`Society
`of Clinical
`Oncology
`Educational
`Book
`
`Fall Education Conf ere nee
`Fall 1998
`
`EDITOR
`Michael C. Perry, MD,
`Columbia, MO
`
`Editorial Projects
`Coordinator
`Rich Harrington,
`Chestnut Hill, MA
`
`© 1998 American
`Society
`of Clinical Oncology
`
`Alexandria, VA
`
`
`
`-
`
`1998 FALL EDUCATION CONFERENCE SUPPORT
`The American Society of Clinical Oncology gratefully acknowledges the
`following companies for their unrestricted educational grants in support
`of the Fall Education Conference:
`
`Amgen
`Berlex Laboratories
`Bristol-Myers Squibb
`Genentech, Inc.
`Medtronic Neurological
`Ortho Biotech, Inc
`
`Please note, the above indicates support committed as of October 10, 1998.
`
`
`
`New Agents-Noncytotoxins, Antiangiogenesis
`
`By Edward A. Sausville, MD, PhD, Session Chair
`
`Abstract: Traditional cytotoxic agents, which cause
`death of tumor cells in tissue culture or in vivo animal
`models, have been the mainstay of agents available
`for clinical trials and have led to useful and relatively
`safe agents for clinical use. However, the continuing
`improvement of our understanding of the biochemical
`pathways that give rise to the malignant phenotype
`have defined new targets for antineoplastic agents
`that may have a cytostatlc, rather than cytotoxic, end
`point. The hope is that such agents may have a greatly
`improved therapeutic index, allowing their long-term
`and perhaps indefinite use, These agents would be of
`
`T HIS SESSION WILL EXPLORE initial ef(cid:173)
`
`forts to define the clinical value of noncyto(cid:173)
`toxic approaches to cancer therapeutics. Current
`cancer drugs that have proven safe and effective
`in a way that leads to regulatory approval are, in
`large measure, cytotoxic compounds, the activity
`of which was fh·st detected in tumor screening
`systems in vivo, such as the murine leukemias,
`L1210 and P388, or murine solid tumors. Yet,
`these drugs use a relatively limited repertoire of
`mechanisms, for the most part targeting either
`some aspect of DNA (eg, structural integrity or
`synthesis) or tubulin. These agents have a set of
`stereotypical side effects, including leukopenia
`and mucositis. Moreover, it is increasingly recog(cid:173)
`nized that the success of these agents in causing
`cell death does not directly relate to their affecting
`an immediate target, eg, topoisomerases or nucle(cid:173)
`otide-synthesizing systems, but rather in induc(cid:173)
`ing the activity of cell death pathways, the action
`of which leads to apoptosis or programmed cell
`death. In that regard, interaction with traditional
`cancer drug targets is necessary, but not sufficient
`for a useful effect of these agents.
`There is no question that clinically used, tradi(cid:173)
`tional cytotoxic agents have saved lives either
`through their direct use in advanced disease or
`through adjuvant treatment programs after pri-
`
`Addres,9 reprint requests to Edward A. Sausville, MD, PhD,
`Developmental Therapeutics Program, Diuision o(Cancer TI·eat(cid:173)
`ment and Diagnosis, National Cancer Institute, Executive
`Plaza North, Suite 843, Bethesda, MD 20892; Email
`sausville@dtpax2.neifcrfgov.
`This is a US government work. There are no restrictions on
`its use,
`1092-9118/98/2-17$0.00I0
`
`112
`
`particular value in the treatment of disease diagnosed
`by molecular markers in a tissue at risk or as an
`adjuvant treatment after primary ablative therapy.
`Inhibitors of signal transduction pathways and antian•
`giogenic agents are two such categories of agent that
`do not have primary cytotoxiclty for the tumor cell.
`17-Allylamino 17-demethoxygeldanamycin (17AAG;
`NSC 330507) and rapamycin analog WAY-130779 (NSC
`683864) represent novel signal transduction inhibitors
`that will enter or have just entered clinical trial, which
`illustrate the thinking that encourages clinical trials
`with this type of agent.
