Scandinavian Journal of Gastroenterology. 2014; 49: 238–245
ORIGINAL ARTICLE
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
Efficacy of argon plasma coagulation in the treatment of radiation-induced hemorrhagic gastroduodenal vascular ectasia
HEE-WON KWAK1, WOO JIN LEE1, SANG MYUNG WOO1, BO HYUN KIM1, JOONG-WON PARK1, CHANG-MIN KIM1, TAE HYUN KIM1, SUNG-SIK HAN1, SEONG HOON KIM1, SANG-JAE PARK1 & MYEONG CHERL KOOK2 1
Center for Liver Cancer, National Cancer Center, Goyang, Korea, and 2Center for Gastric Cancer, National Cancer Center, Goyang, Korea
Abstract Objective. Radiation-induced hemorrhagic gastroduodenal vascular ectasia (GDVE) is rare but difficult to manage. Argon plasma coagulation (APC) has not yet been evaluated in the treatment of radiation-induced hemorrhagic GDVE. The efficacy of APC in patients with radiation-induced hemorrhagic GDVE has been investigated in this article. Material and methods. Eighteen patients with upper gastrointestinal (GI) bleeding caused by radiation-induced GDVE, including 13 with hepatocellular carcinoma, 3 with pancreatic cancer, and 2 with cholangiocarcinoma, were treated with APC. The efficacy of APC was retrospectively evaluated, based on cessation of macroscopic GI bleeding, resolution or stabilization of anemia and transfusion dependence, endoscopic ablation of almost all vascular lesions, complications, and recurrence. Results. Mean patient age was 59 years (range 42–80 years). The median time from radiation to GDVE diagnosis was 4.6 months (range 3.3–21.5 months). The median number of APC sessions per patient was 2.4 (range 1–4). All 18 patients showed an endoscopic response to APC treatment, with sustained increases in mean hemoglobin level, from 6.6 g/dL (range 2.9–9.5g/dL) to 9.7 g/dL (range 7.1–12.7 g/dL) (p < 0.001), and decreased dependence on transfusion, from 9.1 (range 0–30) to 4.1 (range 0–15) units of packed red blood cells per patient (p = 0.038) after last endoscopic eradication by APC treatment. There were no major procedure-related adverse events or deaths. At a median follow up of 4.7 months (range 0.6–24.5 months), none of the patients experienced recurrence of GDVE. Conclusions. APC showed short-term effectiveness and safety in the treatment of radiation-induced hemorrhagic GDVE.
Key Words: argon plasma coagulation, gastroduodenal vascular ectasia, radiation-induced
Introduction Radiation-induced injury to the gastrointestinal (GI) tract is not rare. For example, chronic radiation proctosigmoiditis has been reported in 5–20% of patients who received radiation treatment for pelvic malignancies, such as cervical and prostate cancers, or local radiation therapy for rectal and anal cancers [1]. Radiation-induced hemorrhagic gastroduodenal vascular ectasia (GDVE) is rare because the stomach is seldom within the radiation treatment field. Radiation- induced hemorrhagic GDVE is a serious adverse
event in upper GI radiation treatment which is difficult to manage. Unlike radiation proctopathy, GI bleeding from chronically irradiated GDVE is rare, with only a small number of cases reported to date. Radiation-induced hemorrhagic GDVE has been treated with hyperbaric oxygen therapy [2,3], aminocaproic acid [4], cryotherapy [5], band ligation [6], argon plasma coagulation (APC), and resection [7]. APC treatment has been found to be effective and safe in the management of radiation-induced rectal bleeding [8] and hemorrhagic vascular ectasia [9,10] because such lesions show diffuse and superficial
Correspondence: Woo Jin Lee, MD, Center for Liver Cancer, National Cancer Center, 323 Ilsan-ro, Ilsan dong-gu, Goyang, Gyeonggi 410-769, South Korea. Tel: +82 31 920 1612. Fax: +82 31 920 1138. E-mail: [email protected]
(Received 16 August 2013; revised 4 November 2013; accepted 10 November 2013) ISSN 0036-5521 print/ISSN 1502-7708 online 2014 Informa Healthcare DOI: 10.3109/00365521.2013.865783
Efficacy of APC in the treatment of GDVE involvement confined to the mucosal and submucosal layers. To date, only a few patients with radiationinduced gastritis have been treated with APC [11–17]. Moreover, APC treatment has not yet been evaluated in patients with diffuse radiation-induced hemorrhagic vascular lesions located in the upper GI tract. We, therefore, evaluated the efficacy and safety of APC in patients with radiation-induced hemorrhagic GDVE.
