SEARCH · PubMed Health

Results for “Prognostic model”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 739 records · Page 41Linked to original sources

PubMed2003

Repp86 expression and outcome in patients with neuroblastoma.

PURPOSE: Given the well-known challenges of neuroblastoma prognosis, we investigated whether the expression of restrictedly expressed proliferation-associated protein of 86 kDa theoretical molecular mass (repp86), a proliferation-associated protein expressed in S, G2, and M phases of the cell cycle, correlates with the clinical outcome in patients with neuroblastoma. PATIENTS AND METHODS: 161 children with different stages of neuroblastoma were studied; the median follow-up time was 72.8 months. The patients were staged according to the International Neuroblastoma Staging System, and histologic grading of the tumors was performed according to the criteria of Hughes and those of the International Neuroblastoma Pathology Classification. The MYCN gene copy number was determined by Southern blot analysis or fluorescence in situ-hybridization, and repp86 expression was assessed immunohistochemically by means of monoclonal antibody Ki-S2 on paraffin sections from archival tumor samples. RESULTS: A repp86 labeling index (RI) of more than 10% positive tumor cells significantly predicted a shortened disease-free interval and an increased tumor mortality (both P <.0001). Moreover, the RI allowed the identification of patients with favorable and adverse prognosis in subsets defined by stage, grade, age, and MYCN status. In a multivariate analysis, the RI emerged as the most important predictor of event-free and disease-specific survival with hazard ratios of 11.7 and 10.5, respectively (both P <.0001). CONCLUSION: It seems that repp86 expression is closely associated with the biologic behavior of neuroblastoma. Assessment of the RI might, therefore, considerably refine prognostic models.

Adolescent↗

PubMed2003

Combination of fludarabine and mitoxantrone in untreated stages III and IV low-grade lymphoma: S9501.

PURPOSE: To determine the efficacy of combination fludarabine and mitoxantrone (FN) in untreated stages III and IV low-grade lymphoma. The major end point was to estimate progression-free survival (PFS) in all eligible patients. PATIENTS AND METHODS: Seventy-eight eligible patients were registered. Chemotherapy courses were administered every 4 weeks with mitoxantrone 10 mg/m2 on day 1 and fludarabine 25 mg/m2 on days 1, 2, and 3 for a total of six to eight cycles. Pneumocystis carinii prophylaxis was required. RESULTS: Seventy-three patients (94%) attained an objective response. Complete remission was demonstrated in 34 patients (44%) and partial remission was demonstrated in 39 patients (50%). With a median follow-up time of 5.5 years, the median PFS was 32 months, with a 4-year PFS rate of 38%. Median survival has not been reached and 88% of all patients are alive at 4 years. The application of the International Prognostic Index and serologic staging showed significant differences in PFS in all risk groups, whereas overall survival was markedly worse for the highest-risk group in either prognostic model. Three prior Southwest Oncology Group trials using a regimen of cyclophosphamide, doxorubicin, vincristine, and prednisone or a combination of prednisone, vincristine, methotrexate, cytarabine, cyclophosphamide, etoposide, nitrogen mustard, vincristine, procarbazine, and prednisone in similar patient populations demonstrated comparable clinical outcome, although the 4-year survival for FN was better. FN was well tolerated, but mild to severe reversible myelosuppression was noted. Other complications were rare. CONCLUSION: FN is an effective, safe chemotherapy combination for patients with advanced-stage, low-grade lymphoma. Clinical outcomes were comparable to prior published data using anthracycline-based regimens.

Adult↗

PubMed2004

Serum cardiac troponins and N-terminal pro-brain natriuretic peptide: a staging system for primary systemic amyloidosis.