`
`mary therapy. However, increasing definition of
`the regulatory pathways perturbed in tumor cells
`has raised the question of whether specific target(cid:173)
`ing of these deregulated pathways would define
`agents that do not necessarily result in tumor-cell
`death, but decrease the rate of tumor-cell prolifera(cid:173)
`tion in a way that will be clinically useful. An
`additional hope is that agents directed at these
`deregulated pathways will have an altered spec(cid:173)
`trum of toxicities, with ideally little toxicity for
`nonproliferating normal cells. Tolerable or essen(cid:173)
`tially nonexistent toxicity when administered for
`protracted periods, perhaps chronically, will be a
`necessary feature if these molecules are to achieve
`their broadest utility. A differing school of thought
`would hold that cytostatic agents of the various
`types to be described later may actually magnify
`or augment the action of cytotoxic drugs by dimin(cid:173)
`ishing the capacity of the cell to repair damage
`evoked by cytotoxic agents. This point of view
`would propose that cytostatic agents might be
`best intermittently used as modulators to en(cid:173)
`hance the action of cytotoxic agents.
`Noncytotoxic agents under active development
`have the common property of generally not caus(cid:173)
`ing overt tumor-cell kill after brief exposure of
`cells in conventional screening assays for growth
`inhibition in vitro. This is in contrast to classical
`cytotoxic agents, for which a relatively brief expo(cid:173)
`sure is frequently sufficient to evoke significant
`cell kill manifest, for example, in 18 to 24 hours
`after drug exposure. In many cases, brief exposure
`to noncytotoxic agents may lead to protracted
`cytostasis, which can in some cases be reversed by
`drug removal even after several hours or days in
`cell culture. N oncytotoxic agents include those
`
`
`
`NONCYTOTOXIC AGENTS
`
`that interrupt the propagation of growth factor(cid:173)
`related signals, which result in a block to cell(cid:173)
`cycle progression or entry into the cell cycle;
`differentiation agents, which induce gene activity
`that causes cells to exit from the cell cycle, with or
`without apoptosis; antimetastatic agents, in(cid:173)
`tended to target tumor-cell motility, invasion, and
`the appropriate microenvironment for growth;
`and antiangiogenic agents, which attack not the
`tumor cells directly, but the supporting network of
`blood vessels necessary for progression of a tumor
`from a microscopic to macroscopic growth.
`The presentations to be focused on in this
`session will emphasize signal transduction block(cid:173)
`ers and antiangiogenic agents. Examples of agents
`early in clinical development will be presented in
`the remainder of this section to illustrate features
`common to thinking about signal transduction
`blockers. A subsequent module by Dr Michael
`Cooper will contain an update of different types of
`signal transduction blockers that have proceeded
`to more advanced clinical trials. The final module
`by Dr James Pluda will address recent approaches
`to inhibition of tumor endothelial cell prolifera(cid:173)
`tion. Both modules will touch on the difficulty in
`designing clinical trials that efficiently capture
`valuable clinical evidence of cytostatic activity.
`Figure 1 presents a greatly simplified view of
`the controls operating to regulate cell prolifera(cid:173)
`tion. Growth factor receptors respond to external
`stimuli to activate intermediary signal transduc(cid:173)
`tion cascades. Among the best characterized sys(cid:173)
`tems of growth factor receptors are the tyrosine
`kinase-activating receptors, which include epider(cid:173)
`mal growth factor receptor (EGF-R; c-erbBI) and
`c-erbB2. Transduction of signals results in expres(cid:173)
`sion of genes critical to entry into the cell cycle,
`such as the myc family genes and cyclin D gene
`products. The latter activate the cyclin-dependent
`kinase (CDK) family of cell-cycle regulatory en(cid:173)
`zymes CDKs 4 or 6 and CDK2 to promote progres(cid:173)
`sion through Gl and entry into S phase. The
`action of endogenous inhibitors of CDK action
`(p16 and p21 families) diminishes. Phosphoryla(cid:173)
`tion of the tumor-suppressor gene pRB, first de(cid:173)
`tected as the basis for hereditary retinoblastoma,
`occurs. Following pRB phosphorylation, the E2F
`family of transcription factors is activated, which
`leads to elaboration of enzymes necessary for
`completion of DNA synthesis, ultimately followed
`
`f
`
`i
`I
`l
`l
`l
`f
`
`• •
`~ <~.=A~P ' I Pl•JK/PBS
`'._ Pldln>(4,5)P2 -<
`
`113
`
`'
`
`lnsP3
`
`DAG
`
`a
`
`r
`
`\., ~¢" &iJI
`~ r (§)
`
`/
`
`-
`
`Pldln<(J)P
`
`(MAPK-K)
`\..
