Principles of Chemotherapy

Published on 13/03/2015 by admin

Last modified 13/03/2015

Print this page

This article have been viewed 1401 times

CHAPTER 112 Principles of Chemotherapy

The first line of treatment for primary brain tumors is surgery and radiation, yet these modalities alone are rarely curative for malignant tumors. Malignant tumors are infiltrative, and local therapy alone is not sufficient to eradicate all tumor cells. Cell cycle kinetic studies have shown that the cells that migrate into normal brain are the most viable and have the highest capacity for proliferation.¹^,² The tumors therefore tend to recur after surgery or local radiation. Chemotherapy may be a means of treating these cells that escape local therapy.

It was only recently that chemotherapy became a standardized part of the treatment for newly diagnosed primary tumors. The landmark trial that established radiation and concurrent temozolomide (TMZ) followed by adjuvant TMZ as the current standard of care for newly diagnosed glioblastoma multiforme (GBM) was published by Stupp and colleagues in 2005.³ The median survival of patients receiving radiation and TMZ was significantly improved at 14.6 months compared with 12.1 months with radiation alone.

Before that major breakthrough, the role of chemotherapy was less well defined. A modest improvement in median survival with the addition of single-agent or combination chemotherapy had been demonstrated in several trials.⁴^–⁶ Despite these positive results, it was difficult to ascertain the effectiveness of chemotherapy as adjuvant therapy, because improvements in survival were modest at best, reports had been conflicting, and many studies had been small. However, larger reviews and a meta-analysis demonstrated a survival benefit for those patients with malignant glioma treated with radiation and adjuvant chemotherapy compared with those treated with radiation therapy alone in the pre-TMZ era.⁶^–⁹

Limitations of Chemotherapy

There are numerous obstacles or difficulties in attempting to treat central nervous system (CNS) tumors with chemotherapy. Malignant primary brain tumors represent a heterogeneous group of diseases. A variety of histologic subtypes may be seen within a single tumor (e.g., astrocytic cells adjacent to oligodendroglial components),¹⁰ and histologically similar tumors may not behave alike and often have different molecular and genetic profiles. This makes it difficult to target therapy to a single disease. In addition, the number of chemotherapeutic agents available to treat the variety of CNS tumors is limited. Since the 1960s, only four drugs (TMZ, carmustine [BCNU], lomustine [CCNU], and bevacizumab) have been approved by the U.S. Food and Drug Administration (FDA) for specific use in brain tumor patients.¹¹

Drug resistance, either intrinsic or acquired, is common in tumors of the CNS, and many different mechanisms have been described. The best-characterized mechanism of drug resistance involves P-glycoprotein. P-glycoprotein is a drug efflux pump normally present on the luminal surface of brain capillary endothelial cells.¹²^,¹³ It protects the neuronal environment of the CNS by excluding xenobiotics, including a variety of anticancer drugs, from the CNS.¹⁴^–¹⁶ In addition to its presence on the endothelial cells of the blood-brain barrier (BBB), P-glycoprotein is overexpressed on the surface of some cancer cells, including brain tumors.¹⁷ The presence of P-glycoprotein on a tumor cell may restrict drug entry into the tumor cell or convey drug resistance.¹⁷ P-glycoprotein–mediated multidrug resistance is characterized by cross-resistance of tumor cells to several classes of chemotherapeutic agents.¹⁸

Poly (ADP-ribose) polymerase (PARP) is an abundant nuclear enzyme that mediates repair of DNA and also cell death. It also plays a role in the pathogenesis of cardiovascular and inflammatory diseases. PARP activation through mild to moderate genotoxic stimuli facilitates DNA repair by signaling cell cycle arrest and interacting with DNA repair enzymes such as XRCC1 and DNA-PK. Therapeutic DNA damage induced by certain chemotherapy agents (alkylating agents and topoisomerase I inhibitors) is largely dependent on PARP activation. As a result, DNA damage is repaired, tumor cells are rescued, and chemoresistance is induced. More severe DNA damage induces apoptotic cell death, whereby PARP is cleaved and inactivated by caspases.¹⁹

