Bioconjugate Chem. 2010, 21, 5–13
`
`5
`
`REVIEWS
`
`Antibody-Drug Conjugates: Linking Cytotoxic Payloads to Monoclonal Antibodies
`
`Laurent Ducry* and Bernhard Stump
`
`Lonza Ltd, CH-3930 Visp, Switzerland. Received May 4, 2009; Revised Manuscript Received July 7, 2009
`
`Antibody-drug conjugates (ADCs) combine the specificity of monoclonal antibodies (mAbs) with the potency
`of cytotoxic molecules, thereby taking advantage of the best characteristics of both components. Along with the
`development of the mAbs and cytotoxins, the design of chemical linkers to covalently bind these building blocks
`is making rapid progress but remains challenging. Recent advances have resulted in linkers having increased
`stability in the bloodstream while allowing efficient payload release within the tumor cell.
`
`INTRODUCTION
`After cardiovascular diseases, cancer represents the second
`most common cause of death in the Western world and is a
`major area of focus for the pharmaceutical industry. Surgery
`and radiotherapy are generally used when a tumor is localized
`to a certain tissue, but chemotherapy is needed when metastasis
`has occurred. Despite extensive research, most anticancer drugs
`have nonspecific toxicity. By targeting the cell cycle and thereby
`killing rapidly proliferating cells, they do not explicitly dis-
`criminate between healthy and tumor tissues and only gain a
`limited selectivity for malignant cells. Such cytotoxic drugs have
`a narrow therapeutic window, which limits their efficacy and
`results in severe side effects. Due to a lack of selectivity, drug
`concentrations that would eradicate the tumor can often not be
`used.
`In addition,
`tumors can develop resistance against
`anticancer drugs after prolonged treatment. Therefore, achieving
`improved tumor selectivity through targeting of cytotoxic drugs
`to the cancer cells is needed.
`A promising approach to achieving a more selective treatment
`is targeted prodrug therapy (1). Antibody-drug conjugates
`(ADCs) are ideal candidates for such prodrugs. ADCs are
`monoclonal antibodies (mAbs) linked to cell-killing drugs.
`Thanks to their high binding specificity for tumor-specific
`antigens, mAbs can be used as vehicles to target cell-killing
`payloads to tumor cells. Unique or overexpressed, tumor-specific
`antigens can be found in a wide range of human tumor cells
`(2). Some mAbs have the ability to recognize and specifically
`bind to these tumor-associated antigens. They can be used as
`single agents for the treatment of cancer through binding to
`cancer-cell-specific antigens and induction of an immunological
`response against the target cancer cell (3). However, therapeutic
`efficacy is often limited by the extent to which the antibody
`leads to cell death. Monoclonal antibodies are extremely
`discriminating for their targets but sometimes therapeutically
`ineffective on their own. The insufficient efficiency of most
`naked mAbs in cancer therapy has been circumvented by arming
`the immunoglobulin with radioactive isotopes (4) or cytotoxic
`drugs (5-7) (Figure 1), yielding highly specific ADCs.
`ADCs can be viewed as sophisticated delivery systems for
`antitumor cytotoxic drugs. The mAb guides the toxin precursor
`
`* laurent.ducry@lonza.com.
`
`Figure 1. Schematic representation of an antibody-drug conjugate.
`
`to the target cancer cell, where the prodrug can be converted
`chemically and/or enzymatically to the parent drug and unfold
`its cytotoxic activity. The ADC is exposed to different conditions
`on its journey from the blood vessels to the molecular target in
`the tumor tissue. The mode of action at cellular or molecular
`level is complex, with each step bringing its own challenges:
`a. Circulation. The ADC must behave like a naked antibody
`when circulating in the plasma. In particular, the linker must
`be stable in the bloodstream to limit the damage to healthy
`tissue. Decomposition or decay would release the cytotoxin
`before being delivered to the target site.
`b. Antigen Binding. It is necessary that the conjugated mAb
`retains high immunoaffinity. Attaching the cytotoxic compound
`to the mAb must thus not disturb its binding specificity.
`c. Internalization. A sufficient intracellular concentration of
`the drug must be achieved. This is challenging because antigen
`targets on cell surfaces are often present in limited numbers
`and the internalization process for antigen-antibody complexes
`is frequently inefficient.
`d. Drug Release. Once internalized, the ADC has to ef-
`ficiently release the original cytotoxic drug in its active form
`inside the tumor cell.
