62_03
General Management Trends in the United States
The optimal management of clinically localized prostate cancer remains controversial and is often a source of great frustration and anxiety for many patients who are compelled to make a decision regarding a treatment intervention for their disease. The practitioner must be aware that the natural history of this tumor is variable and influenced by multiple prognostic factors. All of the various forms of therapy for prostate cancer can affect quality of life and sexual function in varying degrees. In the process of counseling and discussing therapeutic options, it is important to present all available data regarding the variable natural history of this disease, prognostic significance of the diagnosis, potential therapeutic benefit of the various modalities, and immediate as well as late treatment-related sequelae. Life expectancy and quality of life considerations should be carefully discussed with the patient and spouse or significant other.
Based on the available data, when comparing patients with similar prognostic features, there are no significant differences in the biochemical and disease-free survival outcomes for patients with early stages of disease treated with RP, high-dose external-beam radiation therapy (EBRT), or interstitial implantation (97,220,457). In the absence of randomized trials demonstrating superiority of one treatment over another, there have been wide geographic variations in the preferred therapeutic intervention for early-stage prostate cancer currently practiced throughout the United States.
Observations from the CapSURE database (Cancer of the Prostate Strategic Urologic Research Endeavor), a registry of 10,000 men accrued from community-based urologic practices across the United States, has shed further light on practice patterns. Cooperberg et al. (86) reported an increasing trend for patients presenting at diagnosis with more favorable risk disease in 2001–2002 (47%) compared to 1989–1990 (31%). A lower percentage of high-risk patients (defined as PSA >20 ng/mL, Gleason 8-10 disease or T3-T4 disease) were noted on initial presentation (15% compared with 41%) for 2000–2001 and 1989–1990, respectively. Practice patterns have also changed in more recent years. The use of expectant management has decreased from 20% in 1993 to 8% in 2001. A slight decrease in the use of prostatectomy and EBRT was reported with a significant increase in brachytherapy and primary androgen-deprivation therapy treatment interventions. Finally, there has been a steady increase in the use of neoadjuvant androgen-deprivation therapy in conjunction with planned radiotherapy. Comparing trends from 1989–1990 and 1999–2001, the use of neoadjuvant androgen deprivation has increased from 10% to 75% for those receiving EBRT, and 7% to 25% for those receiving brachytherapy.
Zelefsky et al. (444) reported significant changes of radiation therapy practice across the United States based on comparisons of the 1994 and 1999 surveys of the American College of Radiology Patterns of Care Survey. Overall, it was observed that, compared with the 1994 survey, there were significant changes in the practice and delivery of radiation therapy during the more recent survey period. Specifically, more patients, especially with favorable-risk disease, were treated with implantation compared with prior years. The implant-treated population was noted to be younger than the patients receiving EBRT. In the 1999 survey, higher doses of EBRT were more frequently delivered using 3DCRT techniques compared with what was reported in the 1994 survey. It was also noted that there was a substantial increase in recent years in the percentage of patients treated with androgen deprivation in conjunction with radiation therapy.
Several authors, including Adolfsson et al. (2) and Johansson et al. (200), have reported on patients, age 60 to more than 80 years, who, on histologic diagnosis of carcinoma of the prostate, were managed conservatively and observed without specific anticancer treatment until symptoms developed. Recently, a phase III randomized trial from the Scandinavian Prostate Cancer Group demonstrated the benefits of treatment intervention compared with an expectant management approach (39). In this study, 695 men were included with early stage T1 or T2 prostate cancer. The median follow-up was 8.2 years, the primary end point was death attributed to prostate cancer, and the secondary end point was overall survival. In the watchful waiting group, 8.9% died of disease compared with 4.6% in the RP group (p = .02). Men assigned to RP had a significantly lower incidence of distant metastases compared with those who underwent watchful waiting. However, the overall survival rates between the two groups were noted to be same.
In today's health environment in the United States, it is often not considered acceptable to delay definitive therapy for most patients with localized carcinoma of the prostate, except in selected elderly patients with low Gleason scores and low-volume disease based on the biopsy findings, as well as patients with significant medical comorbidities. Properly designed prospective clinical trials are critically needed to better define the efficacy and cost-effectiveness of various therapeutic approaches for localized carcinoma of the prostate.
Traditionally, the treatment options for patients with early-stage, clinically localized prostatic cancer (stages T1c or T2) have included RP, EBRT, or permanent interstitial implantation. Surgical techniques have significantly improved with the advent of the nerve-sparing operation popularized by Walsh (416), with a lower incidence of sexual impotence (approximately 30% to 60%, depending on the patient's age, tumor stage, and surgery extent) compared with classic RP (almost 100%), as well as improved methods available to reduce risks of posttreatment urinary incontinence. At the same time, the accuracy and safety of EBRT delivery have significantly improved with the emergence of 3DCRT and IMRT approaches. Permanent interstitial implantation using ultrasound-guided transperineal techniques and the recent developments of intraoperative conformal optimization for prostate brachytherapy have consistently improved the dose distributions for this treatment approach, leading to improved outcomes and decreased toxicities. For patients with intermediate-risk and selected unfavorable-risk features, in
P.1457
addition to high-dose EBRT alone, the combination of EBRT with permanent interstitial implantation or a high–dose-rate (HDR) brachytherapy boost represents appropriate treatment interventions in addition to RP.
