ORIGINAL ARTICLES

Ortofarma – Quality Control Laboratories, Matias Barbosa, MG, Brazil

Compatibility of cholecalciferol, haloperidol, hydrochlo- ride, levodopa/carbidopa, , hydrochloride, tacro- limus monohydrate, terbinafine, hydrochloride and valsartan in SyrSpend® SF PH4 oral suspensions

H. C. POLONINI, S. L. SILVA, C. N. CUNHA, M. A. F. BRANDÃO, A. O. FERREIRA

Received October 21, 2015, accepted December 2, 2015 Ortofarma – Quality Control Laboratories, BR 040, n. 39, Empresarial Park Sul. 36120-000. Matias Barbosa – MG. Brazil [email protected] Pharmazie 71: 185–191 (2016) doi: 10.1691/ph.2016.5177

A challenge with compounding oral liquid formulations is the limited availability of data to support the physical, chemical and microbiological stability of the formulation. This poses a patient safety concern and a risk for medication errors. The objective of this study was to evaluate the compatibility of the following active pharma- ceutical ingredients (APIs) in 10 oral suspensions, using SyrSpend® SF PH4 (liquid) as the suspending vehicle: cholecalciferol 50,000 IU/mL, haloperidol 0.5 mg/mL, imipramine hydrochloride 5.0 mg/mL, levodopa/carbidopa 5.0/1.25 mg/mL, lorazepam 1.0 mg/mL, minocycline hydrochloride 10.0 mg/mL, tacrolimus monohydrate 1.0 mg/ mL, terbinafine 25.0 mg/mL, tramadol hydrochloride 10.0 mg/mL and valsartan 4.0 mg/mL. The suspensions were stored both refrigerated (2 - 8 °C) and at controlled room temperature (20 - 25 °C). This is the first stability study for these APIs in SyrSpend® SF PH4 (liquid). Further, the stability of haloperidol,iImipramine hydrochloride, minocycline, and valsartan in oral suspension has not been previously reported in the literature. Compatibility was assessed by measuring percent recovery at varying time points throughout a 90 days period. Quantification of the APIs was performed by high performance liquid chromatography (HPLC-UV). Given the percentage of recovery of the APIs within the suspensions, the beyond-use date of the final preparations was found to be at least 90 days for most suspensions both refrigerated and at room temperature. Exceptions were: Minocycline hydrochloride at both storage temperatures (60 days), levodopa/carbidopa at room temperature (30 days), and lorazepam at room temperature (60 days). This suggests that compounded suspensions of APIs from different pharmacological classes in SyrSpend® SF PH4 (liquid) are stable.

1. Introduction in Table 1, compounded at a single concentration in SyrSpend® SF Extemporaneous preparation of oral liquid dosage forms is a PH4 (liquid) and stored both refrigerated and at room temperature. ® common and important pharmacy practice for patients that require SyrSpend SF is an internationally available, GMP produced, non-standard doses, experience swallowing difficulties or receive ready-to-use taste-masking oral liquid vehicle. Its suspending medication via enteral feeding tubes (Glass and Haywood 2006). properties are derived from starch without traditionally used excip- Oral liquids are not only common practice in pediatrics (Brion et ients that can have toxicological effects, induce allergic reactions al. 2003; Schirm et al. 2003), but also in the general adult popu- or cause irritation, such as sugar (Hill et al. 1988; Jijo and Flow- lation, where recent studies have demonstrated that up to 22.4 % erlet 2014), ethanol (Zuccotti and Fabiano 2014; Fiocchi et al. have difficulties swallowing (Lau et al. 2015; Marquis et al. 2013). 1999), propylene glycol (Committee on Drugs 1997; Fabiano et Oral liquids are relatively quick and easy to prepare with limited al. 2011), sorbitol (Johnston et al. 1994; Payne et al. 1997), benzyl need for compounding equipment and they allow for flexibility in (Gershanik et al. 1982; Centers for Disease Control 1982; dosage out of a single strength preparation (Brion et al. 2003). Committee on Fetus and Newborn 1983) and common food aller- The main challenge with compounding oral liquid formulations gens (Sakai et al. 2012; Audicana Berasategui et al. 2011). The is the limited availability of data to support the physical, chem- compatibility of SyrSpend® SF with various APIs has already ical and microbiological stability of the formulations (Glass and been demonstrated (Geiger et al. 2012a, 2012b, 2013a, 2013b, Haywood 2006; Brion et al. 2003; Conroy 2003). In a UK survey 2015; Sorenson et al. 2012; Sorenson and Whaley 2012; Voudrie it was found that in more than half (54 %) of the extemporaneous and Allen 2010; Voudrie et al. 2011; Vu et al. 2008; Whaley et al. formulations shelf-life was inadequately supported (Brion et al. 2012a, 2012b; Ferreira et al. 2015). 2003). Due to the limited availability of scientific data, there is little In this study the combined physical-chemical compatibility is harmonization in the concentration or formulation of compounded assessed, as a deficit in either of the two would result in an out oral liquids (Brion et al. 2003; Rood et al. 2014). This poses a of specification during analysis. The concentration for each API patient safety concern and a risk for medication errors (Rood et studied was selected based on commonly prescribed concentra- al. 2014). Demands have been made to publish scientifically veri- tions for children or adults. To the best of the authors’ knowledge, fied, palatable extemporaneous formulations with standardized there is no previous stability study in the literature for haloper- oral liquid concentrations to increase patient safety and adherence idol, imipramine hydrochloride, minocycline hydrochloride and (Brion et al. 2003; Conroy 2003; Rood et al. 2014; Allen 2008). valsartan oral suspensions. No stability studies of the current APIs The objective of this study was to evaluate the physical and chem- compounded in SyrSpend® SF PH4 (liquid) have previously been ical stabilities of the active pharmaceutical ingredients (APIs) listed published. Pharmazie 71 (2016) 185 ORIGINAL ARTICLES

