Although studies have assessed thoroughly the relationship between imatinib plasma concentration and efficacy (3, 4, 12), less is known about the relationship between imatinib plasma concentration and toxicity (6). One study showed that the occurrence and number of side-effects correlated with imatinib total and free plasma concentration in patients with GIST (6). However, the study did not assess imatinib exposure with the grade or type of toxicities, findings that may have been of value in clarifying the dose-dependent toxicities of imatinib. Here we described two patients who experienced intolerable toxicities while on standard-dose imatinib treatment. Both had high imatinib plasma trough concentrations, which appeared to have contributed to the severe fluid retention (e.g.

ascites, edema, and pleural effusion) observed in these patients. Through imatinib plasma monitoring, we reduced the dose of imatinib, consequently decreasing these toxicities, while maintaining sufficient imatinib exposure for adequate efficacy of the treatment. Another interesting aspect of these two patient cases is that they highlight the difficulties of response evaluation for imatinib therapy in GIST patients. Toxicity of imatinib or reaction to imatinib treatment may sometimes be mistaken for disease progression. For example, in the first patient who had peritoneal metastases, ascites was regarded as a sign of disease progression; in the second patient, enlarged reactive mesenteric lymph nodes were mistaken for lymph node metastatses.

Because both patients had no tumor shrinkage in other nodules, which is quite common in patients with advanced GIST on imatinib therapy, it was difficult to determine whether imatinib treatment was effective. 18FDG-PET/CT scan may be considered for accurate assessment of response to imatinib, especially when rapid readout of activity is necessary. When considering imatinib plasma level monitoring to address intolerable adverse events, it is necessary to also consider covariates that may affect imatinib pharmacokinetics, such as demographics, body weight, and other clinical characteristics. Patients of Asian descent, for example, might display different imatinib pharmacokinetics compared with those of Caucasian descent as a result of different physical characteristics.

Whilst there has been no clear elucidation of the influence of Brefeldin_A race on imatinib pharmacokinetics, several reports have suggested a trend towards lower imatinib trough levels in patients with higher body mass index (3, 5, 13). As such it might be assumed that Asian patients would have a tendency towards lower imatinib trough levels compared to their Caucasian counterparts. However, this assumption should be treated with caution given that the reports were associated with only a weak correlation or tendency towards insignificance.