This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Glare, P. A. Articles by McMahon, P. Search for Related Content PubMed PubMed Citation Articles by Glare, P. A. Articles by McMahon, P. Social Bookmarking                            What's this?

Diagnostic Accuracy of the Palliative Prognostic Score in Hospitalized Patients With Advanced Cancer

From the Royal Prince Alfred Hospital, Sydney, Australia

Address reprint requests to Paul Glare, MD, Department of Palliative Care, Royal Prince Alfred Hospital, Missenden Rd, Camperdown, NSW 2050, Australia; e-mail: paul{at}email.cs.nsw.gov.au

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... REFERENCES

 
PURPOSE: To evaluate the predictive accuracy of the Palliative Prognostic (PaP) score in patients with advanced cancer under the care of an oncologist.

PATIENTS AND METHODS: The PaP score was calculated in 100 consecutive patients with advanced cancer hospitalized under the care of a medical or radiation oncologist at a university teaching hospital in Australia. The attending oncologist predicted the survival duration for the purpose of the scoring. The positive predictive value of the PaP score was evaluated. Survival analysis was performed to compare the survival of the three prognostic groups.

RESULTS: Assessable survival data were available for 98 patients. The overall median survival was 12 weeks (interquartile range, 7 to 25 weeks). The PaP score divided the heterogeneous patient sample into three isoprognostic groups related to the chance of surviving 1 month, with 64 patients in group A (> 70% chance), 32 patients in group B (30% to 70% chance), and four patients in group C (< 30% chance). The estimated median survival of the three groups was 17 weeks (95% CI, 12 to 26 weeks), 7 weeks (95% CI, 4 to 12 weeks), and less than 1 week (95% CI, < 1 to 3 weeks), respectively. These survival differences were highly significant (log-rank test of trend, ₁² = 25.65; P < .0001). The 1-month survival of the groups was 97%, 59%, and 25%, respectively.

CONCLUSION: When oncologists' survival estimates are used, the PaP score is able to identify accurately three isoprognostic patient groups, irrespective of the cancer type. The PaP score may help reduce the uncertainty of formulating a prognosis in patients with advanced cancer.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... REFERENCES

 
"Doctor, how much time do I have?" is a central question in the care of patients with advanced cancer. It is also one of the most challenging questions with which oncologists are confronted.^1,2 Accurate prognoses are important for many reasons, including: to provide patients and their families with information about what the future is likely to hold so that they can set their goals, priorities, and expectations for care; to help patients develop insight into their dying; to assist clinicians in their decision making; to compare like patients with regard to outcomes; to establish patients' eligibility for care programs, including timely referral to hospice programs; to establish patients' eligibility for clinical trials; to establish policy-making procedures with respect to appropriate resource use and allocation of support services (eg, frequency of contacts if home care is proposed); and to provide a common language for health care professionals involved in end-of-life care.

A recent systematic review of eight studies investigating the accuracy of physicians' survival predictions in more than 1,500 patients with far-advanced cancer has documented the poor prognostic ability of many physicians; their inaccurate forecasts often are overly optimistic.³ Methods for improving the accuracy of prognostic estimates are needed.

The standard prognostic indicators in oncology, such as tumor size, grade, and stage, or molecular biology, seem to be less relevant in patients with far-advanced cancer.⁴ The Palliative Prognostic (PaP) score, developed for use with palliative care patients, combines performance status, symptoms, and hematologic parameters with the physician's estimate of survival to predict the short-term survival probability as high (> 70% chance of surviving 1 month), intermediate (30% to 70% chance of surviving 1 month), or low (< 30% chance of surviving 1 month), irrespective of the primary site of the cancer.⁵ It has been validated in patients referred to palliative care services both at home in Italy and in an acute care hospital in Australia.^6,7 It has not yet been tested in a general oncology population.

The aim of this study was to evaluate the prognostic accuracy of the PaP score as a clinical prediction rule in the hands of oncologists. The PaP score was applied to a consecutive series of patients with advanced cancer hospitalized under the care of medical or radiation oncologists in an Australian university teaching hospital.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... REFERENCES

 
Participants
One hundred consecutive patients, admitted during a 16-week period to the oncology ward of a university teaching hospital in Sydney, Australia, were recruited onto the study. Inclusion criteria were patient age 18 years or older, proven diagnosis of cancer, cancer at an advanced stage (metastatic or locally advanced), cancer considered incurable with standard treatments, and patient able to understand the participant information statement and give written informed consent. Specifically, patients receiving antineoplastic therapy were eligible for inclusion. Patients with hematologic malignancies were excluded. The hospital's Ethics Review Committee approved the study.

