During in vitro studies as long as the compound is soluble in DMSO it is usually possible to evaluate the compound. However when moving to in vivo studies particularly when thinking about a possible development compound it is important to start looking at pharmaceutical development, in particular consider the likely clinical route(s) of administration the dosing regime and also what will be needed to get adequate exposure in the safety species. In general formulation and pharmaceutical development studies can take longer than most discovery scientists imagine and so a consideration of solubility, salt selection and stability can have a significant impact on the overall timeline. Formulation can have a significant impact on bioavailability, stability and ease of manufacture.
Whilst there are many different routes of administration the two principle choices are oral and intravenous.
First choice should be normal saline or 5% dextrose. Acceptable pH ranges for buffered vehicles are 4 to 8 and for unbuffered vehicles 3 to 9, buffer strength should be kept below 10 millimolar. Vehicles outside this range are likely to be extremely painful a range of acceptable pharmaceutical buffers is shown in the table below. If solubility is an issue the use of co-solvents such as ethanol, propylene glycol or polyethylene glycol 400, should be investigated however take note of the hemolytic potential of such systems (see below). If addition of acid or base is used to improve solubility you must ensure the resulting solution is not hypertonic. As a final resort the use of nanosuspensions or microemulsions can be investigated however the best solution might be to look for another compound!
Since the volume of the infusion is likely to be greater than used in a single bolus injection then the pH and tonicity should be restricted to a range much closer to physiological values (pH 7 to 8, buffer (1-10 millimolar).
The pharmaceutical excipients described above used to increase solubility and stability of active pharmaceutical ingredients may also have undesirable effects these include hemolysis, vascular necrosis, thrombosis and vasculitis. The biocompatibility of these excipients is an important consideration during formulation development and the hemolytic potential and local irritation will need to be assessed. The vascular toxicity will depend on a number of factors including, rate of delivery, volume of solvent administered, number of injections per site, relative concentrations of co-solvents,potential for drug precipitation. Despite the importance of hemolytic potential of parenteral formulations, there is no comprehensive resource that I am aware of. Fort et al (J. Parent. Sci. Technol. 1984 38, 82) described the hemolytic potential of a number of co-solvent systems. The results suggest all concentrations of propylene glycol/ethanol formulations greater than 10% propylene glycol and 30% ethanol produce hemoglobinuria in vivo and hemolysis in vitro. Similarly concentrations of greater than 30% polyethylene glycol 400 and 20% ethanol cause hemolysis. Solutions using saline rather than water as the co-solvent appear to have slightly less hemolytic potential. However these results suggest that the maximum acceptable concentrations of co-solvents is approximatley 25%.
The hemolytic potential of various formulation vehicles in dog, rabbit, and human blood by means of an in vitro hemolysis assay are compared in (In vitro hemolysis: guidance for the pharmaceutical scientist J. Pharm. Sci. 2006 Jun;95(6):1173).
Here is a range commonly used pharmaceutical buffers taken from Y.-C. Lee et al. International Journal of Pharmaceutics 253 (2003) 111–119, these can be used to modify the final pH in order to improve solubility of acids or bases, but remember working at the extremes of these pH ranges may compromise the chemical stability of the drug.
|Buffering agents||pKa||pH range||Commercial products|
|Maleic acid||1.9, 6.2||2–3||Teniposide|
|Tartaric acid||2.9, 4.2||2.5–4||Tolazoline HCl|
|Citric acid||3.1, 4.8, 6.4||3–7||Labetalol HCl, nicardipine HCl|
|Acetic acid||4.75||4–6||Mitoxantrone HCl, ritodrine HCl|
|Sodium bicarbonate||6.3, 10.3||4–9||Cefotetan, cyclophosphamide|
|Sodium phosphate||2.2, 7.2, 12.4||6–8||Warfarin, vecuronium Br|
Precipitation upon dosing remains significant challenge for solution formulations, this risk can be evaluated by using a series of serial dilutions of the dosing solution using phosphate buffer(pH 7.4, Na2HPO4 –NaH2PO4 buffer) with a concentration of 0.067M which would be similar to whole blood.
For salts (See below for salt selection) and water soluble compounds start with aqueous, buffers; if there is no good solubility, move to other solution approaches: cosolvents, cyclodextrins, micelles. Pay special attention to the combined use of pH with these excipients.
