Introduction

Bone metastasis is the most common cause of cancer-related pain [1, 2], and metastatic bone pain (MBP) is not only severe but also progressive in many patients [3]. As the tumor grows and tumor-induced bone remodeling progresses, skeletal-related events (SREs) occur. These include not only pain and “breakthrough pain” but also hypercalcemia, anemia, skeletal fractures, compression of the spinal cord, and decreased mobility and activity [4], all of which compromise the patient’s functional status, quality of life, and survival [5]. Furthermore, chronic pain is frequently associated with anxiety, depression and sleep disorders [6]. All these comorbidities have a negative impact on response to analgesic treatment, which in turn contributes to loss of quality of life. Hence, pain treatment is crucial for these patients [7].

As MBP responds poorly to non-steroidal anti-inflammatory drugs (NSAIDs), the WHO therapeutic guidelines for cancer pain recommend opioids [8]. These include drugs such as oxycodone, which is effective for pain relief in patients with bone metastases [3] and has a specific effect on neuropathic pain [9], a frequent complication in vertebrae metastases [3]. Other therapies are currently used for MBP: surgical therapy [3], palliative radiotherapy, and radiopharmaceuticals [10, 11]; bisphosphonates, which delay the onset and reduce the risk of SREs in patients with bone metastases [12]; denosumab, which is a fully monoclonal antibody recently approved for the prevention of SREs [13]; and rescue medication, i.e., supplemental analgesia for “breakthrough pain” [14].

We studied the association of pain management with evolution of overall Eastern Cooperative Oncology Group (ECOG) status in cancer patients with bone metastases, plus the degree of pain control according to analgesic treatment, to determine the quality of MBP management in the Spanish clinical setting.

Methods

Study Design

This was a 3-month, observational, prospective, multicenter study performed in Spain according to the principles of the revised Declaration of Helsinki in 2013. The protocol was approved by the ethics committee (Comité Ético de Investigación Clínica Regional del Principado de Asturias), and patients provided informed consent before inclusion in the study.

Eligible patients were ≥18 years old and diagnosed with cancer with bone metastases, with a life expectancy ≥3 months and pain intensity ≥4 on a numerical rating scale (NRS). Patients were excluded if they were scheduled for palliative surgery or the investigator considered them to have insufficient cognitive capacity to understand the study.

Outcome Measures

The primary endpoint of this study was to assess the correlation between pain relief and performance status of oncologic patients with bone metastases. Secondary endpoints were to assess the effect of pain control on sleep quality and functionality, as well as pain relief depending on the type of treatment. Basic demographic variables were collected at baseline, along with clinical variables. Patient-reported outcomes were collected at baseline and at months 1 and 3 during the study follow-up, including pain severity and the effect of pain control on sleep and functionality recorded on a 0–10 NRS. Performance status was evaluated at baseline and at months 1 and 3, and rated according to the ECOG scale.

The correlation between ECOG (measured by the change in performance status score and by the proportion of patients with better/stable/worse ECOG compared with baseline) and pain (measured by five pain parameters: (1) degree of change; (2) patients with improved/stable/worsened pain; (3) those with ≥20% of pain improvement; (4) those with ≥50% of pain improvement; and (5) those ending with a NRS ≤ 3) was also assessed.

Statistical Analysis

Statistically significant differences for ECOG, pain, functionality, and sleep between baseline and months 1 and 3 were determined using the Wilcoxon signed-rank test. The influence of pain management on ECOG during follow-up was assessed using the Spearman correlation, Kruskal–Wallis test and the Chi-Square test.

Results

Evolution of ECOG, Pain, Functionality, and Sleep

Overall, 579 patients (mean age 63.2 years; 53.2% male) were included in this study across 56 centers in Spain by October 1, 2010 (Table 1). The patients were sampled sequentially to avoid bias during patient selection. In these patients, performance status, pain, functionality, and sleep improved significantly over the study (Table 2). The correlation between the change in ECOG and pain was significant (p < 0.001), with significant improvements in pain and ECOG outcomes both observed at month 1. All patients with ECOG improvement over baseline also had a better evaluation of pain: most patients with stable ECOG also showed stable pain (75%), and most patients with worse ECOG over baseline also showed worse pain (77%; data not shown).

Table 1 Demographic and clinical characteristics of the study population
Full size table
Table 2 ECOG, pain, functionality, and sleep at different study time points
Full size table

Treatments

As shown in Table 2, seven different treatment modalities were used during the study. Chemotherapy was indicated in 68% of patients before the study: as first line in nearly half of these (43.4%), second line in 20.4%, and third or more lines in 18%. Most common therapies used were vinorelbine (11.7%), cisplatin (11.7%), carboplatin (11.4%), gemcitabine (10.7%) and taxotere (10.4%).