Seven oncologists (six medical oncologists and one radiation oncologist) participated in the study.

Instrument
The PaP score was developed in the 1990s as the end result of a series of a prospective, multicenter Italian studies that aimed to identify soft clinical and biologic factors relevant to the prognosis of patients with far-advanced cancer referred to hospice, and then to produce a composite of those factors as a prognostic index.^5,8,9 Previous retrospective, single-center studies indicated that performance status, symptoms, nutritional parameters, and quality-of-life indices were more relevant than standard oncologic prognostic factors.

Univariate analysis of 36 clinical factors evaluated in the preliminary Italian data found that age, performance status, the physician's prediction of the individual's absolute survival duration (measured in weeks), several treatment characteristics (hospitalization, progestin use, corticosteroid use, blood transfusions), and numerous symptoms (pain, anorexia, dry mouth, dysphagia, weight loss, breathlessness) were associated with survival. Multiple regression analysis revealed that of these, only performance status, the physician's prediction, hospitalization, corticosteroid use, anorexia, and dyspnea were independent predictors of survival.⁸ Similarly, univariate analysis of 19 simple laboratory parameters found that six (high total WBC count, high neutrophil percentage count, high neutrophil percentage, low serum albumin, low pseudocholinesterase, and high proteinuria) were associated with survival.⁹

The results from these two multivariate analyses were then inserted into a multiple regression model to produce the PaP score.⁵ The prognostic factors were handled as categoric variables and a parametric exponential model was fitted. Stepwise backward elimination was used in the multiple regression. The final model contained just four criteria: two symptoms (anorexia and dyspnea), performance status (measured by the Karnofsky performance score [KPS]), WBC abnormalities (elevated total WBC count and lymphopenia), and the physician's survival prediction, measured in weeks.

The regression coefficients for the variables retained in the final model were converted into integers (± 0.5) to obtain the following easy-to-handle partial scores for each prognostic factor: (1) symptoms—anorexia, 1.5 points; dyspnea, 1 point (symptoms absent, 0 points); (2) KPS score¹⁰—10 to 20, 2.5 points; 30 to 100: 0 points; (3) total WBC count—less than 8 x 10^–9/L, 0 points; 8 to 11.5 x 10^–9/L, 0.5 points; more than 11.5 x 10^–9/L, 1.5 points; and (4) lymphocyte percentage—more than 20%, 0 points; 12% to 20%, 1 point; less than 12%, 2.5 points.

Scores are also given for the clinician's estimate of survival, measured in weeks: 1 to 2 weeks, 8.5 points; 3 to 4 weeks, 6.5 points; 5 to 6 weeks, 4.5 points; 7 to 10 weeks, 2.5 points; 11 to 12 weeks, 1.5 points; and more than 12 weeks, 0 points.

These partial scores are combined to give a total score for the patient (possible range, 0 to 17.5). Validated cut points have been established that categorize patients into the three prognostic groups for survival at 30 days, based on the total PaP score^6,7: group A, more than 70% probability of 1-month survival, 0 to 5.5 points; group B, 30% to 70% probability of 1-month survival, 6 to 11 points; group C, less than 30% probability of 1-month survival, 11.5 to 17.5 points.

The survival time of 30 days was chosen on clinical grounds. The cut points were chosen to divide the survival probability into three approximately equal groups.

Methods
Participants were recruited as soon as possible after admission to the hospital. Because all patients were known to have advanced cancer, they had already undergone some level of assessment of the status of their malignancies. After obtaining written informed consent, a research assistant obtained baseline patient characteristics (age, sex, name of treating oncologist, name of family physician) and clinical information (cancer site, date of diagnosis, extent of disease, current anticancer treatment modalities, previous anticancer treatment modalities, date of admission to hospital, reason for admission to the hospital), and then asked the participant the following three questions: "In the past 24 hours, has your appetite been decreased?"; "In the past 24 hours, have you been short of breath?"; and "If you were at home today, what sort of activities would you be able to do?"