Table 1: Concentrations of the suspensions used in the study all conditions. Cholecalciferol and lorazepam decomposed under all stress conditions. After these validations, the stability of the API Concentration Action and use ® in suspension APIs in SyrSpend SF PH4 (liquid) was assessed. In this study we did not evaluate the uniformity of the drug in the Cholecalciferol (vitamin D3) 50,000 IU/mL Vitamin D analogue suspension, and expect that any non-soluble drug will exist as insol- Haloperidol 0.5 mg/mL receptor antag- uble crystals, particulates, or precipitate. According to the Merck onist; neuroleptic Index 14th edition (2006), these are the solubilities of the APIs in Imipramine hydrochloride 5.0 mg/mL Monoamine reuptake water (major component of the used suspending vehicle): Cholecal- inhibitor; antide- ciferol – practically insoluble; haloperidol – 1.4 mg/100mL; imip- pressant ramine hydrochloride – freely soluble; levodopa – 66 mg/40 mL; Levodopa/carbidopa 5.0/1.25 mg/mL Treatment of Parkinson’s lorazepam – 0.08 mg/mL; tacrolimus monohydrate – insoluble disease in water; terbinafine – slightly soluble; tramadol hydrochloride – soluble; and valsartan – soluble. Therefore, we predict that some Lorazepam 1.0 mg/mL Benzodiazepine heterogeneity of drug distribution will exist for cholecalciferol, Minocycline hydrochloride 10.0 mg/mL Tetracycline antibacterial levodopa, lorazepam and tacrolimus monohydrate in the stored Tacrolimus monohydrate 1.0 mg/mL Immunosuppressant suspensions and that unsufficient mixing prior to sampling may Terbinafine 25.0 mg/mL Antifungal lead to increased variance in API percentage recovery. The stability results are shown in Table 5 and are expressed as Tramadol hydrochloride 10.0 mg/mL receptor ; noradrenaline reuptake relative percent of recovery (initial sampling time = 100 %). For inhibitor; analgesic the suspensions to be considered stable, the relative percentage recovery should lie within 90-110 % (USP 2015; BP 2015; EP Valsartan 4.0 mg/mL Angiotensin II (AT ) 1 2015). Figure 1 graphically represents the stability of the APIs in receptor antagonist SyrSpend® SF PH4 (liquid) in terms of absolute nominal concen- tration.

2. Investigations, results and discussion At each sampling time, the visual appearance of the suspensions Validation studies of all methods of analysis (chromatographic was evaluated to verify their homogeneity and physical stability (data not shown). Throughout the whole study, no phenomena such conditions described in Table 2) were performed and all results as precipitation, turbidity, macroscopically visible crystal growth, (Table 3) met the respective acceptance criteria. Stability-indi- odor generation, phase separation, flocculation or caking were cating studies were also conducted. These results are summarized observed, except for minocycline hydrochloride after 60 days of in Table 4. Stability-indicating studies are important to determine storage. No study on the stability of minocycline hydrochloride in if the used methods are fully validated and adequate to identify oral liquids was found, but the Merck Index (2006) states that this decomposition of the APIs by chromatographic analysis. The API is sensitive to light and to surface oxidation (also confirmed in decomposition profile of the APIs notably varied for different our findings in Table 4). As all suspensions were stored in light-re- stressing conditions. Only levodopa was found to be stable under sistant bottles, it is likely that the decomposition of minocycline

Table 2: Chromatographic conditions used in the compatibility study

API Mobile phase composition Work concentration Column Flow UV detection (μg/mL)* (mL/min) wavelength (nm)

Cholecalciferol Hexane and pentanol (997:3) 100.0; 20 μL L3, 4.6-mm × 25-cm; at 2.0 254 (vitamin D3) 25°C1 Haloperidol Methanol and buffer solution (6.8 g/L of monobasic potas- 200.0; 20 μL injection L1, 4.6-mm × 15-cm; at 0.8 247 sium phosphate in water, adjusted with phosphoric acid to a 25°C2 pH of 4.0) (55:45) Imipramine 0.06M sodium perchlorate, acetonitrile and trietylamine 300.0, in water and ace- L1, 3.9-mm × 30-cm; at 1.5 269 hydrochloride (625:375:1) tonitrile (625:375); 20 μL 25°C3 injection Levodopa/ Alcohol and buffer (6.6 g/L of monobasic sodium phos- 250/ 25; 20 μL injection L1, 4.6-mm × 25-cm; at 1.0 280 Carbidopa phate in water, adjusted with phosphoric acid to a pH of 25°C4 2.2) Lorazepam Acetonitrile, glacial acetic acid, and water (45: 0.2: 55) 100.0, in methanol; 20 μL L1, 4.0-mm × 30-cm; at 2.0 254 injection 25°C5 Minocycline, Dimethylformamide, tetrahydrofuran, 0.2M ammonium ox- 500.0; 20 μL injection L1, 4.6-mm × 25-cm; at 1.5 280 hydrochloride alate and 0.01M EDTA (120:80:600:180), with pH adjusted 40°C6 to 7.2 with ammonium hydroxide Tacrolimus Acetonitrile and water (65:35) 50.0; 20 μL injection L1, 4.6-mm × 25-cm; at 1.7 214 monohydrate 70°C7 Terbinafine Acetonitrile and water (2:3), with 0.15 % triethylamine and 5.0; 20 μL injection L1, 4.6-mm × 15-cm; at 0.4 224 0.15 % phosphoric acid 25°C8 Tramadol Acetonitrile and a solution with 20 mM of phosphoric acid 250.0, acetonitrile and wa- L1, 4.6-mm × 25-cm; at 1.0 275 hydrochloride and 4 g/L of sodium 1-hexane sulfonate (50:50) ter (50:50); 20 μL injection 30°C9

