Expert Review
Lipid Injectable Emulsions:Pharmacopeial and Safety Issues
David F.Driscoll 1,2,3,j
Received May 8,2006;accepted May 23,2006;published online August 9,2006
Abstract.Lipid injectable emulsions have been routinely ud in patients worldwide for over 40years as a nutritional supplement in patients requiring parenteral nutrition.They can be given as a parate infusion or added into total parenteral nutrition admixtures.Despite such broad u,no pharmacopeial standards exist with respect to the optimal pharmaceutical characteristics of the formulation.Several attempts to establish standard physical and chemical attributes have been attempted by various pharmacopeias around the world,but without success largely due to technical issues regarding the creation of globule size limits.Recently,the United States Pharmacopeia has revid its previous efforts and developed two methods and criteria (under Chapter <729>)to measure the mean droplet size (Method I),and the large-diameter tail >52m (Method II)of the globule size distribution to verify the stability of lipid injectable emulsions.Importantly,it is the latter size limits of Method II that have the greatest implications for infusion safety.The major safety issues involving lipid injectable emulsio
ns include impairments in plasma clearance in susceptible patients,and the infusion of an unstable emulsion containing large quantities of potentially embolic fat globules.Recent animal studies investigating the toxicity from the infusion of unstable lipid injectable emulsions have shown evidence of oxidative stress and tissue damage to the liver when recommended globule size limits determined by Method II of the USP are exceeded.Adoption of Chapter <729>of the USP ems appropriate at this time.
KEY WORDS:globule size distribution;infusion container;lipid injectable emulsions;safety;volume-weighted PFAT 5.
INTRODUCTION
Lipid injectable emulsions,as soybean oil-in-water for-mulations stabilized by egg phospholipids,have been ud worldwide in the clinical tting as a nutritional supplement for more than 40years.More recently,an array of oils,individually or as mixtures,have been ud in lipid injectable emulsions in addition to soybean oil,including safflower oil,medium chain triglycerides,olive and fish oils.Parenteral emulsions are commonly manufactured via a homogenizer that forces a concentrated oil Y emulsifier Y water mixture through a small orifice at very high pressures.This actio
n is repeated through veral cycles and subjects the mixture to very high shear forces,ideally producing a fine dispersion of submicron droplets with a narrow distribution,or low degree of polydispersity.There are three basic types of pharmaceutical emulsions which can be classified according to their mean droplet size (MDS):(1)a F micro-emulsion _,MDS <0.12m,such as liposomal drug formulations;(2)a F mini-emulsion _,MDS <1.02m,such as lipid injectable emulsions;and (3)a F macro-emulsion _,MDS >12m,such as chemoembolization infusions.The micro-emulsion
is transparent,forms spontaneously and is thermodynamically stable,whereas the latter two are turbid,require energy to be formed,and are thermodynamically unstable.
The clinical u of lipid injectable emulsions in the US has spanned approximately 30years.They have been principally indicated for patients requiring parenteral nutri-tion support as a source of esntial fatty acids (i.e.,linoleic and linolenic acids),and as a den source of F isotonic calories _.More recently they have been ud as drug delivery vehicles for poorly soluble drugs such as anesthetic/dative,propofol.They can be administered via the small veins of the peripheral venous ,relatively low-flow blood vesls (e.g.,basilic or cephalic veins)that are physiologically limited by the final tonicity of the infusion.In general,peripheral vein tolerance is achieved for reasonable periods of time (usual intravascular catheter life:up to 72h),before phlebitis inevitably occ
urs,as long as the osmolarity is kept between 600and 900mOsm/l (1,2).Higher osmolarity (e.g.,1,000Y 3,000mOsm/l)formulations,as encountered in the preparation of total parenteral nutrition admixtures,can be administered for prolonged periods (usual intravascular catheter life:7Y 10days in the hospital,to years in the home infusion tting)via large veins of the central venous circulation (e.g.,via the subclavian vein with radiographic confirmation of the catheter tip in the superior vena cava or SVC).The large volume of blood flow through the SVC reprents approximately one half of cardiac output or about 2,500ml/min,so catheter tip placement in this vesl provides maximal hemodilution of the infusate.Hence,lipid injectable
2010考研英语1959
0724-8741/06/0900-1959/0#2006Springer Science +Business Media,Inc.
