A phase I study of single administration of antibody-directed enzyme prodrug therapy with the recombinant anti-carcinoembryonic antigen antibody-enzyme fusion protein MFECP1 and a bis-iodo phenol mustard prodrug.

PURPOSE: Antibody-directed enzyme prodrug therapy is a two-stage treatment whereby a tumor-targeted antibody-enzyme complex localizes in tumor for selective conversion of prodrug. The purpose of this study was to establish optimal variables for single administration of MFECP1, a recombinant antibody-enzyme fusion protein of an anti-carcinoembryonic antigen single-chain Fv antibody and the bacterial enzyme carboxypeptidase G2 followed by a bis-iodo phenol mustard prodrug. MFECP1 is manufactured in mannosylated form to facilitate normal tissue elimination. EXPERIMENTAL DESIGN: Pharmacokinetic, biodistribution, and tumor localization studies were used to test the hypothesis that MFECP1 localizes in tumor and clears from normal tissue via the liver. Firstly, safety of MFECP1 and a blood concentration of MFECP1 that would avoid systemic prodrug activation were tested. Secondly, dose escalation of prodrug was done. Thirdly, the dose of MFECP1 and timing of prodrug administration were optimized. RESULTS: MFECP1 was safe and well tolerated, cleared rapidly via the liver, and was less immunogenic than previously used products. Eighty-fold dose escalation from the starting dose of prodrug was carried out before dose-limiting toxicity occurred. Confirmation of the presence of enzyme in tumor and DNA interstrand cross-links indicating prodrug activation were obtained for the optimal dose and time point. A total of 28 of 31 patients was evaluable for response, the best response being a 10% reduction of tumor diameter, and 11 of 28 patients had stable disease. CONCLUSIONS: Optimal conditions for effective therapy were established. A study testing repeat treatment is currently being undertaken.

Potentiation of antitumor immunity by antibody-directed enzyme prodrug therapy.

Antibody-directed enzyme prodrug therapy (ADEPT) has displayed antitumor activity in animal models and clinical trials. We examined whether antitumor immunity is generated during ADEPT by employing an immunoenzyme composed of the monoclonal antibody (MAb) RH1 conjugated to beta-glucuronidase to target rat AS-30D hepatocellular carcinoma tumors. A glucuronide prodrug of p-hydroxyaniline mustard was used to treat malignant ascites after immunoenzyme localization at the cancer cells. ADEPT cured more than 96% of Sprague-Dawley rats bearing advanced malignant ascites, and all cured rats were protected from a lethal challenge of AS-30D cells. Immunization with radiation-killed AS-30D cells or AS-30D cells coated with immunoenzyme did not provide tumor protection. Likewise, ex vivo treatment of tumor cells by ADEPT before injection into rats did not protect against a tumor challenge. AS-30D and N1-S1 hepatocellular carcinoma cells but not unrelated syngeneic tumor cells were lysed by peritoneal exudate cells isolated from ADEPT-cured rats. Depletion of CD8(+) but not CD4(+) T cells or natural killer (NK) cells reduced the cytolytic activity of peritoneal lymphocytes. ADEPT did not cure tumor-bearing rats depleted of CD4(+) and CD8(+) T cells even though it was curative when given 7 days after tumor transplantation in rats with an intact immune system, indicating that ADEPT can synergize with host immunity to increase therapeutic efficacy. These results have important implications for the clinical application of ADEPT.

Efficient clearance of poly(ethylene glycol)-modified immunoenzyme with anti-PEG monoclonal antibody for prodrug cancer therapy.

The F(ab')(2) fragment of the anti-TAG-72 antibody, B72.3, was covalently linked to Escherichia coli-derived beta-glucuronidase that was modified with methoxypoly(ethylene glycol). The conjugate (B72.3-betaG-PEG) localized to a peak concentration in LS174T xenografts within 48 h after injection, but enzyme activity persisted in plasma such that prodrug administration had to be delayed for at least 4 days to avoid systemic prodrug activation and associated toxicity. Conjugate levels in tumors decreased to 36% of peak levels at this time. Intravenous administration of AGP3, an IgM mAb against methoxypoly(ethylene glycol), accelerated clearance of conjugate from serum and increased the tumor/blood ratio of B72. 3-betaG-PEG from 3.9 to 29.6 without significantly decreasing the accumulation of conjugate in tumors. Treatment of nude mice bearing established human colon adenocarcinoma xenografts with B72. 3-betaG-PEG followed 48 h later with AGP3 and a glucuronide prodrug of p-hydroxyaniline mustard significantly (p< or =0.0005) delayed tumor growth with minimal toxicity compared to therapy with a control conjugate or conventional chemotherapy.

