- Email: [email protected]
Factors affecting the stability of nutritional emulsion mixtures
CLINICALNUTRlTION(1989)8:167-171 0 Longman Group UK Ltd 1989
Factors Affecting the Stability of Nutritional Emulsion Mixtures J. C. Shelton, A. Smith, and I. R. Thrussell* *Kendall Laboratories Ltd., Basingstoke RG21 2XZ. The Boots Company Plc, Notttingham NG2 3AA (Correspondence Dr. A. Smith)
to
SUMMARY The stability of oil droplets in complex nutritional admixtures is known to be affected by the presence of amino acids, dextrose and electrolytes. Hospital Pharmacists unable to meet demands for TPN admixtures sometimes turn to specialised commercial units but the supply of lipid admixtures by any facility remote from the user raises further challenges to admixture stability from various sources through exposure to temperature variations, transit effects etc. A number of interesting variances have been found in lipid admixtures arising from factors not previously recognised as significant. The batch of emulsion, container type and conditions of storage prior to use have been demonstrated to affect a moderately stable admixture. Whilst requiring further investigation, such factors must be considered for all new regimens and particularly to ensure the robustness of mixtures provided through a Compounding Service for Hospital or Home patients.
INTRODUCTION Intravenous nutritional solutions containing amino acids, dextrose and electrolytes are commonly administered from a single container. Despite well recognised problems related to stability, lipid emulsions are incorporated into such admixtures to provide additional calories and to supply essential fatty acids. The droplets in commercially available emulsions for parenteral nutrition range from a few nanometres in diameter up to several microns [ 1, 21. Provided that gross creaming effects are absent, the coalescence of emulsion droplets is the most important and potentially hazardous form of instability in complex nutritional admixtures [ 1, 21. The effects of amino acids, dextrose and electrolytes on stability are, therefore, widely documented [l-lo] thus establishing conditions whereby mixtures may remain sufficiently stable for periods ranging up to several days or weeks. However, little emphasis has previously been given to other factors such as the container type, variability of the lipid emulsion and the effect of storage conditions, all of which may exert an influence on stability.
Table 1
Freamine III 8S0, Intralipid 20°, dextrose 4096 electrolytes (mmol)
AND METHODS
Znl+ cllz+ MnJ’ MO”
120 20 5 2
Solivito Vitlipid adult
1 vial 10ml
(b) which were 12 and 14 months from expiry respectively, provided from pharmacy stock by a local hospital.
Compounding
and storage
Five admixtures
(Table
2
Container No.
Materials 1 Admixtures were compounded to the formula in Table 1 using a single batch of Freamine III (McGaw). The lipid emulsion included was Intralipid 20% (KabiVitrum) from randomly selected batches 54597 (a) 55350
1000 ml 500 ml 1OOOml 120 100 7.5 7.5 10
Na+ K’ Cal+ Mg=+ phosphate
trace elements (pmol)
Table MATERIALS
Selected formula for investigation
2 3 4 5 (control)
167
2) were compounded
in three-
Details of compounded admixtures
Type
Intralipid batch
pH
PVC EVA PVC EVA
a
6.12
: b b
6.11 6.15 6.13 6.12
PVC
168
FACTORSAFFECTING THE STABILITY OF NUTRITIONAL EMULSION MIXTURES
litre capacity polyvinyl chloride (PVC, Boots) or ethylenevinyl acetate (EVA, Travenol) containers under laminar flow. Electrolytes and trace elements were incorporated in the form of a proprietary dextroseelectrolyte-trace element mixture (Polyfusor, Boots). Lipid emulsion was added last into the well mixed solution. Four of the admixtures (two each in PVC and EVA) were refrigerated for 7 days then transferred to an incubator at 30°C to represent an extreme of practical use. A control admixture in PVC was stored at 30°C during the 2 week study period. Analyses
