Outcome of hepatoblastoma treated with the JPLT-1 (Japanese Study Group for Pediatric Liver Tumor) Protocol-1: A report from the Japanese Study Group for Pediatric Liver Tumor.

PURPOSE: Hepatoblastoma is the most common malignant liver tumor in childhood. Multicenter studies elucidate the optimal pre- or postoperative chemotherapeutic regimens. This report reviews the results of the Japanese Study Group for Pediatric Liver Tumor Protocol-1 (JPLT-1) and compares its outcomes with published reports of other studies. METHODS: From March 1991 to December 1999, 154 patients with malignant liver tumor including 145 cases of hepatoblastomas were enrolled in the JPLT study. Data from 134 cases were analyzed in this study. JPLT-1 protocol 91A was used for patients with stage I or II hepatoblastoma. The chemotherapy regimen consisted of repeated courses of cisplatin (CDDP), 40 mg/m(2), and tetrahydropyranyl (THP)-Adriamycin, 30 mg/m(2). JPLT-1 protocol 91B was administered to patients with stage IIIA, IIIB, or IV hepatoblastoma. The chemotherapy regimen consisted of repeated courses of CDDP, 80 mg/m(2), and THP-Adriamycin, 30 mg/m(2)/day for 2 days. Courses were repeated every 4 weeks as tolerated. RESULTS: Seven patients died of chemotherapy-related side effects. Six of them died of sepsis caused by leukopenia and 1 case of liver failure. Overall survival rate (3-year/6-year) was 100%/100% for stage I (n = 9), 100%/95.7% for stage II (n = 32), 76.6%/73.8% for stage IIIA (n = 48), 50.3%/50.3% for stage IIIB (n = 25), 64.8%/38.9% for stage IV (n = 20), and 77.8%/73.4% overall. For stage IIIA and B disease, intravenous chemotherapy was better than intraarterial chemotherapy (66.4% v 38.1% for event-free survival and 69.3% v. 57.1% for overall survival). Patients less than 1 year of age had a better prognosis than older patients, but age was not a significant prognostic factor by multivariate analysis. CONCLUSIONS: The overall and event-free survival rates of the JPLT-1 study of hepatoblastoma were comparable with the results of other multicenter studies in Europe and the United States. The event-free survival rate at 3 years for stage IIIB and IV disease was under 50%. New treatment strategies are needed for patients with advanced hepatoblastoma.

Experience with International Neuroblastoma Staging System and Pathology Classification.

The International Neuroblastoma Staging System and Pathology Classification were proposed in 1988 and in 1999, respectively, but their clinical value has not yet been fully studied in new patients. Six hundred and forty-four patients with neuroblastoma treated between January 1995 and December 1999 were analysed by these classifications. The 4-year overall survival rate of patients <12 months of age with INSS stages 1, 2A, 2B, 3 and 4S disease was 98.5%, which was significantly higher than the 73.1% rate in stage 4 patients <12 months (P<0.0001). When patients were > or = 12 months, the 4-year overall survival rate of patients with neuroblastoma at 1, 2A, 2B and 3 stages was 100% and that of patients at stage 4 was 48.5% (P<0.0001). As to the International Neuroblastoma Pathology Classification histology, the 4-year overall survival rate was 98.8% in patients with favourable histology and 60.7% in those with unfavourable histology in the <12 months group (P<0.0001). In the > or = 12 months group, the 4-year oral survival of patients with favourable histology was 95.3% and that of patients with unfavourable histology was 50.6% (P<0.0001). Among biological factors, MYCN amplification, DNA diploidy and 1p deletions were significantly associated with poor prognosis in patients <12 months, as were MYCN amplification and DNA diploidy in patients > or = 12 months of age. Multivariate analysis showed that the INSS stage (stage 4 vs other stages) and International Neuroblastoma Pathology Classification histology (unfavourable vs favourable) were significantly and independently associated with the survival of patients undergoing treatment, stratified by age, stage and MYCN amplification (P=0.0002 and P=0.0051, respectively).

Fiberoptic colonoscopic polypectomy in childhood: report and review of cases.

