J Gastrointest Surg (2009) 13:2239–2244 DOI 10.1007/s11605-009-0971-y ORIGINAL ARTICLE Subtotal Gastrectomy as Treatment for Distal Multifocal Early Gastric Cancer Paolo Morgagni & Caterina Marfisi & Andrea Gardini & Daniele Marrelli & Luca Saragoni & Franco Roviello & Giovanni Vittimberga & Domenico Garcea & For the Italian Research Group for Gastric Cancer (I.R.G.G.C.) Received: 21 April 2009 / Accepted: 15 July 2009 / Published online: 12 August 2009 # 2009 The Society for Surgery of the Alimentary Tract Abstract Introduction Multifocal early gastric cancer (MEGC) is frequently observed and represents a serious risk when minimally invasive treatments are performed. Patients and Methods We present the experience of two Italian centers situated in a relatively high incidence area for gastric cancer. Out of a total of 791 surgical resections for EGC carried out in two Italian centers from 1976 to 2006, we identified 98 patients with multifocal EGC (12.3%). Two hundred and sixteen lesions were observed. Generally sited near the principal tumors, secondary lesions were, however, sometimes detected distally from the upper primary lesion. No secondary lesions were detected in the upper third when the principal lesion was sited at the lower third. Results Survival of MEGC patients was not significantly lower than that of patients with monofocal EGC. No cases of gastric remnant relapse were observed at a mean follow-up of 9 years (range 1–28) after subtotal gastrectomy. Discussion When EGCs are detected, the possibility of MEGC must always be investigated by endoscopy and chromoendoscopy. When a MEGC is found in the lower third of the stomach and chromoendoscopy of the upper third has been performed, subtotal gastrectomy can be considered as sufficient treatment. Keywords Early gastric cancer . Multifocal early gastric cancer . Subtotal gastrectomy . Surgical treatment P. Morgagni (*) : C. Marfisi : A. Gardini : G. Vittimberga : D. Garcea Department of General Surgery, Pierantoni-Morgagni Hospital, Via Forlanini 34, 47100 Forlì, Italy e-mail: p.morgagni@ausl.fo.it D. Marrelli : F. Roviello Surgical Oncology Unit, University of Siena, Policlinico Le Scotte, Via le Bracci, 53100 Siena, Italy L. Saragoni Pathology Unit, Morgagni-Pierantoni Hospital, Via Forlanini 34, 47100 Forlì, Italy Introduction Multifocality is a rare condition in advanced gastric cancer, but not unusual in early gastric cancer (EGC) lesions, with an incidence of about 10%. In 1957, Moertel presented the following criteria for the diagnosis of multifocal early gastric cancer (MEGC): each lesion is histopathologically malignant, and each one is separated from the others by a normal gastric wall; lesions are not the result of local extension of or metastasis from another gastric tumor.1 Moertel also affirmed that if the depth of invasion is the same in two or more lesions, the one extending over the greatest area should be regarded as the main lesion, with the others considered as accessories. Although more accurate endoscopy techniques, such as chromoendoscopy and magnified imaging, are frequently used to identify MEGC, this condition is often only diagnosed by a pathologist and, therefore, not before surgical treatment. For this reason, some surgeons submit EGC patients to total 2240 gastrectomy, thus, avoiding the risk of missing gastric remnant lesions. This approach is generally considered too aggressive, especially if we take into account comorbidities and quality of life in patients submitted to total gastrectomy. The aim of this retrospective study was to report the experience of two Italian surgical units situated in two hospitals in north-central Italy. In particular, clinical outcome after subtotal gastrectomy for multifocal EGC was evaluated and compared to that of patients with unifocal EGC. Patients and Methods From 1976 to 2006, 791 patients underwent resection for early gastric cancer in two Italian surgical units (Department of General Surgery, Morgagni-Pierantoni Hospital, Forlì and the Surgical Oncology Unit of Siena University). Ninety-eight of these patients (12.3%) had multifocal lesions according to Moertel’s criteria. All patients were classified according to tumor size, Japanese macroscopic type,2 Lauren’s histological type,3 and TNM classification.4 Patients with synchronous advanced gastric cancer or other tumors were excluded. Age, sex, histologic and macroscopic type, size, and site were compared to identify risk factors for multifocal EGC. Subtotal gastrectomy was performed for tumors located in the lower two thirds of the stomach, with removal of the greater and lesser omentum and gastrojejunal Billroth II reconstruction. Total gastrectomy was carried out for tumors located in