A close visual inspection of all admixtures was carried out at frequent intervals to detect coalescence (free oil) and/or creaming. Droplet size analysis was performed over the size range 0.9-22 urn using the Coulter Counter model TAII fitted with a 50~m orifice tube and calibrated using polystyrene divinyl benzene latex (8.7um). All samples were prediluted with distilled water and analysed in 0.9% sodium chloride solution which had no adverse effects on the droplet size over the period of measurement. Data are presented as a mean size and percentage of droplets above an arbitrary diameter (1Sum). A Zeiss microscope incorporating a haemacytometer cell (Neubauer 0.1 mm Hawksley) was used to supplement Coulter Counter data. Samples from the admixtures were diluted with glycerol 50yb (in 0.9% sodium chloride solution) and droplets sized directly in 2.4 pm ranges by means of a calibrated eyepiece. Ten full fields were assessed and data expressed as the average number of droplets per field. Preliminary studies confirmed that the dilution does not influence stability over the period of measurement. RESULTS
The results of visual inspection are given in Table 3. Table 3 Container No. 1 2 3 4 5
Visual observation of admixtures
1
3
5
7
-
SC SC c SC SC
SC SC c * SC
SC SC * SC
-
- = Homogeneous SC = Slight creaming C = Creaming l = Free oil-admixture discarded T = possible trace free oil
Day 8
9
10
13
SC SC
SC
SC
*
SC
SC
SC
T
l
A slight separation of cream was observed in all containers after 2 days storage but this did not worsen noticeably throughout the study and was redispersed by gentle agitation. Free oil appeared most rapidly in admixtures 3 and 4 containing lipid emulsion batch b. Surprisingly the control at 3O”C, which also contained batch b, was considerably more stable although slight discolouration together with slight creaming made the eventual confirmation of free oil difficult. Admixtures in EVA (2 and 4) produced free oil significantly earlier than comparable mixtures in PVC (1 and 3). Furthermore, in PVC containers free oil occurred as discrete droplets or pools, whereas in EVA, oil adhered to the surface in a film and was difficult to dislodge. In general the results from both Coulter counting (Table 4) and microscope examination (Table 5) support the stability information obtained by gross visual inspection. However, Coulter counting appeared relatively insensitive to progressive changes in the particular emulsions studied until significant coalescence had occurred and free oil was imminent. Microscopic counting was more sensitive to the appearance of large droplets and correlates well with visual observations, thus confirming the relatively poor stability of mixtures containing lipid emulsion batch b and the effect of the container type.
DISCUSSION Stability testing
It has been emphasised previously that reliance on a single method of stability assessment for nutritional emulsions should be avoided if erroneous conclusions are to be prevented [ 11. The present study shows this to be true even when the submicron population of droplets is ignored and attention is confined to the narrower droplet size range where changes are potentially harmful to the patient. Although a convenient and useful technique, Coulter counting requires the use of supporting methods since it is susceptible to sampling problems, insensitive to the appearance of relatively small numbers of large droplets and unable to discriminate between flocculation and coalescence. An excessive reliance on visual inspection is clearly to be avoided [ 1 l] whilst microscopic examination again relies on an extremely small sample. Only through the correct application of a combination of techniques can the proper elucidation of stability proceed.
Effect of electrolytes
It is widely recognised that added electrolytes represent the major factor giving rise to emulsion instability in
CLINICAL NUTRITION
Table 4
Summary of stability data from Coulter Counter
Contpiner No.
1
3
Day 5
Mw I,0
1.04 2.9 1.03 2.3
1.02 1.5 1.03 1.8
3
Mw 00
1.02 2.7
1.04 3.5
4
Mw 00
1.03 3.5
Mw O’”
1.03 3.0
1
Mw 00
2
5
169
7
8
10
13
1.02
1.04
1.02
1.03
-
1.4 1.04 2.2
2.2 1.03 2.0
1.4 -
2.4 _ -
-
1.09 8.0
-
-
-
-
1.03 3.5
-
-
-
-
-
1.03 3.0
1.04 3.6
1.04 3.2
1.05 4.5
1.05 3.8
1.08 6.6
-
Mw = Mean diameter in pm 011= Per cent above 1.5 pm Table 5
Container No.
