Paclitaxel-Based Chemoradiotherapy in Localized Gastric Carcinoma: Degree of Pathologic Response and Not Clinical Parameters Dictated Patien
http://www.100md.com
《临床肿瘤学》
the Departments of Gastrointestinal Medical Oncology, Surgical Oncology, Radiation Oncology, Clinical Pathology, Gastrointestinal Medicine and Nutrition, and Biostatistics, The University of Texas M.D. Anderson Cancer Center, Houston, TX
ABSTRACT
PATIENTS AND METHODS: Patients with operable, localized gastric, or gastroesophageal adenocarcinoma were eligible. Staging included endoscopic ultrasonography (EUS) and laparoscopy. Patients received two 28-day cycles of induction chemotherapy of fluorouracil, paclitaxel, and cisplatin followed by 45 Gy of radiation and concurrent fluorouracil plus paclitaxel. The cancer was restaged and surgery was attempted. Postsurgery pathologic findings and R0 resection were correlated with OS and DFS.
RESULTS: Forty-one patients were enrolled. Most carcinomas were proximal (83%) and pretreatment stage EUST3 (85%). Forty patients (98%) underwent surgery, and 78% had an R0 resection. We observed a pathologic complete response (pathCR) rate of 20% and a pathologic partial response (pathPR) rate of 15% (< 10% residual cancer cells in the resected specimen). No pretreatment parameter (sex, cancer location, baseline T stage, or baseline N stage) predicted the type of postsurgery pathologic response, OS, or DFS. However, pathCR (P = .02), pathCR + pathPR (P = .006), R0 resection (P < .001), and postsurgery T and N stages (P = .01 and P < .001, respectively) were associated with OS. Same parameters were significantly correlated with DFS. Toxicity was manageable.
CONCLUSION: The type of pathologic response but not pretreatment parameters was associated with OS and DFS. Efforts to increase the rate of pathologic response and better systemic cancer control are warranted.
INTRODUCTION
The ideal prognostic circumstances for patients with localized gastric cancer are a low cancer stage and an R0 resection performed by an experienced surgeon. A diagnosis of low-stage gastric cancer in most countries is unlikely, because symptoms usually do not exist at early stage. However, treating patients who have stage II or III gastric cancer with chemoradiotherapy to achieve a clinically significant downstaging before surgery—and thus to create a greater likelihood of R0 resection—may be possible. In a recent study assessing the effects of preoperative chemoradiotherapy for localized gastric cancer, we observed that the overall survival (OS) duration of patients who achieved a pathologic complete response (pathCR) was significantly longer than that of patients who did not have a pathCR.5 We have recently described a revised method for stratifying pathologic responses.6 In this study, we evaluated the impact of preclinical stage and an R0 resection as well.
Laparoscopic evaluation for patients with localized gastric cancer is routine at our institution, and this staging step can uncover peritoneal metastases in up to 20% of patients,7,8 thus avoiding unnecessary major operation. In this study, it provided an opportunity to insert a jejunostomy feeding tube before the initiation of preoperative therapy.
We have conducted the current study to also determine whether preoperative paclitaxel-based chemoradiotherapy can induce a pathologic response in patients with stage II/III gastric or gastroesophageal adenocarcinoma.
In this study, paclitaxel replaced leucovorin in the induction chemotherapy regimen and was added to the chemoradiotherapy regimen. Paclitaxel is an active agent against gastroesophageal cancers9,10 and, when used concurrently with radiotherapy, is a radiation sensitizer.11 The combination of paclitaxel, fluorouracil, and cisplatin is also active against adenocarcinoma of the esophagus or gastroesophageal junction.12 Thus paclitaxel-based chemotherapy may provide patients who have gastric or gastroesophageal adenocarcinoma with more benefits than our previously assessed combination of drugs.
Although the experience with preoperative chemoradiotherapy in localized gastric cancer is very limited, on the basis of the results from our previous study,5 we hypothesized that the pathCR rate would be at least 20%, that OS and disease-free survival (DFS) would depend on the degree of chemoradiotherapy-induced pathologic response, and that pretreatment parameters would not predict OS or DFS.
PATIENTS AND METHODS
Patients ineligible for inclusion were those with T4, M1, or T1N0 carcinoma, peritoneal carcinomatosis (gross or microscopic), or uncontrolled medical conditions (eg, diabetes, hypertension, New York Heart Association class III or IV, or psychiatric illness). Patients who could not comprehend or comply with the study were also ineligible.
Study Design
Patients were to receive one or two cycles of induction chemotherapy and then chemoradiotherapy. The second induction chemotherapy cycle was given if the cancer had not progressed, but if the cancer did progress, then this cycle was skipped and the patient began concurrent chemotherapy and radiotherapy. Patients were taken off protocol if distant metastases developed. Clinical responses were judged using radiographic or endoscopic techniques.
If a patient had an R0 resection, no further therapy was given. In the case of R1 resection (microscopically positive margin), R2 resection (only partial resection), or discovery of M1 carcinoma, the patient was given palliative care.