Treatment Techniques
Radical Prostatectomy
RP, initially described by Young (436) in 1905 and popularized by Jewett (199), is a therapeutic option when the tumor is confined to the prostate; according to most urologists, it has no role in the management of gross extracapsular disease, seminal vesicle involvement, or in the presence of lymph node metastases. Two approaches for the classic RP are used: retropubic and perineal. The procedure consists of complete removal of the prostate and its surrounding capsule together with the seminal vesicles, the ampulla, and the vas deferens. The prostate is removed completely by excision of the urethra at the prostatomembranous junction, leaving no residual prostatic tissue at the apex. The retropubic approach is preferred by many urologic surgeons; this procedure also facilitates access for performing a bilateral pelvic lymph node dissection.
Walsh and Donker (417) described the anatomic basis for sexual impotence after RP and, based on that, reported on techniques to achieve a nerve-sparing approach. Through detailed anatomic studies, they demonstrated that the branches of the pelvic autonomic plexus that innervate the corpora cavernosa are located between the rectum and the urethra, along the lateral aspect of the prostate, and penetrate the urogenital diaphragm near or in the midmuscular wall of the urethra. Later they described the technique for radical retropubic prostatectomy with preservation of the neurovascular pedicle (416). They reported preservation of sexual function in 73% of 250 patients treated with this operation; the incidence of sexual impotence was correlated with the age of the patient at the time of surgery (418).
In the last 5 years there has been an increasing interest and practice of laparoscopic radical prostatectomy both in the United States and Europe. Advantages cited for the use of the laparoscopic procedure versus the open approach include improved visualization of the anatomy optical magnification, less blood loss, less postoperative pain, and more rapid resumption of normal activities (181) Preliminary functional and oncologic outcomes with this approach compare favorably to that achieved with open RP approaches. Robotic approaches are currently being used in some centers. The Da Vinci surgical system (Intuitive Surgical, Inc., Sunnyvale, CA) uses three multijoint robotic arms with one arm controlling the binocular endoscope and the other arms controlling small-wristed instruments. This system is controlled by the surgeon, who can be in a remote location from the patient, seated at an operative console. The stereoscopic view of the operative field provides the surgeon three-dimensional visualization with a 10-fold magnification. Fine and precise movements of the instruments can be achieved, and physiologic tremor can be eliminated. Whether such approaches would lead to significant improvements over surgical outcomes achieved with standard techniques is uncertain and will require prospective randomized studies.
Cryosurgery
Cryosurgery, in which tissues are coagulated by exposing them to very low temperatures with probes implanted through the perineum in the gland under ultrasonographic guidance, has been reintroduced in the treatment of prostate cancer after significant improvements in equipment design. In general, the procedure lasts on average for 2 to 3 hours and patients are discharged with a suprapubic tube or urethral catheter that remains in place for approximately 2 weeks.
Several reports have noted encouraging preliminary PSA relapse–free survival outcomes and posttreatment biopsy results after cryotherapy when used as the primary treatment for clinically localized prostate cancer (26,119,136). Yet there have also been reports documenting increased toxicity with cryotherapy, especially when used as a salvage intervention after a prior course of radiation therapy (194). Technical considerations and recent improvements in cryosurgical techniques have reduced the complications associated with the procedure. These improvements have included the use of multiple freeze–thaw cycles, urethral warming during the procedure, and the placement of thermocouples to maintain the temperatures to <40°C. More recently third-generation cryosurgery techniques have been used that use 17-gauge cryoprobes inserted via transrectal ultrasound guidance through a brachytherapylike template. Multiple mini-ice balls are created that coalesce and potentially provide more precision than standard cryotherapy techniques. Argon gas is used for freezing and helium gas for thawing. Early reports incorporating these enhancements have been promising, with associated decreased toxicity observed (168).
Radiation Therapy Techniques
Conventional External-Beam Radiation Therapy Techniques
Much of the current long-term radiation outcome data for the treatment of clinically localized prostate cancer are derived from patients treated in the 1970s. At that time, the treatment field size and portal configuration for radiation therapy were based on estimations of the anatomic boundaries of the prostate defined by plain-film radiography and by DRE. These techniques were suboptimal compared with the current ability to define the shape and location of the prostate with CT-assisted simulation and MRI. Furthermore, early conventional radiation therapy delivery methods limited the treatment volume to relatively small 6 × 6-cm to 8 × 8-cm fields using rotational arcs or a four-field box technique. Although usually sufficient for the treatment of small T1 and T2a tumors, reconstruction of such fields using CT imaging clearly demonstrated that even an 8 × 8-cm field size would be insufficient to encompass most prostates with locally advanced disease, especially when the seminal vesicles were at risk (401).
