Проверка корейского инструмента гериатрической оценки у пожилых пациентов с множественной миеломой: многоцентральное исследование.

Introduction

Multiple myeloma (MM) is a plasma cell neoplasm that predominantly affects older adults [ 1 ]. As the global population continues to age rapidly, the proportion of older individuals with myeloma is projected to increase substantially in the coming years [ 2 ]. Recent advances in treatment have improved 5-year and 10-year survival rates [ 3 ]. While long-term survival rates for patients under 65 have increased approximately fourfold, there has been no significant improvement for those aged ≥ 75 [ 3 ]. Overall, MM-related deaths are highest among patients aged ≥ 75, and early mortality is most common in those aged ≥ 70 [ 4 ]. First-generation novel agents (bortezomib, thalidomide, and lenalidomide) have substantially reduced early toxic mortality (1%) compared to conventional chemotherapy (10%) [ 5 ]. However, the higher toxicity-related mortality in octogenarians underscores the need for thorough frailty assessments [ 5 ]. To prevent both under- and overtreatment, incorporating an objective frailty assessment into MM guidelines is essential.

A comprehensive geriatric assessment (CGA) is a global evaluation of the health of older individuals, encompassing comorbidities, functional status, cognition, polypharmacy and/or inappropriate medications, social support, and depression and/or psychological distress [ 6 ]. It can predict chemotherapy toxicity and survival in cancer patients [ 7 ]. However, CGA can be time-consuming and labor-intensive, and there is a lack of specific data on CGA in patients with hematologic malignancies. Various tools have been developed and validated to identify “frail” patients. The International Myeloma Working Group (IMWG) frailty score, which considers age, the Charlson Comorbidity Index (CCI), Activities of Daily Living (ADL), and Instrumental Activities of Daily Living (IADL), has been shown to predict survival outcomes, treatment discontinuation, and non-hematologic toxicities [ 8 ]. The revised Myeloma Comorbidity Index (R-MCI), which incorporates factors such as age, renal function, lung function, performance status, the timed up and go test, IADL, subjective fitness, and cytogenetics, effectively classifies patients into fit, intermediate-fit, and frail groups, each demonstrating markedly different survival outcomes [ 9 ]. Facon et al. [ 10 ] introduced a simplified frailty index based on age, the CCI, and Eastern Cooperative Oncology Group (ECOG) status, which is easy to replicate and can reliably predict clinical outcomes.

Frailty assessment tools developed in other countries may not accurately capture the cultural and lifestyle differences characteristic of Korean elderly populations. In a busy outpatient setting, healthcare providers need tools that are both accurate and simple to administer. Kim et al. [ 11 ] developed a geriatric screening tool, the Korean Cancer Study Group Geriatric Score-7 (KG-7), based on items with high sensitivity and specificity across each CGA domain. This tool is simple to apply in the treatment of elderly cancer patients and has had its utility validated [ 11 , 12 ]. In this study, we assessed the predictive role of KG-7 in newly diagnosed elderly MM patients to define a frailty score and evaluate its impact on clinical outcomes and toxicity.

Materials and Methods

1. Participants

This study is a multicenter, longitudinal, and prospective cohort study conducted by the Korean Multiple Myeloma Working Party. It included patients aged ≥ 70 years who had been diagnosed with MM according to standard criteria, underwent initial chemotherapy as determined by physicians, and had not undergone transplantation. The patients’ baseline characteristics obtained for analysis included ECOG performance status, creatinine, β2-microglobulin, albumin, lactate dehydrogenase (LDH) levels, and unfavorable cytogenetics defined as t(4;14), t(14;16), or del17p13. The International Staging System (ISS) and revised ISS were also assessed [ 13 , 14 ].

2. Frailty score assessment

Within 14 days prior to chemotherapy initiation, assessments of the CCI, ADL, IADL, and KG-7 were conducted. The KG-7 includes seven questions evenly distributed across each domain of the CGA ( S1 Table ). KG-7 scores range from 0 to 7, with higher scores indicating better conditions.