Step 1: Chemotherapy
Induction chemotherapy consisted of fluorouracil at 750 mg/m2/d by continuous infusion on days 1 to 5, cisplatin at 15 mg/m2/d intravenously on days 1 to 5, and paclitaxel at 200 mg/m2 intravenously over 24 hours on day 1. Patients received therapy in the outpatient setting with standard premedications and hydration. Daily subcutaneous granulocyte colony-stimulating factor was administered on days 6 to 16 (300 μg for patients ≤ 70 kg and 480 μg for patients > 70 kg). If the cancer had not progressed, then the second cycle was to be repeated 28 days after the first cycle began.
On the basis of the National Cancer Institute Common Toxicity Criteria version 2.0, drug doses were decreased by 25% if grade 3 nonhematologic toxicity or grade 4 hematologic toxicity occurred. CBC counts were performed weekly. Serum chemistries were monitored before each course.
Step 2: Chemoradiotherapy
The interval between day 1 of the last induction chemotherapy cycle and the first day of chemoradiotherapy was ≥ 28 days. Radiation fields included the entire stomach, perigastric extension (if present), and draining lymph nodes (gastric, celiac, porta hepatis, gastroduodenal, splenic-suprapancreatic, and retropancreaticoduodenal). For lesions involving the cardia or gastroesophageal junction, a 5-cm margin of esophagus was included, and for distal lesions at or near the gastroduodenal junction, a 5-cm margin of duodenum was included. Esophagoscopy, barium swallow radiographs, and computed tomography of the chest and abdomen were used to determine the maximum extent of disease relative to the primary tumor and the nodal groups. Idealized fields were modified as needed to shield at least two thirds of one kidney. For proximal lesions, cardiac shielding was recommended, along with evaluation of lateral fields for a component of treatment.
Linear accelerators delivered a dose of 45 Gy (25 fractions of 1.8 Gy) over 5 weeks using either 15- or 18-MV photons and a three-dimensional conformal radiotherapy technique. Typically, anterior and posterior fields were sometimes obliqued slightly to avoid the spinal cord and the right kidney were selected and delivered the majority of the dose.
The concurrent chemotherapy, given in an outpatient setting, consisted of fluorouracil at 300 mg/m2/d by continuous infusion with a portable pump 5 days each week (this treatment usually started on a Monday and ended on Friday, after the radiation therapy) plus paclitaxel at 45 mg/m2 intravenously each Monday for 5 weeks.
Step 3: Surgery
The type of surgery performed depended on the location and extent of the primary cancer. The cancer was resected along with a luminal gastric margin of ≥ 5 cm when feasible. Also when feasible, a 2-cm duodenal margin was obtained for distal cancers and a 3-cm esophageal margin was obtained for proximal cancers. In both cases, frozen-section confirmation of a negative margin was sought. For distal cancers, a subtotal gastrectomy was considered adequate; total gastrectomy was at the discretion of the surgeon. For proximal cancers, total gastrectomy or total esophagogastrectomy was performed. En bloc resection of adjacent organs was preformed when their involvement was questionable. The spleen was preserved when possible. An attempt was made to perform a D2-type nodal dissection. During surgery, a J-tube was placed for temporary nutritional support.
Tumor Response, Toxicity Criteria, and Data Management
Upper gastrointestinal barium radiographs were taken after each 28-day cycle of induction chemotherapy and just before surgery, and esophagogastroduodenoscopy, computed tomography of the abdomen, chest radiography, and all blood tests were repeated before surgery. Previously described criteria for response evaluation5,6 were used. Briefly, pathCR was defined as an absence of carcinoma cells in the primary site, and pathologic partial response (pathPR) was defined as less than 10% residual carcinoma cells in the specimen.
Follow-Up
Each patient was assessed at 3, 6, and 12 months and then every 6 months for 5 years or until the patient's death.
Statistical Methods
One of our hypotheses was that preoperative paclitaxel-based therapy results in a pathCR rate of 20%. Thus Simon's two-stage design was used: 21 patients were to be assessed first, and if more than one pathCR was observed, then 22 additional patients were to be added.
Patient responses (pathCR and any pathologic response [pathCR + pathPR]) were cross-tabulated by sex, cancer location, and baseline T and N stages. Baseline T-stage data were grouped into two categories, T1 plus T2, and T3. The Fisher's exact test was used to assess the association between each of these factors and the response rates. OS was defined as the time from the start of induction chemotherapy to the date of death or date of last follow-up. DFS was defined as the time from the start of induction chemotherapy to recurrence of the cancer. Data from living patients were censored. The log-rank test was used to test for differences in survival distributions by sex, cancer location, baseline and postoperative T and N stages, and T and N downstaging. Univariate Cox proportional hazards models were fit to yield estimates of hazard ratios for each of these factors. All statistical tests were two sided, and significance was set at P ≤ .05.
RESULTS
Response to Induction Chemotherapy
None of the patients had progression of the primary cancer during induction chemotherapy. Some subjective improvement in the ability to eat larger meals, reduction in postprandial discomfort or pain, and weight gain was reported by 49% of patients.
