		Conventionally fractionated dose escalation above 78 Gy for intact prostate cancer has been shown to improve failure- free survival. Previously occult prostatic fossa recurrences can now occasionally be visualized using advanced imaging methods such as MRI. As the disease is no longer “microscopic,” it stands to reason that dose escalation to the recurrent tumor may improve failure-free survival in the salvage setting as well.

		In the current study, the authors report on the feasibility and acute toxicity of the first 14 patients enrolled on a trial randomizing subjects to either standard 68 Gy radiotherapy to the fossa, or the same plus a simultaneous integrated boost of 76.5 Gy at 2.25 Gy/fx (EQD2= 82 Gy, alpha/ beta assumption 1.5).

		In this initial report, the authors showed the dosimetric feasibility via DVH parameters of the approach and that there was clinically no discernible acute toxicity difference between the standard arm and the dose-escalated arm. The trial continues to accrue and we await reports on late toxicity as well as impact on their primary endpoint of biochemical failure-free survival. If the authors ultimately show improved bRFS and similar late toxicity, then dose escalation to visualized fossa lesions should become a new care standard.

	MAPS is the first phase III randomized trial of MRI-mapped, dose-escalated salvage radiotherapy. In this planned feasibility analysis, we ensure dosimetric adequacy of the protocol as it relates to acute toxicity.

	Two intensity modulated radiotherapy plans were generated for each patient. In the standard fraction radiotherapy arm, 68 Gy in 34 fractions was prescribed to ≥95% of the planning target volume. In the simultaneous incorporated hypofractionated boost (SIHB) arm, an additional 2.25 Gy daily SIHB was prescribed to the gross tumor volume (GTV). The trial stipulates that ≤35% and ≤55% of the rectum should receive ≥65 Gy and ≥40 Gy, respectively, and ≤50% and ≤70% of the bladder minus the clinical target volume (B-CTV) should receive ≥65 Gy and ≥40 Gy, respectively. Acute toxicities were recorded per National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE v4.0).

	In all plans, ≥95% of the planning target volume and GTV received the prescribed dose. Dosimetric constraints were achieved for all organs at risk except B-CTV. The highest toxicity recorded was grade 2 gastrointestinal toxicity: 1 episode per arm.

	Dose escalation is achievable with expected variations in cases with small bladders. There was no observed increase in acute toxicity.

Radical prostatectomy (RP) cures the majority of patients with localized prostate cancer. However, within 10 years of diagnosis as many as one-third will develop recurrent disease.

 In the setting of localized recurrence, salvage radiotherapy to the prostate bed is the standard treatment. However, long-term salvage rates remain suboptimal, possibly due to poor patient selection.

Ideally, only those with disease isolated to the prostate bed would receive treatment. Dynamic contrast-enhanced (DCE) MRI has been shown to reveal residual or recurrent disease with specificities of at least 80%, and sensitivities ranging from 67% to 97%.

 This is an improvement over transrectal ultrasound and conventional imaging,

 provides objective evidence for treatment, and identifies a target for dose escalation.

Compared with definitive radiotherapy, salvage doses are lower due to the assumption of only microscopic disease and toxicity limits associated with treating the bladder neck.

However, dose escalation may show benefit,

 and radiobiological modeling predicts that PSA control rates will increase along with increasing doses.

 In terms of toxicity, salvage IMRT up to 76 Gy results in <1% grade 3 GI toxicity, and 3% grade 3 GU toxicity at 5 years.

If there is a potential dose response above what is thought permissible in terms of toxicity, and we have an emerging ability to visualize foci of gross disease in the prostate bed, it follows that the addition of targeted dose escalation may improve outcomes compared with standard salvage radiotherapy alone. This is the focus of the MAPS trial— the first phase III randomized trial of MRI-mapped, doseescalated salvage radiotherapy. The primary objective is to determine the effect of an SIHB to MRI-identified foci on initial complete biochemical response (initial PSA <0.1 ng/mL at 9 months). The trial has a planned accrual of 76 evaluable patients, which would provide 80% power to detect an absolute increase of 27% in the PSA response rate, using a one-sided Fisher’s exact test with 5% significance level. This feasibility analysis details dosimetry and acute toxicity, revealing dose escalation as safely achievable.

		Figure 1. Typical DCE-MRI appearance of the non-contoured (above) and contoured GTV (below).

		DCE-MRI indicates dynamic contrast-enhanced MRI; GTV, gross tumor volume.

