TMCnet News
Creation of 3-Dimensional Prostate Cancer Maps: Methodology and Clinical and Research Implications [Archives of Pathology & Laboratory Medicine](Archives of Pathology & Laboratory Medicine Via Acquire Media NewsEdge) * Context.-The creation of 3-dimensional prostate cancer maps could assist with surgical intervention, radiotherapy treatment planning and for correlative pathology-imaging research. Objectives.-To develop methodology for creating detailed, 3-dimensional, prostate cancer maps (3DPCM) of tumor location, extra prostatic extension sites, and positive margins and to assess the adequacy of current clinical target volumes for postoperative radiotherapy to the prostate using 3DPCM coregistered with preoperative magnetic resonance imaging. Design.-Parallel slices of prostatectomy specimens were created with ProCUT, and 2-dimensional cancer maps were generated as line diagrams after microscopic examination of each slice. The 2-dimensional cancer maps were aligned and stacked to create a 3DPCM, which was coregistered with the preoperative magnetic resonance imaging scan. The map was exported to the radiotherapy planning system and was used to determine the areas at greater risk, which were then compared against the current Radiation Therapy Oncology Group guidelines for contouring postoperative clinical target volumes to assess the adequacy of coverage. Results.-Twenty-eight patients with a mean age of 66 years (range, 52-73) underwent radical prostatectomy and postoperative radiotherapy. Seventeen patients (61%) received adjuvant radiotherapy for pT3 disease and/or positive margins, and the rest underwent salvage radiotherapy. Thirty-nine percent (11 of 28) of the patients had Gleason scores of 8 or 9. The contours based on the Radiation Therapy Oncology Group guidelines for postoperative radiotherapy resulted in inadequate coverage of extraprostatic extensions in 79% (22 of 28) and positive margins in 64% (18 of 28) of the cases. Conclusions.-We have developed a methodology for creation of 3DPCM. Modification of the radiotherapy contours, based on the 3DPCM coregistered with pretreatment magnetic resonance imaging, covers the areas at high risk of recurrence. The 3DPCM could become an important clinical and research tool for urologists, pathologists, radiologists, and oncologists. (Arch Pathol Lab Med. 2014;138:803-808; doi: 10.5858/arpa.2012-0609-OA) Radical prostatectomy (RP) pathology reports use descriptive text to provide details on the gross examination, tumor grade, and disease extent. The extent of tumor, sites of extraprostatic extension (EPE), and positive margins are difficult to visualize as 3-dimensional (3D) based on current pathology reports, which can, therefore, lead to inaccuracies in adjuvant treatment, specifically in postoperative radiation treatment.1 A 3D reconstruction of the pathology specimen, or "3D Prostate Cancer Map," could be beneficial for clinical care, research protocols, and pathology-imaging correlative studies.2-4 In prostate cancer, EPE and positive surgical margins are important prognostic factors.5-9 The College of American Pathologists emphasizes reporting size, location, and focality of those parameters in pathology reports.10 Mapping of prostate sections onto digital photographs is an established practice in select hospitals.11 Detailed mapping of the microscopic extent of tumors, the location of positive margins, and the extent of EPE is not widely practiced. The objectives of this pilot study were to develop detailed 3D prostate cancer maps of tumor location, sites of EPE, and positive margins and to coregister ("fuse")the3D prostate cancer maps with preoperative magnetic resonance imaging (MRI). Use of 3D cancer maps should help to define regions at greatest risk for harboring microscopic disease in patients being considered for postoperative radiation. MATERIALS AND METHODS Patient Cohort This clinical-pathology study was approved by the Research Ethics Board of the Ottawa Hospital (Ottawa, Ontario, Canada). A list of patients with prostate cancer who had undergone RP and adjuvant or salvage radiotherapy at the Ottawa Hospital from May 2007 to January 2012 was compiled. Using that list, the first 28 patients (chosen alphabetically) who had also undergone a preoperative staging pelvic MRI were chosen as the study group. The decision for preoperative MRI staging in the study group was at the discretion of the urologist