Other methods of drug resistance include the ability of tumor cells to repair DNA damage induced by cytotoxic agents or the use of alternative pathways of metabolism.²⁰ O⁶-alkylguanine-DNA-alkyltransferase (AGT) is a DNA repair protein that removes alkyl groups from the O⁶ position of guanine. This is the site on DNA frequently targeted by alkylating agents, including CCNU, BCNU, and TMZ. In a number of human tumor cell lines and animal models, in vitro cytotoxicity of nitrosoureas was inversely related to AGT levels.²¹^–²³ Inactivation of AGT also improves response to temozolomide and nitrosoureas.²⁴^,²⁵

A major obstacle to successful use of chemotherapy for CNS tumors is delivery of drug to the tumor site. CNS tumor response to chemotherapy is limited by inadequate intracellular concentrations of the drug, inadequate drug exposure time, and inadequate drug delivery across the BBB.

Genetic Markers as Predictor of Chemotherapy Response

There is growing evidence that understanding and testing for genetic changes and cellular and molecular mechanisms required for tumor growth can lead to customized treatment for the individual patient by improved selection of therapies.

One example of genetic changes that have affected management of gliomas is the understanding of the role that chromosome 1p-19q codeletion plays in treatment-response and prognosis in patients with oligodendroglial tumors. The frequency of the combined 1p and 19q chromosome deletion in oligodendroglial tumors ranges from 40% to 70%, with higher rates in pure oligodendrogliomas.²⁶ In anaplastic oligodendroglial tumors (AOTs), evidence is accumulating that those with 1p-19q codeletion have improved responsiveness to chemotherapy.²⁷^,²⁸ Two phase III studies suggest that 1p-19q codeletion in AOT patients is an independent prognostic factor of better progression-free survival and overall survival whether they received chemotherapy as neoadjuvant or adjuvant treatment or in recurrence.²⁹^,³⁰ Similarly, 1p-19q codeletion is a strong prognostic factor in low-grade gliomas³¹^,³² and is associated with an improved response to chemotherapy, including procarbazine in combination with CCNU and vincristine (PCV) and TMZ.³³^,³⁴

Another important marker of chemotherapy response in glial tumors is the epigenetic silencing of the AGT gene by promoter methylation. The AGT gene is located on chromosome 10q26 and encodes a DNA-repair protein that removes alkyl groups from the O⁶ position of guanine, an important site of DNA alkylation. As described previously, high levels of AGT create a resistant phenotype to alkylating agents. Conversely, epigenetic silencing of the AGT gene induces low expression of AGT and increased sensitivity to alkylating agents. In the European Organisation for Research and Treatment of Cancer and National Cancer Institute of Canada (EORTC-NCIC) trial, AGT promoter methylation was present in 45% of cases. It was an independent favorable prognostic factor, and those patients benefited the most from the addition of TMZ.²⁵

Principles of Central Nervous System Pharmacology

Systemic Drug Administration

Systemic drugs can be administered by a variety of routes. Some chemotherapeutic agents, such as TMZ, CCNU, and procarbazine, are given orally. Although oral administration is easier, oral therapy is complicated by varying degrees of bioavailability. Bioavailability depends on drug stability in gastric acid, absorption through gastric mucosa, inactivation by intestinal enzymes, hepatic metabolism and biliary excretion, and treatment-induced emesis.³⁵ Administering the drug by an intravenous bolus gives a rapid maximal peak plasma concentration; the concentration then declines rapidly as the drug is distributed, metabolized, and excreted. This is best for agents that are not cell cycle specific (e.g., BCNU, cisplatin). Continuous infusion (intravenous administration for ≥24 hours) is used for agents that are cell cycle specific. In brain tumors, the actively dividing population is relatively small (5% to 10%), so continuous treatment maximizes the number of actively cycling cells that are exposed to sustained effective levels of drug.³⁶

Factors that determine the concentration of drug delivered to a tumor cell in the brain include the rate of blood flow, the amount of drug able to cross the vasculature of the brain, diffusion of the drug across the brain parenchyma, and the amount of drug able to cross the tumor cell membrane. Of these, the BBB may represent the greatest impediment to systemically administered drug delivery to brain tumors.

