Prognostic Nomograms and Algorithms
Prostate cancer nomograms and tables are tools which may be used to asses risk at different levels. They may use preoperative clinical data to estimate the risk of extraprostatic extension or recurrence. Alternatively, they may assess the risk of tumor recurrence based on postoperative pathological variables.
Predictive tables categorize patients based on risk according to preoperative variables such as pretreatment PSA, Gleason score, clinical stage, and biopsy density. They provide a range of risk for patients within that group, and not a specific number of recurrence risk for the individual patient. The most recent version of the Partin Tables, released in 2001, include over 5000 patients who underwent radical prostatectomy at Johns Hopkins by a single surgeon. They provide information regarding the risk of extraprostatic extension, seminal vesicle involvement, or lymph node metastasis. Although its data are biased since they are based on patients with early stage prostate cancer selected by a single surgeon, these were validated at the Mayo Clinic with a cohort of over 2400 patients. Nevertheless, they serve as a useful tool to estimate local advancement of tumors and to guide surgical management. These are available at http://urology.jhu.edu/prostate/partintables.php
Center for Prostate Disease Research Tables. Similar to the Partin Tables, the CPDR Tables (available on PDF format on www.cpdr.com) provide information on pathologic stage based on preoperative clinical parameters and not a tumor progression risk assessment. They are based on a more racially diverse population of patients who underwent surgery at multiple institutions and take into account the density of positive biopsies.
D'Amico Tables. Provide prognostic information on biochemical recurrence for patients treated with radical prostatectomy and radiation therapy. Provides additional information in patients with a high density of positive biopsies when endorectal MRI was used. Its main criticism is the use of a definition of three consecutive rises in PSA for biochemical failure in surgically treated patients. Furthermore, Gleason 7 patients were not subdivided according to their primary Gleason grade (3+4 vs. 4+3) and 2-year PSA recurrence was used as a surrogate for long term outcome.
Other validated classification systems have been developed since the D'Amico. The Cancer of the Prostate Risk Assessment (CAPRA) score is a system being used today.
Nomograms are instruments which predict outcome on individual patients. The data entered into the equation may include preoperative clinical variables or pathological characteristics after treatment (in the case of patients treated with radical prostatectomy). Aside from being accurate, the ideal nomogram should not be culturally biased, pathologist or surgeon dependent, and should apply to the patient in the community and not the well selected patient commonly seen in tertiary care institutions.
Kattan Preoperative Nomogram. This nomogram provides an assessment of risk of recurrence based on clinical variables such as serum PSA, clinical stage, and biopsy Gleason grade. Although some have criticized this nomogram for using a definition of > 0.4 ng/ml for PSA recurrence, this nomogram has been validated with a cohort of over 6700 patients from different institutions and its results were consistently accurate independent of the definition of failure used by the institution. A PDA format of this nomogram is available at http://www.mskcc.org/mskcc/html/10088.cfm
Kattan Postoperative Nomogram. Provides more accurate (area under the curve: 0.89) prognostic information compared with the preoperative nomogram (AUC: 0.74) because of the additional pathologic information obtained from the prostatectomy specimen. Although the data stem from the experience of a single surgeon, the nomogram has been validated on a multi-institutional setting. A PDA format of this nomogram is available at http://www.mskcc.org/mskcc/html/10088.cfm.
Kattan Prognostic Nomogram for Metastatic Prostate Cancer. Provides median survival and 1-year and 2-year survival estimates for patients with castrate androgen levels and progressive metastatic disease. The nomogram is based on patient age, Karnofsky performance status, serum hemoglobin, PSA, LDH, alkaline phosphatase, and albumin. (Smaletz et al, JCO 20(19):3972, 2002).
CaPSURE/CPDR Recurrence Equation. Biostatistical model produced between the Center for Prostate Disease Research and CaPSURE (Cancer of the Prostate Strategic Urologic Research Endeavor), a longitudinal observational study of prostate cancer patients nationwide. The equation offers an assessment of recurrence risk based on pretreatment PSA, highest Gleason sum on prostatectomy specimen, prostatectomy organ confinement status, and ethnicity (Recurrence risk = exp[0.54 x race) + (0.05 x PSA) + (0.23 x postop Gleason) + (0.69 x pathologic stage)]. These data stratified patients into 4 risk groups: very low- (4.7 or less), low- (4.7 to 7.1), high-(7.1 to 16.7), and very high- (greater than 16.7). Patients at risk had 7-year disease-free survival rates of 85.4%, 66.0%, 50.6% and 21.3%, respectively. Refer to Moul et al, J Urol 2001 Oct;166(4):1322-7 for details.