You are here

  • Home
  • Archive
  • Volume 6, Issue 3
  • Characteristics of breakthrough cancer pain and its influence on quality of life in an international cohort of patients with cancer

Article Text

Article menu
Download PDFPDF
Research
Characteristics of breakthrough cancer pain and its influence on quality of life in an international cohort of patients with cancer
  1. Marianne Jensen Hjermstad1,2,
  2. Stein Kaasa2,3,
  3. Augusto Caraceni2,4,
  4. Jon H Loge1,5,
  5. Tore Pedersen6,7,
  6. Dagny Faksvåg Haugen2,8,9 and
  7. Nina Aass1,10
  8. on behalf of the European Palliative Care Research Collaborative (EPCRC)
    1. 1Department of Oncology, Regional Advisory Unit for Palliative Care, Oslo University Hospital, Oslo, Norway
    2. 2Department of Cancer Research and Molecular Medicine, Faculty of Medicine, European Palliative Care Research Centre, Norwegian University of Science and Technology (NTNU), Trondheim, Norway
    3. 3Department of Oncology, St Olavs Hospital, Trondheim University Hospital, Trondheim, Norway
    4. 4Department of Palliative Care, Pain Therapy and Rehabilitation Unit, Fondazione IRCCS, Istituto Nazionale Dei Tumori, Milano, Italy
    5. 5Department of Behavioural Sciences in Medicine, University of Oslo, Oslo, Norway
    6. 6Bjørknes University College, Oslo, Norway
    7. 7National Institute of Occupational Health, Oslo, Norway
    8. 8Regional Centre of Excellence for Palliative Care, Western Norway, Haukeland University Hospital, Bergen, Norway
    9. 9Department of Clinical Medicine K1, University of Bergen, Bergen, Norway
    10. 10Faculty of Medicine, University of Oslo, Oslo, Norway
    1. Correspondence to Dr Marianne Jensen Hjermstad, Department of Oncology, Regional Advisory Unit for Palliative Care, Oslo University Hospital, Ullevål, Box 4956, Nydalen, Oslo 0424, Norway; marianne.j.hjermstad{at}ntnu.no

    Abstract

    Objectives Breakthrough cancer pain (BTP) represents a treatment challenge. Objectives were to examine the prevalence and characteristics of BTP in an international sample of patients with cancer, and to investigate the relationship between BTP and quality of life (QoL).

    Methods This was an observational cross-sectional multicentre study. Participating patients completed self-report questionnaires on a touch-screen laptop computer, including the Brief Pain Inventory, Alberta Breakthrough Pain Assessment Tool (ABPAT) and European Organisation for Research and Treatment of Cancer 30-item Core Quality of Life Questionnaire (EORTC QLQ-C30). The study was performed in 17 centres in 8 countries and involved 4 languages (Norwegian, Italian, German and English).

    Results Records from a convenience sample of 978 patients with advanced cancer were analysed; mean age was 62.2 years, 48.3% were women and 84.4% had metastatic disease. A total of 296 patients (30%) had no pain, defined as worst pain in the past 24 hours <1 on a 0–10 scale. Of the 682 patients with a pain score ≥1, 393 (58%) reported no BTP on the screening item, while 289 (30%) confirmed flare ups of BTP. Patients with BTP reported significantly higher pain intensity scores (<0.001) than patients without BTP; 57.1% of patients rated BTP at its worst as being severe: ≥7 on a 0–10 scale. Time from onset to peak intensity was <10 min for 42.9%, and average time to pain relief was 27.1 min. BTP was commonly triggered by medication wearing off (28%). Patients with BTP had significantly worse mean outcomes on 10 of 15 functional and symptom scales of the EORTC QLQ-C30 (<0.001). Severe pain intensity in the last week was a powerful predictor of BTP (OR 4.1) and poor QoL (OR 1.9).

    Conclusions BTP is highly prevalent with prolonged episodes despite analgaesics, and has a pervasive impact on QoL. Patients reporting high pain intensity should be carefully evaluated for BTP and efficacy of analgaesic treatment, to provide optimal pain management and improve QoL.

    Trial registration number NCT00972634; Results.

