Diagnostic Performance of Screening Tools for Depressive Symptoms in Vulnerable Older Patients with Cancer Undergoing Comprehensive Geriatric Assessment (CGA): Results from the SCREEN Pilot Study.

Depression is a common and disabling disorder in later life, particularly among people with poor physical health. There are many screening tools available that can be used to examine depressive symptoms; however, not all of them may be appropriate or accurate for older adults with cancer. This pilot study was designed to test the diagnostic performance of two screening tools and their short versions in a cohort of vulnerable (G8 score ≤ 14/17) older patients with cancer undergoing comprehensive geriatric assessment (CGA). The prospective analysis covered 50 vulnerable patients with cancer aged ≥70 years. The diagnostic performance of the Geriatric Depression Scale (GDS)-15, GDS-4, Patient Health Questionnaire (PHQ)-9 and PHQ-2 was compared to the 'gold standard' Structured Clinical Interview for DSM-5 Disorders (SCID-5-S) depression module A. The sensitivity and specificity in detecting depressive symptoms were the highest in the case of PHQ-2, with an area under the receiver operating characteristic curve (AUROC) of 92.7%. The AUROC for the 9-item version, PHQ-9, was 90.2%. For the GDS-15 and GDS-4, the AUROC was only 56.2% and 62.0%, respectively. The SCREEN pilot study illustrates the potential benefit of using a shorter screening tool, PHQ-2, to identify older patients with cancer who would benefit from a more in-depth emotional evaluation as part of a CGA.

Underrepresentation of vulnerable older patients with cancer in phase II and III oncology registration trials: A case-control study.

OBJECTIVES: We aimed to determine the proportion of "fit" versus "vulnerable" older patients with cancer included in phase II and III oncology registration trials, as compared to the proportions in a real life oncology setting. METHODS: Trial and patient characteristics of older (≥70years) patients treated at the OECI-designated clinical cancer centre in Kortrijk and included in a phase II or III oncology registration trial were collected retrospectively. These patients were matched individually with randomly-selected patients from the general oncology setting, based on gender, age, tumour type, tumour stage, and treatment intent. Patients' fitness, based on routine Geriatric-8 (G8) screening, was retrieved from prospectively constructed databases. RESULTS: Between November 2012 and October 2018, 218 older patients with cancer were included in a phase II or III oncology registration trial. Of those, 41 cases with a mean age of 76.0years were included in the analyses. A Fisher's Exact Test revealed a statistical significant difference between cases and matched controls, with a higher proportion of "fit" patients included in phase II or III oncology registration trials compared to the proportion in the matched control group (respectively 70.7% and 41.5%, p<.010). DISCUSSION: We provide evidence for the hypothesis that older patients included in phase II or III oncology trials are significantly fitter than the real life oncology population. Some form of geriatric evaluation should be integrated in future cancer clinical trials to enable stratification according to this parameter and allow subgroup analysis. This will broaden the application and interpretation of trial results.

Detection of alcohol abuse in older patients with cancer: The integration of alcohol screening questionnaires in the comprehensive geriatric assessment.

OBJECTIVES: We aimed to evaluate the feasibility of implementing an alcohol screening questionnaire as part of the comprehensive geriatric assessment (CGA) by reporting the prevalence of alcohol abuse in a group of older patients with cancer in a Belgian cancer centre. MATERIALS AND METHODS: Patients were recruited at the Geriatric Oncology Clinic of the Kortrijk Cancer Centre and were evaluated by use of a CGA. Two alcohol screening questionnaires were integrated into the CGA: the Cutdown-Annoyed-Guilty-Eye-opener (CAGE) questionnaire and the Alcohol Use Disorders Identification Test-screening version (AUDIT-C). RESULTS: 193 patients with a mean age of 77.7 years were included in the analyses. Abnormal scores on the CAGE were detected in 6.3% of males and 1.2% of women. Abnormal results on the AUDIT-C were noticed in 30.0% of men, and in 21.7% of women. A regression analysis could not find a significant effect of the CAGE questionnaire when entered as predictor for CGA domain scores. Regarding the AUDIT-C, significant results were detected for predicting the score of the Geriatric-8 questionnaire and polypharmacy in men, and the Independent Activities of Daily Living questionnaire in women. No association with one-year survival was detected for either alcohol screening questionnaire. DISCUSSION: It is feasible to implement an alcohol screening questionnaire as part of a CGA as results indicated a rather high level of alcohol abuse in this cohort.