BACKGROUND: Fiberoptic colonoscopy has been a routine therapeutic modality for colorectal polyps in pediatric patients. Methods of bowel preparation, anesthesia, area of investigation and treatment depending on histopathology are still controversial. In order to clarify the rationale of pediatric colonoscopy the present study was performed. METHODS AND RESULTS: We analyzed the results of colonoscopic examination in 21 patients with colorectal polyps. Mean patient age was 3.7 years, with a range of 1--7 years. Rectal polyps were seen in 10 cases: seven had a solitary polyp (juvenile in six and adenoma in one) and three had multiple polyps (juvenile, lymphoid and Peutz--Jeghers coexisting with hyperplastic polyps). Sigmoid colon polyps were seen in 10 cases: all were solitary juvenile polyps, but one had adenomatous change. Another had multiple Peutz-- Jeghers polyps located in the entire colon. Flexible colonoscopic polypectomy was performed in 16 patients and transanal polypectomy was performed in four patients. Autoamputation was seen in two cases of juvenile polyp (resection was ultimately performed in a case having repeated autoamputation). After removing the polyps, all patients have had no recurrence for a period ranging from 6 months to 15 years, except for one case with Peutz--Jeghers syndrome. CONCLUSIONS: Most polyps are located in the rectum or the sigmoid colon. Although the majority are solitary or juvenile polyps, because histopathologic variety is seen in pediatric colon polyps, histopathologic examination of each polyp is important to detect any dysplastic or adenomatous element with malignant potential and to make a suitable follow-up schedule. Symptomatic polyps should be removed by fiberoptic colonoscopy or transanal resection with total colon endoscopic examination under general anesthesia. Polypectomy using the electrocautery snare and clip is effective and safe and bowel preparation using polyethylene glycol electrolyte solution is sufficient for the procedure.

Treatment of ruptured undifferentiated sarcoma of the liver in children: a report of two cases and review of the literature.

Undifferentiated (embryonal) sarcoma of the liver (USL) is a highly malignant tumor of early life. Treatment choices for USL, especially with intraperitoneal rupture, are uncertain. Outcomes have been almost uniformly poor until recently. We describe two 7-year-old girls treated for ruptured USL. In the more recent patient, operative biopsy was followed by three cycles of cisplatin (CDDP), adriamycin (ADR), and cyclophosphamide (CPM). A fluid-filled cavity in the tumor showed enlargement and was drained. Two cycles of CDDP, ADR, vincristine (VCR), and ifosfamide were accompanied by reduction in tumor size, and trisegmentectomy was performed. She has no evidence of disease 3.5 years after surgery. In the other patient, left lobectomy was followed by a less intensive regimen, including CPM, VCR, and fluorouracil. This patient died of dissemination within 5 months. In 170 reported pediatric patients with USL, the 2-year disease-free survival was 17%. For the 96 such patients reported since 1980, 2-year disease-free survival had improved to 27%. More aggressive chemotherapy has been associated with this change. Of 8 patients with tumor rupture whose details have been reported (including the 2 present patients) after resection of the tumor, 4 died, 1 was alive with disease, and 3 were free of disease at 8, 49, and 58 months, respectively, after diagnosis. Ruptured USL should be treated with combination chemotherapy including CDDP and ADR, as well as with curative resection.

Postoperative fecoflowmetric analysis in patients with anorectal malformation.

Because conventional methods of evaluating anorectal function do not necessarily provide good correlations between investigative results and symptoms in patients who have undergone surgery for an anorectal malformation (ARM), we recently introduced feco-flowmetry (FFM) to simulate natural anorectal evacuation. The purpose of this study was to embody significant parameters to elucidate the dynamics of anorectal activity on FFM. The parameters of FFM were compared with those of manometry and Kelly's clinical score (KCS) in 24 patients who underwent surgery for an ARM. There were three fecoflow patterns, namely, block (B) type, segmental (S) type, and flat (F) type. The B-type or S-type patterns were seen in patients classified as "clinically good." There were close relationships between the fecoflow pattern and both the operative procedure and the KCS (P = 0.01 and 0.001, respectively). Maximum fecal stream flow rate (Fmax) precisely reflected the tolerance rate of intended normal saline solution in the colorectum (TR), the evacuative rate (ER), and KCS. Fmax > 45 ml/s or TR > 70% or ER > 50% was statistically regarded as the borderline of fecal continence. Thus, the fecoflow pattern might reflect the motor activity of the pelvic floor muscle. FFM provided quantiative and qualitative evaluations concerning anorectal motor activity in patients who had undergone surgery for an ARM.

Effects of intestinal muscular wrapping on remnant intestinal motility after massive small bowel resection in conscious canines.