Summary of stability data from microscopic examination: Average number of droplets per field Size range (microns)
1
3
5
1
2.4-4.8 4.8-7.2 7.2-9.6 9.6-12.0
6.0 0.2 0 0
6.3 0.1 0 0
5.8 0.1 0 0
2
2.4-4.8 4.8-7.2 7.2-9.6 9.6-12.0
5.1 0 0 0
4.7 0.2 0 0
10.4 0.2 0 0
3
2.4-4.8 4.8-7.2 7.2-9.6 9.6-12.0
7.4 0 0 0
9.4 1.9 0.5 0.1
20.0 4.0 0.2 0.4
4
2.4-4.8 4.8-7.2 7.2-9.6 9.6-12.0
5.4 0.4 0 0
15.2 1.8 0.3 0.1
5
2.4-4.8 4X4-7.2 7.2-9.6 9.6-12.0
6.2 0 0 0
nutritional mixtures. Electrostatic considerations have been used to propose guidelines for the total acceptable electrolyte load [l, 41. In our laboratories Freaminedextrose-Intralipid mixtures have remained stable for many months in the absence of added electrolyte. Satisfactory stability has also been reported for similar mixtures containing low levels of electrolyte [8]. Although both observations are consistent with theory, Jeppsson and Sjoberg [6] have pointed out that the electrostatic approach does not provide unequivocal evidence of stability. For example, a minimum in the stabilityconcentration profile has been noted which produces systems stabilised against coalescence at high electro-
8.5 0.7 0 0
Day 7
8
9
10
14.5 0.6 0.1 0
18.0 0.9 0.2 0.1
13.3 0.5 0 0.1
13.6 0.8 0.2 0
22.4 0.5 0.1 0
16.4 3.8 0.8 0.4
9.5 0.2 0 0
18.1 0.6 0.1 0
13.3 0.8 0 0
20.1 0.9 0.1 0.1
13
31.3 2.9 0.2 0
lyte concentrations [6, 71. Stable high-electrolyte formulae have also been reported for Freamine [lo]. Whilst such stabilisation has been attributed to charge reversal, neither our own laboratory studies nor the work of Jeppsson and Sjoberg [6] have established this for compounded mixtures at the levels of divalent cations involved. It appears likely that interfacial interactions affecting the structure or viscosity of the stabilising film are also implicated. Since optimum stability may be obtained at very low or high electrolyte levels, it follows that the intermediate region in which many clinically useful regimens fall warrants particular attention. The present
170
FACTORS AFFECTING THE STABILITY OF NUTRITIONAL
investigation of a moderately stable admixture containing electrolytes has revealed several additional factors which may affect stability.
EMULSION MIXTURES
ent
lipophilicity of the EVA surface may be associated with the greater rate of free oil production.
Effect of temperature Effect of emulsion
All the mixtures tested were similar initially with no apparent differences arising from the batch of lipid emulsion used. However, marked differences in stability were observed on storage. Significant coalescence was detected within 3 days in refrigerated admixtures containing emulsion batch b (Table 5) and free oil was evident in as little as 4 days in EVA (Table 3). In contrast, mixtures based upon batch a exhibited much improved stability in either container. Although progressive changes in the droplet size distribution were revealed by microscopy, these were not significant up to 7 days and free oil did not appear unti19 days in EVA and 13 days in PVC (Table 3). Several studies have taken place concerning the effect of ageing on the lipid emulsion [12, 13). Most recently Washington and Davis [ 131 considered the relationship between age and pH, zeta potential and flocculation behaviour. Increased critical flocculation concentrations of calcium chloride in older batches were attributed to the production of free fatty acids. Although the age difference of the present emulsions may not itself be sufficient to account for the observed stability behaviour, interfacial differences at manufacture or arising through variations in storage conditions may be implicated. Whilst it is presently recommended to maintain the lipid emulsion unfrozen below 25”C, the precise conditions of storage may vary considerably. Clearly the substitution of intravenous emulsions from alternative manufacturers in a given mixture would be totally inappropriate without extensive stability testing in support.