The IMWG frailty score, ranging from 0 to 5, was then calculated by combining the individual risk scores from these scales along with patient age, CCI, ADL, and IADL, according to previously established IMWG criteria [ 8 ]. Patients were classified into three categories: fit (0), intermediate-fit (1), and frail (≥ 2). The simplified frailty index, also ranging from 0 to 5, was computed by combining the individual risk scores of scales including age, CCI, and ECOG performance status [ 10 ]. The simplified frailty index uses only two categories: non-frail (0-1) and frail (≥ 2).

The KG-7 was tracked over 24 months to observe changes in frailty status over time. KG-7 scores were recalculated at designated intervals, specifically at 6, 12, 18, and 24 months after diagnosis.

3. Outcomes

The primary outcome was to assess whether the KG-7 index could predict overall survival (OS), defined as the time from diagnosis to death from any cause or to censoring. Key secondary outcomes included evaluating the KG-7 index’s ability to predict progression-free survival (PFS) defined as the time from diagnosis to progression, relapse, death from any cause, or censoring and to predict toxicity. Other secondary outcomes involved comparing the KG-7 algorithm with other conventional algorithms and examining the trajectory of the KG-7 score. Adverse events (AEs) were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events, ver. 4.0.

4. Statistical analysis

Baseline characteristics were compared using the χ 2 test or Fisher’s exact test, as appropriate. Kaplan-Meier estimates were calculated for survival, and the log-rank test was used to compare the curves. Hazard ratios (HRs) for survival were estimated using Cox regression. A Sankey diagram was generated to illustrate the proportion of patients in each frail subgroup at each landmark interval (6-, 12-, 18-, and 24-months post-diagnosis). A p-value < 0.05 was considered significant, and 95% confidence intervals (CIs) were reported. All analyses were performed using STATA version 16 (StataCorp.).

Results

1. Patient characteristics

Between August 2020 and January 2022, 100 MM patients undergoing first-line chemotherapy were enrolled at 15 hospitals affiliated with the Korean Multiple Myeloma Working Party. The demographic and baseline disease characteristics of the patients included in this study are presented in Table 1 . Their median age was 77 years (range, 70 to 88 years), 54.0% were male, and 19.0% had an ECOG performance status ≥ 2 ( Table 1 ). According to the ISS, 73.0% of the patients were in stage 2 or 3. Of the patients, 16% had an ADL score ≤ 4, 26% had an IADL score ≤ 5, and 14% had a CCI ≥ 2. The most common first-line regimen was VMP (bortezomib+melphalan+prednisolone) (48.0%), followed by Rd (lenalidomide+dexamethasone) (41.0%).

2. Frailty assessment

The distribution of the KG-7 score is shown in Fig. 1A . When patients were categorized with a cutoff of 5 points, classified as non-frail (score ≥ 5) and frail (score < 5), the KG-7 index correlated well with the IMWG frailty index and the simplified frailty index (p < 0.001) ( S2 Table ). The differences in proportions of the classified frailty groups based on the three different frailty assessment tools are shown in Fig. 1B . The distribution by IMWG frailty category was as follows: 30% fit, 29% intermediate-fit, and 41% frail. Using the simplified frailty index, 43% of patients were characterized as non-frail, and 57% as frail. Based on the 5-point cutoff of the KG-7 index for frailty, the proportion of patients classified as non-frail and frail was 69% and 31%, respectively.

3. Survival

After a median follow-up of 26.8 months (95% CI, 25.4 to 29.6), 20 patients (20%) had died, and the median survival was not reached. The 3-year OS and PFS rates were 73.0% (95% CI, 62.6 to 85.2) and 31.9% (95% CI, 20.5 to 49.8), respectively ( Fig. 2 ). Fig. 3 illustrates the OS for the cohort, as determined by various frailty assessment instruments. When using the IMWG frailty index, there was no statistically significant association with the risk of death by frailty category (in reference to a status of fit: HR, 2.44 for intermediate-fit; 95% CI, 0.63 to 9.44; HR, 2.84 for frail; 95% CI, 0.78 to 10.4; global p=0.251; Harrell’s C=0.599). When using the simplified frailty index, there was no statistically significant association with the risk of death by frailty category (in reference to a status of non-frail: HR, 2.11 for frail; 95% CI, 0.81 to 5.50; p=0.119; Harrell’s C=0.598). When using the KG-7 index, there was no statistically significant association with the risk of death by frailty category (in reference to a status of non-frail: HR, 1.86 for frail; 95% CI, 0.76 to 4.56; p=0.171; Harrell’s C=0.578). Although not shown in the data, the frailty indices did not predict PFS.