We searched the effect of the muscular valve on the management of short bowel syndrome. The motility of the remnant intestine with a special muscular valve after 80% massive distal small bowel resection (MSBR) was evaluated in conscious dogs. The valve (muscular ring) was made by the autointestinal muscle layer holding vascular pedicle. Interdigestive and postprandial bowel motility using bipolar electrodes and/or contractile strain gauge force transducers 2-4 weeks after the surgery, and data of this group (Group I) were compared to the motility in dogs after MSBR without valve construction (Group II) and in controls (Control). Results; Fasting duodenal migrating myoelectric (or motor) complexes (MMCs) in Group I occurred at longer intervals than in Control and almost similarly to those in Group II. MMCs arising from the duodenum were often interrupted before the jejunum above the valve and the anastomosis. The velocity of duodenal MMC propagation was slowed in every intestinal segment including that from the duodenum to the proximal jejunum, and to the jejunum above the anastomosis. Transit time in MSBR group (I and II) from the duodenum to the terminal ileum was extremely shorter than in Control, but there were no differences between in Groups I and II. The duration of the postprandial period without duodenal MMCs in Group I was significantly prolonged than in Control, but was shorter than that in Group II. The muscular valve was frequently activated, and the jejunum covered with the valve was contracted frequently which synchronized with the valve activity. It seemed the valve worked as sphincter. However, intestinal obstruction was not occurred through the jejunum covered by the valve. In conclusion, changes in gut motility after MSBR with the valve construction compensate for the shortened intestine and maintain the bowel content earlier postoperatively in comparison with the MSBR alone, and also contribute to the adaptive increase in the remnant intestinal absorption.

The Vienna classification of gastrointestinal epithelial neoplasia.

BACKGROUND: Use of the conventional Western and Japanese classification systems of gastrointestinal epithelial neoplasia results in large differences among pathologists in the diagnosis of oesophageal, gastric, and colorectal neoplastic lesions. AIM: To develop common worldwide terminology for gastrointestinal epithelial neoplasia. METHODS: Thirty one pathologists from 12 countries reviewed 35 gastric, 20 colorectal, and 21 oesophageal biopsy and resection specimens. The extent of diagnostic agreement between those with Western and Japanese viewpoints was assessed by kappa statistics. The pathologists met in Vienna to discuss the results and to develop a new consensus terminology. RESULTS: The large differences between the conventional Western and Japanese diagnoses were confirmed (percentage of specimens for which there was agreement and kappa values: 37% and 0.16 for gastric; 45% and 0.27 for colorectal; and 14% and 0.01 for oesophageal lesions). There was much better agreement among pathologists (71% and 0.55 for gastric; 65% and 0.47 for colorectal; and 62% and 0.31 for oesophageal lesions) when the original assessments of the specimens were regrouped into the categories of the proposed Vienna classification of gastrointestinal epithelial neoplasia: (1) negative for neoplasia/dysplasia, (2) indefinite for neoplasia/dysplasia, (3) non-invasive low grade neoplasia (low grade adenoma/dysplasia), (4) non-invasive high grade neoplasia (high grade adenoma/dysplasia, non-invasive carcinoma and suspicion of invasive carcinoma), and (5) invasive neoplasia (intramucosal carcinoma, submucosal carcinoma or beyond). CONCLUSION: The differences between Western and Japanese pathologists in the diagnostic classification of gastrointestinal epithelial neoplastic lesions can be resolved largely by adopting the proposed terminology, which is based on cytological and architectural severity and invasion status.

The role of endoscopic retrograde cholangiopancreatography in infants with cholestasis.

BACKGROUND/PURPOSE: Endoscopic retrograde cholangiopancreatography (ERCP) was assessed in the diagnosis of cholestatic liver disease in infants. METHODS: ERCP was performed in 50 infants who had prolonged cholestasis. Their ages ranged from 25 to 274 days (mean, 69 days), and their weight ranged from 2.6 to 6.7 kg (mean, 4.7 kg). Incomplete visualization of the biliary tree or visualization of only the pancreatic duct was followed by exploratory laparotomy. In those in whom the biliary tree was visualized completely, the caliber of the bile duct was compared with that of the pancreatic duct. RESULTS: ERCP was completed in 43 patients (success rate, 86%) without complications. In the 7 patients in whom ERCP failed, 6 had biliary atresia (BA) diagnosed by exploratory laparotomy. The other patient had congenital biliary dilatation (CBD). In 29 of the 43 patients, the biliary tree was seen partially or only the pancreatic duct was visualized. These patients had BA diagnosed by laparotomy. Complete visualization of the biliary tree was obtained in 14 patients. Of these, 9 had neonatal hepatitis (NH), 2 had a paucity of intrahepatic bile ducts (PIBLD), and 3 had CBD. In all of the patients with NH, cholestasis improved spontaneously. The 2 patients with PIBLD had biopsy-proven disease. The caliber of the bile duct was larger than that of the pancreatic duct in NH. This relationship was not observed in PIBLD. CONCLUSIONS: ERCP is safe in infants. It is useful in the diagnosis of prolonged cholestasis.

