Surg Today (2006) 36:193–197DOI 10.1007/s00595-005-3104-6

Primitive Neuroectodermal Tumor Arising in the Colon:Report of a Case

Kouki Kuwabara1, Hideyuki Ishida1, Kazuo Shirakawa2, Masaru Yokoyama1, Hiroshi Nakada1,Yoichi Hayashi2, Daijo Hashimoto2, Ichiro Miura3, Shinji Itoyama3, and Yuji Heike4

Departments of 1 Digestive Tract and General Surgery, 2 Hepato-Biliary-Pancreatic and Pediatric Surgery, and 3 Pathology,Saitama Medical Center, Saitama Medical School, 1981 Kamoda, Tsujido-cho, Kawagoe, Saitama 350-8550, Japan4 National Cancer Center Research Pharmacology Division Institute, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan

Case Report

A 59-year-old man, who had been followed up for dia-betes mellitus for 3 years by an internist at our institu-tion, was referred to our surgical outpatient clinic forinvestigation and treatment of a tumor in his left flank.Physical examination on admission revealed a hardmass, 10cm in diameter, palpable in the left flank. Hisfamily history was noncontributory, and his medical his-tory was remarkable only for diabetes mellitus. Labora-tory examinations revealed a low hemoglobin level(9.0 g/dl) and a high glycohemoglobin level (HbA1c:8.7%); however, the serum carcinoembryonic antigen(CEA) and carbohydrate antigen 19-9 (CA 19-9)levels were within normal limits (CEA, 4.5ng/ml; cutoff6.7ng/ml; CA 19-9, 5U/ml; cutoff 37U/ml). Barium en-ema showed the “apple core” sign in the descendingcolon (Fig. 1a). Colonoscopy showed a circumferentialtype 3 infiltrating ulcerating tumor3 in the descendingcolon (Fig. 1b). Histological examination of a biopsyrevealed malignant small, round cell proliferation, al-though definite cytological diagnosis was not made. Anabdominal computed tomography (CT) scan showedthe circumferentially thickened wall of the descendingcolon (Fig. 2a). There was no evidence of liver metasta-sis, para-aortic lymph node involvement, or peritonealdissemination. Chest radiography showed no abnor-malities. He was operated on 12 days after admission.Laparotomy revealed a whitish tumor in the descendingcolon growing invasively into the iliopsoas muscle.Because we did not find any noncurative signs in theabdomen, we performed left hemicolectomy withlymph node dissection at the D3-level3 and resection ofthe involved muscles (Fig. 3). The patient had an un-eventful postoperative course and was discharged 30days after surgery.

At the 3-month follow-up, another 9 ¥ 13-cm masswas found in the left flank (Fig. 2b). The patient wasreadmitted and informed about the therapeutic options,

AbstractPeripheral primitive neuroectodermal tumors(pPNETs) are usually found in the soft tissue of theextremities, paravertebral region, and chest wall. Wereport a rare case of a pPNET arising in the colon. A 59-year-old man underwent left hemicolectomy for an infil-trative ulcerating tumor, 11cm long, in the descendingcolon. Histological examination of the resected specimen revealed small, round cell proliferation withrosette-like structures, and confirmed regional lymphnode involvement and peritoneal dissemination nearthe primary tumor. Immunohistochemically, the tumorcells were positive for synaptophysin and MIC2 (CD99). ESW-FLI1 chimeric mRNA was detected in thetumor by reverse transcriptase–polymerase chain reac-tion. The patient underwent resection of recurrence inthe retroperitoneum 3 months later, but metastasis rap-idly developed and he died of the disease 7 months afterhis first operation.

Key words Primitive neuroectodermal tumor · Colon

Introduction

Peripheral primitive neuroectodermal tumors(pPNETs) are most commonly found in the extremities,paravertebral region, or chest wall of children andadolescents.1,2 Although this rare malignant tumor mayarise anywhere in the body, its occurrence in the diges-tive tract is very unusual. We report the case of apPNET arising in the descending colon. To our knowl-edge, this is the first documentation of a pPNET arisingin the colon.