Response to Chemoradiotherapy
Further improvement in symptoms was universal on recovery (approximately 5 to 6 weeks) from chemoradiotherapy. Twenty-nine (71%) of the 41 patients underwent an EUS examination just before surgery. Of the 29 patients, 28 (69%) had T3 cancer and 16 (38%) had N+ cancer.
Surgical Findings and Surgical Pathology
Forty (98%) of 41 patients underwent surgery (one patient died before surgery). Thirty-three patients (80%) had a gastrectomy, and 32 patients (78%) had an R0 resection. Gastrectomy was not done in the remaining seven patients because M1 cancer (four with unsuspected peritoneal carcinomatosis) was discovered at surgery.
A pathCR was noted in eight patients (20%), and a pathPR was noted in six patients (15%). Thus the overall pathologic response rate was 35%. In the 33 patients who had a gastrectomy, the primary carcinoma was T3 in nine patients (27%), T2 in six patients (18%), T1 in nine patients (27%), and T0 in nine patients (27%). Twenty-two (67%) of the 33 patients who underwent gastrectomy had N0 cancer, eight patients (24%) had N1 cancer, and three patients (9%) had N2 cancer. The median number of nodes examined in the 33 gastrectomy specimens was 19 (range, one to 62 nodes). The median number of nodes with carcinoma was 3.5, and the median number of cancer-free nodes was 17.
Clinical Predictors of pathCR and pathPR
No clinical factor (sex, cancer location, baseline T stage, baseline N stage, or baseline EUS stage) was significantly associated with pathCR (Table 2) or any pathologic response (data not shown).
Determinants of Downstaging
T-stage downstaging (ie, a lower T stage after surgery than at baseline) was associated with baseline EUS stage (P ≤ .05) but not with sex, cancer location, or baseline N stage. Similarly, N-stage downstaging (ie, a lower N stage after surgery than at baseline) was associated with baseline EUS stage (P ≤ .05) but not with sex, cancer location, or baseline T stage).
Survival
At a median follow-up of 36+ months (range, 24 to 67 months), 28 (68%) of 41 patients were alive. Thus the median survival time for all 41 patients has not been reached. Among the 13 patients who died, one died before surgery as a result of acute myocardial infarction. The remaining 12 died as a result of their carcinoma.
We examined the effect of 13 factors on OS including sex; among these the following were statistically significantly correlated with OS: R0 resection (P < .001; Fig 1), pathCR (P = .02; Fig 2), any pathologic response (ie, pathCR plus pathPR; P = .006; Fig 3), postsurgery node positivity (P = .02), postsurgery N stage (P < .001), and postsurgery T stage (P = .01; Table 3).
Similarly, DFS was statistically significantly associated with T-stage downstaging (P = .005), pathologic response (pathCR plus pathPR; P = .003), pathCR (P = .008), R0 resection (P < .001), postsurgery node positivity (P = .002), postsurgery N stage (P < .001), and postsurgery T stage (P < .001).
OS Predicted by Postsurgery Outcome
Pretherapy EUS T stage (T1/2 v T3) had no significant influence on OS (P = .42; Fig 4A) but the post-therapy T stage had a highly statistically significant influence on patient survival (P < .001; Fig 4B). Similarly, pretherapy EUS N stage (N0 v N+) had no significant bearing on OS (P = .22; Fig 5A) but the post-therapy N stage had a highly statistically significant influence of OS (P < .001; Fig 5B). Similar influence of post-therapy T and N was seen on DFS (data not shown).
Chemotherapy-Induced Toxicity
No deaths were related to induction chemotherapy. The most common toxic effects were myelosuppression, nausea or vomiting, and fatigue (Table 4).
Chemoradiotherapy-Induced Toxicity
One patient died just before surgery of an acute myocardial infarction that was attributed to therapy. The most common chemoradiotherapy-related toxic effects were nausea or vomiting and fatigue (Table 5).
Surgical Complications
Two patients had a temporary anastomotic leak, and one patient had an abdominal abscess. There were no other major complications from surgery.
DISCUSSION
In the current trial, our goal was to investigate paclitaxel in the induction regimen and concurrently with radiation therapy. We hypothesized that paclitaxel-based chemoradiotherapy would result in a pathCR rate of at least 20% and that would, in turn, be associated with OS and DFS. We also hypothesized that preoperative parameters would not predict OS or DFS. Indeed, in this study, baseline T and N stages were not associated with OS or DFS. Our findings certainly do not diminish the value of adequate pretreatment staging, which is paramount in selecting patients in the clinical trials and will be a necessary stratification tool in future phase III trials. Nevertheless, our results suggest that clinical cancer staging, for all its sophistication, is suboptimal and requires improvement. It may be argued that clinical assessment does not reflect the true biology of the cancer and therapy only helps to unmask its character. However, it may also be argued that clinical stage is more valuable and predictive when surgery is the primary therapy, but once the clinical stage (pretreatment pathologic stage) is altered by ensuing preoperative therapy, it is the final pathologic stage (altered or unaltered) that is predictive of patient outcome. Our study, although small, points in that direction and, if true, then discovery of therapies that favorably alter the pretreatment pathologic stage is warranted.
