Yao and colleagues present a concise, yet complete review and analysis of adjuvant therapeutic approaches for gastric adenocarcinoma. They confirm a fact known to all clinical oncologists who manage patients with resected gastric cancer: No adequate data support the value of postoperative (adjuvant) or preoperative (neoadjuvant) therapy in managing patients with locally advanced adenocarcinoma of the stomach.
Gastric cancer is a very common neoplasm worldwide. Various trials of chemotherapy in metastatic gastric adenocarcinoma have been conducted over the last 40+ years, and adjuvant therapy has been studied for the last 30 years. Given these facts, one could justifiably ask, why is there such a dearth of data on adjuvant and neoadjuvant therapy?
Problem Areas Defined Yao and colleagues clearly define problem areas in the preoperative and perioperative treatment of patients with resectable gastric adenocarcinoma. For example, many of the studies that have been conducted, particularly those performed in western countries, are underpowered. Randomized clinical trials with < 200 total patients and < 100 patients per treatment arm only have sufficient statistical power to detect large therapeutic differences with certainty. Expectations of such large therapeutic benefits are unreasonable with the chemotherapeutic and combined-modality approaches currently available.
Also, differences in the biology of adenocarcinoma of the stomach may be important. Worldwide, there are two major pathologic variants of gastric carcinoma. In countries with the highest frequency of this cancer, Helicobacter pylori infection and intestinal metaplasia lead to intestinal-type adenocarcinoma of the stomach. These tumors tend to occur more commonly in the body and distal part of the stomach.
The other major pathologic type of gastric carcinoma is a diffuse histology that can be associated with poorly differentiated, mucin-producing tumors, which widely infiltrate the stomach. In western countries, these neoplasms may be associated with a higher likelihood of cardioesophageal junction involvement. They can also be indistinguishable from adenocarcinoma of the cardioesophageal junction developing against the background of previously existing Barretts esophagus.
Whether the diffuse and intestinal types of tumors have different prognoses, stage for stage, is unclear. However, as discussed by Yao et al, there is evidence that distinct patterns of metastases may be seen with the different histologic types. For example, liver metastases appear to be associated with the intestinal type, and peritoneal carcinomatosis may be seen with increased frequency in the diffuse type.[1-3]
The issue of histology may be important in the design and analysis of adjuvant therapy protocols if, for example, different treatment modalities are more effective in one histology than the other, or if, indeed, there are major prognostic differences between the histologies. It is possible that variations in the outcomes of adjuvant and neoadjuvant therapy between western and eastern countries may be related to diferences in the relative prevalence of these histologies.
Type of Primary Therapy May Be Important
Another issue that may be of major importance, which is unrelated to adjuvant or neoadjuvant therapy, is the type of primary therapy (ie, surgical resection) used for gastric carcinoma. The recently published report of a randomized trial of D1 vs D2 dissection, by Bonenkamp et al, is discussed in some detail by Yao and colleagues. This study from the Netherlands, which involved over 700 patients, showed no advantage for extended lymph node (D2) dissection over the less extensive (D1) procedure. However, many expert gastric surgeons argue that, in experienced hands, a well-done D2 dissection is a safe, effective procedure that should be considered the standard of care.
Although one could argue that there exist no good data to support significant benefits (ie, increased surgical cure rate) of D2 dissection, impressive data suggest a marked staging benefit of this type of dissection. In 1995, the Dutch investigators performed a very interesting analysis on a subset of data from this ongoing study of D2 vs D1 dissection. Pathologists were asked to review the pathology of patients who had undergone a D2 dissection. First, patients who had received a D2 dissection were reviewed as if a D1 dissection had been performed. Thus, only the N1 nodes (those within 3 cm of the primary tumor) were examined for metastases. A stage was applied to each case as a result of the N1 nodal examination. Subsequently, the whole specimen was examined, the N2 nodes were assessed, and a stage was assigned.