4. Univariate and multivariate analyses of risk factors for survival

Table 2 presents a univariate analysis predicting survival. According to this table, no factors were statistically significant in predicting OS. However, there was a tendency for poorer survival associated with several factors: ECOG performance status ≥ 2 (HR, 2.10; 95% CI, 0.80 to 5.46; p=0.130), creatinine ≥ 2 mg/dL (HR, 2.16; 95% CI, 0.72 to 6.50; p=0.169), albumin < 3.5 g/dL (HR, 2.28; 95% CI, 0.93 to 5.58; p=0.072), increased LDH (HR, 2.56; 95% CI, 0.97 to 6.78; p=0.058), ADL ≤ 4 (HR, 2.24; 95% CI, 0.81 to 6.18; p=0.121), CCI ≥ 2 (HR, 2.36; 95% CI, 0.94 to 5.94; p=0.067), and KG-7 < 5 (HR, 1.86; 95% CI, 0.76 to 4.56; p=0.177). In the multivariate analysis ( Table 3 ), the CCI was the only factor that statistically significantly predicted poor survival outcomes (HR, 2.94; 95% CI, 1.13 to 7.64; p=0.027).

5. Toxicity

Among three frailty indices, the simplified frailty index (HR, 2.49; 95% CI, 1.09 to 5.95; p=0.030) and the KG-7 index (HR, 2.43; 95% CI, 1.03 to 5.86; p=0.043) were associated with incident grade 3-4 non-hematologic toxicity during the first line of chemotherapy ( Table 4 ). However, none of the three frailty indices showed a statistical association with grade 3-4 hematologic toxicities during the first line of chemotherapy. Patients with grade 3-4 non-hematologic toxicity exhibited statistically significantly higher drug discontinuation rates (p=0.016) and lower OS (p=0.037), but these factors were not associated with the best response to first-line chemotherapy (p=0.307) or lower PFS (p=0.357).

6. Serial follow-up of KG-7

Within the cohort, 43 patients did not complete the KG-7 follow-up to 24 months for various reasons ( S3 Fig. ). The primary reason for dropout was patient death (n=16), with specific causes outlined as follows: infection (n=8), disease progression (n=6), hyperkalemia (n=1), and cardiac arrest (n=1). Additionally, 10 patients were lost to follow-up, and six patients were transferred to other facilities, including three to a tertiary hospital and three to hospice care. Six patients dropped out due to individual institutional circumstances. Three patients withdrew from the study at their own request. Lastly, two patients were diagnosed with additional hematologic malignancies: one with myelodysplastic syndrome and another with acute myeloid leukemia.

The trajectories of frailty from baseline over the next 2 years are shown in Fig. 4 . The frailty of the elderly either improved or worsened over time. When the KG-7 was used to monitor 90 patients who had at least one follow-up, 32 individuals (35.6%) showed changes in frailty, while 58 (64.4%) remained unchanged. Considering that death (37.0%) was the most common reason for dropping out, it is likely that changes in frailty were more frequent. Six months after the initial follow-up, compared to the diagnosis, 7% of the patients showed an improvement in frailty status, 15% experienced deterioration, 68% remained the same, and 10% dropped out of the study. At the 12-month follow-up, 8% of the patients showed an improvement in frailty status, 8% experienced deterioration, 59% remained the same, and 25% dropped out of the study. At the 24-month follow-up, 4% of the patients showed an improvement in frailty status, 12% experienced deterioration, 43% remained the same, and 43% dropped out of the study.

Discussion

Although advancements in novel therapeutics are gradually improving survival rates for MM patients, securing better long-term outcomes, especially for the elderly, continues to pose a formidable challenge. Survival disparities among older adults with MM are multifactorial; factors that play a role include differences in treatment allocation, therapy toxicity, drug discontinuation, and the physiologic reserve or fitness of the patient [ 15 ].

