Single-Photon Emission Computed Tomography ...

Original Paper Received: March 26, 2010 Accepted: April 9, 2011 Published online: July 12, 2011

Digestion 2011;84:207–211 DOI: 10.1159/000328389

Single-Photon Emission Computed Tomography Enhanced Tc-99m-Pertechnetate Disodium-Labelled Red Blood Cell Scintigraphy in the Localization of Small Intestine Bleeding: A Single-Centre Twelve-Year Study Jiri Dolezal a Jaroslav Vizda a Marcela Kopacova b

Department of Nuclear Medicine and b Second Department of Internal Medicine, Charles University in Prague, Faculty of Medicine in Hradec Kralove and Charles University Teaching Hospital, Hradec Kralove, Czech Republic

Key Words Gastrointestinal bleeding ⴢ Scintigraphy ⴢ Tc-99m-labelled red blood cells

Abstract Aim: To present our experience with the detection of bleeding in the small intestine by means of scintigraphy with in vivo-labelled red blood cells (RBCs) in the period of 1998– 2009. Materials and Methods: A 12-year prospective study was accomplished with 40 patients (23 men, 17 women, aged 12–91, mean 56 years) who had lower gastrointestinal bleeding (obscure-overt bleeding) and underwent scintigraphy with in vivo-labelled RBCs by means of technetium 99m. The scintigraphy was usually performed after other diagnostic tests had failed to locate the bleeding. Results: A total of 26 patients had a positive scintigraphy with in vivo-labelled RBCs and 14 patients had negative scintigraphy. The final diagnosis was confirmed in 20 of 26 patients with a positive scintigraphy by push enteroscopy (6/20), intraoperative enteroscopy (7/20), surgery (4/20), duodenoscopy (1/20), double-balloon enteroscopy (1/20) and X-ray angiography (1/20). The correct location of the bleeding site was identified by RBC scintigraphy in 15 of 20 (75%) patients with the confirmed source. The locations of the bleeding site identified by scintigraphy and enteroscopy (push, intraoperative) and surgical investigations were highly correlated in patients

© 2011 S. Karger AG, Basel 0012–2823/11/0843–0207$38.00/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/dig

with a positive scintigraphy within the first 3 h. Eleven of the 20 correctly localized studies and none of the incorrectly localized studies were positive in the dynamic phase of imaging. In 5 patients (all erroneously localized), scintigraphy was positive only at a period longer than 18 h. Conclusion: RBC scintigraphy is an effective imaging modality in localizing lower gastrointestinal bleeding in patients for whom other diagnostic tests have failed to locate the bleeding. RBC scintigraphy can be successful in the detection of bleeding sites in the small intestine. Copyright © 2011 S. Karger AG, Basel

Introduction

We present our experience with the detection of bleeding sites in the small intestine by means of scintigraphy with in vivo-labelled red blood cells (RBCs) in the period of 1998–2009. Effective and prompt therapy for acute gastrointestinal (GI) bleeding depends on accurate localization of the site of haemorrhage. The history and clinical examination can often distinguish upper and lower GI bleeding. Upper GI tract and colon haemorrhage can be confirmed and localized using conventional endoscopy (gastroscopy, push enteroscopy and colonoscopy), which is the method of choice [1, 2]. Small intestinal bleeding is more problematic and conventional endoscopy has limJiri Dolezal, MD, PhD Department of Nuclear Medicine, University Hospital Sokolska 581 CZ–50005 Hradec Kralove (Czech Republic) Tel. +420 495 834 552, E-Mail dolezal @ fnhk.cz

Downloaded by: Faculty Hosp Hradec Kralove 195.113.123.86 - 12/15/2014 12:27:15 PM

a

Materials and Methods In vivo labelling of RBCs was used. Stanous chloride (10–20 ␮g/kg) from a commercial pyrophosphate kit was injected intravenously. Tin (2+) entered the RBC membrane and attached to haemoglobin to be ready to act as a reducing agent when 99mTcpertechnetate disodium got in the cell. The dose 11 MBq 99mTcpertechnetate disodium per kg of patient’s body weight in saline solution 0.9% was injected intravenously about 20 min later. 99mTc-pertechnetate crossed the RBC membrane and was reduced by stanous chloride and bound firmly to haemoglobin [7]. Dynamic scintigraphy of the abdomen was started immediately after administration of 99mTc-pertechnetate disodium intravenously. The acquisition protocol of 1-second frames for 2 min and 60-second frames for 10 min was used. Static 5-min images were done every 15 min during the first 3 h and every other hour as necessary. To improve sensitivity, specificity and three-dimensional location of the area of increased activity, abdomen SPECT (matrix 64 ! 64, 3°/20 s, zoom 1.0, filtered back reconstruction) was added in all patients. Rotating, double-head SPECT scintillation gamma cameras with infrared body contouring and a large field of view (Helix, VariCam from Elscint; Infinia from General