`8-G0-G
`1\ 7'
`
`NUCLEUS
`
`Jun
`
`Myc
`
`Fas
`SFIF
`
`-'ii>- TRANSCAIPTIO~
`p(i2TCF
`
`Fig 1. Growth factor-mediated signals targeted by cyto(cid:173)
`static agents. In this simplified depiction, growth factors
`combine with their cognate receptors and cause dimerization,
`with activation of protein tyrosine kinase activity. The phos(cid:173)
`phorylated tyrosine kinase receptor recruits adaptor mol(cid:173)
`ecules, including she, and grb2, to activate ras through
`influence on the GTPase activator proteins (GAPs). Ras in turn
`directly interacts with raf, to augment gene activity through
`the MAP kinase pathway (MAPK-K, MAPKI. Phosphorylation
`on tyrosine of activated receptors activates phospholipase C,
`resulting in calcium mobilization by inositol tris phosphate
`(lnsP3) and the potential for protein kinase C activation by
`diacylglycerol (DAG). Note that additional types of receptors
`not depicted here, eg, 7-transmembrane receptor-G-protein(cid:173)
`coupled receptors can activate calcium mobilization by a
`distinct phospholipase.
`
`by mitosis after the action of CDKl, which trig(cid:173)
`gers entry into mitosis.
`The importance of this pathway to tumorigen(cid:173)
`esis is illustrated by the fact that substantial
`numbers of patients with common adult malignan(cid:173)
`cies have lesions in this pathway. For example,
`substantial fractions of breast and ovarian tumors
`overexpress c-erbB2; 80% of small-cell lung tu(cid:173)
`mors have inactivation of pRB, and of those that
`do not, a high frequency of deletions of the endog(cid:173)
`enous inhibitors to cell-cycle progression can be
`demonstrated. Thus, agents that would return
`these altered pathophysiologic features of tumor
`cells toward normal are attractive in that they are
`addressing the pathophysiologic basis for tumor(cid:173)
`cell growth.
`
`17-ALLYLAMINO
`17-DEMETHOXYGELDANAMYCIN
`This agent is a derivative of the benzoquinoid
`ansamycin, geldanamycin. The parent compound,
`geldanamycin, is similar to the related ansamy-
`
`
`
`114
`
`cin, herbimycin (Fig 2), in its ability to cause
`reversion of fibroblasts transformed by the viral
`oncogene v-src to a more normal phenotype under
`noncytotoxic conditions. 1 This feature was ex(cid:173)
`tended to include cells transformed by a number
`of tyrosine kinase oncogenes, including c-erbB2.
