Abstract
Background
Standard oncology tools are inadequate to distinguish which older patients are at higher risk of developing chemotherapy‐related complications.
Materials and Methods
Patients over 70 years of age starting new chemotherapy regimens were prospectively included in a multicenter study. A prechemotherapy assessment that included sociodemographics, tumor/treatment variables, and geriatric assessment variables was performed. Association between these factors and the development of grade 3–5 toxicity was examined by using logistic regression.
Results
A total of 551 patients were accrued. Chemotherapy doses (odds ratio [OR] 1.834; 95% confidence interval [CI] 1.237–2.719) and creatinine clearance (OR 0.989; 95% CI 0.981–0.997) were the only factors independently associated with toxicity. Only 19% of patients who received reduced doses of chemotherapy and had a creatinine clearance ≥40 mL/minute had grade 3–4 toxicity, compared with 38% of those who received standard doses or had a creatinine clearance <40 mL/minute (p < .0001). However, no satisfactory multivariate model was obtained using different selection approaches.
Conclusion
Chemotherapy doses and renal function were identified as the major risk factors for developing severe toxicity in the older patient. These factors should be considered when planning to initiate a new chemotherapy regimen and should also lead to a closer follow‐up in these patients.
Implications for Practice
Older patients are more vulnerable to chemotherapy toxicity. However, standard tools are inadequate to identify who is at higher risk of developing chemotherapy‐related complications. Chemotherapy doses (standard vs. reduced) and renal function were identified as the major risk factors for developing severe toxicity in the elderly. These factors should be considered when planning to initiate a new chemotherapy regimen and should also lead to a closer follow‐up.
Keywords: Older patient, Chemotherapy, Toxicity, Toxicity risk score, Geriatric assessment
Short abstract
Cancer treatment for older patients remains a challenge. This article describes the use of a geriatric assessment tool to target the associatios between certain factors and the development of grade 3 to grade 5 toxicities in patients older than age 70 years.
Introduction
Cancer is frequently diagnosed in the elderly, with approximately 50% of patients with cancer being over 70 years of age 1. Furthermore, the number of older patients with cancer will significantly increase in the near future owing to the aging of the population.
Despite the magnitude of the problem, cancer treatment in older patients remains a challenge. Although some studies have shown that this population derives the same benefit from chemotherapy as the general population 2, 3, chemotherapy is used less frequently than in other age groups, in both the adjuvant and metastatic settings 4, 5, 6. Factors that influence the resistance to use chemotherapy in older patients include a general lack of studies in this age group as well as the fear that the progressive reduction of functional reserve that occurs in various organs with aging might increase the susceptibility of older patients to adverse effects and comorbidity 7. In fact, about 30%–50% of older patients receiving chemotherapy for a solid tumor will experience a severe (grade 3–5) toxicity over the treatment course 8, 9, 10, 11.
Standard oncology tools are inadequate to distinguish which older patients are at higher risk of developing chemotherapy‐related complications 12. Geriatric assessment (GA) is a multidimensional tool that evaluates aspects of the patient's life that can have an impact on the development of the illness and the response to its treatment. It has been recognized that GA may help clinicians to predict poor treatment outcomes: toxicity, morbidity, and mortality 13. Although some specific tools based on GA have been developed to identify older patients with high risk of developing chemotherapy toxicity 9, 10, 11, 14, 15, 16, none of them are widely used. In some cases, they were designed just for a specific tumor type 14, 15, 16 and the results therefore could not be extrapolated to other tumors. In other cases, they have not been externally validated yet 10, 11, and in others, despite external validation 17, their utility has not been confirmed in other series 18.
We have performed a prospective multicenter study to identify the more relevant clinical, laboratory, and treatment variables related with the development of grade 3–5 toxicity in a cohort of older patients with cancer. Furthermore, we assessed the predictive value of this model for chemotherapy toxicity in comparison to the Cancer and Aging Research Group (CARG) Toxicity Score 8. This tool was chosen because it is the only one that has been validated in an external series 17.
Materials and Methods
Our study was a multicenter study performed in 11 hospitals in Spain. We included 551 patients between February 2014 and June 2018. Inclusion criteria were patients over 70 years of age with histological or cytological confirmation of malignancy in any stage. Other inclusion criteria were (a) initiation of a new chemotherapy regimen, (b) the ability to read Spanish (questionnaires for geriatric assessment were in Spanish), and (c) outpatient condition. Patients with brain metastases were excluded. All patients provided written informed consent to participate in the study. The study was approved by the institutional review board at each participating institution.
Study Schema
Full clinical staging was performed according to routine clinical practice depending on each cancer type. Before starting chemotherapy, patients completed a baseline GA (Table 1). The questionnaire was delivered by a research nurse; one part was performed by the patient and another one by the health professional. The latter included the following items: Eastern Cooperative Oncology Group performance status 19, comorbidities (collected using the Cumulative Illness Rating Scale) 20, Short physical Performance Battery (gait speed, chair rise and progressive Romberg) 21, 22, the body mass index, and the Pfeiffer test 23 (Test de Pfeiffer or Short Portable Mental Status Questionnaire [SPMS] is an easily administered, validated, 10‐item screen for cognitive impairment that has been tested in clinical and institution‐based samples). Test‐retest reliability has been ≥0.82, and the agreement between the SPMS and clinical diagnosis for intact cognition or mild impairment was 82%. It has been established that three or more mistakes suggest cognitive impairment and advise a thorough study of cognitive function in an individual patient, always taking cultural and study levels into consideration. The patient part consisted of self‐reported measures of functional status: Activities of Daily Living (ADL) 24, Instrumental Activities of Daily Living (IADL) 25, number of falls in the last 6 months, medications, nutrition, psychological state 26 and social support/function 27, 28, ability to take medications unassisted, ability to walk one block, and the Vulnerable Elders Survey‐13 (VES‐13; a tool to detect frailty) 29. A member of the health care team assisted those who needed help with completing the questionnaires.
