Background: Neurotoxicity of intrathecal (IT) chemotherapy has been variously attributed to the preservatives, volume, osmolality, and pH of the preparations. There has been little evaluation of how different drug concentrations or diluents can affect the osmolality and pH of the final solution. We conducted a three-part study: survey of cancer centers regarding the drug concentrations and diluent used in preparing IT chemotherapy; review of the literature on common practice of preparing IT chemotherapy; evaluation of the pH and osmolality of commonly used chemotherapy preparations for IT.
Method: We surveyed selected cancer centers to provide information on their standard volume, drug concentrations, and choice of diluents. MEDLINE was searched for clinical reports using the MeSH terms of 'cytarabine,' 'methotrexate,' or 'thiotepa' with the subheading 'Cerebrospinal fluid' and combined with 'intrathecal' in all database fields. Data retrieved included the choice of diluent, volume, and/or drug concentration. We evaluated the pH and osmolality of methotrexate (1, 2, 5, and 10 mg/mL), cytarabine (2, 5, 10, and 25 mg/mL), and thiotepa (1, 2, and 5 mg/mL) in normal saline, sterile water for injection (SWFI), and lactated Ringer's solution.
Results: Nine centers were surveyed (seven in Canada, one in Australia, one in United Kingdom). Most centers used 5 mL of preservative-free normal saline, irrespective of the drug or drug concentration used. Forty-four reports in the literature were reviewed. Most reported 5 mL of preservative-free normal saline. Most information on drug concentrations was provided for methotrexate, with an average concentration of about 1-2.5 mg/ mL. Cytarabine 0.4-20 mg/mL and thiotepa 1 mg/mL were also reported. In our in vitro evaluation, there was a trend of increased pH associated with increasing concentration of methotrexate and cytarabine. There was no apparent impact of thiotepa concentration on the pH values of the final preparations, irrespective of the diluent used. Except for cytarabine 10 and 25 mg/mL, all the tested solutions have pH within 10% of the physiologic range of CSF. There was a concentration-dependent change in osmolality with methotrexate and cytarabine preparations. Osmolality was increased with increased concentrations in all except methotrexate mixed in SWFI and thiotepa mixed in normal saline and lactated Ringer's solution. Except for some thiotepa solutions, all the tested solutions have osmolality within 10% of the physiologic range of CSF.
Conclusions: There is limited published literature on the potential impact of diluent and drug concentration on the pH and osmolality of IT chemotherapy preparation. Most cancer centers conventionally prepare IT chemotherapy with 5 mL of preservative diluent normal saline, irrespective of the specific drug or dose used. The conventional practice means that most methotrexate preparations are likely to have comparable pH and osmolality to CSF. In contrast, cytarabine preparations may show significantly higher pH than the CSF, while thiotepa preparations generally have lower osmolality than the CSF.