Palliation of proximal gastric cancer

552 Letters to the Editor tosis after splenectomy and stasis of blood in the stump of the splenic vein appear to predispose to PVT.2 In Dr Atweh and associates’ patient, underlying diseases including malignancy, coagulation disorder, and thrombocytosis were not detected. At present it should not be concluded that splenectomy itself leads to PVT, because more evidence is required. It is still a controversial issue. Prospective studies are needed to determine the incidence, diagnosis, and treatment of PVT for patients who undergo splenecetomy, as Dr Atweh comments. Considering previously mentioned symptoms and associated conditions like abnormal coagulation, cirrhosis, or trauma, careful followup after splenectomy using Doppler ultrasonography may play an important role in the early detection and appropriate treatment for PVT. REFERENCES 1. Hanazaki K, Kajikawa S, Adachi W, Amano J. Portal vein thrombosis may be a fatal complication after synchronous splenectomy in patients with hepatocellular carcinoma and hypersplenism. J Am Coll Surg 2000;191:341–342. 2. van Riet M, Burger JWA, van Muiswinkel JM, et al. Diagnosis and treatment of portal vein thrombosis following splenectomy. Br J Surg 2000;87:1229–1233. 3. Hassan AM, Al-Fallouji MA, Ouf TI, Saad R. Portal vein thrombosis following splenectomy. Br J Surg 2000;87:362–373. 4. Rattner DW, Ellman L, Warshaw AL. Portal vein thrombosis after elective splenectomy. An underappreciated, potentially lethal syndrome. Arch Surg 1993;128:565–569. Palliation of Proximal Gastric Cancer Nicola Di Lorenzo, MD, PhD, FACS, Giorgio Coscarella, MD, Francesca Lirosi, MS, Achille Gaspari, MD, FACS Rome, Italy We were very interested in the article by Farrokh Saidi and associates, “A new approach to the palliation of advanced proximal gastric cancer.”1 The original technique of reconstruction proposed is very interesting for esophageal cancer, and the functional results, as previously described by Dr Saidi in 1988 and 1991,2 seem to be comparable (if not better) than those of transhiatal esophagectomy. Nevertheless, when applying the same technique for J Am Coll Surg palliation of advanced cancer of the proximal stomach and cardia, some considerations should be pointed out. The authors seem to consider their operation as an alternative to esophagogastrectomy. In our opinion, surgeons should remember that this is not actually the most common resection performed for palliation of proximal gastric cancer. This is especially true when facing patients with “serosal extension, distal peritoneal implants, regional nodes involvement, moderate ascites, and . . . liver metastases.” Most of these patients need a total D1 gastrectomy, but a Roux-en-Y esophagojejunostomy can usually be performed through the abdomen, freeing the esophagus trashiatally. It has been also demonstrated3 that reconstruction of duodenal transit is not necessary, but, if required, it can be realized with a jejunal segment, according to Longmire-Mouchet technique. The authors propose the colonic pull-through in the esophageal remnant, which resembles the Soave operation for Hirschprung’s disease. In this procedure, the muscular tunnel is very short; in Dr Saidi’s operation, the colon is forced for at least 25 cm in the esophageal muscular tube, although this was previously manually stretched, and the blood supply to the bowel could be compromised. It is well known that any intestinal segment, when brought to the neck in the standard fashion, can suffer venous engorgement, so some authors4 have suggested a second anastomosis in the neck to increase blood drainage. This problem could be worsened having a narrower space. The same question was proposed in the comments to the cited articles, but no definitive answer was given. Emphasizing the swallowing effect of the esophageal tube, Dr Saidi reported his excellent clinical results, but said that no manometric study had been done. This investigation, together with perfusion studies, could be very helpful in assessing the physiology of the transposed colon. Finally, we were very impressed by the in-hospital mortality rate (10%) in a selected group of patients with advanced neoplasms (70% of them were TNM Stage IV). As stated in an interesting survey by Gordon and associates,5 cumulative in-hospital mortality rate for total gastrectomy in Maryland, between 1990 and 1997, was 15.5% (range 7.5%–22.9%); these data included all patients resected for all cancer location, at all TNM