Table 1.
Summary of comprehensive geriatric assessment domains and elements
| Domain | Elements of assessment |
|---|---|
| Functional status |
ECOG performance status Activities of daily living 24 Instrumental activities of daily living 25 Physical performance test SPPB 21 No. of falls in the last 6 months |
| Comorbidity |
Cumulative Illness Rating Scale for Geriatrics score 20 |
| Psychological status |
Hospital Anxiety Scale 26 |
| Cognitive status |
Pfeiffer test 23 |
| Social support | |
| Nutritional status |
Body mass index Percent unintentional weight lost in the last 6 months |
Abbreviations: ECOG, Eastern Cooperative Oncology Group; MOS, Medical Outcomes Study; SPPB, Short‐Physical Performance Battery.
The following clinical variables were collected: age, gender, education, marital status, household composition, hearing, cancer subtype and stage, and selected blood tests obtained before treatment (hemoglobin, white blood cell count, platelets, basal creatinine, albumin, liver function, and creatinine clearance) 30, 31. Regarding treatment, chemotherapy schedule, chemotherapy doses (standard vs. reduced), treatment line, and use of granulocyte‐growth factors were recorded.
Potential risk of chemotherapy‐induced toxicity was estimated using the MAX2 index toxicity 32, 33 and was calculated for every chemotherapy regimen administered. The MAX2 index summarizes the overall risk of severe chemotherapy‐induced toxicity based on an average study using data from published clinical trials. In addition, the CARG Toxicity Score was calculated in all patients 9. In Spain, the term “walking one block,” included in the CARG Toxicity Score, is not commonly used, so we asked instead about their capacity to walk 700 meters, according the Wikipedia definition for a standard block in Manhattan.
Oncologists were blinded about the results of the GA and the CARG Toxicity Score.
Patients were prospectively followed until the end of chemotherapy course, and treatment‐related toxicity was collected by the treating oncologist at the beginning of each cycle and at the end of treatment. Unscheduled visits and emergency department admissions were also collected. Toxicity was assessed using the Common Terminology Criteria for Adverse Events (CTCAE) v.4.03 34. Blood values were captured as grade 3–5 toxicity if they met the criteria on the date of scheduled chemotherapy or at the time the patient was seeking attention because of chemotherapy toxicities.
Statistical Analysis
Given that several studies have previously shown a grade 3–5 toxicity rate of 30%–50% 8, 9, 10, 11, a conservative estimation assuming a minimum rate of grade 3–5 toxicity of 30%, with a precision of 4%, and a rate of lost patients of 10%, established a minimum sample size of 550 patients.
The primary outcome was the first occurrence of grade 3–5 toxicity. Descriptive statistics characterizing patient groups were provided. The chi‐square test was used to examine the association between grade 3–5 toxicity and categorical variables and independent t tests for continuous variables.
An evaluation of grade 3–5 toxicity predictors was performed by using logistic regression. Univariate models were first fitted for all prognostic factors. Significant variables at the 5% level were selected for inclusion in the multivariable model. Odds ratios (ORs) were reported with their 95% confidence intervals (CIs). A p value <.05 was considered statistically significant for all comparisons. After identifying continuous variables as predictors, we performed an independent exploratory analysis with each of these variables in order to establish thresholds for discriminating toxicity rate. The amount of accounted variance was determined with the Nagelkerke correlation coefficient (r 2). Model calibration and discrimination were assessed by the Hosmer‐Lemeshow test and the area under the receiver operating characteristic (ROC) curve 35, 36. Analyses were carried out by using SPSS software (version 18; SPSS, Chicago, IL).
Results
Patient Characteristics
Of the 551 patients who completed baseline assessment, 3 patients moved to another center after the first cycle of treatment, 2 died from rapidly progressing cancer prior to treatment, 2 patients withdrew consent early, and 4 were treated with targeted therapies without chemotherapy and were excluded from the analysis. Baseline patient characteristics of the 540 who finally entered in the outcome analysis, including demographics, chemotherapy, laboratory findings, and GA, are shown in Tables 2 and 3. Median age was 77 years (range 70–92); staging was I–III (42%) and IV (58%). The most common tumor types were gastrointestinal (55%), lung (13%), genitourinary (13%), and breast (5%). Fifty‐seven percent of patients received polychemotherapy, 46% received standard doses of chemotherapy, 35% received adjuvant/neoadjuvant chemotherapy, 54% received a first‐line treatment, and 11% received a second or subsequent treatment line. Thirteen percent received primary prophylaxis with granulocyte colony‐stimulating factor growth factors. Twenty‐seven percent of patients received radiation therapy together with chemotherapy.
Table 2.
