Moderate Splenic Injury Caused by Colonoscopy

Abstract

Little is known about iatrogenic splenic injury (SI) as an adverse event after colonoscopy. SI is sometimes fatal because of hemorrhaging. We herein report a man who developed SI after colonoscopy. He recovered conservatively. His history of left hydronephrosis and insertion with a maximally stiffened scope were suspected as possible risk factors. Endoscopists should consider the possibility of SI when they encounter patients suffering from left-sided abdominal pain after colonoscopy. Careful interview concerning the medical history and gentle maneuvering around the splenic flexure can help avoid SI.

Introduction

The most common adverse events (AEs) caused by screening colonoscopy are bleeding and perforation [approximately 0.33% and 0.067%, respectively (1)], and pretreatment before colonoscopy can also cause ileus and electrolyte imbalance. While these AEs are well known to most endoscopists, little is known about splenic injury (SI) induced by colonoscopy.

The spleen is one of the most commonly injured solid organs in patients with abdominal trauma. The management of SI has changed significantly. In the 1950s, splenectomy was performed for cases of ruptured spleen. However, postsplenectomy infection is a well-known complication, and splenic-preserving techniques have been encouraged (2). According to the latest guidelines for splenic trauma, if the hemodynamic status is totally stable, observation is recommended, and when the status can be stabilized by transfusions, splenic embolization is recommended, with splenectomy usually considered the last choice (3). To predict the prognosis, a grading system based on computed tomography (CT) findings has proven very useful (3). Although SI caused by colonoscopy is a rare complication (approximately 0.001%) (4) and the symptoms and their time of onset after colonoscopy vary (5), failure to recognize this rare complication can lead to unexpected death [mortality rate 5% (6)].

We herein report a 62-year-old man who suffered from strong epigastric pain a few hours after colonoscopy. This clinical course may be instructive for all medical staff, especially endoscopists.

Case Report

The patient was a 62-year-old man whose general condition was good. He had a medical history of left-sided urinary tract stones (size range 10-15 mm) and hydronephrosis treated with extracorporeal shock wave lithotripsy (ESWL) 3 times 4 years ago, as well as surgical treatment for hemorrhoids, fibula fracture, and sciatic neuralgia. He had undergone colonoscopy one year earlier for the first time, and a colonic adenoma had been endoscopically resected. There were no detailed records concerning the difficulties encountered or time taken for the procedure.

This time, for the second time in his life, he underwent follow-up colonoscopy in ambulatory practice. Since the patient recalled the previous colonoscopy being painful, pethidine (35 mg) was intravenously administered with butylscopolamine (20 mg) to reduce bowel peristalsis. The endoscopist was an experienced attending physician of the Japan Gastroenterological Endoscopy Society who had graduated from a medical school more than 10 years ago with experience of over 5,000 cases of colonoscopy. Colonoscopy was performed using a CF-HQ290ZI (Olympus, Tokyo, Japan), and CO₂ was supplied for intestinal dilatation. The position was changed from the supine position to the right lateral position to pass through the S/F because of some difficulties, and the colonoscope was eventually inserted without any external pressure maneuver. The endoscopist usually set the stiffness of the scope to maximum hardness during the procedure. It took approximately eight minutes to reach the cecum. One adenoma (5 mm in diameter) was found in the sigmoid colon, and cold snare polypectomy was performed. No other significant findings were observed for the colon. The total time spent on the procedure was approximately 20 minutes.

Since the patient reported no pain during the procedure, no additional sedatives or pain killers were administered. After the procedure, his blood pressure (BP) was 136/92 mmHg, and his pulse was 98 bpm. His consciousness level was normal. He complained of epigastric pain immediately after colonoscopy, and the pain continued for a while. The abdomen was soft, but the doctor considered the abdominal pain to be due to a distended colon. After resting in the recovery room of the hospital for approximately 30 minutes, the pain was relieved, and the patient returned home without any medication having been prescribed.

A few hours later, however, during his dinner, the patient experienced unbearable pain; he returned to the hospital, and the doctor on duty examined him. He underwent contrast-enhanced CT and blood tests. Compared with CT performed one year earlier during follow-up for left urinary tract stone and hydronephrosis (Fig. 1A, 2A), imaging showed liquid retention under the left diaphragm (Fig. 1B, 2B), and a blood test showed an elevated white blood cell count (16,600 /μL, normal range: 3,300-8,600 /μL) without anemia (hemoglobin 14.7 g/dL, normal range for men: 13.7-16.8 g/dL) or elevated C-reactive protein (CRP) levels. The patient was transferred to our hospital with a diagnosis of suspected colonic perforation and peritoneal hemorrhaging.

Figure 1.