208

Digestion 2011;84:207–211

Electric Medical Systems) were used. The gamma cameras were fitted with high-resolution, low-energy collimators with parallel holes. Images were evaluated by the conventional processing systems, Xpert-Pro (Elscint) or Xeleris (General Electric Medical Systems). The patient’s radiation exposure expressed by the effective dose for the administered activity of 750 MBq 99mTc-RBCs was around 4.5 mSv. A total of 40 patients (23 men, 17 women, aged 12–91, mean 56 years) underwent RBC scintigraphy from 1998 to 2009. All patients had timely anamnesis of bleeding in the lower GI tract from an uncertain source (obscure-overt bleeding), melaena and/or haematochezia. The gastroscopy, colonoscopy, enteroclysis or alternatively X-ray angiography did not detect the source of bleeding in any patient. Video capsule or double-balloon enteroscopy may have localized most of the bleeding lesions, but the availability of these technologies in our institution has been limited to the last 6 years and our scintigraphy monitoring of bleeding patients was started in 1998. In cases of serious lower GI bleeding and posthemorrhagic anaemia, an exploratory laparotomy or laparoscopy was directly performed before the scintigraphy era.

Results

RBC scintigraphy was positive in 26 patients and negative in 14 patients. The final diagnosis was established in 20 patients with positive scintigraphy by push enteroscopy (6 patients), intraoperative enteroscopy (7), surgery (4), duodenoscopy (1), double-balloon enteroscopy (1) and X-ray angiography (1) with successful percutaneous transcatheter microembolization treatment. The GI bleeding stopped spontaneously in 6 patients with positive scintigraphy. There was only one episode of bleeding without repetition and there were no alarming symptoms that required further diagnostic procedures, such as push enteroscopy, intraoperative enteroscopy or surgery. The source of the bleeding, therefore, could not be confirmed. The final diagnoses of patients with the positive scintigraphy and the verified source of the bleeding can be seen in table 1. The correct location of the bleeding site was identified by RBC scintigraphy in 15 (75%) of 20 patients with a verified source (fig. 1, 2). The locations of the bleeding site identified upon scintigraphy and enteroscopy (push, intraoperative) as well as surgery investigations were highly correlated in patients with a positive scintigraphy within the first 3 h. Eleven of the 20 correctly localized studies and none of the incorrectly localized studies were positive in the dynamic (continuous) phase of imaging. The accurate spatial localization of the sites of 99mTc-RBC extravasation was allowed by SPECT. The incorrect location of bleeding sites was assessed in 5 (25%) patients with positive RBC scintigraphy and the Dolezal /Vizda /Kopacova


ited value in the small intestine. Innovative but costly small intestine endoscopy methods, such as capsule enteroscopy and double-balloon enteroscopy, are not readily available in all hospitals. Scintigraphy with RBCs labelled by means of 99mTc can help to detect the source of GI bleeding in the small intestine and improves disease management. Scintigraphy can identify the site of bleeding at the rate of 0.1 ml/min or more [3]. Only 2–3 ml of extravasated blood is necessary for detection. This compares favourably with the ability of contrast angiography to detect bleeding rates of about 1 ml/min or greater [4]. Extravasated radiolabelled RBCs within the small intestine are identified as an area of activity that increases in intensity with time, and/or as a focus of activity that moves in a pattern corresponding to the lumen of the small intestine. The small intestinal bleeding usually can be distinguished from colonic bleeding by its rapid serpiginous movement [5]. Steady scintigraphic activity should not be diagnosed as an active bleeding site and usually results from a fixed vascular structure (such as haemangioma, accessory spleen or ectopic kidney) [4]. Single-photon emission CT (SPECT) of the abdomen allows true three-dimensional image acquisition and display. Reconstruction of cross-sectional slices has used filtered back or iterative projection, the same methodology used for CT [5]. New hybrid system SPECT/CT for facilitating imaging interpretation and improving the accuracy of 99mTc-RBC scintigraphy in patients with acute lower GI bleeding can be used [6].

