NOTE- CRITICAL EVALUATION OF PROPHYLACTIC SPLENECTOMY IN TOTAL GASTRECTOMY FOR THE STOMACH CANCER Keizo SUGIMACHI,*2 Yoshifumi KODAMA, Ryunosuke KUMASHIRO, Takashi KANEMATSU, Shoichi NODA, and Kiyoshi INOKUCHI Second Department of Surgery, Kyushu University School of Medicine*1 In gastric cancer surgery, an extended radical operation is commonly performed, and in cases of total gastrectomy, there is a tendency to perform splenectomy at the same time. However, some surgeons have reservations regarding this treatment in connection with the possible preservation of the host resistance. The question arises, is it reasonable to accept, by simple analogy with prophylactic lymphadenectomy the concept of the prophylactic splenectomy? The present study was designed to cast light on this problem. In order to examine the value of prophylactic splenectomy in gastric cancer surgery, a comparable patient group was followed up, and it was found that the non-splenectomized group showed a significantly better late survival rate than the splenectomized group (P<0.05), the 4-year survival rate being 63% in the former group and 36% in the latter group. Although these results do not necessarily contraindicate combined splenectomy, it seems desirable to reappraise the value of prophylactic splenectomy in cases having no metastasis in the splenic hilar and adjacent lymph nodes. Key words: Prophylactic splenectomy - Total gastrectomy - Human stomach cancer In the context of total gastrectomy for gastric cancer, combined splenectomy has been advocated by a number of authors4,5,7) in order to enhance the radical nature of the operation, but other surgeons have expressed reservations in connection with the possible preservation of the host resistance.3) Between 1968 and 1972 we favored combined splenectomy in total gastrectomy for gastric cancer, but the late results in patients treated according to this policy were found to be less favorable than we had anticipated. We therefore came to consider that combined subsequently), most of the total gastrectomies were performed by the same surgeon (K. I.). This therefore provided a useful opportunity to reexamine the value of prophylactic splenectomy; the results are presented here. With regard to regional lymph node eradication in gastric cancer surgery, it has been established that prophylactic lymphadenectomy is very effective in improving the late results. The question then arises, is it reasonable, by simple analogy with prophy- *2 To whom communications should be addressed. Gann
PROPHYLACTIC SPLENECTOMY IN TOTAL GASTRECTOMY lactic lymphadenectomy, to adopt the concept of prophylactic splenectomy? The present report considers this problem. Materials and Methods In a 13-year period between 1964 and 1976 a total of 319 patients underwent total gastrectomy for gastric cancer in the Second Department of Surgery, Kyushu University School of Medicine. These patients were divided into splenectomized (201 patients) and non-splenectomized (118 patients) groups as shown in Table I. When the annual numbers of combined splenectomies in total gastrectomy patients are analyzed, it is apparent that splenectomy was performed both in curative and non-curative surgery during the first period until 1972, with a peak in 1967, but since then, combined splenectomy was carried out less frequently. This was due to a change in operative policy to restrict ourselves to therapeutic splenectomy in total gastrectomy, since it was apparent that the late results of total gastrectomy performed between 1965 and 1972 was less favorable than had been expected. Using these patients as study materials, we attempted to evaluate prophylactic splenectomy. For this purpose, the degree of cancer advancement (stage) should be comparable in the splenectomized and non-splenectomized groups. In order to meet this requirement as far as possible, we reselected only curative cases with gastric cancer localized in the proximal region of the stomach and at the same time without direct invasion to the adjacent organs. This yielded 47 splenectomized (group A) and 30 non-splenectomized (group B) patients (Table II). The distributions of prognostic serosal factor (ps positive and ps negative) and of lymph node metastases were almost the same in the two groups (Table III). However, because only group A contained three patients with positive metastases to the splenic hilar lymph nodes, analysis of the prognosis was performed separately with and without these three patients included. Further, almost all of these patients were operated upon by the same surgeon (K. I.), so variations of skill and procedures, which might have affected the surgical prognosis, could be neglected. With regard to adjuvant chemotherapy, Table IV shows that 42.6% of group A and 43.3% of group B patients received chemotherapy. In both groups the chemotherapeutic regimen included intraoperative mitomycin C and postoperative Futraful (1-(2- tetrahydrofuryl)-5-fluorouracil, Taiho Pharmaceutical Co., Tokyo) and/or mitomycin C administration. Thus, it is assumed that the variations in chemotherapeutic regimen would have had little differential effect on the long-term prognosis in the two groups. None of the 77 patients has been lost during the follow-up, and the survival rate is represented by cumulative survival rate. Terms such as prognostic serosal factor (ps), lymph node metastasis (n0, n1, n2) and others appearing in this report are those defined in the General Rules for the Gastric Cancer Study in Surgery.1) Table I. Patients With or Without Splenectomy Table II. Patients Selected for Study
