A double-blind, randomized, placebo-controlled, crossover trial of pentoxifylline for the prevention of chemotherapy-induced oral mucositis.

Oral mucositis is a frequent side effect of cancer therapy. No effective method of prophylaxis is currently available. We conducted a randomized, double-blind, placebo-controlled, crossover trial of pentoxifylline to evaluate its potential in preventing mucositis in cancer patients receiving chemotherapy. Ten cancer patients were randomized for treatment with a 15-day course of 400 mg of pentoxifylline given orally four times daily. Concurrent chemotherapy consisted of bolus cisplatin and infusional 5-fluorouracil. Mucositis was evaluated with the use of the Oral Assessment Guide developed at the University of Nebraska. Patients completing two cycles of chemotherapy--one with pentoxifylline and one with placebo--were evaluated for prophylaxis efficacy. Comparison of the oral assessment scores of the two cycles with a two-sided Student's t test failed to demonstrate a cytoprotective effect for pentoxifylline over placebo. We conclude that pentoxifylline as given in this study is ineffective for preventing mucositis in patients receiving cisplatin and 5-FU.

Cancer therapy and oral mucositis. An appraisal of drug prophylaxis.

Oral mucositis as a consequence of cytotoxic therapy is a major cause of morbidity in cancer patients. Cancer therapy-induced tissue damage leading to mucositis can occur through either direct or indirect stomatotoxicity. Once mucositis has occurred, treatment consists of measures to palliate symptoms. The prevention of cancer therapy-induced oral mucositis is less standardised. Numerous drugs have been used as prophylactic agents to prevent chemo- and radiotherapy-induced mucositis. Controlled trials have shown some degree of prophylactic efficacy for sucralfate, chlorhexidine and benzydamine. Positive but non-placebo-controlled trials requiring more study have been conducted with dinoprostone (prostaglandin E2), silver nitrate, beta-carotene, pentoxifylline and lozenges containing polymixin B, tobramycin and amphotericin B. Current studies have shown a lack of efficacy with allopurinol and granulocyte colony-stimulating factor (G-CSF). Nonpharmacological methods such as oral cryotherapy and helium-neon laser treatments have shown some promise. At the present time no agent has been shown to be uniformly efficacious and can be accepted as standard therapy. Additional studies combining several agents or incorporating nonpharmacological manoeuvres for mucositis prevention are needed.

Comparative evaluation of serum CA 195 and carcinoembryonic antigen in metastatic carcinoma.

BACKGROUND: Carcinoembryonic antigen (CEA) is a well-described human tumor-associated antigen most useful clinically in colon cancer. However, the clinical usefulness of CEA is limited by the marker's overall poor specificity and low sensitivity in patients with minimal disease. CA 195 is a recently discovered human tumor-associated glycoprotein that can be measured in serum using an immunoradiometric assay. CA 195 is expressed on the membrane of human colon cancer cells and shares an epitope with the Lewis A blood group antigens. The authors initiated a study to compare the clinical utility of serum CA 195 with CEA in patients with advanced cancer. A control population was studied to assess the effects of age, gender, alcohol, and tobacco on the measured levels of serum CA 195. METHODS: Using a solid-phase two-site immunoradiometric assay, serum CA 195 and CEA levels were measured in 71 control subjects and 167 patients with a prior diagnosis of cancer. The tumor histologic types included breast cancer, 49 patients; colon cancer, 38; prostate cancer, 24; lung cancer, 22; gastrointestinal noncolon cancer, 7; and miscellaneous, 27. Among patients with a history of cancer, 124 (74%) had active metastatic disease, and 43 (26%) were without evidence of active disease. The control population was composed of subjects without a history of malignancy. Clinical data collected from them included age, gender, smoking history, and alcohol consumption. RESULTS: In this laboratory, the normal ranges established for CA 195 and CEA in the control group were: 0.0-8.3 U/ml and 0.2-4.2 ng/ml, respectively. In the control subjects, the serum CA 195 level, unlike that of CEA, was not affected by age, gender, alcohol consumption, or tobacco use. In the study population, CA 195 had either equivalent or inferior specificity and sensitivity to CEA in all tumor types. A determination of the additive specificity and sensitivity of CA 195 and CEA did not significantly improve its clinical utility compared with CEA alone. However, CA 195 was significantly elevated in three patients with a prior history of colon cancer thought to be without evidence of active disease. Because all three of these patients had a relapse within the next 1-15 months, CA 195 might identify early relapses of colon cancer in some patients. CONCLUSIONS: Based on these results, it was concluded that CA 195 is not superior to CEA as an indicator of disease activity in advanced colon cancer or other solid tumors. However, studies utilizing CA 195 in the detection of early relapses of colon cancer may be warranted. A review of the English literature revealed that CA 195 might be a useful marker in pancreatic cancer.

