Journal of Pharmaceutical Sciences 106 (2017) 446-456
`
`Contents lists available at ScienceDirect
`
`Journal of Pharmaceutical Sciences
`
`j o u r n a l h o m e p a g e : w w w . j p h a r m s c i . o r g
`
`Review
`
`If Euhydric and Isotonic Do Not Work, What Are Acceptable pH and
`Osmolality for Parenteral Drug Dosage Forms?
`
`Dieter Roethlisberger*, Hanns-Christian Mahler, Ulrike Altenburger, Astrid Pappenberger
`
`F. Hoffmann-La Roche Ltd., Pharmaceutical Development and Supplies, Pharma Technical Development Biologics EU, Basel, Switzerland
`
`a r t i c l e i n f o
`
`a b s t r a c t
`
`Article history:
`Received 6 July 2016
`Revised 29 September 2016
`Accepted 30 September 2016
`Available online 23 November 2016
`
`Keywords:
`formulation development
`sterile products
`local tolerance
`pH
`osmolarity
`osmolality
`tonicity
`drug product appropriateness
`buffer strength
`titratable acidity
`parenterals
`
`Introduction
`
`Parenteral products should aim toward being isotonic and euhydric (physiological pH). Yet, due to other
`considerations, this goal is often not reasonable or doable. There are no clear allowable ranges related to
`pH and osmolality, and thus, the objective of this review was to provide a better understanding of
`acceptable formulation pH, buffer strength, and osmolality taking into account the administration route
`(i.e., intramuscular, intravenous, subcutaneous) and administration technique (i.e., bolus, push, infusion).
`This evaluation was based on 3 different approaches: conventional, experimental, and parametric. The
`conventional way of defining formulation limits was based on standard pH and osmolality ranges.
`Experimental determination of titratable acidity or in vitro hemolysis testing provided additional drug
`product information. Finally, the parametric approach was based on the calculation of theoretical values
`such as (1) the maximal volume of injection which cannot shift the blood's pH or its molarity out of the
`physiological range and (b) a dilution ratio at the injection site and by verifying that threshold values are
`not exceeded. The combination of all 3 approaches can support the definition of acceptable pH, buffer
`strength, and osmolality of formulations and thus may reduce the risk of failure during preclinical and
`clinical development.
`© 2017 American Pharmacists Association®. Published by Elsevier Inc. All rights reserved.
`
`During drug product development, nonclinical safety studies are
`performed to support clinical trials and marketing authorization for
`pharmaceuticals.1,2 As an added precaution, the first clinical trials
`(“entry into human”) usually start with a relatively low systemic
`exposure in a small number of healthy volunteers. The formulation
`that is being brought forward into nonclinical and clinical testing
`needs to take into account physical and chemical stability, manu-
`facturability, and local and systemic tolerability, in addition to
`regulatory and pharmacopeial requirements. Although physico-
`chemical parameters like pH, osmolality, buffer concentration, their
`process-related acceptance criteria, and impact on stability are part
`of a state of the art drug product formulation development, the
`definition of what is accepted in terms of tolerability in a clinical
`setting is more controversial and not straightforward.
`Solutions for injection of infusion may require a pH or osmolality,
`which are clearly outside the physiological (euhydric and isotonic)
`
`Current address for Mahler: Lonza AG, Drug Product Services, Basel, Switzerland.
`* Correspondence to: Dieter Roethlisberger
`(Telephone: þ41-61-68-83663;
`Fax: þ41-61-68-88689).
`E-mail address: dieter.roethlisberger@roche.com (D. Roethlisberger).
`
`range, often for solubility or stability reasons. In those cases, only a
`sound understanding of physiological, anatomic, physical, and
`chemical mechanisms of the parenteral administration at the in-
`jection site and during infusion will provide enough insight into the
`suitability of a formulation for the nonclinical and clinical studies.
`Especially, small molecule parenteral dosage forms are often char-
`acterized by a pH and osmolality significantly deviating from ideal
`target values. Although biologics in most cases can be developed
`toward isotonicity, the pH values may also deviate from euhydric pH
`because of stability reasons. For example, antibodies are often
`formulated around pH 5.5-6.5, and G-CSF is around pH 3-4. Biologic
`formulations often contain a buffer, and buffering capacity thus also
`needs to be considered in connection with target pH. In cases where
`lyophilisates are reconstituted in less volume, for example, to ach-
`ieve higher concentration antibody formulations for administration,3
`the tonicity may also be hypertonic.