Patient characteristics (n = 540)
| Characteristic | Patients, n (%) |
|---|---|
| Age, years | |
| 70–74 | 173 (32) |
| 75–79 | 173 (32) |
| ≥80 | 194 (36) |
| Sex | |
| Male | 335 (62) |
| Female | 205 (38) |
| Educational level | |
| Less than associate/bachelor's degree | 400 (74) |
| Associate/bachelor's degree | 86 (16) |
| Advanced degree | 54 (10) |
| Marital status | |
| Married | 367 (68) |
| Widowed | 108 (20) |
| Single | 43 (8) |
| Divorced/separated | 22 (4) |
| Living arrangements | |
| Lives alone | 92 (17) |
| Living with others | 448 (83) |
| Tumor site | |
| Gastrointestinal | 302 (55) |
| Lung | 71 (13) |
| Genitourinary | 71 (13) |
| Breast | 27 (5) |
| Gynecologic | 22 (4) |
| Other | 55 (10) |
| Metastatic status | |
| M0 | 230 (42) |
| M1 | 318 (58) |
| Chemotherapy | |
| Standard therapy | 245 (43) |
| Reduced therapy or monotherapy | 308 (57) |
| MAX2 index | |
| 0–0.44 | 173 (32) |
| 0.45–0.57 | 313 (58) |
| >0.57 | 54 (10) |
| Line of treatment | |
| Adjuvant/neoadjuvant | 189 (35) |
| First‐line palliative | 292 (54) |
| Subsequent‐line palliative | 59 (11) |
| Primary G‐CSF | |
| No | 470 (87) |
| Yes | 70 (13) |
Abbreviation: G‐CSF, granulocyte colony‐stimulating factor.
Table 3.
Baseline geriatric assessment (n = 540)
| Characteristic | Patients, n (%) |
|---|---|
| ECOG PS | |
| 0 | 135 (25) |
| 1 | 351 (65) |
| 2 | 54 (10) |
| IADL | |
| 8 | 227 (42) |
| ≤7 | 313 (58) |
| ADL | |
| 6 | 432 (80) |
| ≤5 | 108 (20) |
| No. of falls in the past 6 months | |
| None | 454 (84) |
| ≥1 | 86 (16) |
| SPPB | |
| <8 | 135 (25) |
| ≥8 | 405 (75) |
| CIRS‐G score | |
| No grade 3–4 comorbidity | 313 (58) |
| 1 grade 3 comorbidity | 135 (25) |
| ≥2 grade 3 or 1 grade 4 comorbidity | 92 (17) |
| Pfeiffer test | |
| ≥3 errors | 65 (12) |
| 0–2 errors | 475 (88) |
| Hospital Anxiety Scale | |
| <7 | 432 (80) |
| ≥8 | 108 (20) |
| Hospital Depression Scale | |
| <7 | 427 (79) |
| ≥8 | 113 (21) |
| MOS Social Support Survey | |
| ≤15 | 43 (8) |
| >15 | 497 (92) |
| Gijón Social Support Survey | |
| >10 | 65 (12) |
| ≤9 | 475 (88) |
| Body mass index, kg/m2 | |
| >26 | 275 (51) |
| ≤26 | 265 (49) |
| Unintentional weight loss % | |
| ≤10% | 438 (81) |
| >10% | 102 (19) |
| VES‐13 | |
| 0–2 | 270 (50) |
| ≥3 | 270 (50) |
Abbreviations: ADL, Activities of Daily Living; CIRS‐G: Cumulative Illness Rating Scale for Geriatrics; ECOG PS, Eastern Cooperative Oncology Group performance status; IADL: Instrumental Activities of Daily Living; MOS, Medical Outcomes Study, SPPB, Short‐Physical Performance Battery; VES‐13, Vulnerable Elders Survey‐13.
In regard to GA, 17% of patients lived alone and 74% had less than associate/bachelor's degree. According to the Cumulative Illness Rating Scale for Geriatrics score, 42% of patients had at least one grade 3–4 comorbidity. ADL and IADL were dependent in 20% and 58% of patients, respectively. Impairment of cognitive function was detected in 13% of the patients. Fifteen percent had a low body mass index of less than 22, and 19% reported unintentional weight loss of more than 10% of the body weight over the past 6 months. According to the Vulnerable Elders Survey‐13, 50% had a score ≥ 3. In laboratory findings, low hemoglobin (Hb) level (Hb <10 g/dL in females and Hb <11 g/dL in males), hypoalbuminemia (<3.5 g/dL), and a decrease in renal function (creatinine clearance <40 mL/minute) were shown in 23%, 26%, and 27% of patients, respectively.
Chemotherapy Toxicity
The median chemotherapy cycles received was 5 (range 1–19). At least one grade 3–5 toxicity occurred in 33.5% of 540 patients (22% grade 3, 11% grade 4, and 0.5% grade 5; Table 4). Hematologic and nonhematologic grade 3–5 toxicity occurred in 19% and 28%, respectively. The most common grade 3–5 hematologic toxicities were neutropenia (10%) and anemia (7%). The most common grade 3–5 nonhematologic toxicities were fatigue (14%), diarrhea (9%), and neuropathy (7%). One patient died as a result of chemotherapy toxicity (febrile neutropenia and sepsis). Twenty‐eight percent of patients required a dose reduction during therapy, and 37% had a dose delay.
Table 4.
