Early studies of the ability of chemotherapy to kill cancer cells were conducted on leukemia cell lines in the 1960s.³ These studies noted log-kill kinetics, meaning if 99% of cells were killed, tumor mass would decrease from 10¹⁰ to 10⁸ or from 10⁵ to 10³. The fraction of cells killed was proportional, regardless of tumor size; thus, even though a given treatment would appear to have eradicated the tumor, both clinically and radiographically, there would be a high probability of residual cells that would eventually proliferate and show up as a clinically evident tumor (relapse). One explanation for the achievement of sustainable complete remission following this argument would be that other factors such as host immune response may be important at low levels of residual tumoral cells.⁴

Clinical prognostic models are, in part, based on risk of disease relapse and, thus, take into account features that might suggest micrometastatic undetectable disease at the time of diagnosis. As an example, a large tumor may suggest a longer clinically silent tumor lifetime or higher doubling rate. Clinically apparent nodal involvement demonstrates the tumor has gained the capability to spread, at a minimum regionally.

Indications for Chemotherapy

If the decision is made that the patient will benefit from chemotherapy, the treatment strategy devised by the medical oncologist will be largely determined by the stage of the cancer. The initial medical treatment of cancer can be thought of as requiring (1) preoperative (neoadjuvant) chemotherapy, (2) postoperative (adjuvant) chemotherapy, or (3) chemotherapy without localized therapy, either for metastatic, inoperable disease (either due to locally advanced stage and/or comorbid medical conditions) or a hematologic malignancy. Chemotherapy without surgical therapy was historically thought of as a palliative measure; however, improved efficacy of chemotherapy and radiotherapy is changing this notion. Many hematologic and epithelial malignancies are now approached in curative fashion with chemotherapy alone or combined with radiotherapy.

Adjuvant Chemotherapy

When the first chemotherapies were developed, they were utilized only in patients with advanced disease and who were failing other therapies. This was largely due to poor efficacy of therapy, and chemotherapy in this setting was usually associated with significant treatment-related morbidity. The therapeutic index (benefit opposed to morbidity) and associated supportive care measures of chemotherapy today tip this balance in favor of treating patients earlier, even with no objective evidence of postsurgical disease.

Surgical and radiotherapeutic treatments have made miraculous progress in the treatment for localized disease; however, many cancers have metastatic spread at diagnosis. Surgically or radiotherapeutically treated tumors may fail locally, but they often recur at distant sites. When considering the previously discussed undetectable period of tumor growth, it becomes apparent how a completely resected tumor may have significant occult residual disease, either locally or distant spread. In this situation, chemotherapy is given as an adjuvant to augment the effect of surgery; hence, the name adjuvant chemotherapy. Many patients who receive adjuvant therapy are without evidence of disease after local therapy. The pathologic margins of surgical specimens may be negative, and imaging may reveal no abnormality; however, significant relapse potential from residual local disease or micrometastases may exist. Adjuvant therapy aims to eradicate this subclinical disease before it reaches a critical threshold at which cure becomes difficult. Breast, lung, and colon cancer are a few examples of the many cancers that benefit from adjuvant therapy.⁵^–⁷

Neoadjuvant Chemotherapy

Treating with chemotherapy before surgery is a newer concept. With more effective chemotherapy, neoadjuvant treatment approaches are occasionally used in appropriate-stage breast, lung, and resectable metastatic colorectal cancers.⁸^–¹⁰

Neoadjuvant chemotherapy has three main advantages. Micrometastases are exposed to chemotherapy earlier in the treatment course, which may more effectively lead to eradication prior to becoming clinically apparent (based on log-kill theory). Second, a primary lesion that fails to respond indicates micrometastatic disease that is also likely resistant, allowing a change in therapy.¹¹ Without neoadjuvant therapy, this knowledge would become apparent only after micrometastatic disease becomes clinically apparent and, thus, the potential for cure at that time would be very unlikely. Third, a primary tumor may regress sufficiently to facilitate a less morbid surgical procedure or occasionally obviate the need for surgical resection.^12,¹³

Chemotherapy for Metastatic Cancer

The large portion of chemotherapy is given for clinically evident metastatic cancer, often as part of a palliation strategy in order to prolong survival and improve quality of life. However, some malignancies, including lymphoma and testicular cancers, may be cured even when they present with advanced metastatic disease. Other cancers, including ovarian and breast cancer, may demonstrate great sensitivity to chemotherapy in the metastatic setting with long-term disease control or even transient disappearance of all disease. An additional number of cancers, such as lung or pancreatic, may have brief stabilization or minor response to therapy, but long-term control is uncommon.

Chemotherapy Schedules

Dose Intensity

In tumors that follow the Gompertzian growth model, the fraction of cycling cells will increase as a tumor mass shrinks from chemotherapy. Each exposure to chemotherapy creates a selection pressure; thus, the remaining cells are more likely to develop resistance. As a result, adequate dose intensity is required for tumor eradication. If chemotherapy doses are too small or too infrequent, the proportion of cell kill will gradually decrease as resistant clones emerge. If doses are too large or frequent, treatment-related morbidity will limit how much therapy a patient will tolerate, possibly leading to treatment delays and resistant clone development. Dose density strives to increase doses of chemotherapy and frequency so as to find the limits of toxicity with as much antitumor activity as possible. Still, many effective chemotherapeutic agents with myelotoxicity could be made more effective by further increasing their doses, at the risk of ablating the marrow. The use of high-dose chemotherapy with autologous stem cell rescue has been successfully attempted in selected malignancies such as lymphomas and multiple myeloma in order to benefit from this fact. In solid tumors, however, success has been minimal, suggesting there is a threshold at which tumor response becomes nonlinear.