performing surgery. Patient demographics are shown in the Table. Analysis of RP Specimens Radical prostatectomy specimens were sent to the pathology laboratory in 10% neutral-buffered formalin. Dimensions, weight, and volume of the gross specimens were recorded. The seminal vesicles were removed before determining the volume of the prostate (by water displacement). The seminal vesicles were sectioned axially or in a sagittal plane and analyzed for tumor involvement and margin status. For the purposes of the radiotherapy study, the locations of tumor in the seminal vesicles were not mapped because the rectovesical space and any retained seminal vesicles would be routinely covered as part of the target volume. The specimen surface was inked and parallel-sliced (5 mm thick). Since July 2011, surgical specimens have been processed using a commercially available ProCUT device (Milestone Medical Technologies Inc, Kalamazoo, Michigan) to ensure even parallel spacing of slices (Figure 1). Digital photographs were taken of the slices (Figure 2). For submission of histology sections, the base and apex were coned, and the remainder was cut into quadrants. The prostate was submitted in toto, with demarcation of each block on the hard copy of the digital photograph. In the latter stage of the protocol, the processing of the apex and base was modified to improve their 3D reconstruction. Instead of coning, the apex and base were sliced transversely in an anterior-posterior direction. The 2 lateral prostate slices were then cut perpendicular to allow for margin assessment. The orientation of the specimen was maintained as per departmental protocols; matching with the imaging was not necessary at sectioning but was achieved later in the process by manipulation of the recreated 3D volume in the 3D slicer software (open-source, free software; http://www.slicer.org, accessed March 30, 2013). A trocar was used to make 3 reference marks (Figure 2) in the prostate. The apex and base were sliced off the RP specimen. A trocar was then used to core 3 holes in the prostate before slicing with ProCUT (Figure 2). The reference marks assisted investigators with alignment of the prostate slices (see "Creation of the 3D Prostate Cancer Map" below). To ensure that the assessment of the margins was not compromised, care was taken not to penetrate the "capsule" with the trocar during the procedure. In later iterations of the protocol, the trocar was replaced with an 18-gauge spinal needle with a trocar. The needle was inserted and green marking ink was injected into the needle tract during extraction of the needle to create the reference marks. This had the advantage of not disrupting the capsule even if the base and apex were not sliced before creating the reference marks. Submitted sections were sent to the histology laboratory for processing, embedding, and cutting. Sections were mounted on a glass slide and stained with hematoxylin-eosin. Microscopic examination of the slides was conducted in conjunction with a hard copy of the digital photograph. Extent and geographic location of the tumor were marked on a hard copy of the photograph. Location, size, and distance to the margin of all areas of EPE were recorded. Positive margins were mapped onto the photograph, and the parameters were measured on all involved prostate slices. Creation of 2D Cancer Maps Photoshop software (Adobe Systems Inc, San Jose, California) was used to create maps demonstrating geographic location, extent of the tumor, EPE, and positive margins (Figure 3). For initial surgical specimens before 2011, the base and apex were coned but not sliced transversely. In those cases, the base and apex regions of the cancer map showed the percentage of tumor in each quadrant. In the updated protocol, the tumor extent was contoured on all slices including the transverse apex and base sections, enabling more-detailed cancer mapping. Areas of cancer were hand-drawn on each slice and filled with red. Areas of EPE were drawn onto the digital images and were indicated by blue contours. Positive margins were represented by green contours, and their size and location were approximated. Creation of 3D Prostate Cancer Maps The 2D prostate cancer maps were converted into 3D maps with in-house ImageJ (National Institutes of Health, Bethesda, Maryland) macros, which allowed manual alignment and registration of the slices; 3D image stacks were created to allow