Age continues to be a key factor in oncologists’ decision-making [ 16 ]. Given that age is a component of the IMWG frailty score, and that patients older than 80 are automatically categorized as frail at best, it is unclear how effective the IMWG score is for assessing clinical outcomes in an older population [ 8 ]. According to a study by Murillo et al. [ 17 ], the inability to statistically significantly predict survival when applying the IMWG frailty index to elderly patients over 75 with MM suggests that age alone has limitations in assessing frailty in the elderly.

The performance status of patients is one of the most important factors in cancer care, and the ECOG performance status is the most used scale. Performance status has a clear impact on outcomes at all ages in MM, suggesting that physical frailty might be a more important determinant of outcome than age itself [ 18 ]. The simplified frailty index has the advantage of being easy to apply in outpatient settings, but it lacks a validation set. Additionally, by definition, patients older than age 80 are considered ‘frail,’ and this tool categorizes 49% of patients as frail. It is also important to consider that the ECOG performance status is not sufficient for the elderly. Among patients with a good performance status, 13% had two or more comorbidities, 9% had ADL limitations, and 38% had IADL limitations [ 19 ]. In Korea, where resources are limited and patient numbers are high, it is not easy to measure the six items of ADL and eight items of IADL during outpatient visits. The R-MCI includes an accessible calculator that helps predict survival outcomes [ 9 ]. However, accurately assessing and entering various comorbidities takes time and specific medical resources, which can be challenging in resource-limited environments.

The KG-7 screening tool comprises seven questions, each representing a domain of the CGA with high sensitivity and specificity [ 11 ]. These items were selected from a large cohort of elderly cancer patients, making it a user-friendly tool for assessment [ 12 ]. We conducted a pilot study on elderly MM patients aged 70 and older. Although the limited number of results restricts interpretation, the IMWG frailty index, simplified frailty index, and KG-7 index all failed to achieve statistically significant results in predicting survival. Univariate analyses aimed at predicting survival showed that factors such as ECOG performance status, poor renal function (creatinine ≥ 2 mg/dL), low albuminemia (albumin < 3.5 g/dL), elevated LDH, poor ADL (≤ 4), and higher CCI (≥ 2) all tended toward poor survival outcomes. In multivariate analysis, a higher CCI (≥ 2) was statistically significant in predicting survival. This suggests that in a real-world cohort of the elderly, the age factor is not crucial for survival outcomes; instead, organ function, comorbidities, and performance appear to be more important.

Grade 3-4 AEs significantly increased the risk of death and progression [ 5 ]. In the current study, among the frailty indices, the KG-7 and simplified frailty index predicted grade 3-4 non-hematologic toxicities, while the IMWG frailty index did not. Grade 3-4 non-hematologic toxicities were associated with higher drug discontinuation rates and poorer OS.

Frailty in MM patients is dynamic. According to the study by Mian et al. [ 20 ], the frailty categorization of over 93% of patients changed (either improved or deteriorated) during the 3-year follow-up. Additionally, the study highlighted that the current frailty status, rather than the frailty status at diagnosis, is a better predictor of outcomes [ 20 ]. This underscores the importance of dynamically monitoring frailty over time to tailor treatments for elderly patients with MM. In our study, the KG-7 score was monitored every six months from diagnosis up to 24 months. Contrary to the findings of the previous study by Mian et al. [ 20 ], about two-thirds of the patients showed no change in frailty, while about one-third exhibited changes in frailty. It is important to note that there were significant limitations in interpretation due to a high dropout rate of 43%, with death being the primary reason for dropout. Given this, it is likely that the incidence of frailty changes could be higher than observed.