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
Methods Eighteen patients who had undergone threedimensional conformal external beam radiation treatment for hepatobiliary or pancreatic cancer between September 2003 and September 2012 (Table I), and who were subsequently diagnosed endoscopically with radiation-induced hemorrhagic GDVE, were evaluated retrospectively. All patients received transfusions due to overt (melena, hematochezia, or hematemesis) or occult (iron-deficiency anemia) GI bleeding. Hemorrhagic GDVE was diagnosed by endoscopic visualization of oozing, blood clots arising from the coalescence of numerous vascular ectasias, or longitudinal red stripes radiating from the antrum and/or duodenum. Patients with radiation-induced GDVE had a typical glove and stocking distribution pattern along the radiation field, making duodenal involvement relatively common. Other sources of active bleeding were excluded. Bleeding from the varix, portal-hypertensive gastropathy (PHG), and ulcer were excluded by endoscopic findings. Colonoscopy to exclude possible lower GI bleeding was performed in three patients with blood clots only, with all three showing improvements after APC treatment. Radiation-induced hemorrhagic GDVE was diagnosed in each patient by two independent expert endoscopists, with an expert radiation oncologist confirming that the mucosal areas affected were compatible with the radiation field. The demographical and clinical characteristics of the study population were reviewed, as were findings at initial presentation, underlying diseases, total radiation dose due to underlying disease, time between radiation treatment and radiation-induced hemorrhagic GDVE, endoscopic findings, and GDVE pattern (watermelon or honeycomb) at initial diagnosis. The institutional review board of our institution approved this study, and each patient provided written informed consent. APC treatment was administered using an ERBE VIO300D system/APC2 (digital) (ERBE Elektromedizin, Tübingen, Germany) and probes including side-wide beam, side-conical beam, straight-axial beam, and 360 round beam probes (ERBE). APC
239
was usually administered at 1–1.2 L/min with a coagulation power setting of 30–40 W (pulsed mode effect 2). Procedures were performed with the patient under conscious sedation, with midazolam used in the six patients without liver cirrhosis but not in the 12 patients who had liver cirrhosis due to the high risk of hepatic encephalopathy. All 18 patients received 20 mg of hyoscine butylbromide to reduce gastric peristalsis. Each treatment session began from the bulb or pylorus and moved proximally, treating all visible lesions. The duration of each treatment session depended on each patient’s tolerance of the procedure but was not more than 30 min. APC treatment was repeated every 1–2 weeks until all treatment end points were reached. Treatment end points were defined as resolution or stabilization of anemia and transfusion dependence, cessation of macroscopic GI bleeding, and endoscopic ablation of >90% of the vascular lesions [9,18]. Once these treatment end points had been reached, there was no endoscopic follow up unless the patient developed recurrent symptoms or signs of GI blood loss. Following APC treatment, all patients were treated with an oral proton pump inhibitor for 2–4 weeks. Mean number of units of blood transfused before and after APC were calculated [19]. Pretreatment hemoglobin level was defined as the concentration on admission for APC treatment and before any transfusions. Post-treatment hemoglobin level was defined as the concentration 3 months after completion of endoscopic therapy or at last follow-up visit [20]. Changes in mean hemoglobin level and packed RBC transfusion volume were analyzed using the Wilcoxon signed-rank test. All statistical analyses were performed using STATA version 10.0 (Stata Corp, College Station, TX, USA), and values of p < 0.05 was considered statistically significant. Results The 18 patients comprised 12 males and 6 females, of mean age 59 years (range 42–80 years). Thirteen patients (72%) presented initially with overt bleeding, whereas the other five patients complained of dizziness and showed iron-deficiency anemia due to occult bleeding. Underlying diseases included hepatocellular carcinoma (HCC) in 13 patients (72%), pancreatic cancer in 3 patients (16%), and cholangiocarcinoma in 2 patients (11%). All patients had unresectable disease. All HCC patients had portal vein tumor thrombosis. All pancreatic cancer and cholangiocarcinoma patients had locally advanced disease, for which they received radiation treatment in
Melena
Melena
Melena
Melena
Melena
IDA
Melena
IDA
IDA
Hematemesis
IDA
Melena
Melena
Melena
Melena
Melena
IDA
Melena
1. F, 46
2. M, 54
3. M, 44
4. F, 61