Treatment‐related adverse events
| Toxicity type | Grade 3–5, n (%) | Grade 3, n (%) | Grade 4, n (%) | Grade 5, n (%) |
|---|---|---|---|---|
| Hematologic | 103 (19) | 78 (14) | 24 (4) | 1 (1) |
| Neutropenia | 52 (10) | 49 (9) | 3 (1) | |
| Anemia | 39 (7) | 37 (6) | 2 (1) | |
| Thrombocytopenia | 24 (4) | 24 (4) | ||
| Febrile neutropenia | 8 (1) | 5 (1) | 1 (1) | 1 (1) |
| Nonhematologic | 151 (28) | 126 (23) | 24 (4) | 1 (1) |
| Fatigue | 76 (14) | 73 (13) | 3 (1) | |
| Diarrhea | 49 (9) | 40 (7) | 9 (2) | |
| Mucositis | 16 (3) | 16 (3) | ||
| Nausea/vomiting | 22 (4) | 22 (4) | ||
| Neuropathy | 36 (7) | 7 (1) | ||
| Dermatologic toxicity | 21 (4) | 4 (1) | ||
| Others | 48 (9) | 41 (7) | 7 (1) | 1 (1) |
Predictive Variables Associated with Occurrence of Grade 3–5 Toxicity
Univariate analysis was performed to analyze domains of GA, clinical, and laboratory parameters. We identified four factors that were associated with a higher risk of grade 3–5 toxicity: creatinine clearance ≤40 mL/minute, IADL ≤7, VES‐13 ≥ 5, and the administration of standard chemotherapy doses (Table 5).
Table 5.
Association between patient characteristics and toxicity
| Variable | Patients, n (%) | No grade 3–5 toxicity, n (%) | Grade 3–5 toxicity, n (%) | OR (95% CI) | p value |
|---|---|---|---|---|---|
| Age, years | 1.02 (0.98–1.06) | .376 | |||
| 70 to <78 | 289 (53) | 190 (52) | 99 (56) | ||
| ≥78 | 251 (47) | 174 (48) | 77 (44) | ||
| Sex | 1.35 (0.93–1.95) | .107 | |||
| Male | 333 (62) | 233 (64) | 100 (57) | ||
| Female | 207 (38) | 131 (36) | 76 (43) | ||
| Tumor site | 0.68 (0.31–1.44) | .475 | |||
| GI or GU | 373 (68) | 239 (66) | 121 (69) | ||
| Other | 167 (32) | 125 (34) | 55 (31) | ||
| No. of chemotherapy agents | 0.81 (0.44–1.51) |
.797 |
|||
| Monochemotherapy | 232 (43) | 155 (43) | 77 (44) | ||
| Polychemotherapy | 308 (57) | 209 (57) | 99 (56) | ||
| Chemotherapy dosing | 1.51 (1.04–2.181 |
.028 |
|||
| Standard dose | 301 (56) | 191 (52) | 110 (62) | ||
| Reduced dose | 239 (44) | 173 (48) | 66 (38) | ||
| MAX2 index | |||||
| 0–0.44 | 173 (32) | 109 (30) | 62 (35) | Ref | |
| 0.44–0.57 |
313 (58) |
222 (61) |
91 (52) |
0.81 (0.44–1.52) |
.51 |
| >0.57 | 54 (10) | 33 (9) | 23 (13) | 1.42 (0.95–1.68) | .076 |
| Hemoglobin g/dL | 0.94 (0.84–1.05) | .573 | |||
| ≥10 (female), ≥11 (male) |
416 (77) |
283 (78) |
133 (76) |
||
| <10 (female), <11 (male) | 124 (23) | 81 (22) | 43 (24) | ||
| Albumin g/dL | 0.81 (0.55–1.19) | .315 | |||
| >3.5 | 457 (84) | 312 (86) | 145 (82) | ||
| ≤3.5 | 83 (26) | 52 (14) | 31 (18) | ||
| Creatinine clearance, mL/minute | 0.67 (0.46–0.99) | .031 | |||
| ≥40 | 394 (73) | 276 (76) | 118 (67) | ||
| <40 | 146 (27) | 88 (24) | 58 (33) | ||
| ECOG PS | 1.30 (0.87–1.85) | .236 | |||
| 0 | 133 (25) | 94 (26) | 39 (22) | ||
| ≥1 | 407 (75) | 270 (74) | 137 (78) | ||
| CIRS‐G score | 1.02 (0.89–1.17) | .155 | |||
| No grade 3–4 comorbidity |
336 (62) |
234 (64) |
102 (58) |
||
| Any grade 3–4 comorbidity | 204 (48) | 130 (36) | 74 (42) | ||
| IADL | 0.83 (0.73–0.93) | .034 | |||
| ≥7 | 302 (56) | 215 (59) | 87 (49) | ||
| <7 | 238 (44) | 149 (41) | 89 (51) | ||
| ADL | 0.99 (0.98–1.01) | .286 | |||
| 6 | 334 (62) | 230 (63) | 104 (59) | ||
| ≤5 | 206 (38) | 134 (37) | 72 (41) | ||
| Falls in the past 6 months | 1.11 (0.89–1.42) | .193 | |||
| None | 454 (84) | 308 (85) | 143 (81) | ||
| ≥1 | 86 (16) | 56 (15) | 33 (19) | ||
| SPPB | 1.04 (0.97–1.09) | .669 | |||
| ≥8 | 382 (71) | 251 (69) | 131 (75) | ||
| <8 | 144 (27) | 103 (28) | 41 (23) | ||
| Missing | 14 (3) | 10 (3) | 4 (2) | ||
| Pfeiffer test | 0.67 (0.27–1.32) | .366 | |||
| 0–2 errors | 475 (88) | 317 (87) | 158 (90) | ||
| ≥3 errors | 65 (12) | 47 (13) | 18 (10) | ||
| Hearing | 0.65 (0.42–1.01) | .056 | |||
| Excellent/good | 405 (75) | 264 (73) | 141 (80) | ||
| Fair/poor/deaf | 135 (25) | 100 (27) | 35 (20) | ||
| Medications intake | 0.83 (0.56–1.10) | .332 | |||
| No assistance | 491 (91) | 334 (92) | 157 (89) | ||
| Requires assistance | 49 (9) | 30 (8) | 19 (11) | ||
| MOS Social Support Survey | 0.99 (0.94–1.05) | .256 | |||
| ≤15 | 43 (8) | 25 (7) | 18 (10) | ||
| >15 | 497 (92) | 339 (93) | 158 (90) | ||
| Body mass index, kg/m2 | 1.01 (0.96–1.04) | .898 | |||
| <25 | 223 (41) | 151 (41) | 72 (41) | ||
| ≥25 | 317 (59) | 213 (59) | 104 (59) | ||
| Unintentional weight loss % | 1.02 (0.97–1.06) | .518 | |||
| ≤10% | 438 (81) | 298 (82) | 140 (80) | ||
| >10% | 102 (19) | 66 (18) | 36 (20) | ||
| VES‐13 | 1.54 (1.02–2.06) | .032 | |||
| 0–5 | 425 (79) | 296 (81) | 129 (73) | ||
| >5 | 115 (21) | 68 (19) | 47 (27) |
Bolded values are statistically significant.
