The cancer anorexia-cachexia syndrome: myth or reality?

Transcription

1 Support Care Cancer (2010) 18: DOI /s SHORT COMMUNICATION The cancer anorexia-cachexia syndrome: myth or reality? Wael Lasheen & Declan Walsh Received: 23 July 2009 / Accepted: 26 October 2009 / Published online: 24 November 2009 # Springer-Verlag 2009 Abstract Background Controversy exists as what constitutes the cancer anorexia-cachexia syndrome (CACS), and whether it truly is a distinct clinical disorder. In this study, we aimed to: (1) assess if CACS is a distinct clinical disorder, (2) identify the symptoms characteristic of CACS, (3) evaluate CACS impact on patient outcomes (symptom burden and survival time from referral). Methods Consecutive patients referred to palliative medicine were assessed by 38-symptom questionnaire. Demographics, Eastern Cooperative Oncology Group (ECOG), disease and extent, and survival were recorded. CACS, defined as anorexia plus weight loss (>10% of pre-illness weight). For analysis, patients were divided into four groups: (1) group CACS; (2) group A (only anorexia, NO >10% pre-illness weight loss); (3) group WL (weight loss >10% pre-illness weight only but NO anorexia); and (4) group N (NO weight loss >10% pre-illness weight and NO anorexia). Symptoms present in 5%, and patients with complete data were analyzed. Results Four hundred eighty-four patients had complete data, metastatic cancer, and 26 symptoms present in 5%. Groups had significantly different ECOG, symptom burden, and survival. Significantly different symptom prevalence between groups: dry mouth,*early satiety,*constipation, *nausea,*taste changes,*vomiting,*dysphagia,*fatigue, *weak,*lack of energy, insomnia, dyspnea, depression, hoarseness, and anxiety. The nine symptoms with asterisk were CACS specific. Symptom Burden: CACS independently predicted greatest burden. Survival: Group N had significantly longer survival. Conclusions CACS appeared to be a distinct disorder with unique clinical characteristics in our advanced cancer population. Nine other symptoms constituted CACS. CACS independently predicted higher symptom burden. CACS absence predicted longer survival. More evidence is needed to better characterize this syndrome and generate a valid CACS consensus. A comprehensive validated CACS assessment instrument is required. Keywords Cancer. Cachexia. Anorexia. Anorexia-cachexia. Symptoms. Symptom burden. Survival. Prognosis. Palliative care A World Health Organization Demonstration Project in Palliative Medicine. An ESMO Designated Center of Integrated Oncology and Palliative Care. W. Lasheen : D. Walsh The Harry R Horvitz Center for Palliative Medicine, Department of Solid Tumor Oncology, Cleveland Clinic, Taussig Cancer Institute, Cleveland, OH, USA D. Walsh (*) Cleveland Clinic, 9500 Euclid Ave. M76, Cleveland, OH 44195, USA walsht@ccf.org URL: Introduction More than 3,000 years ago, symptoms of what is now referred to as the anorexia-cachexia syndrome were reported. I forget to eat my bread...my knees are weak through fasting...my breath is offensive to my wife... My flesh failed of fatness...i am nothing but skin and bones (Bible) [1] Many clinical features have been associated with this syndrome. One symptom, anorexia, because so synonymous with weight loss that the primary syndrome in

2 266 Support Care Cancer (2010) 18: cancer patients is usually called The Anorexia Cachexia Syndrome (CACS). CACS research is fraught with difficulties [2, 3]: 1. There is no agreed-upon definition for the syndrome; 2. No consensus regarding symptoms and signs comprising CACS; 3. Weight loss duration controversial (weight loss over the past 1 month, 2 months, 6 months, or since disease started). This may render weight loss a static indicator with a less dynamic prognostic impetus; 4. Weight loss degree poorly defined (2 lb., 5%, 10%, etc.); 5. Anorexia severity and duration is another area of contention; 6. Duration of concurrent anorexia and weight loss seldom addressed; and 7. Impact of medications or cancer treatment on the