5. F, 65
6. F, 72
7. M, 54
8. M.64
9. F, 42
10. F, 61
11. M, 44
12. M, 79
13. M, 80
14. M, 60
15. M, 60
16. M, 44
17. M, 58
18. M, 74
HCC
HCC
HCC
Cholangiocarcinoma
Cholangiocarcinoma
HCC
Pancreatic cancer HCC
HCC
HCC
Pancreatic cancer Pancreatic cancer HCC
HCC
HCC
HCC
HCC
HCC
Under lying disease
5000
5250
4400
5500
5040
5400
5500
3600
5400
5000
5400
5040
5040
3600
5100
5400
5500
5400
Total radiation dose (cGy)
4.8
3.3
3.5
6.7
5.0
6.2
21.5
8.2
4.1
4.8
4.1
5.7
9.4
3.7
4.2
4.3
4.2
3.5
Interval between radiation and diagnosis (months)
Oozing
Oozing
Oozing
Oozing
Oozing
Oozing
Oozing
Oozing
Blood clots Oozing
Oozing
Blood clots Blood clots Oozing
Oozing
Oozing
Oozing
Oozing
Endoscopic finding
Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb
Water-melon
GDVE pattern
1
4
2
3
1
1
4
1
4
2
2
3
2
1
4
3
3
4
APC session
4
8
8
9
23
6
18
19
19
0
9
28
1
10
37
16
24
7
Total transfusion volume (packed RBC unit)
9.1
4.7
6.8
5.2
4.9
7.5
6.8
9.3
8.4
9.5
6.8
3.5
5.5
5.1
5.2
8.7
2.9
8.5
Pre-APC
12.1
9.5
7.9
10.7
11
8.4
7.8
10.6
10.4
11.7
8.5
7.1
12.7
8.7
7.2
10.0
9.7
11.1
Post-APC
Hb level (g/dL)
12.8 (Dead) 10.8 (Dead) 5.4 (Dead) 9.5 (Dead) 1.5 (Dead) 18.7 (Dead) 1.2 (Dead) 0.6 (Dead) 3.3 (Dead) 16.0 (Dead) 0.6 (Dead) 4.3 (Dead) 3.7 (Dead) 24.5 (Alive) 4.2 (Alive) 3.7 (Dead) 6.5 (Dead) 5.0 (Dead)
Follow-up (months)
Abbreviations: APC = Argon plasma coagulation; cGy = Centigray; F = Female; GDVE = Gastro-duodenal vascular ectasia; HCC = Hepatocellular carcinoma; IDA = Iron-deficiency anemia; M = Male; No = Number; RBC = Red blood cell; Hb = Hemoglobin.
Mode of presentation
Patient No. Sex, Age (years)
Table I. Clinical characteristics of GDVE.
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
240 H.-W. Kwak et al.
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
Efficacy of APC in the treatment of GDVE A
B
C
D
241
Figure 1. Endoscopic findings in a patient with radiation-induced hemorrhagic GDVE (patient 1): (A) before the first session of APC; (B) immediately after the first session of APC; (C) after the second session of APC; and (D) 3 months after the end of APC are shown.
combination with chemotherapy. Twelve patients (67%) had liver cirrhosis, including two, nine, and one with Child-Pugh classes A, B, and C, respectively. Five patients (28%) had esophageal varices and three patients (17%) had PHG. Mean total radiation dose due to underlying disease was 5032 cGy (range 3600– 5500 cGy) and median time between radiation treatment and radiation-induced hemorrhagic GDVE was 4.6 months (range 3.3–21.5 months). Endoscopically, 15 patients (83%) showed active bleeding with an oozing pattern (Forrest classification Ib), whereas the other 3 patients (17%) had blood clots only (Forrest classification IIb). None had spurting bleeding (Forrest classification Ia) or visible vessels (Forrest classification IIa). Seventeen patients (94%) had a honeycomb pattern of GVDE, and one (6%) had a watermelon pattern. The mean number of APC sessions per patient was 2.6 (range 1–4). At last endoscopic examination, all patients showed cessation of macroscopic GI bleeding and the ablation of almost all vascular lesions (Figures 1,2,3). Mean hemoglobin concentration was significantly higher after (9.7 g/dL; range = 7.1–12.7 g/dL) APC treatment (p < 0.001) compared to before treatment (6.6 g/dL; range = 2.9–9.5 g/dL), whereas mean
packed RBC transfusion volume was significantly lower after treatment (4.1 vs. 9.1 units, p = 0.038). Median follow-up period after last APC treatment was 4.7 months (range 0.6–24.5 months) (Table I). No patient experienced any major APC treatmentrelated adverse event or death. None of the patients experienced a recurrence of GI bleeding, and none required transfusions >3 months after the end of treatment. Discussion The stomach has a thick muscular layer and mucosal lining and is relatively resistant to radiation. However, at higher radiation doses (usually >5000 rad), ulceration may lead to perforation [21,22]. Acute vasculopathy may progress to a prolonged and progressive obliterative endarteritis, vasculitis, and endothelial proliferation, leading to mucosal ischemia, ulceration, and mucosal telangiectasias, and ultimately to fibrosis [23]. Mucosae that sustain late radiation injury appear friable and granular with multiple telangiectasias. Similarly, pelvic irradiation of the rectum, intestine, and bladder results in a similar pathology, appearing as proctitis, colitis, and intravesical hemorrhage,
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
242
H.-W. Kwak et al. A
B
C
D
Figure 2. Endoscopic findings in a patient with radiation-induced hemorrhagic GDVE (patient 8): (A) before the first session of APC; (B) the first session of APC; (C) 1 day after the first session of APC; and (D) 1 week after the end of APC are shown.
