You are here

Neoadjuvant Chemotherapy

Because of the frequent inability of adjuvant postoperative systemic therapy to prolong survival in surgically managed gastric cancer, several investigators have pursued the approach of neoadjuvant (preoperative) chemotherapy in an attempt to increase resectability and improve survival. These studies involve a mix of patients including those determined surgically or clinically unresectable for cure, those with “locally advanced” disease (as defined by the study authors), and those with clinically operable lesions. Some patients were staged clinically by a variety of methods, making it difficult to know which patients were truly resectable before neoadjuvant treatment.

Unresectable Disease


Six different trials have assessed the potential benefit of preoperative chemotherapy for patients with initially “unresectable” stomach cancer. 133,220-225 Each of the trials was small in size. The seven trials combined included 155 patients and demonstrated that preoperative chemotherapy in this subset of patients was feasible and resulted in clinical response rates of 30% to 68%. Curative resections were possible in as few as 8% or as many as 73% of patients. This wide range of resectability probably reflects patient selection rather than superiority of any one regimen. Pathological complete responses (CRs) were uncommon except for the Wilke trial (pathological CR in 5 of 34 patients). 133 More recently, in a single arm trial, Sym and coworkers evaluated the combination of docetaxel, cisplatin, and capecitabine. 226 Thirty-six of 49 patients were able to undergo surgery, with 31 achieving an R0 resection. The median OS was 12.1 months for the entire group. The median OS for patients undergoing surgery had not been reached at the time of the report.

Borderline Resectable and Locally Advanced Disease Three phase II studies have tested the use of preoperative systemic treatment in individuals defined by the study authors as having “locally advanced” stomach cancer. 227-229 Presumably, this represents a mix of patients with clinically resectable and unresectable or borderline resectable cancer. Resectability rates ranged from 60% to 77%, which seems to be higher than for patients with initially unresectable cancers. It is likely that more patients in these trials were potentially (borderline) resectable before neoadjuvant chemotherapy.

Kang and associates have presented an updated report of the only phase III trial of neoadjuvant chemotherapy in locally advanced or borderline resectable gastric cancer. 230 In the trial, 107 patients were randomized to receive two to three cycles of etoposide, 5-FU, and cisplatin followed by surgery versus surgery alone. Of the 53 patients randomized to preoperative treatment, 47 (89%) were explored and 37 (70%) were resected for cure. A 7% complete pathological response rate was noted. In the control group of 54 patients, 100% were explored and 61% curatively resected. Median survival was 43 versus 30 months in favor of neoadjuvant treatment, but this difference did not reach statistical significance (P = 0.114).



Preoperative systemic treatment may have certain advantages, such as the potential of reducing tumor bulk and increasing resectability. Micrometastatic disease may also be addressed earlier using this approach. Potential negatives to preoperative treatment include toxicity, delay in definitive therapy, and possibly increased surgical morbidity and mortality rates. The high response rates achieved with neoadjuvant chemotherapy are of interest, and this form of treatment will undoubtedly be the subject of further investigation as new and more active systemic therapies are developed. Thus far, however, no survival advantage has been demonstrated in phase III trials, and neoadjuvant chemotherapy should be considered investigational. Resectability rates in neoadjuvantly treated patients seem higher than the median rate of 40% from several surgical studies, but the patients in these studies are highly selected. Except for the trials in which some patients had unresectable disease on the basis of prior exploration, the successful operations in these reports may not have been influenced by neoadjuvant chemotherapy. In general, pathological CR rates are low ( ≤ 15%), and no proof exists that clinically staged patients are made more resectable by such treatment. The impact of preoperative systemic treatment on survival is even less clear. The one randomized trial that has been reported shows a nonsignificant improvement in survival for neoadjuvant treatment in borderline resectable/locally advanced disease. 230 Newer technologies such as EUS may identify patients who will do poorly with standard therapy alone and would be reasonable candidates for future neoadjuvant studies. The reports of combined systemic and intraperitoneal approaches are provocative and may warrant future phase III trials, because only through such studies will any impact on survival be determined.