syndrome and its diagnosis. Indeed, there is still controversy as what constitute CACS and whether it is distinct from cachexia [4, 5]. In some reviews and articles, the two terms are used interchangeably, while others avoid using CACS entirely. Cachexia has historically been defined and assessed as involuntary weight loss, yet a recent study of pancreatic cancer found no prognostic or survival value to weight loss [6]. Cachexia was defined in terms of three preselected variables: weight loss, and/or reduced food intake (using a diary), and/or C-reactive protein. This study demonstrates the discordance in the literature, and highlights the complexity of CACS research. Other studies have variously defined cachexia in terms of body fat, fat-free mass, serum albumin, C-reactive protein (erythrocyte sedimentation rate), insulin, daily caloric intake, serum insulin-like growth factor 1, resting energy expenditure, leptin, ghrelin, thyroid hormone, tumor-necrosis factor-α; interleukin-1 (IL-6), and interferon-γ [7, 8]. Advances in symptom pathophysiology will eventually clarify CACS. Once CACS is diagnosed, it is important to identify conveniently assessable symptoms to facilitate identification, treatment, and evaluation in both clinical and research settings. Repeated biochemical or body compartment testing may be impractical in clinical practice or many human research trials, especially in advanced disease. In this study, we evaluated whether CACS is a distinct clinical entity, identified the clinical features, and assessed their impact. Materials and methods This was a post hoc analysis of previously reported data [9]. It was approved by the Cleveland Clinic Institutional Review Board. The study group was a prospective cohort of consecutive patients. One thousand consecutive cancer in- and out-patient consults to the palliative medicine team (PMT) were included. Few patients were receiving active treatment. Data was collected using an eight-page questionnaire during initial consultation. Data included demographics, diagnosis, Eastern Cooperative Oncology Group (ECOG) performance score, and 38 symptoms. The symptoms were empirically selected to provide a comprehensive tool to establish individual symptom profiles for routine medical practice. The questionnaire was verbally administered by a trained PMT member. Survival data was collected later. A higher ECOG score (0 4) indicated worse performance status. Symptoms were noted as absent and rated on a categorical scale if present (none, mild, moderate, and severe). Weight loss was rated as present if there was more than 10% loss of pre-illness weight. In this study, only symptoms with an overall prevalence 5% were analyzed. This was because after the exclusion of those with missing data, we would have had insufficient patients (with complete data set) to analyze. We did not have data on prior anti-tumor therapies. CACS was defined as the combination of both weight loss (>10% of pre-illness body weight) and anorexia of any severity. Weight loss >10% of pre-illness body weight has been used in other studies to identify cancer cachexia [10]. Symptom burden was defined as the sum of all symptom severity scores per patient. Anorexia and weight loss were excluded from the calculated symptom burden because they were later analyzed as symptom burden predictors. The objectives were to: 1. Identify the clinical symptom characteristics of CACS and 2. Evaluate CACS independent impact on patient outcomes, assessed by symptom burden, and survival from the time of referral. These objectives were achieved by dividing the patients into four groups: 1. Group CACS = cancer anorexia-cachexia syndrome group = weight loss (>10% of pre-illness weight) and anorexia (of any severity) 2. Group A = those with anorexia (of any severity) and NO weight loss ( 10% pre-illness weight) 3. Group WL = cachexia = those with weight loss (>10% pre-illness weight) and NO anorexia 4. Group N = those with NO anorexia and NO weight loss ( 10% pre-illness weight) Subsequently, we compared the four groups. We identified the characteristics of each and their impact on symptom burden and survival. Symptoms were

