columns were kept identical. Eluent: A: 0.05 M Kl-l,P0,. pH = 2.5 (H,P0.);
`3; CH,OH Gradient (linear): 0- 35 min from 90% A to 55'/. A, 35-40 min to
`400/. A; 40-60 min to 40% A; 60-70 min to 25% A. 70-00 min to 20"/. A.
`How rate: 0.7 ml/min. Temperature: If not otherwise steated the experiments
`mm performed at room temperature. Other selected temperatures have been
`adjusted by a column Ihemtostal. Sample volume: 20 pl. Ahsorbance: 285 nm.
`HPLC system: Gradient system 600 E (Fa. Millipore Waters). diodenarray
`detector 990 (Fa. Millipore Waters). HPLC-column thermostat 5-85 “C (Fa.
`Knauerl-
`
`3.2.2. Peak idenll/imrion
`Retention times and UV-spectra of reference substances have been used to
`identify the peaks in the chromatograms The dilferent reference substan
`have been checked by 'H- and "C-NMR-spectroscopy.
`
`3.3. Solid phase NMR spectroscopy
`NMR-Unit: Bruker AMX 300 with a CP-MAS Unit.
`
`3.3.1. "Si-CP-MAS-NMR
`Measuring frequency: 59.63 MH1. Proton resonance: 300.14 MH7. Optimal
`contact-time: 20ll)0usec for LiChrosphcr 60 RP-select B. 5000 pace for
`Nucleosil I00 A8.
`
`33.2. "C-Cl’-MAS—NMR
`Measuring frequency: 15.48 MHz. Proton resonance: 300.I4 MHz. Optimal
`contact-time: 20(1) user: for Nucleoxil KI) AB. 2000 usec for LiChrospher 60
`RP-select B. 2(X)O usec for Eurospher 1(X).
`
`Acknowledgement: We are most grateful to our colleague Mr. Lubda (Fa.
`Merck) for running the CP-MAS-NMR spectra.
`
`References
`I Gilpin, R. K.: Burke, M. F.: Anal. Chem. 45. I383 (I973)
`2 Dorsey. J. G.; Dill. K. A.: Chem. Rev. 89. 331 (1989)
`3 Eisenbeill. 57.: Her. Bunsenges Phys. Chem. 93. I019 (I989)
`4 Morterra, C.; Low, M. J. D.: J. Canal. 28. 265 (I973)
`5 Waltsmundzlti. A.: Rudzinski. W.; Suprynowicz. 2.; Lebodn. K.: J. Chroma-
`togr. 92. 9 (1974)
`(Ed.):
`in: Horvath. Cs.;
`6 Karger, B. L.; be Page. J. N.; Tanalra, N.;
`High-Performance Liquid Chromatography. Advances and Perspectives.
`Vol. I. S. I59. Academic Press. New York 1980
`7 Nakum. A.: Horvath. C.: J. Chromatogr. 203. 53 (I98l)
`8 Kfihler. J.; Kirkland. J. J.: Chromatogr. 385. 125 (I987)
`9 Kohler. J.; Chase. D. B.; Farlee. R. D.: Vega. A.
`.l.: Kirkland. J. J.: J.
`Chromatogr. 352. 275 (I986)
`10 Unger. K. K.: Porous Silica, Journal of Chromatography Library. Vol. I6.
`Elsevier, Amsterdam 1979
`ll Pfleiderer. B.; Albert. K.: Bayer. E.: J. Chromatogr. 506. 343 (I990)
`I2 Sinsdorf. D. W.; Maciel, G. E: J. Am. Chem. Soc. I05, I487 (I983)
`I3 Sindorf. D. W.; MacieL G. E.: J. Am. Chem. Soc. I03. 4263 (I98l)
`I4 Bayer, E.: Albert, K.; Reinerg J.; Nieder, M.; Muller, D.: J. Chromatogr.
`264. I97 (I983)
`I5 Szabo. K.; Le Ha. N.: Schneider. P.; Zeltner. P.; Kovuts. E.: Helv. Chim.
`Acta 67, 2I28 (I934)
`16 Hetern. M.; van de Ven. L.; de Haan. J.; Cramers. C.; Albert. K.: Bayer.
`E.: J. Chromatogr. 479. 269 (1989)
`I7 KinkeL J. N.; Unger, K. K.: J. Chromntogr. 316. I93 (I984)
`I8 Sokolowski. A.: Wahlund. K.-G.: J. Chromatogr. 189. 299 ([980)
`I9 Freytag, W. E.: Step)‘. “(.2 Pharm. Ztg. Wiss. 6.]l38, I26 (I993)
`20 Fulde, G.: Dissertation Marburg I993
`21 Jones. R. R.; Hnrkrader. R. 1.; Soulhard. G. L.:J. Nat. Prod. 49. l [09 (I986)
`22 Berthold, A.: J. Chromatogr. 549, I (1991)
`23 Unger. K. K.: Porous Silica. Journal of Chromatography, Library, Vol. 16,
`S. 102. Elsevier. Amsterdam I979
`24 Sander, L C.; Wise. 3 A.: Anal. Chem. 56. 504 (I984)
`
`Received January 27. 1994
`Accepted March 20. 1994
`
`Dr. Hanns Hiiberlein
`lnstitut fur Pltarmnzeutische Biologic
`der Philipps-Universitit Marburg
`Deutschhausstrafle I7‘/,
`D-35037 Marburg
`
`III. Medizinischa Klinik und Poliklinik Abteilung Hématologie’, des Klinikums der Johannes Gutenberg-
`Apothake‘,
`Universitéit, Mainz
`
`Stabilitfit von Bendamustinhydrochlorid in Infusionsliisungen
`
`BIRGIT MAAS‘, C. HUBER’ und IRENE KRAMER‘
`
`Bendamustinhydrochlorid wurde rnittels Reversed Pha-
`se HPLC bestimmt. Die chemische Stabilitfit (I90) des
`Zytostatikums (0,25 mg/ml) betrégt in 0.9% Kochsa.lz-
`losung bei 4 °C 120 h und bei 23 °C 9 h. Stabilititsbeeim
`llussende Faktoren sind pH-Wert, Temperatur und
`Chloridionenkonzentration. Kochsalzlésung 0.9% ist
`als Tragerlésung ft'rr Bendamustin-lnfusionslésungen
`einzusetzen. Die Stabilitfitszeitraume gewihrleisten eine
`unproblematische Aulbewahrung und Applikation in
`der klinischen Praxis.