Abbreviations: ADL, Activities of Daily Living; CI, confidence interval; CIRS‐G, Cumulative Illness Rating Scale for Geriatrics; ECOG PS, Eastern Cooperative Oncology Group performance status; GI, gastrointestinal; GU, genitourinary; IADL, Instrumental Activities of Daily Living; MOS, Medical Outcomes Study; OR, odds ratio; SPPB, Short‐Physical Performance Battery; VES‐13, Vulnerable Elders Survey‐13.
In multivariate analysis, only chemotherapy doses (OR 1.834; 95% CI 1.237–2.719) and creatinine clearance (OR 0.989; 95% CI 0.981‐0.997) were independently associated with toxicity. However, no satisfactory multivariate model was obtained using different selection approaches. Just considering those two variables, only 19% of patients who received reduced doses of chemotherapy and had a creatinine clearance ≥40 mL/minute presented grade 3–4 toxicity, compared with 38% of those patients who received standard doses or had a creatinine clearance <40 mL/minute (p < .0001). The ROC curve was 0.59 (considered as a continuous variable; 95% CI 0.54–0.64).
Ability of CARG Toxicity Score to Predict Toxicity
The median overall risk score was 7. The risk of grade 3–5 according to the three risk categories established was 25% for low risk (0 to 5 points), 35% for intermediate risk (6 to 9 points), and 34% for high risk (10 to 19 points). The area under the ROC curve was 0.54 for the CARG Toxicity Score (considered as a continuous variable; 95% CI 0.49–0.59).
Discussion
Given the greater susceptibility of older patients to develop chemotherapy side effects, some tools are needed to identify those patients with a higher risk of developing toxicity. This could help to avoid the worsening of quality of life induced by treatment, loss of independence, hospital admissions, and the risk of death due to toxicity.
The results of our study suggest that chemotherapy doses (standard vs. reduced) and renal function are the two major factors to identify patients with a higher risk of suffering at least one grade 3–5 toxicity. Doses of chemotherapy as a risk factor for toxicity have also been described by other authors 9, 11. However, in our series, as opposed to some previous studies 9, 10, some aspects related to treatment such as the chemotherapy scheme or tumor type (which often determines type of chemotherapy given) were not related to the development of toxicity. This could suggest that doses administered are more important than the scheme in relation to toxicity in older patients. Similarly, renal function has also been reported by other authors as a predisposing factor for the development of toxicity 9, 37. In fact, it has been noted that for every 10 mL/minute decrease in creatinine clearance, the risk of chemotherapy‐related toxicity increases by 12%. This association was independent of the type of chemotherapy received 38.
Although several authors have pointed out the ability of some geriatric variables to predict the development of toxicity in the older patient, such as the grip strength 39, timed up and go 18, nutritional status 10, 11, 40, cognition 10, 18, 40, IADL 10, social support 9, and others, the lack of consistency of these variables is striking among the different studies that have tried to identify predictive risk factors for chemotherapy toxicity in the older patient 41. In fact, none of the three tools that have been proposed so far to predict toxicity in different tumors include the same geriatric variables 9, 10, 11. In our series, none of the variables included in the geriatric assessment were useful to identify the risk of developing grade 3–5 toxicity as a result of chemotherapy. Although two of these instruments (IADL and VES‐13) were predictive in the univariate analysis, they were not retained in the multivariate analysis. We believe that the low impact of these geriatric assessment variables in our series may be due to the fact that oncologists already adjust for them when treatment doses are planned. Oncologists in the study were experienced in the treatment of older patients with cancer, and although at the time of making the decision about the treatment they did not have the results of the geriatric assessment, we believe that the clinical judgment to fitting standard versus lower chemotherapy doses could influence the selection of treatment and, therefore, the risk of inducing toxicity.
The toxicity rate reported in the current study (33.5%), although similar to what has been previously published by some authors 8, is well less than has been observed in multiple prior studies, where it is around 50%, or even higher 9, 10, 11. This is probably influenced by different chemotherapy schedules, delivered doses, and patients’ characteristics, therefore showing different severe toxicity rates. In fact, prior studies on adverse drug reactions have reported sex differences with women having higher rates 42, and this may also be relevant to the current findings, as the percentage of female patients in our series (38%) is smaller than in other studies 9, 10, although similar to some others 13, 18. Additionally, rates of unintentional weight loss >10% were much less frequent in our study than in Hurria et al. 11 (19% vs. 38%), which may also be relevant for toxicity experience.