A variety of treatment techniques were used in the past, ranging from parallel-opposed anteroposterior portals with a perineal appositional field to lateral portals (box technique) or rotational fields to supplement the dose to the prostate (25,107,257). In general, four fields were used to treat the pelvis and prostate to an initial dose of 45 Gy, with a boost to 70 Gy or higher to the prostate only. For patients with node-positive disease, the initial fields were designed to cover the common iliac lymph nodes. The inferior margin of the pelvic fields was set 1.5 to 2 cm distal to the junction of the prostatic and membranous urethra (usually at or caudad to the bottom of the ischial tuberosities). The lateral margins were typically placed approximately 1 to 2 cm from the lateral bony pelvis (Fig. 62.9). The anterior margin of the lateral fields was placed 1.5 cm posterior to the projection of the anterior cortex of the pubic symphysis while posteriorly, the portals were designed to include the pelvic and presacral lymph nodes above the S3 segment, sparing the posterior rectal wall distal to this level. Some small bowel could usually be spared anteriorly, keeping in mind the anatomic location of the external iliac lymph nodes (Fig. 62.10).
Conformal and Intensity-Modulated Radiation Therapy Techniques
In the early to mid-1980s, three-dimensional (also known as conformal) treatment techniques became increasingly available. Although these techniques vary in some aspects, they
P.1458
share certain common principles that offer significant advantages over conventional treatment techniques. CT-based images referenced to a reproducible patient position are used to localize the prostate and normal organs and to generate high-resolution 3D reconstructions of the patient. Treatment field directions are selected using beam's-eye-view techniques and the fields are shaped to conform to the patient's CT-defined target volume, thereby minimizing the volume of normal tissue irradiated. Conformal radiotherapy simulation, planning, and treatment incorporate various additional maneuvers to reduce treatment uncertainties and enhance setup reproducibility required during a protracted course of therapy.
View Figure
FIGURE 62.9. Diagrams of the pelvis showing volumes used to irradiate the prostate and pelvic lymph nodes. Lower margin is at or 1 cm below ischial tuberosities. At the Mallinckrodt Institute of Radiology, 15 × 15 cm portals at source-skin distance are used to stage A2 and B disease and for selected postoperative patients, whereas for stage C or D1 disease, 18–15 cm portals are used to cover all common iliac lymph nodes up to the bifurcation of the common iliac lymph nodes. Sizes of reduced fields are larger (up to 12 × 14 cm) when seminal vesicles or periprostatic tumors are irradiated compared with prostate boost only (up to 10 × 11 cm).
IMRT is a relatively recent refinement of three-dimensional conformal techniques that uses treatment fields with highly irregular radiation intensity patterns to deliver exquisitely conformal radiation distributions. These intensity patterns are created using special “inverse” or “optimization” computer planning systems, the characteristics of which have been thoroughly summarized in several review articles (45,421). Rather than define each field shape and weight as is done in conventional treatment planning, planners of IMRT treatment specify the desired dose to the target and normal tissues using mathematical descriptions referred to as “constraints” or “objectives.” Sophisticated optimization methods are then used to determine the intensity pattern for each treatment field that results in a dose distribution as close to the user-defined constraints as possible.
View Figure
FIGURE 62.10. Lateral portal used in box technique to irradiate pelvic tissues and prostate. The anterior margin is 0.5 to 1 cm posterior to projected cortex of pubic symphysis. Presacral lymph nodes in included down to S3; inferiorly, the posterior wall of rectum is spared.
IMRT delivery is significantly more complex than conformal delivery as well. Delivery of an IMRT intensity pattern requires a computer-controlled beam-shaping apparatus on the linear accelerator known as a multileaf collimator (MLC). The MLC consists of many small individually moving leaves or fingers that can create arbitrary beam shapes. The MLC is used for IMRT delivery in either a static mode referred to as “step and shoot,” which consists of multiple small, irregularly shaped fields delivered in sequence, or a dynamic mode (dynamic multileaf collimation) with the leaves moving during treatment to create the required irregular intensity patterns (379). Since its inception, IMRT has become a common and important method for treating prostate cancer and, through its ability to tightly conform the radiation to the shape of the target, has facilitated an escalation in dose at several institutions, including MSKCC.
In the following section, the 3DCRT and IMRT techniques used at MSKCC are highlighted with reference to similar techniques developed by others.
Immobilization, Simulation, and Computed Tomography Scanning
On the evening prior to simulation, patients undergo a standard bowel preparation. Immediately before the simulation procedure and CT scan the next day, the patient is asked to void his bladder. To visualize bowel in the vicinity of the prostate and seminal vesicles, a barium sulfate suspension is administered and the rectal lumen is visualized by inserting a rectal catheter.
