Comparison of the Freiburg and Charlson Comorbidity Indices in Predicting Overall Survival in Elderly Patients with Newly Diagnosed Multiple Myeloma

Multiple myeloma occurs primarily in elderly patients. Considering the high prevalence of comorbidities, comorbidity is an important issue for the management of myeloma. However, the impact of comorbidity on clinical outcomes has not been fully investigated. We retrospectively analyzed patients with newly diagnosed myeloma. Comorbidities were assessed based on the Charlson comorbidity index (CCI) and the Freiburg comorbidity index (FCI). The CCI is a summary measure of 19 comorbid conditions. FCI is determined by performance status, renal impairment, and lung disease. This study included 127 patients with a median age of 71 years. Approximately half of the patients had additional disorders at the time of diagnosis, and diabetes mellitus was the most frequent diagnosis (18.9%). The most significant factors for prognosis among patient-related conditions were a history of solid cancer and performance status (ECOG ≥ 2). The FCI score was divided into 3 groups (0, 1, and 2-3), and the CCI score was divided into 2 groups (2-3 and ≥4). FCI was a strong prognostic tool for OS (P > 0.001) and predicted clinical outcome better than CCI (P = 0.059). In conclusion, FCI was more useful than CCI in predicting overall survival in elderly patients with myeloma.


Introduction
Multiple myeloma (MM) is a hematologic malignancy of plasma cells that results in bone destruction, marrow failure, and renal impairment. The median age at the time of diagnosis is 70 years, with 36% of patients younger than 65 years, 27% aged 65 to 74 years, and 37% older than 75 years [1]. Considering the increasing life expectancy of the general population, the number of geriatric patients affected by MM is expected to increase over time.
Risk stratification of myeloma using the international staging system (ISS) and host factors such as age, performance status, and comorbidities are thought to be important for determining prognosis and choosing treatment options [2][3][4]. In 2011, Palumbo et al. suggested that appropriate screening for age (>75 years) and vulnerability, in addition to assessment of cardiac, pulmonary, renal, hepatic, and neurological functions, at the start of therapy allows treatment strategies to be individualized and drug doses to be tailored to improve tolerability and optimize efficacy [5]. Their study emphasized that elderly MM patients are more susceptible to side effects and are often unable to tolerate full drug doses. For these patients, lower-dose intensity regimens improve the safety profile and thus optimize the treatment outcome. However, this study was not based specifically on MM because there were few data on the impact of vulnerability on outcomes in MM patients [5]. Nonetheless, in general, several studies have reported problems related to comorbidity and cancer treatment in elderly patients [6][7][8][9][10].
Kleber et al. developed the Freiburg comorbidity index (FCI) to assess patient-related conditions as a risk factor for MM. The FCI is composed of three comorbidity factors: renal impairment, moderate to severe lung disease, and performance status. Interestingly, the FCI showed strong clinical relevance for overall survival (OS) and progression-free survival (PFS). Moreover, compared with other comorbidity indices, such as the Charlson comorbidity index (CCI), hematopoietic cell transplantation-specific comorbidity index (HCT-CI), Kaplan-Feinstein (KF), and Satariano index (SI), FCI was better able to stratify risk in patients with MM [11].
Although CCI is a widely used tool for assessing comorbidity in malignancy, this comorbidity index is complicated and difficult to apply. Moreover, there is no proven cut-off value that divides patients into low-or high-risk groups. As a result, several studies have determined their own cutoff values [11][12][13][14][15][16]. For example, Offidani et al. suggested a vulnerability score consisting of performance status and comorbidity score of CCI 0 or ≥1 [15].
Because of the increased incidence of multiple myeloma with aging and the fact that elderly patients have more comorbidity than younger patients, in the present study, we assessed comorbidities at diagnosis, the impact of host factors on OS, and compared CCI and FCI as prognostic factors in newly diagnosed elderly patients with MM.

Study Design.
This study was a retrospective, singlecenter case series. OS was calculated as the time from diagnosis to death from any cause. Adverse events were graded according to the National Cancer Institute Common Toxicity Criteria (NCI-CTC) version 4.0. This study was reviewed and approved by the Institutional Review Board.