Combination Chemotherapy

Tumors are more genetically unstable than benign cells, leading to a high rate of random mutation and possible chemotherapy resistance. Mutations occur at a high rate; thus, at the time of a tumor becoming clinically apparent, several drug-resistant clones may already exist. This would explain why a tumor with a dominant clone sensitive to chemotherapy may have initial response but subsequently relapse after therapy. Other tumors are largely chemotherapy-resistant at presentation—for example, melanoma and pancreatic cancer—even with low-volume disease. A possible explanation for de novo resistance is that a slower-growing tumor will take longer to become clinically apparent and, thus, may have undergone a longer undetectable growth phase. This longer period of time may allow more mutation opportunities, and thus once diagnosed, the dominant portion of the tumor may already be resistant to standard therapies. In other malignancies such as gastrointestinal neoplasms, tumors might shed cells into the lumen and, thus, take more time, additional doubling times, and genetic mutations before it achieves a given size.

The rationale for combination chemotherapy was adapted from observations of tuberculosis therapy in the 1950s. If a single agent was used, resistance would eventually develop. If multiple agents with different mechanisms of action were used concurrently, resistance was less common. Frei and coworkers ¹⁴ published a study of combinatorial chemotherapy for leukemia in 1958, one of the first randomized clinical trials. They demonstrated transient responses in adult and pediatric leukemia patients achieved by combining methotrexate and 6-mercaptopurine, thus ushering in the modern era of combinatorial chemotherapy. Today, most patients treated with curative intent are given combination chemotherapy.

Intermittent Chemotherapy

Because tumors have impaired DNA repair mechanisms and are thus more sensitive than normal tissue, they take longer to recover from the insult of chemotherapy. As a result, cytotoxic chemotherapy given at the appropriate interval will allow normal tissue, including hematopoietic stem cells, to recover, whereas tumor will not have sufficient time to recuperate (Figure 3-1). Most cytotoxic therapies are dosed every 2 or 3 weeks based on this observation.

Figure 3-1 The effect of intermittent chemotherapy on tumor and normal cell populations.

Continuous Therapy

Cytostatic agents, such as many of the newer targeted agents, require continuous exposure to maintain disease control. Drugs that have a potent, although short-lived, effect on a tumor will not work well with intermittent dosing because when the drug is removed, the tumor will move from a static to an active phase. Imatinib, an oral agent taken daily, inhibits the BCR-ABL tyrosine kinase, resulting in normalization of blood counts and excellent long-term control in the vast majority of chronic-phase chronic myelogenous leukemia (CML) patients. If the medicine is withdrawn, the majority of patients will again develop the hematologic abnormalities of chronic-phase CML. These agents hold the promise of turning cancer that is currently incurable into a chronic disease, such as diabetes or hypertension.

Route of Administration

The majority of chemotherapy is administered intravenously, in an attempt to eliminate erratic gastrointestinal absorption as well as compliance issues. Chemotherapeutic agents can be given as a bolus (given over a very short interval), short infusion (given over a period of hours), or continuous infusion (usually administered as an inpatient or with a preprogrammed pump for home administration). The advantages of intravenous administration include more standardized absorption, documented compliance, and use of concurrent intravenous hydration or supportive medications.

Many of the newer chemotherapeutic agents target a specific metabolic pathway of the cell, and some of these have eliminated the problems of erratic absorption when administered orally, which has advantages for the patient, including not requiring an intravenous catheter. Other routes of administration include intrathecal (e.g., methotrexate to treat central nervous system [CNS] leukemia relapse), subcutaneous (e.g., alemtuzumab to treat chronic lymphocytic leukemia), intra-arterial (e.g., cisplatin to treat sarcoma), intravesical (e.g., bacilli Calmette-Guérin [BCG] to treat bladder dysplasia), intraocular (e.g., bevacizumab to treat macular degeneration), and intraperitoneal (e.g., to treat ovarian cancer).

Drug Development

Historically, many drugs were developed in the same serendipitous fashion as Alexander Fleming’s penicillin discovery. Currently, the vast majority of new agents are engineered to attack a specific tumor-related target, which requires understanding of disease mechanisms. New agents are created as novel chemical entities owing to knowledge of the molecular basis of their action or by modifying an existing agent. Drug companies perform in vitro screening of multiple similar compounds to evaluate for potential efficacy. Once a lead molecule is identified, nonhuman animal studies are conducted to evaluate for potential toxicity. Once an agent has fulfilled these requirements, clinical trials in humans may begin.

Clinical Development

Phase I studies are the first in human studies of a new agent and have the goal of determining the maximum tolerated dose (MTD) by using dose escalation and dose-limiting toxicities (DLTs). Response rates are typically low because adequate dosing is unknown and patients typically have refractory disease, but a significant portion of patients do receive benefit.¹⁵ Pharmacokinetic and pharmacodynamic studies are performed during early clinical development to better understand the in vivo properties of the drug. Phase II studies use doses and schedules from the phase I data to assess efficacy with response as the primary endpoint. Patients are typically less heavily pretreated and must have measurable disease for response monitoring. Phase II trials typically enroll 20 to 50 patients, are designed for early termination if a significant number of responses are not seen, and may evaluate a new agent alone or in combination with standard chemotherapy.

Phase III trials are larger, typically randomized, and evaluate the experiment treatment against a standard of care regimen. Endpoints are usually progression-free survival (PFS) and/or overall survival (OS). Evaluation of OS may be confounded by patients receiving subsequent effective therapies, and hence, PFS is usually thought to be the more reliable endpoint. Quality of life (QOL) and comparative toxicity data are usually collected. These trials usually require hundreds of patients to achieve their statistical goals and may lead to U.S. Food and Drug Administration (FDA) approval of the new agent.