`Since the content of tyrosine phosphates was
`decreased in drug-treated cells, both herbimycin
`and geldanamycin were used for many years as
`nonspecific inhibitors of tyrosine kinases in labora(cid:173)
`tory experiments. However, efforts to define direct
`inhibition of tyrosine kinases in biochemical reac(cid:173)
`tions outside of cells failed to demonstrate potent
`tyrosine kinase inhibitory capacity. Instead, closer
`scrutiny of the effects of geldanamycin in living
`cells showed that the reason for apparent diminu(cid:173)
`tion of tyrosine kinase activity was that the actual
`mass of several tyrosine kinases was decreased in
`drug-treated cells, and subsequent experiments
`demonstrated accelerated turnover with degrada(cid:173)
`tion of newly synthesized tyrosine kinases. 2·4
`The molecular basis for this behavior became
`apparent when Whitesell et al5 demonstrated that
`geldanamycin actually formed a high-affinity com(cid:173)
`plex with the heat-shock protein (hsp) 90, and
`thus, inhibited the formation of complexes be(cid:173)
`tween hsp and the oncoproteins, including the src
`family oncoproteins. Hsps of various molecular
`weights had long been defined as cellular proteins
`whose level increased in response to a variety of
`
`EDWARD A. SAUSVILLE
`
`nonspecific stresses, including heat, osmotic, toxic,
`and other stresses. The hsp90 family act physi(cid:173)
`ologically as "chaperones" to effect proper folding
`and cellular localization of tyrosine kinases. Thus,
`binding of geldanamycin to hsp90 interferes with
`the proper localization and folding of the oncopro(cid:173)
`tein. The improperly chaperoned tyrosine kinase
`is rapidly degraded, with loss of its signaling
`function. Interestingly, hsp90 forms complexes
`not only with tyrosine kinases, but a variety of
`other molecules important to tumorigenesis, in(cid:173)
`cluding c-raf (an intermediary kinase in the MAP
`kinase signaling pathway), several steroid hor(cid:173)
`mone receptors, and the nuclear oncoprotein, p53.
`Thus, geldanamycin can potentially interfere with
`several signaling systems at once by altering the
`normal association of these molecules with their
`chaperones.
`Initial efforts to define antitumor activity on the
`part of geldanamycin met with, at best, limited
`success, as the drug was toxic at doses that
`afforded little or no efficacy. Efforts to develop
`approaches to treat large animals with geldanamy(cid:173)
`cin failed due to hepatic toxicity. Consideration of
`several analogs led to 17-allylamino 17-demthoxy(cid:173)
`geldanamycin (17AAG; NSC 330507). This drug
`has shown consistent evidence of activity, eg,
`when administered to the MEXF 276 melanoma
`model, T/C (ratio of tumor weight in treated as
`compared with control animals) of 6% to 12% were
`
`ll1CO
`
`Herbimycin
`
`Geldanamycin Ring System
`
`2. Geldanamycin,
`Fig
`17AAG, and herbimycin.
`
`A
`
`0
`11,co~I,_
`~ - -
`
`0
`
`R
`
`H,Nc(o:
`
`Geldanamycin
`NSC 122750
`
`A
`II
`~H~
`
`0
`
`Q
`
`R
`
`ll2NC(O)-
`
`17-allylamino, 17-demethoxy
`geldanamycin(17 AAG)
`NSC 330507
`
`
`
`NONCYTOT0XIC AGENTS
`
`observed on twice-daily for 5-day treatments,
`administered as two courses (E. Sausville, H.
`Fiebig, A. Burger, unpublished results, April 1998).
`However, these are at best tumor growth delays.
`In contrast to geldanamycin, 17 AAG can be given
`to larger animals with less, although not absent,
`hepatotoxicity.
`It is certainly not clear that 17 AAG is a practi(cid:173)
`cal therapeutic agent in humans. Initial clinical
`trials will be possible in late 1998 or early 1999.
`1 7 AAG will allow the clinical research community
`to approach the important scientific question of
`whether hsp90 can be targeted in humans safely,
`as changes in hsp90 level or oncoprotein kinase
`level or activity will be important end points of
`initial clinical trials, in addition to the usual end
`points of toxicity and pharmacology. In this re(cid:173)
`spect, 17 AAG illustrates another important prin(cid:173)
`ciple in considering the development of cytostatic
`agents, namely, the need to define an early marker
`of biologic effect of the drug different than the
`usual criteria of tumor shrinkage. While re(cid:173)
`sponses in traditional terms of complete and
`partial responses (CR and PR, respectively) would
`be welcome if observed, the clinical trials commu(cid:173)
`nity must be open to designing efficient early
`phase studies that will capture evidence of a
`potentially valuable biologic effect mechanisti-
`
`115
`
`cally relatable to the action of the drug. In the case
`of 1 7 AAG, hsp90 and its client molecules will be of
`importance to follow prospectively in tumor or
`surrogate tissues early in clinical trials.
