Mechanisms by which cancer chemotherapeutic drugs induce emesis.

Because the area postrema seems essential for chemotherapy-induced vomiting, both circulating and/or neurally mediated stimuli in this area could trigger the emetic response. In our laboratories results of cross-circulation and direct intracerebroventricular infusion experiments in dogs do not support a role for circulating substances. The large increases in serum vasopressin induced by cisplatin were not blocked by inhibitors of the angiotensin-converting enzyme. In the ferret inhibition of serotonin synthesis with p-chloro-phenylalanine, administration of selective antagonists of 5-hydroxy-tryptamine3 (5-HT3) receptors, or visceral deafferentation inhibited the emetic response evoked by cisplatin or high-dose cyclophosphamide. The results suggest that serotonin plays an important role and that peripheral neural mechanisms are involved in the emetic response. The strong antiemetic efficacy of selective 5-HT3 antagonists also has been confirmed in humans. In cancer patients high-dose cisplatin increased the plasma and urinary levels of 5-hydroxy-indoleacetic acid (5-HIAA), but did not affect platelet and free plasma serotonin. The changes in 5-HIAA levels paralleled the onset and development of vomiting. No evidence of serotonin depletion has been obtained after high-dose cisplatin. Dacarbazine, another strongly emetogenic agent, increased urinary 5-HIAA; however, only small increases in 5-HIAA were produced with low-dose cisplatin or cyclophosphamide-containing regimens. Thus, emetogenicity appears to be directly related to the ability of the cytotoxic agent to release serotonin. In humans, antiemetics such as ondansetron, metoclopramide, and dexamethasone did not effect high-dose cisplatin-induced increases in serotonin metabolism. Therefore, these antiemetics seem not to affect the amount of serotonin released. The mechanism by which chemotherapeutic drugs induce serotonin release is unknown; however, release may occur by direct cytotoxicity on the gastrointestinal mucosa, including the enterochromaffin cells. Delayed emesis appears to be mediated by 5-HT3-independent mechanisms. It is proposed that emesis that develops despite high-dose ondansetron (residual emesis) should be considered delayed emesis. Residual and delayed episodes of emesis have similar time courses, are characterized by very mild emetic episodes and poor response to 5-HT3 antagonists, and are not associated with increases in serotonin metabolism.