The majority of the patients with prostate cancer at MSKCC are treated in a prone position, with the supine position reserved for those who are obese or have difficulty lying prone because of arthritis or other orthopaedic problems. Zelefsky et al. (454) compared treatment plans in both supine and prone positions and found the prone position was more suitable for the majority of the patients undergoing 3DCRT or IMRT treatment. For patients treated to doses ≥75.6 Gy, lower rectal wall doses and displacement of the bowel out of the treatment field were more often observed in the prone position. The prone position was also found to be technically reproducible and well tolerated by most patients. Others, however, have not observed the same superiority of the prone position (32,210,245,259,393). Stroom et al. (393) compared prostate and seminal vesicle movement in the supine and prone positions by evaluating serial CT scans in 15 supine and 15 prone patients. Although the overall variability in target position was slightly less for patients treated in the prone position, the systematic component of the organ motion was larger when patients were prone. Because the margin needed to compensate for systematic error is greater than that for random error, the authors concluded that the margins needed to account for organ motion were similar for the two positions. Malone et al. (245) observed more respiratory-induced prostate motion when patients were in the prone position. Fluoroscopic evaluation of the motion of implanted gold seeds was performed for patients in the supine and prone positions, with and without thermoplastic immobilization devices. Mean superior-inferior and anterior-posterior displacements of 2.9 ± 1.7 mm and 1.6 ± 1.1 mm, respectively, were observed for patients in the prone position with thermoplastic shells in place. Significant motion of ≥4 mm was observed in 23% of the patients. The authors suggested that this motion should be considered when designing treatment plans for patients in this position.
P.1459
For immobilization, a thermoplastic mold is fabricated for simulation, CT scanning, and treatment to ensure that the patient is in the same treatment position during all procedures. The thermoplastic sheet is heated in warm water and molded to the patient's shape from the knees to midabdomen. Small sections of the mold are cut away to provide ports for marking and tattooing. The patient is scanned through an approximately 20 to 30 cm region around the prostate with a slice spacing and thickness of 3 mm. Before starting the CT planning study, several transverse images through the prostate and bladder are obtained to ensure that the rectal lumen is clearly visible, the bladder and rectum are not excessively filled, and the patient is properly positioned within the scan circle. Using the CT dataset, a “virtual simulation” is performed, using digitally reconstructed radiographs to localize the treatment area rather than conventional simulation films. The treatment isocenter is placed according to anatomic landmarks near the center of the prostate gland: midline, at the caudad aspect, and approximately 5 cm posterior to the symphysis pubis. The triangulation points for the isocenter are then tattooed, along with an additional alignment tattoo, along the sagittal line, approximately 10 cm superior to the isocenter. To ensure reproducible leg position, tattoos are placed on the back of the legs at the midshaft level, and the distance between the tattoos is recorded for future reference.
Target and Normal Tissue Contouring
The clinical target volume (CTV) is defined as the prostate and seminal vesicles. The planning target volume (PTV) is defined as the CTV with a margin to account for physical uncertainties including setup reproducibility, inter- and intrafractional organ motion. At MSKCC, a 1-cm margin is added to the CTV to form the PTV in all directions except posteriorly at the interface with the rectum, where the margin is reduced to 0.6 cm. Clinically, these margins were found to provide adequate target coverage based on a serial CT scan study evaluating organ motion during a course of 3DCRT (448). Normal tissues identified on each CT slice include the inner and outer walls of the rectum and bladder, the femoral heads, and the outer skin surface. Portions of the small bowel or sigmoid colon within 1 cm of the PTV are also contoured and taken into consideration, if necessary, during planning. In addition, the central 1-cm diameter portion of the prostate encompassing the prostatic urethra is defined for dosimetric consideration and evaluation during high-dose IMRT planning.
Accurate anatomic delineation of the prostate and, in particular, the prostatic apex, has been a topic of some controversy. Urethrography at the time of simulation as a method to accurately localize the apex has been advocated by some and extensively studied (7,88,280,351,353,394,425). Algan et al. (7) reviewed the location of the prostatic apex in 17 patients for whom MRI scan, retrograde urethrogram, and CT of the pelvis were obtained for 3D treatment planning. The location of the prostatic apex as determined by the urethrogram alone was, on average, 5.8 mm caudad to the location on the MRI, whereas the location of the prostatic apex as determined by CT/urethrogram was 3.1 mm caudad to that on MRI. If the prostatic apex is defined as 12 mm instead of 10 mm above the urethrogram tip (junction of membranous and prostatic urethra), the difference between the urethrogram and MRI locations of the prostatic apex is removed. Milosevic et al. (280) also found differences in the position of the prostatic apex between urethrogram, CT, and MR. In an evaluation of 20 patients, the authors found relatively poor correlation between MR and CT or MR and urethrogram in determining the height of the apex above the tuberosities. In response to concerns that the position of the prostate could be altered by the urethrogram itself, Mah et al. (243) performed sagittal MR scans immediately before and after urethrogram in 13 patients. No significant systematic motion of the prostate itself or the apex was observed, leading the authors to conclude that urethrography during simulation does not introduce localization error.
The contribution of MRI to improved accuracy and reproducibility of target localization in prostate cancer has also been well studied (106,333,341,413). Roach et al. (341) studied 10 patients with both MR and CT images of the prostate and noted that the prostate volume was 32% larger when defined by noncontrast CT than when determined by MRI. Areas of disagreement tended to occur in the posterior and posteroinferior-apical portions of the prostate, the apex (because of disagreement between urethrography and MRI), and the regions corresponding to the neurovascular bundle. Rasch et al. (333) also observed differences in CT- and MR-defined volumes. On average, the prostate and seminal vesicle volume defined on CT was 40% larger than that defined on MR. The CT-defined prostate was 8 mm larger at the base of the seminal vesicles and 6 mm larger at the prostatic apex. This difference was found to be significantly larger than interobserver variation.