A period of refrigerated storage is commonly employed for nutritional emulsions not required for immediate infusion. A low temperature primarily inhibits bacterial growth but has also been regarded as generally favourable to emulsion stability. The present study reveals that the latter supposition is not invariably correct. Relatively low stability was characteristic of those admixtures containing lipid emulsion batch b. Remarkably, the control admixture at 30°C containing this batch remained significantly more stable. One possibility is that the elevated temperature may increase electrolyte dissociation into a region of improved stability. Conversely, refrigeration may reduce the ionised electrolyte level such that admixtures enter a region of reduced stability. The limited evidence (admixture 2) that a temperature change after refrigeration may also affect stability adversely warrants further study, particularly in view of the temperature changes incurred by mixtures in practice.
CONCLUSIONS The stability of oil droplets in complex nutritional admixtures is known to be affected by the presence of amino acids, dextrose and electrolytes. In addition, factors such as the batch of emulsion, container type and conditions of storage prior to use have also been shown to affect emulsion stability and warrant further study. For a Compounding Service providing bags for Hospital or Home patients such considerations must be routine for all new regimens. REFERENCES
Effect of container
As in the present case, we have consistently observed that admixtures compounded in PVC containers remained demonstrably more stable than comparable ones in EVA when stored under identical conditions. This phenomenon does not appear to have been reported previously and its cause remains obscure at present. Plasticizer release will be dealt with in a further publication [14] but it should be noted that we have been unable to demonstrate any stabilising effect from the components of PVC plastic. The possible leaching of low molecular weight fractions from EVA to cause a destabilising effect has not yet been investigated. However, it is pertinent to note the different appearance of free oil in the respective containers. The higher appar-
(11Davis S S 1983 The stability of fat emulsions for
intravenous administration. In Johnson I D A (ed)
Advances in Clinical Nutrition. MTP Press, Lancaster, pp 213-239 PI Whateley T L, Steele G, Unvin J, Smail G A 1984 Particle size stability of Intralipid and mixed total parenteral nutrition mixtures. Journal of Clinical and Hospital Pharmacy 9: 113-126 131Black CD, Popovich N G 1981 A study of intravenous emulsion compatibility: effects of dextrose, amino acids, and selected electrolytes. Drug Intelligence and Clinical Pharmacy 15: 184-193 [41 Bumham W R, Hansrani P K, Knott C E, Cook J A, Davis S S 1982 Stability of a fat emulsion based intravenous feeding mixture. International Journal of Pharmaceutics 13: 9-22 [51 Takamura A, Ishii F, Noro S, Tanifuji M, N&ajima S 1984 Study of intravenous hyperalimentation: effect of
CLINICAL NUTRITION
[6]
[7]
[8]
[9]
selected amino acids on the stability of intravenous fat emulsions. Journal of Pharmaceutical Sciences 73: 91-94 Jeppsson R I, Sjoberg B 1984 Compatibility of parenteral nutrition solutions when mixed in a plastic bag. Clinical Nutrition 2: 149-158 Davis S S, Galloway M 1986 Studies on fat emulsions in combined nutrition solutions. Journal of Clinical and Hospital Pharmacy 11: 33-45 Davis S S, Galloway M, Burnham W R, Stevens L 1986 In vitro and clinical studies on intravenous feeding mixtures comprising fat emulsion, amino acid and electrolytes. Clinical Nutrition 5: 21-27 Jeppsson R I, Tengborn H J 1987 One week’s stability of TPN mixtures in plastic bags. Clinical Nutrition 6: 155-160
Submission date: 23 May 1988.
WI McCormick
D C, Knutsen C V, Wells P A, Kaminski M V 1987 Physical stability of a trisubstrate admixture. Nutritional Support Services 7: 18,27-28 Pll El Eini D, Knott C E 1985 Stability of IV lipid emulsions. Pharmaceutical Journal 235: 170 [=I Boberg J, Hakansson I 1964 Physical and biological changes in an artificial fat emulsion during storage. Journal of Pharmacy and Pharmacology 16: 641-646 C, Davis S S 1987 Ageing effects in P31 Washington parenteral fat emulsions: the role of fatty acids. International Journal of Pharmaceutics 39: 33-39 [14] Smith A, Thrussell I R, Johnson G W 1989 The prevention of plasticizer migration into nutritional emulsion mixtures by use of a novel container. Clinical Nutrition: in press
Accepted after revision: 19 February
1989.