Table 1. Final diagnosis of patients (n = 20) with positive scintigraphy

Cases Diagnosis and site of bleeding

Confirmation

Correct scintigraphy detection of the site of bleeding

6

bleeding from a small intestine arteriovenous malformation

push enteroscopy (4) intraoperative enteroscopy (1) surgery (1)

yes (5) no (1)

2

uraemic enteritis with bleeding erosions in the ileum and jejunum

intraoperative enteroscopy (2)

yes (1) no (1)

1

hereditary haemorrhagic teleangiectasia (Osler-Rendu-Weber disease) in the jejunum

intraoperative enteroscopy

yes

1

pseudocyst of the pancreas with the bleeding vessel communicating to the transverse colon

surgery

no

1

bleeding submucose varix in the jejunum


yes

1

bleeding inflammatory polyp at the ileotransversoanastomosis

surgery

yes

1

bleeding carcinoid of the ileum


no

1

bleeding from the ileosigmoideoanastomosis 6 days after hemicolectomy for Crohn’s disease

surgery

yes

1

bleeding from an ulcer close to the papilla of Vater

endoscopy

no

1

bleeding from an ulcer at the jejunum after previous NSAID treatment push enteroscopy

yes

1

haemangiomatosis of the small intestine


yes

1

bleeding from an ulcer at the jejunum after previous NSAID and Warfarin treatment

push enteroscopy

yes

1

two bleeding ulcers at the distal ileum affected by systemic vasculitis

double-balloon enteroscopy

yes

1

bleeding pseudoaneurysm in the hepatojejunonastomosis 14 days after partial liver resection

therapeutic X-ray angiography with percutaneous transcatheter microembolization treatment

yes

confirmed source of bleeding. In these 5 patients (erroneously localized), scintigraphy was positive only at a period longer than 18 h. Of these 5 patients, the first 3 had the confirmed source of bleeding in the small intestine (AV malformation in distal jejunum, uraemic enteritis with bleeding erosions in ileum and jejunum and bleeding carcinoid tumour of the ileum). The RBC scintigraphy was positive only at a period longer than 18 h. The pool of radioactive RBCs was detected in the colon as late as on the last scintigrams. There was intermittent bleeding in the small intestine during long time intervals between scintigrams and the pool of radioactive RBCs in the small intestinal lumen was moved by the accelerated intestinal peristalsis to the colon. The source of bleeding was incorrectly established in the colon. The possibility of an

incorrect bleeding source was mentioned in the scintigraphy description. The next patient had a pancreatic pseudocyst with an intermittently bleeding vessel and communication with the colon transversum. The pool of radioactive RBCs was detected as late as on the last scintigrams in the colon sigmoideum. The last patient had intermittent bleeding from a peptic ulcer close to the papilla of Vater. The pool of radioactive RBCs was revealed as late as on the last scintigrams in the distal part of the small intestine. During the long time intervals between scintigrams, the pool of radioactive RBCs in the lumen was moved by the accelerated intestinal peristalsis to the distal parts of the small intestine. The GI bleeding stopped spontaneously in 14 patients with negative scintigraphy. There was only one episode of

Localization of Small Intestine Bleeding Using SPECT

Digestion 2011;84:207–211

209


Figures in parentheses are numbers of patients unless otherwise indicated.

1 2

Fig. 1. Eighty-one-year-old woman with

enterorrhagia. 99mTc-RBC scintigraphy. The bleeding in the site of A-V malformation in the proximal jejunum. Fig. 2. Fifty-three-year-old woman with enterorrhagia. 99mTc-RBC scintigraphy. The bleeding in the site of hereditary haemorrhagic teleangiectasia (Osler-Rendu-Weber disease) in the distal jejunum.

