After performing a Medline search using the keyword colonoscopy, splenic rupture, splenic injury, splenic trauma, we found 49 citations, 43 of which in English language, for a total of 54 cases of splenic rupture after colonoscopy. Overall there is a female sex preponderance (sex ratio of 3.8/1) and an average age of 62 years (range 29–85). In 13 cases the patient has a previous history of intra-abdominal operations. Twenty-one of the colonoscopies were simply diagnostic, one was done in conjunction with an upper endoscopy, 3 included biopsies and 15 had concurrent polypectomies. Only three cases were described as modestly difficult.
Only one patient was on oral anticoagulation therapy with Warfarin.
The onset of symptoms was usually immediate (within 24 hours from the exam), but in some cases was delayed by several days [2, 4, 7, 24, 26, 27, 33, 36–38, 46].
Although the diagnosis of the first reported case was made by angiography, the Computed Tomography (CT scan) is the main test utilized now days. In previous series (before 1991) the most common diagnostic modality was exploratory laparotomy (10 cases), followed by ultrasonography and CT scan (2 cases), diagnostic peritoneal lavage (1 case) and post-mortem (1 case). The treatment was by emergent laparotomy in the majority of the cases, whereas 12 cases where treated non-operatively. In one case the treatment was by percutaneous embolization of the splenic artery. One reported case of a patient with a history of Crohn's disease had a combined splenic and liver laceration found on exploratory laparotomy. One patient died in spite of emergent splenectomy.
The reason for splenic rupture after a colonoscopic examination seems to be associated with the alteration of the peritoneal attachments that support the spleen in the left upper quadrant (ligaments gastrolienal, pancreaticolienal, phrenolienal).
Any kind of traction on these ligaments could determine a capsular rupture, which then disrupts a portion of the parenchyma densely adherent to it.
The rupture can be immediate because of damage to the hilum or entire organ disintegration, or delayed. In the first case the clinical presentation is with hypovolemic shock. In the latter the timing of manifestation of the shock varies and it could be gradual and delayed by few hours from the trauma depending upon degree of the parenchymal and capsular lesions. The mechanism is related to the formation of a subcapsular or intra-parenchymal hematoma, that then, instead of organizing itself into a pseudocyst, increases in size until determines the rupture of the capsule (delayed splenic rupture).
Although the exact reasons of development of a subcapsular or intra-parenchymal hematoma after a colonoscopy are not entirely clear, three mechanisms have been postulated:
The first one is related to the sudden trauma when the endoscope traverses the splenic flexure .
The second one is related to the avulsion of the splenic capsule caused by the excessive traction on the spleno-colic ligament during the endoscopic examination .
The third theory implies the traction on the adhesions between the spleen and the colon determined by previous surgical interventions or inflammatory processes .
In any case there are several risk factors that could predict the rupture of the spleen after colonoscopy: coagulopathies, infectious or hematological splenomegalies, specific pharmacological treatments (such as Hematopoyetic Growth Factors), intestinal or pancreatic inflammatory processes and previous intra-abdominal operations [7–9, 11–13].
Other authors blame this complication to the endoscopic manoeuvres utilized to navigate through the splenic flexure (in particular hooking and reduction) or to therapeutic interventions such as polypectomies and biopsies [10, 15].
In our case the colonoscopic exam was easily conducted and there was no therapeutic intervention except for a biopsy of small rectal polyps.
Differently from the cases published in the literature, in which the clinical manifestation appeared between 2 hours and 10 days [13, 29], our patient presented with vague and non specific chest pain soon after the endoscopic exam, but the clinical picture related to the splenic rupture was delayed by 13 days.
In most of the cases the patients report left upper quadrant abdominal pain which can be referred to the left shoulder (Kehr sign). The latter sign is not specific and could also be present after uncomplicated colonoscopies. Only one of the reported cases presented with complete lack of pain . In more typical cases, the physical finding can vary from tenderness localized at the epigastrium and left upper quadrant with reduction or absence of gastric tympanism, to diffuse peritoneal signs. Other common signs are those related to the hemodynamic alterations and vary from pallor, hypotension, tachycardia, dyspnea, to the more dramatic signs of shock.
In other cases the clinical presentation is more subtle. This usually occurs in cases of delayed rupture or in the presence of dense adhesions. The adhesions, in fact, not only can limit the extent of the hematoma, but can also prevent the onset of signs of peritoneal irritation from the hemoperitoneum.
In our case the dense adhesions, result of the previous extensive surgical history, determined the unusual and delayed clinical presentation. We can speculate, in fact, that following the trauma or stretching of the splenic ligaments (physiologic or post-operative), the dense adhesions determined a sort of hemostasis and delayed the splenic rupture with signs of hemodynamic instability only13 days after the event.
The atypical chest pain reported by the patient with the associated fever, leukocytosis and left pleural effusion was initially interpreted as a non specific pleuro-parenchymal lesion due to her previous history of bilateral breast cancer. Since the most common complication after colonoscopy are related to perforation and hemorrhage, the absence of intra-abdominal findings with negative abdominal plain x-rays, led to the exclusion of a complication related to the colonoscopy. It is then possible that a number of self-limited splenic ruptures goes undiagnosed because of lack of important symptomatology. On the other hand it is possible that other cases with an even longer lag of time between the endoscopic procedure and the splenic rupture were interpreted as a consequence of a more recent traumatic event and as such not published in the literature. It is then important to rule out a splenic rupture by U/S or CT scan in every case of abdominal or thoracic pain associated with anemia and lack of rectorrhagia [1, 2, 13, 24].
Since both the plain abdominal x-ray and the U/S give indirect signs of splenic rupture, the CT scan is the diagnostic exam of choice, especially for those patients candidate for non-operative treatment. In the literature 9 patients underwent successful conservative treatment with transfusions and close monitoring [12, 16, 19, 21, 29, 33, 46, 48] and only one required splenic artery embolization .