K. SUGIMACHI, ET AL. Results First, the postoperative survival rate was compared with respect to prognostic serosal factor (ps). As illustrated in Fig. 1, with ps negative patients, the 5-year survival was 80% in both groups, A and B. Among ps positive patients, the survival rates from I to 5 years, respectively, were 78, 53, 42, 36 and 34% in group A and 91, 77, 68, 63 and 50% in group B patients. The difference in survival rate was significant (P<0.05) at the 4th postoperative year in favor of those with the spleen preserved. When the two patients in group A who had metastasis to the splenic hilar lymph nodes were excluded, and patients without splenic hilar metastasis were compared, it was found that the 4-year survival rate in group B was 63% as compared with 37% in group A. Thus the difference (P<0.06) still favored group B against group A (Fig. 2). Survival as analyzed in terms of lymph node metastasis was also considered: in patients with no lymph node metastasis (n0), the survival rates from 1 to 5 years, respectively, were 100, 78, 68, 66 and 59% in group A and 100, 100, 92, 92 and 83% in group B. Thus group B appears to have a better prognosis at the third and 4th years (P<0.08) (Fig. 3). However, in patients with regional lymph node metastasis (n1), the 5-year survival was 62% for group A and 52% for group B, the difference being statistically insignificant (Fig. 3). Among patients with intermediate lymph node positive (n2), all 14 patients belonging to group A had died within 4 years and all 6 patients of group B had done so within 5 years postoperatively. Discussion In gastric cancer surgery, an extended radical operation is being commonly performed,4,5,8) and often in cases of total gastrectomy or cardia resection for gastric cancer, splenectomy is carried out simultaneously. Jinnai,3) on the other hand, adopted a more cautious attitude, and in cases with- Gann
PROPHYLACTIC SPLENECTOMY IN TOTAL GASTRECTOMY Table IV. Number of Patients Receiving Cancer Chemotherapy Fig. 1. Survival rate with or without splenectomy Among prognostic serosal factor (ps) positive patients, the difference in survival rate is significant (P<0.05) at the 4th postoperative year in favor of those retaining the spleen. out direct invasion or lymph node metastasis to the splenic hilar region, he recommends eradication of the hilar lymph nodes after mobilization of the spleen, but not extirpation of the spleen. Some surgeons advocate extended surgery, but others are more reluctant. Thus, it is of prime importance to. know whether these therapeutic policies affect the remote postoperative results. Kanai4) studied the significance of combined resection of the pancreatic body and tail and the spleen, and he reported that 5-year survival after combined resection of the pancreas and spleen was 17.1%, superior to that of 10.4% in the simple total gastrectomy group. He further stated that when only curative surgical patients were compared, the 5-year survival rate in the former was better (30.4%) than in the latter group (16.7%). In his report, however, no mention was made of his policy regarding splenectomy; whether it was done therapeutically, prophylactically or both. Therefore, his results are of uncertain significance in discussing the value of prophylactic sple-
K. SUGIMACHI, ET AL. nectomy. In this respect, it is also unclear in Kanai's study whether the patients were correctly selected for a valid comparison. Orita et al.10) analyzed the 5-year survival rate in gastric cancer patients, and he stated Fig. 2. Survival rate of patients without splenic hilar metastasis (ps positive) The 4-year survival rate of the non-splenectomized group is superior to that of the splenectomized group (P<0.06). that in stage I and II groups, patients undergoing gastrectomy with combined splenectomy had a better 5-year survival rate (100% and 88.9%) than simple gastrectomy patients (87.1% and 77.1%), while with stage III and IV groups, they found no significant difference. It should be pointed out, however, that in their study very small numbers of splenectomized patients (stage I, two patients and stage II, nine patients) were involved and the differences were not very significant statistically (stage I, P<0.5; stage II, 0.5>P>0.1). On the other hand, Yoshino et al.11) analyzed the prognosis of advanced gastric cancer located in the proximal two-thirds of the stomach and compared 123 splenectomized and 125 non-splenectomized patients. They found that the 5-year survival rate was 47% in the former group and 61% in the latter, indicating that the prognosis was significantly improved in patients retaining the spleen. Consequently, they suggested that the value of prophylactic splenectomy in elevating the radical cure rate after gastric Fig. 3. Survival rate in terms of lymph node metastasis Patients with no lymph node metastasis appear to have a better prognosis in the non-splenectomized group than in the splenectomized group at the third and 4th year (P<0.08). However, in patients with lymph node metastasis the difference is statistically insignificant. Gann
PROPHYLACTIC SPLENECTOMY IN TOTAL GASTRECTOMY cancer surgery may not be very significant, and in some instances it was suggested that combined splenectomy may adversely affect the prognosis. In curative surgical cases with gastric cancer located in the proximal portion of the stomach and at the same time having comparable cancer staging, it was found that the group retaining the spleen had statistically better late results than the comparable, but splenectomized group. Thus, the usefulness of combined splenectomy in total gastrectomy is quite differently estimated in different reports. In patients with metastasis to the splenic hilar or adjacent lymph nodes, combined splenectomy seems to be superior. Therefore, appraisal of the value of prophylactic splenectomy should be based on an appropriate selection of study patient material. Viewed from this standpoint, few existing reports offer useful information on the significance of prophylactic splenectomy. The spleen has important immunological functions, including the production of lymphocytes. However, the effect of the spleen on experimental tumors is not consistent. It may have an inhibitory or stimulatory effect, depending on such variables as the type of the experimental tumor, time of inoculation, and the number of inoculated cells.6,9,10) Since splenectomy removes suppressor T-cells contained in the spleen, it can prolong the survival of a cancer-bearing organism,9) though it was also reported that splenectomized patients showed reduced resistance to pneumococcal infection for a prolonged period.2) Thus, the view that prophylactic splenectomy is readily acceptable by analogy with lymphadenectomy in gastric cancer surgery is clearly questionable. Reappraisal and elucidation of this problem will require many further studies, both experimental and clinical. (Received March 28, 1980/Accepted, July 5, 1980) REFERENCES