The growth of a tumor can be simplistically conceptualized as being dependent on several variables: the rate of cell loss, growth fraction (proportion of cells in proliferative phase) and cell doubling time. Several models have been devised to explain the impact of therapeutic regimens on the tumor cell cycle [4].

Skipper et al. originally proposed the log kill model, which was based on the behavior of L1210 leukemia cell lines in rodents [5]. They suggested that the benefit of chemotherapy was due to the cytocidal effects on the cancer cells and that tumor growth and tumor regression in response to chemotherapy were exponential in nature. This model proposed that a drug that would cause a reduction of tumor burden from 10¹² to 10¹¹, if given in the same dose, would also decrease the burden from 10⁶ to 10⁵ (i.e. each reduction reflected a 90 % tumor cell kill). This formed the basis for the use of repeated cycles of chemotherapy to achieve maximal tumor eradication. However, it has become clear that exposure to the same regimen for more than 4–6 doses does not improve outcomes. We now understand the growth models of tumor have also a mixed nature with respect to different cancer cells in vivo [6]. The Gompertzian sigmoid growth curve is characteristic of many solid tumors. Tumors grow most rapidly at smaller sizes and then the growth rate slows, secondary to problems with vascularity, hypoxia and interaction with the other cells in their microenvironment [7–11].

Concepts initially derived from resistance in antibacterial therapy were applied to the study of antimetabolites in the treatment of murine cancer models and cell lines, and it was shown that resistance is acquired at various points during the growth of the tumor. This also occurs in human tumors. Goldie and Coldman suggested that larger tumors have more cells and thus have a greater chance of undergoing spontaneous mutation [12], which may, in turn, confer resistance to chemotherapy. It has thus been suggested that tumors are more effectively treated at a smaller size, before they have the ability to develop mutations and resistance. This has also led to the concept that non-cross resistant multiple agent chemotherapy would have a greater proportionate tumor kill and could thus prevent the development of resistance.

Another fundamental concept of cell cycle kinetics is the so-called Norton-Simon regression hypothesis [13]. Their concept is that administration of chemotherapy at shorter time intervals (with greater dose-intensity) achieves greater proportionate tumor cell kill. As noted above (Fig. 13.2), some tumors do not grow in a simple exponential manner but rather follow a Gompterzian growth pattern [11, 13]. Norton and Simon have also proposed that less time available for tumor regrowth between treatments will improve the chance of cure, and that sequential, dose-dense, non-cross resistant regimens will minimize drug resistance. This approach destroys the dominant tumor population initially via the first component of treatment, with different agents addressing the residual resistant cells, without the potential for significant tumor regrowth.

The Goldie-Coldman and Norton-Simon hypotheses have also been applied to the design of adjuvant chemotherapy regimens, attempting to rationalize the prior empirical approach. This has resulted in improved cure rates, perhaps due to increased cytotoxicity and/or a reduced emergence of drug resistance [13–15].

The clinical pharmacology of antineoplastic agents has been an evolving field for the past century, with the most progress having been made in the past decade through the advent of analytical tools such as liquid chromatography and mass spectroscopy. Most cytotoxic agents offer a very narrow therapeutic index compared with most drugs [16].

Response of tumors can be divided into three broad groups [21, 22]: (a) drug sensitive tumors, with treatment resulting in cure; (b) highly responsive tumors, but with eventual refractoriness to cytotoxic treatment; and (c) tumors with little responsiveness to chemotherapy (Table 13.1).

Drug resistance has been studied in both in vivo and in vitro in a range of models. Multiple reasons involving anatomical, pharmacological and biochemical mechanisms explain various aspects of tumor resistance to cytotoxic chemotherapy:

Approaches to circumvent drug resistance have involved the use of multidrug combinations, dose escalation, agents that reverse increased efflux, cofactors that amplify drug efficacy, and inhibition of drug inactivation. Recently liposomal and nanoparticle albumin- stabilized formulations have been applied to overcome drug resistance, but with limited success in genitourinary malignancy [31]. These increase the delivery of chemotherapy to the tumor cell while minimizing toxicity.

Single agent regimens, with few exceptions (e.g. gestational choriocarcinoma), are rarely able to achieve cure. In view of this and the observations above, combination chemotherapy regimens have been devised to accomplish the major objectives of attaining maximum tumor kill with minimal toxicity and to prevent drug resistance [16, 32]. The era began when numerous active drugs became available simultaneously and were applied to curative management of leukemias and lymphomas in the late 1970s. Fundamental principles used in the selection of the drugs for combination regimens include [32]:

The relationship between doses and combination of these agents is complex [32, 33]. The maintenance of dose intensity has proved to be important in the success of some of these regimens. Reduction of dose can result in significantly decreased cure rates, especially in the more responsive tumors such as germ cell cancer. Thus, although responses may occur despite dose reduction, residual tumor cells often persist, leading to eventual relapse and decreased survival. The concept of relative dose intensity (amount of drug delivered in a given time frame) has evolved over the past two decades, but it remains controversial as to whether this is a major factor in cure [33].

Ideally drugs should be used in their optimal schedule and dosage even when being combined with other agents. However, care must be exercised to avoid over-dosage and excessive normal-tissue toxicity. In general, the interval of drug delivery also needs to be consistent, keeping the treatment free interval to be the shortest time necessary for resolution of any dose limiting toxicity (usually bone marrow toxicity) to maintain dose intensity [32–34].

The selection of patients to receive combination chemotherapy has also undergone refinement. In the present era, we are more cautious than in the past, and many oncologists believe that it may not be appropriate for a patient with a poor performance score (e.g. ECOG level 3–4) to be given a toxic regimen, unless there is substantial evidence that his disease is likely to be highly responsive to the treatment [35]. We believe that it is critically important to discuss the aims and expectations of chemotherapy with all patients, but of even greater importance in the elderly and infirm. By contrast, with better understanding of age-related changes in drug pharmacology and the importance of performance status, more venturesome approaches are now being explored in the chemotherapy of the fit elderly with advanced cancer [36].

High dose chemotherapy involves the use of potentially lethal doses of chemotherapy with or without radiation, followed by rescue with hematopoietic stem cells obtained from the blood or by marrow harvest and storage [37–40]. This is predicated on the concepts that (a) there is a dose response relationship for a specific regimen in certain tumors; (b) there is benefit to dosing at a higher level than can be survived by normal tissues (e.g. bone marrow), provided that re-infusion of bone marrow can avoid the death of the host. This modality, which is predominantly used in hematological malignancies, has been applied to the management of relapsed or poor-risk metastatic germ cell tumors [38]. To date, despite evidence of anti-cancer activity, prospective, randomized trial data have not shown an unequivocal survival benefit from this approach to the management of advanced germ cell tumors, compared to standard dose poor-risk or salvage regimens [38, 41]. There are still advocates of this procedure in the management of poor-risk or relapsed germ cell tumors, but the majority of evidence to support this contention comes from retrospective, historically controlled analyses [42].