the upper gastric third, with Roux-en-Y reconstruction. D1 lymph node dissection was generally performed in elderly or critical patients, while en bloc D2 lymphadenectomy, in accordance with Japanese Gastric Cancer Association recommendations, was preferred for all other patients.2 Although hand-sewn anastomosis was usually performed, we used staplers for esophagojejunal anastomosis and sometimes for duodenal remnants. Death from postoperative complications was considered an event if it occurred during hospitalization. No patients were submitted to adjuvant or neoadjuvant therapy. Followup ultrasonography and serum-marker evaluation were carried out every 6 months for the first 5 years. Endoscopic checkups were carried out annually after surgery for a period of 5 years given the incidence of gastric remnant carcinoma and the importance of evaluating for esophagitis or reflux, both treated pharmacologically. Once 5 years had passed, endoscopy was performed after a further 2 years and every 3 years thereafter. Survival times were measured from the date of surgery until death. Univariate analysis was performed by tracing Kaplan–Meier survival curves,5 and comparison of survival curves was based on the logrank test. Multivariate analysis J Gastrointest Surg (2009) 13:2239–2244 was carried out according to the logistic regression model for categorical variables. All p values were based on twosided testing (threshold value p =0.05) and statistical analysis was carried out using SPSS test (software package 13.0 version Chicago Inc.). Results Ninety-eight MEGC patients were operated on between 1976 and 2006, representing 12.3% of the 791 patients consecutively submitted to surgery for EGC in the two hospital departments. Patient characteristics are summarized in Table 1. The variable most at risk for multifocality in our patients was mucosal EGC<1 cm sited at the lower third, but only tumor site (OR 2.007 [95%CI: 1.15–3.49], p=0.014 ) and T infiltration (OR 1.747 [95%CI: 1.05–2.88] p=0.029) were considered as independent risk factors at multivariate analysis, the former being the most significant (Table 2). Conversely, we did not find any correlation between multifocality and sex, age >70 years, macroscopic type, histologic type, differentiation, or lymphatic diffusion. With regard to tumor size, we also considered 2 or 3 cm as cutoff values, but these did not prove to be statistically significant. Macroscopic type did not represent a risk factor even when polipoid (I, IIa) and ulcerated lesions (IIc, III) were considered together. We generally found two EGCs and only rarely three or more multifocal tumors; principal and secondary lesions were sited at the lower third of the stomach. In particular, of the 216 lesions detected, 82 patients had two lesions, 12 patients had three, and four had four. In four patients the principal lesion was sited at the upper third, in 21 at the middle third and in 73 at the lower third; two secondary lesions were detected at the upper third, 31 at the middle third, and 86 at the lower third (Fig. 1). Chromoendoscopy during endoscopy improved detection of MEGC, but as it has only become standard practice for EGC in the last few years, no definitive conclusions on differences before and after its application can be drawn. Twenty-nine patients presented diffuse gastric cancer and 69 intestinal histologic type. With respect to the known association between familial gastric cancer and multifocality, more frequent in younger adults with diffuse type, four patients from the former group were under 50 years of age, but none had a family history of the disease. Eighty-nine patients with distal multifocal EGC were submitted to subtotal gastrectomy with Billroth II reconstruction, while nine patients with upper third lesions underwent total gastrectomy with Roux-en-Y reconstruction. No patient with distal multifocal early gastric cancer was submitted to total gastrectomy if >2 cm of normal mucosa was observed from the resection line. As the study J Gastrointest Surg (2009) 13:2239–2244 Table 1 Clinicopathological factors of patients with and without multifocal early gastric cancer 2241 Unifocal (%) N=693 Sex Male Female 396 297 (57.1) (42.9) 62 36 (63.3) (36.7) 401 292 (57.9) (42.1) 48 50 (49) (51) 94 55 34 328 168 14 (13.5) (7.9) (4.9) (47.3) (24.3) (2.1) 12 6 3 50 26 1 (12.3) (6.1) (3.1) (51) (26.5) (1) 275 418 (39.7) (60.3) 25 73 (25.5) (74.5) 126 512 55 (18.2) (73.9) (7.9) 21 54 23 (21.4) (55.1) (23.5) 346 347 (49.9) (50.1) 62 36 (63.3) (36.7) 530 158 5 (76.5) (22.8) (0.7) 69 29 (70.4) (29.6) 221 190 272 10 (31.9) (27.4) (39.3) (1.4) 38 24 34 2 (38.8) (24.5) (34.7) (2) 598 95 (86.3) (13.7) 88 10 (89.8) (10.29 ns ns 0.001 0.049 0.002 ns ns ns was