`Parenteral products should aim toward being isotonic and
`euhydric (physiological pH). If this is not achievable, as a general
`rule, excessive values of pH and osmolality should be avoided
`as much as possible to minimize or prevent local damage on
`vascular endothelium and circulating blood cells. However, because
`many parameters play a crucial role in terms of local tolerance
`(e.g., administration site and route of administration, vein selected,
`related venous blood flow, injection volume, infusion time, infusion
`
`http://dx.doi.org/10.1016/j.xphs.2016.09.034
`0022-3549/© 2017 American Pharmacists Association®. Published by Elsevier Inc. All rights reserved.
`
`EXELA 2023
`Nexus Pharmaceuticals v. Exela Pharma Sciences
`PGR2024-00016
`
`1
`
`

`

`D. Roethlisberger et al. / Journal of Pharmaceutical Sciences 106 (2017) 446-456
`
`447
`
`duration, residence time in subcutaneous [s.c.] or intramuscular
`[i.m.] tissue, diffusion into surrounding tissues),4 no well-defined
`and generally recognized pH, buffer strength, and osmolality
`limits are available.
`To evaluate the appropriateness of a formulation with regard to
`pH, buffer, and osmolality related to sufficient systemic and local
`tolerance, 3 different approaches can be considered: conventional,
`experimental, and parametric.
`The conventional approach is knowledge based and consists in
`choosing pH and osmolality within the usual ranges of literature.
`The difficulty of this approach is that those ranges can vary
`considerably depending on the reference chosen as shown in later
`sections. Furthermore, they often do not sufficiently take other
`crucial parameters into consideration such as buffer capacity of the
`formulation, administration site, injection volume, injection dura-
`tion, or frequency of administration.
`A second approach is based on experimental data from either (1)
`titratable acidity measurements or (2) in vitro hemolysis tests. The
`titratable acidity measurements are used to determine the buffer
`strength of a solution at nonphysiological pH values, whereas the
`in vitro hemolysis test characterizes the risk of cell membrane
`damage after contact with a parenteral injection solution. The
`advantage of the experimental approach is that it is product specific
`and, at least for hemolysis testing, includes a biological component
`such as a direct contact with the (red blood) cell membrane.
`Hemolysis testing can be considered as a simple in vitro model for
`local tolerance. However, it provides no information about systemic
`toxicity, for example, interaction with the whole blood volume.
`Finally, the parametric approach is based on the calculation of
`theoretical threshold values such as the maximal volume of injection
`which will still keep the blood's pH or its molarity within the
`physiological range or an estimate of the dilution ratio at the injec-
`tion site. As long as those threshold values are not exceeded, the
`formulation and its administration condition have at least no theo-
`retical concern and may be considered within an experimental
`design for further assessment. The parametric approach provides
`some additional information but does not replace the conventional
`or the experimental approach. In fact, all approaches are comple-
`mentary. Of course, such an evaluation is specific for a given product,
`formulation, and administration scheme, and only nonclinical and
`clinical studies will finally confirm the appropriateness of the drug
`product related to systemic and local tolerance and the assessment
`of safety (vs. efficacy).
`
`Assessment via the Conventional Approach
`
`pH Limits of Small- and Large-Volume Parenterals
`For large-volume intravenous (i.v.) infusion administration, pH
`and osmolality recommendations are summarized in Table 1.5
`For small-volume i.v. injection solutions (<100-mL nominal
`volume), broader pH ranges can be envisaged depending on the
`source of information: pH 4-9,6 3-10.5,7 or 3-11.4,8 On the other
`hand, when the risk of infiltration in subcutaneous tissue cannot be
`
`Table 1
`
`Recommendations of Infusion Nursing Society for Minimization or Prevention of
`Vascular Damage From Extremes in Infusate pH or Osmolarity5
`
`Vessel
`
`Superior vena cava
`Subclavian vein and
`proximal axillary vein
`Cephalic and basilica veins
`in the upper arms
`
`Blood Flow
`(mL/min)
`
`Osmolarity
`(mOsm/L)
`
`Solution pH
`
`2000
`800
`
`>900
`500-900
`
`<5 or >9
`<5 or >9
`
`40-95
`
`<500
`
`5-9
`
`excluded, more restrictive pH ranges such as 5.5-8.59 are suggested,
`to avoid any risk of tissue damage.