`
`Stability of bendamustine hydrochloride in infusions
`The stability of bendamustine hydrochloride (0.25 mg]
`ml) in 0.9% sodium chloride was studied after storage
`at 4 °C and 23 °C using Reversed Phase HPLC. Bend-
`amustinhydrochloride is stable at 4 °C for I20 h and at
`23 “C for 9 h. Temperature, pH and chloride concentra-
`tion are important factors which influence stability.
`Isotonic sodium chloride must be used for the prepa-
`ration of bendamustine hydrochloride infusions.
`
`l. Einleitung
`(Ribomustin". 4-[5-Bis(2-chlorethyl)aminc]-
`Bendamustin
`l-methyl-2-benzimidazolylbutansiiure) ist ein wirksames Che-
`
`Pharmazie 49 (1994). H. 10
`
`motherapeutikum in der Bchandlung maligner Erkrankungen.
`Die Stabilitiit der lyophilisiertcn Trockensubstanz ist bereits
`bekannt [1]. lm Rahmen von Untersuchungen zur Pharmako-
`kinetik Von Bendamustin am Menschen wurden auch in vitro-
`Messungen zur Stabilitiit der gelosten Substanz bei einer
`Temperatur von 37 °C in gepuffertem Medium (pH = 7,5;
`tso = 6,2 min) und in heparinisiertem Plasma (t,,, = 1,67 h)
`durchgeffthrt [2]. Es fehlen aber bislang Angaben zur Halt-
`barkeit Von Bcndamustin-lnfusionslésungen. Da die Her-
`stellung patientenbezogener, applikationsfertiger Zytostatika-
`Injektions- oder -lnfusionsliisungen immer hiuliger durch eine
`zentrale Zytostatikazubereitung in der Krankenhausapotheke
`erfolgt und die Infusion nicht mehr unmittelbar vor ihrer
`Applikation zubereitet wird, sind Kenntnisse zur Stabilitat
`von Bendamustinhydrochlorid in waflriger Lésung erl'order-
`lich [3].
`
`2. Untersuchungen und Ergebnisse
`
`in wiiflriger Lésung sehr in-
`Bendamustinhydrochlorid ist
`stabil. In schwach saurer, neutraler sowie alkalischcr Losung
`erfolgt eine Hydrolyse der Bis-2-Chloretylaminofunktion.
`Analog der Hydrolyse von Melphalan entsteht vermutlich
`fiber ein Aziridium-Kation zunichst das Monohydroxybend-
`amustin und nachfolgend das Dihydroxyderivat (Schema) [4].
`In stark saurer Lésung wird die Hydrolyse durch Proto-
`
`775
`
`FRESENIUS KABI 1009-0001
`
`FRESENIUS KABI 1009-0001
`
`

`
`
`
`nierung des freien Elektronenpaares am Stickstoff erschwert.
`Neben dem pl-I-Wert hat die Chloridionenkonzentration
`einen wesentlichen Einflufl auf die Lage des Hydro|ysegleieh-
`gewichtes. Mit steigender Chloridionenkonzentration sollte
`bevorzugt die Regeneration des Bendamustinhydrochlorids
`erfolgen, wodurch die Bildung des Hydroxyderivates zuriick-
`gedrfingl wird. Ffir Melphalan ist die Abhiingigkeit des
`Hydrolysegleichgewichtes von der Chloridionenkonzentra-
`lion [5] und der Temperatur dokumentiert [4]. In Kenntnis
`dieser Zusammenhéinge fir das strukturverwandte Melphalan
`und unter dem Aspekt der Anwendung am Patienten als
`Infusion crscheint isotonische Natriumchloridlésung als Tra-
`gerlésung ffir Bendamustinhydrochlorid geeignet. Ungiinstig
`ist allerdings die nur mfifiige Léslichkeit der Substanz in
`diesem Medium, so daB die Zersetzung bereits wiihrend des
`zeitaufwendigen Losevorgangs beginnt. In reinem Wasser ist
`Bendamustinhydrochlorid gut léslich. Es empfiehlt sich daher
`das Lésen in einer geringen Menge Wasser {fir Injektions-
`zwecke und das sofortige Verdiinnen mit isotonischer Na-
`triumchloridlésung.