respectively. A study of biopsy samples obtained from five patients showed pathological features typical of vascular ectasia, including hypertrophy of the antral mucosa, dilation of mucosal capillaries with focal thrombosis, and fibromuscular hyperplasia of the lamina propria (Figure 4) [24,25]. However, the absence of these characteristics on biopsy would not preclude a diagnosis of radiation-induced GDVE because biopsies are typically insufficiently deep or the lesions are focal and, therefore, easily missed on biopsy. In addition, biopsy can worsen bleeding. Biopsy is not always required to confirm the diagnosis, and treatment can be started based on typical appearance at initial endoscopy. Radiation-induced GDVE is a diffuse process with multiple bleeding sites leading to significant blood loss and requiring multiple blood transfusions, prolonged hospitalizations, and repeated endoscopies with unsatisfactory control of bleeding. Radiationinduced GDVE may be a subset of gastric vascular ectasia, although the main lesions of the latter are located on the antrum (i.e. gastric antral vascular
ectasia [GAVE] syndrome) [26]. Patients with radiation-induced GDVE have a history of radiation treatment and a glove and stocking distribution pattern along the radiation field, making duodenal bulb involvement relatively common, as observed in our patients. Endoscopically, the lesions take two forms: striped (so-called watermelon stomach) and diffuse, honeycomb, punctate patterns. Liver cirrhosis has been reported to be strongly associated with the development of punctate-type vascular ectasias [27]. Only 1 of our 18 patients showed a striped pattern. The differential diagnosis of radiation-induced GDVE and PHG is essential for correct treatment, since GDVE lesions are completely distinct from PHG [28]. Treatment of portal hypertension with b-blockers or transjugular portosystemic shunting leads to an improvement in PHG patients but has no effect in GDVE patients. Both types of lesions may appear as multiple red spots, but lesions in GDVE patients are located mainly in the antrum, whereas lesions in PHG patients, including those with portal hypertensive duodenopathy and/or jejunopathy,
Efficacy of APC in the treatment of GDVE
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
A
243
B
C
Figure 3. Endoscopic findings in a patient with radiation-induced hemorrhagic GDVE (patient 12): (A) before the first session of APC; (B) the first session of APC; and (C) 2 weeks after the end of APC are shown.
usually show a diffuse mosaic pattern or snake skin appearance in the corpus and fundus of the stomach. Patients with radiation-induced GDVE usually have scattered punctate patterns of vascular ectasias and a A
typical glove and stocking distribution pattern along the radiation field. Biopsy may be useful in distinguishing between these two entities. Of our 13 patients with hepatocellular cancer, 3 had PHG; in these B
Figure 4. Pathologic findings in a patient with radiation-induced hemorrhagic GDVE: (A) dilated vessel and thrombosis and (B) ischemic damage in lamina propria are shown.
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
244
H.-W. Kwak et al.
