Radiobiologic Principles and the Role of Radiotherapy in Hematopoietic Cell Transplant and Chimeric Antigen Receptor T-Cell Therapy

Radiotherapy: Mechanisms of Action

External beam radiation therapy machines produce ionizing radiation in one of two ways. The first is by radioactive decay of a nuclide, historically ⁶⁰ Co. Over the past several decades in developed countries, treatment machines now more commonly generate ionizing radiation using linear accelerators. Linear accelerators accelerate electrons to a high energy, and these electrons are either directed to a scattering foil (for treatment with electrons) or a target with a high number of protons (high Z) to produce x-rays through bremsstrahlung interactions. X-rays produced by these electron interactions are also referred to as photons . Photons are the most common form of therapeutic radiation, in contrast to charged particles (protons, carbon ions) or neutrons, which have a more limited role in the treatment of hematologic malignancies and will not be discussed further in this chapter. Upon exit from the gantry head, the beam is refined using collimators to deliver a conformal dose to the target.

Radiation-mediated damage can be direct, with incident radiation resulting in the formation of a deoxyribonucleic acid (DNA) free radical, or indirect, through the production of an intermediate free radical. This intermediate free radical is most often a hydroxyl radical given the abundance of water molecules. The majority of cell killing by radiation is caused by its indirect effects. Free radicals are unstable and quickly undergo further reactions to gain electrons, or lose unpaired electrons, which results in clusters of DNA damage. Specifically, ionizing radiation leads to single and double strand breaks, DNA-DNA crosslinks, nucleotide loss, and base damage. The sensitivity of individual cells to ionizing radiation depends on the phase of the cell cycle, with cells in M and late G2 more sensitive because of a greater amount of DNA present (approaching twice the normal amount at the time of division), while those in late S phase are the most resistant because of enhanced homologous recombination repair during DNA replication.

These DNA repair mechanisms are complex and allow for sublethal damage recovery both in normal tissue and tumor cells. As repair mechanisms are typically more effective in normal cells compared to tumor cells, fractionation is a primary method of enhancing the therapeutic ratio, balancing tumor control with normal tissue toxicity. The radiobiologic basis for fractionation includes the 4 Rs: repair of sublethal damage, reassortment, repopulation, and reoxygenation. As cells vary in radiosensitivity during their cell cycle, an initial dose of radiation results in an asynchronous population of surviving cells, enriched for those in S-phase. As cells then redistribute, the population as a whole becomes more radiosensitive. Repopulation has been observed in rapidly dividing cells, approximately 4 weeks following initiation of RT, and an additional 0.6 Gy per day is needed to compensate for this phenomenon. Hypoxic tumor cells are more radioresistant, as the presence of oxygen forms peroxides, which enable fixation of radiation damage. This occurs even at very low oxygen concentrations; the k-value of oxygen (oxygen tension at which the radiosensitivity is halfway between that of an anoxic and fully aerobic environment) is approximately 3 mm Hg or 0.5%. In the absence of oxygen, endogenous free radical scavengers such as glutathione can donate hydrogen to quickly repair damage caused by radiation. Cell survival is impacted by all of these variables and can be described for a specific normal tissue or tumor type using the α/β ratio, which is the dose at which the linear and quadratic components of cell killing are equal. The biologically effective dose (BED) is used to compare different radiation treatments, taking into account the α/β ratio, radiation dose and fractionation, as well as dose rate factor to account for intracellular repair for treatments delivered over time. In tissues with large capacity for sublethal damage repair, both fractionation and low dose-rate treatment can decrease treatment related toxicity.

Radiation and the Immune Microenvironment

Radiation has been shown to have an immunomodulatory effect on the tumor microenvironment, with resulting effects on leukocyte recruitment. Double strand breaks as a result of radiation-mediated damage activate DNA damage response with activation of ataxia telangiectasia mutated, and downstream activation of p53 and nuclear factor (NF)-κβ. NF-κβ subsequently induces expression of proinflammatory cytokines including but not limited to tumor necrosis factor (TNF)α, interleukin (IL)-1α, IL-1β, and IL-6, which induce upregulation of adhesion molecules (ICAM-1, VCAM-1, and E-selectin) and facilitate leukocyte infiltration.

