JOURNAL OF HEMATOTHERAPY 1:233-250 (1992)
`Mary Ann Liebert, Inc., Publishers
`
`Hematopoietic Stem Cell Cryopreservation: A Review of
`Current Techniques
`
`SCOTT D. ROWLEY
`
`ABSTRACT
`Hematopoietic stem cells (HSC) can be stored for prolonged periods at cryogenic tempera-
`tures. The techniques currently used were derived from the initial report
`in 1949 of
`cryopreservation of bovine sperm in glycerol. The addition of this penetrating cryoprotectant
`protected the cells from the injury associated with ice formation. Current cryopreservation
`techniques (with minor variations) suspend cells in an aqueous solution of salts, protein, and
`one or more cryoprotectants. Cells are frozen at slow rates and stored generally below
`120°C
`in mechanical freezers or nitrogen refrigerators. That these techniques are successful
`in
`maintaining HSC viability is evident from the engraftment of these cells in patients treated
`with marrow-lethal conditioning regimens. However, issues such as the composition of the
`cryoprotectant solution, cell concentration during freezing, cryoprotectant toxicity, and
`storage temperatures have not been adequately studied, primarily because of a lack of
`appropriate assays for HSC cryosurvival. HSC cryobiology will become an increasingly
`important subject as new HSC collection and processing techniques are developed. Improved
`cryosurvival of HSC using modified cryoprotectant solutions may improve engraftment
`kinetics and decrease the cost and morbidity of autologous transplantation.
`
`—
`
`INTRODUCTION
`
`Cryopreservation of bone marrow or peripheral blood-derived hematopoietic stem cells (HSC) permits
`
`the administration of intensive chemo-radiotherapy to patients with dose-responsive malignancies.
`Although cryopreservation is not a requirement for autologous transplantation, and patients have recovered
`marrow function after reinfusion of marrow cells stored for several days at 4°C, a progressive loss of
`hematopoietic stem cells occurs during nonfrozen storage. Definition of ideal conditions (e.g., temperature,
`initial processing of the cells, additives) for nonfrozen storage of cells may slow this loss. However, merely
`maintaining survival, much less growing additional HSC, is beyond the current capabilities of in vitro cell
`culture techniques. Therefore, most transplant centers cryopreserve HSC intended for autologous transplan-
`tation. Cryopreservation allows administration of multiple-day transplant conditioning regimens as well as
`prophylactic storage for patients to be transplanted months to years later. Although some unavoidable loss of
`
`Fred Hutchinson Cancer Research Center, Seattle, WA 98104.
`
`233
`
`Downloaded by University of Michigan from www.liebertpub.com at 05/27/20. For personal use only.
`
`Novo Nordisk Ex. 2040, P. 1
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`ROWLEY
`HSC occurs with marrow or peripheral blood stem cell (PBSC) processing and cryopreservation, progressive
`loss over months of proper storage is not obvious and may not occur.
`That HSC can be successfully cryopreserved is evident from the success ofautologous HSC transplantation
`in regenerating marrow function after marrow-lethal conditioning regimens. Engraftment failure is generally
`not attributed to HSC cryopreservation, although some investigators correlated poor HSC cryopreservation
`with delayed engraftment after transplantation (Gorin, 1986; Rowley et al, 1989). Most cryopreservation
`laboratories use a variation of the technique outlined in Table 1. Cryopreservation research, however, has not
`been a focus of the transplantation teams, and HSC cryopreservation is one of the least understood aspects of
`HSC processing. This lack of understanding of cryobiology hinders the ability to adapt techniques for
`handling unusually small or large collections of cells or other accidental or intentional deviations from
`laboratory protocol. Although easily performed by a number oftransplant centers, HSC cryopreservation and
`reinfusion are not without risk of toxicity to both the stem cell inoculum and the recipient.
`It is the purpose of this commentary to review the development of our current cryobiology techniques,
`examining the parameters depicted in Table 1. For this review, HSC is intended to refer to both primitive
`hematopoietic stem cells responsible for durable engraftment as well as the more mature progenitor cells that
`can be cultured in vitro, regardless of source (e.g., bone marrow, peripheral blood).