Blood-Brain Barrier

Systemically administered chemotherapeutic agents must cross the capillary endothelial cells of the cerebral vasculature to access tumor cells in the brain. These capillary endothelial cells differ from other endothelial cells in that they have extended tight junctions, lack fenestrations and pinocytotic vesicles, and express specific transport mediators.³⁷ They form a relatively impermeable barrier between the blood and the brain, the BBB. The function of this single layer of specialized vascular endothelial cells is to maintain a precise neuronal environment, and it does so by limiting the entry of toxins into the brain parenchyma. The passage of substances across the BBB is limited by size, lipid solubility, and ionization state. Most hydrophilic substances and large (molecular weight > 180 daltons) lipophilic substances, which include many chemotherapeutic compounds, are restricted from entering the CNS.³⁸ Smaller lipophilic compounds are able to pass through the BBB, penetrating the lipid membranes by simple passive diffusion. TMZ, nitrosoureas (BCNU, CCNU), and procarbazine are small lipophilic substances that can cross the BBB, and these have been the principal agents used to treat primary brain tumors.

Blood-Tumor Barrier

A similar barrier, the blood-tumor barrier, exists in brain tumor capillaries, limiting the delivery of cytotoxic agents to brain tumors.³⁹^,⁴⁰ The vessels of some brain tumors have abnormally tight junctions and disrupted endothelial surfaces, which allow water-soluble contrast agents to cross the blood-tumor barrier; as a result, the tumor enhances on magnetic resonance imaging. The blood-tumor barrier tends to be more permeable than normal brain tissue, but the amount of drug reaching the tumor is still severely restricted. BBB breakdown in the area of the tumor is variable, and large areas exist where the BBB remains intact.³⁹ In addition, the brain adjacent to tumor contains infiltrating tumor cells. These sites may be associated with an intact BBB, and drug delivery to this area is difficult.²⁰ For moderately or even highly active drugs to ultimately affect the prognosis of patients with brain tumors, methods of delivering higher concentrations of chemotherapeutic agents to the tumor and proximal brain tissue must be developed.

Drug Concentration at the Tumor Site

If a drug is able to cross the BBB, it is then distributed across the brain parenchyma based on the laws of diffusion. The final concentration of a drug in brain tissue or tumor is difficult to predict based on the concentration of the drug in plasma. The drug’s concentration in cerebrospinal fluid (CSF) is often used as a surrogate for its concentration in brain parenchyma because it is easier to obtain samples, but the blood-CSF barrier differs from the BBB, and it is unknown how predictive CSF drug levels are of drug levels in brain interstitium.

Most of the commonly used chemotherapeutic agents work by interfering with DNA synthesis or function. The drug that reaches a tumor cell must be in its free (unbound) active form, be present in high enough concentrations to be effective, and be present for a long enough time to interfere with tumor cell division. Exposure to a drug is measured by the area under the concentration-time curve. High peak levels may not be effective if the drug is cell cycle dependent and rapidly cleared from the plasma. Conversely, a drug that has a prolonged half-life and is slowly cleared from the body may increase tumor exposure but be associated with excessive toxicities.

Approaches to Overcome Obstacles of Chemotherapy

P-Glycoprotein Inhibition to Reverse Multidrug Resistance

A variety of drugs that inhibit P-glycoprotein have undergone preclinical and clinical testing as modulators of multidrug resistance.⁴¹^,⁴² Examples of such agents are verapamil, cyclosporine, and cephalosporins. Because P-glycoprotein is normally expressed in several tissues, including the liver and proximal tubules of the kidney, it is thought to play a role in the elimination of toxins.⁴³^,⁴⁴ In phase I studies, P-glycoprotein modulation decreased the clearance of anticancer drugs that are P-glycoprotein substrates and enhanced their toxicity, presumably because P-glycoprotein–dependent drug elimination pathways were inhibited.⁴⁴ In addition, the maximal tolerated dose of the initial modulators was below levels that would cause significant P-glycoprotein inhibition.⁴⁴ Despite these restrictions, objective responses have been demonstrated in phase II studies of lymphomas and myelomas.⁴⁵^,⁴⁶ Newer, more potent, but less toxic P-glycoprotein modulators, such as cyclosporin PSC 833, are undergoing clinical trials. However, although P-glycoprotein is a theoretically attractive target for modulation to improve the delivery of chemotherapeutic agents to the CNS, the true contribution of multidrug resistance genes to brain tumor chemotherapy resistance remains speculative.