The rate of dose reduction during therapy in our study was 28%, and the percentage of patients that required a dose delay was 37%. These data may appear higher than what has been previously published 9, but it should be taken into consideration that 36% of our patients were ≥ 80 years of age, whereas in some other studies this percentage ranges from 7% to 19% 9, 11, 18. In this elderly population, a grade 2 toxicity maintained in time during several cycles can also have a negative impact in patients’ quality of life, and therefore, it may lead to a dose delay. So, apart from dose delays scheduled to enhance recovery from previous grade 3–4 toxicities, there were also some patients that needed a dose delay because of grade 2 toxicities.
Another relevant result of our study is that the CARG Toxicity Score was not a useful predictor to estimate the risk of chemotherapy toxicity in older patients. Although this tool has been subsequently validated in a cohort of 250 patients 17 and in some other small series 43, 44, neither we nor other authors 19 have been able to confirm its usefulness. Our series presents some characteristics that make it different from the study by Hurria et al. 9 and that may explain, in part, these results: older age, higher proportion of male patients, lower proportion of patients with severe weight loss, higher proportion of gastrointestinal tumors, lower percentage of patients receiving standard dose treatment, cultural differences (for instance, adapting the term “walking one block” to Spanish language), and so on. In addition, within the CARG Toxicity Score, the decision to consider the patient fit for chemotherapy at standard doses, reduced doses, or single‐agent chemotherapy in an effort to avoid toxicity is subjective and depends on the clinical judgment. Oncologists participating in our study could have selected patients for chemotherapy in a different way. This not only can alter the score's performance but also can explain the lower incidence of grade 3–5 toxicity observed in our series compared with other similar studies 9, 10, 11.
The main strengths of our study are the multicenter approach and the number of patients. Both features are critical for the external validation of a predictive model 45. In addition, the chemotherapy‐toxicity was collected prospectively, strengthening the outcome data. As for limitations, it should be noted that this is a heterogeneous series, which included patients with different tumors, different lines of treatment, and different therapeutic objectives. Also, we focused on the development of grade 3–5 side effects, but in older patients, it is also interesting to determine the risk of developing less severe toxicities (grade 1–2) because these can also influence the quality of life and morbidity. Additionally, the prediction of other variables such as functional deterioration, impact on quality of life, dependence, and even hospitalization was not assessed. A recent study suggests that the CARG score may predict not only the development of toxicity but also the risk of hospitalization for toxicity 43. Finally, our study did not include the use of targeted therapies or immunotherapy, which are becoming widely used and have specific toxicity profiles. Future studies should address the prediction of toxicity in older patients receiving new drugs.
Conclusion
In our series, the dose of chemotherapy administered and renal function were identified as risk factors for developing severe toxicity in older patients. However, with these variables, no reliable predictive tool could be developed. Therefore, because risk of serious toxicity may determine the therapeutic plan, further research is needed to develop better predictive tools.
Author Contributions
Conception/design: Jaime Feliu, Beatriz Jiménez‐Munárriz, Maite Antonio‐Rebollo, Regina Gironés, María‐Dolores Torregrosa, María José Molina‐Garrido
Provision of study material or patients: Jaime Feliu, Beatriz Jiménez‐Munárriz, Laura Basterretxea, Irene Paredero, Elisenda Llabrés, Maite Antonio‐Rebollo, Beatriz Losada, Enrique Espinosa, Regina Gironés, Ana Belén Custodio, María del Mar Muñoz, Mariana Díaz‐Almirón, Jeniffer Gómez‐Mediavilla, Alvaro Pinto, María‐Dolores Torregrosa, Gema Soler, Patricia Cruz, Oliver Higuera, María José Molina‐Garrido
Collection and/or assembly of data: Jaime Feliu, Beatriz Jiménez‐Munárriz, Laura Basterretxea, Irene Paredero, Elisenda Llabrés, Maite Antonio‐Rebollo, Beatriz Losada, Enrique Espinosa, Regina Gironés, Ana Belén Custodio, María del Mar Muñoz, Mariana Díaz‐Almirón, Jeniffer Gómez‐Mediavilla, Alvaro Pinto, María‐Dolores Torregrosa, Gema Soler, Patricia Cruz, Oliver Higuera, María José Molina‐Garrido
Data analysis and interpretation: Jaime Feliu, Beatriz Jiménez‐Munárriz, Mariana Díaz‐Almirón
Manuscript writing: Jaime Feliu, Beatriz Jiménez‐Munárriz, Enrique Espinosa, Alvaro Pinto
Final approval of manuscript: Jaime Feliu, Beatriz Jiménez‐Munárriz, Laura Basterretxea, Irene Paredero, Elisenda Llabrés, Maite Antonio‐Rebollo, Beatriz Losada, Enrique Espinosa, Regina Gironés, Ana Belén Custodio, María del Mar Muñoz, Mariana Díaz‐Almirón, Jeniffer Gómez‐Mediavilla, Alvaro Pinto, María‐Dolores Torregrosa, Gema Soler, Patricia Cruz, Oliver Higuera, María José Molina‐Garrido
Disclosures
Jaime Feliu: Amgen, Ipsen, Eisai, Merck, Roche, Novartis (C/A), Merck (RF), Amgen, Servier (other). The other authors indicated no financial relationships.
