- Email: [email protected]
Hematological Toxicity After Robotic Stereotactic Body Radiosurgery for Treatment of Metastatic Gynecologic Malignancies
International Journal of
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation: Gynecologic Cancer
Hematological Toxicity After Robotic Stereotactic Body Radiosurgery for Treatment of Metastatic Gynecologic Malignancies Charles A. Kunos, MD, PhD,* Robert Debernardo, MD,y Tomas Radivoyevitch, PhD,z Jeffrey Fabien, MS,* Donald C. Dobbins, RT, CMD,* Yuxia Zhang, MS,* and James Brindle, PhD* Departments of *Radiation Oncology, yObstetrics and Gynecology, and zEpidemiology and Biostatistics, University Hospitals Case Medical Center and Case Western Reserve University School of Medicine, Cleveland, Ohio Received Nov 16, 2011, and in revised form Feb 9, 2012. Accepted for publication Feb 10, 2012
Summary Abdominopelvic relapse of gynecologic cancers after radiation and chemotherapy are difficult to manage due to an adverse hazard of bone marrow toxicity. A total of 61 women with stage IV gynecologic malignancies were found, at retrospective review, to have undergone abdominopelvic stereotactic body radiosurgery safely and without adverse bone marrow toxicity. Findings support the novel integration of stereotactic body radiosurgery into treatment strategies for relapsed abdominopelvic gynecologic cancers.
Purpose: To evaluate hematological toxicity after robotic stereotactic body radiosurgery (SBRT) for treatment of women with metastatic abdominopelvic gynecologic malignancies. Methods and Materials: A total of 61 women with stage IV gynecologic malignancies treated with abdominopelvic SBRT were analyzed after ablative radiation (2400 cGy/3 divided consecutive daily doses) delivered by a robotic-armed Cyberknife SBRT system. Abdominopelvic bone marrow was identified using computed tomography-guided contouring. Fatigue and hematologic toxicities were graded by retrospective assignment of common toxicity criteria for adverse events (version 4.0). Bone marrow volume receiving 1000 cGy (V10) was tested for association with post-therapy (median 32 days [25%-75% quartile, 28-45 days]) white- or red-cell counts, hemoglobin levels, and platelet counts as marrow toxicity surrogates. Results: In all, 61 women undergoing abdominopelvic SBRT had a median bone marrow V10 of 2% (25%-75% quartile: 0%-8%). Fifty-seven (93%) of 61 women had received at least 1 preSBRT marrow-taxing chemotherapy regimen for metastatic disease. Bone marrow V10 did not associate with hematological adverse events. In all, 15 grade 2 (25%) and 2 grade 3 (3%) fatigue symptoms were self-reported among the 61 women within the first 10 days post-therapy, with fatigue resolved spontaneously in all 17 women by 30 days post-therapy. Neutropenia was not observed. Three (5%) women had a grade 1 drop in hemoglobin level to <10.0 g/dL. Single grade 1, 2, and 3 thrombocytopenias were documented in 3 women. Conclusions: Abdominopelvic SBRT provided ablative radiation dose to cancer targets without increased bone marrow toxicity. Abdominopelvic SBRT for metastatic gynecologic malignancies warrants further study. Ó 2012 Elsevier Inc.
Reprint requests to: Charles A. Kunos, MD, PhD, Department of Radiation Oncology, University Hospitals of Cleveland, 11100 Euclid Ave,
Int J Radiation Oncol Biol Phys, Vol. 84, No. 1, pp. e35ee41, 2012 0360-3016/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.ijrobp.2012.02.027
LTR 6068, Cleveland, Ohio 44106 USA. Tel: þ216-844-3103; Fax: þ216844-2005; E-mail: [email protected] Conflict of interest: none.
e36
International Journal of Radiation Oncology Biology Physics
Kunos et al.
Introduction
Table 1 Baseline characteristics of stereotactic body radiosurgery study patients
Robotic stereotactic body radiosurgery (SBRT) distinguishes itself from other radiation delivery systems in its steep radiation dose drop-off, allowing narrow “pencil beam” treatment fields to deliver ablative radiation treatment condensed into a few radiation treatment sessions (1-5). One robotic SBRT system, the Cyberknife system (Accuray [Sunnyvale, CA]), uses an industrial frameless robotic arm mounted by a radiation linear accelerator to administer precise ablative radiation dose while simultaneously tracking and following cancer-target motion in real time (6). Sizeable reductions in radiation field margins can be realized by tracking the movement of the target, in the end giving a reduced planning target volume irradiated with submillimeter accuracy (7). However, ablative SBRT introduces the possibility of greater risk of normal tissue injury manifested in organs such as radiosensitive bone marrow. The ability of SBRT to lower bone marrow radiation dose when targeting abdominopelvic sites of disease has not been fully investigated. Quantifying the impact of SBRT upon bone marrow is needed when targeting nearby anticancer targets, especially in the era of widespread clinical use of and novel testing of anticancer chemotherapeutics that may have bone marrow-taxing effects. The purpose of this study was to characterize, for the first time, the abilities of SBRT to lessen radiation-related hematological toxicities when treating women with metastatic abdominopelvic gynecologic malignancies.
Methods and Materials Patients Inclusion criteria for study were female sex, age 18 years or more, and a diagnosis of metastatic gynecological cancer occurring in an abdominopelvic site. A total of 61 women met these criteria during a retrospective study selection among 69 women treated between 2007 and 2011 within the University Hospitals of Cleveland and Case Western Reserve University (Cleveland, OH) stereotactic body radiosurgery program (Fig. 1). Patient demographics, gynecologic cancer type and number of sites to be treated by SBRT, and administered prior therapies are listed in Table 1. All patients provided written informed consent before SBRT. Institutional review board approval was granted for this study.
Simulation One week before SBRT simulation, women had either direct visual or computed tomography (CT)-guided implantation of at
Fig. 1.
Selection of study patients.
Characteristic Female sex: n (%) 61 (100) Age: years Median 64 Range 27-87 Race/ethnicity: n (%)* White 53 (87) Black or African American 5 (8) Hispanic 3 (5) GOG performance status: no (%)y 0 44 (72) 1 10 (16) 2 7 (12) No. of metastases to be treated by radiosurgery: n (%) 1 25 (41) 2 16 (26) 3 15 (25) 4 5 (8) 4 0 (0) Histopathology: n (%) Cervical/vaginal squamous cell carcinoma 10 (16) Endometrial adenocarcinoma 16 (26) Ovarian adenocarcinoma 34 (56) Vulvar squamous cell carcinoma 1 (2) Prior radiation: n (%) 28 (46) Inclusive of radiosurgery site 23 (38) >5000 cGy delivered pre-therapy 15 (25) to radiosurgery site Prior chemotherapy: n (%) 57 (93) No. of prior courses of chemotherapy for metastases: n (%) 0 5 (8) 1 38 (62) 2 15 (25) 3 3 (5) Platinum-containing regimen: no. (%) 57 (93) Taxane-containing regimen: n (%) 44 (72) Anthracycline-containing regimen: no. (%) 13 (21) Topotecan-containing regimen: n (%) 8 (13) Gemcitabine-containing regimen: n (%) 9 (15) * Race/ethnicity was self-reported. y Gynecologic Oncology Group (GOG) performance status reflects individual daily living activities on a scale of 0 (fully active with symptoms) to 5 (dead).