Valsartan Acetonitrile, glacial acetic acid, and water (500:1:500) 500.0; 20 μL injection L1, 4.6-mm × 15-cm; at 1.0 273 25°C10

* diluted with mobile phase, unless specified otherwise. 1Zorbax Eclipse Plus 5μ (Agilent). 2Gemini 5μ 110Å (Phenomenex). 3Luna 5μ C18 100Å (Phenomenex). 4Zorbax Eclipse XDB 5μ (Agilent). 5Zorbax Eclipse XDB 5μ (Agilent). 6Zorbax Eclipse XDB 5μ (Agilent). 7Zorbax Eclipse XDB 5μ (Agilent). 8Gemini 5μ 110Å (Phenomenex). 9Zorbax Eclipse XDB 5μ (Agilent). 10Gemini 5μ 110Å (Phenomenex).

186 Pharmazie 71 (2016) ORIGINAL ARTICLES

Fig. 1. Plot of APIs in SyrSpend® SF PH4 (liquid) throughout the compatibility study (dashed lines represent the lower and upper limits, corresponding to 90 and 110 % of labeled concentration). Values represents mean ± SD (n=6). A – cholecalciferol 50,000 IU/mL, B – haloperidol 0.5 mg/mL, C – imipramine hydrochloride 5.0 mg/mL, D – levodopa/carbidopa 5.0/1.25 mg/mL, E – lorazepam 1.0 mg/mL, F – minocycline hydrochloride 10.0 mg/mL, G – tacrolimus monohydrate 1.0 mg/mL, H – terbinafine 25.0 mg/mL, I – tramadol hydrochloride 10.0 mg/mL, J – valsartan 4.0 mg/mL. Pharmazie 71 (2016) 187 ORIGINAL ARTICLES

Table 3: Summary of linearity’s study for the validation of the HPLC method

API Linearity Specificity Precision Accuracy

Range (μg/mL) Analytical curve R2 ANOVA’s ANOVA’s LOD LOQ Discrepancy Repeatability Intermediate Recovery significance lack of fit (F) (μg/mL) (μg/mL) (%) (CV, %) precision (%) of regression (F) (CV, %)

Cholecalciferol 70.56 – 131.04 y = 125.08x + 221.76 0.9941 2215.06 3.70 0.02 0.08 |1.13| 0.51 0.92 99.67 (vitamin D3) Haloperidol 140.07 – 260.13 y = 35.89x – 339.96 0.9972 4672.53 3.37 0.01 0.04 |1.22| 0.39 0.28 100.04 Imipramine 210.28 – 390.52 y = 14.47x – 113.14 0.9995 26612.58 1.44 0.001 0.003 |0.79| 0.30 2.54 100.16 hydrochloride Levodopa* 175.56 –326.04 y = 15.92x - 271.22 0.9937 2041.11 2.00 0.01 0.03 |0.56| 0.17 3.22 100.54 Carbidopa* 17.78 –33.02 y = 15.38x – 3.86 0.9952 2675.60 3.53 0.13 0.52 1.64 0.31 2.45 99.75 Lorazepam 35.21 – 65.39 y = 19.44x – 54.86 0.9917 1547.32 0.89 0.05 0.18 |1.49| 0.71 0.56 100.06 Minocycline 350.07 – 650.13 y = 17.43x – 142.99 0.9974 5067.23 2.44 0.01 0.03 |0.21| 0.27 0.43 100.36 hydrochloride Tacrolimus 70.00 – 130.00 y = 6.67x + 23.13 0.9980 6358.90 3.66 0.01 0.03 |0.42| 0.11 1.03 100.01 monohydrate Terbinafine 3.57 – 6.63 y = 183.69x + 8.02 0.9978 5906.16 1.13 0.78 2.61 |1.10| 0.35 0.80 99.49 Tramadol 175.00 – 435.00 y = 6.11x + 31.99 0.9956 3075.87 0.55 0.01 0.05 |0.53| 0.41 0.90 99.88 hydrochloride Valsartan 350 – 650.26 y = 12.64x + 92.60 0.9961 3359.58 3.67 0.01 0.02 |1.82| 0.44 1.17 99.61

*Method validated for the combination of the APIs. LOD: Limit of Detection. LOQ: Limit of Quantification (20 μL injections). CV: coefficient of variation. Acceptance criteria were: R2 > 0.99, F (significance of regression) >> 4.67, F (lack of fit) < 3.71, discrepancy < 2 %, repeatability and intermediate precision < 5 %, and recovery = 100 % ± 2 %. All analytical ranges (μg/mL) were adequate to quantify the APIs in the concentrations used in the suspensions (mg/mL).