Bystander killing of tumour cells by antibody-targeted enzymatic activation of a glucuronide prodrug.

RHI-betaG-PEG, formed by linking poly(ethylene glycol)-modified beta-glucuronidase to Mab RH1, was employed to examine bystander killing of antigen-negative N1S1 rat hepatoma cells by activation of a glucuronide prodrug (BHAMG) of p-hydroxyaniline mustard (pHAM) at antigen-positive AS-30D rat hepatoma cells. Sequential treatment of cells with 10 microg ml(-1) RH1-betaG-PEG and 20 microM BHAMG was not toxic to N1S1 cells but killed 99% of AS-30D cells. Over 98% of N1S1 cells, however, were killed in mixed populations containing as few as 2% AS-30D cells after identical treatment, demonstrating an in vitro bystander effect. Subcutaneous injection of AS-30D and N1S1 cells in BALB/c nu/nu mice produced solid tumours containing both cells. Uptake of radiolabelled RH1-betaG-PEG in solid AS-30D and mixed AS-30D/N1S1 tumours was 11.6 and 9.3 times greater than a control antibody conjugate 120 h after i.v. injection. Intravenous treatment with RH1-betaG-PEG and BHAMG cured seven of seven nude mice bearing solid s.c. AS-30D tumours and significantly delayed, compared with control conjugate and prodrug treatment, the growth of mixed N1S1/AS-30D tumours with one cure, showing that targeted activation of BHAMG kills bystander tumour cells in vivo.

Cure of malignant ascites and generation of protective immunity by monoclonal antibody-targeted activation of a glucuronide prodrug in rats.

We examined the in vivo efficacy of targeting beta-glucuronidase (betaG) to activate a glucuronide prodrug (BHAMG) of p-hydroxyaniline mustard (pHAM) at hepatoma ascites in Sprague-Dawley rats. Injection i.p. of 500 microg RH1-betaG, a conjugate formed between recombinant betaG and monoclonal antibody RH1 with specificity for an antigen expressed on AS-30D rat hepatoma cells, into rats bearing AS-30D ascites resulted in the accumulation of 54 microg conjugate per 10(9) tumor cells after 2 hr. Ascites fluid and serum contained 0.53 and 0 microg/ml, respectively, RH1-betaG 2 hr after injection of the conjugate. Conjugate binding to AS-30D cells was heterogeneous and non-saturated, as determined by flow cytometry. BHAMG was less toxic than pHAM to SD rats based on measures of animal mortality, weight loss and hematological toxicity. Treatment of rats bearing established hepatoma ascites with 500 microg RH1-betaG followed 2 hr later with a single i.p. injection of 30 mg/kg BHAMG or 3 i.p. injections of 10 mg/kg BHAMG 2, 3 and 4 hr later resulted in the cure of 6/8 and 8/8 animals, respectively. Treatment with BHAMG or pHAM alone did not produce cures, whereas treatment with a control antibody-betaG conjugate and BHAMG produced significantly greater hematological toxicity compared to treatment with RH1-betaG and BHAMG. All cured rats were completely protected from rechallenge with 2 x 10(7) AS-30D cells, indicating that successful treatment of animals induced protective immunity.

Anti-tumour effects of an antibody-carboxypeptidase G2 conjugate in combination with phenol mustard prodrugs.

