Conservative Management of Pneumatosis Intestinalis After Allogeneic Hematopoietic SCT

Home , Abdominal pain, Nausea, Pneumatosis intestinalis, Vomiting

LETTER TO THE EDITOR Conservative management of pneumatosis intestinalis after allogeneic hematopoietic SCT

Bone Marrow Transplantation (2014) 49, 1436–1438; doi:10.1038/ chemotherapy, type of transplant, stem cell source, aGVHD, bmt.2014.148; published online 21 July 2014 immunosuppression, CMV status and Clostridium difficile infection. In addition, the onset, management and outcome of PI were recorded. Approval for the study was obtained from our Pneumatosis intestinalis (PI) is characterized by the accumulation Institutional Review Board. During that time period, 1288 patients of gas in the intestinal wall. It is not considered to be a disease underwent alloHSCT at our institution. Sixteen patients were itself, but rather a pathologic and radiographic finding. PI is diagnosed with PI at a median of 108 days after alloHSCT (range, associated with various disorders including chronic obstructive 25–578). Three cases of PI were identified between 2005 and 2010 pulmonary disease, autoimmune disorders and solid organ (for 900 alloHSCT performed) and 13 cases between 2010 and transplantation.1,2 The radiographic imaging is diagnostic, show- 2013 (for 388 alloHSCT performed). This represents a statistically ing air-filled bullae along the intestinal border and occasionally a significant increase in the incidence of PI using Fisher’s exact test pneumoperitoneum.3 PI has been observed and reported after (Po0.0001). All patients had aGVHD at the time of PI diagnosis: 15 allogeneic hematopoietic SCT (alloHSCT).3–7 The occurrence of PI patients had GI +/ − skin aGVHD and 1 patient had liver-only in this setting is thought to be associated with chemotherapy, aGVHD. Two of the patients who developed PI late after alloHSCT acute GVHD (aGVHD), immunosuppression and infections. As have received donor lymphocyte infusion 2 months prior to PI symptoms of PI may mimic gastrointestinal aGVHD (GI-aGVHD), PI development and two additional patients had a biopsy-proven GI- is frequently diagnosed incidentally during imaging performed to aGVHD. Fifteen patients were receiving systemic steroids at the evaluate GI-aGVHD. We describe a series of 16 patients presenting time of PI diagnosis and 11 were also on BUD +/ − BDP. Nobody with PI after alloHSCT. The majority of our patients had GI-aGVHD had active C. difficile or CMV colitis. Twelve patients had evidence treated with prednisone and non-absorbable oral steroids of intraperitoneal or extraluminal free air, however, without beclomethasone dipropionate (BDP) and budesonide (BUD). We clinically evident peritonitis. All patients were managed conserva- discuss the possible role of systemic steroids, as well as BUD, in tively with bowel rest, total parenteral nutrition, antibiotics, the development of PI. temporary discontinuation of BUD/BDP and the reduction of We retrospectively searched our radiology database for the prednisone dose. On repeated imaging, 11 patients had resolution images obtained by performing computerized tomography (CT) of PI, 2 had significant improvement and 3 had no repeated on patients who underwent alloHSCT from January 2005 to imaging. Ten patients have died at the median of 72 days (range, September 2013. Search terms included ‘pneumatosis’, ‘pneuma- 7–798) after PI diagnosis: six had documented resolution of PI, two tosis intestinalis’ and ‘pneumatosis cystoides intestinalis’.An had documented improvement and two had no repeated imaging extensive chart review was performed on identified cases for prior to death. Patients and transplant characteristics are shown in the information on age, gender, diagnosis, conditioning Tables 1 and 2.

Table 1. Patient and transplant characteristics

Pt Age/ Diagnosis Conditioning Donor; stem cell source aGVHD Immunosuppressive therapy Steroid CMV sex regimen dose (mg) viremia

