Randomized Clinical Trial on the Combination of Preoperative Irradiation and Surgery in the Treatment of Adenocarcinoma of Gastric Cardia (AGC) - Report on 370 Patients

Zhang Z et al.
Abramson Cancer Center of the University of Pennsylvania
Ultima Vez Modificado: 1 de noviembre del 2001

Share article

Imprima English

Reviewers: John Han-Chih Chang, MD
Source: International Journal of Radiation Oncology Biology Physics December 1998, Volume 42 (Number 5): pages 929 - 934.


Gastric or stomach cancer is a rather uncommon malignancy in the United States, but has a high incidencein China, Japan and Chile. It is the second most common cause of cancer death in the world. The main riskfactors include nitrosamines in preserved salted or smoked foods. Most cases are advanced at the time ofdiagnosis.

The mainstay of treatment is surgery. However, a study from the University of Minnesota in the early1980's demonstrated a local recurrence rate of nearly 70%. Many other authors concur with the poorefficacy of surgery alone. Additional therapy seems to be needed to improve outcome. Still, numerousstudies and retrospective reviews have been performed with equivocal results on survival with adjuvanttherapy of any kind (ie. chemotherapy or radiation therapy). This article details a prospective randomizedtrial of 370 patients with tumors of the gastric cardia (most cephalad aspect of the stomach ? essentiallyjust caudal to the gastro-esophageal junction) that ran from 1978 to 1989 in China.

Materials and Methods

The departments of radiation oncology and thoracic surgical oncology at the Cancer Institute (Hospital),Peking Union Medical College and Chinese Academy of Medical Sciences began this randomizedprospective trial that enrolled 370 patients. All had adenocarcinomas of the gastric cardia (AGC) withoutevidence of distant metastasis.

The two arms consisted of surgery alone (S) versus preoperative radiation therapy (PrRT) followed bysurgery. It was not specified whether a total or subtotal gastrectomy was performed in either case. Onehundred and ninety-nine patients were randomized to surgery alone, while 171 were randomized topreoperative RT. RT fields included the 5-cm margins around the gross tumor along with distalesophagus, gastric fundus, cardia and lesser curvature including draining lymph nodes. Megavoltage ortelecobalt irradiation delivered 40Gy in 2Gy fractions per day.

Median follow-up was over 10 years in both arms. Only 2-3% lost to follow-up, and 80% came back theinstitution of treatment for follow-up. The rest were followed via correspondence with the local physician.Analysis was by intent-to-treat basis.


The survival of the patients in the PrRT arm was far superior than the S arm. Analysis was byintent-to-treat basis. Figure 2 demonstrates this point. There is a divergence of the survival curve fromnearly the beginning to about 3 years and then continued separation to about 9 years. The 5 and 10 yearsurvival rates were 30% and 20% for PrRT and 20% and 13% for S, respectively (p = 0.0094).

One hundred fifty-three (90%) of the PrRT patients and 158 (80%) of the S patients could be resected.For the PrRT patients, 80% were radically resected and 10% palliatively. The respective numbers for theS arm were 62% and 18%. Comparison of the 5 and 10-year survival rates is depicted in figure 3. Spatients had a 22.5% and 16.6% rate, while PrRT patients had 33% and 25% (p = 0.15 - non statisticallysignificant). The survival curve is similar for the first year but diverges from there.

For those that could only be explored and not resected (18 in PrRT and 41 in the S), survival was againmuch better with PrRT. Figure 4 depicts this in graphical form with the curves diverging from thebeginning. The mean survival was 11 versus 5 months and median of 7 versus 4 months for the PrRT andS arms, respectively. The p-value for this survival difference was 0.008. Regardless, they both do poorlywithout surgical resection (almost all patients die by 3 years).

The lymph node positivity was reduced with PrRT from 85% to 64%. The pathological stage wassignificantly reduced by PrRT as demonstrated in the following chart extracted from Table 1 in the article.

Pathological stage after surgery



1 (0.6%)0


48 (28%)21 (11%)


103 (60%)136 (68%)


19 (11%)42 (21%)

There appears to be an 18% downstaging from III and IV to I and II when comparing PrRT to S arms.

Sixty-six and 75% of those that failed had some component of locoregional recurrence in PrRT and S arms, respectively.Distant metastasis alone was seen in 16% of the PrRT arm and 12% of the S arm, though the overall distant metastasisrate (alone or with locoregional recurrence) was 24% in both arms.

Operative mortalities (mortality within 30 days of the operation) were not different in the PrRT (0.6%) arm versus the S arm(2.5%). Anastomotic leak was seen only in 2% of the PrRT patients, while it was 4% for the S arm.

Discussion and Conclusion

The mainstay of gastric cancer treatment remains surgery. But time and time again, we realize that it is not enough tojust perform radical surgery. A retrospective review from Thomas Jefferson University demonstrated a survival advantage topatients who received adjuvant therapy (RT and/or chemotherapy).

However, randomized trial after randomized trial has failed to truly demonstrate a survival advantage. In the BritishStomach Cancer Group three arm trial, there was no difference in overall or median survival in patients who receivedsurgery alone versus surgery plus chemotherapy or surgery plus RT. A criticism was that over 30% of patients in the RTarm did not get full dose RT. The Gastrointestinal Tumor Study Group (GITSG) ran two prospective randomized trials onlocally advanced gastric carcinoma (resected or unresected) patients attempting to define the role for RT. Thesedemonstrated no advantage to adding RT to Chemotherapy.

The general opinion is that RT adds very little to the survival of gastric carcinoma patients though local control may beimproved. This article is important in that it demonstrates an indication for RT for adenocarcinomas of the gastric cardia.Extrapolating this data for all carcinomas of the stomach may be premature, though. To lend more credence to this study,the authors should publish their stratification criteria, especially the distribution of radiographic and clinical stage prior tosurgery. This would support that PrRT downstaging was a real phenomenon and not just secondary to less bulk ofdisease in the PrRT arm patients. Also, details of what constituted a radical resection should be explained.

Three hundred fourteen patients have died of the original 370 as testament to how far we still have to go with this disease.Perhaps the dose of RT was too low. We know adenocarcinomas of the esophagus get much higher biological doses inan attempt to obtain cure. The same may hold true for the stomach. Chemotherapy to be combined with RT may alsoimprove the outcome. In many of the clinical trials underway, combined modality is utilized to reach synergy of cancerkilling power. This article details a step in the right direction, but we must move forward into new avenues of improvementin this disease. As a Chinese proverb states, learning is like rowing upstream, to stop is to fall back.

Imprima English
CT Scans Can ID Gemcitabine Response in Pancreatic Cancer

Mar 20, 2014 - Transport properties identified on routine computed tomography scans correlate with clinically relevant end points for patients with pancreatic ductal adenocarcinoma who receive preoperative gemcitabine chemoradiotherapy, according to a study published online March 10 in the Journal of Clinical Investigation.

I Wish You Knew

How cancer patients have changed my life

View More

Blogs and Web Chats

OncoLink Blogs give our readers a chance to react to and comment on key cancer news topics and provides a forum for OncoLink Experts and readers to share opinions and learn from each other.

OncoLink OncoPilot

Frente a un nuevo diagnóstico de cáncer o de cambiar el curso de su tratamiento actual? Deje que nuestro personal de enfermería cáncer que ayudan a pasar!

Más información