases in patient selection and data collection. Second, the sample size was relatively small, which may limit the generalizability of our findings. Third, our research did not consider possible variations in regional medical practices that may affect the diagnosis and treatment of AI. Nevertheless, our research has some strengths, including the focus on a specific population and the careful assessment of clinical characteristics, and we believe the findings will increase the understanding of AI due to malignancy.
In conclusion, the results of this study suggest that advanced age (≥ 64 years), chronic symptoms, and AI in the colon are characteristics of AI due to malignancy. For patients exhibiting these factors, a close examination of the cause of AI is necessary, even if AI is resolved with conservative treatment.
Acknowledgments
We thank the Gastroenterology Center, Ehime Prefectural Central Hospital, and Endoscopy Center for their support.
Ethics Statement
The study protocol was developed in accordance with the 1964 Declaration of Helsinki and subsequent versions of the ethical guidelines, and it was approved by the Institutional Ethics Committee of Ehime Prefectural Central Hospital (approval no. 03-75).
Consent
Informed consent was obtained from all participants in the form of opt-out on the website. Patients who refused the use of their data for this study were excluded.
Conflicts of Interest
The authors declare no conflicts of interest.
Open Research
Data Availability Statement
The data supporting the findings of this study are available from the corresponding author upon reasonable request.
