Infections in the immunocompromised host: a prospective multicenter survey in patients receiving chemotherapy for acute leukemia.

Patients receiving chemotherapy for acute leukemia were prospectively followed up to determine the frequency, nature and outcome, of episodes of infection during a 6 or 12 month period at each of the participating centres. A total of 605 cycles of chemotherapy were surveyed. Of these, 490 cycles were received by patients with acute lymphoblastic leukemia (ALL) and 115 by patients with acute non-lymphoblastic leukemia (ANLL). 241 episodes of infection were recorded during the survey: 179 among ALL patients and 62 among patients of ANLL. Infections were more common during chemotherapy for ANLL than for ALL, occurring in 54% (62/115) and 36% (179/490) of chemotherapeutic cycle respectively. A favorable response to empiric antibacterial agents was seen in 39% (23/59) of episodes in ANLL patients and 77% (134/174) of episodes among ALL patients. Infection presented as fever of unidentified origin in an overwhelming majority (63%) of episodes. Gastroenteritis and pneumonia occurred with a frequency of 11% and 10% while the frequency of all other diagnoses was 3% or less. Overall, E coli and Candida were the most frequently isolated organisms while Staphylococcus aureus and Group A Streptococci were the most frequent isolates from blood and throat swabs, respectively. A high degree of resistance to commonly used antimicrobial agents was seen among the most frequently isolated organisms. About 75% of episodes of infection which did not respond to antibacterial agents responded to empiric systemic antifungal therapy; although fungi were mycologically isolated in only a quarter of these instances. Oropharyngeal candidiasis occurred in association with 3% of chemotherapeutic cycles.

Prognostic variables and survival in pediatric acute lymphoblastic leukemias: cancer institute experience.

This presentation is an analysis of front-end prognostic variables in achieving a complete response, a continuous complete remission, and disease-free survival in pediatric acute lymphoblastic leukemia at the Cancer Institute, Madras, India between 1983 and 1988. The clinical characteristics at presentation showed that virtually 100% of patients belong to the poor risk category, age < 3 years of > 6 years 72.2%, WBC > 10,000/mm3 59.8%, blast count > 50% 39.2%, organomegaly 91.8%, and L2 morphology 66.0%. All patients had more than one risk factor. Between 1983 and 1988, 97 children were treated on a pilot protocol designed in collaboration with the Lymphoma Biology Division of the Pediatric Oncology Branch of the National Cancer Institute, Bethesda, Maryland. The protocol was designed for a poor prognostic group. The significance of implicated poor prognostic factors was analyzed using the Cox proportional hazard model. Age at presentation was the only variable that emerged as an independent risk factor, and sex appeared to be a modifier. No other variables attained significance. Survival data were calculated by the Kaplan-Meier method. The relapse-free and event-free survivals up to 10 years were 50.7% and 38.1%, and compare reasonably well with results reported for similar groups elsewhere for the same period.

Immunophenotyping of acute lymphoblastic leukaemia in Madras, India.

At the Cancer Institute, Madras, India, we have performed immunophenotyping in 125 untreated cases of acute lymphoblastic leukaemia using a panel of 16 monoclonal antibodies and the avidin-biotin immunoperoxidase technique in a haematology autoanalyser (Technicon Hi system). Our results demonstrate a marked difference in the phenotypic pattern of ALL compared to Western countries, the predominant finding being a relative excess of T-ALL and a paucity of C-ALL cases. Age distribution of C-ALL reveals a peak at 2-6 years in paediatric ALL cases.

Antibiotics in febrile neutropenia: a randomized prospective comparison of two combinations.

