Introduction

Cachexia is a state of involuntary weight loss that was described centuries ago.1, 2 It is estimated that 5 million people in the United States of America develop cachexia annually,3 with 2 million deaths relating to it.4 Such high prevalence of cachexia, as well as its adverse relation to mortality and therapeutic modalities,5, 6 emphasizes the need for its exact definition.7

Older definitions defined cachexia as a loss of a certain proportion of the body weight in a variable time frame. Newer definition, which was proposed recently, set the loss of 5% of edema—free body weight during the previous 12 months or less as a ‘major criterion’. When a patient meets this criterion he/she needs to fulfill at least three out of five ‘minor criteria’ to be diagnosed with cachexia. Those minor criteria are as follows: decreased muscle strength, anorexia, fatigue, low fat-free index and abnormal biochemistry (C-reactive protein (CRP) >5 mg/l or interleukin-6 >4.0 pg/ml, hemoglobin <12 g/dl and albumin <3.2 g/dl).8 This definition was the last published generic definition, that is, designed to be applicable for all the diseases related to cachexia. More recently, the view of cachexia changed shifting its focus toward condition-specific definitions. New definition of cancer cachexia was published. It also incorporates a principal criterion of weight loss and most of ‘minor criteria’ from the generic definition are regarded as a part of the assessment of patients.9

Principal aim of the present study was to investigate the influence of ‘minor criteria’ on the prevalence of cachexia using the most recent generic definition. We gathered clinical data from 137 consecutive patients with malignant diseases or congestive heart failure hospitalized in a single institution. First, we obtained the prevalence of cachexia using just ‘major criterion’ as a sole diagnostic criterion. We then compared it with the prevalence obtained by using the most recent and complete generic definition, that is, incorporating also ‘minor criteria’. Secondary aim of the study was to compare clinical, laboratory and anthropometric characteristics of patients meeting both major and minor criteria with patients that meet only major criterion—if such were found to exist. Finally, we analyzed survival of patients in respect with fulfillment of additional criteria.

Patients and methods

Patients

The study was conducted at the Department of Internal Medicine of University Hospital Merkur in Zagreb. Data were collected from May 2011 until May 2012 in consecutive patients that were eligible for entry. A diagnosis of solid hematological tumors (non-Hodgkin lymphoma, Hodgkin disease), other solid tumors of any site or chronic heart failure of any etiology was mandatory. Diagnosis of malignant disease had to be proven with adequate histopathological sample. Diagnosis of heart failure was made according to the guidelines criteria.10

Exclusion criteria were as follows: age of patients <18 years, starvation defined as deliberate or unintentional reduced food consumption despite preserved appetite, malabsorption defined as documented or suspected disease related to malabsorption, diarrhea defined as having three or more loose or liquid stools per day or as having more stools than is normal for a given patient, active thyroid disease, depression or other severe psychiatric disease, chronic obstructive pulmonary disease, chronic kidney disease stage ⩾3, myocardial infarction in less than last 12 weeks, liver insufficiency defined by clinical or laboratory signs of its synthetic or metabolic dysfunction or a documented cirrhosis, neuromuscular diseases and alcohol or drug abuse.

The study received previous local ethical board approval and was conducted according to the Declaration of Helsinki and its subsequent amendments. All patients gave written informed consent before participation in the study.

Evaluation of patients

Complete history and full physical examinations were performed at baseline. Baseline blood analyses included complete blood count and complete biochemistry (including values of hemoglobin, CRP and albumin). All measurements were done according to the protocol of institutional laboratory.