171
Factors Affecting the Stability of Nutritional Emulsion Mixtures J. C. Shelton, A. Smith, and I. R. Thrussell* *Kendall Laboratories Ltd., Basingstoke RG21 2XZ. The Boots Company Plc, Notttingham NG2 3AA (Correspondence Dr. A. Smith)
to
SUMMARY The stability of oil droplets in complex nutritional admixtures is known to be affected by the presence of amino acids, dextrose and electrolytes. Hospital Pharmacists unable to meet demands for TPN admixtures sometimes turn to specialised commercial units but the supply of lipid admixtures by any facility remote from the user raises further challenges to admixture stability from various sources through exposure to temperature variations, transit effects etc. A number of interesting variances have been found in lipid admixtures arising from factors not previously recognised as significant. The batch of emulsion, container type and conditions of storage prior to use have been demonstrated to affect a moderately stable admixture. Whilst requiring further investigation, such factors must be considered for all new regimens and particularly to ensure the robustness of mixtures provided through a Compounding Service for Hospital or Home patients.
INTRODUCTION Intravenous nutritional solutions containing amino acids, dextrose and electrolytes are commonly administered from a single container. Despite well recognised problems related to stability, lipid emulsions are incorporated into such admixtures to provide additional calories and to supply essential fatty acids. The droplets in commercially available emulsions for parenteral nutrition range from a few nanometres in diameter up to several microns [ 1, 21. Provided that gross creaming effects are absent, the coalescence of emulsion droplets is the most important and potentially hazardous form of instability in complex nutritional admixtures [ 1, 21. The effects of amino acids, dextrose and electrolytes on stability are, therefore, widely documented [l-lo] thus establishing conditions whereby mixtures may remain sufficiently stable for periods ranging up to several days or weeks. However, little emphasis has previously been given to other factors such as the container type, variability of the lipid emulsion and the effect of storage conditions, all of which may exert an influence on stability.
Table 1
Freamine III 8S0, Intralipid 20°, dextrose 4096 electrolytes (mmol)
AND METHODS
Znl+ cllz+ MnJ’ MO”
120 20 5 2
Solivito Vitlipid adult
1 vial 10ml
(b) which were 12 and 14 months from expiry respectively, provided from pharmacy stock by a local hospital.
Compounding
and storage
Five admixtures
(Table
2
Container No.
Materials 1 Admixtures were compounded to the formula in Table 1 using a single batch of Freamine III (McGaw). The lipid emulsion included was Intralipid 20% (KabiVitrum) from randomly selected batches 54597 (a) 55350
1000 ml 500 ml 1OOOml 120 100 7.5 7.5 10
Na+ K’ Cal+ Mg=+ phosphate
trace elements (pmol)
Table MATERIALS
Selected formula for investigation
2 3 4 5 (control)
167
2) were compounded
in three-
Details of compounded admixtures
Type
Intralipid batch
pH
PVC EVA PVC EVA
a
6.12
: b b
6.11 6.15 6.13 6.12
PVC
168
FACTORSAFFECTING THE STABILITY OF NUTRITIONAL EMULSION MIXTURES
litre capacity polyvinyl chloride (PVC, Boots) or ethylenevinyl acetate (EVA, Travenol) containers under laminar flow. Electrolytes and trace elements were incorporated in the form of a proprietary dextroseelectrolyte-trace element mixture (Polyfusor, Boots). Lipid emulsion was added last into the well mixed solution. Four of the admixtures (two each in PVC and EVA) were refrigerated for 7 days then transferred to an incubator at 30°C to represent an extreme of practical use. A control admixture in PVC was stored at 30°C during the 2 week study period. Analyses