We observed an R0 resection rate of 78%, and patients who had an R0 resection survived substantially longer than patients who did not. Whether this was due to patient selection or true effect of preoperative therapy can only be decided in a randomized trial. Paclitaxel-based preoperative chemoradiotherapy was also the subject of a recently completed Radiation Therapy Oncology Group study (RTOG 99-04) for patients with localized gastric cancer. However, orchestrating preoperative or postoperative therapy is complex and requires continuous participation by professionals from multiple disciplines and by informed caretakers and relatives of the patient. Although surgery remains the primary therapy for patients with localized cancer, multidisciplinary decisions improve the outcome as it results in advanced planning of interventions.
The results from our current study of 41 patients and our previous study of 33 patients5 indicate that pathCR may be a consistent result from preoperative chemoradiotherapy in patients with localized gastric carcinoma. The results from our current study also demonstrate that it is feasible to incorporate newer agents in this strategy. The contribution of paclitaxel cannot be determined by comparing sequential phase II studies, but the median survival beyond 36 months is intriguing. If one wished to increase the R0 resection rate and also the fraction of patients with pathologic response, then preoperative chemoradiotherapy may be one important approach. Induction chemotherapy that produces a high response rate could reduce the bulk of the primary cancer, and the ensuing chemoradiotherapy could be more effective against smaller cancer than it would be against a de novo bulky cancer. This could lead to a successful surgery. However, this notion remains hypothetical. Despite patient selection by laparoscopic staging, peritoneal cavity as a frequent failure site remains a concern.
There is limited information regarding pathologic response and its impact on patient outcome for localized gastric carcinoma. Our study suggests that it is possible to expect a pathCR rate of approximately 20% in a well-staged group of patients with localized gastric cancer. Once the pretreatment stage is altered, it becomes less predictive of outcome. The strategy of preoperative paclitaxel-based induction therapy and chemoradiotherapy is feasible. In the absence of phase III trials, the question remains whether one is altering the biologic behavior of gastric cancer with such maneuvers or simply characterizing it. Nevertheless, efforts to increase the rate of pathologic response and better control of metastases are warranted.
Authors' Disclosures of Potential Conflicts of Interest
NOTES
Supported in part by a grant from the Bristol-Myers Squibb Oncology Division and the Cantu, Dallas, and Capporella families.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Macdonald JS, Smalley S, Benedetti J, et al: Chemoradiotherapy after surgery compared with surgery alone for adenocarcinoma of the stomach or gastroesophageal junction. N Engl J Med 345:725-730, 2001
2. Wanebo HJ, Kennedy BJ, Chmiel J, et al: Cancer of the stomach: A patient care study by the American College of Surgeons. Ann Surg 218:583-592, 1993
3. Bonenkamp JJ, Hermans J, Sasako M, et al: Extended lymph node dissection for gastric cancer. N Engl J Med 340:908-914, 1999
4. Hundahl SA, Phillips JL, Menck HR: The National Cancer Data Base report on poor survival of US gastric carcinoma patients treated with gastrectomy. Cancer 88:921-932, 2000
5. Ajani JA, Mansfield P, Janjan N, et al: Multi-institutional trial of preoperative chemoradiotherapy in patients with potentially resectable gastric carcinoma. J Clin Oncol 22:2774-2780, 2004
6. Chiriec LR, Swisher SG, Ajani JA, et al: Posttherapy pathologic stage and residual carcinoma as predictor of survival in patients with esophageal or esophagogastric junction carcinoma treated with preoperative chemoradiation. Cancer (in press)
7. Lowy AM, Mansfield PF, Leach SD, et al: Laparoscopic staging for gastric cancer. Surgery 119:611-614, 1996
8. Pearlstone DB, Mansfield PF, Curley SA, et al: Laparoscopy in 533 patients with abdominal malignancy. Surgery 125:67-72, 1999
8. Ajani JA, Ilson DH, Daugherty K, et al: Activity of paclitaxel (Taxol) in patients with squamous cell carcinoma and adenocarcinoma of the esophagus. J Natl Cancer Inst 86:1086-1091, 1994
10. Ajani JA, Fairweather J, Pazdur R, et al: Phase II study of Taxol in patients with untreated metastatic gastric carcinoma. Cancer J Sci Am 4:269-274, 1998
11. Safran H, Wanebo HJ, Hesketh PJ, et al: Paclitaxel and concurrent radiation for gastric cancer. Int J Radiat Oncol Biol Phys 46:889-894, 2000
12. Ilson DH, Ajani JA, Bhalla K, et al: A phase II trial of paclitaxel, fluorouracil, and cisplatin in patients with advanced carcinoma of the esophagus. J Clin Oncol 16:1826-1834, 1998
13. Allum W, Cunningham D, Weeden S: Perioperative chemotherapy in operable gastric and lower oesophageal cancer: A randomized, controlled trial (the MAGIC trial, ISRCTN 9379371). Proc Am Soc Clin Oncol 22:249, 2003 (abstr 998)(J.A. Ajani, P.F. Mansfiel)
ABSTRACT
PATIENTS AND METHODS: Patients with operable, localized gastric, or gastroesophageal adenocarcinoma were eligible. Staging included endoscopic ultrasonography (EUS) and laparoscopy. Patients received two 28-day cycles of induction chemotherapy of fluorouracil, paclitaxel, and cisplatin followed by 45 Gy of radiation and concurrent fluorouracil plus paclitaxel. The cancer was restaged and surgery was attempted. Postsurgery pathologic findings and R0 resection were correlated with OS and DFS.
