Clinical Investigation
Distribution of Proliferating Bone Marrow in Adult Cancer Patients Determined Using FLT-PET Imaging

Presented at the 51st Annual Meeting of the American Society of Radiation Oncology, Chicago, IL.
https://doi.org/10.1016/j.ijrobp.2009.11.040Get rights and content

Purpose

Given that proliferating hematopoietic stem cells are especially radiosensitive, the bone marrow is a potential organ at risk, particularly with the use of concurrent chemotherapy and radiotherapy. Existing data on bone marrow distribution have been determined from the weight and visual appearance of the marrow in cadavers. 18F-fluoro-l-deoxythymidine concentrates in bone marrow, and we used its intensity on positron emission tomography imaging to quantify the location of the proliferating bone marrow.

Methods and Materials

The 18F-fluoro-l-deoxythymidine positron emission/computed tomography scans performed at the Peter MacCallum Cancer Centre between 2006 and 2009 on adult cancer patients were analyzed. At a minimum, the scans included the mid-skull through the proximal femurs. A software program developed at our institution was used to calculate the percentage of administered activity in 11 separately defined bony regions.

Results

The study population consisted of 13 patients, 6 of whom were men. Their median age was 61 years. Of the 13 patients, 9 had lung cancer, 2 had colon cancer, and 1 each had melanoma and leiomyosarcoma; 6 had received previous, but not recent, chemotherapy. The mean percentage of proliferating bone marrow by anatomic site was 2.9% ± 2.1% at the skull, 1.9% ± 1.2% at the proximal humeri, 2.9% ± 1.3% at the sternum, 8.8% ± 4.7% at the ribs and clavicles, 3.8% ± 0.9% at the scapulas, 4.3% ± 1.6% at the cervical spine, 19.9% ± 2.6% at the thoracic spine, 16.6% ± 2.2% at the lumbar spine, 9.2% ± 2.3% at the sacrum, 25.3% ± 4.9% at the pelvis, and 4.5% ± 2.5% at the proximal femurs.

Conclusion

Our modern estimates of bone marrow distribution in actual cancer patients using molecular imaging of the proliferating marrow provide updated data for optimizing normal tissue sparing during external beam radiotherapy planning.

Introduction

Decisions regarding treatment with radiotherapy (RT) often hinge on striking an acceptable balance between the potential efficacy and toxicity of treatment. One of the organs at risk as a result of treatment with RT is the bone marrow. The hematopoietic stem cells in the bone marrow that are continually replacing circulating peripheral blood cells are among the most radiosensitive cells in the body. Data exist suggesting that radiation doses as low as 2–4 Gy delivered within 1–3 days can cause a significant reduction in bone marrow cellularity (1) and proliferation (2), and doses of 30–40 Gy in conventional fractionation can lead to complete ablation of the bone marrow 3, 4. In addition to the radiation dose, the volume of bone marrow irradiated is an important factor. As the volume of actively proliferating bone marrow irradiated increases, the total hematopoietic output decreases precipitously. Taken to its extreme, only 4.5 Gy of total body irradiation is estimated to be necessary to cause death in 50% of humans (5). Depending on the dose delivered and volume treated, it can also take a considerable period for the bone marrow to recover. In a study of Hodgkin's patients treated with total nodal irradiation, Rubin et al. (6) found that bone marrow suppression frequently lasted for ≥1 year.

Although many larger volume RT techniques have recently fallen out of favor (e.g., total nodal irradiation, hemibody irradiation, whole abdominal radiation), newer techniques such as intensity-modulated RT (IMRT) are delivering lower radiation doses to ever-enlarging volumes of adjacent normal tissues, including the bone marrow. In addition, myelosuppressive chemotherapy is now frequently delivered concurrently with RT, which, when given together, can place greater stress on the bone marrow than either alone.

