However, I do not accept the argument that further efforts to develop anticancer drugs should be abandoned. Braverman [1] advocates a "biological" approach to cancer treatment. In fact, all treatment research efforts, including current cancer drug discovery and development, are based on the explosive growth of knowledge of cell and molecular biology. Braverman has seriously underestimated the biological knowledge that underlies modern efforts to discover new cancer drugs. Cancer biology is an integral part of any informed approach to cancer drug discovery, whether for biological or chemical agents.

Cancer Biology Is the Basis of Drug Discovery Research

Cancer drug discovery efforts fall into two categories: those that seek new active chemical entities (lead compounds) using high through-put screening and those that are based on a rational and specific concept of drug-target interaction. Both areas of drug discovery have undergone fundamental changes in recent years as a result of the rapid growth of knowledge of cancer biology. Cancer drug screening no longer uses murine tumors but is targeted at specific types of human cancer (such as breast cancer or prostate cancer) or biological targets (such as oncogenes or growth factors) known to be responsible for unregulated growth in human neoplasms. The age of drug screening against murine leukemias is over and has been replaced by screens that use, at the least, panels (groups) of human tumor cells of specific histologic types.

The current National Cancer Institute screening efforts use a panel of 60 human tumor cell lines for primary screening [2, 3]. These cell lines are being characterized for patterns of drug response to known agents and, more importantly, for the expression of a broad range of factors that affect drug resistance [4] and regulate growth, including growth factors and oncogenes. Through computer-based analysis of patterns of cell-line response, it is possible to identify the mechanism of action of a new cytotoxic entity [5], to select for compounds having entirely unique patterns of activity and mechanisms of action [6], and to identify candidate compounds that interact with biological targets such as drug-resistance factors or the ras family of oncogenes.

Grant-supported drug discovery programs use molecular targets (such as the mutated ras oncogene family) associated with human malignancy, or use proteins involved in signal transduction (such as tyrosine kinases), or use protein kinase C as their primary screen for lead compounds. These screens have yielded lead compounds with entirely novel sites of action, such as the tyrosine kinase inhibitors that have activity against prostate cancer [7]. Molecular drug design, based on computer reconstruction of a molecular target and its active site, has led to the synthesis of potent inhibitors of the human immunodeficiency virus protease [8]; similar molecular design efforts are underway for cancer drug screening.

Recent Progress in Cancer Drug Discovery

It would be a mistake to minimize the continuing contribution of traditional drug development programs, which have yielded the drugs that are our only current means of curing childhood malignancies, lymphomas, adult leukemias, and testicular cancer, as well as extending survival of patients with ovarian cancer, myeloma, the chronic leukemias, and other rarer tumors. In the last 3 years, a number of highly useful new drugs have been approved for marketing (Table 1).

Two promising new drug classes, the taxanes and camptothecins, are in early stages of clinical development. Taxol has produced a 30% response rate in refractory ovarian cancer patients [16] and a 50% to 60% response rate in breast cancer patients after failure of primary therapy [17]. CPT-11, the most advanced camptothecin analog, has antitumor activity in colon cancer [18], non-small cell lung cancer [19], as well as cervical and ovarian cancer [20]. These new agents inhibit unique biochemical mechanisms: Taxol inhibits mitosis through its stabilization of microtubules [21], whereas CPT-11 inhibits DNA strand passage mediated by topoisomerase I [22]. These are entirely new agents, with new mechanisms of action and with clinically useful effects.

Cancer Biology and Drug Resistance

Advances in cancer cell biology and growth regulation hold great promise not only for drug discovery efforts but also for improving the efficacy and decreasing the toxicity of currently available drugs. Although Braverman [1] wants us to focus on blocking oncogenes, recent work indicates that the absence or mutation of tumor-suppressor genes such as p53 may confer sensitivity to irradiation and DNA-damaging drugs [23]. Tumor-suppressor genes, including p53, act as checkpoints, preventing normal cells from entering the DNA synthetic phase of the cell cycle until their damaged DNA is repaired. The absence of checkpoint function may explain the mutability of neoplastic cells and their particular ability to generate drug-resistant mutants.