patients, the bleeding focus was far from the PHG lesions, and active bleeding no longer occurred after APC treatment of GDVE, despite the presence of PHG lesions. APC is an electrosurgical endoscopic procedure indicated for hemostasis and devitalization of tissues [29]. APC is an effective treatment for radiation proctitis and gastric vascular ectasia, reducing bleeding and requirements for iron or blood transfusion by cauterizing mucosal telangiectasias [8–10]. There have been few comparisons of APC with other established treatment modalities, such as the heater probe, bipolar coagulation probe, and Nd:YAG laser [30,31]. The advantages of APC include a noncontact mode; axial, radial, and retrograde application; controllable depth of coagulation; and low cost of purchase and maintenance of the instrumentation. APC is associated with a lower risk of perforation than other modalities because it coagulates tissue without contact. Endoscopic ultrasound after APC showed the disappearance of the small hypoechoic spaces within the second and third sonographic layers and no evidence of injury to the fourth sonographic layer, the muscularis propria [32]. In the patients described here, procedures were performed under conscious sedation, but in the absence of midazolam in the 12 patients with liver cirrhosis due to the high risk of hepatic encephalopathy. Propofol sedation may be more comfortable, but APC treatment was not difficult and lasted
ORIGINAL ARTICLE
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
Efficacy of argon plasma coagulation in the treatment of radiation-induced hemorrhagic gastroduodenal vascular ectasia
HEE-WON KWAK1, WOO JIN LEE1, SANG MYUNG WOO1, BO HYUN KIM1, JOONG-WON PARK1, CHANG-MIN KIM1, TAE HYUN KIM1, SUNG-SIK HAN1, SEONG HOON KIM1, SANG-JAE PARK1 & MYEONG CHERL KOOK2 1
Center for Liver Cancer, National Cancer Center, Goyang, Korea, and 2Center for Gastric Cancer, National Cancer Center, Goyang, Korea
Abstract Objective. Radiation-induced hemorrhagic gastroduodenal vascular ectasia (GDVE) is rare but difficult to manage. Argon plasma coagulation (APC) has not yet been evaluated in the treatment of radiation-induced hemorrhagic GDVE. The efficacy of APC in patients with radiation-induced hemorrhagic GDVE has been investigated in this article. Material and methods. Eighteen patients with upper gastrointestinal (GI) bleeding caused by radiation-induced GDVE, including 13 with hepatocellular carcinoma, 3 with pancreatic cancer, and 2 with cholangiocarcinoma, were treated with APC. The efficacy of APC was retrospectively evaluated, based on cessation of macroscopic GI bleeding, resolution or stabilization of anemia and transfusion dependence, endoscopic ablation of almost all vascular lesions, complications, and recurrence. Results. Mean patient age was 59 years (range 42–80 years). The median time from radiation to GDVE diagnosis was 4.6 months (range 3.3–21.5 months). The median number of APC sessions per patient was 2.4 (range 1–4). All 18 patients showed an endoscopic response to APC treatment, with sustained increases in mean hemoglobin level, from 6.6 g/dL (range 2.9–9.5g/dL) to 9.7 g/dL (range 7.1–12.7 g/dL) (p < 0.001), and decreased dependence on transfusion, from 9.1 (range 0–30) to 4.1 (range 0–15) units of packed red blood cells per patient (p = 0.038) after last endoscopic eradication by APC treatment. There were no major procedure-related adverse events or deaths. At a median follow up of 4.7 months (range 0.6–24.5 months), none of the patients experienced recurrence of GDVE. Conclusions. APC showed short-term effectiveness and safety in the treatment of radiation-induced hemorrhagic GDVE.
Key Words: argon plasma coagulation, gastroduodenal vascular ectasia, radiation-induced
Introduction Radiation-induced injury to the gastrointestinal (GI) tract is not rare. For example, chronic radiation proctosigmoiditis has been reported in 5–20% of patients who received radiation treatment for pelvic malignancies, such as cervical and prostate cancers, or local radiation therapy for rectal and anal cancers [1]. Radiation-induced hemorrhagic gastroduodenal vascular ectasia (GDVE) is rare because the stomach is seldom within the radiation treatment field. Radiation- induced hemorrhagic GDVE is a serious adverse
event in upper GI radiation treatment which is difficult to manage. Unlike radiation proctopathy, GI bleeding from chronically irradiated GDVE is rare, with only a small number of cases reported to date. Radiation-induced hemorrhagic GDVE has been treated with hyperbaric oxygen therapy [2,3], aminocaproic acid [4], cryotherapy [5], band ligation [6], argon plasma coagulation (APC), and resection [7]. APC treatment has been found to be effective and safe in the management of radiation-induced rectal bleeding [8] and hemorrhagic vascular ectasia [9,10] because such lesions show diffuse and superficial
Correspondence: Woo Jin Lee, MD, Center for Liver Cancer, National Cancer Center, 323 Ilsan-ro, Ilsan dong-gu, Goyang, Gyeonggi 410-769, South Korea. Tel: +82 31 920 1612. Fax: +82 31 920 1138. E-mail: [email protected]
(Received 16 August 2013; revised 4 November 2013; accepted 10 November 2013) ISSN 0036-5521 print/ISSN 1502-7708 online 2014 Informa Healthcare DOI: 10.3109/00365521.2013.865783
Efficacy of APC in the treatment of GDVE involvement confined to the mucosal and submucosal layers. To date, only a few patients with radiationinduced gastritis have been treated with APC [11–17]. Moreover, APC treatment has not yet been evaluated in patients with diffuse radiation-induced hemorrhagic vascular lesions located in the upper GI tract. We, therefore, evaluated the efficacy and safety of APC in patients with radiation-induced hemorrhagic GDVE.