RESULTS: Forty-one patients were enrolled. Most carcinomas were proximal (83%) and pretreatment stage EUST3 (85%). Forty patients (98%) underwent surgery, and 78% had an R0 resection. We observed a pathologic complete response (pathCR) rate of 20% and a pathologic partial response (pathPR) rate of 15% (< 10% residual cancer cells in the resected specimen). No pretreatment parameter (sex, cancer location, baseline T stage, or baseline N stage) predicted the type of postsurgery pathologic response, OS, or DFS. However, pathCR (P = .02), pathCR + pathPR (P = .006), R0 resection (P < .001), and postsurgery T and N stages (P = .01 and P < .001, respectively) were associated with OS. Same parameters were significantly correlated with DFS. Toxicity was manageable.
CONCLUSION: The type of pathologic response but not pretreatment parameters was associated with OS and DFS. Efforts to increase the rate of pathologic response and better systemic cancer control are warranted.
INTRODUCTION
The ideal prognostic circumstances for patients with localized gastric cancer are a low cancer stage and an R0 resection performed by an experienced surgeon. A diagnosis of low-stage gastric cancer in most countries is unlikely, because symptoms usually do not exist at early stage. However, treating patients who have stage II or III gastric cancer with chemoradiotherapy to achieve a clinically significant downstaging before surgery—and thus to create a greater likelihood of R0 resection—may be possible. In a recent study assessing the effects of preoperative chemoradiotherapy for localized gastric cancer, we observed that the overall survival (OS) duration of patients who achieved a pathologic complete response (pathCR) was significantly longer than that of patients who did not have a pathCR.5 We have recently described a revised method for stratifying pathologic responses.6 In this study, we evaluated the impact of preclinical stage and an R0 resection as well.
Laparoscopic evaluation for patients with localized gastric cancer is routine at our institution, and this staging step can uncover peritoneal metastases in up to 20% of patients,7,8 thus avoiding unnecessary major operation. In this study, it provided an opportunity to insert a jejunostomy feeding tube before the initiation of preoperative therapy.
We have conducted the current study to also determine whether preoperative paclitaxel-based chemoradiotherapy can induce a pathologic response in patients with stage II/III gastric or gastroesophageal adenocarcinoma.
In this study, paclitaxel replaced leucovorin in the induction chemotherapy regimen and was added to the chemoradiotherapy regimen. Paclitaxel is an active agent against gastroesophageal cancers9,10 and, when used concurrently with radiotherapy, is a radiation sensitizer.11 The combination of paclitaxel, fluorouracil, and cisplatin is also active against adenocarcinoma of the esophagus or gastroesophageal junction.12 Thus paclitaxel-based chemotherapy may provide patients who have gastric or gastroesophageal adenocarcinoma with more benefits than our previously assessed combination of drugs.
Although the experience with preoperative chemoradiotherapy in localized gastric cancer is very limited, on the basis of the results from our previous study,5 we hypothesized that the pathCR rate would be at least 20%, that OS and disease-free survival (DFS) would depend on the degree of chemoradiotherapy-induced pathologic response, and that pretreatment parameters would not predict OS or DFS.
PATIENTS AND METHODS
Patients ineligible for inclusion were those with T4, M1, or T1N0 carcinoma, peritoneal carcinomatosis (gross or microscopic), or uncontrolled medical conditions (eg, diabetes, hypertension, New York Heart Association class III or IV, or psychiatric illness). Patients who could not comprehend or comply with the study were also ineligible.
Study Design
Patients were to receive one or two cycles of induction chemotherapy and then chemoradiotherapy. The second induction chemotherapy cycle was given if the cancer had not progressed, but if the cancer did progress, then this cycle was skipped and the patient began concurrent chemotherapy and radiotherapy. Patients were taken off protocol if distant metastases developed. Clinical responses were judged using radiographic or endoscopic techniques.
If a patient had an R0 resection, no further therapy was given. In the case of R1 resection (microscopically positive margin), R2 resection (only partial resection), or discovery of M1 carcinoma, the patient was given palliative care.