The answers were used to determine if the patient was anorexic or dyspneic, and to determine his or her performance status, respectively. The research assistant obtained the most recent hematology results. If there were no results available within the last 7 days, then a hematology workup was performed. As soon as possible after recruitment, the patient's treating oncologist was asked to make a temporal survival prediction, based on his or her opinion of the median survival of similar patients. The oncologist's clinical prediction of survival (CPS) was elicited via a two-step process for the purposes of calculating the PaP score: first, the oncologist was asked to decide if he or she believed the patient's median survival was less than 3 months; if so, the oncologist was asked to express the prognosis in 2-week intervals, up to a maximum of 12 weeks (ie, 1 to 2, 3 to 4, 5 to 6, 7 to 8, 9 to 10, and 11 to 12 weeks).

This information was collated by the research assistant to calculate the PaP score and categorize patients into groups with a high (> 70%), intermediate (30% to 70%), or low (< 30%) probability of surviving 1 month (groups A, B, and C, respectively). Dates of death were obtained from the oncologist, palliative care service, or family physician. The oncologist was blinded to the patient's calculated PaP score.

Statistical Considerations
Sample size. A cohort of 100 consecutive eligible patients was assembled for the purpose of this pragmatic study of the PaP score as a clinical prediction rule. This sample size was chosen because of the limited resources available for this unfunded study.

Data analysis. The positive predictive value (PPV) was used to describe the diagnostic accuracy of the PaP score. The Kaplan-Meier method and log-rank test were used to compare the survival distributions of patients in the three PaP score groups. Survival was censored at the date of last known contact in participants lost to follow-up or at 180 days for participants who were still alive. To determine if PaP score performs similarly according to tumor type, we compared the PPV for the PaP score according to those nonhematologic malignancies that are associated with anorexia-cachexia (upper gastrointestinal, lung) and those that are not (breast, sarcomas, colorectal).¹¹

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... REFERENCES

 
Patient Characteristics
During the study period there were 378 admissions of 250 patients under the medical oncology and radiation oncology services. Patients being treated with curative intent (n = 130) were ineligible. Of the 120 eligible patients, 20 were excluded (10 were too unwell to consent, five refused to consent, and five were unable to be recruited because they were in the hospital less than 24 hours). The 100 patients included 55 males, and the median age was 66 years (range, 20 to 88 years). The clinical characteristics of the patients are listed in Table 1.

Thirty patients had already been referred to palliative care services before the hospitalization during which participation took place. During the hospitalization, another 38 patients were referred to the hospital palliative care service. Sixteen were referred later in their illness, whereas another 16 patients have never been referred to palliative care.

Overall Survival
Survival status was known at 1 month for 98 patients and at 6 months for 93 patients. The median survival was 12 weeks; 82 patients (83%) were alive at 1 month and 25 patients (26%) were alive at 6 months.

PaP Score
The prevalence of the abnormalities contributing to the PaP score are shown in Table 2. Most participants complained of anorexia but less than half reported dyspnea. The median KPS was 60; very few patients had a KPS score of less than 30. Recent hematologic results (collected within the last 48 hours) were available in 78 patients. All but five had blood drawn within the last week. Because of an oversight, a new sample was not collected in one of these patients and the results of the most recent sample (> 1 month previously) had to be used in that patient to calculate the PaP score. WBC abnormalities were uncommon, even though more than half of the patients in the sample were still receiving chemotherapy. The oncologists expected 55 patients to live for more than 3 months and expected 10 patients to live for 1 month. The median CPS category selected was 7 to 8 weeks.

Outcome of PaP Groups
The observed median survival of groups A, B, and C was 17, 7, and less than 1 week, respectively. The PPV of the three PaP groups and their respective estimated median survivals are listed in Table 3. The Kaplan-Meier survival curves are shown in Fig 1. The linear trend for survival across the PaP groups was highly significant (log-rank test for trend ₁², 25.65; P < .0001). Whether the patient had a tumor type commonly associated with the anorexia-cachexia syndrome (ACS; n = 32) or not (n = 29) did not affect the utility of the PaP score. The PPV of PaP group A for being alive at 30 days was 93% (13 of 14) for those associated with ACS and 90% (18 of 20) for those not associated. For PaP group B, the PPV was 88% (seven of eight) and 67% (four of six), respectively.

View larger version (15K): [in this window] [in a new window]   Fig 1. Kaplan-Meier survival curves for patients with advanced cancer.

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... REFERENCES

 
This is the first time that PaP score, originally designed for use as a clinical prediction rule in hospice or palliative care populations, has been tested in patients with less advanced disease still under the care of an oncologist. The participants in this study were typical of medical oncology patients with advanced cancer. The majority of patients were still receiving anticancer treatment, the median performance status was fair, and the median survival was measured in terms of months rather than weeks. Even though the patients were clinically different compared with a palliative care population,^6,7,12,13 the PaP score was able to categorize them into three isoprognostic groups on the basis of the 1-month survival probability.

