Douglass C. Tormey

Toxicity and response criteria of the Eastern Cooperative Oncology Group.

STANDARD CRITERIA FOR TOXICITY and for response to treatment are important prerequisites to the conduct of cancer trials. The Eastern Cooperative Oncology Group criteria for toxicity and response are presented to facilitate future reference and to encourage further standardization among those conducting clinical trials.

Prognostic effect of weight loss prior tochemotherapy in cancer patients

The prognostic effect of weight loss prior to chemotherapy was analyzedusing data from 3,047 patients enrolled in 12 chemotherapy protocols of the Eastern Cooperative Oncology Group. The frequency of weight loss ranged from 31 percent for favorable non-Hodgkins lymphoma to 87 percent in gastric cancer. Median survival was significantly shorter in nine protocols for the patients with weight loss compared to the patients with no weight loss. Chemotherapy response rates were lower in the patients with weight loss, but only in patients with breast cancer was this difference significant. Decreasing weight was correlated with decreasing performance status except for patients with pancreatic and gastric cancer. Within performance status categories, weight loss was associated with decreased median survival. The frequency of weight loss increased with increasing number of anatomic sites involved with metastases, but within categories of anatomic involvement, weight loss was associated with decreased median survival. These observations emphasize the prognostic effect of weight loss, especially in patients with a favorable performance status or a limited anatomic involvement with tumor.

Fluorouracil plus Levamisole as Effective Adjuvant Therapy after Resection of Stage III Colon Carcinoma: A Final Report