Step 1: Chemotherapy
Induction chemotherapy consisted of fluorouracil at 750 mg/m2/d by continuous infusion on days 1 to 5, cisplatin at 15 mg/m2/d intravenously on days 1 to 5, and paclitaxel at 200 mg/m2 intravenously over 24 hours on day 1. Patients received therapy in the outpatient setting with standard premedications and hydration. Daily subcutaneous granulocyte colony-stimulating factor was administered on days 6 to 16 (300 μg for patients ≤ 70 kg and 480 μg for patients > 70 kg). If the cancer had not progressed, then the second cycle was to be repeated 28 days after the first cycle began.
On the basis of the National Cancer Institute Common Toxicity Criteria version 2.0, drug doses were decreased by 25% if grade 3 nonhematologic toxicity or grade 4 hematologic toxicity occurred. CBC counts were performed weekly. Serum chemistries were monitored before each course.
Step 2: Chemoradiotherapy
The interval between day 1 of the last induction chemotherapy cycle and the first day of chemoradiotherapy was ≥ 28 days. Radiation fields included the entire stomach, perigastric extension (if present), and draining lymph nodes (gastric, celiac, porta hepatis, gastroduodenal, splenic-suprapancreatic, and retropancreaticoduodenal). For lesions involving the cardia or gastroesophageal junction, a 5-cm margin of esophagus was included, and for distal lesions at or near the gastroduodenal junction, a 5-cm margin of duodenum was included. Esophagoscopy, barium swallow radiographs, and computed tomography of the chest and abdomen were used to determine the maximum extent of disease relative to the primary tumor and the nodal groups. Idealized fields were modified as needed to shield at least two thirds of one kidney. For proximal lesions, cardiac shielding was recommended, along with evaluation of lateral fields for a component of treatment.
Linear accelerators delivered a dose of 45 Gy (25 fractions of 1.8 Gy) over 5 weeks using either 15- or 18-MV photons and a three-dimensional conformal radiotherapy technique. Typically, anterior and posterior fields were sometimes obliqued slightly to avoid the spinal cord and the right kidney were selected and delivered the majority of the dose.
The concurrent chemotherapy, given in an outpatient setting, consisted of fluorouracil at 300 mg/m2/d by continuous infusion with a portable pump 5 days each week (this treatment usually started on a Monday and ended on Friday, after the radiation therapy) plus paclitaxel at 45 mg/m2 intravenously each Monday for 5 weeks.
Step 3: Surgery
The type of surgery performed depended on the location and extent of the primary cancer. The cancer was resected along with a luminal gastric margin of ≥ 5 cm when feasible. Also when feasible, a 2-cm duodenal margin was obtained for distal cancers and a 3-cm esophageal margin was obtained for proximal cancers. In both cases, frozen-section confirmation of a negative margin was sought. For distal cancers, a subtotal gastrectomy was considered adequate; total gastrectomy was at the discretion of the surgeon. For proximal cancers, total gastrectomy or total esophagogastrectomy was performed. En bloc resection of adjacent organs was preformed when their involvement was questionable. The spleen was preserved when possible. An attempt was made to perform a D2-type nodal dissection. During surgery, a J-tube was placed for temporary nutritional support.
Tumor Response, Toxicity Criteria, and Data Management
Upper gastrointestinal barium radiographs were taken after each 28-day cycle of induction chemotherapy and just before surgery, and esophagogastroduodenoscopy, computed tomography of the abdomen, chest radiography, and all blood tests were repeated before surgery. Previously described criteria for response evaluation5,6 were used. Briefly, pathCR was defined as an absence of carcinoma cells in the primary site, and pathologic partial response (pathPR) was defined as less than 10% residual carcinoma cells in the specimen.
Follow-Up
Each patient was assessed at 3, 6, and 12 months and then every 6 months for 5 years or until the patient's death.
Statistical Methods
One of our hypotheses was that preoperative paclitaxel-based therapy results in a pathCR rate of 20%. Thus Simon's two-stage design was used: 21 patients were to be assessed first, and if more than one pathCR was observed, then 22 additional patients were to be added.
Patient responses (pathCR and any pathologic response [pathCR + pathPR]) were cross-tabulated by sex, cancer location, and baseline T and N stages. Baseline T-stage data were grouped into two categories, T1 plus T2, and T3. The Fisher's exact test was used to assess the association between each of these factors and the response rates. OS was defined as the time from the start of induction chemotherapy to the date of death or date of last follow-up. DFS was defined as the time from the start of induction chemotherapy to recurrence of the cancer. Data from living patients were censored. The log-rank test was used to test for differences in survival distributions by sex, cancer location, baseline and postoperative T and N stages, and T and N downstaging. Univariate Cox proportional hazards models were fit to yield estimates of hazard ratios for each of these factors. All statistical tests were two sided, and significance was set at P ≤ .05.
RESULTS
Response to Induction Chemotherapy
None of the patients had progression of the primary cancer during induction chemotherapy. Some subjective improvement in the ability to eat larger meals, reduction in postprandial discomfort or pain, and weight gain was reported by 49% of patients.
Response to Chemoradiotherapy
Further improvement in symptoms was universal on recovery (approximately 5 to 6 weeks) from chemoradiotherapy. Twenty-nine (71%) of the 41 patients underwent an EUS examination just before surgery. Of the 29 patients, 28 (69%) had T3 cancer and 16 (38%) had N+ cancer.