This analysis found a significant improvement in the precision of staging when the N2 nodes were examined. For example, 49 patients were determined to have stage IIIA gastric cancer as a result of examination of N1 nodes. However, after the N2 nodes were examined, it was determined that only 19 of these patients truly had stage IIIA disease; 29 patients had stage IIIB tumors, and 1 patient had stage IV disease. This represents a 61% change in staging after the N2 nodes were examined. Similarly, in 24 patients deemed to have stage IIIB disease by N1 nodal examination, there was a 75% increase in stage when the N2 nodes were examined.
These data are particularly important because they demonstrate that D2 dissections clearly improve the accuracy of clinicopathologic staging. Thus, when the outcome results from the American vs eastern clinical trials are compared, one must keep in mind that the staging of node-positive patients in the American trials is likely to be incorrect in a minimum of 60% to 70% of patients. Therefore, since the US patients are clearly understaged, stage-for-stage survival would be expected to be quite different in the American vs eastern clinical trials.
As noted by Yao et al, the likelihood of understaging is confirmed by the work of Estes and colleagues. Their work reviewed surgical procedures in the most recent American study (intergroup [INT] 0116). This study, which closed in 1998 after the accrual of approximately 610 patients and is now in the follow-up phase, examined the benefit of adjuvant chemoradiation in patients with resected gastric cancer. Estes and colleagues reviewed the surgical technique used in approximately 400 resections. In this trial, only 6% of patients underwent a formal D2 dissection. Thus, in INT 0116, the surgical technique most widely used (D1 or lesser dissection) will result in unreliable staging. Analysis of the outcomes of this trial should be tempered by the likelihood that a large number of these cases were understaged.
Neoadjuvant therapy for gastric carcinoma is an area of significant current interest. All clinicians who manage patients with gastric cancer can cite examples of patients who have had excellent responses to preoperative combination chemotherapy or chemoradiation. These techniques are typically used in patients who have locally advanced tumors, in whom surgeons are reluctant to proceed with primary surgery. In this situation, cytoreduction from chemotherapy and/or chemoradiation is desired to facilitate a subsequent surgical resection.
It is certainly possible for neoadjuvant therapy to provide such a clinical benefit, as documented in a number of phase II studies.[8-10] However, it should be emphasized that, to date, no randomized studies of neoadjuvant approaches in gastric carcinoma have been conducted. A randomized trial of neoadjuvant therapy was recently performed in esophageal cancer, and a randomized study of neoadjuvant therapy will be high on the priority list for the next national gastric cancer clinical trial conducted in the United States.
Designing New Clinical Trials
As previously noted, a number of factors confound the interpretation of the results of adjuvant and neoadjuvant therapy in gastric cancer. These factors should not be considered obstacles, but rather, should be viewed as opportunities for the design of new clinical trials. For example, in the United States, where patients are understaged and microscopic residual disease is frequently present, one might consider the use of combined-modality approaches using systemic chemotherapy plus regional radiation therapy. Radiation is known to be effective in destroying microscopic residual disease and, therefore, may be an adjuvant strategy particularly well suited to less than optimally resected cases.
Clearly, the performance of sufficiently powered, prospectively randomized trials is a valuable tool for defining the efficacy of any gastric cancer treatment . Because of the proven success of the Gastric Cancer Working Group within the Gastrointestinal Cooperative Group intergroup mechanism, it is possible to accrue 600 to 700 patients to gastric cancer clinical trials in a period of approximately 5 to 7 years. This mechanism bodes well for the success of future clinical trials.
Growing Knowledge of Molecular Carcinogenesis
Finally, the future of oncology is being greatly influenced by a rapidly developing knowledge of molecular carcinogenesis. Again, as noted by Yao et al, there is no question that molecularly directed therapeutic interventions will be important in gastric cancer. Various molecular genetic abnormalities in this cancer have been defined,[12-14] and numerous strategies to explore molecularly targeted therapies are currently being developed and tested in many different neoplasms.
Clearly, such targeted approaches will improve the overall outlook for the treatment of human neoplasia. In gastric cancer, there is no question that neoadjuvant and adjuvant therapies will be improved by the application of such techniques in well-designed, randomized clinical trials.