`As summarized in Table 2, drug products within a broad range of
`pH values (2.55-11.15) are on the market, most likely to overcome
`solubility or stability constraints.
`There are no major warnings about pain or irritation in the
`packaging inserts of the drug products listed in Table 2, although an
`injection solution with an extreme pH value is more likely to induce
`vascular irritation, inflammatory reactions, or pain. However, the
`physiological local reaction cascade depends on too many factors
`(injection volume, infusion rate, local blood rate, duration of infu-
`sion, needle diameter, injection depth, buffering capacity, viscosity,
`active ingredient, cosolvents, and so forth) to permit a direct cor-
`relation between the pH (as isolated parameter) and pain or local
`irritation, unless all other factors would be kept constant in a ho-
`mologous test series. Therefore, in case of a low buffering capacity,
`a low injection volume, a slow injection rate (favoring an rapid
`dilution by blood close to the injection site), and a nonirritating
`active drug substance even an extreme pH (high or low) can be
`locally well tolerated and produce neither pain nor irritation. In
`contrast, for an injection solution such as promethazine hydro-
`chloride, which has a relatively harmless pH (4.0-5.5), adverse re-
`actions including burning, pain, thrombophlebitis, tissue necrosis,
`and gangrene are mentioned in the packaging insert.
`Those examples show that a too restrictive use of pH acceptance
`criteria could unnecessarily jeopardize the feasibility of a paren-
`teral drug product.
`For small-volume injections, not only the dilution factor but also
`the time factor plays an important role with regard to local i.v.
`tolerance. In contrast to the almost instantaneous solubilization of
`lipidic cell membrane components in presence of a high concen-
`tration of cosolvents, acid or basic hydrolysis needs time. This is the
`reason why small-volume bolus injections can often be adminis-
`tered i.v. in a very broad pH range. This has been confirmed by
`animal studies5,10 that concluded that a solution with a pH of 3-11
`did not induce phlebitic changes when drugs were administered
`over a few minutes. However, the same studies showed that the
`local tolerance was highly dependent on the pH in case of a 6-hour
`infusion through peripheral vessels. Indeed, a solution with a pH of
`4.5 resulted in a 100% incidence of severe phlebitic changes, a pH of
`5.9 caused mild-to-moderate phlebitic changes in 50% of the ani-
`mal subjects, a pH of 6.3 still caused mild damage in 20% of those
`subjects, and a pH of 6.5 caused no significant damage.8,11,12
`To keep the risk of local
`irritation low, pH values should
`nevertheless be inside the target pH range of 3.5  pH  9.0
`(according to DailyMed database) unless for very compelling
`reason. Especially, alkaline solutions with significant buffering
`capacity should be preferably be avoided. In case of borderline pH, a
`slower infusion rate (e.g., 5-minute push instead of 1-minute bolus
`injection) will contribute to overcome or reduce the risk of local
`irritation and vein damage. The reason of a better local tolerance in
`that case is related to the infusion duration but to the increased
`drug product dilution by the blood flow at the injection site.
`
`Osmolality Limits of Blood and Infusion Solutions
`With regard to osmolality, hypertonic injection solutions with
`an osmolality >600 mOsm/kg13 have been reported to possibly
`cause crenation (shriveling up) of red blood cells and significant
`pain. Hypotonic solutions with an osmolality about <150 mOsm/kg
`in contrast may cause hemolysis and pain at the site of injection.