This short study also showed that the participating oncologists could discriminate reasonably well among patients likely to survive 1 month, those likely to survive more than 3 months, and those who would not, but their predictions were not well enough calibrated to predict more accurately. In patients given a better prognosis, the predictions tended to be overly optimistic, whereas in those with a poorer outlook, the predictions tended to be overly pessimistic. These results confirm previous findings that oncologists' estimates of the survival of incurable patients are well calibrated but that their individual predictions are imprecise.¹⁴ These results are also consistent with a recent systematic review, which showed that physicians' survival predictions in advanced cancer are inaccurate, but not that they are usually overly optimistic.³ There may be two explanations for this discrepancy. First, many of the studies included in the systematic review were carried out in hospice populations, and the psychology of prognostication may be different in terminally ill patients.¹⁵ Second, recent studies in patients with advanced cancer have shown a tendency for prognoses to be more pessimistic than they were in the older studies.^16,17 No attempt was made to analyze the data according to the oncologists because of the small number involved, the fact that they all work in the same cancer center, and because none had been given specific training in prognostication.

Because prognostication is not explicitly taught in medical oncologic training, oncologists often feel vulnerable when asked to predict a patient's survival. Increasingly, simple, valid prognostic models and nomograms are being developed in medical oncology, but typically relate to specific disease entities or treatment scenarios.^18–21 Simple, valid tools that can be used in routine clinical settings are needed, and in this study the PaP score has demonstrated some ability to do this.

Would application of the PaP score have altered the seven physicians' responses to the theoretical patient who asks, "Doctor, how much time do I have?"¹ We know from the scenario in that discussion that the patient is anorexic but not breathless. We can infer that her KPS is more than 20 because she has been able to come to the physician's office; we do not know the results of her hematology work-up. Using the median survival of her cancer (4 weeks) for the clinician's survival estimate, she has a PaP score of 8 to 9.5 unless her hematology work-up is grossly abnormal, putting her in PaP group B with an intermediate prognosis for 1 month and, on the basis of the survival analysis of the study sample, she has a 25% chance of being alive in 3 months. Irrespective of the manner in which the oncologist decides to communicate this information to the patient, the PaP score is able to reduce much of the uncertainty when formulating a prognosis in patients such as this.

This study suggests the PaP score is to able to predict reasonably well the short-term survival of patients with advanced cancer in oncology, but the validity and usefulness of the PaP score in larger oncologic samples and other clinical settings needs to be tested. One issue that needs further consideration is the relevance of the prognostic factors in the PaP score to the oncology population. In patients with advanced cancer, factors related to the primary cancer at the time of diagnosis, such as site, size, grade, and extent, become less important than patient-related factors such as performance status and quality of life.⁴ In patients with advanced disease, simple prognostic factors measurable on routine clinical evaluation, such as performance status, symptomatology, and basic laboratory tests, are also preferable to invasive procedures on ethical grounds.²²

The interpretation of leukocytosis and lymphopenia in patients receiving anticancer treatment requires some consideration. None of the patients in this study had hematologic malignancies. Leukocytosis has been identified as a poor prognostic marker in patients with various nonhematologic malignancies at the time of diagnosis.²³ The cause is unclear, but various tumor-related causes have been proposed. Although most chemotherapy agents cause marrow toxicity and some can cause profound neutropenia, there was no significant difference in the WBC count in those receiving chemotherapy compared with those who were not. Likewise, the cause of lymphopenia in patients with advanced cancer is unclear, but it has been suggested that it may be a consequence of the ACS.⁹ The fact that the PaP score was effective in a heterogeneous population of patients, approximately half of whom were receiving chemotherapy or radiation therapy, only serves to emphasize its robustness in routine clinical practice. It is noteworthy that the PaP score worked equally well in this study whether or not patients had tumors typically associated with ACS.¹¹

One limitation of the PaP score is that it only predicts the probability of surviving 1 month. This time point may be relevant to some clinical decisions in patients with far-advanced cancer (eg, to avoid starting a new line of chemotherapy in the last month).²⁴ Survival probability at other time points (eg, 3, 6, and 12 months) will also be important in other clinical situations, and tools for predicting survival at a variety of time points are needed. Nutritional indices, patients' self-report data, and novel prognostic factors such as C-reactive protein, interleukin 6, and vascular endothelial growth factor may need to be considered for inclusion in such a model.^25–27

In conclusion, the PaP score is a simple, valid tool for busy oncologists to use at the bedside when faced with formulating a prognosis. Although its clinical usefulness is limited by the fact that it can only predict the probability of surviving 1 month, this is an important milestone for many patients with advanced cancer.