A close visual inspection of all admixtures was carried out at frequent intervals to detect coalescence (free oil) and/or creaming. Droplet size analysis was performed over the size range 0.9-22 urn using the Coulter Counter model TAII fitted with a 50~m orifice tube and calibrated using polystyrene divinyl benzene latex (8.7um). All samples were prediluted with distilled water and analysed in 0.9% sodium chloride solution which had no adverse effects on the droplet size over the period of measurement. Data are presented as a mean size and percentage of droplets above an arbitrary diameter (1Sum). A Zeiss microscope incorporating a haemacytometer cell (Neubauer 0.1 mm Hawksley) was used to supplement Coulter Counter data. Samples from the admixtures were diluted with glycerol 50yb (in 0.9% sodium chloride solution) and droplets sized directly in 2.4 pm ranges by means of a calibrated eyepiece. Ten full fields were assessed and data expressed as the average number of droplets per field. Preliminary studies confirmed that the dilution does not influence stability over the period of measurement. RESULTS
The results of visual inspection are given in Table 3. Table 3 Container No. 1 2 3 4 5
Visual observation of admixtures
1
3
5
7
-
SC SC c SC SC
SC SC c * SC
SC SC * SC
-
- = Homogeneous SC = Slight creaming C = Creaming l = Free oil-admixture discarded T = possible trace free oil
Day 8
9
10
13
SC SC
SC
SC
*
SC
SC
SC
T
l
A slight separation of cream was observed in all containers after 2 days storage but this did not worsen noticeably throughout the study and was redispersed by gentle agitation. Free oil appeared most rapidly in admixtures 3 and 4 containing lipid emulsion batch b. Surprisingly the control at 3O”C, which also contained batch b, was considerably more stable although slight discolouration together with slight creaming made the eventual confirmation of free oil difficult. Admixtures in EVA (2 and 4) produced free oil significantly earlier than comparable mixtures in PVC (1 and 3). Furthermore, in PVC containers free oil occurred as discrete droplets or pools, whereas in EVA, oil adhered to the surface in a film and was difficult to dislodge. In general the results from both Coulter counting (Table 4) and microscope examination (Table 5) support the stability information obtained by gross visual inspection. However, Coulter counting appeared relatively insensitive to progressive changes in the particular emulsions studied until significant coalescence had occurred and free oil was imminent. Microscopic counting was more sensitive to the appearance of large droplets and correlates well with visual observations, thus confirming the relatively poor stability of mixtures containing lipid emulsion batch b and the effect of the container type.
DISCUSSION Stability testing
It has been emphasised previously that reliance on a single method of stability assessment for nutritional emulsions should be avoided if erroneous conclusions are to be prevented [ 11. The present study shows this to be true even when the submicron population of droplets is ignored and attention is confined to the narrower droplet size range where changes are potentially harmful to the patient. Although a convenient and useful technique, Coulter counting requires the use of supporting methods since it is susceptible to sampling problems, insensitive to the appearance of relatively small numbers of large droplets and unable to discriminate between flocculation and coalescence. An excessive reliance on visual inspection is clearly to be avoided [ 1 l] whilst microscopic examination again relies on an extremely small sample. Only through the correct application of a combination of techniques can the proper elucidation of stability proceed.
Effect of electrolytes
It is widely recognised that added electrolytes represent the major factor giving rise to emulsion instability in
CLINICAL NUTRITION
Table 4
Summary of stability data from Coulter Counter
Contpiner No.
1
3
Day 5
Mw I,0
1.04 2.9 1.03 2.3
1.02 1.5 1.03 1.8
3
Mw 00
1.02 2.7
1.04 3.5
4
Mw 00
1.03 3.5
Mw O’”
1.03 3.0
1
Mw 00
2
5
169
7
8
10
13
1.02
1.04
1.02
1.03
-
1.4 1.04 2.2
2.2 1.03 2.0
1.4 -
2.4 _ -
-
1.09 8.0
-
-
-
-
1.03 3.5
-
-
-
-
-
1.03 3.0
1.04 3.6
1.04 3.2
1.05 4.5
1.05 3.8
1.08 6.6
-
Mw = Mean diameter in pm 011= Per cent above 1.5 pm Table 5
Container No.