1 52/M AML Bu/Cy mMUD (9/10); PBSC GI—grade 2, skin Pred, T, MMF, BUD, BDP, Photo 60 + 2 64/F CLL Flu/Bu MUD; PBSC GI—grade 4, skin Pred, T, MMF, BUD, BDP, Photo 100 − 3 20/F AML Bu/Cy mMUD (8/10); PBSC GI—grade 2, skin Pred, T, MMF, BUD, BDP, Photo, Sorafenib 10 − 4 61/M AML Bu/Cy MUD; PBSC GI—grade 2, skin Pred, T, MMF, BUD, Photo 17.5 − 5 45/F AML Bu/Cy mMRD (8/10); PBSC GI—grade 1, skin Pred, MMF, BUD 30 − 6 47/M AML Bu/Cy MUD; BM GI—grade 2 Pred, BUD 30 − 7 64/M DLBCL Flu/Bu/ATG MUD; PBSC GI—grade 4 Methylpred, T, MMF, BUD, BDP 70 − 8 67/M AML Flu/Bu/ATG MUD; PBSC Liver—grade 1 MMF, T n/a + 9 25/M HL Flu/TBI MUD; PBSC GI—grade 2 Methylpred, sirolimus, MMF, BUD, BDP, 80 − 10 45/M NHL Bu/Cy MRD; PBSC GI—grade 1, skin Pred, T, Photo 5 − 11 66/F AML Clo/Bu MRD; PBSC GI—grade 4 Pred, T, MMF, BUD, BDP 25 + 12 31/M AML Bu/Cy MUD; PBSC GI—grade 1 Pred, T, MMF, BUD 80 − 13 57/F CML TBI/Cy MUD; PBSC GI—grade 1 Pred, T, MMF 80 − 14 56/M AML TBI/Cy MUD; PBSC GI—grade 4 Pred, T, MMF 60 + 15 56/F NHL TBI/Cy MUD; PBSC GI—grade 2 Pred, T 60 + 16 49/F ALL Bu/Cy MUD; PBSC GI—grade 2 Methylpred, sirolimus, MMF, BUD, BDP 90 + Abbreviations: aGVHD = acute GVHD; ATG = anti-thymocyte globulin; Cy = cyclophosphamide; BDP = beclomethasone dipropionate; Bu = busulfan; BUD = budesonide; Clo = clofarabine; DLBCL = diffuse large B-cell lymphoma; Flu = ﬂudarabine; GI = gastrointestinal; HL = Hodgkin’s lymphoma; Methylpred = methylprednisolone; MMF = mycophenolate mofetil; MUD = matched unrelated donor; mMRD = mismatched related donor; mMUD = mismatched unrelated donor; MRD = matched related donor; NHL = non-Hodgkin lymphoma; PBSC = peripheral blood stem cells; Photo = photopheresis; Pred = prednisone; Pt = patient; T = tacrolimus; TBI = total body irradiation. Letter to the Editor 1437

Table 2. PI onset and outcome

Pt PI onset after Symptoms at Imaging ﬁndings, affected areas Days to PI Survival after alloHSCT (days) and Survival alloHSCT presentation resolution cause of death after PI (days) (days)

1 105 Abdominal pain Ascending colon; free air Resolved; 77 Alive D +710 605 2 105 Abdominal pain, Ascending and transverse Resolved; 40 Expired D +155; stroke, 50 diarrhea colon; free air aspiration pneumonia 3 265 (94 after Abdominal pain Ascending colon Resolved; 72 Alive D +922 657 DLI) 4 413 (90 after Abdominal pain Ascending and transverse Resolved; 156 Alive D +1019 606 DLI) colon; free air 5 150 Diarrhea, cough Ascending and transverse Resolved; 221 Alive D +762 612 colon; free air 6 99 Abdominal pain Cecum and ascending colon Resolved; 27 Expired D +178; progressive 79 aGVHD; bacteriemia 7 102 Nausea, vomiting, Ileum, transverse and No CT Expired D +149; progressive 47 diarrhea descending colon; portal aGVHD, hemorrhagic stroke venous gas 8 25 Fever Diffuse colon; free air No CT Expired D +94; Klebsiella and 69 Clostridium sepsis 9 578 Abdominal pain, Stomach to rectum; portal Improvement Expired D +589; progressive 11 nausea, vomiting venous gas lymphoma, liver failure 10 111 None Ascending and transverse Resolved; 247 Alive D +909 798 colon; free air 11 135 Abdominal pain Distal ileum, ascending and Resolved; 6 Expired D +158; progressive 23 transverse colon; free air aGVHD, respiratory failure 12 371 Abdominal pain, Cecum and ascending colon; No CT Alive D +446 75 dyspnea free air 13 165 Chest pain Descending, sigmoid colon and Improvement Expired D +196; progressive 31 rectum; free air aGVHD, Aspergillus pneumonia, pulmonary embolism, 14 69 Diarrhea Cecum; free air Resolved; 25 Expired D +77; progressive 8 aGVHD, Klebsiella sepsis 15 41 Abdominal pain Cecum and ascending colon; Resolved; 36 Expired D +300; relapse 259 distension free air 16 38 Dyspnea Diffuse colon; retroperitoneal Resolved; 6 Expired D +45; progressive 7 free air aGVHD, respiratory failure Abbreviations: aGVHD = acute GVHD; CT = computerized tomography; D = day; PI = pneumatosis intestinalis; Pt = patient; No CT = no repeated CT imaging.