(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board
Disclosures of potential conflicts of interest may be found at the end of this article.
No part of this article may be reproduced, stored, or transmitted in any form or for any means without the prior permission in writing from the copyright holder. For information on purchasing reprints contact Commercialreprints@wiley.com. For permission information contact permissions@wiley.com.
References
- 1. Siegel RL, Miller KD, Jemal A. Cancer statistics 2016. CA Cancer J Clin 2016;66:7–30. [DOI] [PubMed] [Google Scholar]
- 2. Haller DG, O'Connell MJ, Cartwright TH et al. Impact of age and medical comorbidity on adjuvant treatment outcomes for stage III colon cancer: A pooled analysis of individual patient data from four randomized, controlled trials. Ann Oncol 2015;26:715–724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Langer CJ, Manola J, Bernardo P et al. Cisplatin‐based therapy for elderly patients with advanced non‐small‐cell lung cancer: Implications of Eastern Cooperative Oncology Group 5592, a randomized trial. J Natl Cancer Inst 2002;94:173–181. [DOI] [PubMed] [Google Scholar]
- 4. Hurria A, Wong FL, Villaluna D et al. Role of age and health in treatment recommendations for older adults with breast cancer: The perspective of oncologists and primary care providers. J Clin Oncol 2008;26:5386–5392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Potosky AL, Harlan LC, Kaplan RS et al. Age, sex, and racial differences in the use of standard adjuvant therapy for colorectal cancer. J Clin Oncol 2002;20:1192–1202. [DOI] [PubMed] [Google Scholar]
- 6.Daniele B, Simmonds PD, Best LY et al. Should chemotherapy be used as a treatment of advanced colorectal carcinoma (ACC) in patients over 70 years age? Eur J Cancer 1999;35:1640–1649. [DOI] [PubMed] [Google Scholar]
- 7. Yancik R, Wesley MN, Ries LA et al. Comorbidity and age as predictors of risk for early mortality of male and female colon carcinoma patients: A population‐based study. Cancer 1998;82:2123–2134. [PubMed] [Google Scholar]
- 8. Luciani A, Biganzoli L, Colloca G et al. Estimating the risk of chemotherapy toxicity in older patients with cancer: The role of the Vulnerable Elders Survey‐13 (VES‐13). J Geriatr Oncol 2015;6:272–279. [DOI] [PubMed] [Google Scholar]
- 9. Hurria A, Togawa K, Mohile SG et al. Predicting chemotherapy toxicity in older adults with cancer: A prospective multicenter study. J Clin Oncol 2011;29:3457–3465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Extermann M, Boler I, Reich RR et al. Predicting the risk of chemotherapy toxicity in older patients: The Chemotherapy Risk Assessment Scale for High‐Age Patients (CRASH) score. Cancer 2012;118:3377–3386. [DOI] [PubMed] [Google Scholar]
- 11. Kim JW, Lee YG, Hwang IG et al. Predicting cumulative incidence of adverse events in older patients with cancer undergoing first‐line palliative chemotherapy: Korean Cancer Study Group (KCSG) multicentre prospective study. Br J Cancer 2018;118:1169–1175. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Feliu J, Heredia‐Soto V, Gironés R et al. Can we avoid the toxicity of chemotherapy in elderly cancer patients? Crit Rev Oncol Hematol 2018;131:16–23. [DOI] [PubMed] [Google Scholar]
- 13. Mohile SG, Dale W, Somerfield MR et al. Practical assessment and management of vulnerabilities in older patients receiving chemotherapy: ASCO Guideline for Geriatric Oncology. J Clin Oncol 2018;36:2326–2347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Aparicio T, Jouve JL, Teillet L et al. Geriatric factors predict chemotherapy feasibility: Ancillary results of FFCD 2001‐02 phase III study in first‐line chemotherapy for metastatic colorectal cancer in elderly patients. J Clin Oncol 2013;31:1464–1470. [DOI] [PubMed] [Google Scholar]
- 15. Freyer G, Geay JF, Touzet S et al. Comprehensive geriatric assessment predicts tolerance to chemotherapy and survival in elderly patients with advanced ovarian carcinoma: A GINECO study. Ann Oncol 2005;16:1795–1800. [DOI] [PubMed] [Google Scholar]
- 16. Maione P, Perrone F, Gallo C et al. Pretreatment quality of life and functional status assessment significantly predict survival of elderly patients with advanced non‐small‐cell lung cancer receiving chemotherapy: A prognostic analysis of the multicenter Italian lung cancer in the elderly study. J Clin Oncol 2005;23:6865–6872. [DOI] [PubMed] [Google Scholar]
- 17. Hurria A, Mohile S, Gajra A et al. Validation of a prediction tool for chemotherapy toxicity in older adults with cancer. J Clin Oncol 2016;34:2366–2371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Moth EB, Kiely BE, Stefanic N et al. Predicting chemotherapy toxicity in older adults: Comparing the predictive value of the CARG Toxicity Score with oncologists’ estimates of toxicity based on clinical judgement. J Geriatr Oncol 2019;10:202–209. [DOI] [PubMed] [Google Scholar]
- 19. Roila F, Lupattelli M, Sassi M et al. Intra and interobserver variability in cancer patients’ performance status assessed according to Karnofsky and ECOG scales. Ann Oncol 1991. 2:437–439. [DOI] [PubMed] [Google Scholar]