least 3 single 1.6 3 mm soft tissue gold seed fiducials within 4-6 cm of the SBRT target(s) under anesthesia (1). For immobile paraortic nodal metastases, rigid bony landmarks of the spine may have been used for tracking purposes. Women were treated in a head-first supine treatment position with arms at their sides or on their chest while lying on a frameless flat tabletop. Two-pin knee sponge tabletop registry was used for reproducible orthogonal body alignment. Evacuated vacuum-bag immobilization was used when it was anticipated that radiosurgical targets would have large intrafraction excursion during radiation dose delivery. To account for target respiratory motion, women wore a firmly conforming vest around the thorax affixed with light-emitting diodes (LEDs) monitored by in-room motion-tracking cameras. If respiration
Volume 84 Number 1 2012 influenced the motion of the target, the Synchrony Motion Tracking system (MTS) was used. Women underwent mid-thorax to mid-thigh noncontrasted contiguous axial CT high-resolution imaging (voxel 0.98 0.98 1.0 mm, voltage 120 kVp, 450 mAs) in the head-first supine position. On the same day, women had axial 18F-deoxyglucose positron emission tomography (18F-FDG PET) images acquired in the head-first supine position (voxel 4.0 4.0 4.0 mm) following institutional protocol (8, 9). CT and 18F-FDG PET images were then imported, digitally overlaid, and coregistered for inverse radiation treatment planning on the MultiPlan 3.5.2 Treatment Planning System (Accuray).
Target and normal tissue definition The contoured SBRT clinical target volume (CTV) consisted of identified cancer target(s), highlighted by CT and 18F-FDG PET and agreed upon by both a treating radiation oncologist and gynecologic oncologist (9). A 3.0-mm margin encompassing the entire CTV was typically applied to each planning tumor volume (PTV) (3). Nearby normal tissue structures such as the small bowel, rectum, bladder, liver, kidneys, lungs, bilateral proximal femurs, vagina, and sacral nerve roots were contoured by the radiation oncologist, medical physicist, or certified medical dosimetrist. For this retrospective study, a bone marrow volume was added back to the original SBRT treatment images. A bone marrow volume was estimated by contouring the internal profile of vertebral bodies (exclusive of intervertebral discs) between the T12 and S3 vertebrae (10-12). For the S1 to S3 sacral vertebrae, the internal contour of the vertebral body spanning each medial edge of the sacroiliac joint was included. The defined volume of bone marrow receiving a dose of 10 Gy or greater (V10) was calculated; the V10 was not a dose-limiting parameter at the time of SBRT treatment.
Baseline and follow-up studies Complete blood counts were obtained pre-therapy and posttherapy by peripheral percutaneous phlebotomy. Automated white blood cells (109/L), red blood cellsl (1012/L), and platelet (109/L) counts, and hemoglobin (mg/dL) and hematocrit (%) levels were abstracted from the medical records. Study endpoints were acute treatment-related bone marrow toxicities occurring within 30 days from the start of stereotactic body radiosurgery. Complications were graded on a scale of 0 (none) to 5 (fatal), according to Common Toxicity Criteria for Adverse Events (version 4.0). Major toxicities were grade 3, 4, or 5. Our clinical program did not require or specify a date for post-therapy complete blood counts to be done. It was recommended in the 30-day post-therapy period that complete blood counts be drawn immediately if a woman reported fever, bleeding, or severe fatigue interfering with activities of daily living. Women were scheduled for post-therapy clinical follow-up with the treating radiation oncologist 30 days from the start of stereotactic body radiosurgery. Subsequent follow-up evaluations were done at 3-month intervals with either the radiation oncologist or gynecologic oncologist. Complete blood counts and diagnostic imaging of the chest, abdomen, and pelvis were done in 3-month intervals at the discretion of the treating physicians.
Hematological toxicity after robotic radiosurgery
e37
Treatment planning and delivery A radiation prescription dose of 3 consecutive daily fractions of 800 cGy per fraction totaling 2400 cGy (prescribed to the 70% isodose line) was used. This dose was decided upon arbitrarily based on previous clinical experience (3, 4). Inverse radiation treatment planning was conducted on the MultiPlan 3.5.2 Treatment Planning System (Accuray). The radiation emitted from a 6-MV SBRT linear accelerator was collimated either by using one of 12 fixed tungsten circular collimators (5-60 mm) or by using a tungsten-copper alloy segmented hexagon IRIS collimator (13). During treatment, implanted gold seed fiducials were tracked either using 3 degrees of freedom (ie, x, y, z) or using all 6 degrees of freedom (ie, x, y, z, pitch, roll, and yaw). Cross-plane radiographic imaging, as part of the Cyberknife’s target localizing subsystem (TLS) was used for target fiducial localization or bony anatomy. Real-time images created by the TLS were automatically registered for comparison to planning-CT digitally reconstructed radiographs (DRRs). If automatic registration indicated fiducials (or bony landmarks) were not aligned within predefined tolerances in any one of the applicable 6 degrees of freedom (x, y, z, pitch, roll, or yaw), treatment automatically paused. Either therapistguided automated couch or manual patient repositioning occurred until tolerances were achieved. For targets moving with respiration, the Synchrony system was used. A correlation model built by the Synchrony system guided synchronous robotic arm motion to follow any respiratory-induced motion of the target during the delivery of any and all treatment beams.
Statistical analysis Descriptive and graphical statistics were computed using SPSS 18.0 software (SPSS Inc, Chicago, IL). Dose-volume histograms for bone marrow were generated using dosimetry calculations. Pearson correlation coefficients were determined for the bone marrow V10 and the computed difference between pre-therapy and post-therapy hematological parameters. (a Z 0.05). Paired t tests and Wilcoxon signed-rank tests were performed on measured hematological parameters (a Z 0.05). For context, comparisons of hematological context were made to cohorts of women treated with extended-field irradiation whose associated hematological grade 3, 4, 5 toxicity rate was 15% (14) and with involved-field radiochemotherapy whose associated hematological grade 3 or 4 rate was 62% (15).
Results Stereotactic body radiosurgery dose distribution for bone marrow SBRT involved individual radiation beams converging on single or multiple closely associated clinical abdominopelvic cancer targets, including 27 pre-sacral or common iliac pelvic lymph node targets, 22 paraortic lymph node targets, and 12 peri-caval hepatic targets (Fig. 1). A complex example case is depicted in Figure 2. The median SBRT dose was 2400 cGy (range, 800-2400 cGy) most often in 3 (range, 1-3) consecutive, daily 800-cGy fractions. Among the 61 women, 2 patients received a single
e38
Kunos et al.
International Journal of Radiation Oncology Biology Physics
Fig. 2. (A) The 102 treatment beams by the Cyberknife radiosurgery accelerator for treatment during the targeting of right-sided paraaortic lymph nodes overlying the right psoas major and minor muscles. (B-D) Axial, coronal, and sagittal projections of radiosurgical treatment. The 2400-cGy prescription isodose is highlighted in red, and the 1000 cGy isodose is outlined in yellow. (E) Dose-volume histograms for the paraortic lymph node target (red), bone marrow (yellow), small bowel (pink), and liver (orange) are plotted. High target coverage was achieved while minimizing bone marrow V10 (in this most extreme example).