Between December 2010 and September 2013, 14 patients who had developed biochemical recurrence after definitive RP were enrolled in our institutional review boardapproved MAPS trial. The trial stipulates that all patients are ≥3 months post RP with the following inclusion criteria: a PSA of ≥0.1 ng/mL and ≤3.0 ng/mL within 3 months of enrollment, an MRI detectable lesion in the prostate bed, no evidence of distant metastasis, a serum total testosterone within 40% of normal assay limits within 3 months prior to enrollment, a BUN and creatinine within 40% of normal assay limits within 3 months prior to enrollment, no current active malignancy aside from nonmetastatic skin cancer or early chronic lymphocytic leukemia, Zubrod performance status <2, and age between 35 and 85 years. Any prior androgen deprivation therapy must be completed >6 months previous to enrollment and be ≤7 months in duration. Patients meeting the criteria were randomized between the SFRT arm and the SIHB arm.

		SFRT indicates standard fraction salvage radiotherapy; PTV, planning target volume; SIHB, simultaneous integrated hypo-fractionated boost; GTV, gross tumor volume; B-CTV, bladder minus clinical target volume.

All patients underwent computed tomography (CT)-based virtual simulation in the supine position with 1.5-2-mm thick slices extending from the top of the sacrum to mid femur. Per institutional protocol, all patients were instructed to eat a diet designed to reduce bowel gas for at least 24 hours prior to simulation, use a rectal enema prior to arrival, and maintain a full bladder. Tattoos were placed at the anterior, right lateral, and left lateral isocenter skin points. Under the same conditions and often during the same encounter, T2, T1 non-contrast, diffusion weighed images (DWIs), and DCE-MRI (3.0 T MRI) scans were obtained at 2.5-mm intervals. Approximately 12 DCE-MRI scans were obtained, beginning precontrast and continuing at approximately 30-second intervals postcontrast. The pelvis and prostate bed were imaged.

		Figure 2. Typical dose distributions for SFRT (left) and SIHB (right) plans.

		SFRT indicates standard fraction salvage radiotherapy; SIHB, simultaneous integrated hypo-fractionated boost.

In order to deliver the daily SIHB to the MRI-identified foci, the anatomical, diffusion, and dynamic MRI sequences were coregistered to the planning CT. Coregistration was performed within the Eclipse software using anatomical matching of the MRI and CT acquisitions in the areas of the prostate bed, penile bulb, and bladder and rectal interfaces in the region of the prostate bed. This resulted in the high quality and reproducible alignment between the two acquisitions necessary to ensure that the SIHB was delivered to the proper location.

Structures were manually contoured on the CT simulation scan using an adaptation of the recommendations of the Radiation Therapy Oncology Group.11 The clinical target volume (CTV) included the prostate bed, extending cranially from the seminal vesicle remnants and caudally to approximately 2.5 mm above the superior aspect of the penile bulb. Borders below the superior border of the pubic symphysis were the posterior aspect of the pubis anteriorly, the rectum posteriorly, and the levator ani laterally. Borders above the pubic symphysis were approximately 1 cm of the posterior bladder wall anteriorly, the mesorectal fascia posteriorly, and the sacrorectogenitopubic fascia laterally. Using CT-MRI coregistration, the MRI revealed that GTV 

 was contoured on the CT simulation scan by using the anatomical, diffusion, and dynamic MRI series. Suspicious lesions are most visible on the early and late DCE-MRI sequences (characterized by early contrast wash-in and late contrast wash-out). However, the T2 and DWI images were also used to confirm the lesion location. There was no expansion on the GTV, as the goal was to increase dose to the general area of the GTV. In addition, while the effect was greatest in the area of the GTV, the presence of a boost volume increased dose to the entire prostate bed. Therefore, adding a margin to the GTV would unnecessarily place critical organs at risk. A PTV around the CTV was defined to account for daily setup errors, and consisted of a 7 mm expansion of the CTV in all directions. The bladder was contoured in its entirety. The rectum was contoured as a whole organ from the rectosigmoid flexion to the bottom of the ischial tuberosities. The femoral heads were outlined from the top of the acetabulum to an area between the lesser and greater trochanters. For evaluation purposes, a bladder minus CTV volume was also created (B-CTV). All contours were reviewed by the principle investigator or co-investigator.

Treatment plans were generated using commercial software (Eclipse, version 11.0, Varian, Palo Alto, California) with heterogeneity correction. Two IMRT plans were generated for each patient data set: 1 for the SFRT arm and 1 for the SIHB arm (regardless of the arm under which the patient was actually treated). The prescription dose was 68 Gy in 34 fractions to ≥95% of the PTV in all plans. In the SIHB arm, a 2.25-Gy daily SIHB was prescribed to the GTV, for an absolute dose of 76.5 Gy in 34 fractions 

. The plans were quality controlled by utilizing the ArcCHECK QA device (Sun Nuclear Corporation, Melbourne, Florida) to compare the calculated and measured doses within the phantom using gamma analysis with 3% dose difference and 3-mm distance- to-agreement criterion.