`The limit of 240 mOsm/kg given in the European Pharmacopoeia
`for monoclonal antibodies14 has already a considerable safety
`margin and seems a bit arbitrary acceptance criterion but is not the
`physiologically lowest limit conceivable for these products. Indeed,
`a hypotonic sodium chloride 0.45% infusion solution (154 mOsm/L)
`
`2
`
`

`

`448
`
`Table 2
`
`D. Roethlisberger et al. / Journal of Pharmaceutical Sciences 106 (2017) 446-456
`
`Marketed Drug Products With Extreme Formulation pH
`
`Active Pharmaceutical Ingredient
`
`Trade Name
`
`Company
`
`Administration
`Route
`
`pHa
`
`Doxycycline hyclate
`Dipyridamole
`Midazolam HCl
`Morphine Sulfate
`Minocycline HCl
`Mycophenolate mofetil HCl
`Nicardipine HCl
`Dolasetron mesylate
`Odansentron HCl
`Dopamine HCl
`
`Vancomycin hydrochloride
`Insulin glargine
`Isoproterenol hydrochloride
`Sumatriptan succinate
`Pentobarbital sodium
`Pantoprazole sodium
`Esomeprazole sodium
`Phenytoin sodium
`
`Doxycycline
`Dipyridamole injection USP
`Midazolam
`Morphine Sulfate
`Minocin
`CellCept
`Nicardipine Hydrochloride
`Anzemet
`Odansentron
`Dopamine Hydrochloride
`And Dextrose
`Vancomycin
`Lantus
`Isuprel
`Sumatriptan succinate
`Pentobarbital sodium
`Pantoprazole Sodium
`Esomeprazole Sodium injection
`Phenytoin Sodium
`
`Mylan Institutional LLC
`Baxter Healthcare Corporation
`West-ward Pharmaceutical Corp.
`Hospira, Inc.
`Rempex Pharmaceuticals, Inc
`Roche
`Emcure Pharmaceuticals Ltd.
`Sanofi Aventis
`Remedyrepack.
`Hospira, Inc.
`
`i.v.
`i.v.
`i.v., i.m.
`i.v.
`i.v.
`i.v.
`i.v.
`i.v.
`i.v.
`i.v.
`
`Pfizer Laboratories
`Sanofi-Aventis
`Hospira
`Teva Parenteral Medicines, Inc
`West-ward Pharmaceutical Corp.
`Akorn, Inc.
`Sun Pharmaceutical Industries Limited
`West-Ward Pharmaceutical Corp.
`
`i.v.
`s.c.
`i.v., i.m., s.c.
`s.c.
`i.v., i.m.
`i.v.
`i.v.
`i.v.
`
`2.55 ± 0.75 (1.8-3.3)
`2.7 ± 0.5 (2.2-3.2)
`3.1 ± 0.6 (2.5-3.7)
`3.25 ± 0.75 (2.5-4.0)
`3.25 ± 0.75 (2.5-4)
`3.25 ± 0.85 (2.4-4.1)
`~3.5
`3.5 ± 0.3 (3.2-3.8)
`3.65 ± 0.35 (3.3-4.0)
`3.5 ± 1.0 (2.5-4.5)
`
`3.5 ± 1.0 (2.5-4.5)
`~4
`4.0 ± 0.5 (3.5-4.5)
`4.75 ± 0.55 (4.2-5.3)
`9.7 ± 0.5 (9.2-10.2)
`9.75 ± 0.75 (9.0-10.5)
`10.0 ± 1.0 (9-11)
`11.15 ± 1.15 (10.0-12.3)
`
`a pH values taken from the online drug product database “DailyMed,” the official provider of FDA label information (package inserts), http://dailymed.nlm.nih.gov/
`dailymed/about.cfm.
`
`could for instance be administered i.v. in case of dehydration.
`Dextrose solution is another hypotonic solution for infusion.
`Because it is a permeant solute (in nondiabetic patients), it will
`rapidly enter cells. This explains why a 5% dextrose is hypotonic
`despite being isosmolar. As a consequence of local hypotonicity and
`increased blood concentration, red blood cells can aggregate and
`form so-called erythrocyte rouleaux.15
`To complete those recommendations regarding osmolality, a
`close look at already marketed formulations and drug products is
`useful. As summarized in Table 3, osmolality values (623-2018
`mOsm/kg) mentioned in the package inserts of some drug products
`are significantly outside of the “expected ranges.”