Discussion

Endoscopy is the method of choice to detect the source of bleeding in the upper GI tract and the colon. RBC scintigraphy is only a complementary imaging method and can be successful for the detection of the bleeding site in the small intestine. In our study, 19 of 20 patients with positive RBC scintigraphy and at a confirmed source had the site of bleeding in the small intestine. The last patient had a pancreatic pseudocyst with an intermittently bleeding vessel and communication with the colon transversum. Video capsule endoscopy or double-balloon enteroscopy may have localized most of these bleeding lesions, but the availability of these technologies in our institution has been limited to the past 6 years and RBC scintigraphy of GI bleeding patients was started in 1998. After starting video capsule endoscopy and double-balloon enteroscopy in our endoscopy unit the number of RBC scintigraphies decreased substantially. Push enteroscopy was not performed before proceeding to RBC scintigraphy because scintigraphy was a non-invasive method displaying the whole small intestine. Two of our patients had bleeding at the ileocolonic anastomosis site (bleeding inflammatory polyp at ileotransversoanastomosis, bleeding from the ileosigmoideoanastomosis 6 days after hemicolectomy for Crohn’s disease). This bleeding was not active at the time of endoscopy. Our findings on RBC scintigraphy in pa210

Digestion 2011;84:207–211

tients for obscure GI bleeding were similar as other investigations in the current literature. Chen et al. [8] evaluated the diagnostic value of the double-balloon enteroscopy in 75 patients for obscure GI bleeding. Of the 75 patients, 44 (58.7%) had positive double-balloon enteroscopy findings and 22 patients had negative findings but had potential bleeding at surgery and capsule endoscopy. These 66 patients were finally diagnosed as obscure GI bleeding which was most commonly caused by small bowel tumours (28.0%), angiodysplasia (18.7%) and Crohn’s disease (10.7%). Esaki et al. [9] analyzed 68 patients for obscure GI bleeding who underwent capsule endoscopy. Capsule endoscopy finding was positive in 36 (53%) patients. Das et al. [10] evaluated the diagnostic value of double-balloon and/or capsule endoscopy in 53 patients for obscure GI bleeding. Double-balloon and/or capsule endoscopy localized the source of bleeding in 43 patients (yield 81%). Angiodysplasias, tumours, Crohn’s disease, intestinal tuberculosis and hookworm infestation were predominant aetiologies. Katsinelos et al. [11] assessed the diagnostic yield of capsule endoscopy in 63 patients with obscure GI bleeding and normal upper and lower endoscopy. The overall diagnostic yield was 44.44% of patients and included findings of angiectasia in 11 patients, non-steroidal anti-inflammatory drugs enteropathy in 6 patients, celiac disease in 3 patients, tumours in 2 patients and a variety of other diagnoses ranging from varices to ulcers. Lee et al. [12] performed 23 RBC scintigraphies in 22 paediatric patients for massive GI bleeding. The scintigraphy was usually performed after other diagnostic tests had failed to locate the bleeding. The scintigraphy was positive in 9 of 23 (39%) patients. The test demonstrated a positive scintigraphy within the first 2 h in 6 patients, and 3 paDolezal /Vizda /Kopacova


bleeding without repetition and there were no alarming symptoms that required further diagnostic procedures, such as push enteroscopy, intraoperative enteroscopy or surgery.

tients had positive results at 24 h. The locations of the lesions were identified upon scintigraphy and surgery was highly correlated in patients with a positive scintigraphy within the first 2 h. Diehl et al. [13] evaluated the clinical use of RBC scintigraphy in patients with acute lower GI bleeding as well as negative endoscopy and multislice CT (MSCT). In 31 patients with acute lower GI bleeding in whom the endoscopy findings were negative or the procedure was not feasible, dual-phase MSCT of the abdomen was performed. The MSCT was followed by RBC scintigraphy for patients in whom no active bleeding was visible by MSCT. Images were acquired within 24 h after administration of the tracer, depending on clinical symptoms. The results of the imaging modalities were correlated with the clinical and surgical treatment. In 20 of 31 patients, MSCT showed no active bleeding and RBC scintigraphy was performed. In 8 of these 20 patients, RBC scintigraphy was also negative. Of these 8 patients, 5 were stable and did not require further diagnostic procedures; in the other 3 bleeding persisted and these patients required surgical treatment. In 12 of 20 patients, active bleeding was demonstrated using RBC scintigraphy. Eight of the 12 patients with positive RBC scintigraphy findings underwent surgery where the site of bleeding was confirmed. To improve sensitivity, specificity and location of the area of increased activity, abdomen SPECT was used in our institution. The added cost of abdomen SPECT in the Czech Republic was around EUR 80 and extended the procedure for about 15 min. Because the RBC scintigra-

phy was performed on double-head SPECT gamma camera no added manpower was used. Because first SPECT/ CT multifunctional imaging scanner was started in our institution last year we presented 12-year experience with RBC scintigraphy and abdomen SPECT. Similar experiences are found in the literature. Schillaci et al. [6] presented 27 patients with acute lower GI bleeding who were studied with dynamic and planar 99mTc-RBC scintigraphy. In 19 patients with a positive scan, an abdomen SPECT/CT was performed. In 7 patients, SPECT/CT had significant impact on the scintigraphic results, and in 6 patients it precisely localized the bleeding foci.