During clinical workup, special attention was given to the minor criteria proposed by the most recent generic definition.8 Muscle strength was obtained using handgrip dynamometer. Three consecutive measurements at least 5 s apart were done. Highest of the measurements was used for analysis. Decreased muscle strength was defined as a result in the lowest tertile for the age and gender.8, 11 Fatigue and anorexia were defined according to the definitions provided by the aforementioned consensus document. Fatigue was defined as physical and/or mental weariness resulting from exertion, an inability to continue exercise at the same intensity with a resultant deterioration in performance. Anorexia was defined as a poor appetite.8 Low fat-free mass was defined as a mid upper arm circumference <10th percentile for age and gender.8, 12, 13 Measurements of skinfolds (biceps, triceps, subscapular and suprailiac) and circumferences (mid-arm, waist and hip) were done using previously published protocols.14 All measurements were done three times at each site. Calculated mean value was used for analysis. Body mass index (BMI) was calculated as weight in kilograms divided with squared height in meters. Survival analyses were done using a National Registry of Deceased Persons. Through that Registry, we were able to gather information about mortality, as well as information about dates of deaths, in our population of patients up to 31 December 2012.

Statistics

The results were expressed as the mean±s.d. or as a proportion of the total number. Differences in proportions (categorical variables) were compared using χ2-test. Mann–Whitney test was used to test the equality of continuous variables. A P-value <0.05 was considered statistically significant. All statistics were performed with the StatView statistical program, version 5.0.1. (SAS Institute, Cary, NC, USA).

Results

Baseline clinical characteristics of patients

One hundred thirty-seven (86 males; mean age 63.4 years) consecutive patients were enrolled. Fifty-four patients were diagnosed with solid hematological tumor (44 had diagnosis of non-Hodgkin lymphoma and 10 had diagnosis of Hodgkin disease). Fourty-one patients had solid tumor of various sites (19 hepatocellular cancer, 8 colorectal cancer, 8 pancreatic cancer, 3 billiary duct cancer, 2 gastric cancer and 1 cancer of unknown primary site) and 42 patients had chronic heart failure (31 had ischemic cardiomyopathy, 6 had dilated cardiomyopathy and 5 had valvular cardiomyopathy). Out of 42 patients with heart failure, four (10%) were in New York Heart Association (NYHA) I stage, 16 (38%) were in NYHA II stage, 18 (42%) were in NYHA III stage and 4 (10%) were in NYHA IV stage.

Prevalence of cachexia

We identified 42 (30.6%) patients that lost >5% of body weight in 12 months or less in our group of patients. In that group, there were 13 patients with heart failure, 14 patients with solid hematological tumors and 15 patients with solid tumor of any site. When we applied new definition, that is, included three out of five minor criteria with loss of 5% of body weight, we identified only 30 (21.8%) patients in our group. In that group, we indentified 10 patients with heart failure, 9 with hematological disease and 11 with solid tumor. Statistical analysis of data showed that applying minor criteria lead to significant decrease in prevalence of cachexia (P=0.0006). Results are shown in Figure 1. We found no significant influence of diagnoses (heart failure vs hematological disease vs solid tumor) on the distribution of patients in the whole group (n=42) with loss of body weight (P=0.42). There was also no such influence in the group (n=30) that met three out of five additional criteria (P=0.75).

Figure 1
figure 1

Prevalence of cachexia. Data showing prevalence, expressed as number, with or without using minor criteria from the new definition.

Full size image

Analysis of clinical, laboratory and anthropometric data

Further evaluation was focused on the subgroup of patients that lost >5% of body weight in 12 months or less. Possible differences between patients (n=30) that met additional three out of five criteria from the new definition and patients that did not meet them (n=12) were sought. Analysis of body weight, BMI, amount of body weight loss, period of body weight loss and intensity (expressed as kilograms/month) of body weight loss did not show any significant differences between two groups. Data are shown in Table 1. Analysis of laboratory parameters included in new definition showed that patients that met three out of five minor criteria had significantly higher levels of CRP (36.7±53.8 vs 16.6±36.3 mg/l; P=0.03) and significantly lower levels of albumin (3.34±0.58 vs 3.89±0.47 g/dl; P=0.007). Data are shown in Table 2. Analysis of anthropometric data showed significantly lower measurements for mid-arm circumference (25.52±4.33 vs 28.67±2.64 cm; P=0.02), triceps skinfold (0.87±0.36 vs 1.25±0.44 cm; P=0.02) and suprailiac skinfold (0.87±0.46 vs 1.19±0.31 cm; P=0.02) in patients that met the above mentioned criteria. Data are shown in Table 3.