Table 4: Summary of the stability-indicating study for the APIs

API HCl NaOH UV Heat H2O2

Area %d* Area %d* Area %d* Area %d* Area %d*

Cholecalciferol 0.00 |-100.00| 207.45 |-98.61| 14960.11 |0.04| 8642.87 |-42.21| 4627.40 |-69.06| (vitamin D3) Haloperidol 6580.82 |0.53| 2320.02 |-64.56| 7146.31 |9.16| 6489.87 |-0.86| 5697.64 |-12.97| Imipramine 4045.03 |-2.96| 3979.50 |-4.53| 3607.85 |-13.45| 4212.22 |1.05| 4191.83 |0.57| hydrochloride Levodopa 3750.61 |1.38| 3715.46 |0.43| 3690.97 |-0.23| 3762.34 |1.69| 3667.44 |-0.87| Carbidopa 282.79 |4.21| 193.85 |-28.56| 278.80 |2.74| 280.33 |3.31| 267.04 |-1.59| Lorazepam 2266.58 |-24.67| 2494.96 |-17.08| 2512.16 |-16.51| 2698.65 |-10.31| 2711.76 |-9.87| Minocycline 8795.76 |1.30| 8786.12 |1.19| 8636.63 |-0.53| 8699.92 |0.20| 0.00 |-100.00| hydrochloride Tacrolimus 655.23 |-1.71| 0.00 |-100.00| 671.86 |0.79| 696.08 |4.41| 597.06 |-10.44| monohydrate Terbinafine 929.78 |-0.39| 931.14 |-0.25| 937.94 |0.48| 906.43 |-2.89| 911.75 |-2.32| Tramadol 1532.61 |-0.18| 1551.04 |1.02| 1580.75 |2.95| 1576.36 |2.67| 1538.22 |0.18| hydrochloride Valsartan 6458.88 |1.45| 1111.01 |-82.55| 6621.43 |4.01| 6483.83 |1.85| 6451.30 |1.33|

(Results presented as average of 3 replicates, at the work concentration) *%d = percentage of discrepancy between the API peak without submission to stressing factors (negative control) and the peak of a sample subjected to one of the cited accelerated-degradation factors. Areas given as mV. Maximum acceptable = 2 % (values higher than this are in bold).

Table 5: Stability of the APIs in SyrSpend® SF PH4 (liquid)

Elapsed time (days) % Recovery Elapsed time (days) % Recovery Refrigerated Temperature Controlled Room Temperature (2-8 ºC) (20-25 ºC) Refrigerated Temperature Controlled Room Temperature (2-8 ºC) (20-25 ºC) T = 7 99.85 ± 0.28 99.85 ± 0.37 Cholecalciferol 50,000 IU/mL T = 14 99.95 ± 0.25 100.06 ± 0.26 T = 0 100 ± 0.11 100 ± 0.11 T = 30 98.70 ± 0.08 99.68 ± 0.28 T = 7 99.79 ± 1.01 100.01 ± 0.48 T = 60 95.86 ± 0.79 96.68 ± 1.81 T = 14 100.58 ± 1.00 101.01 ± 0.60 T = 90 97.59 ± 0.30 98.11 ± 0.28 T = 30 101.72 ± 0.62 100.46 ± 1.02 Imipramine hydrochloride 5.0 mg/mL T = 60 102.00 ± 0.39 100.99 ± 0.63 T = 90 100.83 ± 0.26 101.84 ± 0.28 T = 0 100 ± 0.65 100 ± 0.65 T = 7 100.14 ± 0.88 100.44 ± 0.16 Haloperidol 0.5 mg/mL T = 14 99.47 ± 0.62 99.42 ± 0.55 T = 0 100 ± 0.11 100 ± 0.11 T = 30 99.33 ± 0.62 100.00 ± 0.49

188 Pharmazie 71 (2016) ORIGINAL ARTICLES

Elapsed time (days) % Recovery Elapsed time (days) % Recovery

Refrigerated Temperature Controlled Room Temperature Refrigerated Temperature Controlled Room Temperature (2-8 ºC) (20-25 ºC) (2-8 ºC) (20-25 ºC)