Although our study is observational and the findings are mainly prednisone, contributes to the reduced epithelial healing in these confirmatory, this is the largest reported series of patients patients. Even though we saw a significant increase in the developing PI after alloHSCT and the first report describing the incidence of PI after we began to use BUD for GI-aGVHD, these possible role of BUD in the development of PI. It has become a results should be interpreted with caution. It is possible that the common practice to add non-absorbable oral steroids, BDP and increase in incidence of PI after 2010 is due to the higher 8,9 BUD, to the treatment of patients with GI-aGVHD. Both drugs utilization CT imaging and better supportive care of patients with have a high affinity for the local steroid receptors, are rapidly aGVHD with lower early mortality. Although patients with PI are metabolized by the intestinal cells and have a low bioavailability. frequently asymptomatic, some of the presenting symptoms may BDP is predominantly distributed to the upper GI tract, while BUD be diarrhea, intestinal bleeding, bloating, cramping or pain, works primarily in the ileocolonic region. Steroids that work at the fi fi mimicking GI aGVHD. CT scan is the most sensitive and speci c level of bowel mucosa interfere less with the bene cial graft- radiographic test to diagnose PI; however, it does not have versus-tumor effect compared to systemic immunosuppression. 3,5 prognostic implications. The presence of pneumoperitoneum in The potential mechanism for developing PI after alloHSCT is that these patients does not suggest a full mucosal perforation, but chemotherapy and immunosuppression induce atrophy of the rather a rupture of a pneumatocele. None of our patients required Peyer’s patches, leading to the loss of structural integrity of the bowel mucosa and subsequent gas migration into the submucosal surgical intervention or had a fatal outcome secondary to PI. There and subserosal regions. Mucosal injury and increased intraluminal is no increased morbidity and mortality in patients with aGVHD pressure in the setting of GI-aGVHD could be an additional who had PI compared with those without PI. Historically, the contributing factor. Another hypothesis is that gas-forming mortality in patients with PI has been associated with surgical bacteria invade the bowel wall, producing intramural gas. Steroids complications rather than peritonitis or septic shock. Although it is have known immunosuppressive properties, predispose to infec- possible that discontinuation of BUD might have had a beneficial tion and inhibit wound healing. In several preclinical studies, both effect on resolution of PI findings, improvements in PI could be prednisone and BUD have been found to inhibit intestinal due to the self-limiting nature of PI. Awareness about the epithelial cell restitution and proliferation in vitro.10 Both of these existence of this condition is of paramount importance in processes play an important role in rapid epithelial healing providing appropriate conservative therapy and minimizing following intestinal injury. It is possible that BUD, in addition to unnecessary surgical interventions.

© 2014 Macmillan Publishers Limited Bone Marrow Transplantation (2014) 1436 – 1438 Letter to the Editor 1438 CONFLICT OF INTEREST 4 Hepgur M, Ahluwalia MS, Anne N, Thomas J, Liu H, Schiff MD et al. The authors declare no conﬂict of interest. Medical management of pneumatosis intestinalis in patients undergoing allogeneic blood and marrow transplantation. Bone Marrow Transplant 2011; 46: 876–879. PK Bhamidipati1, A Ghobadi2, S Bauer2, JF DiPersio2 and I Pusic2 1 5 Ade-Ajayi N, Veys P, Stanton M, Drake DP, Pierro A. Conservative management of Division of Hospital Medicine, Department of Internal Medicine, pneumatosis intestinalis and pneumoperitoneum following bone-marrow trans- Washington University School of Medicine, St Louis, MO, USA and plantation. Pediatr Surg Int 2002; 18:692–695. 2 Division of Oncology, Department of Internal Medicine, Washington 6 Laskowska K, Burzynska-Makuch M, Krenska A, Koltan S, Chrupek M, Nawrocka E University School of Medicine, St Louis, MO, USA et al. Pneumatosis cystoides interstitialis: a complication of graft-versus-host E-mail: [email protected] disease. A report of two cases. Polish J Radiol 2012; 77:60–63. 7 Day DL, Ramsay NK, Letourneau JG. Pneumatosis intestinalis after bone marrow transplantation. Am J Roentgenol 1988; 151:85–87. REFERENCES 8 Hockenbery DM, Cruickshank S, Rodell TC, Gooley T, Schuening F, Rowley S et al. A randomized, placebo-controlled trial of oral beclomethasone dipropionate as a 1 Galm O, Fabry U, Adam G, Osieka R. Pneumatosis intestinalis following cytotoxic prednisone-sparing therapy for gastrointestinal graft-versus-host disease. Blood or immunosuppressive treatment. Digestion 2001; 64:128–132. 2007; 109: 4557–4563. 2 Ho LM, Mosca PJ, Thompson WM. Pneumatosis intestinalis after lung transplant. 9 Ibrahim RB, Abidi MH, Cronin SM, Lum LG, Al-Kadhimi Z, Ratanatharathorn V et al. Abdominal Imaging 2005; 30:598–600. Nonabsorbable corticosteroids use in the treatment of gastrointestinal graft- 3 McCarville MB, Whittle SB, Goodin GS, Li CS, Smeltzer MP, Hale GA et al. versus-host disease. Biol Blood Marrow Transplant 2009; 15:395–405. Clinical and CT features of benign pneumatosis intestinalis in pediatric hemato- 10 Jung S, Fehr S, Harder-d'Heureuse J, Wiedenmann B, Dignass AU. Corticosteroids poietic stem cell transplant and oncology patients. Pediatr Radiol 2008; 38: impair intestinal epithelial wound repair mechanisms in vitro. Scand J Gastro- 1074–1083. enterol 2001; 36: 963–970.