BACKGROUND: Problems of initial empirical antibiotic therapy in febrile neutropenia are further complicated by other factors such as cost and the pattern of infective organisms in a particular institution. We, therefore, conducted a randomized study comparing the efficacy of two sets of antibiotics which differed in their spectrum of action, availability and price. METHODS: Sixty episodes of febrile neutropenia in 40 patients who were not on any prophylactic antibiotics were randomized into one of two arms--cefotaxime and gentamicin or ciprofloxacin and gentamicin. Depending upon the response by 72 hours, they were crossed over to the other arm or continued with the same combination. Empirical antifungal therapy was added in those who did not become afebrile. RESULTS: Infection was documented either clinically, bacteriologically or radiologically in 42% of the febrile episodes. The commonest organism isolated was Klebsiella and the commonest organism producing bacteraemia was the Staphylococcus. The temperature was reduced to normal without cross-over in 53% of the febrile episodes with cefotaxime and gentamicin and in 60% with ciprofloxacin and gentamicin (p > 0.05). After cross-over the temperature came down in 30% of the episodes with cefotaxime and gentamicin (initial combination) and 40% with ciprofloxacin and gentamicin (initial combination; p > 0.05). The overall response rate without empirical antifungal therapy was 83% in the patients on cefotaxime and gentamicin (initial combination; p > 0.05). While both the arms of the study had a 100% response rate, there was no significant difference between the efficacy of the antibiotic combinations. The ciprofloxacin-gentamicin combination is one-third as expensive as cefotaxime-gentamicin and is more readily available. CONCLUSION: We recommend the use of ciprofloxacin and gentamicin as the initial drug combination and cefotaxime and gentamicin only when the former is not effective.

Induction chemotherapy in non-metastatic high grade osteo sarcomas--results of pilot study at Cancer Institute (WIA), Madras.

Twelve patients with high grade osteosarcomas of the extremities were treated with two cycles of induction chemotherapy using adriamycin and cis-platinum and sandwich radiation between the two cycles (4000 rads). Ten patients underwent amputation or disarticulation, two patients had wide excision followed by endoprosthesis. The specimen was assessed for grade of necrosis. The Disease Free Survival at a minimum follow-up period of 26 months and median follow-up period of 35.5%. All the five patients who developed distant metastases had shown only a grade I necrosis in the tumour.

Chronic granulocytic leukaemia. A study of 160 cases.

Chronic granulocytic leukaemia (CGL) is the commonest leukaemia among adults in India. Case records of 183 CGL patients diagnosed between 1975 and 1985 were reviewed. The median age at diagnosis was 40.5 years. Most patients presented with weakness, fullness in the left upper abdomen and fever. Splenomegaly and hepatomegaly were present in 90% and 48% respectively. Patients were treated with oral, intermittent busulphan with monitoring of total leucocyte count. Overall, 87 patients expired, including 63 (72%) due to blast crisis. The median survival was 33 months from diagnosis and 44 months from the onset of symptoms.

Levels of detection of tumor cells in human bone marrow with or without prior culture.

With an in vitro culture technique combined with light microscopy, immunocytochemistry and molecular probing, we previously detected occult tumor cells in histologically-normal human bone marrow harvested for autologous transplantation. In this study, we mixed known numbers of malignant lymphoid (Raji and CEM) or breast cancer (MCF-7) cells with normal human bone marrow cells to determine the levels at which tumor cells can be detected before and after culture. Cytocentrifuge preparations were made before culture and after 2 or more weeks of culture and examined by light microscopy. We detected contaminating lymphoma cells at a level of more than 5% before culture, and at a level of 0.01% after culture for 2 or more weeks in 2% human lymphocyte conditioned medium. Before culture, we detected MCF-7 cells at a level of 0.001% using glucose oxidase immunocytochemical staining techniques; these cells were detected at a level of 0.00001% after culture. Since, of necessity, these calibrations rations were performed using cell lines, it is likely that these results overestimate the absolute sensitivity of these methods for detection of tumor cells in patient samples. We found the glucose oxidase immunocytochemical method more specific for detecting occult tumor cells in bone marrow than the immunoperoxidase staining method because of the absence of non-specific staining arising from endogenous peroxidase in bone marrow cells which makes the interpretation of the latter difficult. We conclude that culture techniques can increase the sensitivity of detection of occult tumor cells in human bone marrow about 100-fold.

Breast involvement in acute lymphoblastic leukaemia.

A 35 year old woman who was in clinical remission of acute lymphoblastic leukaemia on standard therapy presented with leukaemic involvement of both breasts. Local radiation therapy with concurrent systemic chemotherapy with vincristine, daunorubicin. L-asparaginase and prednisolone produced complete resolution of the breast lesions.

Ovarian granulocytic sarcoma in chronic granulocytic leukaemia with myeloblastic crisis.

Granulocytic sarcoma is an uncommon tumour composed of granulocytic precursor cells. A rare presentation of granulocytic sarcoma of the ovary occurring in a patient with chronic granulocytic leukaemia in blastic crisis is reported.