`Osmolality is an estimation of the osmolar concentration of
`plasma and is proportional to the number of particles per kilogram
`of solvent; it is expressed as mOsmol/kg. (The SI unit is mmol/kg,
`but mOsmol/kg is still widely used.) According to the USP
`(<785>16), the osmolality of blood ranges between 285 and 310
`mOsmol/kg. The average physiological osmolality is therefore
`about 297.5 mOsmol/kg.
`Infusion solutions with an osmolarity >80017 or >850
`mOsmol/L18 should be infused by the central venous route. This is
`a common practice in parenteral nutrition. In general, infusion
`solutions with an osmolality <600 mOsm/kg and a pH close
`to the physiological value have been suggested to have a low-
`to-moderate risk of phlebitis.5,19
`Osmolality limits have nevertheless to be interpreted with
`caution because not only the selection of the vein and the volume of
`infusion but also the infusion duration will play a role. Thus, the
`tolerance osmolality of peripheral venous endothelial cells has
`been reported to be 820 mOsm/kg for 8 hours, 690 mOsm/kg for
`12 hours, and 550 mOsm/kg for 24 hours, with poor blood flow.20
`
`Injection Volume of Small-Volume Injections (Bolus)
`As long as a parenteral injection has not to be administered
`over a prolonged time as infusion, the quotient of the target dose
`(D) divided by the maximal injection volume (Vinj) corresponds to
`the minimal solubility (Smin) needed to get the drug substance in
`solution.
`The usual injection volume ranges for dosage forms applied as
`bolus injection to adults, depending on the route of administration,
`
`are summarized in Table 4, completed with some pediatric
`examples.
`
`Smin ¼
`
`D
`Vinj
`
`; ½mg=mLŠ ¼
`
`½mgŠ
`½mLŠ
`
`The volume administered by subcutaneous bolus injection is
`typically not exceeding 2 mL in products currently on the market.
`However, the administration of up to 3 mL and more has been re-
`ported23 and has been evaluated to not be increasing injection
`pain.24 Any upper limit of s.c. bolus injection volume is likely
`limited by tissue backpressure25 and injection site induration and
`especially leakage.
`This upper bolus injection volume limit can even be significantly
`extended in case of coadministration or prior administration of a
`recombinant hyaluronidase. This enzyme can temporarily hydro-
`lyze the connective tissue in the subcutaneous space and thus de-
`creases tissue backpressure and ensure spreading of the injected
`volume. This increases the practical s.c. injection volume to tens,
`and possibly hundreds, of mL.21,22
`The maximal injection volume for the different administration
`routes is however not an absolute physiological limit and has to be
`adjusted drug product specifically depending on the local tolerance
`and pain measured in nonclinical and clinical studies.
`Given the maximal injection volumes (Vinj) mentioned previ-
`ously, the minimal solubility of a drug (Smin) needed for s.c.
`administration is the quotient of the target dose (D) divided by Vinj.
`
`Local Tolerance
`The probability of local tissue damage on injection of a drug
`product, although critical to the clinical success of parenteral drug
`products, is relatively thorny to assess being dependent on mani-
`fold parameters such as injection volume, infusion rate, choice of
`the injection site, injection duration, injection technique, and fre-
`quency of administration, as shown in the literature.26-34 Also, the
`active ingredient itself can have an impact on local tolerance and
`injection pain, which is assessed in preclinical and clinical studies.
`The administration route plays an essential role in local tolera-
`bility of injections especially due to the dilution effect by blood flow
`which is missing in s.c. and i.m. administration. Tissue damage,
`inflammation, and pain at the injection site are mainly related to
`
`3
`
`

`

`D. Roethlisberger et al. / Journal of Pharmaceutical Sciences 106 (2017) 446-456
`
`449
`
`Table 3
`
`Marketed Drug Products With Significant Hyperosmolarity
`
`API
`
`Trade Name
`
`Company
`
`Administration
`Route
`
`Osmolality
`
`Reference
`
`Amino Acids
`Dextrose 5% monohydrate
`and sodium chloride 0.9%
`N(2)-L-Alanyl-L-Glutamin
`Aminoven 10%
`Mannitol 20%
`
`Amino acids
`Amino acids
`
`Aminosyn-HBC 7%
`Dextrose 5% and
`sodium chloride 0.9%
`Dipeptiven
`Aminoven 10%
`20% OSMITROL
`
`Aminoven 15%
`Aminomix N 1
`
`Hospira, Inc.