Although the KG-7 screening tool was initially developed for geriatric assessment in older patients with advanced cancer who were planned to undergo first-line palliative chemotherapy, it exhibits several limitations that may affect its efficacy in the elderly MM cohort. In MM, salvage therapies are often effective even in older adults, potentially diminishing the direct impact of baseline frailty on survival outcomes. The relatively short follow-up of 26.8 months may also be insufficient to fully capture differences in PFS or OS, particularly given MM’s relapsing-remitting nature and the availability of multiple lines of therapy. Additionally, because KG-7 was developed for older cancer patients in general, certain items or weighting strategies may not be fully optimized to reflect MM-specific risk factors. Moreover, although KG-7 was originally validated with a 5-point cutoff based on its sensitivity and negative predictive value for abnormal geriatric assessment, our study found its area under the curve for OS to be below 0.5. Consequently, we primarily used the 5-point cutoff to identify severe non-hematologic toxicity risk rather than OS prediction. Another limitation is that our study was conducted during a period when VRd (bortezomib+lenalidomide+dexamethasone) was not reimbursed by national insurance, leading to a higher proportion of patients (48%) receiving VMP. As a result, the treatment regimens may not fully reflect the current standard of care, potentially limiting the applicability of our findings to contemporary clinical practice where VRd is more widely utilized. Finally, the lack of extensive validation across diverse patient cohorts undermines the generalizability and reliability of KG-7’s predictive power. This contributed to KG-7’s failure to discriminate OS or PFS in this pilot study, underscoring the need for larger-scale research, extended follow-up, and possible modifications to enhance its disease-specific predictive utility.

In this study, we assessed the usefulness of the KG-7 as a frailty screening tool in elderly patients with newly diagnosed MM. While the pilot study, which involved a small number of participants, did not demonstrate its utility in predicting survival, it successfully predicted grade 3-4 non-hematologic toxicities. We are planning further studies involving a larger number of elderly MM patients to determine whether KG-7 can aid in tailoring treatments to individual patients.

Footnotes

Ethical Statement

This study was conducted in accordance with the Declaration of Helsinki and the Ethical Guidelines for Clinical Studies. This prospective cohort study was approved by the Institutional Review Board (IRB) of Seoul National Bundang Hospital (IRB No. B-2006-616-309) and the independent ethics committee of each participating hospital. Informed consent was obtained from all participants prior to enrollment.

Author Contributions

Conceived and designed the analysis: Lee JY, Bang SM.

Collected the data: Lee JY.

Contributed data or analysis tools: Lee JY, Kim SA, Koh Y, Yhim HY, Lee GW, Min CK, Do YR, Kim HJ, Bae SH, Eom HS, Jung SH, Park H, Nam SH, Lee JH, Kim SH, Lee HJ, Park YS, Bang SM.

Performed the analysis: Lee JY, Bang SM.

Wrote the paper: Lee JY.

Conflicts of Interest

Conflict of interest relevant to this article was not reported.

Funding

This research was supported by the Seoul National University Bundang Hospital Research Fund (14-2020-0039).

Acknowledgments

The Korean Multiple Myeloma Working Party conducted this study. We would like to express our deepest gratitude to Dr. So-Yeon Ahn for her statistical advice and guidance.

Electronic Supplementary Material

Supplementary materials are available at Cancer Research and Treatment website ( https://www.e-crt.org ).

References

1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin. 2016;66:7–30. doi: 10.3322/caac.21332.

2. Rosenberg PS, Barker KA, Anderson WF. Future distribution of multiple myeloma in the United States by sex, age, and race/ethnicity. Blood. 2015;125:410–2. doi: 10.1182/blood-2014-10-609461.

3. Costa LJ, Brill IK, Omel J, Godby K, Kumar SK, Brown EE. Recent trends in multiple myeloma incidence and survival by age, race, and ethnicity in the United States. Blood Adv. 2017;1:282–7. doi: 10.1182/bloodadvances.2016002493.

4. Kumar SK, Dispenzieri A, Lacy MQ, Gertz MA, Buadi FK, Pandey S, et al. Continued improvement in survival in multiple myeloma: changes in early mortality and outcomes in older patients. Leukemia. 2014;28:1122–8. doi: 10.1038/leu.2013.313.

5. Bringhen S, Offidani M, Palmieri S, Pisani F, Rizzi R, Spada S, et al. Early mortality in myeloma patients treated with first-generation novel agents thalidomide, lenalidomide, bortezomib at diagnosis: a pooled analysis. Crit Rev Oncol Hematol. 2018;130:27–35. doi: 10.1016/j.critrevonc.2018.07.003.

6. Extermann M, Hurria A. Comprehensive geriatric assessment for older patients with cancer. J Clin Oncol. 2007;25:1824–31. doi: 10.1200/JCO.2007.10.6559.