ADEPT is an antibody-based targeting strategy for the treatment of cancer. We have developed two new prodrugs, 4-[N,N-bis(2-chloroethyl)amino]-phenoxycarbonyl-L- glutamic acid (PGP) and (S)-2-[N-[4-[N,N-bis(2-chloroethyl)amino]- phenoxycarbonyl]amino]-4-(5-tetrazoyl)butyric acid (PTP), which are cleaved by the bacterial enzyme CPG2 to release the 4-[N,N-bis(2-chloroethyl)amino] phenol drug. In vitro, both prodrugs are approximately 100- to 200-fold less potent than the parent drug (1 h IC50 = 1.4 microM) in LoVo colorectal tumour cells. These prodrugs have been evaluated for utility in ADEPT when used in combination with a conjugate of CPG2 and the F(ab')2 fragment of the anti-CEA monoclonal antibody, A5B7. The conjugate was shown to localise specifically to established LoVo tumour xenografts growing in nude mice and optimal tumour-normal tissue ratios were achieved after 72 h. Administration of either prodrug, at doses which cause 6-8% body weight loss, 72 h after administration of the A5B7-CPG2 conjugate to the LoVo tumour-bearing mice resulted in tumour regressions and growth delays of 14-28 days. The PTP prodrug in combination with a high dose of conjugate (10 mg kg-1) gave the best anti-tumour activity despite being a 10-fold worse substrate for CPG2 than PGP. Prodrug alone, active drug alone or prodrug in combination with a non-specific conjugate had minimal anti-tumour activity in this tumour model.

In vitro and in vivo activities of monoclonal antibody-alkaline phosphatase conjugates in combination with phenol mustard phosphate.

The prodrug p-[N,N-bis(2-chloroethyl)amino]phenyl phosphate (phenol mustard phosphate, POMP) was prepared from p-[N,N-bis(2-chloroethyl)amino]phenol (phenol mustard, POM) by phosphorylation with phosphoryl chloride, followed by aqueous hydrolysis. It was found that POMP was much less cytotoxic than POM when tested against H2981 human lung and H3396 human breast carcinoma cells in vitro. Pretreatment of the H2981 cells with L6-alkaline phosphatase (L6-AP), a monoclonal antibody conjugate that could bind to cell surface antigens, greatly enhanced the cytotoxic effects of POMP in an immunologically specific manner. Owing to its reduced toxicity in nude mice, larger amounts of POMP compared to POM could be administered. Neither agent exhibited significant in vivo antitumor activity when tested against subcutaneous H2981 tumors in nude mice. However, antitumor activity was observed in animals receiving L6-AP 48 h prior to POMP administration. This level of activity was greater than with the drugs alone, or a combination of 1F5-AP (nonbinding control) with POMP.

DNA-directed alkylating agents. 3. Structure-activity relationships for acridine-linked aniline mustards: consequences of varying the length of the linker chain.

Four series of acridine-linked aniline mustards have been prepared and evaluated for in vitro cytotoxicity, in vivo antitumor activity, and DNA cross-linking ability. The anilines were attached to the DNA-intercalating acridine chromophores by link groups (-O-, -CH2-, -S-, and -SO2-) of widely varying electronic properties, providing four series of widely differing mustard reactivity where the alkyl chain linking the acridine and mustard moieties was varied from two to five carbons. Relationships were sought between chain length and biological properties. Within each series, increasing the chain length did not alter the reactivity of the alkylating moiety but did appear to position it differently on the DNA, since cross-linking ability (measured by agarose gel assay) altered with chain length, being maximal with the C4 analogue. The in vivo antitumor activities of the compounds depended to some extent on the reactivity of the mustard, with the least reactive SO2 compounds being inactive. However, DNA-targeting did appear to allow the use of less reactive mustards, since the S-linked acridine mustards showed significant activity whereas the parent S-mustard did not. Within each active series, the most active compound was the C4 homologue, suggesting some relationship between activity and extent of DNA alkylation.

DNA-directed alkylating agents. 1. Structure-activity relationships for acridine-linked aniline mustards: consequences of varying the reactivity of the mustard.

A series of DNA-targeted aniline mustards have been prepared, and their chemical reactivity and in vitro and in vivo cytotoxicity have been evaluated and compared with that of the corresponding simple aniline mustards. The alkylating groups were anchored to the DNA-intercalating 9-aminoacridine chromophore by an alkyl chain of fixed length attached at the mustard 4-position through a link group X, while the corresponding simple mustards possessed an electronically identical small group at this position. The link group was varied to provide a series of compounds of similar geometry but widely differing mustard reactivity. Variation in biological activity should then largely be a consequence of this varying reactivity. Rates of mustard hydrolysis in the two series related only to the electronic properties of the link group, with attachment of the intercalating chromophore having no effect. The cytotoxicities of the simple mustards correlated well with group electronic properties (with a 200-300-fold range in IC50S). The corresponding DNA-targeted mustards were much more potent (up to 100-fold), but their IC50 values varied much less with linker group electronic properties. Most of the DNA-targeted mustards showed in vivo antitumor activity, being both more active and more dose-potent than either the corresponding untargeted mustards and chlorambucil. These results show that targeting alkylating agents to DNA by attachment to DNA-affinic units may be a useful strategy.

