A phase I dose-finding study of a combination of pegylated liposomal doxorubicin (Doxil), carboplatin and paclitaxel in ovarian cancer.

Standard chemotherapy for advanced epithelial ovarian cancer is a combination of platinum-paclitaxel. One strategy to improve the outcome for patients is to add other agents to standard therapy. Doxil is active in relapsed disease and has a response rate of 25% in platinum-resistant relapsed disease. A dose finding study of doxil-carboplatin-paclitaxel was therefore undertaken in women receiving first-line therapy. Thirty-one women with epithelial ovarian cancer or mixed Mullerian tumours of the ovary were enrolled. The doses of carboplatin, paclitaxel and doxil were as follows: carboplatin AUC 5 and 6; paclitaxel, 135 and 175 mg m(-2); doxil 20, 30, 40 and 50 mg m(-2). Schedules examined included treatment cycles of 21 and 28 days, and an alternating schedule of carboplatin-paclitaxel (q 21) with doxil being administered every other course (q 42). The dose-limiting toxicities were found to be neutropenia, stomatitis and palmar plantar syndrome and the maximum tolerated dose was defined as; carboplatin AUC 5, paclitaxel 175 mg m(-2) and doxil 30 mg m(-2) q 21. Reducing the paclitaxel dose to 135 mg m(-2) did not allow the doxil dose to be increased. Delivering doxil on alternate cycles at doses of 40 and 50 mg m(-2) also resulted in dose-limiting toxicities. The recommended doses for phase II/III trials are carboplatin AUC 6, paclitaxel 175 mg m(-2), doxil 30 mg m(-2) q 28 or carboplatin AUC 5, paclitaxel 175 mg m(-2), doxil 20 mg m(-2) q 21. Grade 3/4 haematologic toxicity was common at the recommended phase II doses but was short lived and not clinically important and non-haematologic toxicities were generally mild and consisted of nausea, paraesthesiae, stomatitis and palmar plantar syndrome.

Pursuit of optimum outcomes in ovarian cancer: methodological approaches to therapy.

The treatment of ovarian cancer is an evolving area and important clinical questions remain unanswered at all stages from early disease to relapse. This review outlines current practice at each disease stage, some of the current unanswered questions and issues surrounding the design of clinical trials to answer these questions. The gold standard test for new treatments must remain the randomised controlled trial with survival as the major endpoint because other outcome measures such as radiological response do not bear a strong relationship to survival. Patient factors greatly influence the likelihood of response to treatment and subsequent survival, hence failure to control for these in trial design may lead to spurious results. Quality of life is an important endpoint but quality-of-life measures in clinical trials should be interpreted carefully. The introduction of novel, target directed anticancer therapies will require new study designs as the phase I/II/III paradigm may not be relevant. It is current practice to offer adjuvant chemotherapy to women with early stage disease who are considered at high risk of relapse despite conflicting evidence from clinical trials. Current questions include the optimal choice of regimen and the duration of treatment. However, the relative rarity of early stage disease and the likely small difference between treatments makes evaluations difficult. Advanced disease is currently treated using a combination of surgical cytoreduction and platinum-paclitaxel chemotherapy. Women with poor risk disease are unlikely to be cured of their disease and the investigation of strategies to minimise treatment may be appropriate. Conversely, women with good risk disease may be better candidates for experimental treatment to increase cure rate. Strategies that have been tried include dose-intensification, high-dose therapy and intraperitoneal therapy. Whereas there is some evidence to support the latter, there is no current evidence for dose-intensification or high-dose therapy, and these must remain areas of investigation. Most current trials investigate the addition of agents to platinum-paclitaxel. Relapsed disease is an important area. Despite this, only 9 randomised controlled trials have been undertaken. Uncertainties exist in the role of surgery, both surgical cytoreduction and palliative surgery. The mainstay of treatment at disease relapse is chemotherapy and the choice of agent revolves around the concept of platinum sensitivity. Many agents are active in platinum-resistant disease, but uncertainties remain about the relative efficacies of each and the place of combination therapy.

Cryptococcal spondylitis: solitary infective bone lesions are not always tuberculous.

Tuberculous spondylitis is a well-recognized cause of back pain and vertebral collapse due to infection and with must not be overlooked even if it is not the most likely diagnosis. If a patient, particularly one of Asian origin, were to present with a solitary destructive bone lesion, without evidence of myeloma or other malignancy, a trial of anti-tuberculous chemotherapy would be one therapeutic approach, even if there was no evidence of tuberculosis elsewhere. However, failure to biopsy the bone lesion and undertake the appropriate microbiology could lead to other important diagnoses being missed. This is illustrated by the case which we report below.

Red cell volume distribution curves in the diagnosis of glomerular and non-glomerular hematuria.

It has been suggested that measurement of red blood cell (RBC) volume, by red cell analyzers (RCA), may be used to distinguish glomerular from non-glomerular hematuria. The RCA measures all urinary particles, including RBC. We found that red cell volume distribution curves (RVDC) obtained with a Coulter S+IV RCA were non-specific in predicting the cause of hematuria. Normal RBC added to urine at about 15,000/mm3 produced the expected non-glomerular RCVDC but at decreasing RBC concentrations mixed and then glomerular RCVDCs were found. Urine from 7 normal subjects (6 without urinary RBC) showed particle volume distribution curves identical to glomerular RCVDC. These urinary particles were examined by scanning electron microscopy (SEM) after filtration (3 microns and 0.8 micron pore size). Hematuric urine from a patient after prostatectomy contained RBC (3.5-9.5 microns diameter) and particulate debris (1.0-7.5 microns diameter). Similar results were obtained for a patient with crescentic glomerulonephritis (particles 1-13 microns diameter). Attempts to separate glomerular urinary RBC from debris by centrifugation against a sucrose polymer were unsuccessful. The shape of the RCVDC obtained with RCA is altered by the presence of debris in urine. The true MCV of glomerular RBC is still unknown. The usefulness of RCA in the diagnosis of hematuria is limited to patients with large numbers of urinary RBC.