The unirradiated tumor microenvironment harbors a baseline inflammatory, immunosuppressive state that facilitates tumor growth and angiogenesis, mediated by specific subsets of myeloid derived cells including tumor associated macrophages (TAMs) and myeloid derived suppressor cells (MDSCs). Both produce inflammatory cytokines and also generate reactive nitrogen and oxygen radicals that disrupt T-cell receptor signaling. TAMs also directly suppress T-cell activity through IL-10 and transforming growth factor (TGF)-β, while MDSCs have a similar effect through secretion of IL-6 as well as by limiting availability of amino acids such as tryptophan and cysteine. A subset of CD4+ T-cells, T regulatory cells (Tregs), also exert an immunosuppressive effect through TGF-β and IL-10, which can directly inhibit effector T-cell activity. Tregs are recruited by CCL22 expressed by tumor cells and TAMs, and in turn stimulate both TAMs and MDSCs.

Radiation enhances T-cell infiltration into the tumor microenvironment through TNF, IL-1, and endothelial cell adhesion molecules, and T-cells themselves secrete interferon (IFN)-γ, which further increases expression of VCAM-1 in a positive feedback loop. Radiation is key to augmenting tumor antigenicity, which can occur both by directly increasing tumor-specific antigen production and presentation, as well as improved recognition by cytotoxic T-cells. The expression of major histocompatibility complex (MHC) I on the cell surface also increases for several days following radiation, which in turn is further upregulated by IFN-β secretion in an autocrine or paracrine manner. This upregulation is critical as peptide-MHC complex recognition by effector T-cells is necessary for cytotoxic, antitumor cell activity. Radiation-mediated cell death also facilitates antigen sampling by dendritic cells as well as dendritic cell maturation. Dendritic cells can induce cytotoxic T-cell responses, while also initiating CD4+ T-cell responses to increase antigen presentation and further stimulate cytotoxic T-cell function.

Radiobiology of Hematopoietic Cells and Organs at Risk

The radiosensitivity of normal and neoplastic cells varies widely. Hematopoietic cells and lymphocytes in particular are very radiosensitive, with D ₀ (defined as the dose to reduce survival by 67% on the linear portion of the cell survival curve) generally estimated to be between 0.5 and 1.5 Gy (though up to 5 Gy has also been reported), compared to 10 Gy for granulocytes, with generally little to no shoulder suggesting limited repair. Cell survival assays of leukemic cells suggest even greater radiosensitivity compared to normal hematopoietic cells, however with some variation based on specific leukemia type. For example, acute myeloid leukemia (AML) cells appears to have very limited repair capacity and thus may be more sensitive to radiation compared to leukemic lymphoid cells, but fractionation for chronic myeloid leukemia (CML) may be associated with increased risk of relapse. Bone marrow stromal cells, which are critical to supporting engraftment, may have sublethal damage repair capacity that would favor fractionation, though the data is not entirely consistent. In myeloablative regimens, doses of 10 to 16 Gy are needed and at these doses, pulmonary, renal, and hepatic toxicity are of key importance. Given the sublethal repair ability of these tissues, for example, the low α/β ratio of 3 to 6 for lung tissue, increasing the number of fractions (commonly to daily or twice daily) and decreasing the dose rate are used to decrease the risk of toxicity and improve the therapeutic index. In general, the volume of these organs receiving low doses, even of 5 Gy, is minimized though a mean dose of 5 to 9 Gy is often permitted.