PARP Inhibition

Because of the involvement of PARP in the repair of DNA damage induced by certain anticancer agents or radiation,⁴⁷^,⁴⁸ PARP inhibitors have been investigated as chemosensitizers and radiosensitizers for cancer treatment. GPI 15427 enhances the antitumor activity of TMZ against GBM, brain lymphoma, and intracranial malignant melanoma. When administered intravenously in combination with TMZ, GPI 15427 prolonged survival of tumor-bearing mice with reduction of the tumor mass and the number of sites of brain infiltration.⁴⁹ Another preclinical study demonstrated that the PARP inhibitor CEP-6800 potentiates TMZ activity against subcutaneous GBM xenograft.⁵⁰ A phase I study of the PARP inhibitor AG-014699 showed safety, significant inhibition of PARP, and antitumor activity in patients with advanced solid tumors.⁵¹ A phase I clinical study on the addition of the PARP inhibitor BSI-201 to radiation therapy and TMZ in patients with newly diagnosed malignant glioma is ongoing (see http://www.clinicaltrial.gov).

AGT Inhibition

Alkylating agents, such as CCNU, BCNU, and TMZ, have a predilection for methylating guanine at the O⁶ position.⁵²^,⁵³ The resultant O⁶-methylguanine can be repaired by AGT, which confers resistance to these drugs, as described earlier. O⁶BG is a modulating agent that irreversibly inactivates and depletes AGT.⁵⁴ In in vitro and in vivo tumor models, O⁶BG was able to potentiate the antitumor effects of TMZ.⁵⁵^,⁵⁶ Early clinical trials have combined O⁶BG with TMZ and showed modest tumor response, depletion of AGT, and toxicity limited to myelosuppression.⁵⁷^,⁵⁸ Clinical trials are ongoing to evaluate the efficacy of O⁶BG added to TMZ in recurrent and TMZ-resistant gliomas. Protracted exposure to TMZ leads to MGMT gene consumption and depletes the cell repair mechanism that may overcome the resistance of the tumor to the chemotherapy.⁵⁹ Preclinical studies have demonstrated progressive depletion of MGMT following prolonged exposure to TMZ.⁶⁰ Clinical trials are currently ongoing to assess whether a more intense schedule of TMZ will enhance its intrinsic cytotoxicity in malignant glioma.

Blood-Brain Barrier Disruption

Nonspecific osmotic disruption of the BBB has been clinically tested for the treatment of CNS malignancies. Neuwelt and coworkers demonstrated the feasibility of temporarily opening the BBB through intracarotid injection of osmotic agents to deliver higher doses of chemotherapeutic agents to the brain.⁶¹ The opening of the BBB is produced by shrinkage of endothelial cells and opening of the tight junctions. Cellular messenger systems such as calcium influx and nitrous oxide, as well as cytoskeletal changes, also contribute to this opening. Animal studies suggest that this method increases drug delivery to the brain 10- to 100-fold over injection into the neck arteries without the osmotic solution.⁶² Although a modest number of responses have been seen,⁶³ the procedure can be associated with considerable toxicity; little additional chemotherapy is delivered to the bulk of the tumor, and more chemotherapy is delivered to normal brain areas, which may potentiate the neurotoxicity of chemotherapeutic agents.⁶⁴^–⁶⁶

Biomodulation of specific receptors on the BBB may be a means of achieving therapeutic drug concentrations in the tumor and the brain around the tumor. Lobradimil (Cereport, RMP-7) is a bradykinin analogue that interacts with the B2 (bradykinin) receptor on the BBB, resulting in selective permeability of the BBB at the blood-tumor interface and a small area around the tumor.⁶⁷^,⁶⁸ Phase I and II clinical trials administering lobradimil with the cytotoxic agent carboplatin, either intravenously or intra-arterially, have being conducted in an effort to deliver more of the chemotherapeutic agent to CNS tumors. Minimal efficacy was seen with this combination.⁶⁹^–⁷³

High-Dose Chemotherapy

High-dose chemotherapy followed by autologous bone marrow transplantation or peripheral stem cell transplantation has been used in an attempt to improve the delivery of chemotherapy to CNS tumors. The rationale is increased delivery of chemotherapeutic agents across the BBB secondary to higher serum levels, resulting in greater tumor cell kill, given the steep linear-log dose-response relationship of alkylating agents ⁷⁴ and their cell cycle–independent cytotoxic properties.