`e. Drug Action. The inherent potency of the released drug
`must be sufficient to kill the tumor cell, even at low concentra-
`tion. To achieve significant cytotoxicity, use of very potent drugs
`with subnanomolar IC50 (as free drug) becomes necessary. It
`has been shown that compounds found to be too toxic when
`tested as a stand-alone chemotherapy agent are suitable candi-
`dates for ADC payloads. These toxins can be 100 to 1000 times
`more cytotoxic than traditional anticancer agents.
`The molar loading of the drug on the mAb can be tuned to
`achieve the desired potency. Because of points (c) and (d), linkage
`of several drugs is generally necessary to achieve significant
`cytotoxicity. However, if too many cytotoxic molecules are attached
`10.1021/bc9002019  2010 American Chemical Society
`Published on Web 09/21/2009
`
`IMMUNOGEN 2128, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`6 Bioconjugate Chem., Vol. 21, No. 1, 2010
`
`Ducry and Stump
`
`Chart 1
`
`Chart 3
`
`Chart 2
`
`once the ADC entered the lower pH environment found inside
`a cell. The focus in the early days of ADC development was
`on acid-cleavable hydrazone linkers that were relatively stable
`at neutral pH (bloodstream pH 7.3-7.5), but were undergoing
`hydrolysis once internalized into the mildly acidic endosomes
`(pH 5.0-6.5) and lysosomes (pH 4.5-5.0).
`In the 1990s, Bristol-Myers Squibb scientists covalently linked
`doxorubicin, an intercalating agent that blocks DNA replication
`and is used as chemotherapeutic agent, to the humanized mAb
`BR96 (10). Doxorubicin was connected over a (6-maleimidoca-
`proyl)hydrazone linker to cysteine residues of the Lewis-Y specific
`mAb. The preclinical activities of BR96-doxorubicin (1, Chart 1)
`were remarkable, but despite the use of eight drugs per mAb
`molecule, the amount of conjugate needed to achieve these effects
`in ViVo were high (>100 mg/kg). This was presumably due to the
`relatively low potency of doxorubicin (IC50 of 0.1-0.2 µM for
`human carcinoma lines (11), whereas subnanomolar activities are
`now typically seen for ADC payloads). Moreover, the half-life of
`the drug in the blood was approximately 43 h, which is short
`compared to the half-lives of several days to weeks of the naked
`BR96 mAb in humans. The compound later failed to show
`sufficient clinical efficacy.
`The first and so far sole mAb linked to a cytotoxic payload that
`has been given regulatory approval is Wyeth’s gemtuzumab
`ozogamicin (2, Mylotarg; Chart 2) (12-14). The drug was
`approved by the US Food and Drug Administration (FDA) in 2000
`for use in patients over 60 suffering from relapsed acute myelocytic
`leukemia (AML), the most common form of leukemia in adults.
`Gemtuzumab ozogamicin (2) consists of N-acetyl-γ-calicheamicin
`(an enediyne antibiotic that binds to the minor groove of DNA
`and causes double-strand breaks leading to apoptosis) covalently
`attached to humanized anti-CD33 IgG4 κ antibody (hP67.6) via a
`bifunctional linker. The 4-(4-acetylphenoxy)butanoic acid moiety
`allows for attachment to surface-exposed lysines of the antibody
`over an amide bond and forms an acyl hydrazone linkage with
`N-acetyl-γ-calicheamicin dimethyl hydrazide. Upon internalization
`of the ADC, the calicheamicin prodrug is released by hydrolysis
`of the hydrazone in lysosomes of the CD33-positive target cells.
`The hydrolyzability of an unconjugated intermediate at 37 °C over
`24 h increased from 6% at pH 7.4 to 97% at pH 4.5 (15). The
`enediyne drug then gets activated by reductive cleavage of the
`disulfide bond. This disulfide linkage has been stabilized by two
`methyl groups to prevent premature release of calicheamicin by
`circulating reduced thiols, such as glutathione. In most preclinical
`models, the hydrazone linkage produced ADCs with higher potency
`than the corresponding amide-bearing conjugate 3, providing
`evidence that with this mAb the disulfide alone is insufficient for
`
`to the antibody, the body may recognize the conjugate as a damaged
`form of the protein and quickly clears it from the system, proving
`points (a) and/or (b) to be problematic. The extent of drug
`substitution may also affect pharmacokinetics, so that most of the
`time 2-4 drugs per antibody give the best therapeutic window (8, 9).