800-cGy ablative radiation dose to a single lumbar vertebral body metastasis. The median prescription isodose line contour was 70% (25%-75% quartile: 70%-75%). Beam collimation was fixed for cancer targets in 33 women (range, 10-50 mm) or was automated (IRIS; range, 7.5-60 mm) in 28 remaining women. In all, 50 (82%) of 61 women had gold seed fiducials used to track cancer targets. Of the 61 women, 15 (25%) had Synchrony tracking used for targets moving with respiratory motion. An 18 F-FDG PET was fused with CT images to guide radiation therapy planning in 54 (89%) of 61 women. The median abdominopelvic cancer planning target volume irradiated was 71 cm3 (25%-75% quartile, 29-158 cm3). Median clinical follow-up was 8 months (25%-75% quartile, 4-15 months) for the 61 women under study. No surviving woman has been lost to followup, but 24 (39%) women have died. Example axial, coronal, and sagittal 1000 cGy radiation dose distributions for SBRT are depicted in Fig. 2B-D. Dose-volume histograms for cancer targets and bone marrow are illustrated in Fig. 2E. This representative case provides an extreme example of a large SBRT cancer target (337 cm3), whose bone marrow V10 was 48% and whose white blood cell (2%), red blood cell (þ2%), and platelet (7%) counts were not substantially lowered by treatment 30 days post-therapy relative to pre-therapy counts. Overall, the median bone marrow V10 was 2% (25%-75% quartile, 0%-8%) of the lumbosacral bone marrow volume among all
61 women. For the 22 women with paraortic or para-vertebral cancer targets, the median bone marrow V10 was 6% (25%75% quartile, 3%-13%). Table 2 lists pre- and post-therapy hematological parameters. Pre-therapy complete blood counts were done a median 28 days (25%-75% quartile, 14-38 days) before the first radiation dose, most often before SBRT simulation and gold seed fiducial placement. Post-therapy complete blood counts were done in all patients and occurred at a median of 32 days (25%-75% quartile, 28-45 days) after SBRT. Bone marrow V10 (Fig. 3) did not correlate with pre-therapyepost-therapy differences in white blood cells (coefficient Z 0.03; PZ.825), red blood cells (coefficient Z 0.10; PZ.447), hemoglobin (coefficient Z 0.13; PZ.328), or platelet counts (coefficient Z 0.03; PZ.825). Abdominopelvic SBRT cancer target volume did not correlate with pre-therapyepost-therapy difference in white blood cells (coefficient Z 0.14; PZ.254), red blood cells (coefficient Z 0.15; PZ.240), hemoglobin (coefficient Z 0.11; PZ.407), or platelet counts (coefficient Z 0.04; PZ.774). For the 22 women with paraortic or para-vertebral targets, no significant correlation was found between white cell, red cell, hemoglobin, or platelet count difference and bone marrow V10 (PZ.738; PZ.238; PZ.251; PZ.443, respectively) or abdominopelvic SBRT cancer target volume (PZ.055 [coefficient Z 0.423]; PZ.942; PZ.937; PZ.965, respectively).
Volume 84 Number 1 2012 Table 2
Hematological toxicity after robotic radiosurgery
e39
Hematological parameters of stereotactic body radiosurgery study patients Stereotactic body radiosurgery parameters
Delivered stereotactic radiation dose (cGy): mean (standard deviation) Total contoured axial bone marrow volume (cm3): mean (standard deviation) Marrow volume in daily 800 cGy isodose line (cm3): mean (standard deviation) V10 stereotactic radiation dose (%): mean (standard deviation) V20 stereotactic radiation dose (%): mean (standard deviation) Hematological parameters 9
White blood cell count (10 /L): median (25%-75th quartile) Red blood cell count (1012/L): median (25%-75th quartile) Hemoglobin (mg/dL): median (25%-75% quartile) Hematocrit (%): median (25%-75% quartile) Platelet count (109/L): median (25%-75% quartile)
Clinical toxicity after stereotactic body radiosurgery Self-reported fatigue within the first 10 days post-therapy was grade 2 in 15 women (25%) and grade 3 in 2 women (3%) among the 61 women treated. Fatigue resolved spontaneously in all 17 women by 30 days post-therapy. Neutropenias or febrile neutropenias were not observed post-therapy. Three women (5%) had a grade 1 decline in serum hemoglobin level to less than 10.0 g/dL, and 3 women (5%) had thrombocytopenia (1 patient, grade 1; 1 patient, grade 2; and 1 patient, grade 3). There have been no greater than 90-day post-therapy hematological toxicities documented.
Discussion This retrospective study showed that robotic SBRT did not lower hematological measures of bone marrow toxicity in women with metastatic abdominopelvic gynecologic cancer. Bone marrow hematological toxicity, which encompasses both post-therapy patient-reported fatigue and objective post-therapy 1-month complete blood counts, was selected as the primary retrospective end-point for this study. This was done on the basis of the hypothesis that robotic SBRT targeting abdominopelvic sites of cancer may exert symptomatic collateral ablative cytotoxic effects in bone marrow cells, lowering white cell counts acutely and red cell counts long-term, especially in the context of preSBRT multi-agent chemotherapy directed at metastatic disease. For this reason, lifestyle-altering fatigue or precipitous drop in blood component counts were regarded as clinically meaningful in assessing outcome after robotic SBRT targeting abdominopelvic cancer sites of disease. Stereotactic radiosurgery has been evaluated in several studies of metastatic gynecologic malignancies and as demonstrated activity at various doses and schedules (3-5). In our study, 2400 cGy divided into 3 consecutive 800-cGy daily fractions was given to take radiobiological advantage of relative radiosensitivity of gynecologic malignancies (1, 2). Given the sub-millimeter accuracy and precision of robotic SBRT (7), collateral ablative doses delivered to bone marrow were low overall (2%) and only slightly higher (6%) when SBRT targets were abutting vertebrae of the thoracolumbar spine or sacrum. For perspective, extended-field
Pre-therapy 5.5 3.9 11.9 35.9 228
(4.2-7.1) (3.6-4.2) (10.7-12.7) (32.6-38.8) (192-293)
2325 cGy 112 cm3 7.3 cm3 7% 4% Post-therapy 5.3 3.9 12.1 35.4 217
(4.3-7.4) (3.6-4.3) (10.7-12.7) (32.9-38.0) (166-286)
(330 cGy) (25 cm3) (4.1 cm3) (11%) (7%) P .465 .360 .344 .905 .305
radiation portals render a grade 3, 4, or 5 hematological toxicity rate of 15% (14). In the instance of abdominopelvic involved-field radiochemotherapy, radiation portals may be associated with a grade 3 or 4 acute hematological toxicity rate of 62% (15). Although, on occasion, ablative doses resulted in substantial bone marrow volumes receiving 1000 cGy or more (as shown in the example in Figure 2), substantial hematological toxicity was not apparent. In a prior study, SBRT-targeting of paraortic nodal sites of disease relapse achieved low bone marrow volume dose (5). Our use of SBRT in 57 women who had undergone multi-agent, multi-cycle chemotherapy was tolerable, as evidenced by manageable fatigue and minimal global alteration in complete blood cell counts. In other studies in which radiation and chemotherapy were co-administered (14, 15), adverse drops in hematologic counts occured in 15%-62% of women, sometimes compromising and stopping desired radiochemotherapy treatment. The fact that the majority of women (93%) in our study had chemotherapy antecedent to SBRT and then tolerated SBRT suggests that SBRT plus novel chemotherapy could be investigated in a phase 1 clinical trial setting. SBRT targeting cancer sites in close proximity to bones harboring marrow led to appreciable fatigue, as reported by one quarter of our patients. We were not able to detect differences in bone marrow constituents having short 6-8 hour (16) peripheral blood turnover (eg, white cells) nor having long 40-60 day (17) peripheral blood lifespans (eg, red cells). It is tempting to speculate that radiation-induced damage of nuclear and mitochondrial DNA in normal and cancer cells activates nucleoside supply de novo (18, 19) and salvage (20) dNTP supply pathways, and that a decrease in salvage reserve in the plasma impacts terminally differentiated muscle cells that depend on nucleoside salvage to replicate their mitochondrial DNA, thus creating symptoms of fatigue. Limitations of this retrospective study include the following: small sample sizes for each individual gynecologic malignancy type, which limits our assessment of clinical benefit and progression-free survival; potential investigator bias in retrospective assignment of toxicity criteria; inconsistencies in the timed collection of complete blood counts, which may underestimate hematological toxicity assessment; and limited overall follow-up. A low observed variation mean V10 for bone marrow in this study, perhaps resulting from few women being treated and from the submillimeter precision of robotic SBRT, raises a statistical hazard for missed association in treatment-related hematological adverse
e40
Kunos et al.