Total Body Irradiation: Myeloablative Regimens

Myeloablative conditioning typically includes alkylating agents with or without total body irradiation (TBI). TBI has been widely used as a component of HCT conditioning regimens for several decades and allows comprehensive treatment of sanctuary sites including the CNS and testes, is agnostic to pharmacokinetics of drug delivery, and avoids cross-resistance in chemorefractory disease. E. Donnall Thomas pioneered the development of TBI, demonstrating the effectiveness of high-dose cytarabine before 10 Gy TBI delivered in a single fraction to patients with AML or acute lymphoblastic leukemia (ALL); 94% of patients engrafted successfully and the majority of patients did not experience relapse. In his early work, the choice of 10 Gy was based on data in canine models that suggested doses of 8 Gy or less were not sufficient for consistent engraftment. However, the majority of patients treated with 10 Gy TBI in a single fraction developed interstitial pneumonitis, characterized by a syndrome of fever, interstitial pulmonary infiltrates, and severe hypoxia, which was the proximate cause of death in over one-third of patients. Graft-versus-host disease (GVHD) was seen in three-fourths of patients and was also implicated as a cause of death in over one-third of patients.

Subsequently, the same group investigated whether radiation, if delivered in a fractionated manner, could decrease transplant-related mortality based on radiobiologic principles suggesting normal tissue repair. They randomized 53 patients in first remission to TBI delivered at 10 Gy in one fraction or 12 Gy in six daily fractions (2 Gy per day) before allogeneic matched-sibling transplant and found that fractionated TBI was associated with improved survival because of decreased risk of nonleukemic death. Due ot limited numbers, the authors were unable to demonstrate differences in rate of pneumonitis, GVHD, or specific causes of death between the two regimens.

However, following similar biologic principles, the effect of dose rate and hyperfractionation have also been explored. Retrospective review of 77 patients demonstrated that low dose rate (< 6 cGy/min) TBI may be associated with decreased pulmonary toxicity without impacting overall survival (OS) or relapse, although total dose and conditioning regimen were also important. Hyperfractionation (2–3 fractions per day) has been investigated on a Phase I/II trial, which did not demonstrate improvement compared to historical results. Intensified hyperfractionated radiation to 10.2 Gy in six fractions delivered twice daily has been prospectively compared to 12 Gy in four daily fractions, with the intensified regimen associated with a decreased risk of progression without increasing the risk of pulmonary or other acute or late toxicity, however with no improvement in OS. A randomized trial comparing 15.75 Gy in seven daily fractions to 12 Gy in six fractions demonstrated that while the higher TBI dose improved relapse-free survival (RFS), this was at the cost of increased acute GVHD, which in turn likely increased transplant-related mortality and offset the benefit in RFS; OS was similar between the two TBI groups. Similar findings have been demonstrated in observational series comparing 12 Gy to 13–14 Gy, with decreased risk of relapse with higher doses, however, higher nonrelapse mortality translating to a lack of OS benefit. This was also seen in a retrospective analysis of over 1000 patients treated in the Netherlands, in which TBI dose was stratified based on BED using an α/β of 10; 6 to 12 Gy were delivered in one or two fractions, with corresponding BEDs of ≤ 14, greater than 14 to ≤ 18, and > 18 Gy. This analysis demonstrated an increase in nonrelapse mortality and decrease in relapse rate at higher doses, with no difference in OS or RFS.

Of note, multiple randomized trials have compared busulfan and cyclophosphamide (Bu-Cy) with cyclophosphamide and TBI (Cy-TBI), with differing results with regard to disease-free survival (DFS) and OS. Updated metaanalysis of the four trials, limited to 488 analyzable patients with CML and AML, demonstrated that there was no statistically significant difference in DFS or 10-year survival between Bu-Cy compared to Cy-TBI before transplant. On multivariable analysis, there was a significant difference in the incidence of cataract formation (hazard ratio [HR], 2.3; P = .004) but no difference in pulmonary complications. Thiotepa-based regimens have been retrospectively compared with Cy-TBI in adults with ALL and may be associated with increased relapse (HR, 1.78, P = .03) but not necessarily a difference in OS.

Based on the aforementioned data, the preferred myeloablative TBI dose at our institution is 12 Gy delivered at 3 Gy per day, once or twice daily (12 Gy over eight fractions if twice daily), depending on the clinical scenario, with lung and kidney blocks for 3 Gy ( Fig. 12.1 ). Additional regimens include 12 Gy delivered at 2 Gy twice daily (4 Gy per day), 12 to 13.5 Gy in 1.5 Gy fractions delivered twice daily, or 12 to 13.2 Gy in 1.2 Gy fractions delivered 3 times daily.