`To avoid the interference with the antigen recognition, the drugs
`are ideally linked at the Fc or constant region of the mAb that
`does not participate in binding to the antigen (point (b)). To regain
`the full cytotoxic potential of the drugs, cleavable linkers releasing
`the drug molecule at the target site in unmodified form were
`developed to address point (d), although these also affect the
`stability of the construct. Therefore, the right balance between
`plasma stability and efficient drug release at the target cell has to
`be found.
`The insights gained during ADC development within the past
`decade have resulted in a number of new strategies and
`drugs (4-7). This review focuses on linker technologies used
`to attach low molecular weight cytotoxic drugs to mAbs. The
`linker is a small but central part of ADCs and as such accounts
`for stability in circulation, favorable pharmacokinetic properties,
`and efficient release of toxic agent at the tumor site.
`
`CHEMICALLY LABILE LINKERS
`Selective linker cleavage and payload release is based upon
`the differential properties between the plasma and some
`cytoplasmic compartments. Linkers were chosen that were stable
`in the blood’s neutral pH environment but were getting cleaved
`
`IMMUNOGEN 2128, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Reviews
`
`Bioconjugate Chem., Vol. 21, No. 1, 2010 7
`
`Scheme 1. Presumed Release Mechanism of MMAE (7) from ADC 6 Incorporating a Valine-Citrulline Peptidic Linker
`
`Table 1. Compared Stabilities of Enzyme-Labile and Chemically Labile Linkers in ADCs with MMAE Conjugated to cBR96 mAb
`
`efficient release of the drug in the target cell. Interestingly, with
`the murine mAb CTM01, ADC 3 showed activities equal to or
`even greater than that of the corresponding hydrazone conjugate 2
`in several in Vitro and in ViVo tumor models, revealing that this
`conclusion is not general (16). In clinical trials, however, compound
`3 had only limited evidence of activity.
`Mylotarg has met limited successes due to a narrow thera-
`peutic window. The linker technology based on pH-dependent
`release mechanism is possibly not sufficiently stable and too
`much of the drug is being released in the bloodstream, as
`pharmacokinetic data have shown that the mean half-life of
`Mylotarg is 72 h (15). Nonetheless, the development of CMC-
`544 (inotuzumab ozogamicin), which is a humanized anti-CD22
`mAb identically attached to N-acetyl-γ-calicheamicin dimethyl
`
`hydrazide via the acid-labile 4-(4′-acetylphenoxy)butanoic acid
`linker is ongoing at Wyeth. Although this ADC is closely related
`to Mylotarg, good stability was shown in both human plasma
`and serum (rate of hydrolysis of 1.5-2%/day over 4 days) (17, 18).
`Disulfide bonds can be an alternative to acid-labile hydrazone
`linkers. In this case,
`the release is attributed to the high
`intracellular concentration of glutathione. This thiol-containing
`tripeptide is present in micromolar concentrations in the blood,
`whereas its concentration in the cytoplasma is in the millimolar
`range (up to 1000-fold higher). This is especially true for tumor
`cells, where irregular blood flow leads to hypoxic state
`(decreased oxygen level), resulting in enhanced activity of
`reductive enzymes and therefore even higher glutathione
`concentrations. In the absence of sulfhydryl groups, disulfide
`
`IMMUNOGEN 2128, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`8 Bioconjugate Chem., Vol. 21, No. 1, 2010
`
`Chart 4
`
`Ducry and Stump
`
`bonds are thermodynamically stable and provide reasonably
`good stability in the bloodstream.
`An example of an intracellularly cleavable disulfide-based
`linker was already discussed above (see “amide” conjugate 3).
`Other examples are taxoid ADC payloads, conjugated via a
`disulfide-bearing 4-mercaptopentanoate linker to mAbs directed
`against the epidermal growth factor receptor (19). However, the
`most significant examples are ADCs where the immunoglobulin
`is connected over the same type of linker to ImmunoGen’s drugs
`DM1 or DM4 (20-22). These potent thiol-containing may-
`tansine analogues are antimitotic agents that bind to tubulin and
`inhibit microtubule assembly. After internalization of the ADC
`via antigen-mediated endocytosis and delivery to lysosomes by
`vesicular trafficking, the mAb is believed to be degraded,
`affording lysine derivatives linked to the toxin (22). Further
`intracellular modifications like cleavage of the disulfide linker
`through a disulfide exchange, and finally thiol methylation
`presumably catalyzed by intracellular methyltransferases, gener-
`ate potent derivatives of DM1 or DM4 (22). These methylated
`drugs are uncharged, enabling them to pass back out through
`cell membranes to broaden their attack; the drug not only kills
`from inside the cell, but also knocks out adjacent cells within
`the tumor that are not targeted by the antibody (23). The linker
`of compound 4 (Chart 3) nevertheless proved to be relatively
`unstable in ViVo (half-life of 47 h in mice) (24). Replacement
`of DM1 by DM4 structurally leads to the introduction of two
`methyl substituents on the carbon atom geminal to the disulfide
`bond, thereby limiting the accessibility for reducing agents.