- 20. Linn BS, Linn MW, Gurel L. Cumulative illness rating scale. J Am Geriatr Soc 1968;16:622–626. [DOI] [PubMed] [Google Scholar]
- 21. Guralnik JM, Simonsick EM, Ferruci L et al. A short physical performance battery assessing lower extremity function: Association with self‐reported disability and prediction of mortality and nursing home admission. J Gerontol 1994;49:85–94. [DOI] [PubMed] [Google Scholar]
- 22. Guralnik JM, Ferruci L, Simonsick EM et al. Lower‐extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med 1995;332:556–561. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Pfeiffer E. A short portable mental status questionnaire for the assessment of organic brain deficits in elderly patients. J Am Geriatr Soc 1975;22:433–441. [DOI] [PubMed] [Google Scholar]
- 24. Katz S, Ford AB, Moskowitz RW et al. Studies of illness in the aged: The index of ADL—A standardized measure of biological and psychosocial function. JAMA 1963;185:914–919. [DOI] [PubMed] [Google Scholar]
- 25. Lawton MP, Brody EM. Assessment of older people: Self‐maintaining and instrumental activities of daily living. Gerontologist 1969;9:179–186. [PubMed] [Google Scholar]
- 26. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scan 1983;67:361–370. [DOI] [PubMed] [Google Scholar]
- 27. Ren XS, Skinner K, Lee A et al. Social support, social selection and self‐assessed health status: Results from the veterans health study in the United States. Soc Sci Med 1999;48:1721–1734. [DOI] [PubMed] [Google Scholar]
- 28. Costa‐Requena G, Salamero M, Gil F. Validity of the questionnaire MOS‐SSS of social support in neoplastic patients. Med Clin 2007;128:687–691. [DOI] [PubMed] [Google Scholar]
- 29. Luciani A, Ascione G, Bertuzzi C et al. Detecting disabilities in older patients with cancer: Comparison between comprehensive geriatric assessment and vulnerable elders survey‐13. J Clin Oncol 2010;28:2046–2050. [DOI] [PubMed] [Google Scholar]
- 30. Jelliffe RW. Estimation of creatinine clearance when urine cannot be collected. Lancet 1971;1:975–976. [DOI] [PubMed] [Google Scholar]
- 31. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31–41. [DOI] [PubMed] [Google Scholar]
- 32. Extermann M, Chen H, Cantor AB et al. Predictors of tolerance to chemotherapy in older cancer patients: A prospective pilot study. Eur J Cancer 2002;38:1466–1473. [DOI] [PubMed] [Google Scholar]
- 33. Extermann M, Bonetti M, Sledge GW et al. MAX2—A convenient index to estimate the average per patient risk for chemotherapy toxicity; validation in ECOG trials. Eur J Cancer 2004;40:1193–1198. [DOI] [PubMed] [Google Scholar]
- 34.Common Terminology Criteria for Adverse Events (CTCAE), Version 4.0. Available at https://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03/CTCAE_4.03_2010-06-14_QuickReference_8.5x11.pdf
- 35. Hosmer DW, Lemeshow S, eds. Applied Logistic Regression. New York: John Wiley and Sons, 1989. [Google Scholar]
- 36. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29–36. [DOI] [PubMed] [Google Scholar]
- 37. Hurria A, Hurria A, Brogan K et al. Effect of creatinine clearance on patterns of toxicity in older patients receiving adjuvant chemotherapy for breast cancer. Drugs Aging 2005;22:785–791. [DOI] [PubMed] [Google Scholar]
- 38. Peterson LL, Hurria A, Feng T et al. Association between renal function and chemotherapy‐related toxicity in older adults with cancer. J Geriatr Oncol 2017;8:96–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Puts MT, Monette J, Girre V et al. Are frailty markers useful for predicting treatment toxicity and mortality in older newly diagnosed cancer patients? Results from a prospective pilot study. Crit Rev Oncol Hematol 2011;78:138–149. [DOI] [PubMed] [Google Scholar]
- 40. Aaldriks AA, Maartense E, Nortier HJ et al. Prognostic factors for the feasibility of chemotherapy and the Geriatric Prognostic Index (GPI) as risk profile for mortality before chemotherapy in the elderly. Acta Oncol 2016;55:15–23. [DOI] [PubMed] [Google Scholar]
- 41. Versteeg KS, Konings IR, Lagaay AM et al. Prediction of treatment‐related toxicity and outcome with geriatric assessment in elderly patients with solid malignancies treated with chemotherapy: A systematic review. Ann Oncol 2014;25:1914–1918. [DOI] [PubMed] [Google Scholar]
- 42. de Vries ST, Denig P, Ekhart C et al. Sex differences in adverse drug reactions reported to the National Pharmacovigilance Centre in the Netherlands: An explorative observational study. Br J Clin Pharmacol 2019;85:1507–1515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Nie X, Liu D, Li Q et al. Predicting chemotherapy toxicity in older adults with lung cancer. J Geriatr Oncol 2013;4:334–339. [DOI] [PubMed] [Google Scholar]
- 44. Nishijima TF, Deal AM, Williams GR et al. Chemotherapy toxicity risk score for treatment decisions in older adults with advanced solid tumors. The Oncologist 2018;23:573–579. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Collins GS, Ogundimu EO, Altman DG. Sample size considerations for the external validation of a multivariable prognostic model: A resampling study. Stat Med 2016;35:214–226. [DOI] [PMC free article] [PubMed] [Google Scholar]