Analysis of the Clinical Management and Quality of Life of Frail Patients with Cancer and Breakthrough Pain in Clinical Practice.


 BackgroundThe purpose of this study was to analyse the clinical management and quality of life of frail patients with cancer, chronic pain and breakthrough pain (BTP) and to assess whether treatment was conditioned by their frailty status.Methods This was an observational study in adult frail patients with cancer, chronic background pain and BTP. Outcomes of interest collected include clinical and sociodemographic data, Karnofsky Performance Status, quality of life (EuroQoL-5D-5L), chronic pain and BTP characteristics, as well as treatments administered for their control.ResultsA total of 222 patients were included with a mean age of 68 years (range 24-91), 60.5% men, with a mean Karnofsky of 63.2%. The number of daily episodes of BTP was 3.8 (95% CI 3.3-4.3), with a duration of 34.6 minutes (95% CI 28.8-40.3), and 56.8% had a gradual onset. Opioids were administered to 88.3% of patients for the chronic pain, and to 83.8% for BTP. The treatment's daily doses administered for chronic pain and BTP did not differ from those usually recommended. Quality of life was significantly worst in frail patients with cancer than no frail patients and was related to performance status (p<0.001) and to the social-familial status (p=0.045). ConclusionsBTP in frail patients with cancer presents with more episodes, of a shorter duration and more gradual onset compared to other patients with BTP, and Quality of life was seriously affected. No relevant differences were seen in the doses or method of administration of treatments for chronic pain and BTP in frail patients with cancer as compared to the standard recommendations for non-frail patients. Trial registration: Not applicable to the study.


Background
Frailty is de ned as a physiologic state with increased vulnerability to stress factors, resulting from the decline in physiologic reserve or dysregulation of multiple physiologic systems [1]. It is a particularly important risk factor for patients with cancer because cancer itself and its treatment adds stressors that can reduce the patient's physiologic reserve. The frequency of frailty in elderly patients with cancer is high, and approximately half of them are frail or pre-frail [2], with a greater risk of postoperative complications, chemotherapy intolerance, disease progression, and death [3][4][5][6][7]. Frailty has been associated in different studies with an increased risk of chemotherapy-related toxicity and poorer tolerance to treatment [8][9]. For these reasons, many studies in patients with different types of cancer now recognize the importance of the evaluation of frailty for the strati cation of risk for patients in the selection of the cancer therapy and for the decision of pharmacological or non-pharmacological treatment like radiotherapy and brachytherapy [7,[10][11][12]. Furthermore, the relationship between frailty and the prognosis of the tumour process has been widely demonstrated, and is, therefore, a factor that should be considered in these patients [13]. Therefore, the International Society of Geriatric Oncology (SIOG) recommends frailty assessment in all cancer patients over 70 years of age to help the oncologist make decisions about the most appropriate treatment for the patient [10].
Most patients with cancer suffer chronic pain, with a prevalence between 33% and 64%, reaching over 70% in patients in advanced stages of the disease [14]. Furthermore, patients with chronic pain may experience, at some point in their course, the onset of breakthrough pain (BTP), which is de ned as a "transient exacerbation of pain occurring spontaneously or related to a speci c predictable or unpredictable trigger, despite stable and adequately controlled background pain" [11,16].
The incidence of BTP varies across studies between 35-95% [17]. In cancer patients, the prevalence of BTP increases as the disease progresses and can reach up to 80% [14]. This frequency increases in patients with low performance status and advanced stages of the disease and is an indicator of poor prognosis [18].
Recognition of frailty status in patient with cancer conditions the decision of the type of treatment for the cancer and its dose, and it is related to its prognosis. However, in the group of frail patients with chronic pain and BTP, it is not known whether the choice of analgesic therapy is conditioned by the presence of frailty. In this regard, the objectives of this study were to analyse the clinical management and quality of life of frail patients with chronic pain and BTP and the treatments administered for pain control in standard clinical practice.

Methods
A cross-sectional observational study was conducted involving 29 investigators from sites in 12 Spanish provinces, including 17 medical oncology units, six pain units, three palliative care units, two geriatric departments, and one home hospitalization unit.
Patients were included from 27-June-2018 to 13-May-2019. The study was approved by the Medicinal Product Research Ethics Committee of HM Hospitales de Madrid (2-April-2018; Minutes 132).
Written informed consent was obtained from all patients. The study protocol was in accordance with the ethical standards described in the Declaration of Helsinki.