The optimal management of clinically localized prostate cancer remains controversial and is often a source of great frustration and anxiety for many patients who are compelled to make a decision regarding a treatment intervention for their disease. The practitioner must be aware that the natural history of this tumor is variable and influenced by multiple prognostic factors. All of the various forms of therapy for prostate cancer can affect quality of life and sexual function in varying degrees. In the process of counseling and discussing therapeutic options, it is important to present all available data regarding the variable natural history of this disease, prognostic significance of the diagnosis, potential therapeutic benefit of the various modalities, and immediate as well as late treatment-related sequelae. Life expectancy and quality of life considerations should be carefully discussed with the patient and spouse or significant other.
Based on the available data, when comparing patients with similar prognostic features, there are no significant differences in the biochemical and disease-free survival outcomes for patients with early stages of disease treated with RP, high-dose external-beam radiation therapy (EBRT), or interstitial implantation (97,220,457). In the absence of randomized trials demonstrating superiority of one treatment over another, there have been wide geographic variations in the preferred therapeutic intervention for early-stage prostate cancer currently practiced throughout the United States.
Observations from the CapSURE database (Cancer of the Prostate Strategic Urologic Research Endeavor), a registry of 10,000 men accrued from community-based urologic practices across the United States, has shed further light on practice patterns. Cooperberg et al. (86) reported an increasing trend for patients presenting at diagnosis with more favorable risk disease in 2001–2002 (47%) compared to 1989–1990 (31%). A lower percentage of high-risk patients (defined as PSA >20 ng/mL, Gleason 8-10 disease or T3-T4 disease) were noted on initial presentation (15% compared with 41%) for 2000–2001 and 1989–1990, respectively. Practice patterns have also changed in more recent years. The use of expectant management has decreased from 20% in 1993 to 8% in 2001. A slight decrease in the use of prostatectomy and EBRT was reported with a significant increase in brachytherapy and primary androgen-deprivation therapy treatment interventions. Finally, there has been a steady increase in the use of neoadjuvant androgen-deprivation therapy in conjunction with planned radiotherapy. Comparing trends from 1989–1990 and 1999–2001, the use of neoadjuvant androgen deprivation has increased from 10% to 75% for those receiving EBRT, and 7% to 25% for those receiving brachytherapy.
Zelefsky et al. (444) reported significant changes of radiation therapy practice across the United States based on comparisons of the 1994 and 1999 surveys of the American College of Radiology Patterns of Care Survey. Overall, it was observed that, compared with the 1994 survey, there were significant changes in the practice and delivery of radiation therapy during the more recent survey period. Specifically, more patients, especially with favorable-risk disease, were treated with implantation compared with prior years. The implant-treated population was noted to be younger than the patients receiving EBRT. In the 1999 survey, higher doses of EBRT were more frequently delivered using 3DCRT techniques compared with what was reported in the 1994 survey. It was also noted that there was a substantial increase in recent years in the percentage of patients treated with androgen deprivation in conjunction with radiation therapy.
Several authors, including Adolfsson et al. (2) and Johansson et al. (200), have reported on patients, age 60 to more than 80 years, who, on histologic diagnosis of carcinoma of the prostate, were managed conservatively and observed without specific anticancer treatment until symptoms developed. Recently, a phase III randomized trial from the Scandinavian Prostate Cancer Group demonstrated the benefits of treatment intervention compared with an expectant management approach (39). In this study, 695 men were included with early stage T1 or T2 prostate cancer. The median follow-up was 8.2 years, the primary end point was death attributed to prostate cancer, and the secondary end point was overall survival. In the watchful waiting group, 8.9% died of disease compared with 4.6% in the RP group (p = .02). Men assigned to RP had a significantly lower incidence of distant metastases compared with those who underwent watchful waiting. However, the overall survival rates between the two groups were noted to be same.
In today's health environment in the United States, it is often not considered acceptable to delay definitive therapy for most patients with localized carcinoma of the prostate, except in selected elderly patients with low Gleason scores and low-volume disease based on the biopsy findings, as well as patients with significant medical comorbidities. Properly designed prospective clinical trials are critically needed to better define the efficacy and cost-effectiveness of various therapeutic approaches for localized carcinoma of the prostate.
Traditionally, the treatment options for patients with early-stage, clinically localized prostatic cancer (stages T1c or T2) have included RP, EBRT, or permanent interstitial implantation. Surgical techniques have significantly improved with the advent of the nerve-sparing operation popularized by Walsh (416), with a lower incidence of sexual impotence (approximately 30% to 60%, depending on the patient's age, tumor stage, and surgery extent) compared with classic RP (almost 100%), as well as improved methods available to reduce risks of posttreatment urinary incontinence. At the same time, the accuracy and safety of EBRT delivery have significantly improved with the emergence of 3DCRT and IMRT approaches. Permanent interstitial implantation using ultrasound-guided transperineal techniques and the recent developments of intraoperative conformal optimization for prostate brachytherapy have consistently improved the dose distributions for this treatment approach, leading to improved outcomes and decreased toxicities. For patients with intermediate-risk and selected unfavorable-risk features, in
P.1457
addition to high-dose EBRT alone, the combination of EBRT with permanent interstitial implantation or a high–dose-rate (HDR) brachytherapy boost represents appropriate treatment interventions in addition to RP.