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
Methods Eighteen patients who had undergone threedimensional conformal external beam radiation treatment for hepatobiliary or pancreatic cancer between September 2003 and September 2012 (Table I), and who were subsequently diagnosed endoscopically with radiation-induced hemorrhagic GDVE, were evaluated retrospectively. All patients received transfusions due to overt (melena, hematochezia, or hematemesis) or occult (iron-deficiency anemia) GI bleeding. Hemorrhagic GDVE was diagnosed by endoscopic visualization of oozing, blood clots arising from the coalescence of numerous vascular ectasias, or longitudinal red stripes radiating from the antrum and/or duodenum. Patients with radiation-induced GDVE had a typical glove and stocking distribution pattern along the radiation field, making duodenal involvement relatively common. Other sources of active bleeding were excluded. Bleeding from the varix, portal-hypertensive gastropathy (PHG), and ulcer were excluded by endoscopic findings. Colonoscopy to exclude possible lower GI bleeding was performed in three patients with blood clots only, with all three showing improvements after APC treatment. Radiation-induced hemorrhagic GDVE was diagnosed in each patient by two independent expert endoscopists, with an expert radiation oncologist confirming that the mucosal areas affected were compatible with the radiation field. The demographical and clinical characteristics of the study population were reviewed, as were findings at initial presentation, underlying diseases, total radiation dose due to underlying disease, time between radiation treatment and radiation-induced hemorrhagic GDVE, endoscopic findings, and GDVE pattern (watermelon or honeycomb) at initial diagnosis. The institutional review board of our institution approved this study, and each patient provided written informed consent. APC treatment was administered using an ERBE VIO300D system/APC2 (digital) (ERBE Elektromedizin, Tübingen, Germany) and probes including side-wide beam, side-conical beam, straight-axial beam, and 360 round beam probes (ERBE). APC
239
was usually administered at 1–1.2 L/min with a coagulation power setting of 30–40 W (pulsed mode effect 2). Procedures were performed with the patient under conscious sedation, with midazolam used in the six patients without liver cirrhosis but not in the 12 patients who had liver cirrhosis due to the high risk of hepatic encephalopathy. All 18 patients received 20 mg of hyoscine butylbromide to reduce gastric peristalsis. Each treatment session began from the bulb or pylorus and moved proximally, treating all visible lesions. The duration of each treatment session depended on each patient’s tolerance of the procedure but was not more than 30 min. APC treatment was repeated every 1–2 weeks until all treatment end points were reached. Treatment end points were defined as resolution or stabilization of anemia and transfusion dependence, cessation of macroscopic GI bleeding, and endoscopic ablation of >90% of the vascular lesions [9,18]. Once these treatment end points had been reached, there was no endoscopic follow up unless the patient developed recurrent symptoms or signs of GI blood loss. Following APC treatment, all patients were treated with an oral proton pump inhibitor for 2–4 weeks. Mean number of units of blood transfused before and after APC were calculated [19]. Pretreatment hemoglobin level was defined as the concentration on admission for APC treatment and before any transfusions. Post-treatment hemoglobin level was defined as the concentration 3 months after completion of endoscopic therapy or at last follow-up visit [20]. Changes in mean hemoglobin level and packed RBC transfusion volume were analyzed using the Wilcoxon signed-rank test. All statistical analyses were performed using STATA version 10.0 (Stata Corp, College Station, TX, USA), and values of p < 0.05 was considered statistically significant. Results The 18 patients comprised 12 males and 6 females, of mean age 59 years (range 42–80 years). Thirteen patients (72%) presented initially with overt bleeding, whereas the other five patients complained of dizziness and showed iron-deficiency anemia due to occult bleeding. Underlying diseases included hepatocellular carcinoma (HCC) in 13 patients (72%), pancreatic cancer in 3 patients (16%), and cholangiocarcinoma in 2 patients (11%). All patients had unresectable disease. All HCC patients had portal vein tumor thrombosis. All pancreatic cancer and cholangiocarcinoma patients had locally advanced disease, for which they received radiation treatment in
Melena
Melena
Melena
Melena
Melena
IDA
Melena
IDA
IDA
Hematemesis
IDA
Melena
Melena
Melena
Melena
Melena
IDA
Melena
1. F, 46
2. M, 54
3. M, 44
4. F, 61
5. F, 65