`CRYOPRESERVATION THEORY
`Preservation of viability after cryopreservation and thawing became possible after the discovery that
`glycerol could be used for the freezing of bovine sperm (Polge et al, 1949). Subsequent experiments
`demonstrated that bone marrow frozen with glycerol could be used to reconstitute the bone marrow of
`irradiated mice (Barnes & Loutit, 1955). Considerable exploration of the cryobiology of various mammalian
`and nonmammalian cells ensued. Those studies defined the mechanism of cell damage incurred during
`cryopreservation, and techniques to moderate that damage. The current understanding of cryobiology
`contends that ice crystal formation during cooling is the primary cause of cell damage (Karow & Webb,
`1965a). Intracellular ice crystals may form at rapid rates of cooling, resulting in mechanical rupture of cell
`structures. At slower rates of cooling, ice crystal formation occurs in the extracellular space, resulting in
`increasing osmolality as free water is incorporated into the ice crystals. This loss of free water results in the
`concentration of extracellular solutes, such as sodium ions, that do not freely penetrate the cell membrane (or
`are actively excluded), extreme hyperosmolality, and dehydration injury. For example, the molality of Na+
`10CC is about 2.8 molal, and at -20°C about 5.6 molal (Mazur, 1970).
`in a saline solution at
`
`-
`
`Cell processing:
`Cell concentration:
`Storage container:
`Sample volume:
`Cryoprotectant:
`Protein:
`Electrolytes:
`Rate of cooling:
`Storage temperature:
`Rate of thawing:
`Post-thaw manipulation:
`
`Table 1. Seattle HSC Cryopreservation Technique
`Buffy-coat or light-density cells
`>2 x 107 cells/ml final concentration
`Blood storage bag tolerant of cryogenic temperatures
`Dependent upon bag size, generally 40-60 ml per bag
`10% (vol/vol) dimethylsulfoxide final concentration
`20% autologous plasma
`Tissue culture medium
`l°C/minute to —40°C with compensation for heat of fusion, then
`10°C/minute to
`-80°C
`135°C in nitrogen vapor-phase refrigerators or mechanical freezers
`Below
`Rapidly (>100°C/minute) in a 37°C water bath
`Addition of ACD (20% vol/vol of product volume), then infusion immediately after
`thawing without further processing; filtration through blood administration set
`(170 u.) if clumped after thawing
`
`—
`
`—
`
`234
`
`Downloaded by University of Michigan from www.liebertpub.com at 05/27/20. For personal use only.
`
`Novo Nordisk Ex. 2040, P. 2
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`—
`
`CRYOPRESERVATION OF HEMATOPOIETIC CELLS
`Intracellular ice formation may be limited by cooling cells slowly. At slow rates, it is probable that ice
`nucleation will first occur in the larger volume outside the cell. Progressive dehydration of the cell will result
`if the rate ofcooling is slow enough to allow water to shift to the extracellular space and be incorporated into
`the growing ice crystals. The permeability ofthe cell to water appears to define the optimal cooling rate for any
`particular cell (Mazur, 1966). Dehydration with concentration of intracellular solutes protects the cells from
`intracellular ice formation and growth.
`Glycerol and dimethylsulfoxide (DMSO) are colligative cryoprotectants that prevent dehydration injury by
`moderating the increased concentration ofnonpenetrating extracellular solutes such as sodium ions during ice
`formation, and by decreasing the amount of water absorbed by (in equilibrium with) the ice crystals at a
`defined temperature. Colligative refers to properties dependent upon the number ofparticles (solute), not the
`composition of the particles. Freezing is the crystallization of water, and the freezing point is the temperature
`at which ice crystals can be sustained in equilibrium with water. The freezing point of water is depressed by
`the addition of solute. For any particular mixture of solute(s) and water, there will be a defined temperature at
`which ice crystals can initially form. Unlike pure water, ice crystal formation and growth in aqueous solutions
`occurs over a temperature range. Growing ice crystals absorb free water and exclude solute particles. The
`incorporation of water into ice results in concentration of the solute and further depression of the freezing
`temperature oftheremaining water, thereby preventing additional ice formation unless further cooling occurs.