Reprint requests to: K. KuwabaraReceived: September 14, 2004 / Accepted: May 24, 2005

194 K. Kuwabara et al.: Primitive Neuroectodermal Tumor of the Colon

which included resective surgery, chemotherapy, radio-therapy, and their combinations. He chose to undergoresective surgery, which was performed 112 days afterthe initial surgery. At the second laparotomy, the recur-rent tumor in the retroperitoneium was found to involvethe jejunum. We performed macroscopic en-blocresecion including the jejunum, but the tumor rapidlyrecurred and invaded the skin (Fig. 2c). The skin tumorwas biopsied for a chemosensitivity test by the CD-geldroplet method,4 but this test was unsuccessful. Thepatient died of the disease 7 months after the laparo-tomy. There were no findings of distant hematogenousmetastasis throughout his clinical course. An autopsywas not performed.

Histological Examinations

Histologically, the resected specimen contained a sheet-like proliferation of spindle-to-polygonal cells with largenuclei and thin vascular stroma (Fig. 4a). Focally,the tumor formed rosette-like structures (Fig. 4b). Thetumor cells invaded the iliopsoas muscle beyondthe colonic wall. Lymph node metastasis was found in theparacolic regional lymph nodes, and peritoneal dissemi-nation was confirmed in the mesentery near the primarylesion. Immunohistochemically, the tumor cells werepositive for chromogranin (Fig 4c) and MIC-2 (CD99)(Fig 4d), but negative for c-kit, CD 34, vimentin, CD56,S-100, and keratin. The positive rate of MIB-1 was 62%.

Fig. 1. a Barium enema showedthe “apple core” sign in the de-scending colon. b Colonoscopyshowed a circumferential type 3tumor in the descending colona b

Fig. 2. Abdominal computed tomography scans showed: a the circumferentiallythickened wall of the descending colon before the first operation, b a 9 ¥ 15-cmmass in the left flank 102 days after the first operation, and c the recurrent tumorinvading the skin 149 days after the first operation

a b

c

195K. Kuwabara et al.: Primitive Neuroectodermal Tumor of the Colon

extracted by the guanidinium thiocyanate–phenol–chloroform extraction method. The primers used were5¢-CCCACTAGTTACCCACCCCAAA-3¢ and 5¢-TGTTGGGCTTGCTTTTCCGCTC-3¢. Thermal cy-cling was performed with initial denaturation at 94°Cfor 45s, 34 cycles of denaturation at 94°C for 45s,annealing at 68°C for 30s, and extension at 72°C for45min. The PCR products were electrophoresed on a2% agarose gel containing ethidium bromide. TheEWS-FLI1 chimeric mRNA gene was expressed in thetumor (Fig. 5).

Fig. 3. The resected specimen of descending colon contained atype 3 tumor

Reverse Transcriptase–Polymerase Chain Reaction(RT-PCR)

This study was approved by the Ethical Committee ofSaitama Medical School. Written informed consent wasobtained from the patient and from two other patientswhose gastrointestinal tumors served as a negative con-trol. The detection of EWS-FLI1 fusion transcripts fromthe tumor was performed according to the method de-scribed by Tokudome et al.5 with some modifications.Briefly, the total RNA of each resected specimen was

a b

c d

Fig. 4. a Histological examina-tion of the resected specimenrevealed a sheet-like prolifera-tion of spindle-to-polygonalcells (H&E, ¥25). b The tumorcontained the focal formation ofrosette-like structures (H&E,¥100). c,d Immunohistochemicalstaining revealed tumor cellspositive for chromogranin (c)(¥100) and MIC-2 (CD99) (d)(¥100)

Fig. 5. Expression of the EWS-FLI 1 chimeric mRNA usingreverse transcriptase–polymerase chain reaction. Lane 1, mo-lecular size marker (100bp DNA ladder, Gibco BRL, GrandIsland, NY, USA); lane 2, peripheral primitive neuroectoder-mal tumor (operatively resected tumor); lanes 3, 4, operativelyresected gastrointestinal stromal tumor. The EWS-FLI1 chi-meric genes were evident in lane 2 (arrow)