International Journal of Radiation Oncology Biology Physics
Fig. 3. Bone marrow volume receiving 1000 cGy radiation (V10) and differences in pre-therapy to post-therapy white blood cells (A), red blood cells (B), hemoglobin (C), and platelets (D). events. A prospective robotic SBRT trial for women with metastatic gynecologic malignancies with a quite similar cohort makeup has finished accrual within our group in an effort to report safety, tolerability, and outcomes measures.
2.
3.
Conclusion 4.
In conclusion, this retrospective study lends support to the concept that SBRT targeting abdominopelvic sites of gynecologic cancer does not generate substantial hematologic toxicity on its own, or when sequenced after chemotherapy. On the basis of these retrospective results, a phase I clinical trial of SBRT and chemotherapy co-administered in a similar population of women with metastatic gynecologic cancer is currently being considered.
5.
6. 7.
References 1. Kunos C, Zhang Y, Brindle J. Stereotactic body radiosurgery in gynecologic carcinomas. In: Ayhan A, Reed N, Gultekin M, et al.,
8.
editors. Textbook of Gynaecological Oncology. Ankara, Turkey: Gu¨nes Publishing; 2011. p. 835-841. Mayr N, Huang Z, Sohn J, et al. Emerging application of stereotactic body radiation therapy for gynecologic malignancies. Expert Rev Anticancer Ther 2011;11:1071-1077. Kunos C, Von Gruenigen V, Waggoner S, et al. Cyberknife radiosurgery for squamous cell carcinoma of the vulva after prior pelvic radiation therapy. Technol Cancer Res Treat 2008;7:375-380. Kunos C, Chen W, DeBernardo R, et al. Stereotactic body radiosurgery for pelvic relapse of gynecologic malignancies. Technol Cancer Res Treat 2009;8:393-400. Choi C, Cho C, Yoo S, et al. Image-guided stereotactic body radiation therapy in patients with isolated para-aortic lymph node metastases from uterine cervical and corpus cancer. Int J Radiat Oncol Biol Phys 2009;74:147-153. Kilby W, Dooley J, Kuduvalli G, et al. The Cyberknife robotic radiosurgery system in 2010. Technol Cancer Res Treat 2010;9:431-438. Wilcox E, Daskalov G. Evaluation of GAFCHROMIC EBT film for Cyberknife dosimetry. Med Phys 2007;34:1967-1974. Kunos C, Radivoyevitch T, Abdul-Karim F, et al. 18F-fluoro-2-deoxyd-glucose positron emission tomography standard uptake value as an indicator of cervical cancer chemoradiation therapeutic response. Int J Gyn Oncol 2011;21:1117-1123.
Volume 84 Number 1 2012 9. Kunos CA, Debernardo R, Fabien J, Dobbins DC, Zhang Y, Brindle J, Faulhaber PF. 18FDG-PET/CT definition of clinical target volume for robotic stereotactic body radiosurgery treatment of metastatic gynecologic malignancies. J Nucl Med Rad Ther 2011. http://dx.doi.org/10. 4172/2155-9619.S4-001. 10. Ellis R. The distribution of active bone marrow in the adult. Phys Med Biol 1961;5:255-258. 11. Mell L, Tiryaki H, Ahn K-H, et al. Dosimetric comparison of bone marrow-sparring intensity-modulated radiotherapy versus conventional techniques for treatment of cervical cancer. Int J Radiat Oncol Biol Phys 2008;71:1504-1510. 12. Roeske J, Lujan A, Reba R, et al. Incorporation of SPECT bone marrow imaging into intensity modulated whole-pelvic radiation therapy treatment planning for gynecologic malignancies. Radiother Oncol 2005;77:11-17. 13. Echner G, Kilby W, Lee M, et al. The design, physical properties and clinical utility of an iris collimator for robotic radiosurgery. Phys Med Biol 2009;54:5359-5380. 14. Varia MA, Bundy BN, Deppe G, et al. Cervical carcinoma metastatic to para-aortic nodes: Extended field radiation therapy with concomitant
Hematological toxicity after robotic radiosurgery
15.
16. 17.
18.
19.
20.
e41
5-fluorouracil and cisplatin chemotherapy: A Gynecologic Oncology Group study. Int J Radiat Oncol Biol Phys 1998;42:1015-1023. Lee YS, Kim YS, Kim JH, et al. Feasibility and outcome of concurrent chemoradiotherapy for recurrent cervical carcinoma after initial surgery. Tumori 2010;96:553-559. Summers C, Rankin S, Condliffe A, et al. Neutrophil kinetics in health and disease. Trends Immunol 2010;31:318-324. Kruse A, Uehlinger D, Gotch F, et al. Red blood cell lifespan, erythropoiesis and hemoglobin control. Contrib Nephrol 2008;161: 247-254. Kunos C, Chiu S, Pink J, et al. Modulating radiation resistance by inhibiting ribonucleotide reductase in cancers with virally or mutationally silenced p53 protein. Radiation Res 2009;172: 666-676. Kunos C, Radivoyevitch T, Pink J, et al. Ribonucleotide reductase inhibition enhances chemoradiosensitivity of human cervical cancers. Radiation Res 2010;174:574-581. Kunos C, Ferris G, Pyatka N, et al. Deoxynucleoside salvage facilitates DNA repair during ribonucleotide reductase blockade in human cervical cancers. Radiat Res 2011;176:425-433.