Hypoxia-selective antitumor agents. 3. Relationships between structure and cytotoxicity against cultured tumor cells for substituted N,N-bis(2-chloroethyl)anilines.

A series of aniline mustards with a wide range of electron-donating and -withdrawing substituents in the 3- and 4-positions has been synthesized and evaluated for cytotoxicity in cell culture to examine the potential of using nitro group deactivated nitrogen mustards for the design of novel hypoxia-selective anticancer drugs (Denny, W. A.; Wilson, W. R. J. Med. Chem. 1986, 29, 879). Hydrolytic half-lives in tissue culture media, determined by bioassay against a cell line (UV4) defective in the repair of DNA interstrand cross-links showed the expected dependence on the Hammett electronic parameter, sigma, varying from 0.13 h for the 4-amino analogue to greater than 100 h for analogues with strongly electron-withdrawing substituents. Cytotoxic potencies in aerobic UV4 cultures showed a similar dependence on sigma. This dependence predicted that the 4-nitroaniline mustard would be 7200-fold less potent than its potential six-electron reduction product, the 4-amino compound, in growth inhibition assays using a 1-h drug exposure. The measured differential was much lower (225-fold) because of the instability of the latter compound, but a differential of 17,500-fold was observed in the initial rate of killing by using a clonogenic assay. The potential for formation of reactive mustards by reduction to the amine or hydroxylamine was demonstrated by the 4-nitroso compound, which had an aerobic toxicity similar to that of the amine. Although these features confirmed the original rationale, the 3-nitro- and 4-nitroaniline mustards had only minimal hypoxic selectivity against UV cells. Toxicity to hypoxic cells appears to be limited by the low reduction potentials of these compounds and consequent lack of enzymatic nitroreduction. However, this study has demonstrated that nitro groups can be used to latentiate aromatic nitrogen mustards and indicates that examples with higher reduction potentials could provide useful hypoxia-selective therapeutic agents.

Quinone-induced DNA damage and its relationship to antitumor activity in L5178Y lymphoblasts.

The presence of a quinone group in the structure of a series of model compounds was shown to produce cell kill by a mechanism involving free radicals and active oxygen species. Furthermore, the ability of the compound to bind to DNA appeared to enhance its cytocidal activity. The same model compounds were used to investigate the effect of the quinone group on cellular DNA. DNA single-strand breaks, DNA double-strand breaks, and DNA-DNA cross-linking induced by the model compounds were measured by elution assays. Hydrolyzed benzoquinone mustard, which contains a quinone group, induced dose-dependent single-strand and double-strand breaks but no DNA cross-linking. Benzoquinone mustard, which possesses both a quinone moiety and an active alkylating group, produced dose-dependent DNA double-strand breaks but no apparent single-strand breaks. However, this compound produced significant levels of DNA cross-linking, a process which interferes with the assay for single-strand breaks. The relative activity of benzoquinone mustard in inducing DNA double-strand breaks was approximately 15,000-fold greater than that of hydrolyzed benzoquinone mustard. Aniline mustard, which has the same alkylating group as does benzoquinone mustard but no quinone function, produced lower levels of DNA-DNA cross-links and no DNA strand breaks. The induction of both DNA single-strand and double-strand breaks by hydrolyzed benzoquinone mustard was significantly inhibited by the cell-protective enzymes superoxide dismutase and catalase. The cytotoxic activity of hydrolyzed benzoquinone mustard appeared to correlate with the induction of DNA single- and double-strand breaks. These studies provided evidence that the presence of a quinone group in the chemical structure of a compound results in the production of DNA strand breaks. DNA damage was inhibited by superoxide dismutase and catalase, suggesting the involvement of free radicals and active oxygen species. The induction of DNA damage appeared to be enhanced by the ability of the compound to bind to DNA. The induction of strand breaks may correlate with the cytotoxic activity of the quinone agents.

Response of a high-glucuronidase human tumour xenograft to aniline mustard.