T = 60 99.49 ± 0.56 100.01 ± 0.73 T = 30 99.95 ± 0.13 101.34 ± 0.26 T = 90 97.59 ± 1.13 99.07 ± 0.58 T = 60 100.33 ± 0.31 100.37 ± 0.60 T = 90 101.50 ± 0.24 100.78 ± 0.60 Levodopa 5.0 mg/mL* *Assayed in a single suspension containing both levodopa and carbidopa. T = 0 100 ± 0.53 100 ± 0.53 T = 7 100.23 ± 0.33 99.16 ± 0.55 (color changes in the suspension, Fig. 2) is due to oxidation reac- T = 14 96.20 ± 0.92 98.62 ± 0.70 tions. Additionally, the suspension thickened after 60 days to a T = 30 100.37 ± 0.36 99.66 ± 0.33 point that a chromatographic injection was not recommended and T = 60 95.47 ± 0.39 95.90 ± 0.33 accurate patient dosing would be challenging. T = 90 96.04 ± 0.36 96.49 ± 0.35 Carbidopa 1.25 mg/mL* T = 0 100 ± 0.32 100 ± 0.32 T = 7 99.25 ± 0.35 97.32 ± 0.46 T = 14 102.69 ± 0.44 98.64 ± 0.33 T = 30 99.27 ± 0.36 98.90 ± 0.30 T = 60 90.93 ± 0.21 67.55 ± 0.38 T = 90 91.22 ± 0.34 66.72 ± 0.64 Lorazepam 1.0 mg/mL T = 0 100 ± 0.17 100 ± 0.17 T = 7 100.01 ± 0.27 100.03 ± 0.18 T = 14 99.82 ± 0.31 99.96 ± 0.27 T = 30 100.68 ± 0.20 100.67 ± 0.36 T = 60 100.22 ± 0.33 99.93 ± 0.17 T = 90 99.24 ± 0.55 83.04 ± 0.47 Minocycline hydrochloride 10.0 mg/mL T = 0 100 ± 0.27 100 ± 0.27 T = 7 99.51 ± 0.05 100.01 ± 0.31 T = 14 100.10 ± 0.73 100.52 ± 0.76 T = 30 101.34 ± 0.31 101.39 ± 0.24 T = 60 100.46 ± 0.30 100.57 ± 0.14 T = 90 Interrupted (physical Interrupted (physical instability) instability) Tacrolimus monohydrate 1.0 mg/mL T = 0 100 ± 0.16 100 ± 0.16 T = 7 100.21 ± 0.33 100.40 ± 0.45 T = 14 100.24 ± 0.15 100.40 ± 0.18 T = 30 99.53 ± 0.12 99.91 ± 0.29 Fig. 2. Minocycline hydrochloride syrup at a) T = 0 and b) T = 90 days. T = 60 98.29 ± 0.66 100.03 ± 0.37 T = 90 99.33 ± 0.32 100.54 ± 0.42 Terbinafine 25.0 mg/mL Cholecalciferol (vitamin D3) was stable for at least 90 days in our study. This is longer than some of the results found by Connors T = 0 100 ± 0.12 100 ± 0.12 et al. (1986), who evaluated cholecalciferol syrups containing T = 7 100.18 ± 1.02 99.12 ± 0.30 different stabilizers (ethyl gallate 0.01 %; butylated hydroxytol- T = 14 100.41 ± 0.65 99.53 ± 0.48 uene 0.01 %; citraconic acid 0.1 % and butylated hydroxytoluene T = 30 97.92 ± 0.45 99.08 ± 0.29 0.01 %; citric acid 0.1 % and butylated hydroxytoluene 0.01 %; T = 60 99.47 ± 0.41 100.27 ± 0.16 butylated hydroxytoluene 0.01 % and ascorbic acid 0.01 %). They T = 90 99.53 ± 0.22 99.60 ± 0.19 found a maximum stability of two months for these syrups when stored at 37 °C. When stored at 17 °C, the syrups were stable for Tramadol hydrochloride 10.0 mg/mL six months, but only with the added stabilizers. T = 0 100 ± 0.38 100 ± 0.38 Lorazepam oral suspension stored at 2-8 °C was stable for at least T = 7 98.66 ± 0.15 98.77 ± 0.18 90 days, while the same suspension stored at 20-25 °C presented T = 14 96.68 ± 0.28 96.68 ± 0.24 a 17 % loss in concentration (average) in the same period. This suggests that the lorazepam oral suspension could be used for up T = 30 98.74 ± 0.24 98.93 ± 0.27 to 60 days after compounding when stored at room temperature or T = 60 96.77 ± 0.19 95.21 ± 0.21 90 days when refrigerated. No report was found evaluating a loraz- T = 90 96.89 ± 0.08 96.65 ± 0.08 epam oral suspension, but a study conducted by Stiles et al. (1996) Varsartan 4.0 mg/mL with a lorazepam injection packaged in polypropylene-pump syringes found that 23-26 % degradation of the API occurred in T = 0 100 ± 1.06 100 ± 1.06 10 days when stored at either ambient or refrigerated temperatures. T = 7 101.62 ± 0.36 101.46 ± 0.19 The combination levodopa/carbidopa stored at 2-8 °C was found T = 14 98.37 ± 0.98 98.57 ± 0.36 to be stable for at least 90 days, while the suspension stored at Pharmazie 71 (2016) 189 ORIGINAL ARTICLES