`
`Schema
`
`R—N
`
`Cl
`
`Cl
`
`=== Cl'+R —N +
`
`Cl
`lH,O
`OH
`H"+R —N/_
`\—C1
`
`+ /-OH
`C1'+R—N\>
`
`H20 R_NF°“ +H.
`¥oH
`
`Aufgrund entsprechender Vorversuche wurden folgende Be-
`dingungen fiir die Stabilititsuntersuchungen gewiihlt: 25 mg
`Bendamustinhydrochlorid (=1 Ampulle Ribomustin“) wur-
`den zunichst in 10 ml Wasser geliist und anschlieflend mit
`03% Natriumchloridlésung zu 100 ml verdiinnt. Die damit
`hergestellte Konzentration Von 0,25 mg/ml entspricht thera-
`peutischen Erfordernissen. Die Stabilititsuntersuchung wurde
`bei Temperaturen von 4°C und 23°C durchgefiihrt. Die
`Bestimmung von Bendarnustinhydrochlorid erfolgt mittels
`Reversed Phase HPLC. Der ffir Bendamustin charakte-
`ristische Peak (Abb.) crscheint bei einer Retentionszeit (RT)
`Von 5,93 min. Das erste Zersetzungsprodukt, verrnutlich das
`Monohydrolyseprodukt, crscheint mit einer RT von 4,14 min.
`Bei dem als “unknown” bezeichneten Peak bei einer RT von
`5.23 min handelt es sich um ein Nebenprodukt aus der
`Synthese. Mittlerweile ist es durch Optimierung des Syn-
`theseverfahrens gelungen, die Verunreinigung véllig zu elimi-
`nieren [6]. Die Fliche dieses Peaks indert sich wihrend der
`gesamten Analysezeit nicht. Nach einer Aufbewahrungszeit
`von 1l9h ist
`im Chromatoyamm neben der deutlichen
`Zunahme des Monohydroxyderivales ein vierter Peak bei
`einer RT von 3,78 min zu erkennen, wobei cs sich urn das
`Dihydroxyderivat handelt. Mit einer Probe der kfuzlich als
`Dihydroxyderivat charakterisierten Verbindung [7] war es
`moglich, diescs Zersetzungsprodukt in unserem HPLC Assay
`zu identifizieren.
`Aufgrund der rasch fortschreitenden Hydrolyse von wfiflrigen
`Bendamustinhydrochloridlésungen dfirfen bei allen chroma-
`tographischen Bestimrnungen und entsprechenden Validie-
`rungen nur frisch hergestellte Liisungen verwendet werden,
`die unmittelbar nach ihrer Zubereitung injiziert werden mfis-
`sen. Simtliche Standardproben zur Validierung der Method:
`wurden mit Ribomustin* durchgeffihrt. Die MeBergebnisse
`erlauben somit cine relative Aussage zur Stabilitfit von Benda-
`
`776
`
`mustin, was hit unsere Frageslellung ausreichend ist.
`Vor Versuchsbeginn hatte die frisch hergestellte Untersu.
`chungslésung einen pH-Wert von 4. Die Bestimmung der Zer.
`setzung von Bendamustinhydrochlorid-Infusionslésungen er.
`folgte zuniichst bei 4 °C. Die Meflergebnisse sind in Tabelle 1
`zusammengefafit. Die Geschwindigkeitskonstante (‘k) und
`die Zeit bis zurn Verlust Von 10% wurden mittels exponen.
`tieller Regression berechnet. Fiir ‘k ergab sich ein Wert Von
`-0,00093 b’ ‘. Der Korrelationskoeffizient wurde mit 0,9892
`ermittelt. Hieraus errechnet sich ffir die Stabilitit (:90) von
`Bendamustinhydrochlorid in Natriumchloridlésung 0,81“/.
`bei 4 °C ein Wert von 118,5 h. Weiterhin wurde die Stabilitit
`von Bendamustinhydrochlorid-lnfusionslosung bei 23 °C be-
`stimmt. Die MeBergebnisse sind in Tabellc 2 zusammengefaflt.
`Die Geschwindigkeitskonstante (‘k) und die Zeit bis zum
`Verlust von 10% (tgo) wurden mittels exponentieller Regres-
`sion berechnet. Fiir ‘k ergab sich ein Wert von —0,0l 165 h‘ 1
`(Korrelationskoefiizient 0,9985). Hieraus ertechnet sich ffir
`die Stabilitfit von Bendamustinhydrochlorid (I90) in Natrium-
`chloridlésung 0,81% bei 23 “C ein Wert Von 9,2 h.