`Consistent with these considerations, ADC 5 exhibits an
`increased half-life of 102 h in mice.
`
`ENZYME-LABILE LINKERS
`Chemically labile linkers often suffer from limited plasma
`stability. An alternative approach to achieve better control
`of the drug release involves linking a drug to an antibody
`via a peptide linkage. The free drug can be specifically
`cleaved from the carrier by the action of lysosomal proteases
`(e.g., cathepsin or plasmin) present at elevated levels in
`certain tumor tissues (25). Peptidic bonds are expected to
`have good serum stability, as proteases are normally not
`active outside cells because of unfavorable pH conditions
`and inhibition by serum protease inhibitors.
`Seattle Genetics is using a maleimido-containing dipeptide
`linker to conjugate cysteine residues of mAbs with mono-
`methyl auristatin E (MMAE, 7) and F (MMAF),
`two
`auristatin derivatives that bind to tubulin and inhibit micro-
`tubule assembly. A valine-citrulline linker was designed to
`provide high plasma stability (half-lives in mice and monkey
`of 6.0 and 9.6 days, respectively) as well as a cleavage site
`for proteases (26). Efficient hydrolysis of the dipeptide linker
`by cathepsin B has been established in Vitro. After enzymatic
`
`Chart 5
`
`cleavage, 1,6-elimination of the strongly electron-donating
`4-aminobenzyl group will occur, releasing MMAE in its
`active form (Scheme 1). This self-eliminating spacer is
`needed to spatially separate the drug from the site of
`enzymatic cleavage (27). Otherwise, the bulky payload has
`a negative influence on the kinetics of the peptide hydrolysis,
`as the access to the enzyme’s active site can be affected.
`Comparison of plasma stabilities of ADCs bearing enzyme-
`labile valine-citrulline or phenylalanine-lysine linkers (6 and
`8, respectively) with ADC 9 equipped with a chemically
`labile hydrazone linker connecting MMAE with the cBR96
`mAb showed impressively the advantage of the peptide
`linkage (Table 1): in human plasma, the ADC 6, featuring a
`peptidic valine-citrulline-derived linker, exhibits an almost
`100-fold increased stability compared to its hydrazone relative
`9 and the in Vitro specificity was increased (26). The peptidic
`linker has a half-life of 6 days in mice compared to 2 days
`for the hydrazone linker (3-fold improvement), resulting in
`a lower in ViVo toxicity (28). Several ADCs incorporating a
`peptidic linker are in clinical development, with the most
`advanced being cAC10-vcMMAE (SGN-35)
`for CD30-
`positive hematologic malignancies (26, 29, 30).
`In addition to N-terminus linked MMAE and MMAF,
`attachment of dipeptide linkers at the C-terminus was studied
`as a way to vary the rate of drug release and thus ADC potency
`and tolerability (26, 31). For MMAF, it could be shown that
`linking the mAb over a dipeptide linker to the C-terminus of
`the drug can lead to further improvement of the therapeutic
`window.
`The superiority of enzyme-labile linkers was also shown for
`doxorubicin derivatives conjugated to mAbs (11, 32). With this
`
`IMMUNOGEN 2128, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Reviews
`
`Bioconjugate Chem., Vol. 21, No. 1, 2010 9
`
`Scheme 2. Presumed Release Mechanism of CBI-Containing Drug 18 from ADC 17
`
`strategy, the immunological specificity of BR96 mAb-peptide-
`doxorubicin conjugates 10a and 10b (Chart 4) was significantly
`improved compared to that of the corresponding hydrazone-
`based conjugate 1 (11), demonstrating again the importance of
`the linker structure for the ADC properties.