conducted over a relatively long period, and as D2 lymphadenectomy was still infrequent during the 1980s, only 24 D2 dissections were performed, with a mean number of 24.9 lymph nodes dissected. Taking into account all the patients, a mean of 17.7 lymph nodes were dissected, Table 2 Logistic regression analysis for variables associated with multifocality Explanatory variable Odds ratio 95% CI P value Site T1a/b Size 2.007 1.747 1.15–3.49 1.05–2.88 0.014 0.029 ns ns not significant P value ns Age (years) <70 >70 Macroscopic type 1 2a 2b 2c 3 Unknown Site Fundus/corpus Antrum Size ≤ 1 cm >1 cm Unknown T1 a b Histological type Intestinal Diffuse/mixed Unknown Histological grade 1 2 3 Unknown Lymphatic diffusion N0 N+ ns not significant Multifocal (%) N=98 18.1 in MEGCs (range 4–54) and 17.2 in single EGCs (range 3–62). The relatively low number of lymph nodes obtained stems from the fact that during the first few years of the study, sampling of fresh lymph node tissue was not performed by the surgeon station by station but rather “en bloc,” with only one piece of formalin-embedded tissue sent to the pathologist. MEGC patients had a lower, nonsignificant number of lymph node metastases than those with EGC, and 10 patients were classified as N1 (11.3% vs.16.2%, respectively; p=0.42). Five- and 10-year survival of MEGC patients was 92.6% and 89.8%, respectively, not significantly different from those with unifocal EGC (93.3% and 89.6%, respectively; p=0.41; Fig. 2). Patients submitted to subtotal gastrectomy had 5- and 10-year survival rates of 91.9% and 88%, respectively, while 2242 J Gastrointest Surg (2009) 13:2239–2244 Discussion 4 1 2 21 16 2 77 13 7 73 Figure 1 Site of secondary lesions related to primary EGC. The boxes show the number of primitive lesions, while the arrows indicate the sites and number of secondary lesions. all patients who underwent total gastrectomy for upper lesions showed 100% 10-year survival. Although this difference is important, it must be remembered that only nine total gastrectomies were performed. No gastric remnant recurrence was observed in MEGC patients after a median follow-up of 9 years (range 1–28), while 10 of the 574 patients with single EGC who underwent subtotal gastrectomies developed a new gastric cancer or had gastric remnant relapse (1.7%) during the same follow-up period. One of the 10 patients was diagnosed with a new, well-differentiated T2 gastric remnant adenocarcinoma (histologically similar to the first one radically resected) only 11 months after the first surgical treatment and was submitted to total gastrectomy. The other nine patients developed gastric remnant recurrence or new gastric cancer after a median follow-up of 7.4 years (range 3–21). 120 Survival Rate (%) 100 Multifocal EGC 80 Monofocal EGC 60 40 20 0 0 4 8 12 16 20 Years Figure 2 Long-term survival of patients with and without multifocal EGC. Multifocality is a condition described in 0.8–22% of EGC.6 Often diagnosed by the pathologist after surgical treatment rather than preoperatively by the endoscopist (about 35% of missed lesions are reported in Eastern studies7 and even more in Western series8), multifocality represents a risk for undertreatment. Generally observed in early lesions, multifocality is not very frequent in advanced gastric cancer. Kitamura, reporting a 7.4% incidence of MEGC but only 3.02% of multifocal advanced gastric cancer, tried to explain this difference by referring to the theory of collision cancer, which suggests that two different early synchronous lesions may fuse together after lateral and vertical growth, becoming a single advanced cancer.9 Similarly, in our experience, MEGC represents around 12% of EGCs and is rare in advanced stages. Some authors consider the large areas of intestinal metaplasia and dysplasia, frequently present in elderly patients, as a precursor of multifocality. Furthermore, an average age of 65 years is generally believed to be a risk factor for MEGC, 5–8 years higher than the age considered most at risk for monofocal lesions.10,11 The male sex, mucosal lesions, differentiated histologic type, elevated macroscopic type, tumor size <2 cm, and lower third site have been found to be the most important risk factors for MEGC. Literature data is, however, still somewhat contradictory. Takeshita found a higher incidence of depressed macroscopic type in his series of 61 MEGCs,7 while Huguier observed a high number of nondifferentiated histologic type EGCs in his relatively small patient population.12 Genetic factors such as germline mutations in the E-caderin gene have also been reported to be involved in a small number of signet ring cell diffuse histotype EGCs with multiple foci.13,14 Twenty-nine of our patients