`Baxter Healthcare
`Corporation
`Fresenius Kabi
`Fresenius Kabi
`Baxter Healthcare
`Corporation
`Fresenius Kabi
`Fresenius Kabi
`
`i.v.
`i.v.
`
`i.v.
`i.v.
`i.v.
`
`i.v.
`i.v.
`
`623 mOsmol/L
`585 mOsmol/L
`
`921 mOsmol/L
`990 mOsmol/L
`1098 mOsmol/L
`
`1505 mOsmol/L
`1826e2018 mosm/kg
`
`a
`
`c
`
`b
`
`b
`
`a
`
`b
`
`b
`
`a Drug Product Database “DailyMed” (USA), http://dailymed.nlm.nih.gov/dailymed/about.cfm.
`b Drug Product Database “Arzneimittelkompendium der Schweiz” (CH), http://compendium.ch/home/de?nocookie¼true.
`c Drug Product Database “Electronic Medicines Compendium” (UK), http://www.medicines.org.uk/EMC/AdvancedSearchPage.aspx.
`
`physical (osmotic pressure) and chemical stress (pH, surfactant
`concentration, solvent content, increased sodium and potassium
`concentration) and their interaction with vascular endothelial cells,
`circulating blood cells, adipocytes, skeletal muscular cells, and
`nerve cells, depending on the route of administration.
`In a vein, this can result in a vein inflammation (phlebitis),
`initiate a coagulation cascade, form a repairing thrombus, and
`evolve in a local inflammation (thrombophlebitis), as well as a
`systemic inflammatory reaction with significantly increased levels
`of white blood cells and plasma C-reactive protein.35
`From the biochemical point of view, an acidic or basic injection
`solution will be harmful for the cell membranes mainly by their
`cytotoxic and cytolytic properties, by protonation or deprotonation
`and destabilization of the extracellular portion of membrane pro-
`teins, and by hydrolysis of both phospholipids and proteins. The
`phospholipid ester bonds connected to free fatty acids and situated
`directly beneath the phosphate group are sufficiently accessible to
`the injection solution to be hydrolyzed. The damage or dissolution
`of cell membranes can result in the release of cellular components
`into plasma,36 pain, vascular irritation, necrosis, and phlebitis.37,38
`In case of injection of alkaline solutions, an additional chemical
`injury is possible: cell membrane fatty acids are saponified causing
`cell disruption and cellular death.39
`Arachidonic acid can for instance be released from the cell
`membranes and be transformed into powerful cellular mediators
`such as thromboxanes, prostaglandins and leukotrienes, hydroxy
`fatty acids, and lipoxins. These compounds beside a broad range of
`activities will initiate a pain signal in nerve cells.40 Some main pain-
`inducing mechanisms are summarized here41:
`
` Damaged tissue releases globulin and protein kinases, which are
`believed to be among the most active pain-producing sub-
`stances. Minute s.c. injections of globulin induce severe pain.
` Tissue damage stimulates the mast cells to release histamine to
`the surrounding area. Histamine excites the nociceptors. Minute
`s.c. injections of histamine elicit pain.
` Most tissue damage results in an increase in extracellular
`potassium ions. There is a good correlation between pain
`intensity and local potassium ion concentration.
`
`Table 4
`
`Typical Injection Volumes of Bolus Injections
`
`Administration
`Route
`
`Injection
`Volume
`
`5 mL
`
`3 mL
`(>>5 mL with
`Halozyme,
`e.g., 11.7 mL for
`Mabthera s.c.)21,22
`
`0.5 mL
`
`5 mL
`(thigh)
`2 mL
`(deltoid)
`
`Subcutaneous
`
`Intravenous Intramuscular Intradermal
`
`i.v. Infusion
`
` After drug-induced intramuscular tissue damage (rhabdomyol-
`ysis), the injured muscle cells leak myoglobin.42
` It has been reported that intramuscular pain is increased
`by temporal
`(repeated injections) and spatial summation
`(injections given at different sites).43
`
`All i.v., i.m., and the s.c. administration routes have of course also
`specific characteristics, among which the most important will be
`recalled in the following sections.