3 Support Care Cancer (2010) 18: considered special to CACS if they were significantly moreprevalentinthegroupcacsascomparedtoeither group A or group WL. Statistical analysis Any patient with missing symptom data was excluded. Descriptive statistics were used to report the mean, standard deviation, median, and range. Demographic variables, symptom frequency, severity, and survival between patient groups were reported and compared using appropriate nonparametric tests (Fisher s test, chi-square, and Kruskal Wallis test). Only significant differences in symptom prevalence between groups were reported in the tables; insignificant differences were reported in the text. A stepwise multiple regression-mixed selection model was constructed with a variable entry criterion of 0.25 and a variable retention criterion of The dependent variable was symptom burden; the independent variables were age, gender, ECOG, primary cancer site, group CACS, group A, group WL, and group N. Survival curves were constructed and survival between groups compared using both the log-rank and Wilcoxon tests. A P value of <0.05 was considered statistically significant. When multiple comparisons were performed, the P value was set at <0.01. All percentages were rounded to the nearest whole number. Analysis was performed with JMP statistical software package (JMP, Version 7. SAS Institute Inc., Cary, NC, USA ). Results Demographic data Initial assessments of 1,000 patients were available for analysis. Symptoms present in 5% were analyzed; there were 26 such symptoms: anorexia, anxiety, belching, bloating, confusion, constipation, cough, depression, dizziness, dry mouth, dyspepsia, dysphagia, dyspnea, edema, insomnia, lack of energy, fatigue, hoarseness of voice, nausea, pain, early satiety, taste changes, vomiting, weakness, wheezing, and weight loss. Those with any missing data were excluded. The total number of patients with complete data in the final analysis was 484. All had metastatic cancer (Table 1). Group CACS, group A, group WL, and group N The groups had similar age and gender distributions but significantly different ECOG, symptom burden, and survival (see Table 1). There were no significant differences in belching, bloating, confusion, cough, dizziness, dyspepsia, edema, pain, and wheezing prevalence between groups. For significant differences in symptom prevalence between groups, see Table 1. CACS-specific symptom profile Nine symptoms were significantly more common in the CACSgroupwhencomparedtogroupAandgroupWL (P<0.01). These were constipation, dysphagia, early satiety, fatigue, lack of energy, nausea, taste changes, vomiting, and weakness. Group WL and group A were significantly different (P<0.01) in two symptoms; dysphagia (more common in group WL) and early satiety (more common in group A). The total number of gastrointestinal (GI) symptoms per patient (Table 1) were significantly (P<0.01) greatest in group CACS (as compared to both group WL and group A). There were no differences between group CACS, group WL, and group A in the total number of non-gi symptoms per patient. Symptom burden CACS independently predicted the greatest symptom burden (Table 2), but group A and group WL did not. Group N predicted the lowest symptom burden. ECOG score and age were independent predictors of symptom burden. Survival There was a statistically significant difference between groups: group N had significantly longer survival (Fig. 1a). There was no significant survival difference between group CACS, group A, and group WL (Fig. 1b). Discussion This is the first report, to our knowledge, attempting to describe the unique symptom characteristics of the cancer anorexia-cachexia syndrome. We did that using a comprehensive 38-symptom survey rather than a potentially biased brief questionnaire. We validated CACS as a distinct entity with a unique clinical profile. This was further confirmed by a higher symptom burden in the CACS group. CACS independently predicted this, but group A and group WL did not (Table 2). Although there was no survival difference between group CACS, group A, and group WL, group N had significantly longer survival. It is noteworthy that group WL had low symptom burden and better performance status (ECOG score). The cause is not immediately evident but it could be due to the