A Varian Trilogy (Varian, Palo Alto, California) linear accelerator delivered volumetric- modulated arc therapy (VMAT) at either 6 MV or 18 MV energies. All patients were treated once daily, 5 times a week. Image guided radiotherapy (IGRT) was accomplished with daily cone beam CT (CBCT), using prostate bed clips and bladder and rectal filling to perform daily alignment. Patients were not treated unless aligned to any clips, with bladder and rectal filling similar to the CT simulation. Patients were instructed to express gas/stool, and urine and drink additional fluids before a repeat CBCT, as needed.

Acute toxicity was defined as during and up to 3 months after treatment. Prostate- related symptoms were assessed before treatment, weekly during treatment, and 6 weeks and 3 months after treatment completion using the NCI CTCAE v4.0. Only patients with data at all above time points were included.

The percentage of the volume of the PTV and GTV receiving the prescribed dose was recorded for the SFRT and SIHB plans. Also recorded were the maximum doses to the OARs, as well as the percentage of volume of the rectum and B-CTV receiving ≥65 Gy and ≥40 Gy. The Shapiro- Wilk normality test was used for the variables of interest. The Student’s T test was used for comparison between the SFRT and SIHB plans when the normality test was not significant. Otherwise, the Kruskal-Wallis test was used for comparing the two plans. A 2-tailed 

 value (<.05) was used to indicate statistical significance.

Fourteen patients have been enrolled on the trial to date, and data from 13 are available for acute toxicity analysis (7 randomized to SFRT and 6 to SIHB treatment). The median follow-up at this time is 15 months.

The mean CTV volume was 143.3 cc, with a range of 84.8-202.7 cc SD, 39.2 cc). The mean PTV volume was 312.8 cc, with a range of 218.3-388.9 cc (SD, 57.4). The mean GTV volume was 2.34 cc, with a range of 0.31-10.4 cc (SD, 2.8 cc). The mean volume encompassed within the 76.5 Gy isodose line was 27.6 cc, with a range of 1.02-151.3 cc (SD, 39 cc).

Dose-volume histograms (DVH) were generated for all plans (Figure 3). In every case, at least 95% of the PTV received 68 Gy, and at least 95% of the GTV received 76.5 Gy, as specified in the constraints 

, with maximum doses to the GTV kept below 115% of the prescription dose. SIHB plans demonstrated an increased integral dose to the prostate bed PTV as shown in Figure 3. While this increase occurred throughout the PTV volume, it was greatest in the region of the GTV.

		Table 2. Dosimetric results for SFRT and SIHB plans.

 P value from Kruskal-Wallis test due to non-normality. R indicates rectum; B-CTV, bladder minus clinical target volume; GTV, gross tumor volume; PTV, planning target volume; SFRT, standard fraction salvage radiotherapy; SIHB, simultaneous integrated hypofractionated boost; NA, not applicable.

Dosimetric values for target volumes and OAR are presented in

 for both SFRT and SIHB plans. We tested normality for each variable using the Shapiro- Wilk test. V76.5 (GTV)% in SIHB and V68 (PTV)% in SFRT were statistically significant (both 

 <.001), ie, not normal. All other variables were assumed to be normal. Dosimetric constraints were achieved for all OARs except for bladder. Five plans had >70% of the bladder receiving ≥40 Gy in both the SFRT and SIHB plans, and 1 plan had >50% of the bladder receiving ≥65 Gy in the SIHB plan only. Review of the plans revealed this to be due to suboptimal bladder filling. In terms of dosimetric constraints for OARs, as well as PTV and GTV coverage, there was no difference between the SFRT and SIHB plans per patient or overall, except in the case of V68 (PTV)% 

		Table 3. Number of radiotherapy-related toxicities per arm.

		SFRT indicates standard fraction salvage radiotherapy; SIHB, simultaneous integrated hypofractionated boost.

		Figure 3. Typical dose volume histogram for SFRT and SIHB plans. All SIHB DVH curves are shifted slightly to the right. The GTV for the SIHB plan is also included. SFRT

		SFRT indicates standard fraction salvage radiotherapy; SIHB, simultaneous integrated hypo-fractionated boost; DVH, dose-volume histogram, GTV, gross tumor volume.

Thirteen patients (7 treated with SFRT and 6 with SHIB) are included in the toxicity analysis. Twelve of these patients experienced toxicities (1 patient in arm 2 did not). 

 details the number of occurrences of all radiation related toxicities by treatment arm. 

 details the highest grade GU and GI toxicity per patient.