`
`Intravenous Administration
`The i.v. administration has the advantage that, in case of emer-
`gency, there is no delay before the drug reaches the circulation and
`a bioavailability of 100% is defined for i.v. administration.44 An i.v.
`injection of a drug product with some anaphylactoid potential will
`in contrast bear a higher risk of severe response than when using
`other administration routes. This risk can be decreased by reduc-
`tion of the infusion rate or of the dose of excipients with anaphy-
`lactoid potential. The local blood flow dilution effect is still a major
`advantage of the i.v. administration and explains why too irritating
`drugs by the i.m. or s.c. route can be given intravenously.26
`A direct i.v. injection (bolus) is in general given over shorter
`periods (e.g., 2-3 minutes), observing the patient and the injection
`site for signs of adverse reaction. The volume of injection is usually
`5 mL. In case of an i.v. infusion, the volume used to dilute ranges
`in general from 50 to 500 mL. In clinical practice, most drugs are
`given in 100 mL and are set to infuse over some more extended
`period, for example, 20-30 minutes (Table 5).
`In the example mentioned previously, the infusion rate limita-
`tion is based on a volume/time base independently of the formu-
`lation and the excipients. The importance of the volumetric
`
`Table 5
`
`Intravenous Injection Techniques
`
`Injection Technique
`
`General Infusion Rates and Points to Consider
`
`i.v. Bolus injection
`
`i.v. Push medication
`
`Infusion rate ¼ 1 mL/10 s or 6 mL/min, that is, 5 mL/50 s,
`duration: 1 min; max. volume about 5 mL;
`No cannula needed
`Low infusion rate (0.6-1 mL/min), over 5-8 minutes;
`max. volume about 5 mL; to avoid pain or injury at
`the injection site due to the needle, the use of a
`catheter and a short tubing are necessarya
`Prolonged infusion time, 15-60 min (or more); infusion
`rate: 30 mL/10 min or, e.g., 180 mL/1 h corresponding
`to 1 mL/20 s (3 mL/min); the infusion rate is
`controlled either by a bag with a drop chamber or by
`an infusion pump or syringe pumpa
`
`a The solution being in contact with an infusion set (tubing, connectors, filter, and
`cannula), the absence of leachables or absorption during application has to be
`verified by means of a standardized compatibility test.
`
`4
`
`

`

`450
`
`D. Roethlisberger et al. / Journal of Pharmaceutical Sciences 106 (2017) 446-456
`
`infusion rate will be discussed more in detail in the parametric
`chapter under “local dilution ratio.”
`There are however special cases where an excipient-specific
`limitation should be considered as well. This is the case with
`polymeric surfactants such as Cremophor EL45-47 or polysorbates48,49
`which are prone to induce pseudoallergic reactions. Such hypersen-
`sitivity reactions do not involve IgE immunoglobulins, arise as a
`consequence of activation of the complement system, are (excipient-)
`dose-dependent, and do not need any presensitization. Therefore, to
`ensure a safe administration of parenteral formulations containing
`surfactants, not only the excipient dose and its concentration in view
`of a good local i.v. tolerability but also the excipient-infusion rate
`(mg/min) should be taken into consideration. A good practice con-
`sists in verifying that the intended excipient infusion rate is
`already covered by a marketed drug product for the same adminis-
`tration route.
`For other excipients such as cosolvents or sucrose, a closer look
`at the excipient infusion rate is also advisable. In 1999, the Food and
`Drug Administration (FDA) issued a warning to physicians about
`the safety of intravenous immunoglobulins (IVIG) products. In this
`document, the FDA also provided advice regarding proper patient
`management surrounding the administration of IVIG. For sucrose-
`containing IVIGs, FDA recommended a maximum infusion rate of
`3 mg sucrose/kg/min.50
`
`Intramuscular Administration
`Absorption by the i.m. route is relatively fast. Nevertheless, a
`delayed onset of action is usually associated with that adminis-
`tration route. Furthermore, bioavailability differences between i.m.