`Thus, temperature (and pressure) defines the equilibrium between ice and the nonfrozen solution. With further
`cooling, a temperature is eventually reached at which the solute itself crystallizes (eutectic point). The
`molality of a solution in equilibrium with ice is determined by the temperature of the solution (and not the
`20°C has a Na+ molality of
`initial concentration of the solute). In our example above, a saline solution at
`about 5.6 molal. For a two-component system such as DMSO and saline, both solutes will contribute to the
`molality of the unfrozen solution. The molality (before freezing) of 10% DMSO (vol/vol) in solution is about
`1.6 molal, about 10 times the molality ofNa+ in the medium. At -20°C, this molal ratio between DMSO and
`Na+ will be maintained. Therefore, the molality of the Na+ will be about 0.56 molal, or a little less than four
`times the molality ofNa+inasolution without ice. The addition of a penetrating cryoprotectant to an aqueous
`saline solution, therefore, reduces the osmotic stress across the cell membrane at
`20°C from about 36 times
`to less than four times that of an ice-free solution. According to this theory, colligative cryoprotectants must
`be capable ofpenetrating the cell to avoid merely contributing to the molality of the extracellular medium (and
`must be nontoxic to the cells at the concentration required for cryopreservation).
`The tolerance of cells to freezing, therefore, depends on the ability of the cells to withstand osmotic stress.
`The osmotic tolerance of granulocytes is much less than that for lymphocytes or mature hematopoietic
`progenitors, probably accounting at least in part for the difference in survival after cryopreservation (Dooley
`& Takahashi, 1981 ; Law et al, 1983). With DMSO concentrations less than 10%, the degree of dehydration
`caused by concentration of the nonpenetrating solutes during freezing will be greater because nonpenetrating
`solutes will contribute proportionately more to the molality of the nonfrozen solution in the extracellular
`medium. With greater concentrations of DMSO,the osmotic stress will be less. The optimal concentration for
`acolligative cryoprotectant, therefore, depends upon the osmotic tolerance of the cell to be frozen, the toxicity
`to the cell of high concentrations of the cryoprotectant itself (Karow & Webb, 1965b; Fahy, 1986; Arakawa
`et al, 1990), and the presence of other cryoprotectants.
`Colligative properties do not explain the cryopreservation achieved by freezing cells in solutions of
`macromolecules such as hydroxyethyl starch (HES). Solutions of high-molecular-weight, polymeric
`cryoprotectants contain relatively few particles and, moreover, do not freely penetrate the cell. These
`cryoprotectants may protect the cell by forming a viscous, glassy shell that retards the movement of water,
`thereby preventing progressive dehydration as water is incorporated into the extracellular ice crystals
`(Takahashi etal, 1988). Solutions of some compounds, when present in sufficiently high concentrations, will
`solidify to an amorphous glass without first forming ice crystals, a process termed "vitrification" (Fahy et al,
`1984). The "glass-transition temperature" (Tg) depends upon both the structure and concentration of the
`solute. Pure water forms a glass at about
`139°C (Grout, 1991). The addition of cryoprotectants such as
`DMSO, glycerol, or HES raises the Tg (Luyet, 1960). At very high concentrations ofcryoprotectants, (6.3 M
`for DMSO [Fahy et al, 1984]), the T is higher (warmer) than the temperature at which ice crystals can form,
`thereby preventing crystallization during cooling and its resulting mechanical and osmotic stresses. The
`235
`
`—
`
`—
`
`Downloaded by University of Michigan from www.liebertpub.com at 05/27/20. For personal use only.
