Image-guided radiation therapy
One of the most crucial aspects of patient treatment involves the setup and target localization before each treatment fraction. New technologic advances have dramatically improved the accuracy of IGRT.17–19 These techniques include daily CT scans acquired by an imaging system mounted on the linear accelerator,20 radio-opaque fiducial markers placed in the prostate and imaged with daily X-rays,21 and radiofrequency beacons that can be implanted, similar to fiducials.22 The position of those radiofrequency beacons can be electronically determined and allows real-time monitoring of the position of the prostate during therapy. This helps reduce rectal injury as the radiation beam can be turned off if the target beacons move outside a predetermined location. Rajendran et al23 evaluated dosimetric consequences of daily isocenter corrections during prostate cancer radiation therapy of 28 men using an electromagnetic beacon system. They calculated the daily corrections in positioning based on real-time data obtained by the beacons during delivery of 79.2 Gy. They reported that, without daily electromagnetic localization, 70% of the rectum received an additional 10 Gy.
To investigate the importance of IGRT in prostate radiation therapy, Chung et al24 examined 25 patients treated with high-dose IMRT. Radio-opaque fiducials were placed into the prostate in 15 patients and were compared to 10 patients without fiducials. The planning target volume dose coverage was not significantly different between the two groups. However, the volume of rectum and bladder receiving ≥40, ≥60, and ≥70 Gy were all significantly less using IGRT (P < 0.001). This correlated with lower acute RTOG Grade 2 rectal toxicities (80% vs 13%, P = 0.004).
Dose–volume histograms (DVHs)
Given the increased use of IMRT, quantitative assessments of the doses delivered to normal organ volumes have become increasingly important. With the use of 3D planning, reliable dose–volume relationships have been developed to reduce toxicity. This has also led to the establishment of normal tissue complication probability (NTCP) models to allow radiation oncologists to protect normal tissues by setting predetermined dose–volume constraints at the time of RT treatment planning.
Fiorino et al25 reviewed the existing literature on NTCP models and dose constraints of pelvic organs and related these constraints to specific pelvic toxicities. The authors found relatively consistent agreement among investigations on the DVH relationships to rectal bleeding. Additionally, the strict application of rectal dose–volume constraints in high-dose IMRT translated into a greatly reduced rate of bleeding.25–32 After reviewing available literature, they found that keeping the volume of rectum receiving >70 Gy and >75 Gy (V70Gy and V75Gy) to <25% and <5%, respectively, reduced the incidence of late rectal bleeding.33–35 In addition to keeping the high-dose volumes (V70 and V75) as low as possible, the moderate dose volumes (V40–V50) can also play a key role in the development of rectal bleeding. The authors found an increased incidence of rectal bleeding when large portions of the rectum received 40–50 Gy.
Two large clinical trials reported correlations between rectal DVHs and rectal incontinence.32,36 The authors suggested keeping the V40Gy to <65–70% to reduce the risk of chronic late incontinence, defined as using pads, to <1.5%. Others reported that large volumes (80–90%) of the rectum receiving 40–50 Gy were predictive of late incontinence.36 They hypothesized that the mechanism involved the reduced ability of the rectal mucosa to absorb water.
The majority of dose–volume investigations have focused on late rectal toxicities and the doses that may predict for these events. While fewer studies examined acute rectal toxicity, they are equally important as acute rectal injuries predict for chronic injuries and can lead to worsened patient QOL and treatment interruptions. Michalski et al37 reported on 262 patients that were enrolled on RTOG 9406, a phase I–II dose escalation trial using 3DCRT for localized prostate cancer. They used a rectal constraint of V65Gy <20%. They found that acute toxicity was remarkably low at 3%, with no Grade 4 or 5 acute toxicities reported. Peeters et al38 examined dose–volume data and correlated this to acute toxicity. They found that both intermediate and high doses delivered to the rectum correlated with GI toxicity. They also found that the rectal lengths irradiated to doses greater than 5, 15, and 30 Gy were correlated with acute side effects.
In a recent study, Schaake et al39 developed multivariable NTCP models for late rectal bleeding, stool frequency, and fecal incontinence in men that received pelvic radiation therapy for prostate cancer. They prospectively analyzed 262 patients. Anorectal toxicity was assessed, and the authors identified and contoured different anatomical subregions within and around the anorectum for dosimetric analyses. They found that rectal bleeding was associated with the anorectum V70. Fecal incontinence was associated with the external sphincter V15 and the iliococcygeal muscle V55. Moreover, they found that increases in stool frequency were associated with the iliococcygeal muscle V45 and the levator ani V40. They did not find significant associations for rectal pain.
Given that most of the aforementioned studies defined dose–volume constraints based primarily on 3D-conformal RT, Mirjolet et al40 studied whether these same constraints applied to IMRT. They retrospectively examined 180 patients with prostate cancer treated with IMRT and looked at standard dose–volume-specific endpoints, specifically the volume of rectum receiving from 25 to 75 Gy (V25–V75), expressed in percentages (%) and in cubic centimeters (cc). They then calculated the area under the DVH curve between 25 and 50 Gy for the rectum (rAUC25–50), which they hypothesized would more accurately predict risk of rectal events. The rAUC25–50 calculated in cubic centimeters correlated with any grade ≥1 acute GI toxicity (P = 0.028). Based on this, the authors recommended that the rAUC25–50 of the entire rectum should not exceed 794 cc.Gy.
The aforementioned studies examined the association between rectal injury and dose to the rectum using whole organ at risk (OAR) models, where the entire rectum is assumed to be homogeneously radiosensitive. However, Dréan et al41 recently published a unique analysis where the authors set out to identify rectal subregions at risk (SRR) that correlated with rectal bleeding and compared these SRRs to conventional whole-organ rectal DVHs. The authors prospectively treated 173 patients with localized prostate cancer and generated 20 geometric rectal delineations for all patients. These delineations included the whole rectum, anterior hemirectum, craniospinal rectum thirds (inferior, medium, and superior), the part of the rectum directly in front of the prostate, and the portions of the rectum localized at <5, 10, 15, and 20 mm from the prostate surface. DVHs were then calculated in three types of rectal subregions: “geometric”, “personalized”, obtained by non-rigid registration followed by voxel-wise statistical analysis (SRRp); and “generic”, mapped from SRRps, located within 8×8 rectal subsections (SRRg). Through this non-rigid registration and voxel-wise statistical analysis, the authors were able to identify a specific SRRp for each patient representing <4% of the absolute rectal volume as a whole. These rectal subregions were primarily located in the subprostatic anterior hemirectum and upper part of the anal canal. The dose delivered to these subregions in patients suffering from rectal bleeding was almost 4 Gy greater (up to 6.8 Gy) than for patients without rectal bleeding.