Pharmaceutical Rearch,Vol.23,No.9,September 2006(#2006)DOI:10.1007/s11095-006-9092-4
1Harvard Medical School,Boston,Massachutts,USA.
slashing2
Department of Medicine,Beth Israel Deaconess Medical Center (BIDMC),Boston.3
To whom correspondence should be addresd.(e-mail:ddriscol@bidmc.harvard.edu)
j Dr.
Driscoll is a Consultant and/or Rearcher in the area of lipids for AstraZeneca,B.Braun,Biolink and Hospira companies.
emulsions may also be administered via the high-flow vesls as well.For most patients requiring total parenteral nutrition(TPN),where the osmolarity is very high from the combination of15Y18crystalline amino acids,dextro,10Y12 electrolytes,12Y13multivitamins and5Y7trace minerals,the u of central venous catheters to deliver this therapy, especially for prolonged periods of time,is mandatory.In addition,20%soybean oil lipid injectable emulsions are commonly added to the TPN formulations(known as all-in-one or total nutrient admixtures),often comprising up to 30%of the total caloric intake.Adding lipids transforms the conventional parenteral solution to an emulsion,thus engen-dering potentially significant stability issues arising from the numerous ionically active components typically found in the extemporaneously prepared infusions.
The United States Pharmacopeia(USP),who mission statement includes that it B promotes the public health and benefits practitioners and patients by disminating authorita-tive standards and inf
ormation’’(3),is responsible for creating official ,drug monographs and chapters)that are recognized and enforceable by the Food and Drug Adminis-tration(FDA).Under usual circumstances,the goal of the USP is to generate an official drug monograph within3Y5 years after its FDA approval(personal communication,Roger Williams,Executive Director,USP,October24,2000).With respect to lipid injectable emulsions,the USP began its first pharmacopeial preview in1991of a monograph entitled B Intravenous Fat Emulsion’’(4)and the associated chapter <728>entitled B Globule Size Distribution in Intravenous Emulsions’’(5),approximately15years after their introduc-tion into the US for clinical u.Several F in-process_revisions followed in1994,1995and1998,without adoption of an official monograph or chapter.The task was referred to another USP expert committee in2000,resulting in the publication of another monograph in2003.Unfortunately, like the previous versions,no globule size limits were suggested.However,without globule size limits and the appropriate instrumentation to define them,the esntial question relating to emulsion stability and subquent safety, could not be answered.In2004,the USP took a different approach and decided to first publish a new version of chapter <729>(formerly<728>)entitled B Globule Size Distribution in Lipid Injectable Emulsions’’(6)which detailed two methods of globule size analysis with specific pharmacopeial limits. Since then,a cond chapter revision has been published in 2005(7)along with a new monograph en
titled B Lipid Injectable Emulsions’’in2005(8)and again most recently in 2006(9),that incorporates the propod globule size limits.
There are two goals in this review.The first will be to detail the key issues regarding the propod globule size limits of USP <729>with respect to the physical stability of lipid injectable emulsions,both in the native state and after inclusion in the TPN bag as a total nutrient admixture(TNA).The cond goal, as a logical extension of the aforementioned stability issue,will be to discuss the implications of globule size for infusion safety. PHARMACOPEIAL ISSUES AND LIPID INJECTABLE EMULSIONS
The current edition of USP chapter<729>(7)and its accompanying monograph(9)makes clear the desirable physical and chemical characteristics of lipid injectable emulsions in meeting pharmacopeial standards.Destabiliza-tion of the mini-emulsions via coalescence is the inevitable outcome of the thermodynamically unstable dosage forms. The goal of the USP in outlining the globule size limits is to ensure that the dosage form does not prematurely progress to a stage where the process of coalescence advances to a critical point before the end of its shelf-life,where the safety of the infusion is ,the formation of potentially embolic fat globules larger than52m).There are a variety of common physical and chemical stress that can accelerate the des
tabilization process.In one ca, destabilization is inadvertent and ,during transportation and storage).US manufacturers of lipid injectable emulsion explicitly state the required storage temperature for the products.They differ,however,in that one stores the product in a conventional glass bottle,while the other stores the lipid injectable emulsion in a newly introduced plastic bag.In addition,they also differ in their storage requirements.For example,in the ca of the plastic product,it is to be stored above25-C’’行政管理专业就业前景
(10),while with the glass product,it is recommended to be
The most common lipid injectable emulsion ud clinically is a soybean oil-in-water formulation.The soybean oil droplets are stabilized by an egg phospholipid emulsifier that coats the submicron droplets.The hydrophobic tails of the phospholipids,containing long-chain fatty acids such as the18-carbon oleic acid,align at the oil droplet surface of the internal pha,while the hydrophilic heads,containing phosphatidic acid,projects out into the external water pha. At or near physiologic pH,the polar phosphate head groups are ionized which establishes electrostatic repulsion between neighboring charged droplets,thereby conferring stability to the formulation.The pharmacopeial limits specifically out-line the physical and chemical attributes of the dosage form that should be maintained throughout its shelf-life.They mainly include the pH,free fatty acids concentration,and globule size limits,and the are listed in Table I.