`
`Novo Nordisk Ex. 2040, P. 3
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`ROWLEY
`practical difficulty with achieving vitrification arises from the necessity to use very high concentrations of
`cryoprotectants. Vitrification has been used for the cryopreservation of murine embryos, for example, using
`a solution of 20% DMSO, 15.5% acetamide, 10% propylene glycol, and 6% polyethylene glycol (Rail &
`Fahy, 1985). Prolonged exposure to this solution was toxic, however, with complete loss ofviability after 30
`min at 4°C. With lower concentrations of cryoprotectants, the crystallization temperature is higher than the
`vitrification temperature and ice crystals will form during cooling. Yet, glass transition may occur even in the
`presence of ice crystallization. With the formation of ice crystals and resulting loss' in free water,
`the
`concentration of the cryoprotectant in the nonfrozen water will increase. With decreasing temperature causing
`increasing cryoprotectant concentration, a point will be reached at which the cryoprotectant forms a glass. At
`this "limiting glass transition temperature" (T'g), the solution suddenly becomes viscous, retarding, if not
`stopping, the flow of water through the extracellular matrix. This has been proposed as the mechanism for
`In one study of
`cryoprotection afforded by the extracellular cryoprotectants (Takahashi et al, 1988).
`peripheral blood monocyte cryopreservation using extracellular cryoprotectants, a limiting glass-transition
`20°C was optimal, and this was achieved using a 20% solution of HES (Takahashi et al,
`temperature of
`1988). Substances forming glasses at higher or lower temperatures were less effective. This model of
`cryopreservation requires adequate dehydration to occur to concentrate intracellular solutes and decrease the
`probability of intracellular ice formation, but glass formation at an appropriate temperature (please recall that
`temperature defines the osmolality of the unfrozen aqueous solution) to prevent excessive dehydration of the
`cell.
`Cells are not frozen in simple, two-component solutions, but rather in complex solutions containing salts,
`sugars, penetrating cryoprotectants with or without extracellular cryoprotectants, and plasma proteins (Table
`2). Phase transition temperatures (such as T'g) for these solutions have not been determined. The improved
`cryosurvival observed with increasing protein concentration (Ragab et al, 1977) and combined penetrating
`and extracellular cryoprotectants (Stiff et al, 1983; Conscience & Fischer, 1985) may be explained at least in
`part by these theories on ice and glass formation and their effects on the cell. Moreover, the existence ofglass
`formation may explain the relationship between storage temperature and survival of HSC over time. Below
`the T'g temperature, ice crystal growth cannot occur and the cells are protected from mechanical disruption by
`growing ice crystals. The optimal storage temperature is below the T'g for the cryoprotectant solution used.
`
`—
`
`Other
`Heparin
`
`Table 2. Published Cryopreservation Techniques
`Final concentration
`End-freeze
`Cell
`Pre-chill
`DMSO
`Protein
`(per ml)
`(%)
`(%)
`Author
`cells
`Medium
`temperature
`0.2-1 x 108
`-120°C
`Weiner & Gross
`25, FFP
`HBSS
`Yes
`10
`4 x 107
`10, AP
`-90°C
`Soken
`Yes
`TCI 99
`10
`-90°C
`5, PPF
`TCI 99
`Janssen & Lee
`Yes
`NS
`10
`<108
`Meagher, Herzig & Herzig
`-90°C
`5, AP
`RPMI-1640
`Yes
`10
`<108
`Hill, Robertson & Dickson
`-100°C
`0.5, HSA
`MEM
`Yes
`10
`-90°C
`NS, AP
`TCI 99
`Hollinsworth
`Yes
`10
`NS
`2-4 x 107
`Dyson, Haylock & To
`-80°C
`20, AP
`HBSS
`Yes
`10
`NS, AP
`Heparin
`Yao
`TCI 99
`NS
`NS
`Yes
`10
`4 x 107
`Bouzgarou et al.
`140°C
`NS, AP
`HBSS
`Yes
`10
`Normosol-R*
`Oldenburg & Stiff
`3.6, HSA
`NA
`6% HES
`NS
`Yes
`Plasma-Lyte"
`<108
`Rosina & Jiang
`5, HSA
`NA
`6% HES
`Yes
`All techniques were recently published in a manual of marrow processing techniques (Areman, Deeg & Sacher, 1992).