Patient selection
Patients were consecutively selected from those who attended to the clinics and who met the screening criteria, completing a single visit. No treatment was administered as a requirement of the study.
The study population consisted of frail patients with a history of controlled background chronic pain and a diagnosis of BTP. Patients with cancer were analysed in this report.
The Frail scale was used to identify frail patients. This is a validated scale consisting of ve questions corresponding to a domain: Fatigue, Resistance, Ambulation, Illness and Weight Loss. Each domain is scored with one point. Patients were classi ed as frail when they scored three or more points, over ve [19].
The Davies criteria and algorithm were used to diagnose BTP [16]. This algorithm determines the existence of BTP with: 1) Presence of baseline pain as persistent pain for 12 or more hours per day, in the week prior to the assessment (or that would exist if analgesics were not taken); 2) Adequately controlled background pain: no pain or mild pain (not moderate or severe) for 12 or more hours per day, during the week prior to the assessment; 3) Presence of transient pain exacerbations: severe or unbearable, with a visual analogue scale score for pain intensity greater than seven points over ten points, occurring spontaneously or related to a speci c, predictable or unpredictable trigger.
The inclusion criteria were: 1) Adult men and women; 2) Frail patients with ≥3 points on the Frail scale [19]; 3) Patients with controlled background pain with a visual analogue scale score for pain intensity ≤4 over ten points. 4) Diagnosis of BTP; 5) Patients who have signed the written informed consent to participate in the study.

Evaluation of study objectives
To assess the primary objective, the treatments received by the patient for chronic pain and BTP and their doses and administration route were recorded.
Information was collected on age, sex, race, weight, height, body mass index (BMI) and occupational status.
Social and family status was assessed using the Gijón scale [20,21]. It is a hetero-administered scale consisting of ve variables (family situation, economic situation, housing, social relations, and social support), each with ve possible categories. The categories are scored from 0 to 4 points, resulting in an overall score ranging from 0 to 20 points. The cut-off point for the detection of social risk is from 16 points on.
Medical history information and the patient's performance status (Karnofsky Performance Status) was recorded. The Karnofsky scale classi es patients into ten categories (0 and 100 points).
The date of cancer diagnosis (date of diagnostic biopsy) and the organ affected by cancer were recorded.
The main characteristics of background pain and BTP were recorded.
Quality of life was assessed using the EuroQoL-5D-5L, a generic questionnaire consisting of ve questions and a visual analogue scale with values between 0-100 millimetres (EQ-VAS). Each question has ve degrees and evaluates one dimension: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression [22].

Determination of sample size
The most restrictive variable for sample size determination was quality of life. The reference value for EQ-VAS in the Spanish population over 18 years of age is 75.0 points (SD 0.4) [23,24]. A sample of 222 patients would allow for the description of quality of life with a precision of 0.05 points and the power to nd differences was 96% with a two-sided alpha error of 0.05 (Sample Power, IBM-SPSS).

Statistical analysis
A descriptive analysis was performed of frequencies and percentages for the qualitative variables, with calculation of the mean, standard deviation, minimum, maximum values, and 95% con dence intervals, for quantitative variables.
Between-group comparisons were made using the Fisher test or the Chi 2 test if the variables were qualitative, and the Student's t-test or Wilcoxon test was used for quantitative variables.
A multivariate linear regression analysis was performed to explore the relationship between the EQ-VAS score and different patient characteristics: age, sex, Gijón scale score, BMI, Karnofsky Performance Status score, Frail scale score characteristics of BTP, time since diagnosis and location of cancer.
Statistical signi cance was considered at a value of 0.05. The IBM-SPSS version 27.0 statistical package was used throughout. The STROBE guidelines (www.strobe-statement.org) were followed to present the results of cross-sectional studies [2].

Results
A total of 240 patients participated in the study, with 222 patients with cancer selected for this analysis Table 1 shows their anthropometric, sociodemographic, and clinical characteristics. The percentage of patients with a Karnofsky score under 50 was 11.3% (n=25).
A total of 96.8% (n=215) received treatment for some comorbidity. About 79.7% (n=177) of the patients with cancer (n=222) were receiving treatment for cancer. The median time since cancer diagnosis was 14.7 months.

Frailty
All patients were frail, with a mean score on the Frail scale of 3.9 points (95% CI 3.8-4). There were no signi cant differences in the scores per cancer location.

Quality of life
The mean EQ-VAS score was 51.3 mm (95% CI 48.5-54), with a median of 50 mm. Figure 1 shows the distribution of patients in the categories of the ve dimensions of the quality-of-life questionnaire. No signi cant differences were observed in the EQ-VAS score between frail patients or according to the cancer location ( Figure 2).