196 K. Kuwabara et al.: Primitive Neuroectodermal Tumor of the Colon

Discussion

Primitive neuroectodermal tumors can be classifiedlargely into two groups: central PNETs, arising fromthe central nervous system, and peripheral PNETs(pPNETs), arising from other sites. According to recentreports, pPNET and Ewing sarcoma (ES) are biologi-cally equivalent because of the following common fac-tors: first, genes such as EWS on chromosome 22, F1I-1on chromosome 11, and ERG on chromosome 21, makeup the chimeric gene;6,7 and second, since the tumorcells express p30/32MIC antigen, a glycogen encoded bythe pseudoautosomal MIC2 gene and clustered as theCD99 antigen, the cell membrane becomes positive forthe anti-MIC2 (CD99) antibody in immune histochemi-cal staining.8,9 Therefore, both diseases are now classi-fied under the Ewing family of tumors.6

Although ES and pPNETs are grouped in the Ewingfamily of tumors, the pPNET is sometimes differenti-ated from ES by identifying clear neural differentia-tion.10,11 Therefore, pPNETs are characterized bypositive staining for neural markers such as chromo-granin and occasional rosette formation, whereas ESbasically shows no such findings.

In our patient’s tumor, the chimeric mRNA corre-sponding to chromosomal translocation in the Ewingfamily of tumors identified by RT-PCR, immunohisto-logic positivity for MIC (CD 99) and chromogranin(a neural marker), and rosette formation all supportthe diagnosis of a pPNET. The most common sites ofpPNETs are the extremities, the paravertebral region,and the chest wall,1,2 although there are sporadic casereports of pPNETs arising in the kidneys,12 gynecologi-cal organs,12 pancreas,12 omentum,13 and the mesentery.5

Conversely, the incidence of a pPNET in the alimentarytract is extremely rare.12,14,15 In fact, we were able to findonly one report of a pPNET occurring in the rectum,16

based on a MEDLINE search up until June 2004. More-over, we found no documentation of a pPNET arising inthe colon, and this case may represent the first.

Peripheral primitive neuroectodermal tumor is highlymalignant and is associated with a poor prognosis. Ittends to recur and metastasize hematogenously early inits development.1,17,18 In fact, Kushner et al.1 reportedthat the 2-year survival rate of patients with tumorsgreater than 5cm was only 25%, and Ushigome et al.18

reported that 21 of 25 patients died within 3–29 monthsafter the diagnosis. Despite the potentially curative re-section in our patient, local recurrence was found withina few weeks of resection. Unless further cases ofpPNET in the colon are reported and analyzed, we willnot know whether local recurrence is characteristic.

Although several treatment modalities have beenreported, namely, surgical resection, high-dose chemo-therapy,17,19 and radiotherapy,17,19 there is still no estab-

lished treatment for pPNETs. It could be argued thatchemotherapy or radiotherapy, or both, should be thefirst choice of treatment for recurrent pPNETsconsidering the speed of their growth. Because there isno reliable and effective treatment for pPNET, we in-formed our patient of the options available and per-formed resective surgery for the recurrent tumor withhis full consent. We still believe that this decision wasappropriate because of the involvement of the jejunum.The subsequent rapid growth of the tumor after thesecond operation prompted us to use chemotherapeuticagents. However, the test for sensitivity was unsuccess-ful, and we did not have the opportunity to give himchemotherapy.

In summary, pPNETs rarely arise in the colon. How-ever, our documentation of this case may lead to therecognition of additional cases in the future.

References

1. Kushner BH, Hajdu SI, Gulati SC, Erlandson RA, Exelby PR,Lieberman PH. Extracranial primitive neuroectodermal tumors.The Memorial Sloan-Kettering Cancer Center experience.Cancer 1991;67:1825–9.

2. Dehner LP. Primitive neuroectodermal tumor and Ewing’ssarcoma. Am J Surg Pathol 1993;17:1–13.

3. Japanese Research Society for Cancer of the Colon and Rectum.General rules for clinical and pathological studies on cancer ofcolon, rectum, and anus in Japan. Tokyo: Kanehara; 1998.