`Only technical details without supporting clinical data documenting cryopreservation efficacy were provided. Abbrevi-
`ations used are: FFP = fresh frozen plasma, AP = autologous plasma, PPF = plasma protein fraction, HSA = human
`serum albumin, NS = not specified, NA = not applicable.
`
`-
`
`236
`
`55
`
`Downloaded by University of Michigan from www.liebertpub.com at 05/27/20. For personal use only.
`
`Novo Nordisk Ex. 2040, P. 4
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`CRYOPRESERVATION OF HEMATOPOIETIC CELLS
`Warmer temperatures may be used, but at risk of cell damage, a risk that is dependent upon the temperature
`of storage and the stability (viscosity) of the solution at that temperature.
`This simplified review of the effects of freezing and the cryoprotectant properties of penetrating and
`polymeric cryoprotectants does not completely explain the processes involved during freezing, and more
`detailed reviews of the freezing of aqueous solutions and mammalian cells have been published (Karow &
`Pegg, 1981; Fuller& Grout, 1991). In addition to mechanical and dehydration injuries, cooling itself may be
`to explain the
`damaging to the cell (Fuller, 1991). Also, colligative effects alone are not sufficient
`cryoprotectant properties of DMSO or glycerol. Other freely penetrating chemicals such as urea and DMS02
`do not function as cryoprotectants for mammalian cells (Vos & Kaalen, 1965; McGann & Walterson, 1987),
`and some chemicals such as ethanol and guanidine may actually function as cryosensitizers (Kruuv et al,
`1990). Obviously, the chemical structure of the cryoprotectant is important in the cryosurvival of mammalian
`cells, and a molecular interaction between the cryoprotectant and protein or lipid molecules appears necessary
`foroptimal cryopreservation (Carpenter & Crowe, 1988; Crowe et al, 1987, 1990; Anchordoguy et al, 1991).
`The damaging effect of freezing on membrane proteins or structure compared to the survival of the cell is
`illustrated by the ability of HSC to proliferate in vitro when stimulated by peripheral blood feeder layers but
`lack of response to GCT-conditioned medium (Gilmore, 1983). Similar "resistance" to recombinant growth
`factors by thawed bone marrow progenitors, suggests a loss of cytokine receptors despite the survival of the
`cell (Rowley & Hattenburg, 1990). The ideal cryoprotectant solution would also protect the various cell
`organdíes and molecular species from sublethal damage.
`
`PROCESSING OF BONE MARROW FOR CRYOPRESERVATION
`important concepts of clinical HSC cryopreservation that must be realized is the
`One of the most
`heterogeneity of marrow and PBSC populations. HSC cryopreservation involves not only the cryobiology of
`the HSC, but also of the mature blood cells contained in the harvested product. The minimum number of cells
`required for marrow reconstitution after autologous transplantation is not known. Generally, 1-3 x 108
`nucleated marrow cells per kilogram of patient weight are harvested. Less certain is the minimum amount of
`cells required for PBSC transplantation; under certain conditions, the quantity collected may far exceed the
`number harvested for marrow transplantation. Regardless, hematopoietic stem cells comprise a very small
`portion (< 1 %) ofeither the marrow orperipheral blood product. Bone marrow consists ofHSC, mature blood
`cells, and noncellular material such as fat. Cryopreservation techniques that are optimal for HSC may destroy
`mature blood cells. The standard technique for red cell cryopreservation uses glycerol (Sputteck & Korber,
`1991); granulocytes cannot be cryopreserved successfully. The presence of mature blood cells affects HSC
`cryopreservation in at least three ways. First, the large proportion of mature blood cells collected may hinder
`the laboratory processing ifclumping before or after thawing is induced by damaged granulocytes or platelets.