PURPOSE: Primary systemic amyloidosis (AL) is a multisystemic disorder resulting from an underlying plasma cell dyscrasia. There is no formal staging system for AL, making comparisons between studies and treatment centers difficult. Our group previously identified elevated serum cardiac troponin T (cTnT) as the most powerful predictor of overall survival. Others have reported that N-terminal pro-brain natriuretic peptide (NT-proBNP) is a valuable prognostic marker. We sought to develop a staging system for patients with AL. PATIENTS AND METHODS: Two hundred forty-two patients with newly diagnosed AL who were seen at the Mayo Clinic between April 1979 and November 2000, and who had echocardiograms and stored serum samples at presentation were eligible for this retrospective review. NT-proBNP measurements were performed on 242 patients in whom cTnT and cardiac troponin I (cTnI) had been previously run. Two prognostic models were designed using threshold values of NT-proBNP and either cTnT or cTnI (NT-proBNP < 332 ng/L, cTnT < 0.035 microg/L, and cTnI < 0.1 microg/L). Depending on whether NT-proBNP and troponin levels were both low, were high for only one level, or were both high, patients were classified as stage I, II, or III, respectively. RESULTS: Using the cTnT+NT-proBNP model 33%, 30%, and 37% of patients were stages I, II, and III, respectively, with median survivals of 26.4, 10.5, and 3.5 months, respectively. The alternate cTnI+NT-proBNP model predicted median survivals of 27.2, 11.1, and 4.1 months, respectively. CONCLUSION: Stratification of AL patients into three stages is possible with two readily available and reproducible tests setting the stage for more consistent and reliable cross comparisons of therapeutic outcomes.

Adult↗

PubMed2004

Prognostic indicators and survival in patients with stage IIIB inflammatory breast carcinoma after dose-intense chemotherapy.

PURPOSE: To improve treatment outcome for patients presenting with inflammatory breast cancer (IBC), we have sequentially developed and tested single and tandem dose-intense chemotherapy regimens (DICT). Tumor- and treatment-related factors were analyzed to generate a prognostic model. PATIENTS AND METHODS: Between May 1989 and April 2002, 120 patients received conventional-dose chemotherapy, surgery, and sequentially developed single- or tandem-cycle DICT. Disease- and treatment-specific features were subjected to univariate and multivariate analysis to correlate with outcome. RESULTS: At a median follow-up of 61 months (range, 21 to 161 months), estimated 5-year relapse-free survival (RFS) and overall survival (OS) were 44% (95% CI, 34% to 53%) and 64% (95% CI, 55% to 73%), respectively. In an age-adjusted multivariate analysis, RFS was better in patients with estrogen receptor (ER)/progesterone receptor (PR)-positive tumors (P =.002), for patients with fewer than four involved axillary nodes before DICT (P =.01), and in patients treated with radiation therapy (P =.001) and tandem DICT (P =.049). OS was improved in patients with ER/PR-positive tumors (P =.002), in those with fewer than four involved axillary nodes before DICT (P =.03), and in patients treated with radiation therapy (P =.002). CONCLUSION: This retrospective analysis suggests that either single or tandem DICT can be administered safely and may benefit selected patients with stage IIIB IBC. Those with receptor-negative IBC and with four or more involved axillary nodes before DICT need improved neoadjuvant and postadjuvant intensification therapy. A prospective randomized trial of single versus tandem DICT would be required to confirm the potential benefit of tandem DICT in the setting of IBC.

Adult↗

PubMed2006

Postoperative nomogram predicting risk of recurrence after radical cystectomy for bladder cancer.

PURPOSE: Radical cystectomy and pelvic lymphadenectomy (PLND) remains the standard treatment for localized and regionally advanced invasive bladder cancers. We have constructed an international bladder cancer database from centers of excellence in the management of bladder cancer consisting of patients treated with radical cystectomy and PLND. The goal of this study was the development of a prognostic outcomes nomogram to predict the 5-year disease recurrence risk after radical cystectomy. PATIENTS AND METHODS: Institutional radical cystectomy databases containing detailed information on bladder cancer patients were obtained from 12 centers of excellence worldwide. Data were collected on more than 9,000 postoperative patients and combined into a relational database formatted with patient characteristics, pathologic details of the pre- and postcystectomy specimens, and recurrence and survival status. Patients with available information for all selected study criteria were included in the formation of the final prognostic nomogram designed to predict 5-year progression-free probability. RESULTS: The final nomogram included information on patient age, sex, time from diagnosis to surgery, pathologic tumor stage and grade, tumor histologic subtype, and regional lymph node status. The predictive accuracy of the constructed international nomogram (concordance index, 0.75) was significantly better than standard American Joint Committee on Cancer TNM (concordance index, 0.68; P < .001) or standard pathologic subgroupings (concordance index, 0.62; P < .001). CONCLUSION: We have developed an international bladder cancer nomogram predicting recurrence risk after radical cystectomy for bladder cancer. The nomogram outperformed prognostic models that use standard pathologic subgroupings and should improve our ability to provide accurate risk assessments to patients after the surgical management of bladder cancer.