4 268 Support Care Cancer (2010) 18: Table 1 Demographic data and symptoms prevalence Total Group A Group WL Group N Group CACS P Value * N (%) 484 (100%) 163 (34%) 46 (10%) 125 (26%) 150 (31%) Age b 65 (21 94) 67 (25 920) 60 (21 83) 66 (25 94) 65 (29 89) NS Gender 259 (54%) 84 (51%) 22 (48%) 66 (53%) 87 (58%) NS ECOG b 3(0 4) 3 (0 4) 2 (0 3) 3 (0 4) 3 (1 4) <0.05 Symptom burden b, c 14.5 (1 39) 16 (1 36) 13.5 (3 31) 14 (1 39) 18 (2 38) <0.05 Survival b, d 1.8 (0 40) 1.8 (0 35) 1.8 (0 17) 2.5 (0 40) 1.4 (0 27) <0.05 Five most common primary cancer sites a Lung 111 (23%) 43 (26%) 8 (17%) 22 (18%) 38 (25%) NS Colorectal 51 (10%) 11 (7%) 6 (13%) 13 (10%) 21 (14%) Breast 49 (10%) 16 (10%) 5 (11%) 14 (11%) 14 (9%) Prostate 39 (18%) 16 (10%) 2 (4%) 13 (10%) 8 (5%) Pelvic 33 (7%) 15 (9%) 4 (9%) 9 (7%) 5 (3%) Gastrointestinal (GI) symptom prevalence a Anorexia 313 (65%) Dry mouth 266 (55%) 94 (58%) 22 (48%) 53 (42%) 97 (65%) <0.05 ** Early satiety 246 (51%) 102 (63%) 18 (39%) 21 (17%) 105 (70%) <0.05 ** Constipation 232 (48%) 82 (50%) 17 (37%) 46 (37%) 87 (58%) <0.05 Weight loss 196 (40%) ** Nausea 169 (35%) 60 (37%) 12 (26%) 31 (25%) 66 (44%) <0.05 ** Taste changes 136 (28%) 52 (32%) 7 (15%) 9 (7%) 68 (45%) <0.05 ** Vomiting 107 (22%) 30 (18%) 5 (11%) 24 (19%) 48 (32%) <0.05 Dyspepsia 93 (19%) 29 (18%) 7 (15%) 26 (21%) 31 (21%) NS Belching 91 (19%) 29 (18%) 10 (22%) 16 (13%) 36 (24%) NS ** Dysphagia 77 (16%) 18 (11%) 13 (28%) 15 (12%) 31(21%) <0.05 Bloating 63 (13%) 22 (14%) 5 (11%) 11 (9%) 25 (17%) NS Total GI symptoms/patient: Median (range) 3 (0 8) 2.5 (0 5) 2 (0 10) 4 (0 5) <0.05 Significant non-gi symptom prevalence a ** Fatigue 308 (64%) 100 (61%) 31 (67%) 54 (43%) 123 (82%) <0.05 ** Weak 292 (60%) 100 (61%) 29 (63%) 51 (41%) 112 (75%) <0.05 ** Lack of energy 260 (54%) 79 (48%) 29 (63%) 44 (35%) 108 (72%) <0.05 Insomnia 226 (47%) 82 (50%) 21 (46%) 43 (34%) 80 (53%) <0.05 Dyspnea 219 (45%) 83 (51%) 18 (39%) 43 (34%) 75 (50%) <0.05 Depression 200 (41%) 67 (41%) 20 (43%) 38 (30%) 75 (50%) <0.05 Hoarseness 131 (27%) 50 (31%) 15 (33%) 20 (16%) 46 (31%) <0.05

5 Support Care Cancer (2010) 18: Table 1 (continued) Total Group A Group WL Group N Group CACS P Value * Anxiety 109 (23%) 41 (25%) 13 (28%) 16 (13%) 39 (26%) <0.05 Total Non-GI symptoms/patient: Median (range) 5 (0 14) 5.5 (1 11) 4 (0 12) 6.5 (0 13) <0.05 Group A those with anorexia (of any severity) and NO weight loss ( 10%of pre-illness weight) Group WL those with weight loss >10% of pre-illness weight and NO anorexia Group N those with NO anorexia and NO weight loss ( 10% of pre-illness weight) Group CACS weight loss >10% of pre-illness weight and anorexia (of any severity) * P<0.05 significance, P value comparing the four groups ** P<0.025 significance, symptoms significantly more common in group CACS when compared to either group A or group WL a N (%); total GI symptom prevalence does not include anorexia and weight loss b Median (range) c Symptom burden; sum of symptoms severity scores per patient d Survival in months from time of referral All percentages rounded to the nearest whole number absence of anorexia (which added an extra burden in the other groups), a different subtype of cachexia (secondary cachexia), or it could reflect a shortcoming of considering only pre-illness weight loss rather than, say, weight loss in the past month or two. Albeit statistically insignificant, we observed some variance by primary cancer site (colon cancer in 14% of group CACS vs. 7% in group A). More importantly, CACS patients had significantly more GI symptoms but not non-gi symptoms compared to group A and group WL. Our data suggests that CACS is a subtype of cachexia and not a