`We observed the loss of about 50% of cells when clumping occurred, and infusion through a standard blood
`administration set may be required to prevent pulmonary embolism. Second, damaged cells may also
`predispose toward infusion-related toxicity. The infusion of marrow that was frozen without depletion of
`mature blood cells was associated with acute renal failure in 3 of 33 patients in one study (Smith et al, 1987);
`this presumably resulted from massive hemolysis of red cells. Other investigators correlated the quantity of
`cells cryopreserved and infused to the post-infusion toxicity observed. In one study, infusion ofcryopreserved
`buffy-coat cells incurred greater toxicity than infusion of light-density cells, which were essentially depleted
`of mature blood cells (Davis et al, 1990). Kessinger et al reported a high incidence of infusion-related toxicity
`for recipients of large quantities of peripheral blood stem cell products (Kessinger et al, 1990). Third, if the
`cell concentration affects HSC cryosurvival, then the presence oflarge numbers of mature blood cells requires
`that the cells be frozen in large volumes. It is possible that the toxicity reported by Kessinger et al and Davis
`et al may relate to the quantity of DMSO infused, and not specifically from lysis of mature blood cells.
`Therefore, cryopreservation of HSC can be facilitated by the prefreeze depletion of mature blood elements.
`Collection of buffy-coat cells is the minimum processing required for cryopreservation of bone marrow. A
`number of apheresis or cell-washing devices are capable of processing the large quantities of cells harvested.
`237
`
`Downloaded by University of Michigan from www.liebertpub.com at 05/27/20. For personal use only.
`
`Novo Nordisk Ex. 2040, P. 5
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`ROWLEY
`Isolation of buffy-coat cells achieves volume reduction and depletion of red cells. Plasma collected during
`fractionation may be used for subsequent processing or cryopreservation. Some apheresis devices may
`provide enrichment of mononuclear cells relative to granulocytes, providing a "cleaner" preparation for
`processing and cryopreservation. Density-gradient separation of light-density cells further enriches for HSC,
`although at the expense of additional cell losses and cost ofprocessing. At the extreme ofprefreeze processing
`is the extensive purification of HSC by isolation of CD34+ cells (Berenson et al, 1991). This technology may
`be especially advantageous in the cryopreservation of peripheral blood-derived HSC collected from
`hematopoietic growth factor-stimulated donors, from whom large quantities of cells are collected with each
`apheresis.
`
`CRYOPROTECTANT SOLUTIONS
`
`Dimethylsulfoxide
`Glycerol and DMSO are colligative cryoprotectants that protect the cell from excessive dehydration as
`extracellular water is drawn into growing ice crystals. The cryoprotectant properties of glycerol were
`described in 1949 (Polge et al, 1949), and those of DMSO 10 years later (Lovelock & Bishop, 1959). Both
`have been used for the cryopreservation of bone marrow. It is because of the rapid diffusion of DMSO through
`the cell membrane, and the difficulty in removing glycerol before reinfusion (infusion of DMSO-containing
`products without washing is generally tolerated), that DMSO became the favored agent for HSC cryopreser-
`vation.
`Dimethylsulfoxide, a byproduct of paper manufacturing,
`is a hygroscopic polar compound developed
`originally as a solvent for chemicals such as insecticides, fungicides, and herbicides (David, 1972). Pure
`DMSO is a colorless, virtually odorless liquid (sp. gr. 1.108, m.w. 78.13 g/mole), although industrial grades
`may have a strong sulfur odor (Willhite & Katz, 1984). The serum half-life of DMSO is about 20 hr, although
`that of dimethylsulfone (DMS02), a renal-excreted metabolite, 72 hr (Willhite & Katz, 1984). A small
`proportion of DMSO is reduced to dimethylsulfide (DMS), which is expired through the lungs for about 24 hr
`after administration, and which is responsible for the characteristic odor resulting from DMSO infusion. The
`LD50 values (amount of DMSO required to kill 50% of test animals) reported for intravenous infusion of
`DMSO are 3.1-9.2 g/kg for mice, and 2.5 gm/kg for dogs (Willhite & Katz, 1984). The acute toxic dose of
`DMSO for humans has not been determined.
`The optimal concentration of either DMSO or glycerol for the cryoprotection of HSC appears to be about
`10%. In their original report, Lovelock and Bishop demonstrated a dose-response with improving red cell
`survival as the concentration ofDMSO was increased to 15% (Lovelock & Bishop, 1959). The effect ofhigher
`concentrations was not reported. Subsequently, Ragab et al studied the survival of hematopoietic progenitor
`cells (CFC) from human donors after freezing with various concentrations of DMSO (Ragab et al, 1977).