Salvage radiotherapy is the only curative treatment for patients with a localized recurrence post RP. Multiple retrospective studies have reported on the efficacy of standard dose salvage radiotherapy in achieving biochemical control, with actuarial 5-year biochemical recurrence free survival (bRFS) rates ranging from 10% to 66%.

 Advances in IMRT and IGRT have now made possible the delivery of definitive-range radiotherapy doses. Dose escalation in the salvage setting has been shown to improve biochemical control in various retrospective studies.

King et al compared 38 patients treated with 60 Gy with 84 patients treated with 70 Gy, demonstrating a significant improvement in 5-year bRFS from 25% to 58% with the higher dose. Another study reviewed 364 men who received salvage radiotherapy at 1 of 3 dose levels (low, <64.8 Gy; moderate, 64.8-66.6 Gy; high, >66.6 Gy) and found that doses greater than 66.6 Gy resulted in improved bRFS.

 A review of published reports by Ohri et al in 2012 examined the role of salvage radiotherapy dose and timing on biochemical control and found that bRFS increased with salvage radiotherapy dose by 2.5% per Gy and decreased with pre-salvage radiotherapy PSA by 18.3% per ng/mL (

 <.001). Radiobiological models predicted that an increase in the pre-salvage radiotherapy PSA from 0.4 to 1 ng/mL would increase the salvage radiotherapy dose required to achieve a 50% bRFS rate by 60 to 70 Gy.

 While randomized trial data are not yet available, these studies demonstrate the potential for improved outcomes with dose escalated salvage radiotherapy.

The benefits of dose escalation to the prostate bed are limited by the possible increase in side effects. Three-dimensional conformal radiotherapy (3D-CRT) first allowed for escalation to biologic equivalent doses of 68 Gy16 without an increase in toxicity compared with 64 Gy delivered by conventional radiotherapy (<5% late grade 3 GU toxicity and <5% late grade 3 GI toxicity).

However, escalation >68 Gy using 3D-CRT resulted in a 5-year risk of late grade 3 GU toxicity of 16%.

 The model by Ohri et al estimated that severe late toxicity rates may reach 10% at doses of approximately 70 Gy;

 however, toxicity data used for this model came from series employing 2-D or 3-D treatment techniques. In fact, it was recently reported that IMRT to a median dose of 76 Gy resulted in a 5-year risk of late grade 3 GU toxicity of 3%, and late grade 2 to 3 GI toxicity of 8%,

 which is comparable to toxicity reported with doses of 64 Gy using conventional radiotherapy.

 In our trial, with only acute toxicity available at this time, the highest reported toxicity included 1 episode of grade 2 gastrointestinal toxicity in each arm. Finally, be aware that the CTCAE grading system used in our trial may report lower grade toxicity for similar side effects compared with other scoring scales, such as that used by the Radiation Therapy Oncology Group (RTOG).

With retrospective and radiobiologic modeling evidence for improved bRFS with dose escalation,

 reservations regarding potential increases in toxicity,

 and the ability to identify gross disease in the prostate bed using DCEMRI, 

 it follows that dose escalation to the gross disease alone has the potential to improve the therapeutic ratio of salvage radiotherapy.

In this study, we have demonstrated that even though areas of gross disease in the prostate bed are close to critical structures, dose escalation to these areas results in good coverage of the PTV (range: 95%-97.9%) and GTV (range: 95%-100%) without exceeding conservative dosimetric constraints specified for the rectum in all cases and the bladder in the majority of cases. This is occasionally unavoidable, as the bladder neck must be pulled into the CTV to account for the absence of the intraprostatic urethra, and some patients are unable to achieve or maintain a full bladder (due to anatomic or physiologic limitations). In cases like this, bladder constraints were unachievable in both the SFRT and SIHB plans. Per the protocol, a primary variation will be noted if up to an additional 7.5% of the B-CTV volume exceeds the dose constraint. Beyond this constitutes a secondary protocol variation. At the least, a portion of the bladder is included in the PTV by necessity; these constitute protocol variations, not violations. Therefore, none of the plans would have resulted in a protocol violation. Regardless, the clinical acute toxicity was comparably low in both arms.

		Table 4. Highest grade GU and GI toxicity per patient according to NCI CTCAE v4.0.

Articles

Treating MRI-identified prostate lesions to definitive doses using an SIHB is possible without significantly increased dose or acute toxicity to critical structures

Alan Pollack, MD, PhD

Articles

A Phase III Randomized Trial of MRI-Mapped, Dose-Escalated Salvage Radiotherapy Post-Prostatectomy: The MAPS Trial