presented diffuse carcinoma, but as no hereditary cancers were hypothesized, genetic studies were not conducted. Our data identified mucosal EGC sited in the antrum as an independent risk factor for MEGC. Tumors <1 cm were considered a significant, albeit not independent, risk factor. Although MEGC was also rarely diagnosed preoperatively in our series (50%), the increasing use of chromoendoscopy and magnifying endoscopy gradually led to a higher number of endoscopic diagnoses during the study period. With regard to secondary lesions, two are generally observed, although four or five are not uncommon and as many as 43 synchronous lesions have also been reported.6,9,10,15,16 In our series, 12 patients presented three lesions and only four had four lesions. Secondary lesions are generally observed in the same gastric third as the first tumor or in a lower third and are only rarely located in the proximal third. Kodama noticed that when the major lesion was located in the upper gastric third, the others were sited in the J Gastrointest Surg (2009) 13:2239–2244 lower one, whereas when the first lesion was in the lower third, the secondary lesions were generally in the same third.6 In our case series, we likewise observed that secondary lesions were usually located near the principal one and were generally observed in the lower gastric third. In particular, 79.8% of the lesions were sited in the same area, 18.2% in an area near the principal lesion, and only 2% of main upper lesions showed secondary antral lesions (Fig. 1). Secondary lesions, when differentiated, generally present the same histologic characteristics. Takeshita reported that in his case series, 61% of well-differentiated main lesions had well-differentiated secondary lesions, and only 13% of undifferentiated lesions showed the same histologic characteristics as the accessory lesions.7 In the literature, treatment of multifocal early gastric cancer does not differ from that of monofocal lesions, and endoscopic mucosal resection or subtotal gastrectomy are recommended if indications for each tumor are satisfied.7,16 We agree with this approach and currently perform, when possible, endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD). When criteria for EMR or ESD are not satisfied and distal MEGC is detected, we generally carry out subtotal gastrectomy for the following reasons: secondary lesions are generally sited in the lower third near the main lesion; secondary EGC sited in the upper gastric third are rare; randomized and retrospective studies have shown that patients submitted to subtotal gastrectomy have lower morbidity and mortality, a better quality of life and similar survival with respect to those who undergo total gastrectomy, in both early and advanced antral disease.17,18 The high incidence of misdiagnosis in MEGC makes it important to perform chromoendoscopy, a simple and inexpensive technique that allows small, hidden lesions to be detected in all EGC patients. Such a strategy could help reduce the incidence of missed lesions and gastric remnant carcinoma, reported in 1.1–2% of EGC patients submitted to subtotal gastrectomy.10,19 The incidence of gastric remnant cancer in MEGC is no higher than that of unifocal EGC. Kodera and coworkers, in a series of 2061 EGC, observed a 2% incidence of gastric remnant cancer after a follow-up of 16 years, and only one (0.8%) patient in this subgroup had been diagnosed with MEGC.10 Although we did not observe any relapses or metachronous gastric remnant cancers in patients operated on for MEGC, 10 (1.7%) of the 574 patients who underwent subtotal gastrectomies for monofocal EGC had a gastric remnant cancer incidence similar to that reported in the literature. One of these was a patient treated for a single EGC of the lower third who was diagnosed with a second lesion in the upper third only 11 months after subtotal gastrectomy. In our opinion, this was a case of a missed synchronous lesion. The other relapses were detected over a 2243 period ranging from 3.5 to 16 years after subtotal gastrectomy. Five- and 10-year survival rates in patients with distal MEGC treated with subtotal gastrectomy do not differ greatly from those with single EGC,10 and MEGC is not considered as a prognostic factor.20,21 The results from the present study would seem to confirm this observation, with survival curves similar for the two groups. Our findings also indirectly point to the effectiveness of subtotal gastrectomy in distal MEGC. Although patients with upper lesions treated with total gastrectomy presented better 5- and 10-year survival, few patients were treated, and data are not significant. However, we must underline that this is a retrospective study, and a