Mean doses of the four most frequent opioids used were: fentanyl 56.6 μg/h every 3 days; tramadol 101.7 mg every 8 or 12 h; oxycodone 22.1 mg every 12 h and morphine 46 mg every 12 h.

Bisphosphonates, almost exclusively zoledronic acid (96.7%), were received by 51.7% of patients before the study. Palliative radiotherapy to alleviate pain was given to 20.9%, with the spinal column being the most frequent application area (56.2%), followed by pelvis (30.6%) and femur (10.7%). Median dose was 30 Gy, with 10× 300 cGy in 57.9% of cases (data not shown). Finally, surgery and non-pharmacological treatment were considered for 4.1% and 0.7% of patients before the study, respectively.

Change in Treatment During Follow-Up

Chemotherapy and bisphosphonates remained stable over the study, and had no effect on ECOG and pain (p = 0.801 and p = 0.172; and p = 0.462 and p = 0470, respectively); however, there was an increase in palliative radiotherapy at baseline and a decrease at the following months, with no effect on ECOG or pain (p = 0.321 and p = 0.715; Table 2).

Non-opioids or coadjuvant administration decreased over the study, while opioids almost doubled at baseline and remained at the same high levels thereafter (Table 2). Mean doses of fentanyl and oxycodone increased slightly from 56 to 62 μg/h and 22 to 25 mg, respectively.

Rescue treatment was prescribed in 71.4% of patients at baseline (immediate release [IR] oxycodone capsules 60.4%, IR oxycodone solution 11.7%, morphine 13.8%, fentanyl 7.5%, and others 6.6%) and in 44.4% at month 1 (mainly oxycodone 66.8% or morphine 16%) and in a significantly smaller proportion of patients at month 3 (35%; p = 0.001). Laxatives and antiemetics were prescribed in 71.5% and 54.7% of patients, respectively, at baseline and in 71.8% and 44.4%, respectively, at month 1. Treatment at month 3 was similar to the treatment received at month 1.

Treatment compliance was high at the 1- and 3-month visits. At month 1, 76.5% of patients complied with radiotherapy, 92.6% with chemotherapy, 86.8% with bisphosphonates, and 95.3% with analgesic medication. At month 3, percentages were: radiotherapy 79.5%, chemotherapy 86.5%, bisphosphonates 79.9% and analgesic medication 93.2%.

Regarding adverse events (AE) to analgesic treatment, 37 (6.4%) patients experienced some baseline AEs, 53 (9.8%) patients had AEs at month 1, and 44 (9.3%) patients had AEs at month 3. The most frequent AEs, for the first and third months, were constipation (27.4% and 36.5%), nausea (23.8% and 19.5%), and somnolence (12.3% and 14.6%), respectively (Supplementary Figure 1). There were no grade 4 AEs, six patients were grade 3 (15.8%), with opioids AEs being the most common.

ECOG Evolution According to Treatment

Patients treated with a specific analgesic treatment (prescribed at baseline and with confirmed compliance at the specific time point analyzed) were compared with those who did not receive it in terms of ECOG. At both time points, patients treated with opioids showed significantly more favorable ECOG outcomes than those who did not (p < 0.001). Similarly, patients who did not need rescue medication showed more favorable ECOG outcomes than those who did (p < 0.001; Table 3).

Table 3 Patients (%) with better, stable or worse ECOG over baseline, according to whether or not they received opioid treatment or rescue treatment
Full size table

Pain Evolution According to Treatment

The pain parameters of patients treated with a specific analgesic treatment (prescribed at baseline and with confirmed compliance at the specific analyzed time point) were compared with those that did not. Opioids and rescue medication showed significant differences depending on whether they were used or not. Since opioids seemed to have an effect on pain evolution, the two most commonly used (oxycodone and fentanyl) were analyzed specifically (Table 4).

Table 4 Significance (p-value) of the difference in pain parameters between patients receiving opioid and rescue treatment or not
Full size table

At month 1, there was a significant difference in pain between patients taking and not taking opioids. Similarly, there were significant differences in all pain parameters in favor of patients not taking rescue medication. Mean decrease in pain (−3.9 vs. −3.2; p = 0.003) and the proportion of patients with ≥50% improvement in pain (72.9% vs. 59.3%; p = 0.007) were significantly larger in patients taking oxycodone than in those taking other opioids. The proportion of patients with ≥50% improvement in pain was significantly lower in patients taking fentanyl than in those taking other opioids (59.1% vs. 71.5%; p = 0.041).

