3082
`
`MlNIMALLY IMMUNOGENIC Ab BY ABBREVIATED CDR GRAFTING
`
`Table II. Reactivity of the variant mm with patients' sera”
`
`Competitor Ab
`
`HuCOL—l
`24’25’27L
`61H
`24’25’27L/61H
`
` 3M
`(nM)
`
`6.6
`10.7
`4.9
`9.1
`
`JS
`(nM)
`
`2.0
`2.5
`2.8
`4.1
`
`MB
`(nM)
`
`5.5
`15.2
`10.5
`14.4
`
`0 Competitor Ab concentrations required for IC50 of binding of serum from pa-
`tients EM, JS, and MB to HuCOL-l were calculated.
`
`cols for therapeutic efliciency of the radiolabeled mCOL—l has
`been demonstrated in an animal model (30). In addition to causing
`a possible allergic reaction, the HAMA response results in rapid
`clearance of the Ab from the circulation, preventing the Ab from
`reaching the targeted tumor sites. In a phase I clinical trial of
`131I—labeled mCOL—l, the onset of HAMA response prevented ad—
`ministration of a second dose of the radiolabeled Ab in two pa—
`tients who were otherwise eligible for further treatment. HAMA
`response was also reported in a phase II clinical trial, when IFN
`was combined with 131I—labeled COL—l and 131I—labeled CC49, a
`murine Ab against another pancarcinoma Ag,
`tumor—associated
`glycoprotein—72, to achieve more eificient targeting of human colo—
`rectal cancer (31).
`Development of a mouse—human chimeric mAb is an approach
`that has been widely used to minimize HAMA response (52). Ac—
`cordingly, a cCOL—l mAb was developed by replacing the C re—
`gions of the L and H chains of the mCOL—l with the human K and
`yl constant regions. The Ag—binding affinity of the cCOL—l
`(3.45 X 108 M71) was found to be comparable to that of murine
`Ab (5.17 X 108 M71). A chimeric Ab, though likely to be less
`immunogenic than the murine Ab, may evoke anti—V region re—
`sponse in patients, because the V region of a murine Ab is poten—
`tially immunogenic. Anti—V region responses have been reported
`after the administration of chimeric Abs in patients (53, 54; re—
`viewed in Ref. 55). To attempt
`to reduce this problem, mAb
`COL—1 has been humanized following a procedure that involves
`grafting of the CDRs of a xenogeneic Ab onto the human Ig frame—
`works (reviewed in Ref. 8).
`The most important consideration in humanizing an Ab is the
`preservation of its Ag—binding property, which depends on the
`
` 3
`
`180
`
`”i UGO
`
`10
`
`Com peflmr cancentcatian {13M}
`
`FIGURE 6. Serum reactivity, by SPR, of mCOL—l and the engineered
`Abs derived from it. Increasing concentrations of mCOL—l (I), HuCOL—l
`(A), and 24’25’27L/61H (C) were used to compete with the anti—V region Abs
`to COL—1 present in serum from patient MB for binding to mCOL—l
`im—
`mobilized on a sensor chip. Percent binding of the sera to mCOL—l was
`calculated from the sensograms and plotted as a function of the concen—
`tration of the competitor.
`
`3!.
`
`a}
`E'16
`.5I;
`
`10;} 7777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777
`
`
`
`50 q
`
`so
`
`:50
`
`3’
`
`% b
`
`e 4D
`
`0 .
`'1
`
`7
`1339
`V
`V m
`{tampefitar c-annenimfioa {n M}
`
`1000
`
`
`Serum reactivity, by SPR, of HuCOL—l and its variants.
`FIGURE 5.
`
`
`
`Increasing concentrations of HuCOL—l (A), 24’25’27L (O), 61H (
`), and
`24’25’27L/61H (C) mAbs were used to compete with the anti—V region Abs
`to COL—1 present in sera from patients EM (A), JS (B), and MB (C) for
`binding to HuCOL—l immobilized on a sensor chip. Percent binding of the
`sera to HuCOL—l was calculated from the sensograms and plotted as a
`function of the concentration of the competitor.
`
`of the binding of the sera anti—V region Abs to mCOL—l immobi—
`lized on the sensor chip (Fig. 6). Sera from two other patients, JS
`and EM, were used to compare serum reactivity of mCOL—l and
`HuCOL—l. Although the IC50 value of HuCOL—l was ~6—fold
`higher than that of mCOL—l for JS serum, it was not possible to
`evaluate the difference in the reactivity of the two Abs to EM
`serum; even 1000 nM HuCOL—l was unable to attain 50% inhi—
`bition of the binding of the serum to mCOL—l (data not shown).
`
`Discussion
`The HAMA response is a major impediment to the clinical use of
`the murine mAbs, particularly for the dose fractionation clinical
`protocols that are likely to be used for greater therapeutic eflicacy
`and reduced toxicity. The advantage of dose fractionation proto—
`
`