Table I.Lipid Injectable Emulsions10,20or30%w/v Physicochemical Attribute Pharmacopeial Limits pH Between6.0and9.0 Mean droplet size e500nm
PFAT5e0.05%
Free fatty acid e0.07mEq/g
1960Driscoll
pH Limits
The propod pharmacopeial pH range for lipid injectable emulsions is between6.0and9.0,whether its intended u is for nutrition support or as a drug delivery vehicle,and this range should be maintained throughout its shelf-life.During the normal shelf life of lipid injectable ,between18 and24months),the initial pH is nearer9.0,whereas at the end of its shelf life,it approaches6.0.This is to be expected as a result of the hydrolytic degradation of the long-chain trigylcer-ide to its constituent free fatty acids over time.As the pH drops, the stabilizing anionic electrostatic charge conferred to the droplets by the egg phospholipids moieties is reduced,as the latter become less ionized,thereby reducing the stability of the emulsion.In the extreme ca,the effect of an increasingly acidic pH will eventually neutralize the electrostatic charge residing on the lipid droplets at a pH of3.2(12).Even with emingly modest reductions in the propod pharmacopeial pH range of a lipid injectable emulsion formulation,the detrimental effects on stability can be en.For example,a generic1%propofol formulation in10%lipid injectable emulsion that is prerved with sodium metabisulfite,thus requiring a unique final acidic pH range(4.5Y6.6),has been shown to be unstable during its manufacturer-assigned shelf-life,as evidenced by a growing(coalescing)large-diameter tail, measured by light obscuration or extinction employing a singl
e-particle optical nsing(LE/SPOS)technique(13).In other cas,when the emulsion is extemporaneously com-pounded into TNAs for adults producing admixtures with a pH of approximately5.7,the short-term exposure to the acidic conditions does not appear to adverly affect emulsion stability over the period of clinical ,up to30h at room temperature)(14).For physical mixtures of lipid injectable emulsions containing medium-chain triglycerides(MCTs) along with long-chain triglycerides(LCT)provided as soybean oil,physical stability has been maintained for up to48h in neonatal/infant TNA admixtures with final pH values as low as 5.0(15).The stabilizing influence of MCTs when combined with LCTs as physical mixtures have been shown to produce more stable TNAs than tho prepared from LCTs alone(16). Free Fatty Acid Limits
Free fatty acid concentrations in lipid injectable emul-sions reflect the relative stability of the phospholipid emulsifier due both to its hydrolysis during heat sterilization, as well as from the breakdown of the triglyceride over the normal time cour of the shelf-life for the formulations (17,18),i.e.,18Y24months.The principal long-chain fatty acids found in soybean oil include(in decreasing order) linoleic,oleic,palmitic,linolenic and stearic acids.From a quantitative standpoint over time,the10,20or30%w/v triglyceride concentrations found in lipid injectable emul-sions are major contributors to the formation of free fatty acids over time compared to the amounts in the egg
phospholipids ,0.74Y1.8%w/v)V hence,the pharmacopeial limit of e0.07mEq/g largely pertains to changes in the stability of the triglyceride.Moreover,the free fatty acid limit is also bad on some concerns of systemic toxicity.In dogs,parenteral administration of free fatty acids from hydrolyzed lecithin produced blood and liver abnormalities(19),while in rabbits,intravenous infusion of free fatty acids caud pulmonary edema and ventilatory defects(20).In fact,an oleic acid-induced lung injury model has been ud in animals to evaluate potential treatments of lung injury in the critically ill(21).In humans,displacement of bilirubin by free fatty acids from rum albumin occurs (22),and may be a factor in the development of kernicterus in premature infants.In some cas,the do of lipid injectable emulsions may need to be reduced by one half of the usual lipid do until the bilirubin level is lowered(23). Thus,limiting the amount of free fatty acids in commercial lipid injectable emulsions is primarily intended as a measure to improve the stability of the main emulsion components,as well as to minimize exposure to the hydrolysis byproducts upon intravenous administration and reduce subquent potentially adver clinical conquences.