`Peptide-containing linkers have been successfully used with
`CC-1065-derived toxins. CC-1065, duocarmycins, and other
`cyclopropapyrroloind-4-one (CPI, 11; Chart 5) derivatives are
`potent minor-groove binding DNA alkylating agents. According
`to Boger et al., cyclopropabenzindol-4-one analogues (CBI, 12)
`are chemically more stable, biologically more potent, and
`synthetically more accessible than their parent compounds
`incorporating the natural CPI alkylation subunit (33). In analogy
`to a duocarmycin prodrug (34), the seco-CBI precursor 13 was
`developed. This compound, which was postulated to show
`improved pharmacokinetics, will provide the CBI analogue 12
`via Winstein cyclization (35). A further improvement was
`achieved by protecting the phenolic hydroxyl group with an
`enzyme-labile carbamate, as in derivative 14, decreasing the
`toxicity of the prodrug while increasing its water solubility by
`introduction of an additional basic nitrogen (36).
`Using their UltiMAb technology, Medarex conjugated com-
`pound 14 to a mAb via a peptidic linker (37). Conceptually,
`this approach is a double prodrug strategy. It does not only
`improve the drug’s therapeutic efficacy by conjugating it to a
`tumor-specific mAb, but additionally employs a prodrug of the
`cytotoxic agent that has reduced potency and thus minimizes
`side effects before being released intracellularly by carboxyes-
`terases. The two activating steps needed to trigger the release
`of the active drug (Scheme 2) may lead to an enlarged
`therapeutic window.
`Tietze et al. have reported that the CBI prodrugs 15 and
`16 exhibit lower toxicities due to the reduced reactivity of
`the secondary chloride compared to the primary chloride
`in 13 and 14 (39-41). Such derivatives with
`present
`secondary chlorides (see, for example, ADC 19; Chart 6)
`are used in Syntarga’s SpaceLink conjugation technology
`(42). To our knowledge, no data are yet published on the
`linker stability.
`To extend its ADC technology to drugs with a comple-
`mentary mode of action, Seattle Genetics has developed
`
`ADCs 20 and 21 (Chart 6) containing an amino-CBI and a
`hydroxy aza-CBI payload, respectively (43). Because of the
`hydrophobicity of this class of molecules, attention was
`focused on developing hydrophilic peptide-linker derivatives
`that prevented aggregation. This was achieved by use of a
`more hydrophilic valine-lysine sequence (compared to valine-
`citrulline as in ADC 6) and incorporating a tetra(ethylene
`glycol) unit between the mAb and the peptide linker. The
`direct attachment of the linker to the amine of the CBI
`building block (as in 20) or over a self-immolative p-
`aminobenzyl spacer to the corresponding hydroxyl group of
`the aza-CBI toxin (as in 21) prohibits the spontaneous
`formation of the active toxin in the plasma. Only the
`enzymatic cleavage of the linker after internalization of the
`ADC into the cancer cell triggers the release of the prodrug
`that can now be transformed into the DNA-alkylating
`cyclopropyl derivative via a Winstein cyclization.
`Another type of enzyme-labile linker was recently reported
`by Seattle Genetics (44). In this case, a carbohydrate moiety
`serves as the cleavage site. Its hydrophilic nature should limit
`aggregation. Interestingly, the (cid:2)-glucoronide moiety in ADC
`22 is not directly linking the ADC with the payload. Neverthe-
`less, cleavage by (cid:2)-glucuronidase triggers the 1,6-elimination
`of the spacer liberating the free drug 23 (Scheme 3). This
`carbohydrate motif was also used in ADCs bearing Camptoth-
`ecin analogues (45).
`
`NONCLEAVABLE LINKERS
`
`ImmunoGen has reported the fortuitous discovery of a
`noncleavable thioether bond to link DM1 to mAbs (22, 46).
`Derivative 24 was actually prepared as a control experiment,
`but surprisingly, this noncleavable thioether linked ADC also
`proved very potent. The release mechanism is believed to
`occur via internalization of the ADC followed by degradation
`of the mAb component in the lysosome, resulting in the re-
`lease of the maytansinoid drug 25 still attached via the linker
`to a lysine residue. This chemical modification of the drug
`did not diminish its cytotoxic potential. This form of the drug
`is, however, charged and presumably not able to diffuse into
`neighboring cells. Hence, it cannot kill adjacent tumor cells
`
`IMMUNOGEN 2128, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`10 Bioconjugate Chem., Vol. 21, No. 1, 2010
`
`Chart 6
`
`Ducry and Stump
`
`that do not express the target antigen (antigen-negative cells).