During the past decade, great progress was made in defining the genetics and biochemistry of drug resistance in experimental tumors. A single gene, mdr-1, was identified and shown to produce a broad-based resistance to the vinca alkaloids, etoposide, doxorubicin, and other natural products; it codes for a cell-membrane pump (the P-glycoprotein) that actively extrudes toxic drugs from tumor cells [24]. A variety of drugs, including calcium-channel blockers, cyclosporins, and calmodulin antagonists, block the P-glycoprotein and reverse drug resistance in experimental tumors [25]. The discovery of mdr-1 and its product has opened a new area that may have important ramifications for drug discovery strategies, clinical chemotherapy, and even gene therapy. Recent work [26] indicates that mdr-1-mediated resistance probably plays a role in chemotherapy failure in certain types of leukemia and in patients with non-Hodgkin lymphoma [27], multiple myeloma [28], and neuroblastoma [29], and is suspected to influence response in breast cancer [30]. Trials examining reversal of multidrug resistance are underway for many of these malignancies, and new resistance-modulating agents are entering clinical evaluation. These attempts at clinical reversal of resistance are at an early stage; plasma concentrations of agents such as verapamil are probably not adequate to do the job. Trials have not yet consistently identified and targeted an appropriate patient population expressing mdr-1. New and more potent reversing agents will enter clinical trial shortly. Experiments will soon begin in which the mdr-1 gene is inserted into human bone marrow stem cells to confer drug resistance on the host and to allow nonmyelosuppressive chemotherapy of breast cancer. Multidrug resistance with mdr-1 is the first of many new leads in the effort to understand drug resistance in clinical chemotherapy.

"Biologicals" Enhance Effectiveness of Drugs

Basic biological research has contributed to decreasing the toxicity of chemotherapy. Colony-stimulating factors [31, 32] and bone marrow stem cell harvesting and reinfusion [33] have decreased the risk for bone marrow transplantation and intensive chemotherapy. Similar rescue strategies are required for protection of megakaryocytes and intestinal epithelium, but important lead biologicals that target these tissues have been identified, including interleukin-6 for megakaryocytes [34] and tumor growth factor ß for bone marrow stem cells [35] and for intestinal epithelial cells [36]. In the future, we will have the ability to protect these vital organs from the acute toxicity of chemotherapy.

Increasing Support for Biological Therapies

Despite the advances in understanding drug resistance and the potential of exploiting biological targets for drug development, the search for cancer therapies must not be confined to traditional drugs. Thus, the National Cancer Institute is investing a much greater percentage of its support in basic research than in drug discovery and has greatly expanded its support of the new fields of immunotherapy and gene therapy. Therapeutic research in biological compounds, which started in 1981 with the interest in -interferon, now has a rapidly growing budget that is half of that for cytotoxic chemotherapy.

Despite this considerable investment, biological therapies have had a limited impact thus far on the practice of medical oncology. Alpha-interferon has been a useful therapy for hairy cell leukemia [37] but has been supplanted by deoxycoformycin and by 2-chlorodeoxyadenosine, which produce long-lasting complete responses in most patients. Alpha-interferon produces occasional responses in other malignancies but has not yet had a major impact on specific forms of cancer. Interleukin-2, either alone or with tumor-infiltrating lymphocytes, produces occasional long-lasting responses in some patients with renal cell carcinoma or melanoma but causes considerable morbidity. Unarmed monoclonal antibodies have produced responses in malignant melanoma and lymphoma and neuroblastoma, whereas antibodies conjugated to radionuclides or toxins have shown activity in lymphoma and breast cancer [38]. However, the effectiveness of monoclonal antibodies has been hampered by some unique problems, including host antibody responses to the foreign antigen, and by problems that afflict drugs and biologicals in common: 1) toxic effects on normal host tissues; 2) tumor mutability, leading to mutation or loss of the biological target; and 3) inadequate penetration of the drug or biological into the tumor.