Globule Size Limits
In any emulsion,determination of stability is demon-strated by maintenance of the globule size distribution (GSD)within defined ,no growth in the extreme population of large-diameter fat g
lobules).Alterations in the GSD reveal a change in the stability of the lipid injectable emulsion.The destruction of an emulsion system is man-ifested by the fusion of droplets that ultimately parate from the disperd pha as enlarged fat globules via a process known as coalescence.From a clinical perspective,globule size limits for lipid injectable emulsions are most important, as they ultimately reflect the safety of the formulation.
The USP propos globule size limits using two methods, applying two criteria to measure the mean droplet size (MDS),and the large-diameter tail of the GSD to verify the stability of lipid injectable emulsions.The first of the measures,the intensity-weighted MDS,expresd in nm,is an important qualitative measure of the homogenization process,and the USP specifies that the MDS cannot exceed 500nm,irrespective of the final concentration of the disperd lipid ,10,20or30%w/v).The technique for determining MDS,as per Method I of USP<729>,can be accomplished using either dynamic light scattering or Classi-cal Mie or F static_light scattering.Clearly,for lipid injectable emulsions that meet USP<729>limits,the greatest mass of the disperd lipid droplets resides below12m,but assuming a normal or Gaussian droplet size distribution for stable lipid injectable emulsions,there will be droplet extremes on both the left-(sub-micron)and right-side(up to and above12m) tails of the distribution.Assuming a normal distribution of
droplets,Fig.1illustrates the relative distribution of droplet sizes in stable vs unstable lipid injectable emulsions.For example,if a stable lipid injectable emulsion has an MDS of 300T60nm(2T A),the proportion of droplet sizes of375nm (x0)or larger can be estimated by calculating its z-value,and then identifying the corresponding area under the(Normal Distribution)curve(AUC)from standard statistical tables.In the above example,the calculated z-value is:
z¼
x0À"
'
complement¼
375À300
60
¼1:25
A z-value of1.25corresponds to AUC=0.7887,with one half the remaining area of the curve lying to the right of
1961
Lipid Injectable Emulsions:Pharmacopeial and Safety Issues
,10.5%of the droplets are above 375nm,or,alterna-tively,89.5%of droplets are below 375nm.At 2A (i.e.,x 000=420nm)above the mean diameter,with an AUC =0.9545,2.28%of the droplets are above 420nm.Finally,at 3A (i.e.,x 000=480nm)above the mean diameter with an AUC =0.9973,0.135%of MDS values are above 480nm.Obviously,there are dimin-ishing droplet numbers at either extreme of the GSD of stable lipid injectable emulsions.In particular,the population of large-diameter fat globules larger than 12m is vanishingly small,but nonetheless,they are certainly prent.In 1980,the British Pharmacopeia attempted to t limits for lipid injectable emulsions and stated no fat globules could exceed 52m in size (24).This requirement,of cour,was subquently dropped,as it was bad on flawed technological capabilities at that time,rather than actual measurements of statistically irrelevant,but potentially clinically significant,large-diameter fat globules.Method I of the USP <729>is not capable of discerning the large-diameter population of the statistical F outliers _,despite their pathophysiological significance.