High-dose chemotherapy has shown encouraging results, primarily for pediatric patients with relapsed medulloblastoma and intracranial germ cell tumors.⁷⁴ It has also been used in young children to avoid the toxic effects of craniospinal radiation.⁷⁴ The major problem with high-dose chemotherapy is the increased systemic toxicity. Although myelosuppression may be treated with the infusion of stem cells, other toxicities are then unveiled, including neurotoxicity associated with high-dose thiotepa, renal and ototoxicity associated with carboplatin, cardiovascular toxicity associated with cyclophosphamide, mucositis associated with etoposide, and pulmonary and hepatic toxicity associated with the nitrosoureas.⁷⁴ The toxic mortality rate of high-dose chemotherapy for children with brain tumors ranges from 5% to 20%.⁷⁴

High-dose chemotherapy with autologous stem cell rescue is significantly less promising in adults with malignant glioma. In general, response rates and overall survival have not been impressive, and toxicity has been substantial.⁷⁵ This is not particularly surprising, given the fact that high-dose chemotherapy and autologous bone marrow transplantation have proved to be effective only in the most chemosensitive tumors. Unfortunately, most adult gliomas cannot be considered among this group.

Intra-arterial Therapy

Infusion of chemotherapeutic agents directly into the arterial supply of a tumor can provide a pharmacokinetic advantage by enhancing drug delivery to the tumor without increasing systemic drug exposure.⁷⁶^,⁷⁷ This method may be more beneficial with drugs that undergo a large first-pass effect when given systemically and with drugs that have a high total-body clearance.⁷⁸ Local plasma peak drug concentrations and local exposure are augmented compared with an equivalent intravenous dose, with less systemic toxicity.

A study by Levin and colleagues using radiolabeled BCNU demonstrated a fourfold increase in BCNU concentration in the brain after intracarotid administration compared with intravenous administration.⁷⁹ In a study involving 12 adult patients with recurrent glioblastoma, intra-arterial BCNU infusion resulted in a median survival time of 54 weeks after recurrence (92 weeks after diagnosis), but no major improvement in survival was demonstrated in a group of 43 patients treated adjuvantly after receiving irradiation.⁸⁰ It was hypothesized that the group receiving adjuvant intra-arterial BCNU had radiation-related changes that limited drug penetration into tissue,⁸⁰ although the effects of radiation on drug delivery across the BBB have been controversial.

Nitrosoureas, cisplatin, and carboplatin have been given intra-arterially, with some responses noted.⁸¹^–⁸³ However, the responses have been modest, and the technique is associated with significant toxicity. In a study combining both intra-arterial and systemic chemotherapy in patients with grade II to grade IV gliomas, response rates up to 54% were demonstrated, but 31% of patients developed neurotoxicity, and 12% developed serious permanent local toxicity.⁸⁴ No large, randomized trial has yet demonstrated that intracarotid administration of chemotherapy is more advantageous than systemic administration.⁸⁵^,⁸⁶

One difficulty with intra-arterial administration is that there may be nonuniform drug distribution within the brain or the tumor after intracarotid infusion owing to poor mixing or streaming of the drug solution within the artery.⁸⁷ Tissue regions that receive high concentrations of drug are at risk for toxicity, whereas low drug concentrations in other tissue regions may be subtherapeutic.⁸⁸ In addition to increasing drug levels in the tumor, this technique increases drug levels in the area of normal brain supplied by the artery, leading to greater neurotoxicity. For example, intracarotid administration of BCNU has been associated with ocular toxicity, strokes, and an encephalitic picture.⁸⁹