A case of Currarino triad with familial sacral bony deformities.

We describe a male patient presenting with Currarino triad: a recto-urethral fistula, sacral bony deformity, and a presacral teratoma. Clinical screening of his family revealed three additional cases with incomplete forms of this association. Cytogenetic findings in the patient and his mother were normal. This case suggests that the occurrence of an anorectal malformation together with a sacral bony deformity should raise a physician's index of suspicion for associated presacral tumors, and that screening of the patient's family members with sacral radiographs is necessary.

Effects of trimebutine on intestinal motility after massive small bowel resection.

Effects of trimebutine maleate (TM) on intestinal motility in short bowel syndrome (SBS) were studied in conscious canines in both acute and chronic phases following 80% massive distal small bowel resection (MSBR). TM was administered orally to beagles with MSBR or as controls in the postprandial and fasting states, and given simultaneously with meals. Intestinal motility was measured using bipolar electrodes for approximately 1 month after the electrodes were implanted in each beagle and the data compared between treatment groups. When TM was given with meals, the postprandial period without duodenal migrating myoelectric (or motor) complexes (MMCs) was shorter than in those given meals only. When TM was given in the postprandial state in short bowel beagles, the initial duodenal MMCs occurred earlier, i.e. the postprandial period was shorter. Diarrhea did not occur in these beagles. When TM was given in the fasting state, duodenal MMCs occurred and propagated to the distal intestine. In conclusion, oral TM administration can produce a more appropriate intestinal condition for the next food intake and make enteral nutrition possible even in the acute phase after MSBR. Such feeding can be carried out without overloading gut function as a result of the modulation of gastrointestinal motility by TM.

Isopower mapping of electrogastrograms in short-bowel syndrome.

Methods for making topographic or isopower electrogastrographic (EGG) maps and for obtaining maximum power foci (MPFs) by means of 27-channel EGG recordings are briefly described. The methods are applied to short-bowel syndrome (SBS). The gastro-intestinal tract is traced by videofluorograms after X-ray-opaque barium has been ingested. The MPFs are generally located on the trace of the gastro-intestinal tract. The gastric area is occupied by 3 cycles x min(-1) (up to 28% of total MPFs) and 6 cycles x min(-1) MPFs (26%). The trace of the small intestine is occupied mainly by 8 cycles x min(-1) (39%) and 10 cycles x min(-1) (43%) MPFs. The trace of the colon is occupied almost evenly by all five spectral frequency groups, that is, by 1 (58%), 3 (53%), 6 (48%), 8 (57%) and 10 cycles x min(-1) (42%). Most interestingly, the numbers of 8-10 cycles x min(-1) power foci found on EGG maps (including MPFs and relatively higher power foci) are proportional to the remaining length at operation and to the length of the remaining small intestine in the long-term, over 6 years postoperatively. It is therefore possible to follow approximately the trace of the gastero-intestinal tract.

Treatment of childhood renal cell carcinoma with lymph node metastasis: two cases and a review of literature.

Standard treatment for renal cell carcinoma (RCC) is radical nephrectomy with lymph node dissection. Stages I and II have encouraging prognoses, but Stage III with regional lymph node metastasis can be unfavorable. Adjuvant therapy for pediatric patients with advanced RCC with lymph node involvement or metastatic lesion has not been defined. Advanced pediatric RCC is reported in two patients (boys, aged 6 and 9 years: Stage IIIs, Robson; Stage III and IV, pTNM classification) treated by nephrectomy and lymph node dissection followed by postoperative interferon-alpha (IFN), that can be used as an adjuvant therapy with side effects such as fever, bone marrow suppression, or decreased liver function. One is doing well for 7 years, another is suffered from lung metastases at 3 years after surgery. Although immunotherapy is expected to improve survival in pediatric patients with advanced RCC, surgical resection of renal and metastatic tumors remains the standard treatment.

Multiple K-ras mutations in hyperplasia and carcinoma in cases of human pancreatic carcinoma.