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation: Gynecologic Cancer
Hematological Toxicity After Robotic Stereotactic Body Radiosurgery for Treatment of Metastatic Gynecologic Malignancies Charles A. Kunos, MD, PhD,* Robert Debernardo, MD,y Tomas Radivoyevitch, PhD,z Jeffrey Fabien, MS,* Donald C. Dobbins, RT, CMD,* Yuxia Zhang, MS,* and James Brindle, PhD* Departments of *Radiation Oncology, yObstetrics and Gynecology, and zEpidemiology and Biostatistics, University Hospitals Case Medical Center and Case Western Reserve University School of Medicine, Cleveland, Ohio Received Nov 16, 2011, and in revised form Feb 9, 2012. Accepted for publication Feb 10, 2012
Summary Abdominopelvic relapse of gynecologic cancers after radiation and chemotherapy are difficult to manage due to an adverse hazard of bone marrow toxicity. A total of 61 women with stage IV gynecologic malignancies were found, at retrospective review, to have undergone abdominopelvic stereotactic body radiosurgery safely and without adverse bone marrow toxicity. Findings support the novel integration of stereotactic body radiosurgery into treatment strategies for relapsed abdominopelvic gynecologic cancers.
Purpose: To evaluate hematological toxicity after robotic stereotactic body radiosurgery (SBRT) for treatment of women with metastatic abdominopelvic gynecologic malignancies. Methods and Materials: A total of 61 women with stage IV gynecologic malignancies treated with abdominopelvic SBRT were analyzed after ablative radiation (2400 cGy/3 divided consecutive daily doses) delivered by a robotic-armed Cyberknife SBRT system. Abdominopelvic bone marrow was identified using computed tomography-guided contouring. Fatigue and hematologic toxicities were graded by retrospective assignment of common toxicity criteria for adverse events (version 4.0). Bone marrow volume receiving 1000 cGy (V10) was tested for association with post-therapy (median 32 days [25%-75% quartile, 28-45 days]) white- or red-cell counts, hemoglobin levels, and platelet counts as marrow toxicity surrogates. Results: In all, 61 women undergoing abdominopelvic SBRT had a median bone marrow V10 of 2% (25%-75% quartile: 0%-8%). Fifty-seven (93%) of 61 women had received at least 1 preSBRT marrow-taxing chemotherapy regimen for metastatic disease. Bone marrow V10 did not associate with hematological adverse events. In all, 15 grade 2 (25%) and 2 grade 3 (3%) fatigue symptoms were self-reported among the 61 women within the first 10 days post-therapy, with fatigue resolved spontaneously in all 17 women by 30 days post-therapy. Neutropenia was not observed. Three (5%) women had a grade 1 drop in hemoglobin level to <10.0 g/dL. Single grade 1, 2, and 3 thrombocytopenias were documented in 3 women. Conclusions: Abdominopelvic SBRT provided ablative radiation dose to cancer targets without increased bone marrow toxicity. Abdominopelvic SBRT for metastatic gynecologic malignancies warrants further study. Ó 2012 Elsevier Inc.
Reprint requests to: Charles A. Kunos, MD, PhD, Department of Radiation Oncology, University Hospitals of Cleveland, 11100 Euclid Ave,
Int J Radiation Oncol Biol Phys, Vol. 84, No. 1, pp. e35ee41, 2012 0360-3016/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.ijrobp.2012.02.027
LTR 6068, Cleveland, Ohio 44106 USA. Tel: þ216-844-3103; Fax: þ216844-2005; E-mail: [email protected] Conflict of interest: none.
e36
International Journal of Radiation Oncology Biology Physics
Kunos et al.
Introduction
Table 1 Baseline characteristics of stereotactic body radiosurgery study patients
Robotic stereotactic body radiosurgery (SBRT) distinguishes itself from other radiation delivery systems in its steep radiation dose drop-off, allowing narrow “pencil beam” treatment fields to deliver ablative radiation treatment condensed into a few radiation treatment sessions (1-5). One robotic SBRT system, the Cyberknife system (Accuray [Sunnyvale, CA]), uses an industrial frameless robotic arm mounted by a radiation linear accelerator to administer precise ablative radiation dose while simultaneously tracking and following cancer-target motion in real time (6). Sizeable reductions in radiation field margins can be realized by tracking the movement of the target, in the end giving a reduced planning target volume irradiated with submillimeter accuracy (7). However, ablative SBRT introduces the possibility of greater risk of normal tissue injury manifested in organs such as radiosensitive bone marrow. The ability of SBRT to lower bone marrow radiation dose when targeting abdominopelvic sites of disease has not been fully investigated. Quantifying the impact of SBRT upon bone marrow is needed when targeting nearby anticancer targets, especially in the era of widespread clinical use of and novel testing of anticancer chemotherapeutics that may have bone marrow-taxing effects. The purpose of this study was to characterize, for the first time, the abilities of SBRT to lessen radiation-related hematological toxicities when treating women with metastatic abdominopelvic gynecologic malignancies.
Methods and Materials Patients Inclusion criteria for study were female sex, age 18 years or more, and a diagnosis of metastatic gynecological cancer occurring in an abdominopelvic site. A total of 61 women met these criteria during a retrospective study selection among 69 women treated between 2007 and 2011 within the University Hospitals of Cleveland and Case Western Reserve University (Cleveland, OH) stereotactic body radiosurgery program (Fig. 1). Patient demographics, gynecologic cancer type and number of sites to be treated by SBRT, and administered prior therapies are listed in Table 1. All patients provided written informed consent before SBRT. Institutional review board approval was granted for this study.
Simulation One week before SBRT simulation, women had either direct visual or computed tomography (CT)-guided implantation of at
Fig. 1.
Selection of study patients.
Characteristic Female sex: n (%) 61 (100) Age: years Median 64 Range 27-87 Race/ethnicity: n (%)* White 53 (87) Black or African American 5 (8) Hispanic 3 (5) GOG performance status: no (%)y 0 44 (72) 1 10 (16) 2 7 (12) No. of metastases to be treated by radiosurgery: n (%) 1 25 (41) 2 16 (26) 3 15 (25) 4 5 (8) 4 0 (0) Histopathology: n (%) Cervical/vaginal squamous cell carcinoma 10 (16) Endometrial adenocarcinoma 16 (26) Ovarian adenocarcinoma 34 (56) Vulvar squamous cell carcinoma 1 (2) Prior radiation: n (%) 28 (46) Inclusive of radiosurgery site 23 (38) >5000 cGy delivered pre-therapy 15 (25) to radiosurgery site Prior chemotherapy: n (%) 57 (93) No. of prior courses of chemotherapy for metastases: n (%) 0 5 (8) 1 38 (62) 2 15 (25) 3 3 (5) Platinum-containing regimen: no. (%) 57 (93) Taxane-containing regimen: n (%) 44 (72) Anthracycline-containing regimen: no. (%) 13 (21) Topotecan-containing regimen: n (%) 8 (13) Gemcitabine-containing regimen: n (%) 9 (15) * Race/ethnicity was self-reported. y Gynecologic Oncology Group (GOG) performance status reflects individual daily living activities on a scale of 0 (fully active with symptoms) to 5 (dead).