Treatment Techniques
Radical Prostatectomy
RP, initially described by Young (436) in 1905 and popularized by Jewett (199), is a therapeutic option when the tumor is confined to the prostate; according to most urologists, it has no role in the management of gross extracapsular disease, seminal vesicle involvement, or in the presence of lymph node metastases. Two approaches for the classic RP are used: retropubic and perineal. The procedure consists of complete removal of the prostate and its surrounding capsule together with the seminal vesicles, the ampulla, and the vas deferens. The prostate is removed completely by excision of the urethra at the prostatomembranous junction, leaving no residual prostatic tissue at the apex. The retropubic approach is preferred by many urologic surgeons; this procedure also facilitates access for performing a bilateral pelvic lymph node dissection.
Walsh and Donker (417) described the anatomic basis for sexual impotence after RP and, based on that, reported on techniques to achieve a nerve-sparing approach. Through detailed anatomic studies, they demonstrated that the branches of the pelvic autonomic plexus that innervate the corpora cavernosa are located between the rectum and the urethra, along the lateral aspect of the prostate, and penetrate the urogenital diaphragm near or in the midmuscular wall of the urethra. Later they described the technique for radical retropubic prostatectomy with preservation of the neurovascular pedicle (416). They reported preservation of sexual function in 73% of 250 patients treated with this operation; the incidence of sexual impotence was correlated with the age of the patient at the time of surgery (418).
In the last 5 years there has been an increasing interest and practice of laparoscopic radical prostatectomy both in the United States and Europe. Advantages cited for the use of the laparoscopic procedure versus the open approach include improved visualization of the anatomy optical magnification, less blood loss, less postoperative pain, and more rapid resumption of normal activities (181) Preliminary functional and oncologic outcomes with this approach compare favorably to that achieved with open RP approaches. Robotic approaches are currently being used in some centers. The Da Vinci surgical system (Intuitive Surgical, Inc., Sunnyvale, CA) uses three multijoint robotic arms with one arm controlling the binocular endoscope and the other arms controlling small-wristed instruments. This system is controlled by the surgeon, who can be in a remote location from the patient, seated at an operative console. The stereoscopic view of the operative field provides the surgeon three-dimensional visualization with a 10-fold magnification. Fine and precise movements of the instruments can be achieved, and physiologic tremor can be eliminated. Whether such approaches would lead to significant improvements over surgical outcomes achieved with standard techniques is uncertain and will require prospective randomized studies.
Cryosurgery
Cryosurgery, in which tissues are coagulated by exposing them to very low temperatures with probes implanted through the perineum in the gland under ultrasonographic guidance, has been reintroduced in the treatment of prostate cancer after significant improvements in equipment design. In general, the procedure lasts on average for 2 to 3 hours and patients are discharged with a suprapubic tube or urethral catheter that remains in place for approximately 2 weeks.
Several reports have noted encouraging preliminary PSA relapse–free survival outcomes and posttreatment biopsy results after cryotherapy when used as the primary treatment for clinically localized prostate cancer (26,119,136). Yet there have also been reports documenting increased toxicity with cryotherapy, especially when used as a salvage intervention after a prior course of radiation therapy (194). Technical considerations and recent improvements in cryosurgical techniques have reduced the complications associated with the procedure. These improvements have included the use of multiple freeze–thaw cycles, urethral warming during the procedure, and the placement of thermocouples to maintain the temperatures to <40°C. More recently third-generation cryosurgery techniques have been used that use 17-gauge cryoprobes inserted via transrectal ultrasound guidance through a brachytherapylike template. Multiple mini-ice balls are created that coalesce and potentially provide more precision than standard cryotherapy techniques. Argon gas is used for freezing and helium gas for thawing. Early reports incorporating these enhancements have been promising, with associated decreased toxicity observed (168).
Radiation Therapy Techniques
Conventional External-Beam Radiation Therapy Techniques
Much of the current long-term radiation outcome data for the treatment of clinically localized prostate cancer are derived from patients treated in the 1970s. At that time, the treatment field size and portal configuration for radiation therapy were based on estimations of the anatomic boundaries of the prostate defined by plain-film radiography and by DRE. These techniques were suboptimal compared with the current ability to define the shape and location of the prostate with CT-assisted simulation and MRI. Furthermore, early conventional radiation therapy delivery methods limited the treatment volume to relatively small 6 × 6-cm to 8 × 8-cm fields using rotational arcs or a four-field box technique. Although usually sufficient for the treatment of small T1 and T2a tumors, reconstruction of such fields using CT imaging clearly demonstrated that even an 8 × 8-cm field size would be insufficient to encompass most prostates with locally advanced disease, especially when the seminal vesicles were at risk (401).