`
`Tnbelle I: Stnbilililsdnten flir Bendamlmlnllydrodilorld be! 4 °C
`Ztil
`Miuzhlerl der
`Sundard-
`Variation;
`Gehall III
`[In]
`gemmeneu
`abweichung
`koeflizienl
`Bendnmuuilr
`Konzmln-
`[%|
`HCI |%|
`lionen an
`(I, - 100%)
`BendarmIs(in-
`HCI [mg/ml]
`IT" n - 9
`
`0.00
`19.77
`42.57
`71,63
`92,55
`1 18.27
`
`0.2499
`0.2478
`0.2422
`0,2369
`0,2316
`0.2232
`
`0.00160
`0.00175
`0.(X)230
`0,(XX)47
`0.111052
`0,(XX)49
`
`0,64
`0,71
`0,95
`0.20
`0,22
`0,22
`
`100
`99,14
`96,22
`94,79
`92,68
`89,29
`
`-IDIvolt! m
`
`Mittelwen der
`gemesenen
`Konzenlrationen
`an Bendnmuslin-
`HCI [mg/ml]
`fiir n = 9
`
`Stnndnrd-
`Vnrintions- Gelull
`abweichnng koeflizient
`an Benda-
`[Va]
`mustin-HCI
`['/a]
`(l,, = HI)"/u)
`
`Abb.: HPLC-Chromalogramm von Bendamustinhydrochlorid (0,25 mg/rnl in
`NaCl 0.81% unmiuelbar nach Verdflnnung
`
`Pharmazie 49 (1994), H. 10
`
`FRESENIUS KABI 1009-0002
`
`FRESENIUS KABI 1009-0002
`
`

`
`1-.|,¢|Ie2: Stalnllititsdaten fllr Bendllntlnhydrochlorid bei 23 ‘C
`Geluli nn
`Bendamuuim
`HCI [96]
`(.0 3 mo-/.)
`
`SundIrd-
`Ibweiehung
`
`VnriIIions-
`koelfiziail
`|'/-|
`
`an
`M
`
`Milldwerl der
`pmenenen
`Konaenua-
`(ion in
`Bendniniulim
`HCI lmglmll
`filr n = 3
`
`0 00
`0,2525
`0,00107
`0,43
`100
`0,2508
`0,(X)2l2
`0,85
`99,31
`0,2401
`0.00130
`0,54
`95,06
`0.2375
`0,002l I
`0.89
`94.04
`0.2243
`0,00404
`I,80
`88.84
`0.1966
`0.00139
`0,71
`77,87
`
`H,SO.). Methanol 40/60 (V/V). Fluflmle: 0.3 nil/min. Welienlinge: 254 mm.
`lnjeklionsvolumenz 10 pl, Mefllemperalurz 4 “C bzw. 23 °C.
`
`4.3. Prizision des Gerites und der Melhode
`
`Sechs Einzelinjektionen nus sech: jeweils friach hergeslelllen Referenzldsungen
`der Konzentmion 0.25 mg/ml warden beziiglich erminelter Konzemration und
`Relenlionszeil nusgewerleL _Die Ergebnisse sind in Tnbelle 3 dargmlelll.
`
`TabelIe3: Prin'sion des Gerites und (let Methode
`Probe Nr.
`
`Men;
`lmsl
`
`Relenlicnneil
`[Ins]
`
`0,2499
`0,2494
`0,2497
`0,2493
`0,2486
`0,2492
`0,2494
`0,000433
`0.2%
`
`5,93
`5.92
`5,92
`5.94
`5.93
`5,93
`5,93
`0,0063
`0.1%
`
`Mittelwert
`Slandardabweichung
`Variationskoeffizient
`
`4.4. Eichung des Syslems
`Die Eichung dc: Systems eifolgl miuels Mehrpunkleiehung mil exlemem
`Standard. Es wenden je drei Injektionen von I0‘/u. IN‘/u und 200% det zu
`bestimmenden Konzenlmion nusyfiihn (Tubelle 4). Dazu wird eine Ampulle
`Ribomustin‘ zunichsl in I0 ml H,0 geléul und anschlieflend mil NaC1 0.9‘/u
`zu 0,025 mg/ml, 0,25 mg/ml und 0,50 mg/ml verdflnnl. Die Eichgernde ergibl
`xieh nus den berechneten Peakflichen, nufgetrngen gegen die theorelische
`Konzenuution. Der Kotmlationskoeflizienl betrigl 0.9997.
`
`Tabe|Ie4: Eicllulg des Systems
`Theoteliaell
`Minelwen der
`Konzsenlnnion
`gefundenen
`lfionzenlnliorlen
`lml/ml]
`Wu’ - 3
`[MB/Ml]
`
`0.025
`0.25
`0.50
`
`0,0266
`0,2499
`0,5080
`
`Vnrielionr
`koeflizient
`[Va]
`
`l,4
`0.6
`0,4
`
`4.5. Probenvorbereitung
`Jeweils eine Ampulle Ribomustin' (4.l.) wird in I0 ml Wasser fiir lnjeklions-
`zwecke gelésl und anschlieflend mil NACI-Lésung nus Viaflex‘ Beuteln der
`Fimia Baxter (pH-Wen: 5.0-7.0) zu I00 ml verdfinnl. Daraus iesulliert eine
`NaCl-Konzenlmion von 0.81‘/s. Die Zubeneilung der Standardproben zur
`Erstellung der Eichgenden erfolgl enupnechend.
`
`4.6. Auswertung
`Die Auswenung erfolgi dutch aulomatische Peakllicheninlegnlion minels
`Eichgernde (Software Waters Maximn 820).