Chronological changes in CT findings (axial view). (A) Plain CT performed one year earlier for follow-up of the left ureteric stone and hydronephrosis. (B) Contrast CT was performed at the onset after colonoscopy to evaluate abdominal pain. (C) Contrast CT performed two months later to observe the change after splenic injury. CT: computed tomography, S/F: splenic flexure of the colon. Black arrows, spleen; black arrowheads, splenic subcapsular hematoma; white arrows, splenic flexure of the colon.

Figure 2.

Chronological changes in CT findings (coronary view). (A) Plain CT performed one year prior for follow-up of the left ureteric stone and hydronephrosis. (B) Contrast CT was performed at the onset after colonoscopy to evaluate abdominal pain. (C) Contrast CT performed two months later to observe the change after splenic injury. CT: computed tomography, S/F: splenic flexure of the colon. Black arrows, spleen; black arrowheads, splenic subcapsular hematoma; white arrows, splenic flexure of the colon; white arrowheads, pelvic fluid collection.

In our hospital, the findings on contrast CT (Fig. 1B, 2B) were interpreted as splenic injury with subcapsular hematoma involving more than 50% of the surface area, without active bleeding. There was displacement of the splenic flexure (S/F) of the colon downward caudally due to the pressure of the splenic hematoma. Accumulation of fluid in the pelvis was also observed. This corresponded to grade III (moderate) damage according to the American Association for the Surgery of Trauma (AAST) imaging criteria (7). Abdominal contrast CT was performed again at our hospital four hours later to examine the acute change in splenic injury; the images did not show extravasation or any other changes from the earlier CT findings. Since his vital signs were stable (BP 169/80 mmHg with a heart rate of 78 bpm), and no apparent active bleeding was detected, he was admitted to our hospital with a plan to follow a wait-and-watch approach. The patient was given bed rest without any additional treatment or blood transfusion.

The next day, the vital signs had stabilized; there was no sign of active bleeding or progression to anemia. Abdominal ultrasonography was performed to monitor the size of the hematoma in the spleen, demonstrating stable circulation in the parenchyma of the spleen with no significant change in the surrounding subcapsular hematoma (Fig. 3). The abdominal pain gradually subsided and completely diminished on the fifth day. The patient was discharged on day six of hospitalization.

Figure 3.

Abdominal ultrasonography with Doppler imaging the day after the onset of splenic injury.

Two months later, the patient underwent follow-up CT at the previous hospital, and the images showed that the splenic subcapsular hematoma had decreased in size (Fig. 1C, 2C). The patient has not reported any AEs since his discharge.

Discussion

We encountered a man who developed moderate SI after colonoscopy and required noninvasive management, eventually recovering after bed rest for a few days in the hospital, along with the chronological changes in the CT findings of SI.

Colonoscopy is a relatively safe procedure, and fatal complications are rare. According to a report based on the Japan Endoscopy Database (JED), the overall rate of occurrence of complications of colonoscopy was 0.43% (1). Among them, bleeding (58.8%), perforation (11.8%), abdominal pain (11.8%), hypotension (11.8%), and hyperpnea (5.9%) were included (1); no cases of SI were reported. In general, the incidence of SI due to colonoscopy is estimated to be very low (1 per 100,000 colonoscopies) (8). However, Rex reported that the rate was approximately 1 per 6,000 colonoscopies based on 2000-2007 U.S. Nationwide Inpatient Sample for ICD-9 codes and warned that “colonoscopy-related SI appears greatly underreported” (7). In Japan, 11 cases of SI after colonoscopy were reported between 1998 and 2021 (9). Therefore, it is possible that, in Japan, the actual occurrence of SI by colonoscopy is very low, or endoscopists do not notice the injury itself, partly due to a lack of knowledge.

SI caused by colonoscopy is seldom mentioned in the guidelines or books for maneuvers in Japan. In our case, if the symptoms had been milder, CT might not have been performed, and SI would not have been detected. This implies that some cases of postcolonoscopy SI with spontaneous recovery might be overlooked. Since SI can be fatal if the diagnosis is delayed, vital signs need to be monitored routinely during the procedure, and SI should be included in the representative AEs of colonoscopy to alert every endoscopist about possible occurrence.