The available data regarding the distribution of actively proliferating bone marrow in adults might not be entirely accurate. Most of these data were studied >50 years ago and were determined by weighing bones from cadavers before and after heating and visual inspection of the color of the marrow 7, 8, 9, 10, 11, 12. Although data have been published using radionuclides such as 59Fe and 52Fe to estimate the erythroid component and 99mTc-labeled nanocolloids to image the distribution of reticuloendothelial cells, most of these data either provided no quantitative information regarding the distribution of the bone marrow or reflected only a component of hematopoietic marrow function 6, 13, 14.

To date, most positron emission tomography (PET) in oncology has used 18F-flurodeoxyglucose (FDG), which is taken up more readily by metabolically active tissues, including most malignancies. FDG probably does reflect marrow proliferative activity to some extent, as demonstrated by the increased uptake in response to growth factor stimulation, including the use of granulocyte colony-stimulating factor. However, areas known to harbor proliferating bone marrow generally have relatively low FDG uptake, limiting its ability to quantify the bone marrow distribution. More recently, researchers using PET imaging in oncology have begun to explore the use of other radiotracers, including 18F-fluoro-l-deoxythymidine (FLT) 15, 16, 17, 18. FLT is phosphorylated by the enzyme thymidine kinase 1, which leads to intracellular trapping (19). Because the thymidine kinase 1 concentration increases almost 10-fold during the S phase of the cell cycle, during which DNA synthesis occurs, the FLT uptake intensity on PET imaging reflects the proliferation rate of both normal and malignant cells (20), including the proliferating bone marrow (2).

In addition to an increase in the range of radiotracers available for use during PET scanning, advances have also been made in how the PET data are collected. In particular, most current PET scans are now obtained with computed tomography (CT) (i.e., PET/CT). This advance has allowed, not only for better PET images as a result of improved tissue attenuation correction, but also, and equally important, the ability to link radiotracer uptake to specific anatomic structures, including bone marrow cavities.

Using existing pretreatment, whole-body FLT PET/CT scans performed on cancer patients, we sought to use the distribution and intensity of uptake in the skeleton to quantify the distribution of actively proliferating bone marrow in defined bony regions in adult cancer patients to provided updated data for optimizing normal tissue sparing during external beam RT planning.

Section snippets

Methods and Materials

The FLT PET/CT scans of eligible patients who had undergone scanning at the Peter MacCallum Cancer Centre were analyzed. To be eligible, the patients had to be ≥18 years old, have a pathologically diagnosed, or high clinical suspicion (e.g., FDG-PET–positive solitary pulmonary nodule thought to be consistent with non–small-cell lung cancer) of a, nonhematologic malignancy, and to have already undergone a pretreatment whole-body FLT PET/CT scan (i.e., from at least the base of skull through the

Results

The characteristics of our sample population are listed in Table 1. The study population consisted of 13 patients, 6 of whom were men. At FLT scans, they ranged in age from 36 to 85 years (median, 61). Of the 13 patients, 9 had non–small-cell lung cancer (Stage I in 4, Stage III in 3, and Stage IV in 2), 2 had Stage IV colon cancer, and 1 each had Stage IV melanoma and leiomyosarcoma. In 3 of the 4 patients with Stage I non–small-cell lung cancer, the diagnosis had not been confirmed

Discussion

Although previous estimates of bone marrow distribution were determined using the weight and visual appearance of bone marrow in cadavers, our estimates were determined from molecular imaging of proliferating bone marrow in actual cancer patients. Therefore, they are likely to be more applicable to patients undergoing planning for treatment with external beam RT.

It is illustrative to contrast our use of functional imaging with the approaches taken in previous analyses of the distribution of

Acknowledgments

The authors would like to acknowledge the assistance provided by Steven Kronenberg in the preparation of the figures for this report.

References (27)

Cited by (0)

Note—An online CME test for this article can be taken at http://asro.astro.org under Continuing Education.

Conflict of interest: none.

View full text