Summary of stability data from microscopic examination: Average number of droplets per field Size range (microns)
1
3
5
1
2.4-4.8 4.8-7.2 7.2-9.6 9.6-12.0
6.0 0.2 0 0
6.3 0.1 0 0
5.8 0.1 0 0
2
2.4-4.8 4.8-7.2 7.2-9.6 9.6-12.0
5.1 0 0 0
4.7 0.2 0 0
10.4 0.2 0 0
3
2.4-4.8 4.8-7.2 7.2-9.6 9.6-12.0
7.4 0 0 0
9.4 1.9 0.5 0.1
20.0 4.0 0.2 0.4
4
2.4-4.8 4.8-7.2 7.2-9.6 9.6-12.0
5.4 0.4 0 0
15.2 1.8 0.3 0.1
5
2.4-4.8 4X4-7.2 7.2-9.6 9.6-12.0
6.2 0 0 0
nutritional mixtures. Electrostatic considerations have been used to propose guidelines for the total acceptable electrolyte load [l, 41. In our laboratories Freaminedextrose-Intralipid mixtures have remained stable for many months in the absence of added electrolyte. Satisfactory stability has also been reported for similar mixtures containing low levels of electrolyte [8]. Although both observations are consistent with theory, Jeppsson and Sjoberg [6] have pointed out that the electrostatic approach does not provide unequivocal evidence of stability. For example, a minimum in the stabilityconcentration profile has been noted which produces systems stabilised against coalescence at high electro-
8.5 0.7 0 0
Day 7
8
9
10
14.5 0.6 0.1 0
18.0 0.9 0.2 0.1
13.3 0.5 0 0.1
13.6 0.8 0.2 0
22.4 0.5 0.1 0
16.4 3.8 0.8 0.4
9.5 0.2 0 0
18.1 0.6 0.1 0
13.3 0.8 0 0
20.1 0.9 0.1 0.1
13
31.3 2.9 0.2 0
lyte concentrations [6, 71. Stable high-electrolyte formulae have also been reported for Freamine [lo]. Whilst such stabilisation has been attributed to charge reversal, neither our own laboratory studies nor the work of Jeppsson and Sjoberg [6] have established this for compounded mixtures at the levels of divalent cations involved. It appears likely that interfacial interactions affecting the structure or viscosity of the stabilising film are also implicated. Since optimum stability may be obtained at very low or high electrolyte levels, it follows that the intermediate region in which many clinically useful regimens fall warrants particular attention. The present
170
FACTORS AFFECTING THE STABILITY OF NUTRITIONAL
investigation of a moderately stable admixture containing electrolytes has revealed several additional factors which may affect stability.
EMULSION MIXTURES
ent
lipophilicity of the EVA surface may be associated with the greater rate of free oil production.
Effect of temperature Effect of emulsion
All the mixtures tested were similar initially with no apparent differences arising from the batch of lipid emulsion used. However, marked differences in stability were observed on storage. Significant coalescence was detected within 3 days in refrigerated admixtures containing emulsion batch b (Table 5) and free oil was evident in as little as 4 days in EVA (Table 3). In contrast, mixtures based upon batch a exhibited much improved stability in either container. Although progressive changes in the droplet size distribution were revealed by microscopy, these were not significant up to 7 days and free oil did not appear unti19 days in EVA and 13 days in PVC (Table 3). Several studies have taken place concerning the effect of ageing on the lipid emulsion [12, 13). Most recently Washington and Davis [ 131 considered the relationship between age and pH, zeta potential and flocculation behaviour. Increased critical flocculation concentrations of calcium chloride in older batches were attributed to the production of free fatty acids. Although the age difference of the present emulsions may not itself be sufficient to account for the observed stability behaviour, interfacial differences at manufacture or arising through variations in storage conditions may be implicated. Whilst it is presently recommended to maintain the lipid emulsion unfrozen below 25”C, the precise conditions of storage may vary considerably. Clearly the substitution of intravenous emulsions from alternative manufacturers in a given mixture would be totally inappropriate without extensive stability testing in support.