In adults, more brain tumors are supratentorial and are supplied by the carotid arteries, but more than 50% of childhood brain tumors are infratentorial.⁹⁰ Infratentorial tumors are supplied by the vertebrobasilar arterial system, which is unique because of the convergence of the two vertebral arteries into the basilar artery. Cervical spinal cord toxicity has been described with carboplatin and etoposide injected in the vertebral artery in combination with osmotic BBB disruption.⁹¹

Interstitial or Intratumoral Drug Delivery

Given that most CNS tumors recur at, or very close to, the original tumor site,⁹² improving local control may translate into improved survival. Interstitial therapy involves administering drug directly into or adjacent to the brain tumor, circumventing the BBB. The theoretical advantages are increased local drug concentrations; decreased systemic exposure; and high, prolonged levels of the chemotherapeutic agent. Different strategies for intratumoral drug delivery include topical applications, Ommaya therapy, indwelling catheters, implantable drug delivery systems, and biodegradable polymers (wafers).

Ommaya reservoirs are subcutaneously implanted devices that have a catheter outlet in the CNS, usually within the tumor bed or in the ventricular system. Chemotherapeutic agents can be administered percutaneously into the reservoir, which is then manually compressed to deliver the drug to the catheter tip in the tumor. Although convenient, this system does not allow for continuous drug delivery to the tumor bed, and it is also subject to problems such as clotting and infections.

Biodegradable polymer matrices are molded wafers impregnated with a chemotherapeutic agent that is released in an active form for sustained periods in the area of the tumor. This technique increases a drug’s therapeutic efficacy by producing high local tissue concentrations over an extended period, with minimal systemic exposure.⁹³ To effectively deliver drugs to the brain, several characteristics of the polymer are necessary: a biocompatible and biodegradable matrix, dependable and reproducible drug release, the ability to maintain the bioactivity of the therapeutic agent, and ease of surgical handling. Given the short plasma half-life of BCNU and its significant systemic toxicity, BCNU-impregnated biodegradable wafers have been studied.⁹⁴ Wafers containing 3.8% BCNU allow sustained-release delivery and higher local concentrations for extended periods. Preclinical studies demonstrated that BCNU could be released from polymer disks for up to 21 days in rat and rabbit brain.⁹⁵^,⁹⁶ Rats with implanted 9L gliomas that received BCNU-impregnated polymers demonstrated improved survival compared with controls with empty polymers.⁹⁷

Gliadel wafers (MGI Pharma, Eisai, Woodcliff Lake, NJ) are now a surgical option for patients with newly diagnosed and recurrent single lesions compatible with malignant glioma without any communication with the ventricular system for which a nearly gross-total resection is possible.⁹⁸ After pathologic confirmation of malignant glioma and maximal surgical tumor resection, up to eight wafers can be implanted in a patient depending on the size of the surgical cavity. Adjuvant treatment can be used in combination with the wafers (see later).

In a phase I trial of BCNU loaded in the PCPP-SA polymer in patients with recurrent glioma, the overall median survival was 46 weeks after implantation (87 weeks after initial diagnosis), with 86% of patients alive more than 1 year after diagnosis.⁹⁹ No significant systemic toxicities were observed during this study. One major complication was the later development of neurological deterioration in a significant number (47%) of patients, which was thought to be due to development of necrosis around the wafer.

Gliadel has been further evaluated both as a therapy for newly diagnosed GBM ¹⁰⁰^,¹⁰¹ and as a salvage therapy for recurrent GBM.¹⁰² A recent meta-analysis of the use of Gliadel for high-grade glioma concluded that the addition of Gliadel to radiotherapy led to increased survival relative to radiotherapy alone for newly diagnosed tumors but that there was no significant increase in survival with Gliadel therapy for recurrent disease.¹⁰³ In a phase III trial randomizing patients with recurrent glioma to 3.85% BCNU polymer or empty polymer, there was a minimal but statistically significant improvement in median survival in the BCNU-treated group compared with the control group (31 weeks versus 23 weeks).¹⁰² Significant complications due to necrosis did not occur in this study.¹⁰⁴

Buy Membership for Neurosurgery Category to continue reading. Learn more here