For insoluble compounds emulsion or suspension formulations may be an option (See below for excipients), there are several suspension alternatives: conventional suspension, micronized suspension, nanoparticles. Formulations using reduced particle size and increased homogeneity may provide potential for improved dissolution as well as control of batch reproducibility. Emulsions can be dosed in soft gel capsules but remember the patient will probably balk at taking anything larger than 1 mL capacity so the concentration of the emulsion may limit dose.
The selection of the best salt form can be a time-consuming process but some simple rules can help. Usually you need 2 pH units difference between acid and base pKa, if you are using volatile counter ions you may need a larger difference (triazole fungicide (pKa 2), Crystalline HCl salt: 12% loss of chloride after 6 hrs at 60ºC). Solubility is very difficult to predict and most algorithms require melting point so you need to make a sample anyway. Salts of mineral acids are often microcrystalline and hygroscopic. See Also (H.G. Brittain, ―Strategy for the Prediction and Selection of Drug Substance Salt Forms‖, Pharm. Tech. (2007) 31(10): 78-88).
For each salt form check:-
- Particle morphology (Microscopy)
- Crystallography (either optical crystallography or x-ray powder diffraction)
- Thermal Methods of Analysis (melting behavior, DSC, TG)
- pH dependence of aqueous solubility
- Solid-state spectroscopy (FTIR-ATR, Raman, and/or SS-NMR)
- Evaluation of hygroscopicity through exposure to controlled relative humidities
Top ten salts used in commercial products
Also remember that a new salt form with improved stability, solubility or bioavailability may allow for additional patent protection. For more information see also Handbook of Pharmaceutical Salts
A list of some common excipients and the range for early formulations. Whilst this gives some information about the concentration range remember there maybe a maximum tolerated dose for the excipient. See Also (Handbook of Pharmaceutical Excipients (Rowe, Handbook of Pharmaceutical Excipients)).
|Excipient||Route and Conc range|
|5% dextrose in water||i.v.|
|Buffered solutions pH: 2–8||p.o.,i.v|
|Cosolvent||N-methylpyrrolidone (NMP)||10–20% (oral, i.v.)|
|dimethyl sulfoxide (DMSO)||10–20% (oral or i.v.)|
|N,N-dimethylacetamide (DMA)||10–30% (i.v.)|
|Ethanol||10% (oral, i.v.)|
|propylene glycol (PG)||30–60% (oral, i.v.)|
|polyethylene glycol 400||40–100% (oral, i.v.)|
|diethylene glycol monoethyl ether||30% (oral)|
|80 5–10% (oral, i.v.)|
|polyoxyl-35 castor oil |
|5–10% (oral, i.v.)|
|polyoxyl 40 hydrogenated castor oil |
|5–10% (oral, i.v.)|
|caprylocaproyl macrogol-8-glycerides |
|40–60% (oral); 20–40% (i.v.)|
|Lipid||Soybean oil||50–100% (oral)|
|polyxoyethyllated oleic glycerides |
|30–60% (oral, i.v.)|
|medium chain mono- and diglycerides |
|30–60% (oral, i.v.)|
Other routes of administration
Intramuscular, subcutaneous and intraperitoneal
Aqueous solutions described above may be used, in addition solution in oils may be useful. Soybean oil, peanut oil, sesame oil or Miglyol 810 (medium chain triglyceride) have all been used, in addition up to 0.5% w/w TWEEN 80 or SPAN 80 can be used to improve solubility. Bear in mind that the rate of drug release from these systems may be very slow. This is due to the viscosity of the system and the need for the drug to diffuse from the oil bolus before it can distribute into other tissues. Suspensions of drug may also be used in lab experiments but it is important to control particle size to avoid sedimentation, vehicle should preferably be saline but addition of 0.5% sodium carboxymethylcellulose may also help.
If none of the above solutions work then for preclinical work IMWITOR/TWEEN 80 combinations, or LABRAFIL M2125CS with or without TWEEN 80 may provide the required level of solubility
Only aqueous solutions are acceptable and they must be buffered to within a pH range of 4 to 9, but closer to physiological pH the better. Buffer concentrations should be as low as possible (10 millimolar).
Topical transdermal administration
The vehicles of choice are propylene glycol, isopropyl myristate, PEG 300 or 400 or mixtures of these containing isopropyl alcohol or ethanol. When using volatile solvents care needs to be taken to ensure the drug does not crystallize as the solvent evaporates.
Updated 17 February 2012