stages. Considering that, although not specified, reconstruction is usually performed by esophagojejunostomy Roux-en-Y, we could suppose that mucosal esophageal Vol. 192, No. 4, April 2001 stripping and colon interposition could not only improve postoperative nutritional problems, but also postoperative mortality. Most probably, we could state that this very interesting procedure, which is very effective in Dr Saidi’s experience, needs to be more extensively applied by other surgeons in a prospective, multicentric randomized trial to assess its real impact on treatment of advanced proximal gastric cancer. REFERENCES 1. Saidi F, Keshoofy M, Azizollah A, et al. A new approach to the palliation of advanced proximal gastric cancer. J Am Coll Surg 1999;189:259–268. 2. Saidi F, Azizollah A, et al. Endothoracic endoesophageal pullthrough. A new approach to cancer of the esophagus and proximal stomach. J Thorac Cardiovasc Surg 1991;102:43–50. 3. Basso N, Materia A, et al. Nutritional effects of total gastrectomy. A prospective randomized study of Roux-en-Y vs LongmireMouchet reconstruction. Ital J Surg Sci 1985;15:335–340. 4. Marukami M, Sugiyama A, Ikegami T, et al. Revascularization using the short gastric vessels of gastric tube after subtotal esophagectomy for intrathoracic esophageal carcinoma. J Am Coll Surg 2000;190:71–77. 5. Gordon T, Bowman H, Bass E, et al. Complex G.I. surgery: impact of provider experience on clinical and economic outcomes. J Am Coll Surg 1999;190:46–56. Reply Farrokh Saidi, MD, FACS Tehran, Iran As Drs Di Lorenzo and associates rightly point out, the Roux-en-Y esophagojejunostomy remains the standard manner of reconstruction after palliative total gastrectomy. This is understandable because surgeons are reluctant to enlarge the scope of the operation when survival of patients with advanced stomach cancer is short. But some of these posttotalgastrectomy patients live longer than anticipated. And when they become incapacitated by relentless weight loss and severe postprandial distress, the surgeon regrets having removed the stomach in the first place. As a gastric reservoir, the left colon is large enough to allow patients to eat comfortably full course meals, until Letters to the Editor 553 the inevitable terminal phase of their disease. The question is how to mobilize and use the left colon as a gastric substitute in the least traumatic manner. The endoesophageal route is the shortest way for transposing the colon to the neck. This manner of reconstruction after a palliative total gastrectomy may lower postoperative morbidity because it dispenses with the cumbersome esophagojejunostomy, especially if an attempt is made to fashion a jejunal loop. We start jejunal feeding on the first or second postoperative day, and this may lower postoperative morbidity. The manner of reconstruction has, of course, no bearing on length of survival. On the technical side, the likelihood of leaving small strips of esophageal mucosa behind, was worrisome at first. Nothing serious happened, however, even when the esophageal mucosal sleeve did not come out intact in one piece. The possibility of strangling the colon segment in the neck loomed more portentous. This fear also proved groundless. We have not observed any venous congestion of the colon once it has reached the neck. The one episode of total colon infarction was found to be from thrombosis beginning at the origin of the left colic artery. Transecting both strap muscles in the neck creates plenty of room in the neck. It is essential, also, to manually overcome the diffuse spasm of the esophageal muscle, the moment its mucosal layer has been stripped off. Clearly, as was indicated by Drs Di Lorenzo and associates, the proposed technique of reconstruction after total gastrectomy must earn its value in the hands of other surgeons. What happens to the esophageal muscular layer in the long run? We have had the opportunity to inspect this layer at 1 and 5 years after the operation. On both occasions the esophageal muscular tunnel wrapped around the transposed colon in the posterior mediastinum was nothing more than a thin film of tissue. The endoesophageal colon pull-through technique is no more time consuming than the standard esophageojejunostomy. But it must be performed in an unhurried manner and with exactitude. I have regretted at times, halfway through a transhiatal esophagectomy, not having started off with the endoesophageal pull-through, which can always be converted into the former.