visualization in all 3 planes. The prostate slices were manually aligned based on the external contour, urethra, and reference marks. Reference marks helped to minimize the risk of slice misalignment, including translation and rotation errors. The 3D prostate cancer map permits the clinician to visualize in all 3 planes and to verify and evaluate the map in axial, coronal, and sagittal views. Fusing the 3D Prostate Cancer Map to the Preoperative MRI In our cohort, the maps were registered with preoperative MRIs (Figure 4, A and B) and a 3D slicer 3.6 (http://www.slicer.org, accessed March 4, 2011; correlative research results to be reported separately). The map is registered with rigid-body and size adjustments (affine transformations) because the prostate is deformable. The external contour of the prostate and urethra are used to align the map with the MRI. To assist with registration, the imaging and pathology map can be viewed and manipulated in all 3 planes in the 3D slicer (Figure 5, A through D). Use of the 3D Prostate Cancer Map for Postoperative Radiotherapy A label map was created on the preoperative MRI of the tumor, the EPE, and the involved margins, based on the pathology map. The map was exported, in a Digital Imaging and Communications in Medicine (Rosslyn, Virginia) format, into the radiotherapy treatment planning system (XIO/Focal CMS, Elekta, Stockholm, Sweden) and coregistered with the treatment planning CT (computerized tomography) scan. The prostate tumor bed was contoured on axial CT images by the radiation oncologist to generate a clinical target volume (CTV). The CTV was used as a baseline and compared with CTV contours as recommended by the Radiation Therapy Oncology Group (RTOG) guidelines (hereafter called CTV RTOG). Radiotherapy plans were applied to CTV RTOG to assess the incidence and location of a "geographic miss" for sites of EPE and positive margins. The objective of the radiotherapy study12 was to determine whether the contours done per RTOG guidelines would cover the entire prostate bed and areas at high risk of local recurrence (areas of EPE and positive margins) as determined by the CTV based on the 3D map. RESULTS Twenty-eight patients with a mean age of 66 years (range, 52-73 years) underwent RP and radiotherapy. Seventeen patients (61%) received adjuvant radiotherapy for pT3 disease and/or positive margins, and the rest (39%) underwent salvage radiotherapy on prostate-specific antigen relapse. Fifty-four percent (15 of 28) of the patients were Gleason 7, and the rest (13 of 28; 46%) were Gleason 8 or 9. Mean prostate weight was 41 g (range, 19-84 g). In all cases (28 of 28; 100%), the sites of tumor in the prostate (as suggested by areas of T2 hypointensity and dynamic contrast enhancement) on MRI were confirmed pathologically, that is, the MRI had 100% specificity and sensitivity in this cohort. The contours based on CTV RTOG resulted in a geographic miss of EPE in 79% (22 of 28) of the cases and sites of positive margins in 64% (18 of 28) of the cases. The predominant site of the "geographic miss" was along the superior aspect of the prostate bed. In 3 of the 22 cases (14%) of an EPE miss, the geographic miss was secondary to a major ( . 1 cm) shift of the rectum (preoperative versus postoperative imaging). In the remaining cases (19 of 22; 86%), the geographic miss was due to inadequate CTV coverage of the base region of the prostate bed. Modification of the contours based on the 3D cancer maps coregistered with the pretreatment MRI generated a new CTV at the Ottawa Hospital that covers the areas at high risk of recurrence, as published in our previous results.1 COMMENT In prostate cancer, EPE and positive surgical margins are important prognostic factors.5-9 The College of American Pathologists protocols emphasize reporting size (in millimeters), location, and the number of blocks and focality of those parameters in pathology reports. Surgical margin status in RP specimens is a known prognostic parameter for postoperative biochemical recurrence and disease progression.13-17 Progression-free probability for men with surgical margin positivity on RP ranges from 58% to 64%, which contrasts with 81% to 83% for patients whose RP specimens are margin negative.18,19 Furthermore, the location, extent, and presence of EPE at each margin should be reported as millimeters of involvement. At the 2009 International Society of Urological Pathology