least 3 single 1.6 3 mm soft tissue gold seed fiducials within 4-6 cm of the SBRT target(s) under anesthesia (1). For immobile paraortic nodal metastases, rigid bony landmarks of the spine may have been used for tracking purposes. Women were treated in a head-first supine treatment position with arms at their sides or on their chest while lying on a frameless flat tabletop. Two-pin knee sponge tabletop registry was used for reproducible orthogonal body alignment. Evacuated vacuum-bag immobilization was used when it was anticipated that radiosurgical targets would have large intrafraction excursion during radiation dose delivery. To account for target respiratory motion, women wore a firmly conforming vest around the thorax affixed with light-emitting diodes (LEDs) monitored by in-room motion-tracking cameras. If respiration
Volume 84 Number 1 2012 influenced the motion of the target, the Synchrony Motion Tracking system (MTS) was used. Women underwent mid-thorax to mid-thigh noncontrasted contiguous axial CT high-resolution imaging (voxel 0.98 0.98 1.0 mm, voltage 120 kVp, 450 mAs) in the head-first supine position. On the same day, women had axial 18F-deoxyglucose positron emission tomography (18F-FDG PET) images acquired in the head-first supine position (voxel 4.0 4.0 4.0 mm) following institutional protocol (8, 9). CT and 18F-FDG PET images were then imported, digitally overlaid, and coregistered for inverse radiation treatment planning on the MultiPlan 3.5.2 Treatment Planning System (Accuray).
Target and normal tissue definition The contoured SBRT clinical target volume (CTV) consisted of identified cancer target(s), highlighted by CT and 18F-FDG PET and agreed upon by both a treating radiation oncologist and gynecologic oncologist (9). A 3.0-mm margin encompassing the entire CTV was typically applied to each planning tumor volume (PTV) (3). Nearby normal tissue structures such as the small bowel, rectum, bladder, liver, kidneys, lungs, bilateral proximal femurs, vagina, and sacral nerve roots were contoured by the radiation oncologist, medical physicist, or certified medical dosimetrist. For this retrospective study, a bone marrow volume was added back to the original SBRT treatment images. A bone marrow volume was estimated by contouring the internal profile of vertebral bodies (exclusive of intervertebral discs) between the T12 and S3 vertebrae (10-12). For the S1 to S3 sacral vertebrae, the internal contour of the vertebral body spanning each medial edge of the sacroiliac joint was included. The defined volume of bone marrow receiving a dose of 10 Gy or greater (V10) was calculated; the V10 was not a dose-limiting parameter at the time of SBRT treatment.
Baseline and follow-up studies Complete blood counts were obtained pre-therapy and posttherapy by peripheral percutaneous phlebotomy. Automated white blood cells (109/L), red blood cellsl (1012/L), and platelet (109/L) counts, and hemoglobin (mg/dL) and hematocrit (%) levels were abstracted from the medical records. Study endpoints were acute treatment-related bone marrow toxicities occurring within 30 days from the start of stereotactic body radiosurgery. Complications were graded on a scale of 0 (none) to 5 (fatal), according to Common Toxicity Criteria for Adverse Events (version 4.0). Major toxicities were grade 3, 4, or 5. Our clinical program did not require or specify a date for post-therapy complete blood counts to be done. It was recommended in the 30-day post-therapy period that complete blood counts be drawn immediately if a woman reported fever, bleeding, or severe fatigue interfering with activities of daily living. Women were scheduled for post-therapy clinical follow-up with the treating radiation oncologist 30 days from the start of stereotactic body radiosurgery. Subsequent follow-up evaluations were done at 3-month intervals with either the radiation oncologist or gynecologic oncologist. Complete blood counts and diagnostic imaging of the chest, abdomen, and pelvis were done in 3-month intervals at the discretion of the treating physicians.
Hematological toxicity after robotic radiosurgery
e37
Treatment planning and delivery A radiation prescription dose of 3 consecutive daily fractions of 800 cGy per fraction totaling 2400 cGy (prescribed to the 70% isodose line) was used. This dose was decided upon arbitrarily based on previous clinical experience (3, 4). Inverse radiation treatment planning was conducted on the MultiPlan 3.5.2 Treatment Planning System (Accuray). The radiation emitted from a 6-MV SBRT linear accelerator was collimated either by using one of 12 fixed tungsten circular collimators (5-60 mm) or by using a tungsten-copper alloy segmented hexagon IRIS collimator (13). During treatment, implanted gold seed fiducials were tracked either using 3 degrees of freedom (ie, x, y, z) or using all 6 degrees of freedom (ie, x, y, z, pitch, roll, and yaw). Cross-plane radiographic imaging, as part of the Cyberknife’s target localizing subsystem (TLS) was used for target fiducial localization or bony anatomy. Real-time images created by the TLS were automatically registered for comparison to planning-CT digitally reconstructed radiographs (DRRs). If automatic registration indicated fiducials (or bony landmarks) were not aligned within predefined tolerances in any one of the applicable 6 degrees of freedom (x, y, z, pitch, roll, or yaw), treatment automatically paused. Either therapistguided automated couch or manual patient repositioning occurred until tolerances were achieved. For targets moving with respiration, the Synchrony system was used. A correlation model built by the Synchrony system guided synchronous robotic arm motion to follow any respiratory-induced motion of the target during the delivery of any and all treatment beams.
Statistical analysis Descriptive and graphical statistics were computed using SPSS 18.0 software (SPSS Inc, Chicago, IL). Dose-volume histograms for bone marrow were generated using dosimetry calculations. Pearson correlation coefficients were determined for the bone marrow V10 and the computed difference between pre-therapy and post-therapy hematological parameters. (a Z 0.05). Paired t tests and Wilcoxon signed-rank tests were performed on measured hematological parameters (a Z 0.05). For context, comparisons of hematological context were made to cohorts of women treated with extended-field irradiation whose associated hematological grade 3, 4, 5 toxicity rate was 15% (14) and with involved-field radiochemotherapy whose associated hematological grade 3 or 4 rate was 62% (15).
Results Stereotactic body radiosurgery dose distribution for bone marrow SBRT involved individual radiation beams converging on single or multiple closely associated clinical abdominopelvic cancer targets, including 27 pre-sacral or common iliac pelvic lymph node targets, 22 paraortic lymph node targets, and 12 peri-caval hepatic targets (Fig. 1). A complex example case is depicted in Figure 2. The median SBRT dose was 2400 cGy (range, 800-2400 cGy) most often in 3 (range, 1-3) consecutive, daily 800-cGy fractions. Among the 61 women, 2 patients received a single
e38
Kunos et al.
International Journal of Radiation Oncology Biology Physics
Fig. 2. (A) The 102 treatment beams by the Cyberknife radiosurgery accelerator for treatment during the targeting of right-sided paraaortic lymph nodes overlying the right psoas major and minor muscles. (B-D) Axial, coronal, and sagittal projections of radiosurgical treatment. The 2400-cGy prescription isodose is highlighted in red, and the 1000 cGy isodose is outlined in yellow. (E) Dose-volume histograms for the paraortic lymph node target (red), bone marrow (yellow), small bowel (pink), and liver (orange) are plotted. High target coverage was achieved while minimizing bone marrow V10 (in this most extreme example).
