Commentary (Rowinsky): New Anticancer Agents in Clinical Development

For nearly two decades, the development of novel cytotoxic agents progressed at a snail's pace. In the 1970s and 1980s--decades of instant gratification in which massive expenditures and manpower were expected to have immediate payoffs, the future of new cytotoxic drug discovery looked dismal.

Further pessimism was generated due to the naive, overexaggerated expectations of programs, such as the war on cancer, which were often inappropriately contrasted to the space program, which accomplished staggering feats of engineering in order to land a man on the moon. However, the rational development of novel cancer therapeutics requires that the basic mechanisms of cancer cell dysregulation be unraveled before new strategic cellular targets can be discovered and rational therapeutics designed. Unlike the engineering achievements of the space program, understanding these processes may be considered tantamount to understanding life itself, and certainly cannot be accomplished in such a short time, even if expenditures are vast.

Identification of Novel Subcellular Targets

Most major therapeutic advances in any single cancer have resulted from the discovery of novel therapies, particularly cytotoxics. This process has largely been serendipitous, generated by the costly, large-scale, random, and probably inefficient screening of natural products and compounds with unique chemical structures. In a roundabout manner, drug discovery by serendipity has led to the identification of novel subcellular targets, which have unexpectedly increased our knowledge of cellular biology and neoplasia.

Such has been the case with the taxanes and camptothecins, which were identified through a large-scale "serendipitous" screening program. Largely as a consequence of these "accidents," new strategic targets for cancer therapeutics--tubulin polymerization and topoisomerase I--were elucidated. Just as we were noticing some real progress, however, the inherent inefficiency and staggering costs of such drug discovery processes led to the scaling down and phasing out of screening programs, particularly in the private sector. The potential impact of these decisions is a major concern.

The report by Eckardt et al clearly indicates that the 1990s may be the decade in which we begin to appreciate the payoff and investments that began several decades ago, as well as the need for persistence and patience before recklessly abandoning ship. Not only have drug screening programs identified prototypic compounds, such as paclitaxel and camptothecin, the identification of these agents, coupled with a renewed interest in natural products and advances in natural product chemistry, has resulted in the synthesis of analogs, such as docetaxel (Taxotere), topotecan, irinotecan, 9-aminocamptothecin, and GI147211, as well as other taxane and camptothecin analogs currently in preclinical development.

Traditionally, analog development has not substantially broadened the antitumor spectra of prototypic compounds or produced major therapeutic advances. However, the results of phase I and II studies with taxane and camptothecin analogs indicate that these agents may possess unique toxicologic and antitumor spectra, which may result in widely disparate therapeutic indices and applications. More importantly, these differences raise the prospect of discovering more potent analogs, possibly compounds in preclinical development, on the shelf, or not yet conceived.

The success with the taxanes and camptothecins also raises questions about our predilection for developing water-soluble compounds due to their ease of administration and our relative lack of progress in developing feasible clinical formulation vehicles for hydrophobic agents, which may possess optimal pharmacologic characteristics and superior therapeutic indices. However, times are changing, as exemplified by the decision to develop 9-aminocamptothecin using a new colloidal dispersion formulation and the pursuit of evaluations of oral formulations of camptothecin, topotecan, and 9-aminocamptothecin.

Rational Drug Design Finally Bearing Fruit

Rational drug design is also alive and well, and appears finally to be bearing fruit, particularly due to improvements in powerful analytical techniques for characterizing strategic subcellular targets. Eckardt et al discuss exciting clinical results with Tomudex and other thymidylate synthase (TS) inhibitors. The molecular definition of this target, coupled with the clinical results with Tomudex in colorectal cancer, has provided the impetus to rationally design the lipophilic TS inhibitors AG-331 and AG-337. These agents, which traverse cell membranes by passive diffusion and do not possess terminal glutamate moieties, were designed after the crystal structure of TS was characterized using x-ray crystallography, and are the first anticancer drugs so designed to reach clinical trials.

Until recently, the enzyme dihydropyrimidine dehydrogenase also had received little attention as a vital subcellular target, despite the fact that it is the rate-limiting enzyme involved in the catabolism of fluorouracil. Intraindividual and intratumor differences in the activity of this enzyme may be responsible for substantial interindividual variability in the toxicologic and pharmacologic profiles of the fluoropyrimidines and may contribute to their broad range of antitumor activity.

In addition, substantial variability in the activity of dihydropyrimidine dehydrogenase has been the principal reason why administration of fluorouracil by the oral route is not feasible. The potential for the enzyme inactivator 776C85 to substantially affect clinical practice with respect to the administration of fluorouracil is enormous. To date, preclinical results in xenograft models indicate that 776C85 may increase the therapeutic index of fluorouracil, and preliminary clinical data suggest that the enzyme inactivator may permit reliable oral dosing of fluorouracil as an alternative to intravenous bolus and protracted infusion schedules.

Traditional Drug Discovery Programs Still Important

Although the vast amount of information about the molecular biology of neoplasia that has recently been generated will undoubtedly result in many successes in the rational design of cancer therapeutics in the future, traditional drug discovery programs are currently yielding important compounds, which, in turn, are leading to the elucidation of new targets and progress, albeit slow, in cancer therapeutics. Therefore, traditional drug discovery programs should be kept healthy and should not be abandoned, at least until newer "rationally guided" approaches consistently yield results of major therapeutic importance.