different syndrome. Just like insulin- and noninsulin-dependent diabetes; they are both subtypes of diabetes, yet essentially distinct. Cachexia has been classified [5] into (1) primary cachexia or CACS and (2) secondary cachexia. This approach is supported in our study by a distinct clinical symptom profile, and significant differences in symptom burden between CACS and group WL. Cancer anorexia cachexia is a metabolic syndrome which appears to consist of progressive wasting and loss of lean body mass, with or without loss of fat mass. Weight loss is the hallmark of cachexia in adults [11], yet how much weight loss constitutes cachexia is unclear. Cachexia has been variously defined as weight loss of: 1. >10% of pre-illness weight [12, 13] 2. >10% in the past 6 months [14] 3. >5% of pre-illness weight [15] 4. >5% in the previous 6 months [5] 5. >5% in the previous 6 weeks [16] 6. >5 lb in the last 2 months [17]. In general, some try to capture the velocity of weight loss over an extended period (pre-illness or past 6 months), others choose a more short-term weight change. Any weight loss, using pre-illness weight as a reference point, has been associated with poor response to chemotherapy, increased toxicity, and poor survival [18]. Short-term weight loss was associated with poor survival in advanced cancer [19]. It seems sensible to include both short- and long-term weight losses in a single definition; however, this needs further validation. We used >10% weight loss (of pre-illness weight) as it is widely recognized and used as an indicator of severe weight loss [12, 13]. Similarly, anorexia, which constitutes the other half of the syndrome, is poorly defined. Some suggest an anorexia score of more than three 10 cm. on a visual analog scale is relevant, others accept any anorexia [5]. Additionally, the importance of anorexia duration, chronicity, and course has not been addressed. This area needs further research. A number of symptoms have previously been described as components of CACS. These have included anorexia, asthenia, chronic nausea, changes in body

6 270 Support Care Cancer (2010) 18: Table 2 Stepwise multivariate analysis: predictors of symptom burden Independent variables Symptom burden (dependent variable) Estimate F ratio P value Group N < Group CACS < ECOG <0.01 Age <0.05 Group A NS Group WL NS Symptom burden sum of symptom severity scores per patient Group N those with NO anorexia and NO weight loss ( 10% of preillness weight) Group CACS weight loss >10% of pre-illness weight and anorexia (of any severity) Group A those with anorexia (of any severity), and NO weight loss ( 10%of pre-illness weight) Group WL those with weight loss >10% of pre-illness weight and NO anorexia image, and reduced food intake [5].Thesewerelargely based on clinical observation rather than systematic analysis in a clinical trial. In our study, constipation, dysphagia, early satiety, fatigue, lack of energy, nausea, taste changes, vomiting, and weakness were all more common (P<0.01) in the CACS group (as compared to either group A or group WL) and seems to constitute a part of this unique syndrome s presentation. Our findings do not necessarily diagnose CACS, but rather describe its clinical presentation and perhaps severity. It underscores the importance of assessment and monitoring of these symptoms to better evaluate treatment in practice and help clinical studies. The paucity of the evidence base to what constitutes CACS renders efforts to reach a consensus difficult. A recent consensus conference [20] proposed a new cachexia definition: weight loss of at least 5% of body weight over at most the past 12 months and at least three of the following: 1. Decreased muscle strength 2. Fatigue 3. Anorexia 4. Low fat-free mass index or 5. Abnormal biochemistry: as increased inflammatory markers (CRP and IL-6), anemia, or low serum albumin. The major flaw is the absence of coherent clinical evidence to guide these statements. Weaknesses, we believe, include grouping cancer and non-cancer etiologies in one