A variety of treatment techniques were used in the past, ranging from parallel-opposed anteroposterior portals with a perineal appositional field to lateral portals (box technique) or rotational fields to supplement the dose to the prostate (25,107,257). In general, four fields were used to treat the pelvis and prostate to an initial dose of 45 Gy, with a boost to 70 Gy or higher to the prostate only. For patients with node-positive disease, the initial fields were designed to cover the common iliac lymph nodes. The inferior margin of the pelvic fields was set 1.5 to 2 cm distal to the junction of the prostatic and membranous urethra (usually at or caudad to the bottom of the ischial tuberosities). The lateral margins were typically placed approximately 1 to 2 cm from the lateral bony pelvis (Fig. 62.9). The anterior margin of the lateral fields was placed 1.5 cm posterior to the projection of the anterior cortex of the pubic symphysis while posteriorly, the portals were designed to include the pelvic and presacral lymph nodes above the S3 segment, sparing the posterior rectal wall distal to this level. Some small bowel could usually be spared anteriorly, keeping in mind the anatomic location of the external iliac lymph nodes (Fig. 62.10).
Conformal and Intensity-Modulated Radiation Therapy Techniques
In the early to mid-1980s, three-dimensional (also known as conformal) treatment techniques became increasingly available. Although these techniques vary in some aspects, they
P.1458
share certain common principles that offer significant advantages over conventional treatment techniques. CT-based images referenced to a reproducible patient position are used to localize the prostate and normal organs and to generate high-resolution 3D reconstructions of the patient. Treatment field directions are selected using beam's-eye-view techniques and the fields are shaped to conform to the patient's CT-defined target volume, thereby minimizing the volume of normal tissue irradiated. Conformal radiotherapy simulation, planning, and treatment incorporate various additional maneuvers to reduce treatment uncertainties and enhance setup reproducibility required during a protracted course of therapy.
View Figure
FIGURE 62.9. Diagrams of the pelvis showing volumes used to irradiate the prostate and pelvic lymph nodes. Lower margin is at or 1 cm below ischial tuberosities. At the Mallinckrodt Institute of Radiology, 15 × 15 cm portals at source-skin distance are used to stage A2 and B disease and for selected postoperative patients, whereas for stage C or D1 disease, 18–15 cm portals are used to cover all common iliac lymph nodes up to the bifurcation of the common iliac lymph nodes. Sizes of reduced fields are larger (up to 12 × 14 cm) when seminal vesicles or periprostatic tumors are irradiated compared with prostate boost only (up to 10 × 11 cm).
IMRT is a relatively recent refinement of three-dimensional conformal techniques that uses treatment fields with highly irregular radiation intensity patterns to deliver exquisitely conformal radiation distributions. These intensity patterns are created using special “inverse” or “optimization” computer planning systems, the characteristics of which have been thoroughly summarized in several review articles (45,421). Rather than define each field shape and weight as is done in conventional treatment planning, planners of IMRT treatment specify the desired dose to the target and normal tissues using mathematical descriptions referred to as “constraints” or “objectives.” Sophisticated optimization methods are then used to determine the intensity pattern for each treatment field that results in a dose distribution as close to the user-defined constraints as possible.
View Figure
FIGURE 62.10. Lateral portal used in box technique to irradiate pelvic tissues and prostate. The anterior margin is 0.5 to 1 cm posterior to projected cortex of pubic symphysis. Presacral lymph nodes in included down to S3; inferiorly, the posterior wall of rectum is spared.
IMRT delivery is significantly more complex than conformal delivery as well. Delivery of an IMRT intensity pattern requires a computer-controlled beam-shaping apparatus on the linear accelerator known as a multileaf collimator (MLC). The MLC consists of many small individually moving leaves or fingers that can create arbitrary beam shapes. The MLC is used for IMRT delivery in either a static mode referred to as “step and shoot,” which consists of multiple small, irregularly shaped fields delivered in sequence, or a dynamic mode (dynamic multileaf collimation) with the leaves moving during treatment to create the required irregular intensity patterns (379). Since its inception, IMRT has become a common and important method for treating prostate cancer and, through its ability to tightly conform the radiation to the shape of the target, has facilitated an escalation in dose at several institutions, including MSKCC.
In the following section, the 3DCRT and IMRT techniques used at MSKCC are highlighted with reference to similar techniques developed by others.
Immobilization, Simulation, and Computed Tomography Scanning
On the evening prior to simulation, patients undergo a standard bowel preparation. Immediately before the simulation procedure and CT scan the next day, the patient is asked to void his bladder. To visualize bowel in the vicinity of the prostate and seminal vesicles, a barium sulfate suspension is administered and the rectal lumen is visualized by inserting a rectal catheter.
The majority of the patients with prostate cancer at MSKCC are treated in a prone position, with the supine position reserved for those who are obese or have difficulty lying prone because of arthritis or other orthopaedic problems. Zelefsky et al. (454) compared treatment plans in both supine and prone positions and found the prone position was more suitable for the majority of the patients undergoing 3DCRT or IMRT treatment. For patients treated to doses ≥75.6 Gy, lower rectal wall doses and displacement of the bowel out of the treatment field were more often observed in the prone position. The prone position was also found to be technically reproducible and well tolerated by most patients. Others, however, have not observed the same superiority of the prone position (32,210,245,259,393). Stroom et al. (393) compared prostate and seminal vesicle movement in the supine and prone positions by evaluating serial CT scans in 15 supine and 15 prone patients. Although the overall variability in target position was slightly less for patients treated in the prone position, the systematic component of the organ motion was larger when patients were prone. Because the margin needed to compensate for systematic error is greater than that for random error, the authors concluded that the margins needed to account for organ motion were similar for the two positions. Malone et al. (245) observed more respiratory-induced prostate motion when patients were in the prone position. Fluoroscopic evaluation of the motion of implanted gold seeds was performed for patients in the supine and prone positions, with and without thermoplastic immobilization devices. Mean superior-inferior and anterior-posterior displacements of 2.9 ± 1.7 mm and 1.6 ± 1.1 mm, respectively, were observed for patients in the prone position with thermoplastic shells in place. Significant motion of ≥4 mm was observed in 23% of the patients. The authors suggested that this motion should be considered when designing treatment plans for patients in this position.