control group (distal MEGC treated with total gastrectomy) is lacking. In conclusion, MEGC is a rare occurrence, generally involving the lower gastric third and usually not detected by preoperative endoscopy. It is more frequent in patients with small differentiated mucosal EGC. The prognosis of MEGC is good, and treatment does not differ from that of monofocal EGC. For this reason, subtotal gastrectomy could be considered adequate treatment for lower and middle third MEGC. However, secondary lesions in the upper gastric third may be present, making preoperative chromoendoscopy and postoperative endoscopic follow-up strongly indicated. Conflict of interest The authors have no conflicts of interest. References 1. Moertel CG, Bargen JA, Soule EH. Multiple gastric cancer: review of the literature and study of 42 cases. Gastroenterology 1957;32:1095– 103. 2. Japanese Gastric Cancer Association. The new Japanese Classification of Gastric Carcinoma. 2nd Engl. ed. Gastric Cancer 1998;1:10– 24. 3. Lauren P. The two histological main types of gastric carcinoma: diffuse and so-called intestinal-type carcinoma. Acta Pathol Microbiol Scand 1965;64:31–49. 4. Sobin LH, Wittekind CH. TNM classification. In International Union Against Cancer (UICC). TNM classification of malignant tumours, 5th ed. New York: Wiley-Liss, 1997, pp 59–62. 5. Kaplan EL, Meier P. Non parametric estimation from incomplete observation. J Am Stat Assoc 1958;53:457–481. 6. Kodama M, Tur GE, Shiozawa N, Koyama K. Clinicopathological features of multiple primary gastric carcinoma. J Surg Oncol 1996;62:57–61. 7. Takeshita K, Tani M, Honda T, Saeki I, Kando F, Saito N, Endo M. Treatment of primary multiple early gastric cancer: from the viewpoint of clinicopathologic features. World J Surg 1997;21:832– 836. 8. Borie F, Plaisant N, Millat B, Hay JM, Fagniez PL, De Saxce B. French Associations for Surgical. Research Treatment and prognosis of early multiple gastric cancer. Eur J Surg Oncol 2003;29:511–514. 9. Kitamura K, Yamaguchi T, Okamoto K, Otsuji E, Taniguchi H, Hagiwara A, Sawai K, Takahashi T. Clinicopathologic features of 2244 10. 11. 12. 13. 14. 15. 16. synchronous multifocal early gastric cancers. Anticancer Res 1997;17:643–646. Kodera Y, Yamamura Y, Okamoto K, Otsuji E, Taniguchi H, Hagiwara A, Sawai K, Takahashi T. Incidence, diagnosis and significance of multiple gastric cancer. Br J Surg 1995;82:1540–1543. Noguchi Y, Ohta H, Takagi K, Takahashi T, Ohashi I, Kuno K, Kajitani T, Kato Y. Synchronous multiple early gastric carcinoma: a study of 178 cases. World J Surg 1985;9:786–793. Huguier M, Ferro L, Barrier A. Early gastric carcinoma: spread and mulcentricity. Gastric Cancer 2002;5:125–129. Pedrazzani C, Corso G, Marrelli D Roviello F. E-caderin and hereditary diffuse gastric cancer. Surgery 2007;142:645–657. Charlton A, Blair V, Shaw D, Parry S, Guilford P, Martin IG. Hereditary diffuse gastric cancer: predominance of multiple foci of signet ring cell carcinoma in distal stomach and transitional zone. Gut 2004;53:814–820. Baba Y, Ishikawa S, Ikeda K, Honda S, Miyanari N, Iyama K, Baba H. A patient with 43 synchronous early gastric carcinomas with a krukenberg tumor and pericardial metastasis. Gastric Cancer 2007;10:135–139. Otsuji E, Kuriu Y, Ichikawa D, Okamoto K, Hagiwara A, Yamagishi H. Clinicopathologic characteristics and prognosis of synchronous multifocal gastric carcinomas. Am J Surg 2005;189:116–119. J Gastrointest Surg (2009) 13:2239–2244 17. Bozzetti F, Marubini E, Bonfanti G, Miceli R, Piano C, Gennari L. Subtotal versus total gastrectomy for gastric cancer. Five-year survival rates in a multicenter randomized Italian trial. Ann Surg 1999;2:170–178. 18. De Manzoni G, Verlato G, Rovello F, di Leo A, Marrelli D, Morgagni P, Pasini F, Paragoni L, Tomezzoli A. Italian Research Group for Gastric Cancer. Subtotal versus total gastrectomy for T3 adenocarcinoma of the antrum. Gastric Cancer 2004;7:183–184. 19. Sano T, Sasako M, Kinoshita T, Maruyama K. Recurrence of early gastric cancer. Follow-up of 1475 patients and review of the Japanese literature. Cancer 1993;72:3174–3178. 20. Folli S, Morgagni P, Roviello F, De Manzoni G, Marrelli D, Saragoni L, Di Leo A, Gaudio M, Nanni O, Carli A, Cordiano C, Dell'Amore D, Vio A. Italian Research Group for Gastric Cancer (IRGGC). Risk factors for lymph node metastases and their prognostic significance in early gastric cancer for the Italian Research Group for Gastric Cancer. Jpn J Clin Oncol 2001;31:495– 499. 21. Roviello F, Rossi S, Marrelli D, Pedrazzani C, Corso G, Vindigni C, Morgagni P, Saragoni L, de Manzoni G, Tomezzoli A. Number of lymph node metastases and its prognostic significance in EGC: a multicenter Italian study. J Surg Oncol 2006;15:275–280.