`
`Limerltw
`
`I Ribomuslin' Slandardinfonnalion fir Krankenheusapotheker April 1993
`2 Preiu, R.; Sohr. R.; Matthias, M.: Brockmann. B.: Hfiller. H.: Pharmnzie
`40. 782 (I985)
`3 Kriimet. I.: Krnnkenhnuspharmnzie 8, 325 (I990)
`4 Sloul, S. A.; Riley, C. M.: Int. J. Pharm. 24, I93 (I985)
`5 Chang. S. Y.: Evans, T. S.; Albens. D. S.; J. Phann. Pharmacol. 31. 853
`(I979)
`6 Persénliche Mineilung der Firma ribosepharm 1994
`7 Gun, R.: Unveréffenllichie Ergebnisse (Inslitul fiir Phannnzie. Universilil
`Regensburg)
`8 Flora, K. P.; Smith. S. L.; Cndock. J. C.: J. Chronuxogr. I77, 9| (I979)
`
`Eingegangen am 8. Juni I994
`Angenornmen am 4. Juli I994
`
`Dr. Ilene Krilner
`Apolheke der Klinikums
`Langenheckslr. I
`D-55l3I Mainz
`
`4,17
`6,33
`10,25
`21,50
`
`3. Diskon
`
`Die Messungen aeigen, dafl sowohl bei 4 “C als auch 23 °C
`die hydrolytische Zersetzung wiflriger Bendamustinhydro-
`chloridlésungen in dem gemessenen Bereich linear verliiuft.
`Ffir die Zersetzung von Melphalan ist eine Kinetik pseudo-
`erster Ordnung beschrieben [8], die aufgrund des identischen
`Hydrolysemechanismus auch ffir das
`strukturverwandte
`Bendamustin angcnommen werden kann. Die mittels expo-
`_/ nentiellcr Reyession berechneten Korrelationskoeflizienten
`ff unterstfitzen diese Annahme. Neueste Untersuchungen zur
`I-Iydrolysekinetik von Bendamustin in Wasser bei Raum-
`Iemperatur fiber den gesamten Zersetzungszeitraum verifi-
`zieren die Zersetzungskinetik pseudo-erster Ordnung [7]. Alle
`Untersucliungslésungen wiesen den stabilitiitsbegfinstigenden
`pH-Wen von 4 auf. In Vorversuchen wurde die I-Iydrolyse
`von Bendamustinhydrochlorid in Wasser
`fiir
`Injektions-
`zwecke fiberprfift, wobei gezeigt werden konme, dam in diesem
`chloridfreien Medium das Ausmafl der Hydrolyse trotz
`gleichen pH-Wencs wesentlich gr6Ber ist. Ffir Lésungen der
`Ausgangskonzentration 0,25 mg/ml wurde bei Raumtempe-
`ratur exemplarisch eine tgo von 4,2 h bestimmt. Dies belegt
`den stabilitfitserhéihenden Einflufl der Chloridionen. Als Tr5-
`germedium {fir Bendamustiuzubereitung ist ausschliefllich
`isotonische Natriumchloridlésung zu verwenden.
`Die Hydrolyse von Bendamustinhydrochlorid ist erwartungs-
`gemiifl stark temperaturabhéngig. Die Stabilitit der Lésung
`mit einer Konzentration von 0,25 mg/ml
`ist bei 4°C (t,°
`= 118,5 I1) um mehr als das Zehnfache gr6Ber als bei 23 °C
`(:90 = 9,2 h).
`I-‘fir die klinische Praxis bedeutet dies, daB
`Bendamustin-Infusionslésungen kiihl zu lagem sind. Mit
`physiologischer Kochsalzlésung als Trfigermedium und unter
`Aulbewahrung im Kfihlschrank ist fiir Bendamustinlésungen
`der Konzentration 0,25 mg/ml eine Hallbarkeit von vier d
`gegeben, so daB eine Versorgung mit applikationsfertigen
`Bendamustin-Infusionslésungen im Rahmen einer zentralen
`Zytostatikazubereitung dutch die Krankenhausapotheke
`sicher méglich ist. Ffir die empfohlene Applikation als I(urz-
`zeitinfusion fiber 30 min sind ebenfalls keine Stabilit§ts-
`probleme zu erwarten, da bei Raumtemperatur fiir diese
`Bendamustinzubereitungen eine Stabilitit von 9h gegeben
`1st.
`
`4. Experimenteller Teil
`4.]. Subsuinzen
`Die Bestimrnungen wurden mi! Ribomustin-Ampullen der Firme riboee-
`phann (Charge 0190393) durchgefiihrl. Jede Durchslechllasehe mi! 55 mg
`Trockensubslanz enlhill 25 mg Bendamustinhydrochlorid (Hill'nIoII': Man-
`niloly
`
`4.2. Chromnlographisehe Bedingungen
`Die Messung erfolgle mi! einer Millipore/Weten HPLC-Anlage.
`Pumpe: Waters 510 HPLC-Pump. Delektor: UV-Deleklor Walets 490 E.
`Probengeber: Wnlers 717-Aulosempler. Slulez Waters Novapek C 18, 60A,
`4pm. 3.9x I50 mm. Flieflmiuel: wiiflriye N:,S0.-Lésung 20mM (pH 3.
`
`Pbarmazie 49 (1994), H. 10
`
`FRESENIUS KABI 1009-0003
`
`FRESENIUS KABI 1009-0003
`
`

`
`Pharmacy1, III. Medical Clinic and Polyclinic, Division of Hematology2, of the University Medical Center of the
`Johannes Gutenberg University Mainz
`
`Stability of Bendamustine Hydrochloride in Infusions
`BIRGIT MAAS1, C. HUBER2 AND IRENE KRAMER1
`
`Bendamustine hydrochloride was identified by means of
`reverse-phase HPLC. The chemical stability (t90) of the
`cytostatic (0.25 mg/mL) in 0.9% sodium chloride is 120 h
`at 4°C and 9 h at 23°C. The pH value, temperature, and
`chloride ion concentration are influencing factors for
`stability. 0.9% sodium chloride solution must be used as
`the carrier solution for bendamustine infusion solution.