`
`RAPAMYCIN ANALOG WAY-130779
`Rapamycin is a drug at an advanced stage of
`clinical development as an immunosuppressant.6
`However, it has long been known to possess
`antitumor activity in a variety of murine and
`human tumor systems.7 Best activity was ob(cid:173)
`served in murine models using parenteral routes
`of administration. Therefore, definition of an eco(cid:173)
`nomical, reliable parenteral form of the molecule
`required careful comparison of a variety of ana(cid:173)
`logs. WAY-130779 (NSC 683864; Fig 3) had the
`desired physicochemical properties, and as shown
`in Table 1, demonstrates easily definable tumor
`growth delays at several nontoxic dose levels in,
`for example, the prostate xenograft model PC3.
`Additional preclinical activity has been ob(cid:173)
`served in glioma, breast carcinoma, and lym(cid:173)
`phoma models. Interestingly, while the parent
`compound's immunosuppressant effects are mani(cid:173)
`fest and indeed require continuous administra(cid:173)
`tion schedules, antitumor activity is manifest on
`an intermittent schedule, where one would antici(cid:173)
`pate little persisting effect on the immune system.
`
`Rapamycln
`NSC 226080
`
`CCl•TT9
`NSC 683884
`
`Fig 3. Rapamycins.
`
`
`
`116
`
`Table 1. Activity of WAY-130779 Against PC3 Prostate Cancer
`Xenografts (advanced-stage model)
`
`Schedule
`
`%TIC
`
`IP daily x 5
`45 mg/kg
`30 mg/kg
`20 mg/kg
`
`IP every 2 days x 5
`45 mg/kg
`30 mg/kg
`20 mg/kg
`
`25
`28
`43
`
`45
`37
`47
`
`Start treatment day 12, assess on day 25
`
`Abbreviation: IP, intraperitoneal.
`
`Nonetheless, the antitumor effect in conventional
`terms is rather modest, and as with other cyto(cid:173)
`static molecules there is little evidence of overt
`tumor-cell kill.
`Rapamycin8 congeners can affect cell growth by
`affecting at least two important intracellular tar(cid:173)
`gets. The first is the PHAS family of translational
`repressors. Rapamycins are better considered pro(cid:173)
`drugs for a complex formed between the respec(cid:173)
`tive rapamycin and FKBP12, 9 a family of intracel(cid:173)
`lular adaptor molecules originally defined by their
`ability to bind to the structurally related immuno(cid:173)
`suppressant FK506. Once bound to FKBP12, the
`rapamycin-FKBP12 complex can act efficiently to
`inhibit the action of the kinase mTOR (mamma(cid:173)
`lian target of rapamycin). 8 This results in a de(cid:173)
`creased phosphorylation of PHAS, allowing the
`efficient binding to PHAS of the 4E family of
`translation initiators. Initiation factor 4E binds to
`and promotes the translation of a variety of
`capped RNAs important to progression through
`Gl. When PHAS is phosphorylated, which occurs
`after the activation of a variety of growth factor
`receptors, 4E is freed and binds to mRNA and
`promotes translation. Thus, rapamycins act to
`affect the phosphorylation state of PHAS, ulti(cid:173)
`mately causing G 1 arrest, as critical RN As for
`
`EDWARD A. SAUSVILLE
`
`cell-cycle progression cannot be translated. An
`additional target for rapamycin effect, the molecu(cid:173)
`lar basis of which is not clear at this time, is the
`ribosome protein S6 protein kinase activity. 10 This
`effect results in a decrease in the degree of phos(cid:173)
`phorylation of ribosomal proteins, again with de(cid:173)
`creased efficiency of protein translation. This oc(cid:173)
`curs by a separate mechanism from that effected
`by the PHAS protein. Thus, rapamycin may be
`thought of as decreasing the translation of genes
`critical for cell-cycle progression by a variety of
`mechanisms. Interestingly, both S6 kinase and
`PHAS are prominent effectors of growth factor(cid:173)
`mediated signals, particularly insulin and insulin(cid:173)
`like growth factors.
