Adjuvant Therapy for Gastric Carcinoma: Closing out the Century
Adjuvant Therapy for Gastric Carcinoma: Closing out the Century
Gastric cancer is a global health problem.
Although the incidence of this cancer is declining in many
industrialized nations, gastric cancer remains the second most
frequently diagnosed malignancy worldwide. It accounts for 9.9% of
all new cancer diagnoses and is responsible for 12.1% of all
cancer-related deaths. In the United States, it is estimated that
21,900 cases of gastric cancer will be diagnosed in 1999 and 13,500
persons will die of this disease.
Gastric cancer is often advanced and unresectable at diagnosisa
fact that contributes to its high morbidity and mortality. The
stomach is a hollow organ, and the abdominal cavity is large and
compliant to distention. Because of the large potential space,
patients often develop symptoms only when the cancer is far advanced.
Outside of Japan, an industrialized nation where gastric cancer is
common, early detection of gastric cancer is not attempted.
Reports based on large German and Korean databases show that 57% to
63% of gastric cancer patients undergoing resection of the primary
tumor have lymph node metastases.[3,4] In the United States, data
suggest that 85% of patients enrolled in a recently completed
adjuvant trial had lymph node metastases [personal communication, J.
S. Macdonald, MD, November 1998].
Worldwide, large amounts of resources have been expended in the
search for an effective adjuvant therapy to reduce the risk of
postoperative relapse. Numerous phase III clinical trials have been
published over the past few decades. The results of these trials,
however, have often been disappointing or equivocal and are sometimes
conflicting. Adjuvant therapy trials in western countries and Asia
published between 1984 and 1997 were recently reviewed by Shimada and
Ajani. In this article, we review various approaches to adjuvant therapy.
Selection of appropriate patients for adjuvant therapy is extremely
important. In addition to TNM stage, recent advances in the field of
molecular diagnostics will likely affect the selection of gastric
cancer patients for such therapy, as well as identify new prognostic
markers. Current studies correlating clinical outcomes with the
status of genes, such as p53, thymidylate synthase, ERCC1, and many
others, may help direct therapy for individual patients.
Based on TNM staging criteria, the depth of invasion (T), presence of
lymph node metastases, and number of lymph nodes involved (N) predict
the risk of relapse. This was confirmed in a large German multicenter
trial published by Siewert et al in 1998. The study involved 1,654
patients undergoing surgical resection of gastric tumors.
A significant survival advantage was found for the 1,182 patients
(71.5%) who underwent an R0 resection (ie, no residual disease and no
cancer cells at the resection margins). In addition to the depth of
invasion and nodal status, the ratio between involved and removed
lymph nodes proved to be an important independent prognostic factor.
The prognostic importance of R0 resection, depth of invasion, nodal
status, and the ratio between involved nodes and removed nodes was
confirmed in a large Korean study involving 10,783 patients. In
this study, an R0 resection was achieved in 4.8% of patients.
The American College of Surgeons also has published a large analysis
of gastric cancer patients. This study was based on a tumor
registry review of 25 consecutive patients from each participating
institution seen in the years 1982 and 1987. Data on a total of
18,365 patients were analyzed.
Despite the large number of patients, however, the results of this
analysis are not easy to interpret. This was a retrospective study
that involved a large
number of heterogeneous institutions. Symptoms at presentation may
have important prognostic value. Weight loss was the most common
symptom, reported in 61.6% of patients. However, the presence or
absence of weight loss was not correlated with clinical outcome, and
the degree of weight loss was not specified. Similarly, 42% of
patients who underwent surgical resection with clear margins received
adjuvant chemotherapy or radiation therapy. However, patient
characteristics and type of therapy given were not specified. The
5-year survival rates based on pathologic stage after resection were
reported as 50%, 29%, 13%, 3% for stages I, II, III, and IV,
respectively. These survival rates are significantly worse in every
stage category than those reported in the German and Korean studies.
Histologic tumor type may influence patterns of failure. The liver is
a frequent site of failure in intestinal-type tumors, whereas
peritoneal carcinomatosis is more common with the diffuse type.[7-9]
Diffuse histology is on the rise throughout the world.
Advances in molecular diagnostics have opened new avenues for
predicting clinical outcome. One area of active research is the
determination of p53 mutations. In early-stage, T1 tumors, p53
overexpression correlates with depth of invasion and lymph node involvement.