Importantly,the u of light obscuration or extinction,employing a single-particle optical nsing (LE/SPOS)tech-nique as described in Method II of <729>,has been shown to reproducibly measure this extreme globule outlier population in a ries of commercially available,stable lipid injectable emulsions.(25)Moreover,this study also showed that the LE/SPOS technique could be validated in terms of sizing and counting accuracy using certified (traceable to the National Institute of Standards Technology)calibrator polymer micro-spheres.By comparison,lar diffraction showed nonlinear respons to the same lipid admixtures and varying concen-trations of calibrator microspheres as shown in Fig.2.(26)A summary of the findings and application of various sizing methods for lipid injectable emulsions were recently reviewed.(27)Method II of <729>specifies that the volume-weighted percentage of fat >52m or PFAT 5cannot exceed 0.05%of the total disperd pha,irrespective of the final lipid concentration.Ideally,such a limit should be extended to apply to extemporaneously prepared lipid injectable emulsion-containing TPN admixtures during their period of clinical u (28).Of note,the same technique in Method II of <729>has also rved as the reference method for USP <788>entitled B Particulate Matter in Injections’’(29)for more than 20years.
Table II illustrates the ability of LE/SPOS to routinely measure this population in a variety of commer
cially available and stable lipid injectable emulsions of varying oil composition (26).By comparison,Table III depicts the F stable population _of large-diameter fat globules when lipid injectable emulsions are admixed as TNA dosage forms (28).Thus,it is evident that the method is capable of measuring very small amounts of fat globules >52m in stable emulsion formulations.As human capillaries have an internal diameter between 4and 92m,USP <729>also specifies a physiolog-ically relevant dimension of 52m,a size where such globules may begin occlusion of the microvasculature.Destabilization of the lipid injectable emulsion will result in an increasing population of the large-diameter tail of the GSD,through coalescence,recognizing that changes detected in the large-diameter tail (PFAT 5)will have virtually no measurable effect on the MDS (14),until very late in the destabilization process (i.e.,when obvious pha paration occurs with free oil that is easily detectable by the naked eye).Hence,Method II is the stability-indicating measurement indicated in <729>for lipid injectable emulsions.
PHARMACEUTICAL EXAMPLES OF APPLICATION OF SINGLE-PARTICLE COUNTING/SIZING TO LIPID INJECTABLE EMULSIONS
longerthe bird and the beeMost would agree that a stability-indicating method for determining the pharmaceutical integrity of lipid injectable emulsion requires quantification (i.e.,single-particle or globule counting)of the large-di
ameter fat globules that have formed as a result of coalescence.There are basically three ways to accomplish this task,including,microscopy,electrical resistive pore method or electrical zone nsing,and the
optical
Fig. 2.Light obscuration (LO )three different lar diffraction instruments (LD1,LD2,LD3)versus ideal correlation (X =Y )of varying concentrations of 5-2m calibrator spheres in lipid injectable emulsions*.
*Adapted from reference 27
.
Fig. 1.Normal probability curve and relevant droplet/globule populations for lipid injectable emulsions*.*Previously published from reference 26.
1962Driscoll
equivalent of this technique,known as light obscuration or extinction.They are listed above in order of increasing statistical relevance with respect to the typical number of fat globules that can be counted and sized in practice from a given emulsion sample.It should be noted,for example,that in a stable native lipid injectable emulsion containing 300-nm droplets,an estimated 1012globules/ml can be calculated to be prent in the emulsion,knowing the density of oil,the mass or concentration of lipids and the volume of a sphere.As the emulsion is not a monodisper formulation,a range of droplet sizes spanning a wide range of sphere volumes,is prent,and thus the F total _number of droplets per milliliter is not easily calculated nor determined.For clinically important large-diameter fat globules,Tables II and III show the number of globules per ml in stable lipid injectable emulsions (i.e.,PFAT 5<0.05%),and above 52m,there are approximately between 104and 105/ml.Hence,even in this F remote _population of the large-diameter tail of the distribution of droplet sizes,the are
Table II.Physical Characteristics of Commercially Available Lipid Injectable Emulsions
Product
Lot No.Months to ED GN 1.8GN 5GN 10PFAT 1.8PFAT 5PFAT 10MDS Soybean oil only Intralipid 10%12202-5191224718751487740.0240.0090.0010286Intralipid 20%10776-716298365586451350.0170.0050.0008340Intralipid 30%16115-51172017816125046080.0480.0070.0020420Liposyn III 10%45-351-DE 184827977545653120.0220.0130.0040263Liposyn III 20%43-440-DE 126740987382223200.0100.0050.0007307Liposyn III 30%41-395-DE 102184390340158409840.0400.0290.0160301Lipofundin-N 10%8085A831532192338561750.0110.0010.0005272Lipofundin-N 20%8082A841525257206750839780.0160.0050.0020332Soybean oil mixtures Liposyn II 20%47-412-DE 167448694563718930.0090.0040.0010278ClinOleic 20%980137616701530115987850.0040.0010.0005276Structolipid 20%18417-515122249112366147730.0180.0090.0020276Lipoplus 20%
9235A321518167378364259270.0190.0080.0040263Lipofundin MCT 10%8042A81134387574493027310.0140.0080.0030266Lipofundin MCT 20%8075A8115123049011429957080.0160.0090.0030287Lipovenous MCT 20%KK1569205304751
54831090.0040.0010.0005275Critilip 20%
KV1249B
17
9548816
205183
3723
0.051
0.012
0.0020
330
Table adapted and expanded from data in (26)and from (27).