Carcinoma of the colon is one of the most common malignant diseases afflicting Western civilization. Most patients with this condition present with disease that is grossly completely resectable and in which any residual cancer is microscopic in nature. This is an ideal setting for reducing cancer mortality by adding adjuvant therapy to potentially curative surgery; pursuing the hope of such a reduction, clinical efforts at adjuvant therapy after surgery for colon cancer began more than 35 years ago and have involved numerous randomized trials enrolling several thousand patients treated with cytotoxic drugs, nonspecific immune stimulants, or various combinations thereof. Despite these efforts, no convincing evidence for the effectiveness of any regimen has been shown. Indeed, in a meta-analysis of all patients with colon cancer who were entered into trials of adjuvant therapy, Buyse and colleagues [1] found that the odds for dying were 8% higher for treated than for untreated patients. In 1989, a study by the North Central Cancer Treatment Group suggested that levamisole, a drug long used as an anthelmintic and presumed to have immunostimulatory activity [2], may be valuable. In this relatively small trial, levamisoleparticularly when used in combination with fluorouracilwas found to significantly reduce recurrence rates (P = 0.04) in patients with surgically treated stage II and stage III colorectal cancer. However, such therapy did not confer a statistically significant survival advantage, although the subset analysis suggested possible benefit for those patients with stage III disease. These results, although not convincing, were sufficiently intriguing to stimulate two large national intergroup trials. The first, which examined stage III disease, compared both postoperative levamisole alone and the postoperative combination of fluorouracil plus levamisole with surgery alone. The second trial, which examined stage II disease, compared only the postsurgical combination therapy with surgery alone. In 1990, we reported the early results of the stage III trial, which showed that patients treated with fluorouracil plus levamisole had a striking 41% reduction in the recurrence rate (P < 0.0001) and a 33% reduction in the mortality rate (P = 0.006) when compared with untreated controls [3]. On the basis of these results, a Consensus Panel convened by the National Institutes of Health recommended fluorouracil plus levamisole as standard therapy for patients with surgically treated stage III colon cancer [4]. Marketing of levamisole for this purpose was approved by the Food and Drug Administration. Our results, however, had been reported relatively early; the median follow-up time was only 3 years, and only a few patients had been followed for more than 5 years. Although the evidence for therapeutic benefit was strong, it was possible that we were only prolonging the interval to cancer recurrence and death rather than actually improving cure rates. In this report, we document our mature study with all patients able to be followed for more than 5 years. Methods Our methods have been described previously [3]. Within each of the three participating cooperative groups, patients with stage III (Dukes stage C) colon cancer who were fully recovered from surgery were grouped according to interval since surgery, depth of invasion of the primary tumor, and number of lymph nodes with metastasis. They were then randomly assigned to no further therapy, to treatment with levamisole alone, or to treatment with fluorouracil plus levamisole. Treatments could not be blinded, so no placebos were used. Levamisole was initiated 7 to 35 days after surgery at a dose of 50 mg administered orally three times/d for 3 days, repeated every 2 weeks for 1 year. Patients assigned to the combination therapy received the same regimen of levamisole plus fluorouracil at a dose of 450 mg/m2 body surface area, given by rapid intravenous injection daily for 5 days. Fluorouracil and levamisole were initiated simultaneously, 21 to 35 days after surgery. Twenty-eight days after the start of chemotherapy, weekly treatment with fluorouracil was begun at an intravenous dose of 450 mg/m2 body surface area and was continued for 48 additional weeks. Appropriate adjustments in dosage were made according to toxicity. Patients were followed with periodic medical examinations, which included blood counts, blood chemistries, chest radiographs, and imaging of the bowel using colonoscopy or proctoscopy plus radiography of the colon. Carcinoembryonic antigen assays were optional. Statistical analyses were done according to the procedures of the Statistical Analysis System [5]. Progression and survival curves were generated using the Kaplan-Meier method [6]; the log-rank statistic [7] was used to compare the distributions of survival times. The proportional hazards model [8] was used to determine the ratios of relapse and survival rates and to do all multivariate analyses. Backward regression was used to find the significant prognostic factors; variables were progressively eliminated on the basis of the maximum partial-likelihood statistics. To adjust for covariates when evaluating treatments, we kept the variable of treatment in the model and used backward regression for other covariates, keeping those whose maximum partial-likelihood statistics satisfied the criterion of P < 0.01. All P values reported are two-sided. The protocol was approved by the respective review boards of the participating institutions, and all patients gave written informed consent. Results A total of 971 patients were randomized in our stage III study. Forty-two (4.3%) of these were ineligible, in most cases because they had disease at a more advanced stage than that allowed by the protocol; these patients were excluded from the analysis. Fourteen patients refused to accept their treatment assignment. Because this withdrawal could be, and undoubtedly was, biased by treatment assignment, these patients were included in all analyses according to treatment assigned. At present, all patients entered into the study can potentially be followed for more than 5 years after surgery; the actual median follow-up time is 6.5 years. Ten patients have been lost to follow-up before 5 years but after a median time of 53 months (range, 9 to 59 months). As described in our earlier report [3], the clinical and pathologic characteristics of our patients were balanced among the study arms, except that more women were treated with fluorouracil plus levamisole, fewer patients receiving levamisole had tumors of the sigmoid colon and the rectosigmoid, and fewer patients receiving fluorouracil plus levamisole were 60 years of age or younger or had lesions invading adjacent organs. Cancer Recurrence Among the 315 patients assigned to no additional treatment, 177 have had proven recurrence. Among the 310 assigned to levamisole alone, 172 have had recurrence. In contrast, however, only 119 of the 304 patients assigned to fluorouracil plus levamisole have had recurrence. Figure 1 shows recurrence-free intervals according to treatment arm. Clearly, therapy with levamisole alone produced no benefit, whereas patients treated with fluorouracil plus levamisole have a highly significant advantage (P < 0.0001). These curves have no tendency to converge during long-term follow-up. Figure 1. Recurrence-free interval according to treatment arm. Figure 2 shows patterns of initial sites of recurrence according to treatment arm. It can be seen that recurrence in all common sites of metastasis was reduced in patients receiving fluorouracil plus levamisole when compared with untreated controls and with those receiving levamisole alone. It is noteworthy, however, that therapy had only a minimal effect on local or regional recurrence. Figure 2. Patterns of recurrence sites according to treatment arm. Table 1 shows the influence of patient and pathologic characteristics on recurrence. Depth of primary tumor invasion, number of metastatic lymph nodes, adhesion to or invasion of adjacent structures, regional implants, histologic differentiation, and preoperative carcinoembryonic antigen levels were all found to be determinants of recurrence (P < 0.01). After adjustment for the minor imbalances in prognostic variables among treatment arms, therapy with fluorouracil plus levamisole was again found to have an advantage over observation (40% reduction in recurrence rate; P < 0.0001). Levamisole alone had no detectable advantage (2% reduction in recurrence rate; P = 0.86). Table 1. Prognostic Factors for Recurrence Survival Four hundred forty-seven patients have died: 168 of the 315 patients assigned to observation only, 158 of the 310 receiving levamisole alone, and 121 of the 304 receiving fluorouracil plus levamisole. The survival curves shown in Figure 3 are similar to the recurrence curves. The survival pattern for patients receiving levamisole alone overlaps with that of the untreated control patients. Again, fluorouracil plus levamisole shows statistically and clinically significant advantages. The curves have no tendency to converge, which provides considerable evidence that cancer-related deaths have been prevented rather than simply delayed. These survival patterns were seen in each of the three cooperative groups. Figure 3. Survival according to treatment arm. The relations between patient or tumor characteristics and survival are shown in Table 1. Depth of primary tumor invasion, invasion of adjacent structures, regional implants, number of metastatic lymph nodes, histologic differentiation, and preoperative carcinoembryonic antigen level were each found to have prognostic significance (P < 0.01). After correction for the influence of prognostic factors through the use of a proportional hazards model, patients receiving fluorouracil plus levamisole were again found to have a significant survival advantage when compared with patients assi