Subjective, but not objective, cognitive complaints impact long-term quality of life in cancer patients.

OBJECTIVES: Cognitive complaints, of objective or subjective nature, may negatively impact cancer patients' quality of life (QoL). Further, the early detection of cognitive alterations may lead to an improved QoL. However, the content of such screening is yet unclear. This paper presents long-term QoL data of cancer patients treated with curative intent and its relation with objective and subjective cognitive complaints, and patient-reported outcome measures (PROMs). METHODS: QoL data, measured by the EORTC QLQ C-30, were obtained at baseline, 6 (T1), 12 (T2), and 24 months (T3) after treatment start, and compared between patients with and without objective and subjective cognitive complaints. The predictive value of PROMs was also examined. RESULTS: QoL data at baseline was collected in 125 patients. Response rates at T1, T2, and T3 were 84.7%, 81.5%, and 83.1%, respectively. Eighty-nine patients returned their QoL questionnaires at all times. Baseline subjective cognitive complaints had a stronger association with worse scores on patients' overall QoL and QoL subscale scores than objective cognitive complaints. An exploratory analysis into the value of PROMs in predicting long-term QoL at T3 revealed a significant effect for the Hospital Anxiety and Depression Scale-Depression and FACIT Fatigue scale. CONCLUSIONS: Self-perceived cognitive alterations are negatively associated with patients' overall QoL. As these troubles may already be present at baseline, oncology nurses should screen for the early signs of subjective cognitive complaints by use of PROMs, in order to refer the patient to proper intervention programs which may lead to an improved long-term QoL and faster reintegration into society.

The added value of an assessment of the patient's hand grip strength to the comprehensive geriatric assessment in G8-abnormal older patients with cancer in routine practice.

OBJECTIVES: A comprehensive geriatric assessment (CGA) is the key treatment approach to guide decisions in older patients with cancer. In this paper, the added value of an assessment of the patient's hand grip strength to predict survival in patients with an abnormal G8-questionnaire (G8) score is investigated. MATERIALS AND METHODS: Patients were screened by the G8, followed by a CGA in case of an abnormal screening (≤14.0). Hand grip strength was assessed by use of the JAMAR® hydraulic hand dynamometer. Cut-offs were applied according to the Fried frailty criteria. The survival rate was calculated twelve months after the CGA date. RESULTS: We retrospectively reviewed data of 2071 patients who were treated at the Kortrijk Geriatric Oncology Clinic (General Hospital Groeninge, Belgium) between November 2012 and December 2016. Of those, 944 patients with a mean age of 79.6 years were included in the analyses. 64.2% of patients presented an abnormal hand grip strength score. A log rank test revealed a statistical significant result between patients when accounting for the hand grip strength score (p < .01). When added to a Cox regression model, a significant result was found (p < .01). However, this added only 0.4% to the explained variance of the model. DISCUSSION: While a statistically significant result was detected, when adding the hand grip strength score to a regression model for survival, our data indicate that such assessment may clinically be less relevant when included in an already extensive test battery and may therefore provide only limited information in terms of patient survival.

The use of uHear™ to screen for hearing loss in older patients with cancer as part of a comprehensive geriatric assessment.

OBJECTIVE: We previously validated uHear™ to screen for hearing loss in older patients with cancer without a known hearing loss, as part of a comprehensive geriatric assessment (CGA). In view of low specificity, we tested a new modified uHear™ scoring system as described by Handzel. METHODS: Patients, aged ≥70 years, were evaluated by uHear™ and conventional audiometry, which is considered the gold standard, as part of a CGA. The pass or fail screening cut-off for uHear™ was defined as having ≥2 consecutive hearing grades starting from the moderate-severe threshold zone ranging from 0.5 to 2.0 kHz (modified Handzel-uHear™ scoring system). To accept the modified Handzel-uHear™ as screening tool, it was predefined that the combined sensitivity (S) and specificity (Sp) of the test (S + Sp/2) was at least 80% and that an actual combined (S + Sp)/2 of 90% would be found. RESULTS: Ninety ears (45 subjects) were tested. Of those ears, 24.4% were identified as impaired by conventional audiometry. Modified Handzel-uHear™ identified 26.7% of tested ears as impaired. The combined (S + Sp)/2 of the modified Handzel-uHear™ was calculated as 77.5%, while in previous cohort, this was retrospectively calculated as 94.6%. A new uHear™ scoring system was proposed and tested in current and previous cohort. A (S + Sp)/2 of 80.2 and 78.8%, respectively, were obtained. CONCLUSION: uHear™ is a feasible tool for use within the CGA and shows promising results. However, further research is warranted to optimize the cut-off method before it could be routinely implemented within geriatric oncology.