`and i.v. administration may be observed. After drug-induced
`intramuscular tissue damage (rhabdomyolysis) beside the mani-
`fold inflammatory mediators,51 the injured muscle cells leak in
`addition myoglobin.42
`To ensure an i.m. administration of a drug, a sufficient needle
`length is essential. Suboptimal hospital practice of i.m. injections can
`indeed result in a majority of unintentional s.c. administration.52
`The i.m. site is more often associated with pain on injection
`compared to i.v. or s.c. administration. This most likely results from
`the prevalence of nerves in muscle tissue compared to subcutaneous
`tissue and the rapid dilution of the drug by blood when administered
`i.v., which may limit the concentration at the injection site.26 Intra-
`muscular injections require a longer needle than s.c. injections to
`reach the desired muscle. In addition, the needle diameter is in
`general larger (22-25 G) as shown in vaccination practice. Although
`it is generally assumed that larger needles mean larger pain sensa-
`tion in patients, the contrast may also be true. Contrary to the belief
`that smaller needles cause less pain, study findings revealed how-
`ever that vaccination with smaller bore needle (25 G) caused more
`pain compared to larger needles (23 G).53 This could be explained by
`a higher jet pressure created through thinner needles in comparison
`to larger ones. To reduce this pressure-induced pain and muscle
`trauma, an i.m. injection should be administered slowly (e.g. 1 mL in
`10 or 20 seconds).54 However, in matter of i.m. injection technique
`and vaccination, there is no general consensus. A systematic review
`of randomized controlled trials has for instance shown that rapid
`injection might be less painful.55
`
`Subcutaneous Administration
`In s.c. administration, absorption is slower than by the i.m. route
`but, nevertheless, can be prompt with some drugs. The slower rate
`of absorption by the s.c. route is usually the reason for choosing the
`route. Drugs such as insulin and sodium heparin for instance are
`given by this route because it is desirable to distribute their phar-
`macological effect over several hours, to avoid frequent injections.
`Irritant medications should not be administered s.c. as they may
`
`cause tissue necrosis or a sterile abscess. Not only increased in-
`jection volume but also the buffering strength at a nonphysiological
`pH or the buffer choice56 can be related to pain perception at the
`injection site.26
`A sound understanding and standard ranges of pH, buffer
`strength, osmolality, and injection volume are essential to develop
`parenteral formulations for the different administration routes
`with a high likelihood of good local and systemic tolerance.
`Complementary quantitative parameters can further facilitate the
`selection of the right composition will be discussed in the para-
`metric approach.
`
`Assessment via the Experimental Approach
`
`The term experimental approach is used to describe additional
`assays allowing better prediction of either the likelihood of a sys-
`temic pH shift outside the physiological pH range or the damaging
`potential toward cell membranes rather than the common exper-
`imental physicochemical measurements such as pH and osmolality.
`
`Titratable Acidity Limits for Infusion Solutions
`The difference between the pH of a parenteral injection solution
`and the physiological pH and also the buffer concentration of the
`injection solution can be quantified by means of the measurement
`of the titratable acidity which is defined as the number of mmoles
`of NaOH required to titrate an acidic solution back to a pH of 7.4. For
`injection solutions having a pH >7.4, a titratable basicity can be
`defined in a similar way using an HCl solution as titrant. For slightly
`acidic large-volume infusion solutions which are infused over a
`prolonged duration,
`the titratable acidity is a good marker
`of a potential risk of local inflammation reactions and phlebitis.
`A higher titratable acidity will also increase the risk of irritation of
`endothelial cells over a longer distance from the catheter tip.
`Therefore, it is common practice to mention titration acidity on
`package inserts of large-volume infusion solutions (Table 6).
`The lower the titration acidity, the better. However, the titrat-
`able acidity is not the only factor influencing the time to return to
`pH homeostasis because downregulation of carbon dioxide excre-
`tion in the lung and increased reabsorption of bicarbonate and
`excretion of acids in the kidney will support the regulation of the
`acidebase balance of blood.
`There are 2 important disadvantages with this experimental
`parameter: the titratable acidity is rarely mentioned in parenteral
`drug product information sheets and there are no generally
`recognized threshold values. The parametric approach is therefore
`easier to handle especially during early formulation development.
`
`Table 6
`
`Titratable Acidity of Hyperosmolar Dr