Annual hazard rates of recurrence for breast cancer after primary therapy.

PURPOSE To determine if the long-term increase of recurrence for breast cancer is stable or slowly decreasing, or if it ever reaches zero; and to determine the effect of prognostic factors on the hazard of recurrence. METHODS All patients entered onto the seven completed and unblinded Eastern Cooperative Oncology Group (ECOG) coordinated studies of postoperative adjuvant therapy for breast cancer were analyzed in terms of annual hazard of recurrence of breast cancer. RESULTS For the entire group, the peak hazard of recurrence occurred in the interval of 1 to 2 years. The hazard decreased consistently in the interval of 2 to 5 years. Beyond 5 years, the hazard of recurrence decreased very, very slowly through year 12. The average hazard of recurrence between years 5 and 12 for the entire population was 4.3% per year. The pattern of a peak hazard of recurrence during the first 5 years with a slowly decreasing hazard of recurrence beyond 5 years was also observed to varying degrees in most subsets. Higher risk subsets such as patients with more than three nodes positive had a higher hazard of recurrence at all time intervals, while lower risk subsets such as patients with negative nodes had a lower hazard of recurrence in all time periods. CONCLUSION Patients 5 years postsurgery for breast cancer appear to have a very slowly decreasing hazard of recurrence. The mean hazard of recurrence between years 5 to 12 postsurgery is 4.3% per year. This group of patients may be well suited for trials evaluating cytostatic drugs or differentiating agents.

HER-2/neu in node-negative breast cancer: prognostic significance of overexpression influenced by the presence of in situ carcinoma.