Conclusion

RBC scintigraphy is an effective imaging modality in the localization of lower GI bleeding for patients in whom other diagnostic tests have failed to locate the bleeding or are not available. RBC scintigraphy can be successful for detecting the bleeding site in the small intestine. The location of the bleeding site identified upon scintigraphy and enteroscopy (push, intraoperative) and surgery were highly correlated in patients with a positive scintigraphy within the first 3 h. We suggest that positive bleeding scans only be reported in the first 3 h of scanning given the high false-positive rate at later hours. To improve sensitivity, specificity and location of the area of increased activity the abdomen SPECT or SPECT/CT are recommended.

References

Localization of Small Intestine Bleeding Using SPECT

7 8

9

10

in the study of lower gastrointestinal bleeding with technetium-99m red blood cells. Q J Nucl Med Mol Imaging 2009;53:281–289. Krishnamurthy GT, Krishnamurthy S: Nuclear Hepatology. Berlin, Springer-Verlag, 2000, pp 52–53. Chen LH, Chen WG, Cao Hj, Zhang H, Shan GD, Li L, Zhang BL, Xu CF, Ding KL, Fang Y, Cheng Y, Wu CJ, Xu GQ: Double-ballon enteroscopy for obscure gastrointestinal bleeding: a single centre experience in China. World J Gastroenterol 2010; 16: 1655– 1659. Esaki M, Matsumoto T, Yada S, Yanaru-Fujisawa R, Kudo T, Yanai S, Nakamura S, Iida M: Factors associated with the clinical impact of capsule endoscopy in patients with overt obscure gastrointestinal bleeding. Dig Dis Sci 2010;55:2294–2301. Das K, Sarkar R, Dasgupta J, Ray S, Ghatak S, Das K, Mridha AR, Dhali GK, Chowdhury A: Obscure GI bleeding in the tropics: im-

pact of introduction of double-balloon and capsule endoscopies on outcome. Gastrointest Endosc 2010; 72: 292–300. 11 Katsinelos P, Chatzimavroudis G, Terzoudis S, Patsis I, Fasoulas K, Katsinelos T, Kokonis G, Zavos C, Vasiliadis T, Kountouras J: Diagnostic yield and clinical impact of capsule endoscopy in obscure gastrointestinal bleeding during routine clinical practice: a singlecentre experience. Med Princ Pract 2011; 20: 60–65. 12 Lee J, Lai MW, Chen CC, Chen SY, Chao HC, Chan SC, Kong MS: Red blood cell scintigraphy in children with acute massive gastrointestinal bleeding. Pediatr Int 2008; 50: 199– 203. 13 Diehl SJ, Ko HK, Dominquez E, Kaara Tesdal I, Kähler G, Böhm C, Düber G: Negative endoscopy and MSCT findings in patients with acute lower gastrointestinal haemorrhage. Value of 99mTc erythrocyte scintigraphy. Radiologe 2007;47:64–70.

Digestion 2011;84:207–211

211


1 Loffelf RJLF, van der Putten ABMM: The completion rate of colonoscopy in normal daily practice: factors associated with failure. Digestion 2009;80:267–270. 2 Ziegler M, Schubring-Giese B, Bühner M, Kolligs FT: Attitude to secondary prevention and concerns about colonoscopy are independent predictors of acceptance of screening colonoscopy. Digestion 2010;81:120–126. 3 Ryan P, Styles CB, Chmiel R: Identification of the site of severe colon bleeding by technetium-labelled red-cell scan. Dis Colon Rectum 1992;35:219–22. 4 Thrall JH, Ziessman HA: Nuclear Medicine – The Requisites, ed 2. St. Louis, Mosby, 2001, pp 280–287. 5 Ziessman HA, O’Malley JP, Thrall JH: Nuclear Medicine. The Requisites in Radiology. Philadelphia, Elsevier Mosby, 2006, pp 53–63. 6 Schillaci O, Spanu A, Tagliabue L, Filippi L, Danieli R, Palumbo B, Del Sole A, Madeddu G: SPECT/CT with a hybrid imaging system