Adenocarcinoma↗

PubMed2004

The high risk of stroke immediately after transient ischemic attack: a population-based study.

BACKGROUND: The risk of stroke is elevated in the first 48 hours after TIA. Previous prognostic models suggest that diabetes mellitus, age, and clinical symptomatology predict stroke. The authors evaluated the magnitude of risk of stroke and predictors of stroke after TIA in an entire population over time. METHODS: Administrative data from four different databases were used to define TIA and stroke for the entire province of Alberta for the fiscal year (April 1999-March 2000). The age-adjusted incidence of TIA was estimated using direct standardization to the 1996 Canadian population. The risk of stroke after a diagnosis of TIA in an Alberta emergency room was defined using a Kaplan-Meier survival function. Cox proportional hazards modeling was used to develop adjusted risk estimates. Risk assessment began 24 hours after presentation and therefore the risk of stroke in the first few hours after TIA is not captured by our approach. RESULTS: TIA was reported among 2,285 patients for an emergency room diagnosed, age-adjusted incidence of 68.2 per 100,000 population (95% CI 65.3 to 70.9). The risk of stroke after TIA was 9.5% (95% CI 8.3 to 10.7) at 90 days and 14.5% (95% CI 12.8 to 16.2) at 1 year. The risk of combined stroke, myocardial infarction, or death was 21.8% (95% CI 20.0 to 23.6) at 1 year. Hypertension, diabetes mellitus, and older age predicted stroke at 1 year but not earlier. CONCLUSIONS: Although stroke is common after TIA, the early risk is not predicted by clinical and demographic factors. Validated models to identify which patients require urgent intervention are needed.

Aged↗

PubMed2002

Preoperative chemoradiotherapy for squamous cell carcinoma and adenocarcinoma of the esophagus: a phase II study.

OBJECTIVES: This study evaluated the concurrent treatment of chemoradiation followed by esophagectomy in the management of locoregional esophageal carcinoma. The main end points were to determine the resectability of the tumor and the pathologic tumor response. An accessory aim was to evaluate the survival rate. PATIENTS AND METHODS: Thirty-nine patients were treated as follows: 5-fluoruracil, 1,000 mg/m(2), by 24-h IV infusion for 4 days, and cisplatin, 100 mg/m(2), on day 1. Concurrent radiotherapy was delivered at a total dose of 40 Gy in daily fractions of 2 Gy five times per week. The performance of an esophagectomy was planned 4 weeks after induction treatment and restaging. RESULTS: All patients completed the preoperative treatment. A potentially radical resection was performed in 29 patients, and a complete or partial histologically proven response was observed in 9 patients (23%) and 20 patients (51%), respectively. The 3-year overall survival rate was 40%. The 3-year rates of overall survival and disease-free survival were 88% and 76%, respectively, in patients with complete response (p < 0.0012), and 16% and 17%, respectively, in patients with partial response (p < 0.0013). Age, histology, and response represented the best prognostic model related to survival. CONCLUSIONS: The results of this combined approach appear to be better than those reported with surgery alone. Despite the small number of patients in the series and the inclusion of patients with different histotypes, we concluded that patients with the squamous histotype show a better outcome than those with adenocarcinoma.

Adenocarcinoma↗

PubMed2004

How to understand a complex reality: statistical issues in microarray experiments. Interview by Nadine Lemmens.

Dr Mansmann is Deputy Director of the biostatistic unit of the Institute of Medical Biometry and Informatics at the University of Heidelberg Medical School, Germany. He is head of the working group in quantitative methods in molecular medicine. His research interests are prognostic models for cancer patients and establishing methodological standards for reproducible research in molecular medicine. As a statistician, he is involved in research on the use of Bayesian methodology for complex stochastic systems.

Humans↗

PubMed2005

Recent advances in the molecular diagnosis of diffuse large B-cell lymphoma.