6. F, 72
7. M, 54
8. M.64
9. F, 42
10. F, 61
11. M, 44
12. M, 79
13. M, 80
14. M, 60
15. M, 60
16. M, 44
17. M, 58
18. M, 74
HCC
HCC
HCC
Cholangiocarcinoma
Cholangiocarcinoma
HCC
Pancreatic cancer HCC
HCC
HCC
Pancreatic cancer Pancreatic cancer HCC
HCC
HCC
HCC
HCC
HCC
Under lying disease
5000
5250
4400
5500
5040
5400
5500
3600
5400
5000
5400
5040
5040
3600
5100
5400
5500
5400
Total radiation dose (cGy)
4.8
3.3
3.5
6.7
5.0
6.2
21.5
8.2
4.1
4.8
4.1
5.7
9.4
3.7
4.2
4.3
4.2
3.5
Interval between radiation and diagnosis (months)
Oozing
Oozing
Oozing
Oozing
Oozing
Oozing
Oozing
Oozing
Blood clots Oozing
Oozing
Blood clots Blood clots Oozing
Oozing
Oozing
Oozing
Oozing
Endoscopic finding
Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb Honey comb
Water-melon
GDVE pattern
1
4
2
3
1
1
4
1
4
2
2
3
2
1
4
3
3
4
APC session
4
8
8
9
23
6
18
19
19
0
9
28
1
10
37
16
24
7
Total transfusion volume (packed RBC unit)
9.1
4.7
6.8
5.2
4.9
7.5
6.8
9.3
8.4
9.5
6.8
3.5
5.5
5.1
5.2
8.7
2.9
8.5
Pre-APC
12.1
9.5
7.9
10.7
11
8.4
7.8
10.6
10.4
11.7
8.5
7.1
12.7
8.7
7.2
10.0
9.7
11.1
Post-APC
Hb level (g/dL)
12.8 (Dead) 10.8 (Dead) 5.4 (Dead) 9.5 (Dead) 1.5 (Dead) 18.7 (Dead) 1.2 (Dead) 0.6 (Dead) 3.3 (Dead) 16.0 (Dead) 0.6 (Dead) 4.3 (Dead) 3.7 (Dead) 24.5 (Alive) 4.2 (Alive) 3.7 (Dead) 6.5 (Dead) 5.0 (Dead)
Follow-up (months)
Abbreviations: APC = Argon plasma coagulation; cGy = Centigray; F = Female; GDVE = Gastro-duodenal vascular ectasia; HCC = Hepatocellular carcinoma; IDA = Iron-deficiency anemia; M = Male; No = Number; RBC = Red blood cell; Hb = Hemoglobin.
Mode of presentation
Patient No. Sex, Age (years)
Table I. Clinical characteristics of GDVE.
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
240 H.-W. Kwak et al.
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
Efficacy of APC in the treatment of GDVE A
B
C
D
241
Figure 1. Endoscopic findings in a patient with radiation-induced hemorrhagic GDVE (patient 1): (A) before the first session of APC; (B) immediately after the first session of APC; (C) after the second session of APC; and (D) 3 months after the end of APC are shown.
combination with chemotherapy. Twelve patients (67%) had liver cirrhosis, including two, nine, and one with Child-Pugh classes A, B, and C, respectively. Five patients (28%) had esophageal varices and three patients (17%) had PHG. Mean total radiation dose due to underlying disease was 5032 cGy (range 3600– 5500 cGy) and median time between radiation treatment and radiation-induced hemorrhagic GDVE was 4.6 months (range 3.3–21.5 months). Endoscopically, 15 patients (83%) showed active bleeding with an oozing pattern (Forrest classification Ib), whereas the other 3 patients (17%) had blood clots only (Forrest classification IIb). None had spurting bleeding (Forrest classification Ia) or visible vessels (Forrest classification IIa). Seventeen patients (94%) had a honeycomb pattern of GVDE, and one (6%) had a watermelon pattern. The mean number of APC sessions per patient was 2.6 (range 1–4). At last endoscopic examination, all patients showed cessation of macroscopic GI bleeding and the ablation of almost all vascular lesions (Figures 1,2,3). Mean hemoglobin concentration was significantly higher after (9.7 g/dL; range = 7.1–12.7 g/dL) APC treatment (p < 0.001) compared to before treatment (6.6 g/dL; range = 2.9–9.5 g/dL), whereas mean
packed RBC transfusion volume was significantly lower after treatment (4.1 vs. 9.1 units, p = 0.038). Median follow-up period after last APC treatment was 4.7 months (range 0.6–24.5 months) (Table I). No patient experienced any major APC treatmentrelated adverse event or death. None of the patients experienced a recurrence of GI bleeding, and none required transfusions >3 months after the end of treatment. Discussion The stomach has a thick muscular layer and mucosal lining and is relatively resistant to radiation. However, at higher radiation doses (usually >5000 rad), ulceration may lead to perforation [21,22]. Acute vasculopathy may progress to a prolonged and progressive obliterative endarteritis, vasculitis, and endothelial proliferation, leading to mucosal ischemia, ulceration, and mucosal telangiectasias, and ultimately to fibrosis [23]. Mucosae that sustain late radiation injury appear friable and granular with multiple telangiectasias. Similarly, pelvic irradiation of the rectum, intestine, and bladder results in a similar pathology, appearing as proctitis, colitis, and intravesical hemorrhage,
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
242
H.-W. Kwak et al. A
B
C
D
Figure 2. Endoscopic findings in a patient with radiation-induced hemorrhagic GDVE (patient 8): (A) before the first session of APC; (B) the first session of APC; (C) 1 day after the first session of APC; and (D) 1 week after the end of APC are shown.