The distress thermometer predicts subjective, but not objective, cognitive complaints six months after treatment initiation in cancer patients.

OBJECTIVES: Research has indicated that cancer-related cognitive impairments (CRCI) may be influenced by psychosocial factors such as distress, worry and fatigue. Therefore, we aimed to validate the distress thermometer (DT) as a screening tool to detect CRCI six months post-treatment-initiation in a group of general cancer patients. METHODS: Patients (≥18 years, n = 125) with a histologically confirmed diagnosis of a solid cancer or hematological malignancy, scheduled for a curative treatment, were evaluated at baseline (T0) and six months post-treatment-initiation (T1) for CRCI by a neuropsychological assessment, including patient-reported outcome measures (PROMs). Assessed cognitive domains included premorbid intelligence, attention, processing speed, flexibility, verbal and visual episodic memory and verbal fluency. PROMs entailed distress (DT, cut-off ≥4, range 0-10), anxiety and depression, fatigue (FACIT-fatigue scale) and subjective cognitive complaints. RESULTS: At T0, 60.4% of patients showed a DT score of ≥4, whereas 50% met this criterion at T1. According to the definition of the International Cognition and Cancer Task Force, 25.5% and 28.3% of patients presented with a CRCI at T0 and T1, respectively. When evaluating the DT as a screening tool for CRCI at T1, data showed an inverse relationship between the DT and CRCI. ROC-curve analysis revealed an AUC <0.5. ROC-curve analyses evaluating the DT and FACIT-fatigue scale as screening tools for subjective cognitive complaints showed an AUC ± SE of, respectively, 0.642 ± 0.067 and 0.794 ± 0.057. CONCLUSIONS: The DT at T0 cannot be used to screen for objective CRCI at T1, but both the DT and FACIT-fatigue scale at T0 showed potential as screening tools for subjective cognitive complaints at T1.

Predictors of baseline cancer-related cognitive impairment in cancer patients scheduled for a curative treatment.

INTRODUCTION: Recent research in the field of cancer-related cognitive impairments (CRCI) has shown CRCI presentation prior to treatment initiation. Some have attributed these problems to worry and fatigue, whereas others have suggested an influence of age, IQ, and other psychosocial and medical factors. METHODS: Patients (≥18 years) with a histologically confirmed diagnosis of a solid cancer or hematological malignancy, scheduled for a curative treatment, were evaluated with a baseline neuropsychological assessment including Patient-Reported Outcome Measures (PROMs). PROMs entailed distress, anxiety and depression, fatigue, and cognitive complaints. The neuropsychological assessment comprised several cognitive domains such as premorbid IQ, attention, processing speed, flexibility, verbal and visual episodic memory, and verbal fluency. RESULTS: Cross-sectional data of 125 patients were collected. Patients had a mean age of 60.9 years (range: 30.0-85.0) and comprised primarily females (65.6%). Patients presented with cancer of following sites: breast (44.0%), digestive (28.8%), urological (11.2%), gynecologic (8.0%), hematologic malignancy (4.8%), and lung (3.2%). Patients presented with a premorbid IQ of 105.3 (range: 79.0-124.0). In 29.6% of patients, a CRCI was detected. Binary logistic regression analyses showed that a lower premorbid IQ (ß = -.084, P < .01) and a higher level of fatigue (ß = -.054, P < .05) predicted baseline CRCI. Premorbid IQ also predicted performance on individual cognitive domains. Some domains were also influenced by age, gender, having a breast cancer diagnosis, and an active treatment for hypertension. CONCLUSION: Premorbid IQ and fatigue are important predictors of baseline CRCI. Therefore, we advise researchers to implement a short IQ test when conducting clinical trials on CRCI.

Implementation of uHear™--an iOS-based application to screen for hearing loss--in older patients with cancer undergoing a comprehensive geriatric assessment.