Patients.
A total of 127 consecutive patients aged 65 years and older who were newly diagnosed with symptomatic MM at the Samsung Medical Center, Seoul, South Korea, between January 1, 1999, and June 30, 2011, met the inclusion criteria for this study. We excluded patients with amyloidosis, those who were suitable for autologous stem cell transplantation (ASCT), and those who were lost to follow-up within 6 months from the time of diagnosis due to any cause except death. The last follow-up date was March 31, 2013. (Table 1) is a summary measure of 19 comorbid conditions that are each weighted from 1 to 6 based on disease severity. This measure provides a total score ranging from 0 to 37 [17]. Information on comorbidity was extracted from a detailed review of each patient's medical records and laboratory values at the time of diagnosis. In the original study, MM was included in the classification of lymphoma for convenience. In this study, lymphoma was defined, as in the original study, except that the definition did not include myeloma.

Charlson Comorbidity Index (CCI). The CCI
In addition, according to the original study, each decade of age over 40 would add 1 point to the risk value (i.e., 50 years = +1 point), and the age point would be added to the score of the comorbidity index. In this study, we used the method described by Kleber et al., which adds the age point to the CCI score [11].

Freiburg Comorbidity Index (FCI).
Renal impairment is defined as estimated glomerular filtration rate (eGFR) ≤30 mL/min/1.73 m 2 , based on the modification of diet in renal disease (MDRD) study equation [18]. Poor performance status is based on a Karnofsky performance status (KPS) score ≤70. Moderate or severe lung disease is defined in the same manner as in the CCI [11]. Each of the variables contributes 1 point and the FCI is a summation of these points; thus, the FCI value ranges from 0 to 3.

Statistical
Analysis. Numerical variables are summarized by median and range and categorical variables are described by count and relative frequency (%) of subjects in each category. Comparison of the distribution of categorical variables in the different groups was performed with either Fisher's exact test or the 2 test. OS was estimated using the Kaplan-Meier methodology. Uni-and multivariate Cox regression analyses were applied to assess factors affecting OS. Components with a P value less than 0.05 in univariate analysis were included in the subsequent multivariate analysis.
These analyses were performed using PASW statistics 18.0.0 (WinWrap, IBM, New York, USA). Null hypotheses of no difference were rejected if P values were less than 0.05.

Patient Characteristics.
During the study period, a total of 159 patients aged 65 years or older were newly diagnosed with symptomatic MM. Among them, 22 patients were excluded because of a combined diagnosis of amyloidosis, early followup loss, or ASCT. Therefore, data from 127 patients were included in the analysis. Table 2 shows baseline characteristics of patients at the time of diagnosis. The median overall survival of all patients was 34.1 months, and the median follow-up duration for the surviving patients was 46 months. The median age of the patients was 71 years and 26.8% of the patients were aged 75 years or older. Performance status was evaluated by ECOG status. In FCI, performance status was assessed by KPS, and, by definition, ECOG grade 2 is interchangeable with KPS grade 70 [19][20][21].
The prevalence of comorbidity at the time of diagnosis was 48.8% (Table 3). The most frequent comorbid condition was diabetes without end organ damage ( = 24, 18.9%). The median CCI score, including age points, was 3 (range: 2 to 13). The CCI score was divided into two groups based on the median score; CCI scores of 2-3 were classified as the low CCI score group, and CCI scores ≥4 were classified as the high CCI score group. In this study, 53 and 74 patients belonged to the low and high CCI score groups, respectively. When using the FCI classification, 59, 54, and 12 patients belonged to the 0, 1, and 2-3 score groups, respectively.

Host Factors and Comorbidity Indices as Prognostic
Factors. We analyzed the impact of host factors, such as age, sex, performance status, and each of the comorbidities on OS (Table 4). The results showed that all of the factors constituting FCI, such as performance status, chronic lung disease, and eGFR, were significant for OS. In addition, any tumor, metastatic solid tumor, cerebrovascular disease, and ISS each had statistical significance. When multivariate analysis was conducted with these factors, only performance status and previous cancer history (regardless of metastasis) remained significant. Figures 1, 2, 3, and 4 show Kaplan-Meier survival curves for both comorbidity indices. FCI effectively predicted the OS of the three different groups (P < 0.001). The median survival times were 55.0 months, 29.5 months, and 19.5 months for FCI scores of 0, 1, and 2-3, respectively. Although the P value was not statistically significant, CCI also distinguished between the two score groups for OS (44.8 months versus 34.7 months, P = 0.059). However, OS based on CCI without age points did not demonstrate clinical relevance (P = 0.147). In contrast, FCI was significant in subgroup analysis for age groups (65-74 and ≥75 years; P < 0.001 and 0.04, resp.)