20-25 °C showed unacceptable degradation (>10 %) of Carbidopa 30 min, immediately before use. All volumetric glassware and analytical balance used after 30 days of storage. This instability in an aqueous system was were previously calibrated. already reported by Pappert et al. (1996), with losses in concen- tration of up to 60 % independently of the storage temperature. 3.2. Equipment ® Interestingly, the SyrSpend SF PH4 (liquid) vehicle was able to HPLC analyses were performed on a qualified and calibrated chromatography system increase the stability of this combination of APIs even though it is (Young Lin, Anyang, Korea) composed of a quaternary gradient pump (YL 9110), a an aqueous vehicle. Nahata et al. (2000) also studied a levodopa/ photodiode array (PDA) detector (YL 9160), a 96-vial programmable autosampler carbidopa oral suspension with the same concentration as the (YL 9150), a column oven compartment (YL 9130), a variable sample loop up to 200 μL and a software controller (Clarity). current study, but using a mixture of equal parts of the vehicles Ora-Plus® and Ora-Sweet®. This suspension remained stable for 42 days of storage when refrigerated, and the addition of ascorbic 3.3. Chromatographic conditions acid (2 mg/mL) increased the decomposition rate of the product. The chromatographic determinations were based upon USP methods for the APIs Nahata and colleagues also noticed a darker yellow color appearing or their final products, with minor modifications when necessary. The exact chro- in the product during longer storage. It seems that SyrSpend® matographic conditions used for each API are stated in Table 2. The columns were connected with a pre-column with the same packing (4.0 × 3.0 mm, 5 μm) from the SF PH4 (liquid) was able to maintain the potency and the phys- same vendor of the columns. ical characteristics of the Levodopa/Carbidopa suspension for a longer period, in comparison with the Ora-Plus® and Ora-Sweet® suspending vehicles. 3.4. Validation of the HPLC method A longer beyond-use date was also found for tacrolimus mono- Validation protocol and the acceptance criteria were established based upon USP ® (2015) and ICH (International Conference on Harmonization) (2005) guidelines. hydrate in SyrSpend SF PH4 (liquid) (at least 90 days at both Specificity of the method was determined by running HPLC analyses of a standard studied temperatures) when compared to a previous study by solution, a SyrSpend® SF PH4 (liquid) blank solution, and a mobile phase/diluents Jacobson et al. (1997). In this work, a tacrolimus 0.5 mg/mL oral blank solution. The acceptance criterion was defined as a percentage of discrepancy | suspension was prepared from capsules using a mixture of equal {[(standard area – sample area) / standard area] x 100} | between the peak areas of less ® than 2 %. In addition, the specificity of the method was obtained through comparison parts of Ora-Plus and simple syrup. This suspension was found to of standard chromatograms with and without the SyrSpend® SF PH4 (liquid) matrix. be stable for 56 days when stored at 25 °C, compared to 90 days All analyses were run in triplicate. found in SyrSpend® SF PH4 (liquid). Precision was evaluated as repeatability and intermediate precision. Repeatability was Terbinafine hydrochloride in SyrSpend® SF PH 4 (liquid) showed determined by consecutively analyzing six replicates by a single analyst in a single a stability of at least 90 days when stored at both refrigerated and day. Intermediate precision was also performed in six replicates, but over two days, by different analysts. An injection precision of more than 95 % (coefficient of variation, at room temperature, while Abdel-Rahman and Nahata (1999) CV) was considered acceptable. reported a maximum stability of 42 days for syrups of this API The accuracy of the method was determined through spike-recovery of the SyrSpend® compounded with Ora-Plus® and Ora-Sweet® (1:1, v/v) in the same SF PH4 (liquid) matrix, diluted within the range used for final sample measurements concentration and stored at 4 °C and 25 °C. (to the calibration curves). Percent recovery was calculated from the concentration ® measured relative to the theoretical concentration spiked. Tramadol Hydrochloride 10 mg/mL in SyrSpend SF PH 4 (liquid) For linearity, concentrations from 70-130 % of the working concentration of the API presented similar results to other works, although all these studies in SyrSpend® SF PH4 (liquid) were prepared, and analyzed. The data from each evaluated products with concentrations lower than 10 mg/mL. experiment was fitted by ordinary least squares method and was evaluated by analysis Wagner et al. (2003) evaluated 5 mg/mL syrups compounded of variance (ANOVA). ® The limit of detection (LOD) and limit of quantification (LOQ) were determined from with equal parts mixtures of Ora-Plus with strawberry syrup or three standard calibration curves of the APIs in the presence of the SyrSpend® SF PH4 Ora-Sweet®. No significant loss of API in either formulation at (liquid) matrix and were calculated as shown in Eqs. (1) and (2), respectively: two storage conditions (3-5 °C and 23-25 °C) over 91 days was 3 reported. Johnson et al. (2004) evaluated the same suspending LOD s (1) a vehicles for compounding a combination of tramadol hydrochlo- ride 7.5 mg/mL and acetaminophen 65 mg/mL, and found equal 10 LOQ s (2) stability results up to 90 days. These two studies, however, were a conducted using syrups compounded from commercial tablets. where a is the slope of the calibration curve, and s is the standard deviation of the Finally, for haloperidol, imipramine hydrochloride, minocycline, y-intercept. The LOD and LOQ were confirmed by the analysis of chromatograms and valsartan oral liquids, the authors found no previous reports in generated by injecting solutions in their respective limit concentrations. the literature. This makes this work the first report of the stability of these APIs in compounded oral liquids. 3.5. Preparation of API suspension samples In conclusion, this study demonstrates that most of the aforemen- ® The API suspensions were prepared using the following general protocol: (i) the tioned suspensions of 11 APIs in SyrSpend SF PH4 (liquid) were required quantity of each ingredient for the total amount to be prepared was calcu- stable for at least 90 days after preparation, when stored both at lated; (ii) each ingredient was accurately weighed; (iii) the API was placed in a mortar refrigerated and at room temperature. Exceptions with shorter and triturated until a fine powder was obtained; (iv) a small amount of the SyrSpend® stability are: Minocycline hydrochloride at both storage tempera- SF PH4 (liquid) was added to the powder and mixed to form a uniform paste; (v) the SyrSpend® SF PH4 (liquid) was further added in approximately geometric portions tures (60 days), levodopa/carbidopa at room temperature (30 days), almost to volume, mixing thoroughly after each addition; (vi) sufficient SyrSpend® and lorazepam at room temperature (60 days). This indicates that SF PH4 (liquid) was added to bring the volume to 300 mL, and then mixed well; (v) SyrSpend® SF PH4 (liquid) is a suitable vehicle for compounding the final product was packaged in low-actinic, light-resistant prescription bottles and with a wide range of APIs from diverse pharmacological classes. labeled. The final concentrations in the bottles are summarized in Table 1. The suspensions were then immediately assayed at T = 0, and then separated into two It also suggests probable success for validating the APIs evaluated different 150 mL bottles: one sample was stored at refrigerated (2-8 ºC) and the other in this study for use in multiple-use oral suspensions, likely to be at controlled room temperature (20-25 ºC), for the duration of the study (temperature used in clinical applications by pharmacists or drug manufacturers. and humidity were checked in real time throughout the whole experiment, using a calibrated, digital thermo-hygrometer (Incoterm)). 3. Experimental 3.6. Forced-degradation studies: stability-indicating characteristics 3.1. Reagents, reference standards and materials API samples were subjected to the following stressing conditions to determine the All API raw materials and SyrSpend® SF PH4 (liquid) (batch number 14F02-U59- capacity of the HPLC method to detect any possible degradation products that may 019404) were obtained from Fagron (St. Paul, MN, USA). HPLC-grade reagents arise during storage of the oral suspension: (i) dilution in acid (0.1M HCl, at 25 °C); (Panreac, Barcelona, Spain) were used. Ultrapure water obtained with an Aqua- (ii) dilution in base (0.1M NaOH, at 25 °C); (iii) exposure to ultraviolet light at 365 nm