Consensus Conference on Handling and Staging of RP Specimens,10 the consensus was that the extent of a positive margin should be reported as millimeters of involvement. The creation of cancer maps on hard copies of digital photographs is used in a few pathology laboratories.11 It is also used by some pathologists for other types of cancer, including breast and colorectal cancers. However, mapping the microscopic extent of tumors, the location and extent of positive margins, and the EPE is not a common practice to date. There is little research using detailed pathology information to create 3D cancer maps.2-4,20,21 This study provides such a methodology using commercially available equipment and free software for creation of 3D prostate cancer maps after in-depth evaluation of tumor extent and sites of EPE and positive margins. The study also describes a method for fusing the cancer maps with preoperative MRIs for correlative research. The pathology protocol developed and refined differs from routine practice. The protocol optimizes the accuracy of the 3D cancer map while avoiding compromising the assessment of EPE and margin status. In the current version of the protocol, the apex and base are sliced transversely instead of being coned. The transverse slices are easier to work with in developing the 3D cancer model. To enhance the geometric and spatial accuracy of the maps, it was important to ensure even, parallel spacing of the prostate slices. Traditionally, in our institution, a RP specimen was sliced following the axis of the urethra. That conventional processing technique results in uneven slice thickness (some slices are thinner anteriorly). The use of the commercially available ProCUT results in parallel, 5-mm slices. Detailed 2D cancer maps were created in Photoshop. The maps include the external contour of the prostate and the urethra as well as a detailed contour set of the entire tumor, sites of EPE, and positive margins. Reference marks in combination with the external contour and urethra all aid in alignment of the prostate slices to create a 3D stack in ImageJ. The 3D stack results in a full 3D prostate cancer map that can be visualized in all 3 planes. Clinicians verify and evaluate the map by scrolling through it in each coordinate plane. The overall extra processing time, beyond traditional processing, was about 45 minutes, but with automation of some of the image processing steps (which are currently time consuming), that time could be reduced significantly. Creation of 3D prostate cancer maps should assist urologists in the management of prostate cancer. These maps provide a detailed representation of tumor extent and areas of EPE and positive margins. As such, 3D prostate cancer maps could help to refine operative techniques and to evaluate new surgical approaches, minimizing the risk of positive margins. Furthermore, the maps could be important teaching aids for residents and be included in clinical studies evaluating new surgical techniques. The 3D prostate cancer maps can be fused with diagnostic imaging (eg, MRI) to evaluate the sensitivity and specificity of imaging protocols. In our protocol, the cancer maps were fused in a 3D slicer for pathology-imaging correlative studies (research in progress) and to assist with planning of radiotherapy treatment. Current imaging sequences have modest sensitivity for detecting sites of EPE, and there is a need for better imaging protocols.22 The cancer maps can be used to evaluate new imaging sequences and imaging modalities. In addition, 3D cancer maps could be developed for many types of cancer for clinical care, correlative pathology-imaging studies, and research purposes. The 3D prostate cancer maps should help radiation oncologists define optimal target volumes (CTVs) for postoperative radiotherapy. To define a CTV, information should be employed from all available sources, including imaging, surgical, and pathology reports. Mature clinical trials of adjuvant radiotherapy have indicated reduced risk of relapse for high-risk disease.23-25 Despite the benefits of adjuvant radiotherapy, a significant percentage of patients still relapse with the predominant site of failure being local.23 Potential reasons for local failure include inadequate radiation dose or inadequate CTV definition. There are, to our knowledge, 4 published consensus guidelines that define CTV for postoperative radiotherapy. There are significant