Tumor control probability for selective boosting of hypoxic subvolumes, including the effect of reoxygenation

Abstract

$\text{Purpose}$: To study the effect on tumor control probability of selectively boosting the dose to hypoxic subvolumes.

$\text{Methods and Materials}$: A Monte Carlo model was developed that separates the tumor into two compartments, one of which receives a primary dose, and one of which receives a higher boost dose. During radiation delivery, each compartment consists of three clonogen subpopulations: those that are well oxygenated, those that are temporarily hypoxic (geometrically transient hypoxia), and those that are permanently hypoxic (geometrically stable hypoxia). The spatial location of temporary hypoxia within the tumor volume varies over time, whereas, the spatial location of permanent hypoxia does not. The effect of reoxygenation was included. Clonogen proliferation was not included in the model.

$\text{Results}$: A modest boost dose (120%–150% of the primary dose) increases tumor control probability to that found in the absence of permanent hypoxia. The entire hypoxic subvolume need not be included to obtain a significant benefit. However, only tumors with a geometrically stable hypoxic volume will have an improved control rate.

$\text{Conclusions}$: Tumors with an identifiable geometrically stable hypoxic volume will have an improved control rate if the dose to the hypoxic volume is escalated. Further work is required to determine the spatiotemporal evolution of the hypoxic volumes before and during the course of radiotherapy.

Introduction

Traditionally, tumor volumes have been irradiated to uniform doses. However, recent theoretical work 1, 2, 3 and clinical experience with brachytherapy and stereotactic techniques have shown that uniform dose distributions are not necessary to obtain good clinical results. It has been demonstrated that modest underdoses to small portions of a tumor volume do not produce a significant reduction in tumor control probability (TCP) (2). Clinical experience demonstrates that “hot spots” can be tolerated if isolated from critical structures; furthermore, theoretical work suggests that hot spots may increase the TCP (2). In addition, the tumor itself is not homogeneous and contains variation in clonogen density, radiosensitivity, and metabolic rate. Radioresistant subpopulations, particularly hypoxic clonogens, may determine the outcome of radiotherapy (4). Recent advances in functional imaging of tumors have made it possible to identify regions that are at a higher relative risk for recurrence after radiation therapy 5, 6. Such functional techniques include PET imaging with ¹⁸F-labeled fluorodeoxyglucose to determine metabolic rate and thus tumor burden (or clonogen density) (7), magnetic resonance spectroscopy to measure the choline:citrate ratio, which may be indicative of active tumor 8, 9, and PET or SPECT imaging with ¹⁸F-labeled fluoromisonidazole to determine hypoxia (10). The advances in functional imaging, coupled with clinical observations and theoretical work, have prompted a re-examination of the goal of homogeneous dose distributions.

Potential targets for increased dose are the hypoxic subvolumes of a tumor. It has been demonstrated that solid tumors contain regions of low oxygen partial pressures (Po₂) 11, 12, 13, resulting from the rapid proliferation of tumor cells. Such rapid proliferation produces an increased demand for oxygen that inefficient tumor microvasculature is unable to adequately deliver 13, 14. It is well established that hypoxic cells are less sensitive to radiation (15). Furthermore, it has been shown for cervix cancer that low median tumor Po₂ is a prognostic indicator for lower overall survival and recurrence-free survival (12). There is evidence that modification of hypoxia significantly improves locoregional tumor control after radiotherapy. Overgaard and Horsman (16) performed a detailed survey of 83 clinical trials that encompassed a range of approaches to hypoxia modification, including hyperbaric oxygen, hypoxic radiosensitizers, oxygen or carbogen breathing, and blood transfusion. The survey contained 10,779 patients and a variety of tumor sites. The main conclusion of the survey is that treatment of hypoxia improves local control and overall survival, particularly for squamous cell carcinomas of the head and neck.

Several radiolabeled pharmaceuticals that are selectively taken up by hypoxic cells have been investigated (17), including ¹⁸F-labeled fluoromisonidazole (18), ¹⁸F-labeled fluoroerythronidazole (19), ¹²³I-labeled iodoazomycin (20), and ⁶⁴Cu-labeled Cu(II)-diacetyl-bis(N⁴-methylthiosemicarbazone) (21). Functional imaging of the hypoxic volume allows delineation of a hypoxic biologic target volume (hBTV) that may be dose escalated relative to nonhypoxic tumor using intensity modulated radiotherapy. The feasibility of using intensity modulated radiotherapy to deliver a boost to a hypoxic volume has been demonstrated by Chao et al. (22). It is important to consider that hypoxic volumes may not be spatially fixed over time and that a single pretherapy imaging study cannot distinguish geometrically transient from geometrically stable hypoxia. Therefore, an hBTV defined using a single imaging study will contain volumes that were hypoxic at the time of the imaging study, but not at the time of therapy. At the time of therapy, a volume of the tumor not contained within the hBTV may be hypoxic. The geometric stability of the hypoxic volumes will influence the outcome of radiotherapy.

Several authors have examined the consequences of radiation therapy to heterogeneous tumors in which the cell radiosensitivity or clonogen density varies within the tumor volume 1, 2, 3, 23, 24, 25. Tome and Fowler have examined the boosting of subvolumes for tumors in which the radiosensitivity of the clonogens varies throughout the volume. They conclude that modest boost doses to an arbitrary subvolume of the tumor produce an improvement in TCP (3). To obtain a significant improvement, approximately 50% or more of the volume must receive the boost dose. Motivated by the recent advances in imaging and treatment techniques, we investigated the selective boosting of hypoxic subvolumes. We include the effect of reoxygenation, which changes the radiosensitivity over the course of radiotherapy. Hypoxic volumes may not be spatially fixed over time, and so the effect of geometrically transient vs. geometrically stable hypoxia was investigated.