`The stability times ensure unproblematic storage and
`application in clinical practice.
`
`Stability of bendamustine hydrochloride in infusions
`The
`stability of bendamustine hydrochloride
`(0.25
`mg/mL) in 0.9% sodium chloride was studied after storage
`at
`23°C using
`reverse-phase HPLC.
`4°C and
`Bendamustine hydrochloride is stable at 4°C for 120 h and
`at 23°C for 9 h. Temperature, pH, and chloride
`concentration are key influencing factors for stability.
`Isotonic sodium chloride must be used for the preparation
`of bendamustine hydrochloride infusions.
`
`1.
`Introduction
`Bendamustine
`4-[5-bis(2-chlorethyl)amino]-1-
`(Ribomustin®,
`methyl-2-benzimidazolyl
`butyric
`acid)
`is
`an
`effective
`chemotherapeutic drug in the treatment of malignant diseases.
`
`The stability of the lyophilized dry substance is already known
`[1]. In studies of the pharmacokinetics of bendamustine in
`humans, in vitro measurements of the dissolved substance were
`carried out at a temperature of 37°C in a buffered medium (pH =
`7.5; t50 = 6.2 min) and in heparinized plasma (t90 = 1.67 h) [2].
`However, data is still lacking on the stability of bendamustine
`infusions. Since patient-specific, ready-to-use cytostatic infusions
`or injections are increasingly prepared in a central cytostatic
`preparation facility in the hospital pharmacy and the infusion is
`no
`prepared
`immediately
`before
`administration,
`longer
`information on the stability of bendamustine hydrochloride in an
`aqueous solution is imperative [3].
`
`2. Tests and results
`Bendamustine is very unstable in an aqueous solution. In a
`weakly acid, neutral, and alkaline solution, hydrolysis of the bis-
`2-chlorethylamino function occurs. As in the hydrolysis of
`melphalan, monohydroxy bendamustine presumably develops
`first via an aziridium cation, and then the dihydroxy derivative
`develops (see diagram) [4]. In a highly acidic solution, hydrolysis
`is inhibited by protonation of the free electron pair on the
`nitrogen.
`In addition to the pH value,
`the chloride ion
`concentration has a significant
`impact on the hydrolysis
`equilibrium. It is preferred that the bendamustine hydrochloride
`be regenerated as the chloride ion concentration increases, which
`suppresses
`the formation of
`the hydroxyl derivative. The
`
`Pharmazie 49 (1994), H. 10
`
`775
`
`FRESENIUS KABI 1009-0004
`
`

`
`for the stability (t90) of bendamustine hydrochloride in a 0.81%
`sodium chloride solution at 4°C. The stability of bendamustine
`hydrochloride infusion solution was also determined at 23°C. The
`measurement results are shown in Table 2. The rate constant (1k)
`and time to 10% loss (t90) were calculated by means of
`exponential regression. A value of -0.01165 h-1 was obtained for
`1k (coefficient of correlation 0.9985). This calculated to a value of
`9.2 h for the stability of bendamustine hydrochloride (t90) in a
`0.81% sodium chloride solution at 23°C.
`
`Table 1: Stability data for bendamustine hydrochloride at 4°C
`Standard
`Coefficient of
`Content of
`Time [h]
`Mean
`Deviation
`Variation [%]
`Bendamustine
`Concentration
`HCl [%]
`of
`(t0 = 100%)
`Bendamustine
`HCl [mg/mL]
`Measured for
`n = 9
`0.2499
`0.2478
`0.2422
`0.2369
`0.2316
`0.2232
`
`0.00
`19.77
`42.57
`71.63
`92.55
`118.27
`
`0.00160
`0.00175
`0.00230
`0.00047
`0.00052
`0.00049
`
`0.64
`0.71
`0.95
`0.20
`0.22
`0.22
`
`100
`99.14
`96.22
`94.79
`92.68
`89.29
`
`dependence of the hydrolysis equilibrium on the chloride ion
`is documented for
`concentration [5] and temperature [4]
`melphalan. On account of these relationships for the structurally
`related melphalan and the fact that it is administered to the patient
`as an infusion, isotonic sodium chloride solution seems to be
`suitable as a carrier solution for bendamustine hydrochloride. The
`merely moderate solubility of the substance in this medium is,
`however, unfavorable, because decomposition already begins
`during the time-consuming dissolution process. In pure water,
`bendamustine has good solubility. It is therefore recommended
`that it be dissolved in a small amount of water for injection and
`then immediately diluted with an isotonic sodium chloride
`solution.