After 3 months, mean reduction in pain was significantly greater in patients treated with opioids (−4.4 vs. −3.6; p = 0.014). Mean decrease in pain, and the proportion of patients with ≥20% improvement, ≥50% improvement, and those ending with NRS ≤ 3 were significantly in favor of patients not taking rescue medication. Mean decrease in pain (−4.7 vs. −3.6; p < 0.001) and the proportion of patients with ≥50% improvement (84.6% vs. 71.3%; p = 0.002) were significantly larger in patients taking oxycodone than in those taking other opioids. In terms of fentanyl, 83% of patients taking this drug decreased pain by ≥20%, and 66% by ≥50% vs. 93.2% (p = 0.015) and 83.1% (p = 0.005) of patients taking other opioids, respectively (Table 4).

Discussion

Good management of bone metastases has been shown to considerably relieve pain [7]. However, while baseline pain values in this study (mean NRS = 6.5) were moderate–severe, and up to 43.5% of patients had neuropathic pain, which is difficult to treat [15], only 59% of patients were being treated with opioids. The WHO analgesic ladder recommends administration of a minor opioid in the second step (for moderate pain) and a major opioid in the third step (for severe pain) [16].

During the study, the percentage of patients treated with chemotherapy and bisphosphonates hardly varied, whereas palliative radiotherapy and non-opioids or coadjuvants decreased from baseline to follow-up visits. It has been reported that bisphosphonates and radiation therapy have beneficial effects on reducing metastatic bone pain [17]. However, in the present study, even though the number of patients using these therapies reduced with treatment, these changes were not statistically significant.

Opioids have been successfully used for neuropathic pain since the early 1990s [18], with oxycodone specifically being reported as a beneficial drug in benign neuropathic pain, and as efficacious as morphine in treating oncologic pain [9]. Although the European Society for Medical Oncology (ESMO) still recommends morphine as the first line opioid for cancer patients with moderate to severe pain [19], current European Association for Palliative Care (EAPC) recommendations suggest that the analgesic efficacy and adverse event profile of morphine, oxycodone and hydromorphone are similar in these patients [20]. The results of the present study show that the percentage of patients treated with opioids almost doubled during the study compared with previous treatments, baseline, and follow-up.

Significant improvements in pain, sleep, and functionality found during the first month of treatment in the present study were similar to another Spanish oncologic neuropathic pain study [21]. In that study, 88% of patients were treated with oxycodone, and pharmacologic cancer pain treatment had a greater effect in patients with metastases than in those with no metastases, evidencing its importance in metastatic cancer pain. Analgesic treatment response depended on the specific treatment used, with patients on oxycodone or other opioids showing a greater improvement in pain than patients on treatment other than opioids.

In addition to performance status being dependent on patients’ overall response, our results show that ECOG improvement is closely related to pain. The response to analgesic treatment for palliation of MBP has been found to be significantly related to performance status [22]. Furthermore, the association between pain and performance status seems to be reciprocal, since patients with poor performance status (ECOG ≥ 2), among other variables, are more likely to have SREs, including pain [23]. The change from previous treatment consisted in a decrease in non-opioids or coadjuvants and a 2-fold increase in opioids, due to oxycodone increase (6-fold). Therefore, pain and ECOG improvements observed at months 1 and 3 seem to be the result of the increase in opioid administration. Despite the statistically significant difference in the performance status and pain scores with treatment, the results need to be interpreted with caution considering the validity of measurement methods used. Furthermore, although radiotherapy treatment increased at baseline, it decreased subsequently, and hence would not explain the improvement observed during the follow-up period.

The strengths of this study include the number of patients included in this analysis as well as the observational nature of this study, which allows for the determination of this effect in a real-world setting. However, the unselected nature of the population could mean that there are confounding factors influencing the patients’ perception of their functional status. A major limitation of our study is the lack of information on the physicians' clinical judgments for the choice of treatment and that nearly 80% of patients received baseline oxycodone during the study, indicating that most patients were switched from other opioids to oxycodone which may have introduced bias. Although no differences were observed between oxycodone and the rest of opioids for ECOG, significant differences were observed in pain outcomes. Thus, decrease in pain was significantly greater in patients taking oxycodone, while the proportion of patients with a ≥50% improvement in pain was significantly larger in patients taking oxycodone, and significantly smaller in those taking fentanyl. This suggests that, in patients with MBP, oxycodone, specifically, might improve pain more than other opioids. Furthermore, most of patients received non-opioids or adjuvant analgesics and either chemotherapy or radiotherapy during the study. However, prescription changes of these treatments throughout the observation period did not show statistically significant effects on pain improvement, whilst only the increment of opioid treatment revealed a statistically significant impact on pain relief. Another limitation of this study is that the effects of treatment were only investigated for 3 months; studies investing the long-term influence of pain treatment on clinical outcomes in patients with MBP are warranted.

In conclusion, the current management of MBP in Spain is poor. In fact, pain, sleep, functionality, and ECOG improvements were achieved after changing treatment to a wider range of opioids. Pain and ECOG outcomes are directly related. Opioid treatment seems to have a favorable effect on ECOG and pain evolution in patients with MBP, while the opposite is true of the need for rescue treatment.