Table 1 Body weight, body mass index, amount of weight loss, period of weight loss and intensity of weight loss for patients with 5% of body weight loss
Full size table
Table 2 Values of C-reactive protein, albumin and hemoglobin for patients with 5% of body weight loss
Full size table
Table 3 Values of anthropometric measurements for patients with 5% of body weight loss
Full size table

Survival analysis

Mortality data were gathered through National Registry as described in Patients and methods section. Median time of follow-up until December 31st was 394.0±135.9 days. We observed 14 (47%) deaths among patients fulfilling additional criteria. In the group of patients that did not meet three out of five additional criteria, we observed 5 (42%) deaths. This difference was not statistically significant (P=0.76). Kaplan–Meier analysis of survival also did not reveal any significant differences (P=0.11) between the groups (Figure 2).

Figure 2
figure 2

Kaplan–Meier analysis of survival of patients with the respect of fulfillment of three out of five additional criteria. w/o, without.

Full size image

Discussion

Weight loss is an independent predictor of mortality both in patients with cancer, as well as in patients with chronic heart failure.15, 16 It is also a powerful predictor of adverse events during cancer chemotherapy5 and surgical procedures.6 Because of it, weight loss was used as a sole criterion for diagnosis of cachexia. On the other hand, experimental data led some authors to claim that ‘weight loss alone does not identify the full effect of cachexia on physical function and is not a prognostic variable’.17 This disagreement was a foundation to a consensus document that brought last generic definition of cachexia. Cachexia is now considered as a complex metabolic syndrome, incorporating both weight loss and other prominent clinical and laboratory features. Weight loss is considered to be a major, but not the sole criterion for diagnosis. Additional, so-called minor, criteria were introduced. When at least three out of five minor criteria, together with weight loss, are met, the patient can be diagnosed with cachexia.8 More recently, an international consensus document that defines cancer cachexia was published. It also incorporates weight loss or a reduced BMI as a principal criterion. Most of the minor criteria from the last generic definition are now found in the assessment part of this new definition and are not regarded obligatory for diagnosis.9 It also introduces an exciting concept of precachexia, which was defined in earlier work,18 as well as refractory cachexia, in that way trying to stage cachexia,9 highlighting the need for earlier treatment.19 This staging concept led to some interesting clinical work that tried testing the concept and proved some of its benefits.20 Our working group also strongly believes in cachexia being a disease-specific condition.21 Yet, our clinical practice consists of patients with various diseases related to cachexia. This makes generic definitions attractive and more practical to us. Cancer-specific definition of cachexia still incorporates weight loss as a major criterion together with most of minor criteria from the generic definition. Those minor criteria are not regarded obligatory, by this condition-specific definition, but their evaluation is strongly encouraged.9 On the other hand, there are no condition-specific definitions for diseases other than cancer. It all led us to evaluate the generic definition,8 that is, the influence of its minor criteria on the prevalence of cachexia. Given the relatively small number of patients in our study, and given the fact that we used a generic definition, we resisted the temptation of dividing the patients into any other subgroups. Even dividing patients into cancer and non-cancer ones would give us relatively unbalanced groups because of the diversity of diagnoses in the cancer group. In that way, we would jeopardize the principal aim of the study, which was to evaluate the value of the generic definition on any given patient that is admitted to our institution. The principal aim of the study was not to see the differences between cancer and cardiac patients. Still, we are very well aware that different influences of different diseases may have somewhat confounded the data. Possible differences could be related to different pathophysiological mechanisms, such as production of lipid mobilizing factor or proteolysis inducing factor found exclusively in cancer patients.22