The HT29R colonic adenocarcinoma xenograft has been shown to be rich in the enzyme beta-glucuronidase. Experiments in rodent systems have demonstrated a marked anti-tumour effect of the drug aniline mustard (AM) on tumours with high levels of this enzyme (e.g. the plasmacytomas PC5 and PC6). We have found that AM is no more effective than its analogue paramethyl aniline mustard (PMAM) or other alkylating agents against the HT29R xenograft. Amongst the possible explanations for this may be: (1) The wide shoulder on the cell-survival curve shown for exposure to alkylating agents of HT29R in vivo. (2) Lack of correlation between physiological availability of beta-glucuronidase and the high levels measured by the standard assay. (3) Increased beta-glucuronidase levels in host mouse marrow, making the latter potentially more susceptible to AM damage.

Enhanced cytostatic effectiveness of aniline mustard against 7,12-dimethylbenz[a]anthracene-induced rat mammary tumors during regression in response to ovariectomy.

Sprague-Dawley rats bearing 7,12-dimethylbenz[a]anthracene (DMBA)-induced mammary tumors were treated with either of two aromatic alkylating agents, aniline mustard or melphalan, alone or combined with ovariectomy. Both drugs were applied once a week for 8 weeks. Eight-four percent of the tumors responded to ovariectomy, 38% regressing completely and 46% regressing partially. Aniline mustard, though virtually ineffective as a single agent, appeared synergistic with ovariectomy: a 100% regression rate (72% complete, 28% partial) was observed for this combination. Treatment with melphalan was as effective as ovariectomy, but the combination of melphalan with ovariectomy was no more effective than either treatment alone. The end product of aniline mustard metabolism, p-hydroxyaniline mustard O-glucuronide, may be more extensively activated by beta-glucuronidase in hormonally regressing than in growing or stationary tumors. Intratumoral levels of beta-glucoronidase occurring in DMBA-induced tumors 4 days after ovariectomy were found to be similar to those in the aniline mustard-sensitive mouse plasma cell tumor ADJ/PC6. It remains to be more extensively studied whether an effect of endocrine treatment on tumor beta-glucuronidase levels, and possibly on intracellular distribution of enzyme, could be used therapeutically. An effectively scheduled cytostatic treatment (with a drug conjugate such as that formed metabolically from aniline mustard) in conjunction with ovariectomy might be effective in the treatment of hormone-responsive breast cancer.

Comparison of quantitation methods for L-1210 cell populations and evaluation of selected cytotoxic agents in leukemic mice.

A more rapid cell-counting technique using an electronic cell counter was developed as an improvement over the slower hemocytometer method. The electronic counting method produced cell counts that had a smaller standard deviation but were not significantly different from the hemocytometer method. The acetate, hexanoate, and decanoate esters of p-N,N-bis(2-chloroethyl)aminophenol were investigated for acute toxicity in mice, effects on survival times, and effects on L-1210 cell populations in L-1210 leukemic mice. The decanoate ester was the least toxic compound and was most effective in lengthening the lifespan of the mice. The acetate and hexanoate esters were more effective in reducing L-1210 cell populations.

Structure-activity relationships in antitumor aniline mustards.

Quantitative structure-activity relationships (QSAR) have been formulated for the hydrolysis of aniline mustards and their antitumor activity against Walker 256 tumor and L1210 and P388 leukemia. In general, the antitumor activity parallels hydrolysis under the conditions defined by Ross; toxicity (LD50) parallels antitumor efficacy. Chlorambucil is an exception. A most important finding is that ideal lipophilicity for effectiveness against Walker tumor appears to be much higher than for the leukemias which suggests that solid tumors may, in general, require more lipophilic drugs than leukemias.

Therapeutic trial of aniline mustard in patients with advanced cancer. Comparison of therapeutic response with cytochemical assessment of tumor cell beta-glucuronidase activity.

Seventy-eight patients with advanced cancer received an adequate therapeutic trial with aniline mustard (NSC 18429). Significant anticancer activity with clinical benefit was demonstrated in five patients with cancer of the prostate and one patient with renal cancer. beta-glucuronidase levels in aspirate and imprint preparations of tumor cells were assessed by a timed cytochemical technique. A partial correlation appeared to exist between very intense glucuronidase staining and tumor regression in prostate and kidney lesions; however, these high levels were observed only rarely. Sequential observations in two patients demonstrated loss of enzymatic activity concomitant with development of clinical relapse.