The field of molecular diagnostics is changing and adapting to new information obtained from genetic, genomic and proteomic profiling of diseases. One of the novel technologies that has made significant impact on the molecular diagnosis of lymphoid malignancies is DNA microarray technology. It has allowed the profiling of the most common types of lymphomas, identifying distinct molecular signatures of these diseases as well as novel subtypes that cannot otherwise be identified by conventional methods. In addition, it has also allowed the construction of molecularly defined prognostic models for various types of lymphomas and to better understand the molecular mechanisms that determine the behavior of the tumor. In this review, recent advances in the molecular diagnosis of diffuse large B-cell lymphoma are highlighted, using examples of how gene expression profiling has been used in disease classification and outcome predictions. The future development of this field and its applications in the clinical arena will also be discussed.

Animals↗

PubMed1995

Prediction of progression of the curve in girls who have adolescent idiopathic scoliosis of moderate severity. Logistic regression analysis based on data from The Brace Study of the Scoliosis Research Society.

In a study conducted by the Scoliosis Research Society, 159 girls with a mean age of thirteen years (range, ten to fifteen years) who had adolescent idiopathic scoliosis were followed prospectively until skeletal maturity or until the curve had increased 6 degrees or more. All patients had had an initial curve of 25 to 35 degrees and an apical level between the eighth thoracic and first lumbar vertebrae, inclusive. Of the 159 patients, 120 were observed without treatment and thirty-nine were managed with lateral electrical surface stimulation. The curve progressed at least 6 degrees in eighty patients. There was no apparent difference in the outcome between the patients who were managed with observation only and those who were given electrical stimulation. Logistic regression analysis was performed to determine which of eleven factors were predictive of progression of the scoliotic curve. A Risser sign of 0 or 1, an apical level cephalad to the twelfth thoracic vertebra, and an imbalance of ten millimeters or less were found to be independently prognostic of progression of more than 6 degrees. A prognostic model that included these three factors and chronological age allowed correct classification of the curve as either progressive or non-progressive in 81 per cent of these patients who had a thoracic or thoracolumbar adolescent idiopathic scoliosis. The positive predictive value was 82 per cent, the negative predictive value was 80 per cent, and the sensitivity and specificity were each 81 per cent.

Adolescent↗

PubMed2006

[Individualisation of therapy in acute nonlymphoblastic leukaemia].

Acute nonlymphoblastic leukaemia involves the dynamic and individual coupling of four groups of significant prognostic factors: biological potential of the patient, leukaemic clone, normal haematopoiesis, and therapy. The active, dynamic, and timely prognosis of an unfavourable outcome represents a solid basis for the individual adaptation of antileukaemic and supportive therapy. A part of the ANLL NS 03 programme for the individualised therapy of acute nonlymphoblastic leukaemia in patients no older than 60 years will be described. The programme is based on a network of prognostic models, identifying predictors and preventive measures against an unfavourable outcome. The crucial point of the ANLL NS 03 programme is to determine the optimal timing for the transplantation of allogeneic and autologous haematopoietic stem cells. Indicators of early death include the age of the patient, infection, and hemorrhagic syndrome. According to our models, the predictors of fatal bleeding and fatal infection are hyperleukocytosis, leucopenia, and granulocytopenia, respectively. Resistance to cytostatics can be predicted on day 14 from the onset of therapy using two original cytological-mathematical parameters: the absolute blast count (ABC) forming the intensity dimension, and parameter S forming the selectivity dimension, of the early effects of the first induction treatment. The ABC and S values determine the structure and timing of the second induction treatment. Transplantation of autologous and allogeneic haematopoietic stem cells within the ANLL NS 03 programme is applied selectively during the early stages of the first remission in patients at high risk of an early relapse. Predictors of early relapse are leukocyte counts higher than 30x10(9)/l, remission induction during the second treatment, and the presence of myelodysplasia. In all other patient categories and in patients with cytogenetically favourable forms of acute nonlymphoblastic leukaemia, transplantation is postponed until the second remission of the disease.

Humans↗

PubMed1991

Maternal factors associated with high birth weight.

Maternal characteristics associated with high birth weight were studied in 473 mothers delivered of singleton infants at term with a birth weight of 4500 g or more. The controls were mothers who gave birth to singleton infants at term, with a normal birth weight +/- 1 SD for Swedish newborns. In the multivariate analysis the maximum symphysis-fundus height measurement and gestational duration were strongly significant (p less than 0.001), after correction for other variables, for the probability of being delivered of an infant of high birth weight. Maternal height, weight at beginning of pregnancy, total gestational weight increase and previous live birth of an infant weighing greater than or equal to 4500 g were also important (p less than 0.05) for high birth weight. The maternal characteristics included were evaluated in a prognostic model. With symphysis-fundus height measurement included, the sensitivity increased from 80.3 to 83.3% and specificity from 78.8 to 85.6%, compared with a model where symphysis-fundus measurement was not available.