`They found a significant increase in CFC recovery when the concentration ofDMSO was increased from 7.5%
`to 10%. No further improvement was found with an increase to 12.5%. The improvement in CFC recovery
`resulted from a 2-fold increase in nucleated cell recovery from 17.2% to 35.6% and 32.1 %, respectively, after
`washing. Cell recovery before the post-thaw wash did not differ, nor did the numbers of colonies per 105 cells
`plated. The loss ofcells during the wash steps may not reflect direct intravenous infusion, however, and these
`data suggest that HSC tolerate at least a limited range in DMSO concentration.
`High concentrations ofcryoprotectants may incur direct toxicity to the cells being cryopreserved (cf. effects
`In general, cells are more tolerant
`of vitrification solutions on murine embryos discussed above).
`to
`cryoprotectants at reduced temperatures. For example, DMSO is concentrated also during the formation of
`ice. Optimal cryopreservation requires a balance between protection from freeze damage and the occurrence
`of cryoprotectant-induced toxicity. The toxicity of DMSO to the HSC or to the patient has not been fully
`defined. Using in vitro cultures, Douay et al described a 23.5% recovery of CFU-C after 60-min exposure (at
`4°C, without cryopreservation) and a 15% recovery after 120 min (Douay et al, 1982). We were unable to
`confirm their findings of DMSO-induced progenitor cell loss in similar studies (Rowley, 1991a). Recoveries
`of nucleated cells, granulocyte-macrophage colony-forming units (CFU-GM), and erythroid burst-forming
`238
`
`Downloaded by University of Michigan from www.liebertpub.com at 05/27/20. For personal use only.
`
`Novo Nordisk Ex. 2040, P. 6
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`CRYOPRESERVATION OF HEMATOPOIETIC CELLS
`units (BFU-E) were virtually 100% after exposure to 10% DMSO at either 4°C or 37°C. Only at 20%
`concentration of DMSO did cell clumping cause a decrease in cell recovery to 21.7% (and total CFU-GM
`quantity to 27.1%). The numbers ofmyeloid and erythroid progenitors per 5 x 104 cells plated dropped only
`after exposure to 40% DMSO (3.3% of control after 60 min incubation). Similarly, we found no loss of
`progenitor cells after thawing if the removal of DMSO was delayed, even up to 60 min. One possible
`explanation for the differences between these studies is the purity ofDMSO, a strong solvent. Others have also
`reported no toxicity to hematopoietic progenitors with short exposure to DMSO (Ma et al, 1982), although
`direct addition of DMSO to the culture medium at
`1 % or greater concentration was toxic to cell culture
`(Goldman et al, 1978). On the basis of these studies, we no longer precool the freezing chamber or chill the
`cells before addition of the cryoprotectant solution (although that solution is cooled because of the exothermic
`reaction caused by the initial dilution of DMSO in aqueous solutions).