`Initial clinical trials with WAY-130779 have
`commenced in mid 1998. Critical to the drug's
`development will be early consideration of how it
`might be combined with conventional agents. In
`this way, the beneficial effects of initial cytoreduc(cid:173)
`tion may be extended in time by decreasing the
`efficient progression of remaining cells through
`the cell cycle. As with 17 AAG, efficient demonstra(cid:173)
`tion of drug effect on mTOR, S6 kinases, or their
`substrates will be of importance in assuring that
`the drug is actually having the desired effect at
`the cellular level.
`Both 17AAG and WAY-130779 serve to illus(cid:173)
`trate the emerging concept that effective (at least
`in terms of animal models) cytostatic agents can
`target a number of biochemical pathways or tar(cid:173)
`gets; a balance must be struck between frequency
`of administration and the emergence of undesired
`toxicity (hepatic in the case of 17 AAG and immu(cid:173)
`nosuppression in the case of WAY-130779). While
`conventional tumor models show, at best, modest
`but reproducible growth delays, great interest in
`exploring strategies to optimize their beneficial
`use in humans will hopefully occupy the attention
`of clinicians in the coming years.
`
`REFERENCES
`A: Reduction of total phosphotyrosine levels due to reduced
`1. Uehara Y, Hori M, Takeuchi, et al: Phenotypic change
`kinase activity and increased turnover of pp60v-HrcI. Cancer
`from transformed to normal induced by benzoquinoid ansamy(cid:173)
`cins accompanies inactivation of p60 91'' in rat kidney cells
`Res 49:780-785, 1989
`3, Miller P, Di Orio C, Moyer M, et al: Depletion of the erbB2
`infected with Rous sarcoma virus. Mo! Cell Biol 6:2198-2206,
`1986
`gene product p185 by benzoquinoid ansamycins, Cancer Res
`2. Uehara Y, Murakami Y, Sugimoto Y, et al: Mechanism of
`54:2724-2730, 1994
`reversion of Rous sarcoma virus t1·ansformation by herbimycin
`4, Sepp-Lorenzino L, Ma Z, Lebwohl DE, et al: Herbimycin
`
`
`
`N0NCYTOTOXIC AGENTS
`
`117
`
`A induces the 20S proteasome- and ubiquitin-dependent degra(cid:173)
`dation of receptor tyrosine kinases. J Biol Chem 270:16580-
`16587, 1995
`5. Whitesell L, Mimnaugh EG, De Costa B, et al: Inhibition
`of heat shock protein HSP90-pp6ov-src heteroprotein complex
`formation by benzoquinone ansamycins: Essential role for
`stress proteins in oncogenic transformation. Proc Natl Acad
`Sci USA91:8324-8328, 1994
`6. Sehgal SN, Molnar-Kimber K, Ocain TD, et al: Rapamy(cid:173)
`cin: A novel immunosuppressive macrolide. Med Res Rev
`14:1-22, 1994
`7. Morris RE: Rapamycins: Antifungal, antitumor, antipro-
`
`liferative, and immunosuppressive molecules. Transplant Rev
`6:39-87, 1992
`8. Brunn GJ, Hudson CC, Sekulic A, et al: Phosphorylation
`of the translational repressor PHAS-I by the mammalian
`target ofrapamycin. Science 277:99-101, 1997
`9. Sabers CJ, Martin MM, Brunn GJ, et al: Isolation of a
`protein target of the FKBP12-rapamycin complex in mamma(cid:173)
`lian cells. J Biol Chem 270:815-822, 1995
`10. Chung J, Kuo CJ, Crabtree GR, et al: Rapamycin(cid:173)
`FKBP specifically blocks growth-dependent activation of and
`signalling by the 70 kd S6 protein kinases. Cell 69:1227-1236,
`1992
`
`.E
`
`n
`I•
`e
`8
`
`I•
`
`d
`e
`s
`if
`d
`
`- - - - - - - - - - I i i , ,
`
`