PURPOSE Amplification and/or overexpression of the HER-2/neu oncogene have been shown to correlate with poor clinical outcome in patients with axillary node-positive breast cancer. In contrast, the prognostic significance of HER-2/neu in node-negative disease is controversial. This study was undertaken to evaluate further the relationship between HER-2/neu and clinical outcome in node-negative disease. PATIENTS AND METHODS Overexpression of HER-2/neu was evaluated by permanent-section immunohistochemistry in tumors from 613 patients with long-term clinical follow-up enrolled in the Intergroup Study 0011. Patients were stratified into low-risk (n = 307) and high-risk (n = 306) groups on the basis of tumor size and estrogen-receptor (ER) status. Low-risk patients were defined as having small (less than 3 cm), ER-positive tumors and were observed without additional treatment after initial surgery. High-risk patients had either ER-negative or large (greater than or equal to 3 cm), ER-positive tumors and were randomized to be observed (n = 146) or to receive adjuvant chemotherapy (n = 160) after surgery. RESULTS The rate of HER-2/neu overexpression was 14.3% in all tumors combined and was higher in invasive carcinomas with (21.5%) than without (11.2%) a significant noninvasive or in situ histologic component (P less than .0001). There was no relationship between overexpression and clinical outcome in the natural history setting of combined low-risk and high-risk patients not receiving adjuvant therapy (n = 453). Based on the reasoning that the influence of HER-2/neu may have been obscured by high-risk features and/or the presence of noninvasive carcinoma, we also analyzed the subset of patients with low-risk lesions not containing a significant in situ component (n = 179). Patients of this group with HER-2/neu-positive tumors showed only 40% disease-free survival (DFS) at 5 years, compared with over 80% in patients with HER-2/neu-negative tumors (P less than .0001). A similar inverse correlation was observed between overexpression and overall survival in the same group of patients (P = .0001). In a separate analysis involving patients receiving adjuvant chemotherapy, those with HER-2/neu-negative tumors showed significantly improved DFS in response to therapy compared with patients with HER-2/neu-positive tumors. CONCLUSION Overexpression of HER-2/neu is associated with poor clinical outcome in a subset of node-negative patients with small, ER-positive, predominantly invasive tumors and may play a role in resistance to adjuvant chemotherapy.

Locoregional Failure 10 Years After Mastectomy and Adjuvant Chemotherapy With or Without Tamoxifen Without Irradiation: Experience of the Eastern Cooperative Oncology Group

PURPOSE To assess patterns of failure and how selected prognostic and treatment factors affect the risks of locoregional failure (LRF) after mastectomy in breast cancer patients with histologically involved axillary nodes treated with chemotherapy with or without tamoxifen without irradiation. PATIENTS AND METHODS The study population consisted of 2,016 patients entered onto four randomized trials conducted by the Eastern Cooperative Oncology Group. The median follow-up time for patients without recurrence was 12.1 years (range, 0.07 to 19.1 years). RESULTS A total of 1,099 patients (55%) experienced disease recurrence. The first sites of failure were as follows: isolated LRF, 254 (13%); LRF with simultaneous distant failure (DF), 166 (8%); and distant only, 679 (34%). The risk of LRF with or without simultaneous DF at 10 years was 12.9% in patients with one to three positive nodes and 28.7% for patients with four or more positive nodes. Multivariate analysis showed that increasing tumor size, increasing numbers of involved nodes, negative estrogen receptor protein status, and decreasing number of nodes examined were significant for increasing the rate of LRF with or without simultaneous DF. CONCLUSION LRF after mastectomy is a substantial clinical problem, despite the use of chemotherapy with or without tamoxifen. Prospective randomized trials will be necessary to estimate accurately the potential disease-free and overall survival benefits of postmastectomy radiotherapy for patients in particular prognostic subgroups treated with presently used and future systemic therapy regimens.

Venous and arterial thrombosis in patients who received adjuvant therapy for breast cancer.

The records of 2,673 patients randomized according to seven consecutive Eastern Cooperative Oncology Group (ECOG) studies of adjuvant therapy for breast cancer were reviewed for the occurrence of vascular complications. All protocols opened and closed between June 1977 and July 1987. The objectives of the present study were (1) to compare the frequency of vascular complications among patients who received adjuvant therapy for breast cancer with patients on observation, and (2) to estimate the contribution of chemotherapy and hormonal therapy to the occurrence of venous and arterial thrombi. The frequency of thrombosis, both venous and arterial combined, was 5.4% among patients who received adjuvant therapy and was 1.6% among patients on observation (P = .0002). Premenopausal patients who received chemotherapy and tamoxifen had significantly more venous complications than those who received chemotherapy without tamoxifen (2.8% v 0.8%, P = .03). Postmenopausal patients who received tamoxifen and chemotherapy had significantly more venous thrombi than those who received tamoxifen alone (8.0% v 2.3%, P = .03) or those who were observed (8.0% v 0.4%, P less than .0001). Premenopausal patients who received tamoxifen and chemotherapy had a 1.6% frequency of arterial thrombosis, significantly more than patients who received chemotherapy alone (1.6% v 0.0%, P = .004). The frequency of arterial thrombosis among postmenopausal patients was not significantly correlated with adjuvant therapy. In conclusion, patients who received adjuvant therapy for breast cancer had a 5.4% frequency of thromboembolic complications, significantly more than those who were observed. The combination of chemotherapy and tamoxifen was associated with more venous and arterial thromboembolic complications than chemotherapy alone in premenopausal patients and with more venous thrombi than tamoxifen alone among postmenopausal patients.