Surgical Findings and Surgical Pathology
Forty (98%) of 41 patients underwent surgery (one patient died before surgery). Thirty-three patients (80%) had a gastrectomy, and 32 patients (78%) had an R0 resection. Gastrectomy was not done in the remaining seven patients because M1 cancer (four with unsuspected peritoneal carcinomatosis) was discovered at surgery.
A pathCR was noted in eight patients (20%), and a pathPR was noted in six patients (15%). Thus the overall pathologic response rate was 35%. In the 33 patients who had a gastrectomy, the primary carcinoma was T3 in nine patients (27%), T2 in six patients (18%), T1 in nine patients (27%), and T0 in nine patients (27%). Twenty-two (67%) of the 33 patients who underwent gastrectomy had N0 cancer, eight patients (24%) had N1 cancer, and three patients (9%) had N2 cancer. The median number of nodes examined in the 33 gastrectomy specimens was 19 (range, one to 62 nodes). The median number of nodes with carcinoma was 3.5, and the median number of cancer-free nodes was 17.
Clinical Predictors of pathCR and pathPR
No clinical factor (sex, cancer location, baseline T stage, baseline N stage, or baseline EUS stage) was significantly associated with pathCR (Table 2) or any pathologic response (data not shown).
Determinants of Downstaging
T-stage downstaging (ie, a lower T stage after surgery than at baseline) was associated with baseline EUS stage (P ≤ .05) but not with sex, cancer location, or baseline N stage. Similarly, N-stage downstaging (ie, a lower N stage after surgery than at baseline) was associated with baseline EUS stage (P ≤ .05) but not with sex, cancer location, or baseline T stage).
Survival
At a median follow-up of 36+ months (range, 24 to 67 months), 28 (68%) of 41 patients were alive. Thus the median survival time for all 41 patients has not been reached. Among the 13 patients who died, one died before surgery as a result of acute myocardial infarction. The remaining 12 died as a result of their carcinoma.
We examined the effect of 13 factors on OS including sex; among these the following were statistically significantly correlated with OS: R0 resection (P < .001; Fig 1), pathCR (P = .02; Fig 2), any pathologic response (ie, pathCR plus pathPR; P = .006; Fig 3), postsurgery node positivity (P = .02), postsurgery N stage (P < .001), and postsurgery T stage (P = .01; Table 3).
Similarly, DFS was statistically significantly associated with T-stage downstaging (P = .005), pathologic response (pathCR plus pathPR; P = .003), pathCR (P = .008), R0 resection (P < .001), postsurgery node positivity (P = .002), postsurgery N stage (P < .001), and postsurgery T stage (P < .001).
OS Predicted by Postsurgery Outcome
Pretherapy EUS T stage (T1/2 v T3) had no significant influence on OS (P = .42; Fig 4A) but the post-therapy T stage had a highly statistically significant influence on patient survival (P < .001; Fig 4B). Similarly, pretherapy EUS N stage (N0 v N+) had no significant bearing on OS (P = .22; Fig 5A) but the post-therapy N stage had a highly statistically significant influence of OS (P < .001; Fig 5B). Similar influence of post-therapy T and N was seen on DFS (data not shown).
Chemotherapy-Induced Toxicity
No deaths were related to induction chemotherapy. The most common toxic effects were myelosuppression, nausea or vomiting, and fatigue (Table 4).
Chemoradiotherapy-Induced Toxicity
One patient died just before surgery of an acute myocardial infarction that was attributed to therapy. The most common chemoradiotherapy-related toxic effects were nausea or vomiting and fatigue (Table 5).
Surgical Complications
Two patients had a temporary anastomotic leak, and one patient had an abdominal abscess. There were no other major complications from surgery.
DISCUSSION
In the current trial, our goal was to investigate paclitaxel in the induction regimen and concurrently with radiation therapy. We hypothesized that paclitaxel-based chemoradiotherapy would result in a pathCR rate of at least 20% and that would, in turn, be associated with OS and DFS. We also hypothesized that preoperative parameters would not predict OS or DFS. Indeed, in this study, baseline T and N stages were not associated with OS or DFS. Our findings certainly do not diminish the value of adequate pretreatment staging, which is paramount in selecting patients in the clinical trials and will be a necessary stratification tool in future phase III trials. Nevertheless, our results suggest that clinical cancer staging, for all its sophistication, is suboptimal and requires improvement. It may be argued that clinical assessment does not reflect the true biology of the cancer and therapy only helps to unmask its character. However, it may also be argued that clinical stage is more valuable and predictive when surgery is the primary therapy, but once the clinical stage (pretreatment pathologic stage) is altered by ensuing preoperative therapy, it is the final pathologic stage (altered or unaltered) that is predictive of patient outcome. Our study, although small, points in that direction and, if true, then discovery of therapies that favorably alter the pretreatment pathologic stage is warranted.