Underlying mechanisms of cachexia are not well understood. Yet, conceptual representation of the new definition recognizes most important pathophysiological processes involved. Those processes are responsible for weight loss. On the other hand, they are also involved in other clinical features that are now recognized as minor criteria.8 It is therefore possible that most of the patients, if not all, that meet major criterion would also meet minor criteria from the definition. If this was to be true, those minor criteria would in fact be redundant. Some of the minor criteria, such as anorexia and fatigue, are easily assessed. On the other hand, assessing muscle strength and fat-free mass index is less easy. Actually, if one does not have adequate medical equipment or specific percentile tables, assessing those two are impossible. Possible redundancy as well as obvious difficulties in assessing some of the minor criteria was a principal reason for our study. We gathered data from 137 consecutive patients with malignant diseases or chronic heart failure. Owing to organizational/structural issues in our outpatient clinic during the time frame of our interest, we gathered all data from hospitalized patients. In this group of patients, we identified 42 patients (30.6%) that had lost at least 5% of body weight in last 12 months or less. Only 30 patients (21.8%) met additional three out of five minor criteria from the new definition. This difference was found to be statistically significant. Such a result led us to conclude that applying minor criteria to the patients reduces the prevalence of cachexia.

What does this reduced prevalence mean? There are several possible explanations. Such a reduced prevalence enables us to focus on patients with worse prognosis. Looking at the data, which show relationship between the presence of some of the minor criteria and adverse prognosis,23, 24 this conclusion seems logical. Applying additional criteria, as set by the definition, recognizes patients that need our full attention and prompt institution of corrective measures. On the other hand, most of the studies investigated the impact of single phenomenon, that is, just reduced muscle strength or just enhanced inflammatory response. Yet, the new definition states that a patient needs at least three of those criterias. What about the patients that meet only one or two of those criteria? What if the criteria that they meet are the ones proven to be predictors of mortality? Such are previously mentioned reduced muscle strength or increased inflammatory response.23, 24 If this is to be true, and it seems probable, such a reduced prevalence would in fact mean that we are, by applying new definition, missing some of the high-risk patients. If we would not consider them cachectic, we would probably not treat them for cachexia. Low success rate of current cachexia treatment modalities8 makes this problem smaller. Yet, if future therapeutic options prove to be better, this problem will indeed rise. In the end, there will be some patients that lost weight but that do not meet any of the minor criteria. Excluding them would also reduce prevalence of cachexia. Looking at the huge amount of experimental data, that showed body weight loss to be independent predictor of mortality,15, 16 one should probably conclude that those patients are also at higher risk.

Aside from the questions raised, we could speculate that our results could contribute to the concept of staging of the cachexia syndrome.9 Using three out of five additional criteria as a cutoff could indeed provide us with two groups of patients within the cachexia range. Patients that meet those criteria could intuitively be regarded as high-risk patients. Unfortunately, being very well aware of its limitations, our survival analysis did not show any differences in the survival between the two groups.

Trying to further evaluate our population, we tried to look for some clinical or anthropometric differences between the two aforementioned groups within cachexia range. We found that the two groups did not differ in most of the investigated characteristics. Yet, we found some differences. Patients that met the criteria of new definition had higher levels of CRP and lower levels of albumin. They also had lower measurements of their mid-arm circumference, triceps and suprailiac skinfolds. Observed differences may imply some differences in underlying mechanism in the two groups. Known correlation of high levels of CRP and low levels of albumin, with worse prognosis of both cancer and cardiac patients,5, 25 could be a sign of higher risk of those patients. We observed reduction of mid-arm circumference, which correlates with fat-free mass, that is, skeletal muscle mass, in patients that met all the required additional criteria. This reduced skeletal muscle mass was found to be an independent prognostic factor,26 which could imply worse prognosis in that group of our patients. Again, our survival analysis did not support this notion.

Principal limitation of our study is a relatively low number of patients. We also did not investigate possible pathophysiological differences between the groups of patients that we identified. Taking it all together, our findings need larger studies with prognostic and pathophysiological insights to be confirmed.

Conclusions

Applying minor diagnostic criteria from the latest generic definition to the principal criterion of the loss of body weight reduces the prevalence of cachexia. This leads to identification of two populations of patients within the cachexia range. Some clinical and anthropometric differences between the two groups can be observed. Those differences could imply differences in underlying mechanisms, as well as differences in prognosis, between the groups. Our survival analysis did not find any differences between the two groups. Further studies with larger number of patients and more balanced groups are warranted to confirm our findings.