variations among the guidelines, and previous research from our institution indicate that the guidelines miss portions of the prostate bed in most cases.1,12 The 3D prostate cancer maps help define an appropriate CTV. At Ottawa Hospital, we have developed a new CTV (termed CTV TOH) that incorporates information from preoperative imaging and the 3D prostate cancer maps. In the postoperative radiotherapy protocol, the 3D cancer map is fused with the preoperative MRI and imported into the radiotherapy treatment planning system for improved definition of the prostate bed. The cancer map allows the oncologists to better define the areas at greatest risk for harboring microscopic disease (sites of EPE and positive margins).1,12 Adjuvant and salvage-radiotherapy outcomes in patients with prostate cancer remain poor.23-25 The maps can be coregistered with MRI and CT treatment-planning images, which are inherently 3-D volumes, for selective radiotherapy dose escalation (ie, cover the entire tumor bed at low dose and deliver greater doses to sites of the greatest tumor burden or greatest risk, guided by the 3D prostate cancer map. All trials to date in this patient population indicate high rates of relapse.23-25 This methodology is investigational but may improve long-term outcomes for this patient population. This would have to be confirmed in a prospective clinical trial documenting safety and improved outcomes for prostate-specific antigen control, disease-free survival, and overall survival. We acknowledge that the small cohort and short clinical follow-up limit the power of this study. CONCLUSIONS In this study, we created detailed 3D prostate cancer maps that define the tumor extent, the location of EPE sites, and the positive margins. The cancer map provides the clinician with supplemental information beyond traditional pathology reports. Our protocol represents a variation on the traditional technique of processing RP specimens; however, the methodology allows for creation of 3-D cancer maps without compromising margin assessment. The detailed maps could become important clinical, educational, and research-oriented tools for pathologists, radiologists, urologists, and oncologists. The maps provide a tool for researchers to evaluate the sensitivity and specificity of new imaging protocols. The maps can be used to evaluate new surgical techniques. In addition, the 3D prostate cancer maps have recently been used at the Ottawa Hospital to optimize CTVs for postoperative radiation therapy. They offer an opportunity for better definition of the tumor bed and of areas at greatest risk of harboring residual, microscopic disease in patients needing adjuvant or salvage radiotherapy. References 1. Croke J, Malone S, Roustan-Delatour N, et al. Postoperative radiotherapy in prostate cancer: the case of the missing target. Int J Radiat Oncol Biol Phys. 2012; 83(4):1160-1168. 2. Crawford ED, Hirano D, Werahera PN, et al. Computer modeling of prostate biopsy: tumor size and location-not clinical significance-determine cancer detection. J Urol. 1998;159(4):1260-1264. 3. Hirano D, Werahera PN, Crawford ED, Lucia MS, DeAntoni EP, Miller GJ. Morphological analysis and classification of latent prostate cancer using a 3-dimensional computer algorithm: analysis of tumor volume, grade, tumor doubling time and life expectancy. J Urol. 1998;159(4):1265-1269. 4. Narayanan R, Werahera PN, Barqawi A, et al. Adaptation of a 3D prostate cancer atlas for transrectal ultrasound guided target-specific biopsy. Phys Med Biol. 2008;53(20):N397-N406. 5. Epstein JI, Amin M, Boccon-Gibod L, et al. Prognostic factors and reporting of prostate carcinoma in radical prostatectomy and pelvic lymphadenectomy specimens. Scand J Urol Nephrol Suppl. 2005;216:34-63. 6. Watson RB, Civantos F, Soloway MS. Positive surgical margins with radical prostatectomy: detailed pathological analysis and prognosis. Urology. 1996.48(1):80-90. 7. Swindle P, Eastham JA, Ohori M, et al. Do margins matter? The prognostic significance of positive surgical margins in radical prostatectomy specimens. J Urol. 2005;174(3):903-907. 8. Hull GW, Rabbani F, Abbas F, Wheeler TM, Kattan MW, Scadino PT. Cancer control with radical prostatectomy alone in 1,000 consecutive patients. J Urol. 2002;167(2 pt1):528-534. 9. Epstein JI, Pound CR, Partin AW, Walsh PC. Disease progression following radical prostatectomy in men with Gleason score 7 tumor. J Urol. 1998;160(1): 97-100. 