definition and apparent incongruity with the Fig. 1 Kaplan Meir survival curves for the different patient groups. a Survival differences between group A, group WL, group N, and group CACS. b Survival differences between group A, group WL, AND group CACS. Group A (N=163) those with anorexia (of any severity), and NO weight loss ( 10%of pre-illness weight). Group WL (N=46) those with weight loss >10% of pre-illness weight and NO anorexia. Group N (N=125) those with NO anorexia, and NO weight loss ( 10% of pre-illness weight). Group CACS (N=150) weight loss >10% of pre-illness weight and anorexia (of any severity) available clinical evidence. For example, a person with weight loss (5%), anorexia, fatigue, and anemia would be defined as cachectic; not uncommon observations in cancer patients. In contrast, the strength of our study is that our conclusions are based on clinical evidence. Another study [11] used a similar approach as ours. Patients were divided into four groups based on degree of weight loss and presence/absence of anorexia, early satiety, and fatigue. There were differences in symptom prevalence, intensity, ECOG scores, and nutritional risk between groups. However, these finding are not necessarily generalizable because: 1. Two-thirds of patients were on active anti-tumor treatment (perhaps itself the cause of weight loss and various symptoms)

7 Support Care Cancer (2010) 18: Loosely defined precachexia (<10% weight loss); 1.0% weight loss could constitute precachexia 3. Only three symptoms were assessed. Our study had several limitations. It was a post hoc analysis of a previous study. Cross-sectional weight loss was defined as >10% decrease of pre-illness weight, so perhaps we missed the impact of less severe weight loss. The study was not longitudinal. Our findings apply only to those with advanced disease. This is the first time, to our knowledge, that CACS clinical characteristic and presentations have been described based on objective clinical data. The clinical relevance of these findings is that it helps characterize this syndrome, providing more knowledge to gage it. They also point the way for future research and therapeutic directions. We believe they have clinical validity as they closely resemble older anecdotal clinical findings of CACS. A number of questionnaires have been used to assess for CACS or its effects: The Edmonton Symptom Assessment System, Functional Assessment of Anorexia/Cachexia Therapy (FAACT; anorexia/cachexia), and QLQ-C30 [21 23]. Of these, only the FAACT is validated. Unfortunately, none include all the symptoms that characterized CACS according to our analysis. A comprehensive validated tool is still needed to monitor CACS course and response to treatment. CACS seems to be a distinct entity different from the groups with weight loss alone or anorexia alone. This was validated by comparison of the symptom profiles of the four groups and demonstrated by the independent effect of CACS on symptom burden. In addition, we described its clinical characteristics. This could be invaluable in CACS identification, treatment, and assessment. It can also guide the development of a more comprehensive CACS questionnaire. Our findings suggest that CACS is a distinct syndrome, however they confirm and further validate. Conclusions The cancer anorexia-cachexia syndrome appeared to be a distinct disorder with unique clinical characteristics in our advanced cancer population. In addition to anorexia and weight loss, nine other symptoms (constipation, dysphagia, early satiety, fatigue, lack of energy, nausea, taste changes, vomiting, and weakness) appear to constitute CACS. The presence of CACS as defined by anorexia and severe weight loss independently predicted higher symptom burden; weight loss alone or anorexia alone did not predict symptom burden. The absence of CACS predicted significantly longer survival. More research is needed to distinguish this syndrome. A comprehensive validated CACS assessment instrument is required. References 1. Accessed 07/ Loprinzi CL, Sloan