Treatments for pain
For the treatment of background pain, 27 different drugs were administered in 215 patients with a total of 316 administrations. Table 3 details the drugs used and summarizes the mean daily doses administered to patients by compound and route of administration. A total of 65.8% of the active substance administered were opioids (208/316) and they were administered to 196 patients (88.3%).
The drugs used for the treatment of BTP, the doses administered, and the administration routes are described in Table 4 [27].
Predictable BTP was found in 35.1% of our frail patients, alike data (30.5%) described by Mercadante et al. 30 Difference on BTP mechanism was observed in both populations. Neuropathic pain was more frequently observed among our frail patients (16.7% vs 8.1%), while mixed pain was more common in those cancer patients of the referred study (71.8%, vs 31.1%) [26].
The higher the number of BTP episodes, the shorter the duration and the more gradual onset seen in frail patients. A pathophysiological process related to frailty status might underly these observations since a greater frequency of BTP episodes has been reported in patients with worse performance status [27]. Interestingly, the performance status of patients observed in both studies was similar, 63.2 (95% CI 61-65.4) versus 61.8 (SD 18.73) [26].
Among our frail patients, 79.7% of them were in active treatment for cancer, likewise 78% of patients in the reference study. This situation should be considered, since receiving this type of treatment could condition the administration of other treatments, such as, for example, for background pain and BTP [26].
Drugs for the treatment of chronic pain and BTP administrated to frail patients did not substantially differ on their doses and the frequency of administration (Tables 3 and 4), to the standard treatment for chronic pain. Regarding BTP treatment, in our study, 83.8% of frail patients with cancer received opioid treatment for BTP control, just as patients with cancer included in other studies [26,27]. Transmucosal immediate release fentanyl was administered to 68.5% of frail patients with cancer, the treatment of choice for BTP in cancer patients [14]. The doses of the treatments were within the range recommended in their prescribing information. On the other hand, our attention was caught by the fact that 10 patients used drugs that are not usually indicated for BTP but for basal pain therapy in standard practice, such as dexketoprofen, ibuprofen, metamizole or paracetamol, frequently complementary to the opioids administrated for the background chronic pain. In the Table 4 the number of administrations for such treatments were 28, but only 6 patients received only these drugs.
We observed BTP interference with daily activities in 93.7% of frail patients ( Figure 1) while Mercadante et al reported that in 86% of their patients [32]. Furthermore, they found that age, Karnofsky, BTP severity, short onset, and longer duration of BTP signi cantly interfered with daily activities of cancer patients. In our study, we were able to relate a poorer quality of life score (EQ-VAS) to a worse Karnofsky score in frail patients, already described in different studies [26,28,29]. In addition, our study found a signi cant association between quality of life and social exclusion in frail patients with cancer but due to the cross-sectional design we cannot know which one was the rst event.
The mean EQ-VAS score, as a measure of patient-perceived quality of life, was 51.3 mm (95% CI 48.5-54) in frail patients with BTP in our study, values well below those seen in the general Spanish population of the same age group, 69 mm ( Figure 2) [23]. In addition, this value is much more affected than in cancer patients, analysed in a review of 32 studies, who presented a value of 68.6 mm [31].
It is known that the occurrence of BTP has a signi cant impact on patient quality of life [32]. In the case of patients in our study, the low quality of life levels observed could be due to both BTP and frailty status, or to the interaction of both factors.
In Spain, the prevalence of frailty has been studied in six cohort studies, ranging from 2.4% to 27.3% in patients over 65 years of age [24]. However, in the setting of our study, conducted mainly in medical oncology units and in some pain and geriatric units, the prevalence of frailty is much higher [7,10,13], so the results of this study are relevant to standard clinical practice in these units.
This study has several limitations inherent to the cross-sectional observational design, which prevents causal relationships from being established.
Patients were classi ed as frail using the Frail scale [19], following national recommendations [24], so frail patients could be classi ed differently from other classi cation scales not based on the Fried criteria [1]. As this is a cross-sectional study, patients were included at different follow-up times after the onset of BTP. For this reason, we could not analyse which were the rst treatments for BTP in frail patients or their doses, which could differ at the start, and then be adjusted. Other treatments for patients with cancer such radiotherapy and brachytherapy play an important role in this setting of patients for the treatment of pain with many advantages in elderly and frail patients, but not collected in our study that was focused on pharmacological treatment [11,12] Conclusions This study concluded that some characteristics of BTP in frail patients with cancer differ from those reported in other studies. In addition, it has been seen that the treatments used for both chronic pain and BTP in frail patients are like those commonly prescribed in non-frail patients. Quality of life in frail patients has also been found to be severely impaired as compared to the population of patients with cancer and was related to a poorer performance status of the patient and poorer social-familial status. This relationship between frailty and impaired quality of life highlights the importance of the frailty assessment in all patients with BTP. Written informed consent was obtained from all patients. The study protocol was in accordance with the ethical standards described in the Declaration of Helsinki, and all participants' privacy rights were respected.

Consent for publication
Not applicable.

Availability of data and material
The study data is available upon reasonable request to the corresponding author.

Funding
The study was sponsored by Kyowa Kirin Farmacéutica, S.L. The sponsor was involved in the design of the study and the decision to submit the manuscript for publication.
Author's contributions GSG, JPC, SFS, contributed to the study concepts, the study design, data acquisition and manuscript review. AJJL, ACA and IHG contributed to the study concepts and design and the manuscript review. BSL contributed to the study design, quality control of data, statistical analysis, and manuscript preparation.
All authors have read and approved the manuscript.    Mean score of the quality-of-life EQ-VAS in frail patients with cancer and breakthrough pain. Comparison by cancer location and with the general healthy population for the same age group. __ __ __Mean EQ-VAS score in the general healthy population of the 65-74 years-old age group [30]. _ _ _ _ _ Mean EQ-VAS score for patients with cancer [31]. EQ-VAS: EuroQoL-5D-5L health-related quality of life questionnaire visual analogue scale; Value 0 means worst quality of life and value 100 means the best quality of life