Mucous cell hyperplasia (MCH) has been considered an important precursor of pancreatic ductal carcinoma based on histological and molecular research, although various K-ras mutations rates are seen among cases with pancreatic carcinoma, chronic pancreatitis and normal pancreas, with a wide range of histological characters. To investigate the premalignant potential of MCH and the multicentricity of pancreatic carcinoma, we analyzed K-ras mutation at codon 12 in carcinoma foci of 82 cases of surgically-resected pancreatic carcinoma [67 solid-type carcinomas (SCs) and 15 ductectatic-type carcinomas (DCs)], as well as in both MCH and carcinoma foci in 42 cases (30 SCs and 12 DCs), using an enriched polymerase chain reaction (PCR)-enzyme linked mini-sequence assay (ELMA). K-ras mutation was recognized in 85% (57/67) of SCs and 73% (11/15) of DCs, and multiple K-ras mutations in 12% (8/67) of SCs and in 20% (3/15) of DCs. Multiple K-ras mutations were also recognized in MCHs in 47% (14/30) of SCs and in 42% (5/12) of DCs. Moreover, the same sequence at K-ras codon 12 in MCH and carcinoma was identified in 76% (32/42) of carcinoma cases and it was more frequently recognized in hyperplasias with histological atypia (51%, 37 of 72 foci) than those without atypia (24%, 16 of 68 foci) (P<0.0007). These results further support the idea of multicentric carcinogenesis and premalignant potential of atypical hyperplasia in the human pancreas, although about half of the hyperplasias around carcinomas were not thought to be direct precursors.

Long-term survivors of advanced neuroblastoma with MYCN amplification: A report of 19 patients surviving disease-free for more than 66 months.

PURPOSE: According to initial reports, stage 4 neuroblastoma patients with amplification of the MYCN proto-oncogene developed progressive disease within 8 months. The prognosis for such patients, however, should now be reevaluated in light of recent results achieved with up-to-date combination chemotherapy. PATIENTS AND METHODS: Patients with stage 3, 4, and 4S neuroblastoma and more than 10 copies of MYCN received induction chemotherapy, which from January 1985 to February 1991 consisted of regimen A(1 )(cyclophosphamide 1,200 mg/m(2) on day 1, vincristine 1.5 mg/m(2) on day 1, pirarubicin 40 mg/m(2) on day 3, and cisplatin 90 mg/m(2) on day 5) and from March 1991 to September 1993 consisted of regimen A(3 )(cyclophosphamide 1,200 mg/m(2) on days 1 and 2, pirarubicin 40 mg/m(2) on day 3, etoposide 100 mg/m(2) on days 1 through 5, and continuous infusion cisplatin 25 mg/m(2) on days 1 through 5). Most of these patients underwent radical surgery to remove the original tumor and local metastases, irradiation, and supralethal preconditioning regimens, followed by blood stem-cell transplantation (SCT). Data on the patients were collected in December 1998, and the factors contributing to disease-free survival were analyzed. RESULTS: During the study period, 66 patients with more than 10 copies of MYCN were treated. Five of nine patients with stage 3 disease, 13 of 55 with stage 4, and one of two with stage 4S survived for at least 66 months. It is interesting that all but one patient who survived for more than 66 months underwent SCT, in contrast with only five of 45 patients who died. CONCLUSION: Not all patients with advanced neuroblastoma who have more than 10 copies of MYCN will die. The requisites for survival in such patients seem to be intensive induction chemotherapy, effective surgery, irradiation, and the use of SCT.

A study of therapy for pediatric hepatoblastoma: prevention and treatment of pulmonary metastasis.

Our results of treatment for pediatric hepatoblastoma are presented with special emphasis on pulmonary metastasis. The pulmonary metastasis rate of hepatoblastoma was 44% (11/25). In 19 patients with resected hepatoblastomas, the 5-year survival rate without pulmonary metastasis was 90% (9/10); while with pulmonary metastasis it was 22% (2/9). Six patients with unresected hepatoblastomas all died within 4 months regardless of chemotherapy and/or metastasis. To improve survival in patients with hepatoblastoma, preoperative or postoperative chemotherapy was thought to be essential for tumors extending over 2 hepatic segments and having predictable factors for pulmonary metastasis (large size or histological evidence of capsular invasion). A long-term multidisciplinary approach including hepatic lobectomy, current multiagent chemotherapy (including CDDP, THP-ADR), and partial pulmonary resection for localized lung areas with metastases would ultimately be needed.

Sacrococcygeal teratoma: a series of 19 cases with long-term follow-up.

