The author explains the commonality between both chemo regimes and why they may work equally well.
In the first study, which is well-known, intraperitoneal vs intravenous cisplatin and paclitaxel were compared. Intraperitoneal therapy showed a tremendously impressive extension of not only the progression-free interval, but also overall survival. The median progression-free survival was increased from 18.3 months to 23.8 months (P = .05) and the median survival difference was 49 months vs 65 months (P = .03), a difference of almost 14 months, which is a whopper in any kind of drug study.
Had this study been not on just a procedure, a technique of intraperitoneal devices, but on a drug of some kind, showing this degree of improvement in survival, it would have been an instantaneous hit and the drug probably would be used even without waiting for completion of an additional confirmatory study. There have been at least three randomized trials that have demonstrated superiority of intraperitoneal over intravenous delivery of platinum-based chemotherapy in patients with optimally debulked advanced ovarian cancer, but intraperitoneal therapy hasn’t really caught on.
There has always been a question in my mind as to why this should be the case. Then, about a year ago, the Japanese GOG reported results from what I consider to be a very well-designed phase III trial showing the benefit of dose-dense, once weekly paclitaxel. When compared with the patients assigned to the conventional regimen of paclitaxel and carboplatin, those on the dose-dense paclitaxel plus carboplatin regimen showed improvement of both the progression-free interval and overall survival. Progression-free survival was 28.0 months and 17.2 months, respectively (hazard ratio [HR] 0.71; 95% CI 0.58–0.88; P = 0.0015). The 3-year overall survival rate was 72.1% vs 65.1% (HR, 0.75; CI, 0.57–0.98; P = 0.03).
A very interesting correlation can be made concerning the possible mechanism of intraperitoneal therapy and also the weekly dosing investigated by the Japanese study. If we consider the reasons that intraperitoneal therapy works, it becomes obvious that it is not simply just the fact that the drug is delivered directly into the abdominal cavity, where most of the ovarian cancers reside. The major mechanism may, perhaps, be a depot, slow release effect by which the intraperitoneal drug, paclitaxel or cisplatin, actually circulates into the blood systemically over time. We know that paclitaxel, when given intraperitoneally, has a very, very long half-life, hence a “dose-dense” effect.
The dosing in the Armstrong study is actually similar to the dose-dense chemotherapy in the Japanese study that was given weekly. Armstrong and colleagues gave 80 mg of paclitaxel per week for 3 weeks. Carboplatin was given at the standard dose, AUC of 6. If we add up the three treatments of paclitaxel, the total dose was 240 mg, which is higher than what we typically use as a pulsed dose of 175 mg. So, by distributing the dose over time, we may have achieved a similar mechanistic effect to that of the intraperitoneal depot effect.
Of course, one of the issues here is that hematologic toxicity is frequently dose-limiting. Hematologic toxicity, particularly thrombocytopenia, is carboplatin-related. Somehow, with intraperitoneal delivery or dose-dense therapy, we can actually reduce the incidence of this toxicity.
Another issue is higher hematologic toxicity with intravenous chemotherapy than with intraperitoneal chemotherapy. The real question is whether we are able to use these two studies to advance the field by looking at a different regimen, as well as a more convenient way of administering therapy instead of intraperitoneally, by actually using intravenous therapy.
Currently, GOG is conducting a trial, GOG 252, in which they compare the different intraperitoneal dose-dense regimens and, then, just to do the fashionable thing, will add bevacizumab to all of the regimens, which will be followed by bevacizumab maintenance therapy. It does not directly compare intraperitoneal and intravenous delivery; nevertheless, we may get a useful result. In short, there is a significant contribution to be made by really looking at the therapy that we already have on hand. For example, in pediatric acute lymphoblastic leukemia, significant improvement in survival was achieved in the past several decades without the introduction of any new drugs.
There are several things to consider in these two studies. One common criticism of both studies is the high dropout rate. Of interest, both studies have very similar dropout rates. In fact, in the intraperitoneal study, only 42% of patients received all six cycles of chemotherapy and in the experimental arm of the Japanese study, 48% received all six cycles. However, if all the patients had completed the six-cycle treatment in either of these two studies, the result would have been even more dramatic. The high noncompletion rate was caused, mainly, by severe hematologic and neurologic toxicity, pain, and fatigue.
Another criticism is that the Japanese population might be different from others, so there may be ethnic or racial differences. For example, there is a difference in histologic distribution of tumors, with higher incidence of clear cell carcinoma in the Asian population; clear cell carcinoma carries a poorer prognosis. However, when the Japanese analyzed only the serous papillary tumors, they still observed an improvement in survival. In fact, in their study, using the dose-dense schedule did not show an obvious impact on outcomes in patients with clear cell carcinoma.
There are other differences in the inclusion criteria, such as the inclusion of stage II disease in the dose-dense trial, whereas, only stage III patients were studied in the intraperitoneal trial. There might also be variation in the surgical debulking techniques used. The GOG 252 trial probably will be one of the studies in which these questions can be answered.
There are other intriguing mechanisms involved. For example, paclitaxel, given at a low dose may have, in itself, an anti-angiogenesis effect, which may be synergistic with the chemotherapy regimen, especially when paclitaxel is given in a dose-dense manner. One wonders if this would explain why in the phase III GOG 218 trial, which looked at bevacizumab in women with advanced ovarian cancer, an improvement in progression-free survival was only observed in patients who received continued bevacizumab maintenance therapy. Women treated with carboplatin and paclitaxel, with or without bevacizumab, showed no difference in progression-free survival.4 The anti-angiogenesis effect may already be maximized by paclitaxel, such that the addition of bevacizumab to chemotherapy showed no significant improvement.
By better dosing, which could be a lower dose of both carboplatin and paclitaxel, the hematologic and other toxicities can be reduced. The kinetics may be more similar to that of the traditional low dose, continuous metronomic chemotherapy. Although the verdict is still out on whether these traditional therapies should be pushed aside in favor of biologics, consideration must be given to the side effects of the newer agents, especially longer term and also a serious issue relative to cost.
These two seemingly very different types of regimen may share a common mechanistic fundamental. The result of one study actually affirmed the other. There were two different methodologies, evaluated in very different patient populations, yet showing comparable superior survival curves. Further investigation of the old drugs with innovative dosing and scheduling, and also delivery, may be in order.