800-cGy ablative radiation dose to a single lumbar vertebral body metastasis. The median prescription isodose line contour was 70% (25%-75% quartile: 70%-75%). Beam collimation was fixed for cancer targets in 33 women (range, 10-50 mm) or was automated (IRIS; range, 7.5-60 mm) in 28 remaining women. In all, 50 (82%) of 61 women had gold seed fiducials used to track cancer targets. Of the 61 women, 15 (25%) had Synchrony tracking used for targets moving with respiratory motion. An 18 F-FDG PET was fused with CT images to guide radiation therapy planning in 54 (89%) of 61 women. The median abdominopelvic cancer planning target volume irradiated was 71 cm3 (25%-75% quartile, 29-158 cm3). Median clinical follow-up was 8 months (25%-75% quartile, 4-15 months) for the 61 women under study. No surviving woman has been lost to followup, but 24 (39%) women have died. Example axial, coronal, and sagittal 1000 cGy radiation dose distributions for SBRT are depicted in Fig. 2B-D. Dose-volume histograms for cancer targets and bone marrow are illustrated in Fig. 2E. This representative case provides an extreme example of a large SBRT cancer target (337 cm3), whose bone marrow V10 was 48% and whose white blood cell (2%), red blood cell (þ2%), and platelet (7%) counts were not substantially lowered by treatment 30 days post-therapy relative to pre-therapy counts. Overall, the median bone marrow V10 was 2% (25%-75% quartile, 0%-8%) of the lumbosacral bone marrow volume among all
61 women. For the 22 women with paraortic or para-vertebral cancer targets, the median bone marrow V10 was 6% (25%75% quartile, 3%-13%). Table 2 lists pre- and post-therapy hematological parameters. Pre-therapy complete blood counts were done a median 28 days (25%-75% quartile, 14-38 days) before the first radiation dose, most often before SBRT simulation and gold seed fiducial placement. Post-therapy complete blood counts were done in all patients and occurred at a median of 32 days (25%-75% quartile, 28-45 days) after SBRT. Bone marrow V10 (Fig. 3) did not correlate with pre-therapyepost-therapy differences in white blood cells (coefficient Z 0.03; PZ.825), red blood cells (coefficient Z 0.10; PZ.447), hemoglobin (coefficient Z 0.13; PZ.328), or platelet counts (coefficient Z 0.03; PZ.825). Abdominopelvic SBRT cancer target volume did not correlate with pre-therapyepost-therapy difference in white blood cells (coefficient Z 0.14; PZ.254), red blood cells (coefficient Z 0.15; PZ.240), hemoglobin (coefficient Z 0.11; PZ.407), or platelet counts (coefficient Z 0.04; PZ.774). For the 22 women with paraortic or para-vertebral targets, no significant correlation was found between white cell, red cell, hemoglobin, or platelet count difference and bone marrow V10 (PZ.738; PZ.238; PZ.251; PZ.443, respectively) or abdominopelvic SBRT cancer target volume (PZ.055 [coefficient Z 0.423]; PZ.942; PZ.937; PZ.965, respectively).
Volume 84 Number 1 2012 Table 2
Hematological toxicity after robotic radiosurgery
e39
Hematological parameters of stereotactic body radiosurgery study patients Stereotactic body radiosurgery parameters
Delivered stereotactic radiation dose (cGy): mean (standard deviation) Total contoured axial bone marrow volume (cm3): mean (standard deviation) Marrow volume in daily 800 cGy isodose line (cm3): mean (standard deviation) V10 stereotactic radiation dose (%): mean (standard deviation) V20 stereotactic radiation dose (%): mean (standard deviation) Hematological parameters 9
White blood cell count (10 /L): median (25%-75th quartile) Red blood cell count (1012/L): median (25%-75th quartile) Hemoglobin (mg/dL): median (25%-75% quartile) Hematocrit (%): median (25%-75% quartile) Platelet count (109/L): median (25%-75% quartile)
Clinical toxicity after stereotactic body radiosurgery Self-reported fatigue within the first 10 days post-therapy was grade 2 in 15 women (25%) and grade 3 in 2 women (3%) among the 61 women treated. Fatigue resolved spontaneously in all 17 women by 30 days post-therapy. Neutropenias or febrile neutropenias were not observed post-therapy. Three women (5%) had a grade 1 decline in serum hemoglobin level to less than 10.0 g/dL, and 3 women (5%) had thrombocytopenia (1 patient, grade 1; 1 patient, grade 2; and 1 patient, grade 3). There have been no greater than 90-day post-therapy hematological toxicities documented.
Discussion This retrospective study showed that robotic SBRT did not lower hematological measures of bone marrow toxicity in women with metastatic abdominopelvic gynecologic cancer. Bone marrow hematological toxicity, which encompasses both post-therapy patient-reported fatigue and objective post-therapy 1-month complete blood counts, was selected as the primary retrospective end-point for this study. This was done on the basis of the hypothesis that robotic SBRT targeting abdominopelvic sites of cancer may exert symptomatic collateral ablative cytotoxic effects in bone marrow cells, lowering white cell counts acutely and red cell counts long-term, especially in the context of preSBRT multi-agent chemotherapy directed at metastatic disease. For this reason, lifestyle-altering fatigue or precipitous drop in blood component counts were regarded as clinically meaningful in assessing outcome after robotic SBRT targeting abdominopelvic cancer sites of disease. Stereotactic radiosurgery has been evaluated in several studies of metastatic gynecologic malignancies and as demonstrated activity at various doses and schedules (3-5). In our study, 2400 cGy divided into 3 consecutive 800-cGy daily fractions was given to take radiobiological advantage of relative radiosensitivity of gynecologic malignancies (1, 2). Given the sub-millimeter accuracy and precision of robotic SBRT (7), collateral ablative doses delivered to bone marrow were low overall (2%) and only slightly higher (6%) when SBRT targets were abutting vertebrae of the thoracolumbar spine or sacrum. For perspective, extended-field
Pre-therapy 5.5 3.9 11.9 35.9 228
(4.2-7.1) (3.6-4.2) (10.7-12.7) (32.6-38.8) (192-293)
2325 cGy 112 cm3 7.3 cm3 7% 4% Post-therapy 5.3 3.9 12.1 35.4 217
(4.3-7.4) (3.6-4.3) (10.7-12.7) (32.9-38.0) (166-286)
(330 cGy) (25 cm3) (4.1 cm3) (11%) (7%) P .465 .360 .344 .905 .305
radiation portals render a grade 3, 4, or 5 hematological toxicity rate of 15% (14). In the instance of abdominopelvic involved-field radiochemotherapy, radiation portals may be associated with a grade 3 or 4 acute hematological toxicity rate of 62% (15). Although, on occasion, ablative doses resulted in substantial bone marrow volumes receiving 1000 cGy or more (as shown in the example in Figure 2), substantial hematological toxicity was not apparent. In a prior study, SBRT-targeting of paraortic nodal sites of disease relapse achieved low bone marrow volume dose (5). Our use of SBRT in 57 women who had undergone multi-agent, multi-cycle chemotherapy was tolerable, as evidenced by manageable fatigue and minimal global alteration in complete blood cell counts. In other studies in which radiation and chemotherapy were co-administered (14, 15), adverse drops in hematologic counts occured in 15%-62% of women, sometimes compromising and stopping desired radiochemotherapy treatment. The fact that the majority of women (93%) in our study had chemotherapy antecedent to SBRT and then tolerated SBRT suggests that SBRT plus novel chemotherapy could be investigated in a phase 1 clinical trial setting. SBRT targeting cancer sites in close proximity to bones harboring marrow led to appreciable fatigue, as reported by one quarter of our patients. We were not able to detect differences in bone marrow constituents having short 6-8 hour (16) peripheral blood turnover (eg, white cells) nor having long 40-60 day (17) peripheral blood lifespans (eg, red cells). It is tempting to speculate that radiation-induced damage of nuclear and mitochondrial DNA in normal and cancer cells activates nucleoside supply de novo (18, 19) and salvage (20) dNTP supply pathways, and that a decrease in salvage reserve in the plasma impacts terminally differentiated muscle cells that depend on nucleoside salvage to replicate their mitochondrial DNA, thus creating symptoms of fatigue. Limitations of this retrospective study include the following: small sample sizes for each individual gynecologic malignancy type, which limits our assessment of clinical benefit and progression-free survival; potential investigator bias in retrospective assignment of toxicity criteria; inconsistencies in the timed collection of complete blood counts, which may underestimate hematological toxicity assessment; and limited overall follow-up. A low observed variation mean V10 for bone marrow in this study, perhaps resulting from few women being treated and from the submillimeter precision of robotic SBRT, raises a statistical hazard for missed association in treatment-related hematological adverse
e40
Kunos et al.
