[CITATION][C] Trastuzumab in CSF

BC Pestalozzi, S Brignoli - Journal of Clinical Oncology, 2000 - ascopubs.org
BC Pestalozzi, S Brignoli
Journal of Clinical Oncology, 2000ascopubs.org
To the Editor: The randomized clinical trial in ovarian cancer conducted by the Gynecologic
Oncology Group (GOG132) 1 recently published has provoked much interest. The most
provocative finding is that study results seem to suggest that cisplatin alone at 100 mg/m2 is
as effective as cisplatin at 75 mg/m2 combined with paclitaxel at 175 mg/m2 over 3 hours.
However, as the authors and others have pointed out, there is some difficulty interpreting the
results of this trial, because approximately one half of the patients on each arm received …
To the Editor: The randomized clinical trial in ovarian cancer conducted by the Gynecologic Oncology Group (GOG132) 1 recently published has provoked much interest. The most provocative finding is that study results seem to suggest that cisplatin alone at 100 mg/m2 is as effective as cisplatin at 75 mg/m2 combined with paclitaxel at 175 mg/m2 over 3 hours. However, as the authors and others have pointed out, there is some difficulty interpreting the results of this trial, because approximately one half of the patients on each arm received some nonprotocol treatment before clinical progression, and it is suggested that this additional treatment is a crucial factor in the results. Closer inspection of these data reveals that of those patients randomized to receive single-agent cisplatin, 26% received paclitaxel before progression. This undoubtedly makes interpretation difficult. However, given the large benefit for this cisplatin/paclitaxel regimen demonstrated in previous trials, 2, 3 we find it difficult to believe that the use of paclitaxel in just 26% of patients allocated to receive singleagent cisplatin is sufficient explanation for the equivalent (or perhaps even slightly better) survival and progression-free survival observed for this group as a whole compared with the 100% of patients receiving combined paclitaxel and cisplatin as first-line treatment. One explanation may be that cisplatin at 100 mg/m2 is in fact as effective as cisplatin 75 mg/m2 plus paclitaxel 175 mg/m2. The authors consider this explanation, but in this eventuality they consider that because of the toxicity profiles, the combination of cisplatin and paclitaxel is preferred.
However, again taking an alternative view, there is good evidence that carboplatin and cisplatin are equally effective. In particular, three trials have compared single-agent cisplatin at 100 mg/m2 with singleagent carboplatin at doses of 350 to 400 mg/m2, and the results of these trials, both individually and combined, suggest that these regimens are equivalent. 4 Although it is difficult to translate carboplatin dosed according to body surface area to the more modern area under the concentration-time curve method, doses of approximately five areas under the concentration-time curve give an average body surface area dose of approximately 350 mg/m2. Unlike cisplatin, carboplatin at these sorts of doses is extremely well tolerated. It thus seems possible that single-agent carboplatin could prove to be as good as the combination of paclitaxel plus cisplatin. This is of course only a hypothesis. There are trials that will report further results in the next year or so that may assess the validity of this hypothesis. These include the three trials comparing cisplatin plus paclitaxel with carboplatin plus paclitaxel, which have randomized some 2,000 patients in total, 5-7 and a trial comparing single-agent carboplatin or cyclophosphamide, doxorubicin, and cisplatin with carboplatin plus paclitaxel, which also includes some 2,000 patients. 8 Further results from these trials are eagerly awaited.
ASCO Publications