Max-Ultra 370 Series (Young Lin, Anyang, Korea) (18.2 MΩ·cm resistivity at 25°C (at 25 °C); (iv) heating at 70 °C; and (v) dilution in H2O2 35 % (v/v) (at 25 °C). These and <10 ppb total organic carbon) was used throughout the experiments. The reference solutions were prepared for each API at its respective work concentration by means standards used were all work standards obtained using primary USP (Rockville, MD, of serial dilution from a stock-solution and using suitable diluents (see Table 2). The USA) reference materials. All the mobile phases and receptor media were filtered stock-solutions were sonically dispersed by 10 minutes and the final solutions were through a 0.45 μm filter membrane (RC-45/15 MS, Chromafil, Düren, Germany) and filtered (15 mm regenerated cellulose syringe filters, with 0.45 μm pore size) before degassed using an ultrasonic apparatus (model 1600A, Unique, Indaiatuba, Brazil) for injection onto the HPLC system. Any extraneous peaks found in the chromatograms 190 Pharmazie 71 (2016) ORIGINAL ARTICLES

were labeled. A resolution of 1.5 between the peaks of the degradation products and Glass BD, Haywood A (2006) Stability considerations in liquid dosage forms extem- the API was considered full separation. Also, a discrepancy greater than 2 % between poraneously prepared from commercially available products. J Pharm Pharm Sci the stressed sample peak and the standard, non-stressed sample peak was considered 9: 398-426. indicative of API decomposition. Hill EM, Flaitz CM, Frost GR (1988) Sweetener content of common pediatric oral liquid medications. Am J Hosp Pharm 45: 135-142. ICH - International Conference On Harmonisation Of Technical Requirements For 3.7. Stability study Registration Of Pharmaceuticals For Human Use (2005). Validation of Analytical Procedures: text and methodology Q2(R1). The API samples were assayed by HPLC at pre-determined time points to verify Jacobson PA, Johnson CE, West NJ, Foster JA (1997) Stability of tacrolimus in an the stability of the API in SyrSpend® SF PH4 (liquid). Before analyses, the bottles extemporaneously compounded oral liquid. Am J Health Syst Pharm 54: 178-180. were shaken until the API was uniformly dispersed by visual inspection. Aliquots for Jijo A, Flowerlet M (2014) Taste masking of peadiatric formulation: a review on tech- quantification (variable for each API) were withdrawn from the middle of the bottles, nologies, recent trends and regulatory aspects. Int J Pharm Pharm Sci 6: 12-19. without contact with the inner surface of the bottle, and diluted in order to obtain work Johnson CE, Wagner DS, DeLoach SL, Cichon-Hensley BK (2004) Stability of solutions in the concentration described in Table 2. Sampling times were: initial (T = tramadol hydrochloride-acetaminophen (Ultracet) in strawberry syrup and in a 0), 7 days (T = 7), 14 days (T = 14), 30 days (T = 30), 60 days (T = 60) and 90 days sugar-free vehicle. Am J Health Syst Pharm 61: 54-57. (T = 90). All suspensions were immediately assayed six times (6 aliquots) at each Johnston KR, Govel LA, Andritz MH (1994) Gastrointestinal effects of sorbitol as an time point (samples were diluted, sonicated for 10 minutes and then filtered in 15 mm additive in liquid medications. Am J Med 97: 185-191. regenerated cellulose syringe filters, with 0.45 μm pore size, before injection onto Lau ETL, Steadman KJ, Mak M, Cichero JAY, Nissen LM (2015) Prevalence of swal- the HPLC system). The evaluation parameter was the percent recovery with respect lowing difficulties and medication modification in customers of community phar- to T = 0, using the HPLC method (results given as percentage ± standard deviation). macists. J Pharm Pract Res 45: 18-23. Marquis J, Schneider MP, Payot V, Cordonier AC, Bugnon O, Hersberger KE (2013) References Swallowing difficulties with oral drugs among polypharmacy patients attending community pharmacies. Int J Clin Pharm 35: 1130-1136. Abdel-Rahman SM, Nahata MC (1999) Stability of terbinafine hydrochloride in an The Merck Index (2006) 14th ed. Merck & Co., Inc. Whitehouse Station, NJ. extemporaneously prepared oral suspension at 25 and 4 °C. Am J Health Syst Nahata MC, Morosco RS, Leguire LE (2000) Development of two stable oral suspen- Pharm 56: 243-245. sions of levodopa-carbidopa for children with amblyopia. J Pediatr Ophtalmol Allen LV (2008) Dosage form design and development. Clin Ther 30: 2102–2111. Stabismus 37: 333-337. Audicana Berasategui MT, Barasona Villarejo MJ, Corominas Sánchez M De Barrio Pappert EJ, Lipton JW, Goetz CG, Ling ZD, Stebbins GT, Carvey PM (1996) The Fernández M, García Avilés MC, García Robaina JC, Gastaminza Lasarte G, stability of carbidopa in solution. Movement Disorders 11: 24-26. Laguna Martínez JJ, Lobera Labairu T, López San Martín M, Martín Lázaro J, Payne ML, Craig WJ, Williams AC (1997) Sorbitol is a possible risk factor for diar- Moreno Rodilla E, Ortega Rodríguez N, Torres Jaén MJ (2011) Potential hyper- rhea in young children. J Am Diet Assoc 97: 532-534. sensitivity due to the food or food additive content of medicinal products in Spain. Sakai S, Adachi R, Miyazaki T, Aso Y, Okuda H, Teshima R (2012) Studies on J Investig Allergol Clin Immunol. 