Adult↗

PubMed1993

C-reactive protein (CRP) levels do not reflect disease status in patients with multiple myeloma.

It has recently been suggested that a combination of C-Reactive Protein (CRP) and beta-2-microglobulin (beta 2M) can be used to devise a simple prognostic model for patients with multiple myeloma. In this study we have measured serum beta 2M, CRP and thymidine kinase (STK) in a series of 215 samples to determine their value as a monitor of disease status. A longitudinal study was also performed with 6 individual patients. CRP levels did not correlate with disease status. The mean of the stable (23.62 mg/L) and the progressive disease 23.64 mg/L) groups were almost identical (t = 0.003; p = NS) with ranges of < 5-150 mg/L and < 5-100 mg/L respectively. There was no correlation between CRP and STK (r = 0.11) or CRP and beta 2M (r = 0.05). In longitudinal studies, CRP did not necessarily reflect changes in disease activity. We conclude that CRP measurements are not valuable as a monitor of disease activity in patients with myeloma.

C-Reactive Protein↗

PubMedDate not supplied

Assessing the clinical impact of a predictive system in severe head injury.

Many computer-assisted diagnostic or prognostic systems have been described in the literature, but as yet such systems have had little impact upon routine clinical practice. There are many possible explanations for the limited success which these systems have enjoyed, but one of the most significant must be the dearth of well-designed and carefully executed trials to assess their performance. Many evaluations only address the scientific question of whether the diagnoses or prognoses are accurate; they stop short of answering the crucial pragmatic question of whether the system benefits patients, i.e. is clinical care improved by the use of such a system? This brief paper describes a predictive system in severe head injury which has been developed by an international group based in the Institute of Neurological Sciences at the Southern General Hospital, Glasgow. The various steps involved were first the establishment of a clinical databank according to a strict protocol and with carefully designed scales to quantify the impairment of consciousness and outcome, the development of prognostic models relating outcome to initial severity, the comparison of the performance of this model with predictions made by clinicians, and finally a multi-centre prospective trial assessing the clinical impact of the system. The results of this trial are not yet available, but a consideration of the design of the trial raises important points which are relevant to many other clinical fields where 'decision aids' need to be evaluated.

Craniocerebral Trauma↗

PubMed1988

DNA cytometry of osteosarcoma.

The relationship between cytochemical features and histomorphology in osteosarcoma, and the clinical significance of DNA content were investigated by microspectrophotometry (MSP) of tissue sections and flow cytophotometry (FCM) of cell suspensions. MSP of tissue sections entails the methodological error of determining the DNA content of sectioned cell nuclei. By analyzing 184 normal mesenchymal cell populations, an upper limit of diploidy (normal DNA content) was deduced. Applying this upper limit for 42 sarcomas, 6 were diploid and 36 hyperploid. Comparative analysis of the same lesions by MSP of imprint preparations and by FCM disclosed complete agreement in ploidy classification (diploid versus hyperploid). Retrospective MSP analysis of bone tumors is often impeded by previous demineralization in acid, which destroys DNA. EDTA as an alternative was found to slightly reduce Feulgen DNA stainability of osteosarcomas, but did not affect tumor ploidy determination. Hence, EDTA offers a means of retaining DNA stainability of bone tumors requiring demineralization. MSP analysis of different histologic areas, and comparative FCM analysis of biopsy and surgical specimens, dislosed that individual osteosarcomas are cytochemically uniform despite morphologic heterogeneity. Hence, a single tumor sample for DNA analysis can be relied upon as representative for the tumor as a whole. In a consecutive series of 83 osteosarcoma patients treated by surgery and adjuvant Interferon, the 7-year survival rate was 0.44. MSP DNA analysis gave no significant prognostic information. Multivariate analysis identified 3 risk factors for tumor related death, i.e., male sex, proximal tumor location, and histologic grade IV. In a prognostication model, the 7-year survival rates, for patients with 0, 1, 2, or 3 risk factors, were 0.80, 0.59, 0.42, and 0.13, respectively. Hence, it is possible to identify subgroups of high grade osteosarcoma patients with different prognosis. In a study of 166 primary bone tumors, the applicability of DNA analysis for differential diagnostic purposes was investigated. The series included high grade osteosarcomas, parosteal osteosarcomas and benign bone tumors, which may be mixed up histologically with osteosarcoma. Out of 166 tumors, 149 (90%) were histologically noncontroversial, whereas 17 (10%) posed diagnostic difficulties. In the diagnostically noncontroversial group, all benign tumors and parosteal osteosarcomas were diploid, whereas 97 of 102 osteosarcomas were hyperploid. Hence, hyperploidy seems to be a characteristic feature of high grade osteosarcoma.(ABSTRACT TRUNCATED AT 400 WORDS)