`In contrast, a high incidence ofgenerally mild, infusion-related morbidity with the reinfusion ofeither bone
`marrow or peripheral blood-derived stem cells has been reported by several centers (Davis et al, 1990;
`Kessinger et al, 1990; Stroncek et al, 1991). DMSO itself has a variety of pharmacologie effects (David,
`1972), which may be compounded by the presence of lysed blood cells and foreign proteins or contaminants
`from nonpharmaceutical grades ofreagents used in the processing. Rarely, anaphylaxis or cardiac dysfunction
`may occur. Sudden and severe hypotension can result from the intravenous infusion of DMSO, presumably
`from histamine-induced vasodilatation (David, 1972). Skin flushing, dyspnea, abdominal cramping, nausea,
`and diarrhea, reported to varying degrees after HSC reinfusion, can also all be attributed to DMSO-induced
`histamine release. In a series of 82 patients who were premedicated with diphenhydramine, Davis et al
`observed increased blood pressure and decreased heart rate, which were maximal about
`1 hr after the
`completion of the marrow infusion (Davis et al, 1990). This probably resulted from the cardiac and peripheral
`vascular effects of DMSO. Cardiac dysrythmias were not observed in that study, although one subsequent
`patient developed a transient 2° heart block about 1 hr after the marrow infusion that persisted for about 6 hr
`before spontaneous resolution (unpublished observation). Published reports of cardiac or pulmonary arrest
`during marrow infusion are rare (Vriesendorp et al, 1984; Rapoport etal, 1991). Painful irritation may occur
`if thawed cells are infused through a peripheral vein. Marrows frozen in 10% DMSO have an average
`osmolality of 1794 mOsm/kg H20 after thawing (Rowley, 1992b), and infusion through a central venous
`catheter is preferred. Despite the hyperosmolality of the products, serum osmolality was not greatly affected
`for patients receiving less than about 1 gram of DMSO per kilogram of body weight (Table 3). Although
`intravascularhemolysis has been reported after DMSO administration to cats (DiStefano & Klahn, 1965), and
`possibly humans (Samoszuk etal, 1983), Davis et al did not detect significant hemolysis in their clinical study
`as reflected by a major change in hematocrit. Hemoglobinuria occurs frequently (Kessinger et al, 1990),
`
`Patient
`1
`2
`3
`4
`5
`6
`7
`8
`
`Table 3. Serum Osmolality Changes Associated With Cryopreserved HSC Reinfusion
`Osmolality
`(mOsmlkg H20)
`Patient weight
`Infusion duration
`Volume
`DMSO
`(kg)
`(gm)
`(ml)
`(min)
`Pre
`Post
`59.9
`100
`289
`293
`11.1
`17
`67.2
`9
`100
`11.1
`291
`303
`80.4
`100
`282
`14
`277
`11.1
`100
`64.3
`290
`31
`11.1
`277
`87.0
`116
`95.0
`856
`315
`ND
`96.4
`100
`28
`11.1
`ND
`277
`79
`95.0
`857
`ND
`297
`147
`96.8
`303
`150
`48.3
`436
`ND
`Blood for serum osmolality testing was obtained immediately before initiation of HSC reinfusion (before mannitol
`administration for patient 4) and immediately (at 6 hours for patient 5) after the completion of the infusion. All patients
`were premedicated with 12.5 gm of mannitol, 50 mg of diphenhydramine, 250 mg of hydrocortisone, and saline
`hydration. Serum osmolality was determined by measuring freezing-point depression. ND = not done.
`239
`
`Downloaded by University of Michigan from www.liebertpub.com at 05/27/20. For personal use only.
`
`Novo Nordisk Ex. 2040, P. 7
`Mylan Institutional v. Novo Nordisk
`IPR2020-00324
`
`

`

`ROWLEY
`presumably resulting from hemolysis of red cells within the HSC inoculum. If large quantities of cells frozen
`in DMSO are to be infused, the infusion can be separated over 2 days to avoid complications from infusion of
`excessive amounts of DMSO.
`Hydroxyethyl starch
`HES is a polymeric substance containing chains of different molecular weights. Initially explored as a
`cryoprotectant for red blood cells, it was found also to be effective for a variety of other cells (Ashwood-Smith
`etal, 1972; Takahashi etal, 1988). The major focus in the study of HSC cryopreservation using extracellular
`cryoprotectants has been their use in combination with penetrating cryoprotectants. In one early study, the
`addition ofpolyvinylpyrrolidone (PVP) to glycerol or DMSO improved the cryopreservation of murine cells
`compared to the use of a penetrating agent alone (van Putten, 1968). Stiff et al froze human cells in a
`combination of 5% DMSO, 6% HES, and 4% human serum albumin, and reported improved progenitor cell
`survival as determined using in vitro cultures (Stiff et al, 1983). They subsequently successfully used this
`mixture of cryoprotectants to cryopreserve the bone marrow of 60 patients (Stiff etal, 1987). No engraftment
`failure was attributed to this technique, or after the cryopreservati