A randomized comparative trial of adriamycin versus methotrexate in combination drug therapy

A prospective randomized trial was conducted comparing the clinical response of 78 previously untreated patients with advanced metastatic breast cancer to a combination of cyclophosphamide, methotrexate, and 5‐fluorouracil (CMF) or to a combination of cyclophosphamide, adriamycin, and 5‐fluorouracil (CAF). Sixty‐two percent of the patients receiving CMF responded to treatment compared to an 82% response rate for the patients receiving CAF. Although within acceptable limits, hematologic and GI toxicity was greater with CAF. There was no significant difference in the duration of response to the two regimens. Therefore, the therapeutic difference between the two therapies is a higher initial response rate to the adriamycin containing regimen.

Identification of a subgroup of patients with breast cancer and histologically positive axillary nodes receiving adjuvant chemotherapy who may benefit from postoperative radiotherapy.

Risk factors for isolated local-regional (LR) recurrence following mastectomy for breast cancer were analyzed in a review of 627 women entered into Eastern Cooperative Oncology Group (ECOG) adjuvant chemotherapy trials between 1978 and 1982. Premenopausal patients were randomized to cyclophosphamide, methotrexate, and fluorouracil (5-FU) (CMF), cyclophosphamide, methotrexate, 5-FU, and prednisone (CMFP), or cyclophosphamide, methotrexate, 5-FU, prednisone, and tamoxifen (CMFPT). Postmenopausal patients were randomized to observation, CMFP, or CMFPT. Median follow-up time was 4.5 years. At 3 years, 225 patients relapsed and in 70 (31% of failures, 11% of all patients) the initial site was LR without distant metastases. In a multivariate analysis, the risk of an isolated LR recurrence significantly correlated with the number of positive axillary nodes, the primary tumor size, the presence of tumor necrosis, and the number of axillary nodes examined. Factors that significantly discriminated between an isolated LR recurrence and distant metastasis were the number of positive nodes and primary tumor size. Patients with four to seven positive nodes or tumor size greater than or equal to 5 cm had a chance of developing an isolated LR recurrence almost equal to the risk of distant metastases. These findings suggest a potential for improved survival in this subset of patients with the addition of postmastectomy radiation to chemotherapy, and continue to emphasize the presence of a group of patients at high risk for isolated LR recurrence despite adjuvant chemotherapy.

Bone mineral density in women with breast cancer treated with adjuvant tamoxifen for at least two years

SummaryWhile in limited animal studies tamoxifen is reported to protect against loss of bone mineral, data in humans are lacking. We measured bone mineral density (BMD) using single photon absorptiometry at the radius and dual photon absorptiometry at the lumbar spine in breast cancer patients treated with chemotherapy at our institution. In this group, 37 women were not treated with tamoxifen (NT) and 48 women were treated with tamoxifen (T) for at least two years. Younger age, greater weight and height, premenopausal status, and shorter time since menopause were found to be significant predictors of greater BMD. Tamoxifen-treated women had been postmenopausal for more years (p = 0.012). Regression analyses used to adjust for differences in risk of bone loss did not reveal significant differences in BMD between the two groups of women. For the postmenopausal women (27 NT and 34 T subjects), the adjusted mean BMD (g/cm2) at the spine was 1.11 (NT), 1.11 (T) (p = 0.93); and at the radius 0.63 (NT), 0.62 (T) (p = 0.30). This limited retrospective study suggests that tamoxifen does not have ‘anti-estrogenic’ effects on BMD.