Serious Adverse Events.
We defined serious adverse events as grade ≥4 for hematologic adverse events and grade ≥3 for nonhematologic adverse events, according to NCI-CTC version 4.0. The most frequent serious adverse event was infection ( = 35, 30.0%) followed by neutropenia and anemia ( = 15, 12.7% for both). Grade ≥3 nonhematologic adverse events occurred in 50% of patients, whereas grade ≥4 hematologic adverse events occurred in 22.9% of patients.  Grade 5 adverse events due to any cause occurred in 6.8% of the patients, as shown in Table 5.

Discussion
This study assessed comorbidities at diagnosis of MM, the impact of host factors on overall survival, and compared CCI and FCI as prognostic factors in newly diagnosed elderly patients.
Univariate analysis revealed that performance status, ISS, and several comorbid conditions such as chronic lung disease, azotemia (eGFR < 30 mL/min/1.73 m 2 ), presence of any tumor, metastatic solid tumor, and cerebrovascular disease were significant factors. However, azotemia as defined by CCI (serum creatinine ≥ 3 mg/mL) was not a prognostic factor. In multivariate analysis, azotemia, as defined by impaired eGFR or chronic lung disease, was not shown to be a significant risk factor in our study. In contrast, a history of cancer, regardless of whether metastasis occurred, was the strongest prognostic factor for elderly patients with myeloma. Unfortunately, use of novel agents over conventional drugs did not significantly improve OS, although this might reflect the relatively short period of use of novel agents. Although two components of FCI-renal impairment and moderate or severe lung disease failed to demonstrate significance in multivariate analysis, when we compared both comorbidity indices and overall survival, the FCI showed a greater ability to separate OS among the three score groups (P < 0.001). The CCI score including age points was not statistically significant but was still valuable and superior to the CCI without age points. As briefly mentioned above, the CCI score without age points did not discriminate for OS.
FCI provides a clear definition of each component and all three components were statistically significant, at least in univariate analysis. In contrast, CCI is more subjective and only 4 among 19 conditions were significant. Most importantly, at the present time, CCI does not have any standard cut-off value. Various studies have divided CCI scores into groups of 0, 1-2, and ≥3; 0, 1, and ≥2; or 0 and ≥1. Some studies included an age point, but others did not [11][12][13][14]. These variations might explain why FCI is more predictable than CCI. Moreover, FCI is also very simple to apply. FCI consists of performance status, moderate or severe lung disease, and azotemia, and each of these factors is worth 1 point. The FCI score is, therefore, a simple summation of these three factors. In contrast, CCI consists of 19 comorbid conditions, and, within the same disease, scores are weighted based on severity ranging from 1 to 6 points. In addition, an age point is calculated and added to the CCI score. Comorbidity definitions frequently use a symptomatic grade.
In this study, all 118 patients who were treated with chemotherapy received a full dose of chemotherapeutic agents as scheduled. Interestingly, the profile for serious adverse events showed that treatment was relatively safe and adverse events were easily controllable. In fact, since the approval of various novel agents, clinical outcomes such as survival and toxicity profiles have improved in transplant-ineligible elderly patients with multiple myeloma [22][23][24][25][26][27]. Thus, we propose that the full dose of chemotherapy might be tolerated, regardless of the presence of comorbid conditions, even though this is against the recommendation that chemotherapy dose reduction is required for patients 75 years or older or those with cardiac, pulmonary, hepatic, renal, or neurologic dysfunctions [5]. Furthermore, this finding might be important evidence for preventing chemotherapy dose reduction because of physician bias.
There are some limitations in this study. First, the followup duration was short and the sample size was small. Second, this is a retrospective single center study. Third, there were no patients with peripheral vascular disease, dementia, DM with end organ damage, or AIDS. Despite these limitations, this study successfully applied the FCI and the CCI to newly

Conclusions
In this study, approximately 50% of elderly patients with newly diagnosed multiple myeloma had at least one comorbid disease at the time of diagnosis. Among host factors tested, performance status and a history of malignancy were the most important prognostic factors. The Freiburg comorbidity index is very simple to use and predicts overall survival better than the Charlson comorbidity index.