International Journal of Radiation Oncology Biology Physics
Fig. 3. Bone marrow volume receiving 1000 cGy radiation (V10) and differences in pre-therapy to post-therapy white blood cells (A), red blood cells (B), hemoglobin (C), and platelets (D). events. A prospective robotic SBRT trial for women with metastatic gynecologic malignancies with a quite similar cohort makeup has finished accrual within our group in an effort to report safety, tolerability, and outcomes measures.
2.
3.
Conclusion 4.
In conclusion, this retrospective study lends support to the concept that SBRT targeting abdominopelvic sites of gynecologic cancer does not generate substantial hematologic toxicity on its own, or when sequenced after chemotherapy. On the basis of these retrospective results, a phase I clinical trial of SBRT and chemotherapy co-administered in a similar population of women with metastatic gynecologic cancer is currently being considered.
5.
6. 7.
References 1. Kunos C, Zhang Y, Brindle J. Stereotactic body radiosurgery in gynecologic carcinomas. In: Ayhan A, Reed N, Gultekin M, et al.,
8.
editors. Textbook of Gynaecological Oncology. Ankara, Turkey: Gu¨nes Publishing; 2011. p. 835-841. Mayr N, Huang Z, Sohn J, et al. Emerging application of stereotactic body radiation therapy for gynecologic malignancies. Expert Rev Anticancer Ther 2011;11:1071-1077. Kunos C, Von Gruenigen V, Waggoner S, et al. Cyberknife radiosurgery for squamous cell carcinoma of the vulva after prior pelvic radiation therapy. Technol Cancer Res Treat 2008;7:375-380. Kunos C, Chen W, DeBernardo R, et al. Stereotactic body radiosurgery for pelvic relapse of gynecologic malignancies. Technol Cancer Res Treat 2009;8:393-400. Choi C, Cho C, Yoo S, et al. Image-guided stereotactic body radiation therapy in patients with isolated para-aortic lymph node metastases from uterine cervical and corpus cancer. Int J Radiat Oncol Biol Phys 2009;74:147-153. Kilby W, Dooley J, Kuduvalli G, et al. The Cyberknife robotic radiosurgery system in 2010. Technol Cancer Res Treat 2010;9:431-438. Wilcox E, Daskalov G. Evaluation of GAFCHROMIC EBT film for Cyberknife dosimetry. Med Phys 2007;34:1967-1974. Kunos C, Radivoyevitch T, Abdul-Karim F, et al. 18F-fluoro-2-deoxyd-glucose positron emission tomography standard uptake value as an indicator of cervical cancer chemoradiation therapeutic response. Int J Gyn Oncol 2011;21:1117-1123.
Volume 84 Number 1 2012 9. Kunos CA, Debernardo R, Fabien J, Dobbins DC, Zhang Y, Brindle J, Faulhaber PF. 18FDG-PET/CT definition of clinical target volume for robotic stereotactic body radiosurgery treatment of metastatic gynecologic malignancies. J Nucl Med Rad Ther 2011. http://dx.doi.org/10. 4172/2155-9619.S4-001. 10. Ellis R. The distribution of active bone marrow in the adult. Phys Med Biol 1961;5:255-258. 11. Mell L, Tiryaki H, Ahn K-H, et al. Dosimetric comparison of bone marrow-sparring intensity-modulated radiotherapy versus conventional techniques for treatment of cervical cancer. Int J Radiat Oncol Biol Phys 2008;71:1504-1510. 12. Roeske J, Lujan A, Reba R, et al. Incorporation of SPECT bone marrow imaging into intensity modulated whole-pelvic radiation therapy treatment planning for gynecologic malignancies. Radiother Oncol 2005;77:11-17. 13. Echner G, Kilby W, Lee M, et al. The design, physical properties and clinical utility of an iris collimator for robotic radiosurgery. Phys Med Biol 2009;54:5359-5380. 14. Varia MA, Bundy BN, Deppe G, et al. Cervical carcinoma metastatic to para-aortic nodes: Extended field radiation therapy with concomitant
Hematological toxicity after robotic radiosurgery
15.
16. 17.
18.
19.
20.
e41
5-fluorouracil and cisplatin chemotherapy: A Gynecologic Oncology Group study. Int J Radiat Oncol Biol Phys 1998;42:1015-1023. Lee YS, Kim YS, Kim JH, et al. Feasibility and outcome of concurrent chemoradiotherapy for recurrent cervical carcinoma after initial surgery. Tumori 2010;96:553-559. Summers C, Rankin S, Condliffe A, et al. Neutrophil kinetics in health and disease. Trends Immunol 2010;31:318-324. Kruse A, Uehlinger D, Gotch F, et al. Red blood cell lifespan, erythropoiesis and hemoglobin control. Contrib Nephrol 2008;161: 247-254. Kunos C, Chiu S, Pink J, et al. Modulating radiation resistance by inhibiting ribonucleotide reductase in cancers with virally or mutationally silenced p53 protein. Radiation Res 2009;172: 666-676. Kunos C, Radivoyevitch T, Pink J, et al. Ribonucleotide reductase inhibition enhances chemoradiosensitivity of human cervical cancers. Radiation Res 2010;174:574-581. Kunos C, Ferris G, Pyatka N, et al. Deoxynucleoside salvage facilitates DNA repair during ribonucleotide reductase blockade in human cervical cancers. Radiat Res 2011;176:425-433.