`Besides this by-stander effect (23), there is an additional
`reason why cleavable linkers may have a broader efficacy.
`Although internalization is often the common initial activation
`process for both cleavable and noncleavable linkers (22),
`ADCs with cleavable linkers may also be active even when
`they are poorly internalized (47). One possible explanation
`is that, when the linker is cleaved in extracellular space, the
`free drug subsequently permeates the cell to reach its target.
`This mechanism is not operating for ADCs incorporating
`noncleavable linkers, which must be internalized and de-
`graded within the cell
`to become active. They are thus
`dependent on the biology of the target cell.
`The in ViVo stability of the thioether-linked ADC 24
`proved, with a half-life of 134 h in mice, superior to that of
`compounds 4 and 5 (47 and 102 h, respectively). ADCs with
`
`noncleavable linkers were found to be better tolerated (47),
`presumably resulting in an improved therapeutic index. The
`use of noncleavable linkers has thus become an important
`feature of
`ImmunoGen TAP (tumor-activated prodrug)
`conjugation technology. This is best exemplified by the
`promising clinical results obtained by Genentech’s trastu-
`zumab-DM1 (T-DM1) for HER2-positive metastatic breast
`cancer (48).
`Seattle Genetics made the same finding with MMAF,
`namely, that the cleavable dipeptide linker (as in 6) could
`be replaced by a noncleavable thioether bond (as in ADC
`26; Chart 7) while retaining the in ViVo activity (49). Mass
`spectrometry showed that the released drug was a cysteine-
`adduct of MMAF, presumably resulting from mAb degrada-
`tion within lysosomes as proposed for maytansinoid conjugate
`24. MMAE, although closely related structurally, was not
`
`Chart 7
`
`IMMUNOGEN 2128, pg. 6
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Reviews
`
`Bioconjugate Chem., Vol. 21, No. 1, 2010 11
`
`Scheme 3. Postulated Release Mechanism of CBI-Containing Drug 23 from ADC 22
`
`Scheme 4. Release of Maytansine Drug-Derivative 25 through
`Lysosomal Degradation of ADC 24
`
`active when attached in this manner. Only MMAF can sustain
`significant modification to the N-terminal position and still
`retain its potency. Recently, it was shown that a slow transfer
`of the drug from the ADC to albumin cysteines in the plasma
`occurs with alkyl-maleimide ADCs (50). To counteract this
`effect, the linker stability was enhanced by replacing the
`maleimide (as in ADC 26) with an acetamide functionality
`(as in ADC 27). This type of ADC led to no apparent
`degradation in the bloodstream for up to 14 days.
`
`CONCLUSION
`A promising approach to improve the potency of antibody-
`based cancer therapies is to conjugate the mAb with cytotoxic
`chemotherapeutic drugs. In the foreseeable future, ADCs will
`eventually fulfill the promise of specific delivery of cytotoxic
`drugs to tumor cells (targeted therapy), thus avoiding the
`dose-limiting toxicity of chemotherapy that occurs as a result
`of
`its effects on normal cells. The linker
`technology
`profoundly impacts ADC potency, specificity, and safety.
`Early ADC linkers were commonly derived from acid-labile
`hydrazones, which were designed to be cleaved inside target
`cancer cells, but inevitably underwent cleavage at nontarget
`sites. Newer linkers feature hindered disulfides and peptidic
`
`moieties that have achieved greater in ViVo stability. To spare
`nontarget tissues from chemotherapeutic damage, noncleav-
`able linkers were recently developed. They are thus far the
`most stable linkers used in ADCs. This recent development
`may shift some of the future challenge in ADC development
`from the linker technology toward finding cytotoxic drugs
`that retain their potency when bearing a covalently attached
`linker. The improved linker stability and reduced toxicity,
`however, come together with greater target dependence than
`in the case of cleavable linkers. ADCs with noncleavable
`linkers must be internalized and degraded within the cell,
`whereas compounds with cleavable linkers may be active
`against targets that are poorly internalized through extracel-
`lular drug release and drug entry into tumor cells. Similarly,
`killing of bystander antigen-negative cells through targeting
`of antigen-positive cells (collateral toxicity) is presumably
`only possible with cleavable linkers. The Lonza ADC-team
`thus believes that no general linker design exists for all ADCs
`or cancer types. Despite significant progress, each mAb must
`be examined separately in order to optimize a specific
`delivery system.