	Authors' Disclosures of Potential Conflicts of Interest

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... REFERENCES

 
The authors indicated no potential conflicts of interest.

	Acknowledgment

 
We thank Associate Professor Michael Boyer, Sydney Cancer Center, Camperdown, Australia, and Dr Marco Maltoni, Divisione di Oncologica Medica, Ospedale Pierantoni, Forli, Italy, for their assistance in preparing the article.

	NOTES

 
Authors' disclosures of potential conflicts of interest are found at the end of this article.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION Authors' Disclosures of... REFERENCES

 
1. Loprinzi CL, Johnson ME, Steer G: "Doc, how long have I got? " J Clin Oncol 18:699–701, 2000[Free Full Text]

2. Schapira L, Eisenberg PD, Macdonald N, et al: A revisitation of "Doc, how much time do I have?" J Clin Oncol 21:8S–11S, 2003[CrossRef][Medline]

3. Glare P, Virik K, Jones M, et al: A systematic review of physicians' survival predictions in terminally ill cancer patients. BMJ 327:195–198, 2003[Abstract/Free Full Text]

4. Mackillop WJ: The importance of prognosis in cancer medicine, in Gospodarowicz, M (ed): Prognostic Factors in Cancer (ed 2). New York, NY, Wiley-Liss, 2001, pp 3–14

5. Pirovano M, Maltoni M, Nanni O, et al: A new palliative prognostic score: A first step for the staging of terminally ill cancer patients. J Pain Symptom Manage 17:231–239, 1999[CrossRef][Medline]

6. Maltoni M, Nanni O, Pirovano M, et al: Successful validation of the Palliative Prognostic Score in terminally ill cancer patients. J Pain Symptom Manage 17:240–247, 1999[CrossRef][Medline]

7. Glare P, Virik K: Independent prospective validation of the PaP score in terminally ill patients referred to a hospital-based palliative medicine consultation service. J Pain Symptom Manage 22:891–898, 2001[CrossRef][Medline]

8. Maltoni M, Pirovano M, Scarpi E, et al: Prediction of survival in terminally ill cancer patients: Results of an Italian prospective, multicentric trial. Cancer 75:2613–2622, 1995[CrossRef][Medline]

9. Maltoni M, Pirovano P, Nanni O, et al: Biological indices predictive of survival in 519 Italian terminally ill cancer patients. J Pain Symptom Manage 13:1–9, 1997[Medline]

10. Karnofsky DA, Burchenal JH: The clinical evaluation of chemotherapeutic agents in cancer, in MacLeod CM (ed): Evaluation of Chemotherapeutic Agents. Symposium, Microbiology Section, New York Academy of Medicine. New York, NY, Columbia University, 1949, pp 191–205

11. De Wys WD, Begg D, Lavin PT: Prognostic effect of weight loss prior to chemotherapy. Am J Med 69:491–499, 1980[CrossRef][Medline]

12. Jenkins CA, Schulz M, Hanson J, et al: Demographic, symptom, and medication profiles of cancer patients seen by a palliative care consult team in a tertiary referral hospital. J Pain Symptom Manage 19:174–184, 2000[CrossRef][Medline]

13. Llobera J, Esteva M, Rifa J, et al: Terminal cancer: Duration and prediction of survival time. Eur J Cancer 36:2036–2043, 2000

14. Mackillop WJ, Quirt CF: Measuring the accuracy of prognostic judgments in oncology. J Clin Epidemiol 50:21–29, 1997[CrossRef][Medline]

15. Christakis NA. Death foretold. Chicago University Press, Chicago, IL, 1999, pp 64–83

16. Morita T, Tsunoda J, Inoue S, et al: Improved accuracy of physicians' survival predictions for terminally ill cancer patients using the Palliative Prognostic Index. Palliat Med 15:419–424, 2001[Abstract/Free Full Text]

17. Faris M: Clinical estimation of survival and impact of other prognostic factors on terminally ill cancer patients in Oman. Support Care Cancer 11:30–34, 2003[Medline]