巴拿马文
Months to ED Months to expiration date at time of test,GN globule number per milliliter,PFAT percentage (volume-weighted)of fat determined by LE/SPOS,MDS mean droplet size (intensity-weighted)in nanometers determined by DLS.
Table III.Globule Size Distribution Data (Mean T SD,n =6replicates per time interval)for TNAs Studied
音频小说
Formula (kg)
Time GN 1.8GN 5GN 10PFAT 1.8PFAT 5PFAT 10MDS 40
0249252T 3712539336T 22151951T 2250.042T 0.0040.028T 0.0020.008T 0.001269.8T 0.8
6226761T 3299030756T 1109928T 1300.032T 0.0030.019T 0.0010.004T 0.00124165409T 2183616880T 2149377T 560.019T 0.0020.010T 0.0010.001T 0.00130162701T 2922514477T 1870348T 1150.017T 0.0030.008T 0.0010.001T 0.000269.7T 0.850
0270482T 119345761T 11362194T 2300.047T 0.0020.031T 0.0020.009T 0.001270.3T 2.9
6248338T 334437056T 12991131T 550.037T 0.0010.022T 0.0010.005T 0.00124178941T 42192066
0T 2390487T 880.022T 0.0010.011T 0.0010.002T 0.00030178152T 281617734T 2702336T 500.020T 0.0020.010T 0.0010.002T 0.001269.8T 2.860
0316026T 9538344800T 32682132T 1520.048T 0.0070.030T 0.0020.009T 0.001274.4T 2.2
6278232T 8298529205T 1199624T 530.032T 0.0040.016T 0.0010.002T 0.00124220450T 7956117148T 878338T 200.021T 0.0040.009T 0.0010.001T 0.00130210269T 7629214263T 867261T 200.019T 0.0040.007T 0.0010.001T 0.001270.3T 4.770
celestial
0256674T 349341901T 14081776T 1670.040T 0.0020.026T 0.0010.007T 0.001271.0T 5.2
6235143T 459531939T 1662745T 1370.030T 0.0010.017T 0.0010.003T 0.00124175895T 213217134T 1387326T 270.018T 0.0010.009T 0.0010.001T 0.00030170892T 560314803T 1967270T 650.017T 0.0010.008T 0.0010.001T 0.000271.7T 3.380
0325270T 7685252727T 35563397T 2350.056T 0.0060.039T 0.0020.015T 0.001271.9T 1.6
6300842T 7164439817T 18591245T 960.039T 0.0050.023T 0.0010.005T 0.00124229026T 6291921429T 913485T 410.023T 0.0030.011T 0.0010.002T 0.00030
219017T 56038
18285T 1053
343T 47
0.021T 0.002
0.009T 0.001
0.001T 0.000
270.8T 2.1
Table adapted from (28).
GN Globule number per milliliter greater than 1.8,5or 10um,PFAT percent fat (volume-weighted)greater than 1.8,5or 10um (boldface values vs .USP <729>for PFAT 5<0.05%),MDS mean droplet size in nanometers (boldface values vs .USP <729>for lipid injectable emulsion of <500nm).
1963
Lipid Injectable Emulsions:Pharmacopeial and Safety Issues参与英文