`
`LITERATURE CITED
`
`(1) Kratz, F., Mu¨ller, I. A., Ryppa, C., and Warnecke, A. (2008)
`Prodrug strategies in anticancer chemotherapy. ChemMedChem
`3, 20–53.
`(2) Panchal, R. G. (1998) Novel therapeutic strategies to selectively
`kill cancer cells. Biochem. Pharmacol. 55, 247–252.
`(3) Zangemeister-Wittke, U. (2005) Antibodies for targeted cancer
`therapy - technical aspects and clinical perspectives. Pathobiology
`72, 279–286.
`(4) Monoclonal antibodies conjugated with radioisotopes do not have
`to release their payload at the target site and need thus completely
`different linkers that mAb conjugated with cytotoxins. They are not
`reviewed in this article. For a review, see Wu, A. M., and Senter,
`P. D. (2005) Arming antibodies: Prospects and challenges for
`immunoconjugates. Nat. Biotechnol. 23, 1137–1146.
`(5) Ricart, A. D., and Tolcher, A. W. (2007) Technology insight:
`Cytotoxic drug immunoconjugates for cancer therapy. Nat. Clin.
`Pract. Oncol. 4, 245–255.
`(6) Carter, P. J., and Senter, P. D. (2008) Antibody-drug conjugates
`for cancer therapy. Cancer J. 14, 154–169.
`(7) Rohrer, T. (2008) New weapons for the oncology arsenal.
`Chem. Rundschau 9, 26–28.
`(8) Hamblett, K. J., Senter, P. D., Chace, D. F., Sun, M. M. C.,
`Lenox, J., Cerveny, C. G., Kissler, K. M., Bernhardt, S. X.,
`Kopcha, A. K., Zabinski, R. F., Meyer, D. L., and Francisco,
`J. A. (2004) Effects of drug loading on the antitumor activity of
`
`IMMUNOGEN 2128, pg. 7
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`12 Bioconjugate Chem., Vol. 21, No. 1, 2010
`
`Ducry and Stump
`
`a monoclonal antibody drug conjugate. Clin. Cancer Res. 10,
`7063–7070.
`(9) McDonagh, C. F., Turcott, E., Westendorf, L., Webster, J. B.,
`Alley, S. C., Kim, K., Andreyka, J., Stone, I., Hamblett, K. J.,
`Francisco, J. A., and Carter, P. (2006) Engineered antibody-drug
`conjugates with defined sites and stoichiometries of drug
`attachment. Prot. Eng. Des. Sel. 19, 299–307.
`(10) Trail, P. A., Willner, D., Lasch, S. J., Henderson, A. J.,
`Hofstead, S., Casazza, A. M., Firestone, R. A., Hellstro¨m, I.,
`and Hellstro¨m, K. E. (1993) Cure of xenografted human
`carcinomas by BR96-doxorubicin immunoconjugates. Science
`261, 212–215.
`(11) Dubowchik, G. M., Firestone, R. A., Padilla, L., Willner, D.,
`Hofstead, S. J., Mosure, K., Knipe, J. O., Lasch, S. J., and Trail,
`P. A. (2002) Cathepsin B-labile dipeptide linkers for lysosomal
`release of doxorubicin from internalizing immunoconjugates:
`Model studies of enzymatic drug release and antigen specific in
`Vitro anticancer activity. Bioconjugate Chem. 13, 855–869.
`(12) Hamann, P. R., Hinman, L. M., Beyer, C. F., Lindh, D.,
`Upeslacis, J., Flowers, D. A., and Bernstein, I. (2002) An anti-
`CD33 antibody-calicheamicin conjugate for treatment of acute
`myeloid leukaemia. Choice of linker. Bioconjugate Chem. 13,
`40–46.
`(13) Hamann, P. R., Hinman, L. M., Hollander, I., Beyer, C. F.,
`Lindh, D., Holcomb, R., Hallett, W., Tsou, H.-R., Upeslacis, J.,
`Shochat, D., Mountain, A., Flowers, D. A., and Bernstein, I.
`(2002) Gemtuzumab ozogamicin, a potent and selective anti-
`CD33 antibody-calicheamicin conjugate for treatment of acute
`myeloid leukemia. Bioconjugate Chem. 13, 47–58.
`(14) Hollander, I., Kunz, A., and Hamann, P. R. (2008) Selection
`of reaction additives used in the preparation of monomeric
`antibody-calicheamicin conjugates. Bioconjugate Chem. 19, 358–
`361.