21: 496-506. the food allergenic proteins contained in pharmaceutical excipients. Kokuritsu BP – British Pharmacopoeia Commission Office (2015). British Pharmacopoeia Iyakuhin Shokuhin Eisei Kenkyusho Hokoku 130: 58-65. 2015. London: The Stationery Office. Schirm E, Tobi H, de Vries TW, Choonara I, De Jong-van den Berg LTW (2003) Brion F, Nunn AJ, Rieutord A (2003) Extemporaneous (magistral) preparation of oral Lack of appropriate formulations of medicines for children in the community. Acta medicines for children in European hospitals. Acta Paediatr 92: 486-90. Paediatr 92: 1486-1489. Centers for Disease Control (1982) Neonatal deaths associated with use of benzyl Sorenson B, Whaley P (2012) Stability of Rifampin in SyrSpend SF. Int J Pharm alcohol- United States. MMWR 31: 290-291. Compd 17: 162-164. Committee on Drugs, American Academy of Pediatrics (1997) “Inactive” Ingredients Sorenson B, Voudrie MA, Gehrig D (2012) Stability of in SyrSpend SF. in Pharmaceutical Products: Update (Subject Review). Pediatrics 99: 268-78. Int J Pharm Compd. 16: 347-349. Committee on Fetus and Newborn, Committee on Drugs, American Academy of Rood JM, Engels MJ, Ciarkowski SL, Wagenknecht LD, Dickinson CJ, Stevenson JG Pediatrics (1983) Benzyl alcohol: Toxic agent in neonatal units. Pediatrics. 72: (2014) Variability in compounding of oral liquids for pediatric patients: A patient 356-358. safety concern. JAPhA 54: 383-389. Connors KA, Amidon GL, Stella VJ (1986) Chemical stability of pharmaceuticals, Stiles ML, Allen LC, Prince SJ (1996) Stability of deferoxamine mesylate, floxu- 2nd ed., New York: John Wiley and Sons. ridine, fluororacil, hydromorphone hydrochloride, lorazepam, and Conroy S (2003) Extemporaneous (magistral) preparation of oral medicines for chil- hydrochloride in polypropylene infusion – pump syringes. Am J Health Syst dren in European hospitals. Acta Paediatr 92: 408-410. Pharm 53: 1583-1588. EP – Council of Europe (2015). European Pharmacopoeia 8.0. USP – United States Pharmacopeial Convention (2015) Inc. <1225> Validation of Fabiano V, Mameli C, Zuccotti GV (2011) Paediatric : Remember the Compendial Procedures. United States Pharmacopeia 38–National Formulary 33, excipients. Pharmacol Res 63: 362-365. v. 1, p. 1445-1450. Rockville, MD: US Pharmacopeial Convention, Inc. Ferreira AO, Polonini HC, Silva SL, Patrício FB, Brandão MA, Raposo NR (2016) Voudrie MA, Allen DB (2010) Stability of oseltamivir phosphate in SyrSpend SF, Feasibility of besylate, chloroquine phosphate, dapsone, , cherry syrup, and SyrSpend SF (for reconstitution). Int J Pharm Compd 14: 82-85. pyridoxine hydrochloride, sulfadiazine, sulfasalazine, tetracycline hydrochloride, Voudrie MA, Alexander B, Allen DB (2011) Stability of hydrochloride trimethoprim and in SyrSpend® SF PH4 oral suspensions. J Pharm compared to sorbitol containing syrup and suspending vehicles. Int J Pharm Biom Anal 118: 105-112. Compd 15: 255-258. Fiocchi MD, Riva E, Giovannini M (1999) Ethanol in medicines and other products Vu NT, Aloumanis V, Ben M (2008) Stability of metronidazole benzoate in SyrSpend intended for children. Nutr Res 19: 373-379. SF one-step suspension system. Int J Pharm Compd 12: 558-564. Geiger CM, Voudrie MA, Sorenson B (2012a) Stability of ursodiol in SyrSpend SF Wagner DS, Johnson CE, Cichon-Hensley BK, DeLoach SL (2003) Stability of oral cherry flavored. Int J Pharm Compd 16: 510-512. liquid preparations of tramadol in strawberry syrup and sugar-free vehicle. Am J Geiger CM, Voudrie MA, Sorenson B (2012b) Stability of hydrochloride Health Syst Pharm 60: 1268-1270. in SyrSpend SF. Int J Pharm Compd. 16: 513-515. Whaley PA, Voudrie MA, Sorenson B (2012a) Stability of omeprazole in SyrSpend Geiger CM, Sorenson B, Whaley PA (2013a) Stability of captopril in SyrSpend SF. Int SF Alka (reconstituted). Int J Pharm Compd 16: 164-166. J Pharm Compd 17: 336-338. Whaley PA, Voudrie MA, Sorenson B (2012b) Stability of vancomycin in SyrSpend Geiger CM, Sorenson B, Whaley PA (2013b) Stability of midazolam in SyrSpend SF SF. Int J Pharm Compd 16: 167-169. and SyrSpend SF Cherry. Int J Pharm Compd 17: 344-346. Zuccotti GV, Fabiano V (2011) Safety issues with ethanol as an excipient in drugs Geiger CM, Sorenson B, Whaley P (2015) Stability assessment of 10 active Pharma- intended for pediatric use. Expert Opin Drug Saf 10: 499-502. ceutical ingredients compounded in SyrSpend SF. Int J Pharm Compd 19: 427-435. Gershanik J, Boecler B, Ensley H, McCloskey S, George W (1982) The gasping syndrome and benzyl alcohol poisoning. New Engl J Med 307: 1384-1388.

Pharmazie 71 (2016) 191