Another area of active research is the correlation of in vivo
response to chemotherapy with genetic phenotype. Fluorouracil, which
is the most commonly used agent for gastric cancer, targets the
enzyme thymidylate synthase. Lenz et al found that low
thymidylate synthase messenger RNA (mRNA) expression was predictive
of response and survival in patients treated with fluorouracil and
cisplatin (Platinol). More recently, however, Fata et al found
high thymidylate synthase mRNA expression to be a predictor of
The excision repair cross-complementing gene (ERCC1) was studied in
33 patients treated with preoperative fluorouracil and cisplatin. In
this study, ERCC1 expression correlated with response to therapy.
Furthermore, when thymidylate synthase and ERCC1 expression were both
low, 11 (85%) of 13 patients responded to therapy; in contrast, when
both thymidylate synthase and ERCC1 expression were high, only 2
(20%) of 10 patients responded.
The current National Comprehensive Cancer Network (NCCN) practice
guidelines for gastric cancer recommend the following tests and
examinations as the minimum preoperative work-up: history, physical
examination, complete blood counts (including platelets), SMA-12,
computed tomography (CT) of the abdomen, chest roentgenogram,
esophagogastroduodenoscopy, and, in female patients, CT or ultrasound
of the pelvis.
Increasingly, laparoscopy also is being recommended prior to a major
resection. The use of laparoscopy may be supported by the fact that
an R0 resection cannot be performed in 30% to 40% of patients
undergoing surgery. In the German study, 29.3% of the patients had M1
disease at laparotomy. In these patients, median survival was less
than 12 months.
At the University of Texas M. D. Anderson Cancer Center, 83 patients
were enrolled in preoperative chemotherapy trials, 73% of whom
underwent a curative resection. In the group who could not receive a
curative resection, 55% of patients avoided nontherapeutic laparotomy
because of an extensive preoperative work-up.
A number of controversies exist in the surgical management of gastric
cancer. The most important of these is the extent of lymphadenectomy.
A D1 resection entails a gastrectomy with the removal of all
perigastric nodes and the removal of the greater and lesser omenta.
In addition to these structures, for a D2 dissection the surgeon
removes the omental bursa portion of the transverse mesocolon and the
nodes along the left gastric, celiac , and splenic arteries. For a D3
dissection, in addition to the standard
D2 dissection, lymph nodes in the hepatoduodenal ligament, along the
superior mesenteric vein, posterior to the common hepatic artery, and
on the posterior surface of the pancreatic head are also removed. A
D4 dissection involves the removal of lymph nodes around the
abdominal aorta, in addition to all of the structures mentioned above.
In the East, extensive lymphadenectomy (D2 through D4) is commonly
practiced without excessive complications. Recently, the Japan Clinical
Oncology Group conducted a randomized study of D2 dissection with
or without para-aortic node dissection; these researchers reported no
treatment-related deaths but a 6% rate of major complications. In the
West, the extent of lymph node dissection is more controversial.
Western surgeons usually do not perform D2 dissections due to the
high rate of complications.
The Dutch Gastric Cancer Group randomized 711 patients with all
stages of disease to either a D1 or D2 dissection. Patients who
underwent a D2 resection had a higher rate of complications, more
postoperative deaths, longer hospital stays, and no improvements in
5-year survival. However, assuming that the ratio of involved
lymph nodes to resected nodes predicts outcome, removal of more
negative nodes, which occurs during an extended node dissection,
may improve long-term survival.
Subgroup analyses of the German and Korean databases show that
D2 dissection benefited patients with stage II disease in both
studies and benefited patients with stage IIIA disease in the Korean
study. However, subgroup analyses are fraught with difficulties. The
level of surgical expertise needed to perform an extended node
dissection appears to have an impact on surgical morbidity and mortality.
An analysis by Estes et al shows that operative documentation
needs improvement. These investigators reviewed operative reports
from more than 300 surgeons given a checklist for documentation prior
to participating in a national protocol. Inadequate documentation was
common. The status of the primary tumor, lymph nodes, liver,
peritoneum, and omentum was not stated in 6%, 10%, 17%, 28%, and 28%
of operative reports, respectively. Also, the reports often lacked
sufficient information about the extent of dissection.
At present, D1 dissection should be considered the minimum resection
for patients with potentially curable gastric carcinoma. Estes et al
found that 54.2% patients undergoing curative resection
of gastric cancer had a D0 dissection. This indicates that adequate
resection and staging are not being practiced. D2 resection, a
proper oncologic surgical procedure, should be considered by
experienced surgeons who frequently perform gastric cancer surgery.