P.1459
For immobilization, a thermoplastic mold is fabricated for simulation, CT scanning, and treatment to ensure that the patient is in the same treatment position during all procedures. The thermoplastic sheet is heated in warm water and molded to the patient's shape from the knees to midabdomen. Small sections of the mold are cut away to provide ports for marking and tattooing. The patient is scanned through an approximately 20 to 30 cm region around the prostate with a slice spacing and thickness of 3 mm. Before starting the CT planning study, several transverse images through the prostate and bladder are obtained to ensure that the rectal lumen is clearly visible, the bladder and rectum are not excessively filled, and the patient is properly positioned within the scan circle. Using the CT dataset, a “virtual simulation” is performed, using digitally reconstructed radiographs to localize the treatment area rather than conventional simulation films. The treatment isocenter is placed according to anatomic landmarks near the center of the prostate gland: midline, at the caudad aspect, and approximately 5 cm posterior to the symphysis pubis. The triangulation points for the isocenter are then tattooed, along with an additional alignment tattoo, along the sagittal line, approximately 10 cm superior to the isocenter. To ensure reproducible leg position, tattoos are placed on the back of the legs at the midshaft level, and the distance between the tattoos is recorded for future reference.
Target and Normal Tissue Contouring
The clinical target volume (CTV) is defined as the prostate and seminal vesicles. The planning target volume (PTV) is defined as the CTV with a margin to account for physical uncertainties including setup reproducibility, inter- and intrafractional organ motion. At MSKCC, a 1-cm margin is added to the CTV to form the PTV in all directions except posteriorly at the interface with the rectum, where the margin is reduced to 0.6 cm. Clinically, these margins were found to provide adequate target coverage based on a serial CT scan study evaluating organ motion during a course of 3DCRT (448). Normal tissues identified on each CT slice include the inner and outer walls of the rectum and bladder, the femoral heads, and the outer skin surface. Portions of the small bowel or sigmoid colon within 1 cm of the PTV are also contoured and taken into consideration, if necessary, during planning. In addition, the central 1-cm diameter portion of the prostate encompassing the prostatic urethra is defined for dosimetric consideration and evaluation during high-dose IMRT planning.
Accurate anatomic delineation of the prostate and, in particular, the prostatic apex, has been a topic of some controversy. Urethrography at the time of simulation as a method to accurately localize the apex has been advocated by some and extensively studied (7,88,280,351,353,394,425). Algan et al. (7) reviewed the location of the prostatic apex in 17 patients for whom MRI scan, retrograde urethrogram, and CT of the pelvis were obtained for 3D treatment planning. The location of the prostatic apex as determined by the urethrogram alone was, on average, 5.8 mm caudad to the location on the MRI, whereas the location of the prostatic apex as determined by CT/urethrogram was 3.1 mm caudad to that on MRI. If the prostatic apex is defined as 12 mm instead of 10 mm above the urethrogram tip (junction of membranous and prostatic urethra), the difference between the urethrogram and MRI locations of the prostatic apex is removed. Milosevic et al. (280) also found differences in the position of the prostatic apex between urethrogram, CT, and MR. In an evaluation of 20 patients, the authors found relatively poor correlation between MR and CT or MR and urethrogram in determining the height of the apex above the tuberosities. In response to concerns that the position of the prostate could be altered by the urethrogram itself, Mah et al. (243) performed sagittal MR scans immediately before and after urethrogram in 13 patients. No significant systematic motion of the prostate itself or the apex was observed, leading the authors to conclude that urethrography during simulation does not introduce localization error.
The contribution of MRI to improved accuracy and reproducibility of target localization in prostate cancer has also been well studied (106,333,341,413). Roach et al. (341) studied 10 patients with both MR and CT images of the prostate and noted that the prostate volume was 32% larger when defined by noncontrast CT than when determined by MRI. Areas of disagreement tended to occur in the posterior and posteroinferior-apical portions of the prostate, the apex (because of disagreement between urethrography and MRI), and the regions corresponding to the neurovascular bundle. Rasch et al. (333) also observed differences in CT- and MR-defined volumes. On average, the prostate and seminal vesicle volume defined on CT was 40% larger than that defined on MR. The CT-defined prostate was 8 mm larger at the base of the seminal vesicles and 6 mm larger at the prostatic apex. This difference was found to be significantly larger than interobserver variation.
posted by stefanoxx at 7:56 AM
0 Comments:
Post a Comment
<< Home