4. Araki Y, Isomoto H, Matsumoto A, Kaibara A, Yasunaga M,Hayashi K, et al. An in vitro chemosensitivity test for colorectalcancer using collagen-gel droplet embedded cultures. KurumeMed J 1994;46:163–6.

5. Tokudome N, Tanaka K, Kai MH, Sueyoshi K, MatsukitaS, Setoguchi T. Primitive neuroectodermal tumor of the trans-verse colonic mesentery defined by the presence of EWS-FLI1chimeric mRNA in a Japanese woman. J Gastroenterol 2002;37:543–9.

6. de Alva E, Gerald WL. Molecular biology of the Ewing’s sar-coma/primitive neuroectodermal tumor family. J Clin Oncol 2000;18:204–13.

7. May WA, Denny C. Biology of EWS/FLI and related fusion genesin Ewing’s sarcoma and primitive neuroectodermal tumor. CurrTop Microbiol Immunol 1997;220:143–50.

8. Ambros IM, Ambros PF, Strehl S, Kovar H, Gadner H, Salzer-Kuntschik M. MIC2 is a specific marker for Ewing’s sarcoma andperipheral primitive neuroectodermal tumors. Cancer 1991;67:1886–93.

9. Fellinger EJ, Garin-Chesa P, Triche TJ, Huvos AG, Rettig WJ.Immunohistochemical analysis of Ewing’s sarcoma cell surfaceantigen p30/32MIC2. Am J Pathol 1991;139:317–25.

10. de Alava E, Antonescu CR, Panizo A, Leung D, Meyers PA,Huvos AG, et al. Prognostic impact of p53 status in Ewing’ssarcoma. Cancer 2000;89:783–92.

11. Kovar H. Ewing’s sarcoma and peripheral primitive neuroecto-dermal tumors after their genetic union. Curr Opin Oncol 1998;10:334–42.

12. O’Sullivan MJ, Perlman EJ, Furman J, Humphrey PA, DehnerLP, Pfeifer JD. Visceral primitive peripheral neuroectodermaltumors: a clinicopathological and molecular study. Hum Pathol2001;32:1109–15.

13. Tanida S, Tanioka T, Inukai M, Yoshioka N, Saida Y, Imai K,et al. Ewing’s sarcoma/peripheral primitive neuroectodermal

197K. Kuwabara et al.: Primitive Neuroectodermal Tumor of the Colon

tumor (pNET) arising in the omentum as a multilocular cyst withintracystic hemorrhage. J Gastroenterol 2000;35:933–40.

14. Graham DK, Stork LC, Wei Q, Ingram JD, Karrer FM, MierauGW, et al. Molecular genetic analysis of a small bowel primitiveneuroectodermal tumor. Pediatr Dev Pathol 2003;6:2787–9.

15. Kie JH, Lee MK, Kim CJ, Lee K, Kwon KW, Yang WI. PrimaryEwing’s sarcoma of the duodenum: a case report. Int J Pathol2003;11:331–7.

16. Drut R, Drut M, Muller C, Marron A. Rectal primitive neuro-ectodermal tumor. Pediatr Pathol Mol Med 2003;22:391–8.

17. Jurgens H, Bier V, Harms D, Beck J, Brandeis W, Etspuler G,et al. Malignant peripheral neuroectodermal tumors. A retro-spective analysis of 42 patients. Cancer 1988;61:349–57.

18. Ushigome S, Shimoda T, Nikaido T, Nakamori K, Miyazawa Y,Shishikura A, et al. Primitive neuroectodermal tumors of boneand soft tissue with reference to histologic differentiation in pri-mary or metastatic foci. Acta Pathol Jpn 1992;42:483–93.

19. Zimmermann T, Blutters-Sawatzki R, Flechsenhar K, PadlbergWM. Peripheral primitive neuroectodermal tumor: challenge formultimodal treatment. World J Surg 2001;25:1367–72.