`(15) van der Velden, V. H. J., te Marvelde, J. G., Hoogeveen, P. G.,
`Bernstein, I. D., Houtsmuller, A. B., Berger, M. S., and van
`Dongen, J. J. M. (2001) Targeting of the CD33-calicheamicin
`immunoconjugate Mylotarg (CMA-676) in acute myeloid leu-
`kemia: in ViVo and in Vitro saturation and internalization by
`leukemic and normal myeloid cells. Blood 97, 3197–3204.
`(16) Hamann, P. R., Hinman, L. M., Beyer, C. F., Lindh, D.,
`Upeslacis, J., Shochat, D., and Mountain, A. (2005) A cali-
`cheamicin conjugate with a fully humanized anti-MUC1 antibody
`shows potent antitumor effects in breast and ovarian tumor
`xenografts. Bioconjugate Chem. 16, 354–360.
`(17) DiJoseph, J. F., Dougher, M. M., Kalyandrug, L. B., Armel-
`lino, D. C., Boghaert, E. R., Hamann, P. R., Moran, J. K., and
`Damle, N. K. (2006) Antitumor efficacy of a combination of
`CMC-544 (inotuzumab ozogamicin), a CD22-targeted cytotoxic
`immunoconjugate of calicheamicin, and rituximab against non-
`Hodgkin‘s B-cell lymphoma. Clin. Cancer Res. 12, 242–249.
`(18) DiJoseph, J. F., Khandke, K., Dougher, M. M., Evans, D. Y.,
`Armellino, D. C., Hamann, P. R., and Damle, N. K. (2008) CMC-
`544 (inotuzumab ozogamicin): A CD22-targeted immunocon-
`jugate of calicheamicin. Hematology Meeting Reports 5, 74–77.
`(19) Ojima, I., Geng, X., Wu, X., Qu, C., Borella, C. P., Xie, H.,
`Wilhelm, S. D., Leece, B. A., Bartle, L. M., Goldmacher, V. S.,
`and Chari, R. V. J.
`(2002) Tumor-specific novel
`taxoid-
`monoclonal antibody conjugates. J. Med. Chem. 45, 5620–5623.
`(20) Ranson, M., and Sliwkowski, M. X. (2002) Perspectives on
`anti-HER monoclonal antibodies. Oncology 63 (suppl. 1), 17–
`24.
`(21) Widdison, W. C., Wilhelm, S. D., Cavanagh, E. E., Whiteman,
`K. R., Leece, B. A., Kovtun, Y., Goldmacher, V. S., Xie, H.,
`Steeves, R. M., Lutz, R. J., Zhao, R., Wang, L., Bla¨ttler, W. A.,
`and Chari, R. V. J. (2006) Semisynthetic maytansine analogues
`for the targeted treatment of cancer. J. Med. Chem. 49, 4392–
`4408.
`(22) Erickson, H. K., Park, P. U., Widdison, W. C., Kovtun, Y. V.,
`Garrett, L. M., Hoffman, K., Lutz, R. J., Goldmacher, V. S.,
`and Bla¨ttler, W. A. (2006) Antibody-Maytansinoid conjugates
`are activated in targeted cancer cells by lysosomal degradation
`
`and linker-dependent intracellular processing. Cancer Res. 66,
`4426–4433.
`(23) Kovtun, Y. V., Audette, C. A., Ye, Y., Xie, H., Ruberti, M. F.,
`Phinney, S. J., Leece, B. A., Chittenden, T., Bla¨ttler, W. A., and
`Goldmacher, V. S. (2006) Antibody-drug conjugates designed
`to eradicate tumors with homogeneous and heterogeneous
`expression of the target antigen. Cancer Res. 66, 3214–3221.
`(24) Xie, H., Audette, C., Hoffee, M., Lambert, J. M., and Bla¨ttler,
`W. A. (2004) Pharmacokinetics and biodistribution of the
`antitumor immunoconjugate, cantuzumab mertansine (huC242-
`DM1), and its two components in mice. J. Pharmacol. Exp. Ther.
`308, 1073–1082.
`(25) Koblinski, J. E., Ahram, M., and Sloane, B. F. (2000)
`Unraveling the role of proteases in cancer. Clin. Chim. Acta 291,
`113–135.
`(26) Doronina, S. O., Toki, B. E., Torgov, M. Y., Mendelsohn,
`B. A., Cerveny, C. G., Chace, D. F., DeBlanc, R. L., Gearing,
`R. P., Bovee,