10. Magi-Galluzzi C, Evans AJ, Delahunt B, et al; ISUP Prostate Cancer Group. International Society of Urological Pathology (ISUP) Consensus Conference on Handling and Staging of Radical Prostatectomy Specimens: working group 3: extraprostatic extension, lymphovascular invasion and locally advanced disease. Mod Pathol, 2011;24(1):26-38. 11. Bostwick D, Cheng L. ed. Urologic Surgical Pathology. 2nd ed. Maryland Heights, MO: CV Mosby Co; 2008:475-478. 12. Malone S, Croke J, Roustan-Delatour N, et al. The case of the missing target: comparison of RTOG CTV to "prostate cancer maps" and pre-operative MRI. Paper presented at: ASTRO-the 53rd Annual Meeting of the American Society for Radiation Oncology; October 2-6, 2011; Miami Beach, FL. Abstract 2355. 13. Weldon VE, Tavel FR, Neuwirth H, Cohen R. Patterns of positive specimen margins and detectable prostate specific antigen after radical perineal prostatectomy. J Urol. 1995;153(5):1565-1569. 14. Wieder JA, Soloway MS. Incidence, etiology, location, prevention and treatment of positive surgical margins after radical prostatectomy for prostate cancer. J Urol. 1998;160(2):299-315. 15. Cheng L, Darson MF, Bergstralh EJ, Slezak J, Myers RP, Bostwick DG. Correlation of margin status and extraprostatic extension with progression of prostate carcinoma Cancer. 1999;86(9):1775-1782. 16. Pettus JA, Weight CJ, Thompson CJ, Middleton RG, Stephenson RA. Biochemical failure in men following radical retropubic prostatectomy: impact of surgical margin status and location. J Urol. 2004;172(1):129-132. 17. Saether T, Sorlien LT, Viset T, Lyderson S, Angelsen A. Are positive surgical margins in radical prostatectomy specimens an independent prognostic marker? Scand J Urol Nephrol. 2008;42(6):514-521. 18. Kattan MW, Wheeler TM, Scardino PT. Postoperative nomogram for disease recurrence after radical prostatectomy for prostate cancer J Clin Oncol. 1999;17(5):1499-507. 19. Kausik SJ, Blute ML, Sebo TJ, et al. Prognostic significance of positive surgical margins in patients with extraprostatic carcinoma after radical prostatectomy. Cancer. 2002;95(6): 1215-1219. 20. Drew B, Jones EC, Reinsberg S, Yung AC, Goldenberg SL, Kozlowski P. Device for sectioning prostatectomy specimens to facilitate comparison between histology and in vivo MRI. J Magn Reson Imaging . 2010;32(4):992-996. 21. Jhavar SG, Fisher C, Jackson A, et al. Processing of radical prostatectomy specimens for correlation of data from histopathological, molecular biological, and radiological studies: a new whole organ technique. J Clin Pathol. 2005;58(5): 504-508. 22. Kirkham PA, Emberton M, Allen C. How good is MRI at detecting and characterizing cancer within the prostate? Eur Urol. 2006;50(6):1163-1174. 23. Thompson IM Jr, Tangen CM, Paradelo J, et al. Adjuvant radiotherapy for pathologically advanced prostate cancer a randomized clinical trial. JAMA. 2006; 296(19):2329-2335. 24. Thompson IM Jr, Tangen CM, Paradelo J, et al. Adjuvant radiotherapy for pathologically advanced prostate cancer a randomized clinical trial. JAMA. 2006; 296(19):2329-2335. 25. Wiegel T, Bottke D, Steiner U, et al. Phase III postoperative adjuvant radiotherapy after radical prostatectomy compared with radical prostatectomy alone in pT3 prostate cancer with postoperative undetectable PSA: ARO 96-02/ AUO AP 09/95. J Clin Oncol. 2009;27(18):2924-2930. Shawn Christopher Malone, MD, FRCPC; Anoop Haridass, MRCP, FRCR; Balasz Nyiri, PhD; Jennifer Croke, MD; Colin Malone; Rodney H. Breau, MD; Christopher Morash, MD, FRCSC; Leonard Avruch, MD, FRCPC; Manijeh Daneshmand, MD; Kyle Malone; Nicolas Roustan Delatour, MD; Ileyaz Ahmed, MD; Eric Belanger, MD Accepted for publication July 24, 2013. From the Divisions of Radiation Oncology (Drs S. C. Malone, Haridass, Croke, and C. Malone, and Mr K. Malone) and Urology (Drs Breau and Morash) and the Departments of Medical Physics (Dr Nyiri), Radiology, (Dr Avruch), and Pathology (Drs Daneshmand, Ahmed, and Belanger), Ottawa Hospital, Ottawa, Ontario, Canada; and the Department of Medical Biology, Montfort Hospital, Ottawa, Ontario, Canada (Dr Delatour). The authors have no relevant financial interest in the products or companies described in this article. Reprints: Shawn Christopher Malone, MD, FRCPC, Division of Radiation Oncology, Ottawa Hospital Cancer Centre, 501 Smyth Rd, Ottawa, ON K1H 8L6, Canada (e-mail: [email protected]). (c) 2014 College of American Pathologists |