This prospective pilot study was carried out according to the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) statement [15] and was approved by our Institutional Review Board.

Inclusion criteria for this study were: (1) histologically confirmed primary malignant bone tumor in the sacrum for which surgery and radiation or radiation alone was planned, (2) age 18 years or older, (3) normal organ and marrow function (for surgical arm). Normal organ and marrow function was defined as: total bilirubin within normal institutional limits, AST (SGOT)/ALT (SGPT) < 2.5 × institutional upper limit of normal, and creatinine within normal institutional limits or creatinine clearance > 60 mL/min/1.73 m2 for subjects with creatinine levels above institutional normal limit, (4) ability to understand and willingness to sign a written informed consent.

As tetracyclines and ionizing radiation are harmful to the developing human fetus, women of child-bearing potential had to agree to use adequate contraception (hormonal or barrier method of birth control; abstinence) prior to study entry and until after the last study-related CT scan.

Exclusion criteria for this study were: (1) history of surgery, chemotherapy, or RT of the sacrum prior to the study, (2) history of allergic reactions attributed to compounds of similar chemical or biologic composition to tetracyclines (for surgical arm), (3) pregnancy or nursing, (4) uncontrolled illness including, but not limited to ongoing or active infection, symptomatic congestive heart failure, unstable angina pectoris, cardiac arrhythmia, or psychiatric illness/social situations that would limit compliance with study requirements.

Patients were enrolled between September 2015 and April 2018 at a single tertiary referral center, prior to the start of treatment for a primary malignant sacral tumor. According to their treatment plan (high-dose RT alone, or a combination of preoperative high-dose RT and surgery followed by additional postoperative RT), determined by the treating physicians and the patients themselves, patients were either enrolled in the non-surgical arm or the surgical arm (Fig. 1). The non-surgical arm consisted of a period of seven weeks of RT (77.4–79.2 Gy), while the surgical arm received pre-operative RT (50.4 Gy in all patients) and post-operative RT (19.8–27 Gy depending on final margins). RT consisted of a combination of protons and/or photons. Radiation treatment clinical target volume (CTV) consisted of a gross target volume (GTV) including MRI and CT extent of gross tumor plus expansion margins to include vertebral body and spinal canal extension of one level above and below to account for potential microscopic extension through the dorsal venous plexus or spinal canal, plus 2 to 3 cm of margin into involved muscles such as erector spinae, gluteus muscles, and/or piriformis muscles (Fig. 2). All subjects underwent two QCTs, and for the surgical arm two biopsies. All patients in the non-surgical arm were followed until after their last radiation treatment, while all patients in the surgical arm were followed until after their last surgery date.

Timeline for both the non-surgical and the surgical arm

Radiation field of A preoperative 50.4 Gy RT for sacral chordoma followed by resection and B definitive RT for sacral chondrosarcoma

All subjects underwent QCT of the L1 and L2 vertebra and the sacrum by using a second-generation dual-source 128-row multidetector CT scanner (Somatom Definition Flash; Siemens Medical Solutions, Forchheim, Germany). Subjects were positioned on a QCTPro calibration phantom (Mindways Software, Inc., Austin, TX). Helical scans were performed at 80 and 140 kV by using 210 and 80 mA, respectively. Other scanning parameters included 1-s gantry rotation time, 0.9:1 pitch, and 64 × 0.6-mm detector configuration with double z-sampling. The images were reconstructed at 2-mm section thickness and 2-mm section interval by using the I31f reconstruction kernel with a sinogram-affirmed iterative reconstruction (SAFIRE; Siemens Healthcare). Single-energy volumetric BMD of the lumbar spine and sacrum was assessed using the 140-kV data set. Three-dimensional reconstructive analysis was performed by using QCT PRO software version 5.1 (Mindways Software, Inc., Austin, TX).

For patients in the surgical arm, QCT was performed prior to the start of pre-operative RT and after completion of pre-operative RT (50.4 Gy). For patients in the non-surgical arm, QCT scan was performed prior to the start of their RT and at approximately 3 months after RT.

In the surgical arm, patients self-administered two cycles of the bone-labeling drug demeclocycline (150 mg, four times a day by mouth) 18 days before starting RT, using a standard 3 days-on, 12 days-off, 3 days-on regimen. Twenty-four days before surgery, shortly after completion of 50.4 Gy preoperative RT, patients self-administered two cycles of the second bone-labeling drug, tetracycline (250 mg, four times a day by mouth), also using a standard 3 days-on, 12 days-off, 3 days-on regimen. The use of the quadruple labeling technique allowed longitudinal assessment of the dynamic parameters of bone formation before and after RT in a single biopsy sample [16, 17]. During the two-stage surgery, biopsies were taken: a biopsy from the non-irradiated iliac crest during the first stage and a biopsy from the radiated sacrum during the second stage.

Samples were prepared and analyzed as previously described [16, 17]. Briefly, biopsy samples were fixed in 70% ETOH, dehydrated, and embedded in methyl methacrylate. Thin sections were cut at 7 μm prior to staining with Goldner's trichrome and toluidine blue for analysis of static parameters, and at 20 μm, unstained, for analysis of dynamic parameters. All histomorphometric analysis was performed by the same individual (HZ) using OsteoMeasure software (OsteoMetrics Inc., Decatur, GA). All indices were calculated and according to the recommendations of the ASBMR Nomenclature Committee [18]. One patient in the surgical arm did not take the second two cycles of tetracycline, which made it impossible to obtain post-treatment bone formation parameters by bone histomorphometry.

Berthold and Haras published a range of normal reference values for the trabecular bone mineral density measured by quantitative CT of the lumbar spine in young adults [19]. In healthy males and females the mean trabecular bone mineral density was 150 mg/mL with a standard deviation of 20 mg/mL at the start of puberty. Assuming that in our population this is at least 1 standard deviation less that of a normal young adult, we get a reference estimate of 130 mg/dL at the start of treatment in all study patients. An ante-hoc power calculation determined that a sample of 8 patients would provide 89% of statistical power (alpha 0.05) to detect a difference of 20 mg/cm3 assuming a standard deviation of the change is 15 mg/cm3.

Wilcoxon signed-rank tests were performed to assess the difference between variables before and after radiation. A two-sided p value of < 0.05 was considered significant. All statistical analyses were performed using Stata Version 13.0 (Stata Corp, College Station, TX, USA).