OBJECTIVE: Validation of uHear™ as a screening tool to detect hearing loss in older patients with cancer without a known diagnosis of presbycusis, as part of a Comprehensive Geriatric Assessment (CGA). MATERIALS AND METHODS: Patients (≥70 years) with a histologically confirmed diagnosis of cancer, were enrolled at the time of CGA screening. Patients were evaluated by uHear™, which was compared to conventional audiometry as gold standard. We defined a pure-tone average (PTA) of ≥40dB HL as the pass or fail screening cut-off. Validation of uHear™ was defined in terms of diagnostic accuracy through Receiver Operating Characteristics (ROC)-analysis. To accept uHear™, we estimated that the Area Under the ROC-curve (AUC) had to differ significantly from 0.50 with an AUC of at least 0.70. The Whispered Voice Test and Hearing Handicap Inventory for the Elderly were also administered. RESULTS: Thirty-three patients consented for participation. In one patient, the results of one ear were excluded from the analysis as the patient was documented with a known hearing disorder in that ear. Significant hearing loss, defined by a PTA of ≥40dB HL calculated from the air conduction thresholds at 0.5, 1.0 and 2.0kHz, was found in 15.4% of tested ears. uHear™ showed excellent diagnostic accuracy with an AUC±SE of 0.98±0.14. It provided maximum sensitivity (100.0%) but poor specificity (36.4%) at our predefined cut-off score of ≥40dB HL. CONCLUSION: uHear™ can be implemented as a screening tool to detect hearing loss in older patients with cancer within a CGA.

Validation of the Freund Clock Drawing Test as a screening tool to detect cognitive dysfunction in elderly cancer patients undergoing comprehensive geriatric assessment.

OBJECTIVE: We aimed to validate the Freund Clock Drawing Test (CDT), with its predefined cutoff score of ≤4, as a screening tool to detect elderly cancer patients in need of a more in-depth cognitive evaluation within a comprehensive geriatric assessment (CGA). METHODS: Patients aged 70 years or older with a histologically confirmed diagnosis of cancer were evaluated with a full CGA, including CDT and Folstein Mini Mental State Examination (MMSE) as gold standard. Validation of the Freund CDT was defined in terms of diagnostic accuracy of the test through receiver operating characteristics (ROC)-analysis. To accept the Freund CDT as a screening tool, we estimated that the area under the ROC curve (AUC) had to differ significantly from 0.70 with an AUC of at least 0.85. RESULTS: Two hundred elderly cancer patients with a mean age of 79.0 years were included. Four patients were excluded from the analyses because of invalid results. Potential cognitive impairment (MMSE ≤23) was observed in 27.0% of patients. Based on of the AUC ± SE, the Freund CDT showed excellent diagnostic performance (0.95 ± 0.17). Furthermore, it provided excellent sensitivity (94.3%) and high specificity (87.4%). CONCLUSIONS: Our results indicate that the Freund CDT can be used as an initial screening tool to detect elderly cancer patients in need of a more in-depth cognitive assessment within CGA, instead of the MMSE.

Use of the Freund clock drawing test within the Mini-Cog as a screening tool for cognitive impairment in elderly patients with or without cancer.

OBJECTIVES: We aimed to determine an optimal cut-off score for the Clock Drawing Test (CDT), scored by the scale of Freund, for efficient screening for cognitive impairment in elderly (cancer) patients within a Comprehensive Geriatric Assessment (CGA) and to compare the Freund CDT to the Mini-Cog. MATERIALS AND METHODS: Data of 221 elderly (≥70years) patients, comprising of an OncoGeriatric (OG) and General Geriatric (GG) group, were retrospectively reviewed. All patients were evaluated with both the CDT and Mini Mental State Examination (MMSE) as the gold standard. Receiver Operating Characteristics (ROC) analysis was used to determine diagnostic performance. A pre-established algorithm was applied to retrieve Mini-Cog results through a combination of the CDT and the 3-Word Delayed Recall (3-WDR) test (included within MMSE). RESULTS: Data of 105 OG and 116 GG patients were evaluated. Potential cognitive impairment (MMSE≤23) was detected in 29.5% and 65.8% of patients, respectively. The CDT showed good diagnostic accuracy in the OG (0.88±0.03) and GG (0.85±0.03) group, based on the area under the ROC curve (AUC±SE). CDT (cut-off≤4) provided good sensitivity (80.7%) and specificity (81.1%) in the OG group and excellent sensitivity (89.6%) and moderate specificity (51.3%) in the GG group. Addition of the 3-WDR test, to form the Mini-Cog, resulted in similar positive and negative predictive values for the OG group and higher negative predictive value for the GG group. CONCLUSION: These data suggest that the Freund CDT, at the cut-off score of ≤4, is promising for use within a CGA. The Mini-Cog might be preferable in the GG population.