7. Hurria A, Togawa K, Mohile SG, Owusu C, Klepin HD, Gross CP, et al. Predicting chemotherapy toxicity in older adults with cancer: a prospective multicenter study. J Clin Oncol. 2011;29:3457–65. doi: 10.1200/JCO.2011.34.7625.

8. Palumbo A, Bringhen S, Mateos MV, Larocca A, Facon T, Kumar SK, et al. Geriatric assessment predicts survival and toxicities in elderly myeloma patients: an International Myeloma Working Group report. Blood. 2015;125:2068–74. doi: 10.1182/blood-2014-12-615187.

9. Engelhardt M, Domm AS, Dold SM, Ihorst G, Reinhardt H, Zober A, et al. A concise revised Myeloma Comorbidity Index as a valid prognostic instrument in a large cohort of 801 multiple myeloma patients. Haematologica. 2017;102:910–21. doi: 10.3324/haematol.2016.162693.

10. Facon T, Dimopoulos MA, Meuleman N, Belch A, Mohty M, Chen WM, et al. A simplified frailty scale predicts outcomes in transplant-ineligible patients with newly diagnosed multiple myeloma treated in the FIRST (MM-020) trial. Leukemia. 2020;34:224–33. doi: 10.1038/s41375-019-0539-0.

11. Kim JW, Kim SH, Kim YJ, Lee KW, Kim KI, Lee JS, et al. A novel geriatric screening tool in older patients with cancer: the Korean Cancer Study Group Geriatric Score (KG)-7. PLoS One. 2015;10:e0138304. doi: 10.1371/journal.pone.0138304.

12. Kim JW, Kim SH, Lee YG, Hwang IG, Kim JY, Koh SJ, et al. Prospective validation of the Korean Cancer Study Group Geriatric Score (KG)-7, a novel geriatric screening tool, in older patients with advanced cancer undergoing first-line palliative chemotherapy. Cancer Res Treat. 2019;51:1249–56. doi: 10.4143/crt.2018.451.

13. Greipp PR, San Miguel J, Durie BG, Crowley JJ, Barlogie B, Blade J, et al. International staging system for multiple myeloma. J Clin Oncol. 2005;23:3412–20. doi: 10.1200/JCO.2005.04.242.

14. Palumbo A, Avet-Loiseau H, Oliva S, Lokhorst HM, Goldschmidt H, Rosinol L, et al. Revised international staging system for multiple myeloma: a report from International Myeloma Working Group. J Clin Oncol. 2015;33:2863–9. doi: 10.1200/JCO.2015.61.2267.

15. Rosko A, Giralt S, Mateos MV, Dispenzieri A. Myeloma in elderly patients: when less is more and more is more. Am Soc Clin Oncol Educ Book. 2017;37:575–85. doi: 10.14694/EDBK_175171.

16. Glatzer M, Panje CM, Siren C, Cihoric N, Putora PM. Decision making criteria in oncology. Oncology. 2020;98:370–8. doi: 10.1159/000492272.

17. Murillo A, Cronin AM, Laubach JP, Hshieh TT, Tanasijevic AM, Richardson PG, et al. Performance of the International Myeloma Working Group myeloma frailty score among patients 75 and older. J Geriatr Oncol. 2019;10:486–9. doi: 10.1016/j.jgo.2018.10.010.

18. Pawlyn C, Cairns D, Kaiser M, Striha A, Jones J, Shah V, et al. The relative importance of factors predicting outcome for myeloma patients at different ages: results from 3894 patients in the Myeloma XI trial. Leukemia. 2020;34:604–12. doi: 10.1038/s41375-019-0595-5.

19. Repetto L, Fratino L, Audisio RA, Venturino A, Gianni W, Vercelli M, et al. Comprehensive geriatric assessment adds information to Eastern Cooperative Oncology Group performance status in elderly cancer patients: an Italian Group for Geriatric Oncology Study. J Clin Oncol. 2002;20:494–502. doi: 10.1200/JCO.2002.20.2.494.

20. Mian H, Wildes TM, Vij R, Pianko MJ, Major A, Fiala MA. Dynamic frailty risk assessment among older adults with multiple myeloma: a population-based cohort study. Blood Cancer J. 2023;13:76. doi: 10.1038/s41408-023-00843-5.

Associated Data

Supplementary Materials