Phase I study of low-dose cyclophosphamide and recombinant interleukin-2 for the treatment of advanced cancer.

We conducted a phase I study of low-dose cyclophosphamide and recombinant interleukin-2 (rIL-2) in 66 patients with advanced cancer resistant to standard therapy. All patients were evaluable for toxicity and 46 patients were evaluable for antitumor response. Patients evaluable for antitumor response included 23 with malignant melanoma, 10 with renal cell carcinoma, 4 with colon cancer, and 9 with various other solid tumors. All patients received i.v. cyclophosphamide (350 mg/m2) on day 1 followed by rIL-2 via 15 min i.v. infusion on days 4-8 and 11-15. The doses of rIL-2 ranged from 6.0 to 36.0 x 10(6) IU/m2. Each treatment cycle consisted of 21 days and a total of 113 cycles was administered. The number of treatment cycles administered per patient ranged from 1 to 8. The dose-limiting toxicities associated with rIL-2 included altered mental status, arthralgias, diarrhea, fatigue, fever, hypotension, nausea/vomiting, and peripheral edema. Twelve patients (18%) were removed from the study secondary to toxicity. Among the evaluable patients, 2 (4%) (malignant melanoma, renal cell carcinoma) developed a partial remission, 13 (29%) maintained stable disease, and 31 (67%) developed progressive disease. We conclude that the combination of low-dose cyclophosphamide and rIL-2 is tolerable in most patients but our data do not suggest an improved response rate for the combination vs. rIL-2 alone.

Immunophenotyping to detect and characterize acute lymphocytic leukemia in testicular biopsies.

This study establishes the utility of immunophenotyping testicular biopsy specimens in patients with acute lymphoid leukemia. The value of immunophenotyping in detecting or excluding leukemic testicular infiltration is demonstrated in six children with acute lymphocytic leukemia. A panel of monoclonal antibodies was employed on snap-frozen testicular biopsies, allowing both detection and immunologic characterization of four neoplastic lymphocytic infiltrates. Two samples were proven both histologically and phenotypically negative for leukemic infiltration. One of the four leukemic cases was clinically silent and might have escaped detection except for phenotyping. One leukemic infiltrate was also suspected to possess a multidrug-resistant phenotype (p-glycoprotein +); the latter possibility was excluded by an absence of reactivity with anti-p-glycoprotein monoclonal antibody. Thus, three clinically useful applications are demonstrated: (1) confirmation of testicular leukemic relapse, gaining assertion in histologically uncertain cases; (2) exclusion of clinically suspected disease relevant to cessation of therapy, and (3) detection/exclusion of drug-resistant phenotypes. Unexpectedly, we found expression of plasma cell-associated antigen in testicular germ cells, which may prove to be diagnostically useful in the future evaluation of germ cell tumors.

Liver biopsy immunotyping to characterize lymphoid malignancies.

To determine the feasibility of liver biopsy immunotyping to characterize hepatic lymphoid malignancies, we employed a panel of monoclonal antibodies on snap-frozen hepatic tissue from 18 patients. Six patients proved to have a histologic diagnosis of lymphoid malignancy. By using free avidin and biotin-blocking reagents to block endogenous biotin, followed by standard immunochemistry, immunotyping was successful in all six cases. Serial section typing allowed delineation of complex B cell phenotypes. Furthermore, architecture was preserved allowing discernment of disease patterns (e.g., sinusoidal, hairy cell leukemia vs. portal, follicular small-cleaved cell lymphoma. Unexpectedly, we found striking expression of common ALL antigen in normal bile canaliculi, which may prove of diagnostic or therapeutic relevance. This study establishes the utility of immunohistochemical techniques applied to hepatic biopsies as a valuable adjunct to histologic diagnosis as well as a tool in revealing the immunobiology of the liver.