Bone Neoplasms↗

PubMed1989

Prognosis in high-grade soft tissue sarcomas. The Scandinavian Sarcoma Group experience in a randomized adjuvant chemotherapy trial.

From 1981 to 1986, 240 patients with primary, malignancy grade III or IV soft-tissue sarcoma were entered into a randomized adjuvant chemotherapy multicenter trial, conducted by the Scandinavian Sarcoma Group. After a median follow-up time of 46 (2-97) months, a multivariate analysis of risk factors for metastases was performed in 138 radically operated on patients with tumors of the extremities. Adjuvant single-agent doxorubicin did not improve the metastasis-free survival. Histologic malignancy grade IV, tumor size greater than 10 cm, vascular invasion by tumor, and male sex were identified as risk factors. Patients with no or one risk factor had a 5-year metastasis-free survival of 0.7, with two risk factors 0.5, and with three or four risk factors 0.2. The combination of different risk factors provides a prognostic model for soft tissue sarcomas, which could be a basis for therapeutic trials.

Adolescent↗

PubMed1991

DNA content prognostic in soft tissue sarcoma. 102 patients followed for 1-10 years.

In a prospective study of 102 patients with soft tissue sarcoma, the prognostic importance of DNA content and clinicopathologic features was analyzed. Based on DNA flow cytometry, 37 lesions were diploid (normal DNA content) and 65 were nondiploid (abnormal DNA content). The 5-year metastasis-free survival rate of the whole series was 0.59. The survival rate was 0.77 for patients with diploid tumors and 0.48 for those with nondiploid tumors (P = 0.01). Multivariate analysis identified two independent metastatic risk factors: increasing tumor size and nondiploidy. Unexpectedly, high malignancy grade (III-IV) was not found to be an independent risk factor for metastasis. The risk of metastasis was strongly related to the number of risk factors present. Thus, the 5-year survival for the 76 patients with no or one risk factor was 0.69, as compared with 0.30 for the 26 patients with two risk factors (P less than 0.0001). Our study shows that metastatic disease in soft tissue sarcoma is closely related to nondiploidy. A prognostication model based on DNA content and tumor size was found to discriminate between patients with a good and a poor prognosis after surgical treatment. The model can be used to identify patients who should be excluded from trials with adjuvant chemotherapy.

Adult↗

PubMed2001

Prognostic markers in malignant lymphoma: an analysis of 1,198 patients treated at a single centre.

The prognostic significance of 20 putative markers has been assessed in a consecutive series of 1,198 patients with malignant lymphoma seen by the Sheffield Lymphoma Group over three decades. Univariate analysis disclosed that ten factors for both Hodgkin's disease (HD) and non-Hodgkin's lymphoma (NHL) Grade I, and twelve factors for NHL Grade II had prognostic significance. However, multivariate analysis selected only three (age, serum albumin and lymphocyte count) for HD, one (serum albumin) for NHL Grade I and five (age, stage, erythrocyte sedimentation rate, serum albumin and serum lactate dehydrogenase) for NHL Grade II as independent predictors for survival. Risk adjusted prognostic models were derived for Hodgkin's disease and NHL Grade II. For Hodgkin's disease the presence of 3 risk factors predicted for only 35% long-term survival for this group of patients. For NHL Grade II the group with 3-5 risk factors present had a median survival of less than 2 years compared to a 9-year median survival in patients with 1 risk factor present. Whilst these models are being validated on a larger series of patients and will also be tested prospectively, new markers are needed to facilitate decisions on treatment for individual patients.

Adult↗