JA, Rowland KM (2003) Methodologic issues regarding cancer anorexia/cachexia trials. In: Portenoy RK, Bruera E (eds) Issues in palliative care research. Oxford University Press, Inc, New York, pp Bennani_Baiti N, Walsh D (2009) What is the cancer anorexiacachexia syndrome? J R Coll Physicians Edinb 39: Grosvenor M, Bulcavage L, Chlebowski RT (1989) Symptoms potentially influencing weight loss in a cancer population. Correlations with primary site, nutritional status, and chemotherapy administration. Cancer 63(2): Strasser F, Bruera E (2002) Cancer anorexia cachexia syndrome: epidemiology, pathogenesis and assessment. In: Ripamonti C, Bruera E (eds) Gastrointestinal symptoms in advanced cancer. Oxford University, New York, pp Fearon KC, Voss AC, Hustead DS, Cancer Cachexia Study Group (2006) Definition of cancer cachexia: effect of weight loss, reduced food intake, and systemic inflammation on functional status and prognosis. Am J Clin Nutr 83: Fouladiun M, Korner U, Bosaeus I, Daneryd P, Hyltander A, Lundholm KG (2005) Body composition and time course changes in regional distribution of fat and lean tissue in unselected cancer patients on palliative caredcorrelations with food intake, metabolism, exercise capacity, and hormones. Cancer 103(10): Perboni S, Mantvovani G, Inui A, Takimoto Y (2005) Anorexia: central and peripheral mechanisms. In: Hofbauer KG, Anker S, Inui A, Nicholson JR (eds) Pharmacotherapy of Cachexia. CRC Press, Boca Raton, pp Walsh D, Donnelly S, Rybicki L (2000) The symptoms of advanced cancer: relationship to age, gender, and performance status in 1, 000 patients. Support Care Cancer 8(3): Blackburn GL, Bistrian BR, Maini BS, Schlamm HT, Smith MF (1977) Nutritional and metabolic assessment of the hospitalized patient. J Parenter Enteral Nutr 1: Bozzetti F, Mariani L (2009) Defining and classifying cancer cachexia: a proposal by the SCRINIO Working Group. JPEN J Parenter Enteral Nutr 33(4): Giacosa A, Frascio F, Sukkar SG et al (1997) Changes of nutritional and psychological status after megestrol acetate treatment of cancer cachexia. Rivista Italian di Nutrizione Parenterale ed Enterale 15: Feliu J, González Barón M, Berrocal A et al (1992) Usefulness of megestrol acetate in cancer cachexia and anorexia. A placebo controlled study. Am J Clin Oncol 15: Fietkau R, Riepl M, Kettner H et al (1997) Supportive use of megestrol acetate in patients with head and neck cancer during radio (chemo)therapy. Eur J Cancer 33: Bruera E, Neumann CM, Pituskin E, Calder K, Ball G, Hanson J (1999) Thalidomide in patients with cachexia due to terminal cancer: preliminary report. Ann Oncol 10(7): Rowland KM Jr, Loprinzi CL, Shaw EG et al (1996) Randomized double blind placebo controlled trial of cisplatin and etoposide plus megestrol acetate/placebo in extensive stage small cell lung cancer: A North Central Cancer Treatment Group Study. J Clin Oncol 14: Loprinzi CL, Bernath AM, Schaid DJ et al (1994) Phase III evaluation of 4 doses of megestrol acetate for patients with cancer anorexia and/or cachexia. Oncology 51(Suppl 1): Dewys WD, Begg C, Lavin PT et al (1980) Prognostic effect of weight loss prior to chemotherapy in cancer patients. Eastern Cooperative Oncology Group. Am J Med 69:

8 272 Support Care Cancer (2010) 18: Glare P, Sinclair C, Downing M, Stone P, Maltoni M, Vigano A (2008) Predicting survival in patients with advanced disease. Eur J Cancer 44(8): Evans WJ, Morley JE, Argilés J, Bales C, Baracos V et al (2008) Cachexia: a new definition. Clin Nutr 27(6): Ribaudo JM, Cella D, Hahn EA, Lloyd SR, Tchekmedyian NS, Von Roenn J, Leslie WT (2000) Re-validation and shortening of the Functional Assessment of Anorexia/Cachexia Therapy (FAACT) questionnaire. Qual Life Res 9(10): Bruera E, Kuehn N, Miller MJ, Selmser P, Macmillan K (1991) The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J Palliat Care 7(2): Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ, Filiberti A, Flechtner H, Fleishman SB, de Haes JC et al (1993) The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 3 85(5):