A period of refrigerated storage is commonly employed for nutritional emulsions not required for immediate infusion. A low temperature primarily inhibits bacterial growth but has also been regarded as generally favourable to emulsion stability. The present study reveals that the latter supposition is not invariably correct. Relatively low stability was characteristic of those admixtures containing lipid emulsion batch b. Remarkably, the control admixture at 30°C containing this batch remained significantly more stable. One possibility is that the elevated temperature may increase electrolyte dissociation into a region of improved stability. Conversely, refrigeration may reduce the ionised electrolyte level such that admixtures enter a region of reduced stability. The limited evidence (admixture 2) that a temperature change after refrigeration may also affect stability adversely warrants further study, particularly in view of the temperature changes incurred by mixtures in practice.
CONCLUSIONS The stability of oil droplets in complex nutritional admixtures is known to be affected by the presence of amino acids, dextrose and electrolytes. In addition, factors such as the batch of emulsion, container type and conditions of storage prior to use have also been shown to affect emulsion stability and warrant further study. For a Compounding Service providing bags for Hospital or Home patients such considerations must be routine for all new regimens. REFERENCES
Effect of container
As in the present case, we have consistently observed that admixtures compounded in PVC containers remained demonstrably more stable than comparable ones in EVA when stored under identical conditions. This phenomenon does not appear to have been reported previously and its cause remains obscure at present. Plasticizer release will be dealt with in a further publication [14] but it should be noted that we have been unable to demonstrate any stabilising effect from the components of PVC plastic. The possible leaching of low molecular weight fractions from EVA to cause a destabilising effect has not yet been investigated. However, it is pertinent to note the different appearance of free oil in the respective containers. The higher appar-
(11Davis S S 1983 The stability of fat emulsions for
intravenous administration. In Johnson I D A (ed)
Advances in Clinical Nutrition. MTP Press, Lancaster, pp 213-239 PI Whateley T L, Steele G, Unvin J, Smail G A 1984 Particle size stability of Intralipid and mixed total parenteral nutrition mixtures. Journal of Clinical and Hospital Pharmacy 9: 113-126 131Black CD, Popovich N G 1981 A study of intravenous emulsion compatibility: effects of dextrose, amino acids, and selected electrolytes. Drug Intelligence and Clinical Pharmacy 15: 184-193 [41 Bumham W R, Hansrani P K, Knott C E, Cook J A, Davis S S 1982 Stability of a fat emulsion based intravenous feeding mixture. International Journal of Pharmaceutics 13: 9-22 [51 Takamura A, Ishii F, Noro S, Tanifuji M, N&ajima S 1984 Study of intravenous hyperalimentation: effect of
CLINICAL NUTRITION
[6]
[7]
[8]
[9]
selected amino acids on the stability of intravenous fat emulsions. Journal of Pharmaceutical Sciences 73: 91-94 Jeppsson R I, Sjoberg B 1984 Compatibility of parenteral nutrition solutions when mixed in a plastic bag. Clinical Nutrition 2: 149-158 Davis S S, Galloway M 1986 Studies on fat emulsions in combined nutrition solutions. Journal of Clinical and Hospital Pharmacy 11: 33-45 Davis S S, Galloway M, Burnham W R, Stevens L 1986 In vitro and clinical studies on intravenous feeding mixtures comprising fat emulsion, amino acid and electrolytes. Clinical Nutrition 5: 21-27 Jeppsson R I, Tengborn H J 1987 One week’s stability of TPN mixtures in plastic bags. Clinical Nutrition 6: 155-160
Submission date: 23 May 1988.
WI McCormick
D C, Knutsen C V, Wells P A, Kaminski M V 1987 Physical stability of a trisubstrate admixture. Nutritional Support Services 7: 18,27-28 Pll El Eini D, Knott C E 1985 Stability of IV lipid emulsions. Pharmaceutical Journal 235: 170 [=I Boberg J, Hakansson I 1964 Physical and biological changes in an artificial fat emulsion during storage. Journal of Pharmacy and Pharmacology 16: 641-646 C, Davis S S 1987 Ageing effects in P31 Washington parenteral fat emulsions: the role of fatty acids. International Journal of Pharmaceutics 39: 33-39 [14] Smith A, Thrussell I R, Johnson G W 1989 The prevention of plasticizer migration into nutritional emulsion mixtures by use of a novel container. Clinical Nutrition: in press
Accepted after revision: 19 February
1989.
171