A series of 19 cases of sacrococcygeal teratoma (SCT) with follow-up of 5 to 25 years is presented. Twelve patients were neonates, age 0 to 26 days (5 immature teratomas and 7 mature teratomas, representing 3, 6, 2, and 1 Altman's type I, II, III, and IV tumors, respectively), four were infants, age 1 to 6 months (all mature teratomas, representing 1, 1, and 2 Altman's type I, II, and IV tumors), and 3 were children, age 1 to 4 years (all malignant teratomas, all Altman's IV tumors). Eight babies were delivered by elective caesarean section (CS). Though the mean gestational age at CS was 34.3 weeks in our series, we now believe that CS often must be performed earlier, depending on a tumor size or fetal condition. Eleven neonates and 4 infants were operated upon using a sacral approach in 10 and an abdominosacral approach in 5, and all survived. However, 4 patients had neurogenic bladder and were treated by urinary catheterization or vesicostomy for 2 to 5 years after surgery. Postoperative urogenital sequelae are seen in patients with a large tumor, urethral compression, urinary retention, or edema of the lower body. Malignant tumors usually had metastasized by the time of diagnosis, but the prognosis for outcome has been improved following surgery and combination chemotherapy.

Treatment results of advanced neuroblastoma with the first Japanese study group protocol. Study Group of Japan for Treatment of Advanced Neuroblastoma.

PURPOSE: To elucidate the efficacy of intensive induction and consolidation chemotherapy regimens (Study Group of Japan for Advanced Neuroblastoma [JANB] 85) for patients with advanced neuroblastoma aged 1 year or older. PATIENT AND METHODS: One hundred fifty-seven patients with newly diagnosed advanced neuroblastoma were entered into this study between January 1985 and December 1990. Eligible patients were 12 months old or older with stage III or IV disease. The patients first received six cyclic courses of intensive induction chemotherapy (designated regimen A1) consisting of cyclophosphamide (1,200 mg/m2), vincristine (1.5 mg/m2), tetrahydro-pyranyl Adriamycin (pirarubicin; 40 mg/m2), and cisplatin (90 mg/m2). The patients were further treated with three different consolidation protocols: 3-[(4-amino-2-methyl-5-pyrimidinyl)methyl]-1-(2-chloroethyl)-1-nitrosour ea, dacarbazine, and bone marrow transplantation. RESULTS: Overall survival rates for patients with stage III disease without reference to the consolidation protocols were 80.8%, 76.9%, and 66.3% at 2, 5, and 10 years, respectively. The overall survival rates for patients with stage IV disease were 58.8%, 34.4%, and 28.9% at 2, 5, and 10 years, respectively. There were no statistically significant differences between the three consolidation treatment groups. Patients who did not achieve complete remission (CR) with induction chemotherapy and surgery all died, suggesting that CR is essential for the cure of advanced neuroblastoma. The overall 5-year survival rate of the 24 patients with N-myc amplified stage III and IV disease was 33.3%, and the longest survival time of a relapse-free patient was 103 months. CONCLUSION: The intensive induction chemotherapy regimen used in this study may be of significant value in increasing the CR rate and survival for patients with N-myc amplified and nonamplified advanced neuroblastoma.

Problems arising from eastern and western classification systems for gastrointestinal dysplasia and carcinoma: are they resolvable?

It has become increasingly apparent that the Japanese and Western systems of classifying dysplasia and carcinoma in the gastrointestinal tract are not the same. The implication of these differences is that in an article in a Western journal on gastrointestinal 'cancer' originating from Japan, it is often impossible to repeat the study to confirm or refute it, because of these differences in definitions. Although there is no reason why it should not be just as much of a problem if Western research is published in Japanese journals, this is currently not a major problem. The terminological differences do not mean that one is right and one wrong; they are simply different. We had an opportunity to look at these differences in detail in a series of 100 gastric biopsies. In this review, we outline these differences and indicate the similarities, differences and problems between these two systems. 'Carcinoma' is diagnosed in Japan by virtue of its structural and cytological features, but by invasion in the Western system. Adenoma does not mean a dysplastic lesion in the Japanese system (although it can) but in most cases is similar to low-grade dysplasia irrespective of the macroscopic/endoscopic appearances (hence flat and depressed adenomas in the Japanese system); however, most examples of high-grade dysplasia in the Western system, as well as some low-grade dysplasia, becomes 'cancer' in the Japanese system. Conversely, both have conceptual areas that are useful in the other's systems. Because we were ultimately able to use each other's systems, we were able to show where these systems come together and diverge. These provide hope that a common classification encompassing both systems that allows communication between both systems is feasible.