We observed an R0 resection rate of 78%, and patients who had an R0 resection survived substantially longer than patients who did not. Whether this was due to patient selection or true effect of preoperative therapy can only be decided in a randomized trial. Paclitaxel-based preoperative chemoradiotherapy was also the subject of a recently completed Radiation Therapy Oncology Group study (RTOG 99-04) for patients with localized gastric cancer. However, orchestrating preoperative or postoperative therapy is complex and requires continuous participation by professionals from multiple disciplines and by informed caretakers and relatives of the patient. Although surgery remains the primary therapy for patients with localized cancer, multidisciplinary decisions improve the outcome as it results in advanced planning of interventions.
The results from our current study of 41 patients and our previous study of 33 patients5 indicate that pathCR may be a consistent result from preoperative chemoradiotherapy in patients with localized gastric carcinoma. The results from our current study also demonstrate that it is feasible to incorporate newer agents in this strategy. The contribution of paclitaxel cannot be determined by comparing sequential phase II studies, but the median survival beyond 36 months is intriguing. If one wished to increase the R0 resection rate and also the fraction of patients with pathologic response, then preoperative chemoradiotherapy may be one important approach. Induction chemotherapy that produces a high response rate could reduce the bulk of the primary cancer, and the ensuing chemoradiotherapy could be more effective against smaller cancer than it would be against a de novo bulky cancer. This could lead to a successful surgery. However, this notion remains hypothetical. Despite patient selection by laparoscopic staging, peritoneal cavity as a frequent failure site remains a concern.
There is limited information regarding pathologic response and its impact on patient outcome for localized gastric carcinoma. Our study suggests that it is possible to expect a pathCR rate of approximately 20% in a well-staged group of patients with localized gastric cancer. Once the pretreatment stage is altered, it becomes less predictive of outcome. The strategy of preoperative paclitaxel-based induction therapy and chemoradiotherapy is feasible. In the absence of phase III trials, the question remains whether one is altering the biologic behavior of gastric cancer with such maneuvers or simply characterizing it. Nevertheless, efforts to increase the rate of pathologic response and better control of metastases are warranted.
Authors' Disclosures of Potential Conflicts of Interest
NOTES
Supported in part by a grant from the Bristol-Myers Squibb Oncology Division and the Cantu, Dallas, and Capporella families.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Macdonald JS, Smalley S, Benedetti J, et al: Chemoradiotherapy after surgery compared with surgery alone for adenocarcinoma of the stomach or gastroesophageal junction. N Engl J Med 345:725-730, 2001
2. Wanebo HJ, Kennedy BJ, Chmiel J, et al: Cancer of the stomach: A patient care study by the American College of Surgeons. Ann Surg 218:583-592, 1993
3. Bonenkamp JJ, Hermans J, Sasako M, et al: Extended lymph node dissection for gastric cancer. N Engl J Med 340:908-914, 1999
4. Hundahl SA, Phillips JL, Menck HR: The National Cancer Data Base report on poor survival of US gastric carcinoma patients treated with gastrectomy. Cancer 88:921-932, 2000
5. Ajani JA, Mansfield P, Janjan N, et al: Multi-institutional trial of preoperative chemoradiotherapy in patients with potentially resectable gastric carcinoma. J Clin Oncol 22:2774-2780, 2004
6. Chiriec LR, Swisher SG, Ajani JA, et al: Posttherapy pathologic stage and residual carcinoma as predictor of survival in patients with esophageal or esophagogastric junction carcinoma treated with preoperative chemoradiation. Cancer (in press)
7. Lowy AM, Mansfield PF, Leach SD, et al: Laparoscopic staging for gastric cancer. Surgery 119:611-614, 1996
8. Pearlstone DB, Mansfield PF, Curley SA, et al: Laparoscopy in 533 patients with abdominal malignancy. Surgery 125:67-72, 1999
8. Ajani JA, Ilson DH, Daugherty K, et al: Activity of paclitaxel (Taxol) in patients with squamous cell carcinoma and adenocarcinoma of the esophagus. J Natl Cancer Inst 86:1086-1091, 1994
10. Ajani JA, Fairweather J, Pazdur R, et al: Phase II study of Taxol in patients with untreated metastatic gastric carcinoma. Cancer J Sci Am 4:269-274, 1998
11. Safran H, Wanebo HJ, Hesketh PJ, et al: Paclitaxel and concurrent radiation for gastric cancer. Int J Radiat Oncol Biol Phys 46:889-894, 2000
12. Ilson DH, Ajani JA, Bhalla K, et al: A phase II trial of paclitaxel, fluorouracil, and cisplatin in patients with advanced carcinoma of the esophagus. J Clin Oncol 16:1826-1834, 1998
13. Allum W, Cunningham D, Weeden S: Perioperative chemotherapy in operable gastric and lower oesophageal cancer: A randomized, controlled trial (the MAGIC trial, ISRCTN 9379371). Proc Am Soc Clin Oncol 22:249, 2003 (abstr 998)(J.A. Ajani, P.F. Mansfiel)