For other biological targets, such as the deficient or mutated suppressor genes mentioned by Braverman [1], therapeutic strategies are not obvious: How do we replace the function of a missing tumor suppressor protein? Essentially, we must find a small molecule (a drug) that will penetrate tumor cells and mimic the DNA-binding functions of the p53 gene product. This is a challenging task and is much more complex than identifying an inhibitor of an enzymatic reaction or a new antibiotic. Although the basic strategy of finding a molecule to replace p53 function has great merit, its practical implementation is not yet possible.

New Avenues for Therapy: Apoptosis and Differentiation

Braverman [1] realizes that antitumor therapy has expanded beyond our previous concepts of cytotoxic or biological therapies and now includes agents that induce differentiation or programmed cell death. The phenomenon of apoptosis [39], or programmed cell death, may be fundamental to the understanding of malignancy and may underlie the effectiveness of chemotherapy as well as interferons. In experimental settings, tumor cells as well as normal cells can be induced to undergo DNA fragmentation and cell death. This process, called apoptosis, is probably responsible for the normal elimination of cells with damaged DNA or other potentially dangerous cells, such as autoreactive lymphocyte populations. A loss of apoptosis may cause follicular lymphomas as the result of the characteristic 14:18 chromosomal translocation in that disease [40]. Interestingly, many of the current cancer drugs activate the process of apoptosis, probably through injury to DNA [41]. New therapies may arise from exploiting apoptosis using noncytotoxic pathways.

Differentiation is an alternative to the induction of programmed cell death. Malignant cells may be induced to differentiate into mature, nonreplicative cells through exposure to various entities, including interferons, retinoids, and low concentrations of cytotoxic drugs [42]. The first clinical breakthrough in the application of differentiation therapy, as cited by Braverman [1], occurred with the use of all-trans-retinoic acid for induction of complete remission in patients with acute promyelocytic leukemia. This therapy produces complete responses even in drug-resistant patients [43]. The best way to use this therapy is still unclear, but it will probably be used either in combination or in sequence with chemotherapy. Retinoids prevent the occurrence of second primary carcinomas in patients with head and neck cancer [44], and with -interferon, produce response in squamous cell carcinomas of the skin [45] and cervix [46]. Although the biology underlying differentiation therapy is not completely understood, strong indications now exist that it will yield new drugs. On this point, Braverman and I are certainly in agreement.

Tumor Vaccines and Gene Therapy

Tumor vaccines and gene therapy represent the newest forays into biological therapy, have great potential, and will soon have practical application. Great progress has been made in defining antigens unique to tumors [47, 48], including oncogene products such as immunoglobulins [49] unique to B-cell lymphocytic tumors and the melanoma-associated antigen [47]. We now know that tumor cells have such unique antigens and present them to the immune system on the surface of the tumor cell. Why these antigens fail to generate a strong antitumor response de novo is not understood. However, the pathways of antigen presentation and immune response are being rapidly dissected, and measures designed to enhance steps in this pathway are being evaluated in both animals and humans [50].

Insertion of new genes into tumors, particularly genes that enhance recognition of tumor-associated antigens, can augment the host response [51, 52]. Pardoll and associates [53] showed that insertion of the gene coding for granulocyte macrophage colony-stimulating factor (GM-CSF) into tumor cells greatly enhances the attraction of antigen-presenting cells to the tumor cell surface (where tumor-specific antigens are recognized), and a rapid antitumor reaction mediated by T lymphocytes is elicited. Clinical experiments to test this finding are being planned. The field of tumor immunology and immunotherapy will probably soon contribute to cancer treatment.

Thus, I agree with Braverman's position [1] that our increasing knowledge of cancer biology will yield better therapies. However, the applications of cancer biology will occur not only in the field of biotherapy but will also have an equal effect on conventional drug discovery and will allow us to use drugs more safely and with greater effect.