`
`Diagram
`
`following
`the
`Based on corresponding preliminary tests,
`stability testing: 25 mg of
`conditions were selected for
`bendamustine hydrochloride (= 1 ampoule of Ribomustin®) was
`dissolved first in 10 mL of water and then diluted with 0.9%
`sodium chloride solution to 100 mL. The produced concentration
`of 0.25 mg/mL meets therapeutic requirements. The stability test
`temperatures of 4°C and 23°C. The
`was performed at
`bendamustine hydrochloride was identified by reverse-phase
`HPLC. The characteristic peak of bendamustine (see figure)
`occurs at a retention time (RT) of 5.93 min. The first
`decomposition product, presumably the monohydrolysis product,
`appears at an RT of 4.14 min. The “unknown” peak at an RT of
`5.23 min is a by-product from synthesis. By optimizing the
`synthesis procedure in the meantime, we have been able to
`completely eliminate the impurity [6]. The area of this peak does
`not change during the entire time of analysis. After a storage time
`of 119 h, in addition to the clear increase in the monohydroxy
`derivative, a fourth peak can be seen at an RT of 3.78 min, which
`is the dihydroxy derivative. With a sample of the compound
`recently characterized as a dihydroxy derivative [7], we were able
`to identify this decomposition product in our HPLC assay.
`the
`rapid
`hydrolysis of
`aqueous bendamustine
`to
`Due
`hydrochloride solutions, only freshly prepared solutions which
`must be injected immediately following their preparation may be
`used in chromatographic determinations and corresponding
`validations. All standard samples for validation of the method
`were performed with Ribomustin®. The measurement results thus
`allow us to determine the relative stability of bendamustine,
`which is sufficient for our investigation.
`Before the start of testing, the freshly prepared test solution had a
`pH of 4. The decomposition of bendamustine hydrochloride
`infusions was determined first at 4°C. The measurement results
`are shown in Table 1. The rate constant (1k) and the time to 10%
`loss were calculated by means of exponential regression. A value
`of -0.00093 h-1 was obtained for 1k. The coefficient of correlation
`was determined as 0.9892. This calculated to a value of 118.5 h
`
`Figure: HPLC chromatogram of bendamustine hydrochloride (0 25
`mg/mL in 0.81% NaCl immediately after dilution
`
`[Translator’s note: See Table 1 for translation of figure.]
`
`776
`
`Pharmazie 49 (1994), H. 10
`
`FRESENIUS KABI 1009-0005
`
`

`
`Table 2: Stability data for bendamustine hydrochloride at 23°C
`Standard
`Coefficient of
`Content of
`Time [h]
`Mean
`Deviation
`Variation [%]
`Bendamustine
`Concentration
`HCl [%]
`of
`(t0 = 100%)
`Bendamustine
`HCl [mg/mL]
`Measured for
`n = 9
`0.2525
`0.2508
`0.2401
`0.2375
`0.2243
`0.1966
`
`0.00
`0.67
`4.17
`6.33
`10.25
`21.50
`
`0.00107
`0.00212
`0.00130
`0.00211
`0.00404
`0.00139
`
`0.43
`0.85
`0.54
`0.89
`1.80
`0.71
`
`Equipment and method precision
`4 3
`Six different injections from six freshly prepared reference solutions with
`a concentration of 0.25 mg/mL were analyzed for concentration obtained
`and retention time. The results are shown in Table 3.
`
`Retention Time
`[mg]
`5.93
`5.92
`5.92
`5.94
`5.93
`5.93
`5.93
`0.0063
`0.1%
`
`Table 3: Precision of equipment and method
`Sample No
`Quantity
`[mg]
`0.2499
`0.2494
`0.2497
`0.2493
`0.2486
`0.2492
`0.2494
`0.000433
`0.2%
`
`ean value
`Standard deviation
`Coefficient of variation
`
`123456M
`
`System calibration
`4.4
`The system was calibrated by means of multiple-point calibration with an
`external standard. Three injections of 10%, 100%, and 200% of the
`concentration to be determined were performed (Table 4). For this
`purpose one ampoule of Ribomustin® was first dissolved in 10 mL of
`H2O and then diluted with 0.9% NaCl to 0.025 mg/mL, 0.25 mg/mL, and
`0 50 mg/mL. The calibration curve was derived from the calculated peak
`areas plotted against the theoretical concentration. The coefficient of
`correlation was 0.9997.
`
`Table 4: System calibration
`Theoretical
`Mean Value of
`Concentration
`Concentrations
`[mg/mL]
`Found for n = 3
`[mg/mL]
`0.0266
`0.2499
`0.5080
`
`0.025
`0.25
`0.50
`
`Standard
`Deviation
`
`Coefficient of
`Variation [%]
`
`0.00036
`0.0016
`0.0021
`
`1.4
`0.6
`0.4
`
`Sample preparation
`4 5
`Each ampoule of Ribomustin® (4.1) was dissolved in 10 mL of water for
`injection and then diluted to 100 mL with NaCl solution from ViaFlex
`bags from the company Baxter (pH 5.0–7.0). This resulted in an NaCl
`concentration of 0.81%. The standard samples for the calibration curves
`were prepared in the same way.
`
`4.6 Analysis
`The analysis was performed with automatic peak area integration using
`the calibration curve (Waters Maxima 820 software).