respectively. A study of biopsy samples obtained from five patients showed pathological features typical of vascular ectasia, including hypertrophy of the antral mucosa, dilation of mucosal capillaries with focal thrombosis, and fibromuscular hyperplasia of the lamina propria (Figure 4) [24,25]. However, the absence of these characteristics on biopsy would not preclude a diagnosis of radiation-induced GDVE because biopsies are typically insufficiently deep or the lesions are focal and, therefore, easily missed on biopsy. In addition, biopsy can worsen bleeding. Biopsy is not always required to confirm the diagnosis, and treatment can be started based on typical appearance at initial endoscopy. Radiation-induced GDVE is a diffuse process with multiple bleeding sites leading to significant blood loss and requiring multiple blood transfusions, prolonged hospitalizations, and repeated endoscopies with unsatisfactory control of bleeding. Radiationinduced GDVE may be a subset of gastric vascular ectasia, although the main lesions of the latter are located on the antrum (i.e. gastric antral vascular
ectasia [GAVE] syndrome) [26]. Patients with radiation-induced GDVE have a history of radiation treatment and a glove and stocking distribution pattern along the radiation field, making duodenal bulb involvement relatively common, as observed in our patients. Endoscopically, the lesions take two forms: striped (so-called watermelon stomach) and diffuse, honeycomb, punctate patterns. Liver cirrhosis has been reported to be strongly associated with the development of punctate-type vascular ectasias [27]. Only 1 of our 18 patients showed a striped pattern. The differential diagnosis of radiation-induced GDVE and PHG is essential for correct treatment, since GDVE lesions are completely distinct from PHG [28]. Treatment of portal hypertension with b-blockers or transjugular portosystemic shunting leads to an improvement in PHG patients but has no effect in GDVE patients. Both types of lesions may appear as multiple red spots, but lesions in GDVE patients are located mainly in the antrum, whereas lesions in PHG patients, including those with portal hypertensive duodenopathy and/or jejunopathy,
Efficacy of APC in the treatment of GDVE
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
A
243
B
C
Figure 3. Endoscopic findings in a patient with radiation-induced hemorrhagic GDVE (patient 12): (A) before the first session of APC; (B) the first session of APC; and (C) 2 weeks after the end of APC are shown.
usually show a diffuse mosaic pattern or snake skin appearance in the corpus and fundus of the stomach. Patients with radiation-induced GDVE usually have scattered punctate patterns of vascular ectasias and a A
typical glove and stocking distribution pattern along the radiation field. Biopsy may be useful in distinguishing between these two entities. Of our 13 patients with hepatocellular cancer, 3 had PHG; in these B
Figure 4. Pathologic findings in a patient with radiation-induced hemorrhagic GDVE: (A) dilated vessel and thrombosis and (B) ischemic damage in lamina propria are shown.
Scand J Gastroenterol Downloaded from informahealthcare.com by Purdue University on 08/30/14 For personal use only.
244
H.-W. Kwak et al.
patients, the bleeding focus was far from the PHG lesions, and active bleeding no longer occurred after APC treatment of GDVE, despite the presence of PHG lesions. APC is an electrosurgical endoscopic procedure indicated for hemostasis and devitalization of tissues [29]. APC is an effective treatment for radiation proctitis and gastric vascular ectasia, reducing bleeding and requirements for iron or blood transfusion by cauterizing mucosal telangiectasias [8–10]. There have been few comparisons of APC with other established treatment modalities, such as the heater probe, bipolar coagulation probe, and Nd:YAG laser [30,31]. The advantages of APC include a noncontact mode; axial, radial, and retrograde application; controllable depth of coagulation; and low cost of purchase and maintenance of the instrumentation. APC is associated with a lower risk of perforation than other modalities because it coagulates tissue without contact. Endoscopic ultrasound after APC showed the disappearance of the small hypoechoic spaces within the second and third sonographic layers and no evidence of injury to the fourth sonographic layer, the muscularis propria [32]. In the patients described here, procedures were performed under conscious sedation, but in the absence of midazolam in the 12 patients with liver cirrhosis due to the high risk of hepatic encephalopathy. Propofol sedation may be more comfortable, but APC treatment was not difficult and lasted
