Metronomic dosing — long-term, low-dose, frequent administration of chemotherapeutic drugs — reduces the toxic side effects of traditional chemotherapy. Rather than directly killing cancer cells, it prevents blood-vessel formation by blocking endothelial-cell growth. Why the endothelial cells of new blood vessels are specifically targeted by this dosing strategy is a mystery that Guido Bocci, Robert Kerbel and colleagues are on the way to solving — the angiogenesis inhibitor thrombospondin-1 (TSP1) thinks less is best.

Bocci et al. began by analysing gene-expression profiles of microvascular endothelial cells and found that long-term exposure to the antitumour agent BAL-9504 increased TSP1 expression. Further in vitro investigations confirmed this observation, showing that TSP1 expression increased in the drug-treated cells, which also secreted the protein into the culture medium. The effects of the metronomic chemotherapy — blocked proliferation and reduced cell survival — were partially reversed by TSP1-neutralizing antibodies, indicating that TSP1 regulates the metronomic dosing response in vitro. But does it have a similar effect in vivo?

They administered a previously well-characterized, low-dose, daily cyclophosphamide treatment to wild-type and Tsp1-null mice and assessed in vivo angiogenesis. After 7 days, the cyclophosphamide treatment had significantly reduced neovascularization in the wild-type mice, but not in the Tsp1-null mice, indicating that Tsp1 is required to mediate the anti-angiogenic effects of the metronomic dosing regimen in vivo. To establish whether lack of Tsp1 also affected tumour growth, fast-growing tumour cells were injected into wild-type and Tsp1-null mice. The mice were initially treated with the maximum-tolerated dose of cyclophosphamide, which slowed growth of the tumours in both mice strains. This was followed by low-dose cyclophosphamide treatment, which reduced tumour growth in only wild-type mice. So, although lack of Tsp1 does not affect the response to the maximum-tolerated dose of cyclophosphamide, it does prevent the effects of metronomic chemotherapy.

As soluble, circulating TSP1 was observed after in vitro metronomic dosing, it might be a useful surrogate marker for monitoring the clinical outcome of metronomic chemotherapy treatments. Human-tumour-bearing immunodeficient mice were treated with various metronomic regimens and after 20 days antitumour effects were seen in all treated animals. This antitumour response correlated with a 2–6-fold increase in Tsp1/tumour volume ratio, indicating that increased Tsp1 coincides with a decrease in tumour size.

This work provides an initial insight into the mechanisms that regulate the antitumour and anti-angiogenesis effects of metronomic therapies. TSP1 might also be a useful clinical tool for monitoring how patients are responding to this treatment strategy.