`
`Literature
`1 Ribomustin® Standardinformation für Krankenhausapotheker, April
`1993
`2 Preiss, R ; Sohr, R ; Matthias, M ; Brockmann, B.; Hüller, H. Pharmazie
`40, 782 (1985)
`3 Krämer, I.: Krankenhauspharmazie 8, 325 (1990)
`4 Stout, S.A ; Riley, C.M.: Int. J Pharm 24,193 (1985)
`5 Chang, S.Y.; Evans, T.S.; Alberts, D.S.: J Pharm. Pharmacol. 31, 853
`(1979)
`6 Personal memo from the Ribosepharm company, 1994
`7 Gust, R.: Unpublished results (Institute of Pharmacy, University of
`Regensburg)
`8 Flora, K.P ; Smith, S.L.; Cradock, J.C.: J Chromatogr. 177, 91 (1979)
`
`Submitted on June 8, 1994
`Accepted on July 4, 1994
`
`Dr. Irene Krämer
`Apotheke des Klinikums
`Langenbeckstr. 1
`D-55131 Mainz
`
`777
`
`100
`99.31
`95.06
`94.04
`88.84
`77.87
`
`3. Discussion
`The measurements show that the hydrolytic decomposition of
`aqueous bendamustine hydrochloride solutions is linear over the
`range measured at both 4°C and 23°C. A pseudo first-order
`kinetic reaction is described for melphalan [8], which can also be
`assumed for the structurally related bendamustine because of the
`identical hydrolysis mechanism. The coefficients of correlation
`calculated by exponential regression support this assumption. The
`latest studies of the hydrolysis kinetics of bendamustine in water
`at room temperature verify the pseudo-first-order decomposition
`kinetics [7]. All test solutions had the stability-promoting pH of 4.
`In
`preliminary
`tests,
`the
`hydrolysis
`of
`bendamustine
`hydrochloride was tested in water for injection and,
`in this
`chloride-free medium, the degree of hydrolysis was substantially
`higher despite having the same pH value. Typically for solutions
`with the initial concentration of 0.25 mg/mL, a t90 of 4.2 h was
`determined at room temperature. This is evidence of the stability-
`increasing effect of the chloride ions. Isotonic sodium chloride
`solution must be used exclusively as the carrier medium for
`bendamustine preparation.
`As expected, the hydrolysis of bendamustine hydrochloride is
`highly temperature-dependent. The stability of the solution with a
`concentration of 0.25 mg/mL is more than ten times greater at
`4°C (t90 = 118.5 h) than at 23°C (t90 = 9.2 h). For clinical practice,
`this means that bendamustine infusions must be kept cool. If a
`physiological salt solution is used as the carrier medium and the
`infusions are stored in a refrigerator, the storage time is four days
`for bendamustine solutions with a concentration of 0.25 mg/mL,
`so a supply with ready-to-use bendamustine infusions is certainly
`possible within the framework of central cytostatic preparation in
`the
`hospital
`pharmacy. Likewise
`for
`the
`recommended
`administration as a short
`infusion over 30 min, no stability
`problems are expected, since these bendamustine preparations
`have a stability of 9 h at room temperature.
`
`4. Experiment
`4.1
`Substances
`The determinations were performed using Ribomustin ampoules from the
`company Ribosepharm (batch 0190393). Each vial with 55 mg of dry
`substance contains 25 mg of bendamustine hydrochloride (excipient:
`mannitol).
`
`Chromatographic conditions
`4.2
`The measurement was performed using a Millipore/Waters HPLC system.
`Pump: Waters 510 HPLC pump; detector: Waters 490E UV detector;
`proton donor: Waters 717 autosampler; column: Waters Novapak C18,
`60Å. 4µm, 3 9 x 150 mm; solvent: aqueous Na2SO4 solution 20 mM (pH
`3, H2SO4), methanol 40/60 (v/v); flow rate 0.3 mL/min; wavelength: 254
`mm; injection volume: 10 µL; measuring temperature: 4 C and 23 C.
`
`Pharmazie 49 (1994), H. 10
`
`FRESENIUS KABI 1009-0006
`
`

`
`E H S!
`
`4752 41°‘ Ave sw. Ste B
`Seattle WA. 981 15
`
`Phone: 206-938-3600
`
`AFFIDAVIT OF ACCURACY
`
`I, Amanda Olson, hereby declare as follows:
`
`I am a translator fluent in the German and English languages. and am authorized to provide this
`affidavit on behalf of Gcnnan Language Services.
`I am over eighteen years of age and fully
`competent to make this affidavit.
`I have personal knowledge of the infonnation contained in this
`affidavit, and it is true and accurate to the best of my knowledge.
`
`I have reviewed the German original and the English translation of Maas et aI., Pharmazie, vol.
`49, No. 10, pp. 291-299 (1994) (“Maas"). To the best of my knowledge and belief, the English
`translation is a complete and accurate translation of Maas.
`
`In signing this affidavit, I understand that the affidavit will be filed as evidence in a petition
`before the Patent Trial and Appeal Board of the United States Patent and Trademark Office.
`I
`declare under penalty of perjury of the laws of the United States of America that the foregoing
`information is true and accurate to the best of my knowledge. I understand that willful false
`statements and the like are punishable by fine or imprisonment, or both (18 U.S.C. § 1001) and
`may jeopardize the validity of the petition.
`
`Executed June 30, 2015 in Seattle, WA.
`
`Signed
`
`g
`
`)
`
`.
`
`NOTARY‘S DECLARATION
`
`On this 30th day of June, 2015, in King County, Seattle, WA, USA, Amanda Olson identified
`
`herself to me as the person who signed the declaration above.
`
`(Notary Pub c)
`
`Signed
`
`
`FRESENIUS KABI 1009-0007
`
`FRESENIUS KABI 1009-0007