Among the cases of iatrogenic SI, colonoscopy is by far the most frequent cause, although other procedures, including abdominal surgeries, pulmonary surgery, and even endoscopic retrograde cholangiopancreatography, esophagogastroduodenoscopy, or ESWL, can also cause SI, albeit very rarely (8). Aubrey-Bassler et al. reported that colonoscopy was the medical procedure reported most frequently as a cause of splenic rupture (77.7%, 87 of 112 cases) (10). According to previous reports (5,11), the major risk factors for SI are female sex, a history of abdominal surgery (adhesions related to previous operations), splenomegaly, smoking, repetitive colonoscopy, treatment colonoscopy, inflammatory bowel disease, inflammatory disease including diverticulosis, anticoagulant use, difficult colonoscopy, rushing during the procedure, and insufficient visualization due to inadequate bowel cleansing. From an anatomical point of view, both pulling the splenocolic ligament and pushing the spleen using an endoscope can be triggers; colonoscopy techniques, such as hooking the S/F of the colon to straighten the lower colon, pushing by hand on the left hypochondrium, sliding by advancement of colonoscopy, and even the alpha loop method are maneuvers reported to be responsible for such cases (8). The endoscopist in the present case was well experienced [more colonoscopies and a longer period since MD graduation were reported to be correlated with fewer severe AEs (12)], and the colonoscope used in this case was a CF-HQ290ZI, one of the safest scopes available, with an adjustable stiffness. However, the endoscopist habitually favors maximally hard shaft stiffness over a softer stiffness to facilitate smooth insertion. Given that changing position from supine to right lateral position was necessary in our case to pass through the S/F, SI might have occurred when the endoscopist experienced some difficulties around the S/F in the supine position.

Follow-up CT performed 2 months later showed the proximity of the S/F and the residual hematoma in the spleen (Fig. 1C, 2C). In addition, our patient had a history of repetitive ESWL for left-sided urinary tract stones and hydronephrosis four years prior, which might have been additional risk factors causing inflammation around the left kidney, possibly leading to peritoneal adhesion and decreased mobility around the spleen due to its proximity. Taken together, these findings suggest that the splenocolic ligament, which might have been stiffened due to the previous urinary tract stone-related inflammation, may have been excessively stretched or the spleen directly injured by the extremely stiff scope. Therefore, in addition to taking a patient's detailed history concerning abdominal diseases, carefully maneuvering around the S/F with a less-stiff scope might be appropriate precautions to take to prevent SI.

Several other risk factors associated with SI after colonoscopy should also be mentioned. According to a previous report, one-third of SI cases were associated with a biopsy or polypectomy (5). In the present case, adenoma in the sigmoid colon was resected, and the direct effect seemed to be very low due to its distance. Conscious sedation for colonoscopy is reported to be associated with a higher frequency of complications (13). In the present case, the patient was administered pethidine, an opioid pain-relief medicine. Therefore, although he was fully awake during the examination, the pain immediately after the onset of SI might have been suppressed by the drug. The position of the spleen might also have influenced the occurrence of SI. Anatomically, the spleen is fixed retroperitoneally by ligaments, and anatomical changes can make some clinical procedures difficult to perform, allowing complications to occur (14). In cases where the spleen is located cranially right under the diaphragm, the splenocolic ligament and S/F also tend to be located cranially, and the S/F angle tends to be sharp. With a sharp-angled S/F, the scope must acutely bend to fit into the S/F curve, requiring more force than usual to advance, which can lead to direct trauma to the spleen. In addition, since the spleen falls posteriorly away from the S/F in the supine position, insertion of the colonoscope in the supine position might exert opposing traction on the splenocolic ligament and increase the risk of SI (5). Since body weight is reported to be related to the location of the spleen (14), evaluating body size as a possible potential risk factor might be valuable in the future.

The majority of patients develop symptoms within the first 24 hours after colonoscopy. However, in some cases, the symptoms cannot be detected until more than 48 hours after colonoscopy and can be lethal (5). The mortality of SI related to colonoscopy is reported to be approximately 5% (6). Enhanced abdominopelvic CT is recommended to assess the severity of SI, but an assessment with sonography is also useful for a rapid check (15). The treatment should be chosen according to the vital signs; if the vital signs are stable with or without transfusion, spleen-preserving methods, such as intravascular embolization, should be attempted before splenectomy (3). In our case, the CT images in our hospital showed liquid retention in the inferior area of the liver and pelvis, with a partly lacerated spleen and an intraparenchymal hematoma larger than 5 cm. Therefore, we diagnosed the patient with grade III severity on the AAST Spleen Injury Scale. Given that no active bleeding was detected, his vital signs were normal, and he had no anemia, we decided to take a wait-and-watch approach while being fully prepared for any emergency situation based on previous reports of grade III cases where splenectomies were performed (8).

Conclusion

Iatrogenic SI caused by colonoscopy is rare in Japan. However, some cases may have been overlooked due to a lack of knowledge, especially late-onset cases. If a patient complains of left-sided abdominal pain after colonoscopy, an SI should be considered as a differential diagnosis. Since some clinical events causing peritoneal adhesion (left urinary tract stone and hydronephrosis in this case) can be risk factors, endoscopists must ensure that they take a detailed history in each patient concerning any past illnesses, determine the most appropriate endoscope to use to reduce the risk of SI, and perform gentle maneuvers around the S/F with a thin and less-stiff scope. Other modalities, such as CT colonography and capsule colonoscopy, might also be substitutable, especially for high-risk patients. The further accumulation of data concerning risk factors, including endoscopy and anatomical predisposing factors, is expected to help clarify the occurrence of SI.

The authors state that they have no Conflict of Interest (COI).