18. Tan CK, Law NM, Ng HS, et al: Simple clinical prognostic model for hepatocellular carcinoma in developing countries and its validation. J Clin Oncol 21:2294–2298, 2003[Abstract/Free Full Text]

19. Sloan JA, Loprinzi CL, Laurine JA, et al: A simple stratification factor prognostic for survival in advanced cancer: The good/bad/uncertain index. J Clin Oncol 19:3539–3546, 2001[Abstract/Free Full Text]

20. Chow E, Fung K, Panzarella T, et al: A predictive model for survival in metastatic cancer patients attending an outpatient palliative radiotherapy clinic. Int J Radiat Oncol Biol Phys 53:1291–1302, 2002[CrossRef][Medline]

21. Smaletz O, Scher HI, Small EJ, et al: Nomogram for overall survival of patients with progressive metastatic prostate cancer after castration. J Clin Oncol 20:3972–3982, 2002[Abstract/Free Full Text]

22. Calman KC, Hanks GWC: Clinical and health services research in palliative care, in: Doyle D, Hanks GWC, Macdonald N (eds): Oxford Textbook of Palliative Medicine. Oxford University Press, Oxford, United Kingdom, 1993, pp 73–76

23. Shoenfeld Y, Tal A, Berliner S, et al: Leukocytosis in non-hematological malignancies: A possible tumor-associated marker. J Cancer Res Clin Oncol 111:54–58, 1986[CrossRef][Medline]

24. Earle CC, Neville BA, Landrum MB, et al: Trends in the aggressiveness of cancer care near the end of life. J Clin Oncol 22:315–321, 2004[Abstract/Free Full Text]

25. Loprinzi CL, Laurie JA, Wieand HS, et al: Prospective evaluation of prognostic variables from patient-completed questionnaires: North Central Cancer Treatment Group. J Clin Oncol 12:601–607, 1994[Abstract]

26. Shadbolt B, Barresi J, Craft P: Self-rated health as a predictor of survival among patients with advanced cancer. J Clin Oncol 20:2514–2519, 2002[Abstract/Free Full Text]

27. Walsh D, Mahmoud F, Barna B: Assessment of nutritional status and prognosis in advanced cancer: Interleukin-6, C-reactive protein, and the prognostic and inflammatory nutritional index. Support Care Cancer 11:60–66, 2003[Medline]

28. Lutgendorf SK, Johnsen EL, Cooper B, et al: Vascular endothelial growth factor and social support in patients with ovarian carcinoma. Cancer 95:808–885, 2002[CrossRef][Medline]

Submitted December 9, 2003; accepted September 17, 2004.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				E. Finlay and D. Casarett Making Difficult Discussions Easier: Using Prognosis to Facilitate Transitions to Hospice CA Cancer J Clin, June 17, 2009; (2009) caac.20022v1. [Abstract] [Full Text]

				M. P. Davis, T. Yavuzsen, D. Khoshknabi, J. Kirkova, D. Walsh, W. Lasheen, R. Lagman, and M. T. Karafa Bioelectrical Impedance Phase Angle Changes During Hydration and Prognosis in Advanced Cancer American Journal of Hospice and Palliative Medicine, June 1, 2009; 26(3): 180 - 187. [Abstract] [PDF]

				M. A. Marinella On the Hippocratic Facies J. Clin. Oncol., July 20, 2008; 26(21): 3638 - 3640. [Full Text] [PDF]

				S. Gripp, S. Moeller, E. Bolke, G. Schmitt, C. Matuschek, S. Asgari, F. Asgharzadeh, S. Roth, W. Budach, M. Franz, et al. Survival Prediction in Terminally Ill Cancer Patients by Clinical Estimates, Laboratory Tests, and Self-Rated Anxiety and Depression J. Clin. Oncol., August 1, 2007; 25(22): 3313 - 3320. [Abstract] [Full Text] [PDF]

				P. Stone and S Lund Predicting prognosis in patients with advanced cancer Ann. Onc., June 1, 2007; 18(6): 971 - 976. [Abstract] [Full Text] [PDF]

				B. Aabom, J. Kragstrup, H. Vondeling, L.S. Bakketeig, and H. Stovring Defining Cancer Patients As Being in the Terminal Phase: Who Receives a Formal Diagnosis, and What Are the Effects? J. Clin. Oncol., October 20, 2005; 23(30): 7411 - 7416. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Glare, P. A. Articles by McMahon, P. Search for Related Content PubMed PubMed Citation Articles by Glare, P. A. Articles by McMahon, P. Social Bookmarking                            What's this?