Jerry Glenn

From July 1975 to December 1982, 563 patients were referred to the Surgery Branch of the National Cancer Institute with the diagnosis of soft-tissue sarcoma. Three hundred and seven of these patients had fully resectable, localized high-grade soft-tissue sarcomas and were treated at the National Cancer Institute using standard protocols with surgery alone, or in combination with chemotherapy and/or radiotherapy. An aggressive surgical approach was undertaken in the management of patients who subsequently developed recurrent disease. These 307 cases have been reviewed, with a median duration of follow-up of 30 months, to determine the frequency of recurrent disease, the patterns of recurrence, and the impact of surgery on the survival of patients who developed recurrent disease. Disease recurred in one hundred seven patients (107/307, 35%), with a median disease-free interval of 18 months (range, 0.5 to 72.0 months). The frequency of recurrence by site of primary sarcoma was extremity, 31% (65/211); head and neck, 33% (4/12); trunk, 40% (17/42); retroperitoneum, 47% (17/36); and breast, 67% (4/6). Isolated pulmonary metastatic disease was the most common pattern of initial recurrence (56/107, 52%) followed by isolated local recurrence (21/107, 20%). Single other sites of recurrence and multiple concurrent sites of recurrence each accounted for 14% (15/107) of all initial recurrences. The relative frequency of each of these four patterns of recurrence varied with the site of the primary sarcoma. The outcome for patients with recurrent disease depended on the site of recurrence, rather than on the site of the primary sarcoma. Sixty-six patients (66/107, 62%) with recurrent disease were rendered surgically disease-free with the first recurrence, including 40 (40/56, 72%) patients with isolated pulmonary metastases, 20 patients (20/21, 96%) with isolated local recurrences, five patients (5/15, 33%), with isolated other sites of recurrence and one patient (1/15, 7%) with multiple sites of initial recurrence. Following surgical resection, the actuarial three-year survival for the 66 patients rendered disease-free was 51%. The median survival for the 41 patients not rendered surgically disease-free with the first recurrence was only 7.4 months. Thirty of the sixty-six patients (30/66, 45%) rendered disease-free with the first recurrence remained disease-free at follow-up, with a median follow-up of 28 months from the time of resection of the first recurrence. The remaining 36 patients (36/66, 55%) subsequently recurred, with a median disease-free interval of 7.3 months.(ABSTRACT TRUNCATED AT 400 WORDS)

Thirty-seven patients with resectable retroperitoneal sarcomas were studied prospectively to determine the efficacy of aggressive multimodality treatments. No patients was lost to follow-up, which ranged from 11 to 85 months (median 29 months). All patients received radiotherapy and some received postoperative chemotherapy (doxorubicin, cyclophosphamide, and high-dose methotrexate). A subset of 15 patients were entered into a prospective, randomized study testing the efficacy of adjuvant chemotherapy (eight received chemotherapy; seven did not). Two-year actuarial survival rates were inferior in the chemotherapy arm (100% versus 47%; p = 0.06), but the small number of patients precluded drawing definitive conclusions from this randomized study alone. Among the entire 37 patients (21 received chemotherapy; 16 did not) the actuarial 3-year survival rate was 43% and appeared unaffected by chemotherapy. Two patients suffered doxorubicin infiltration, three sustained cardiac toxicity, two developed cyclophosphamide-induced cystitis, and three withstood transient, severe bone marrow suppression. Eight patients suffered severe radiation enteritis, and one patient died after bowel resection for this problem. Thus the chemotherapy regimen we administered did not appear to improve survival but was associated with major morbidity. Radiotherapy was also associated with major complications, and since all patients received radiotherapy, it remains to be established if this modality is beneficial in improving survival.

Since 1977, 31 patients were entered in a randomized, prospective study testing the efficacy of adjuvant chemotherapy after aggressive local treatment of high-grade sarcomas of the head, neck, breast, and trunk (excluding retroperitoneal sarcomas). All patients had complete resection of gross tumor and underwent postoperative radiotherapy (6000-6300 rads over 7-8 weeks). Seventeen patients received adjuvant chemotherapy consisting of doxorubicin (less than or equal to 550 mg/m2), cyclophosphamide (less than or equal to 5500 mg/m2), and methotrexate (less than or equal to 1000 mg/kg). Three-year actuarial disease-free survival in the chemotherapy arm was 77%, compared to 49% in the no-chemotherapy arm (P = 0.075). Three-year overall actuarial survivals in the two treatment arms, however, were 68% and 58%, respectively (P = 0.38). Considering only patients with tumors of the trunk (22 patients), 3-year actuarial disease-free survival in the chemotherapy arm was 92%, compared to 47% in the no-chemotherapy arm (P = 0.006). Actuarial 3-year overall survival in the chemotherapy arm was 82%, compared to 61% in the no-chemotherapy arm (P = 0.18). An additional 26 patients were treated in an identical fashion, but were not part of the randomized trial because of contraindications to chemotherapy, refusal to enter the randomized trial, or because they were treated before 1977 in a trial in which all patients received chemotherapy. Considering the entire group of 57 patients, follow-up ranged from 10 to 86 months (median, 35 months). Local control was achieved in 46 patients (81%); 3-year actuarial disease-free and overall survivals were 67% and 77%, respectively. A tendency toward improved disease-free survival was apparent among patients treated with chemotherapy (P = 0.018), but there was no statistically significant improvement in overall actuarial survival (P = 0.46). The subgroup of patients with sarcomas of the trunk (39 patients) demonstrated the greatest benefit from chemotherapy, with regard to disease-free survival (P less than or equal to 0.001). The most significant toxicity associated with chemotherapy was doxorubicin-induced cardiomyopathy, which resulted in clinically apparent congestive heart failure in five patients. Thus, the use of chemotherapy when combined with aggressive local measures appears to improve disease-free survival, but additional patients and longer follow-up are necessary to determine if improved overall survival will result.

By radioimmunoassay we determined circulating levels of a tumor-associated antigen, CA 19-9, in 47 patients with pancreatic adenocarcinoma, to learn if serial testing was useful in predicting prognosis or in detecting disease progression. Before treatment, 42 (89%) had an abnormal serum level, and 45 (96%) had an abnormal level at some time during the disease course. A pretreatment value of less than 1,000 U/mL (normal, less than or equal to 37 U/mL) was found in 38 patients; 20 (53%) had resectable disease. One of nine patients (11%) with a pretreatment value greater than 1,000 U/mL had resectable disease (P2 = .05). Among 14 patients who underwent pancreatectomy and were studied serially, the CA 19-9 level normalized in eight; seven (88%) survived greater than or equal to 18 months. Six patients whose levels did not normalize after pancreatectomy all died in less than 12 months (P2 less than .005). Greatly elevated levels occurred in 11 patients after pancreatectomy 1 to 7 months before clinically apparent recurrence. The other three patients without significant elevations remain clinically free of disease. The data suggest that serial determination of serum CA 19-9 levels are useful as a prognostic indicator and in detecting disease recurrence following pancreatectomy. Concurrent determinations of carcinoembryonic antigen (CEA) levels showed abnormal preoperative values in 28 of 46 patients tested (61%). Concurrent serial postoperative determinations of CEA were available in ten patients. Whereas CA 19-9 values clearly indicated eight recurrences, CEA was helpful in only four. In this small group of patients, CA 19-9 was a better predictor of recurrence.