    • Cancer
    • Pain
    • Quality of life
    • Clinical assessment
    • Breakthrough cancer pain
    • Pain assessment

    https://doi.org/10.1136/bmjspcare-2015-000887

    Statistics from Altmetric.com

      Request Permissions

      If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

      Introduction

      Control of pain and other symptoms in patients with limited life expectancy is critical in cancer care. Breakthrough cancer pain (BTP) or transient episodes of severe pain represent a particular treatment challenge. A range of definitions of BTP have been presented, but one of the yet most frequently cited is from 1990, defining BTP as ‘a transitory increase in pain to greater than moderate intensity, which occurs on a baseline pain of moderate intensity or less’.1 Another definition is one developed by the Association for Palliative Medicine of Great Britain and Ireland (APM): ‘a transient exacerbation of pain that occurs either spontaneously, or in relation to a specific and predictable or unpredictable trigger, despite relatively stable and adequately controlled background pain’.2

      A literature review showed that BTP was described in relation to a stable, treated background pain of mild or moderate intensity, in the majority of the 51 papers that were reviewed.3 Terms such as transient pain, intermittent pain, episodic pain and incident or incidental pain, are also used, but the review found that BTP was by far the most common term.3 BTP is associated with more advanced disease, higher pain intensity (PI) and poorer quality of life (QoL).4–6 The reported prevalence rate of breakthrough cancer pain varies across studies, from 19% to more than 80%.4 ,5 ,7–9 A recent systematic review presented pooled analysis of the published literature and concluded that 50% of patients with cancer with pain also had BTP, but also that the heterogeneity was great, across settings and patient groups.10

      Pain is understood to be a multidimensional symptom, and varying assessment of different pain dimensions may contribute to this variation. An Expert Working Group of the European Association for Palliative Care (EAPC)11 has recommended the use of multidimensional tools, such as the Brief Pain Inventory—Short Form (BPI-SF),12 to promote a standardised approach to pain assessment in advanced cancer, in line with later proposals.13

      The temporal aspect of pain is poorly captured in general pain assessment tools,3 ,14 ,15 despite the recommendations of including this dimension.3 ,13–17 However, only a few specific BTP tools exist. The temporal dimension was not included in the initial tool: the Breakthrough Pain Questionnaire, developed by Portenoy and Hagen.1 Therefore, the Alberta Breakthrough Pain Assessment Tool (ABPAT) was developed, and partly validated, to remediate this shortcoming.14 The ABPAT encompasses important aspects of BTP, such as frequency, severity and location, as well as multiple items on the temporal variations that help distinguish this type of pain from uncontrolled pain and other clinical scenarios of pain, each of which requires different treatment approaches.

      The current study was undertaken by the European Palliative Care Research Collaborative (EPCRC) (2006–2010),18 aiming to examine the feasibility of using computers for registration of medical and patient-reported data; the EPCRC Computerised Symptom Assessment study (EPCRC-CSA).19 A large number of assessments was included, as a first step in the development of a shorter, dynamic tool, as described elsewhere.19 The main objective of the present report was to describe the prevalence and characteristics of BTP in a large international sample of patients with advanced cancer, using patient-reported responses to questions in the ABPAT. Secondary aims were to compare QoL, symptoms and functioning in patients with BTP compared with patients reporting pain but not BTP, and to identify other medical and sociodemographic factors associated with BTP and QoL.

      Methods

      The international multicentre study recruiting a convenience sample of patients started in Norway, in October 2008, with England, Austria, Germany, Switzerland, Italy, Canada and Australia consecutively entering the study, until December 2009. Seventeen centres participated: 11 university clinics, 4 regional hospitals, 1 local hospital and 1 hospice. Recruitment took place in general oncology, medical, palliative care (PC) and hospice units. Potentially eligible inpatients or outpatients meeting the inclusion criteria: verified cancer diagnosis, incurable metastatic/locally advanced disease, age ≥18 years and ability to provide written informed consent, were approached by the study personnel, for information and consent. Patients with language problems and/or severe cognitive impairment as judged by the study personnel at this point, were not included.19

      Data collection was performed using touch-sensitive computers, and consisted of two parts, as described in detail previously.19 The health personnel part included sociodemographic and medical variables, Karnofsky Performance Status20 and Mini-Mental State Examination (MMSE) for cognitive function.21 The MMSE maximum score is 30 points, with <24 indicating a potential cognitive impairment. To avoid subgroup analyses on small samples, we did not adjust for educational background. The patient part had 71 questions, among others, the European Organisation for Research and Treatment of Cancer 30-item Core Quality of Life Questionnaire (EORTC QLQ-C30)22 and two PI questions from the BPI,12 regarding worst and average pain in the past 24 hours (see online supplementary figure S1). These were supplemented by worst and average pain in the past 2 days and last week;12 all scored on a 0–10 numerical rating scale: no pain to pain as bad as you can imagine. If the patient had a worst pain score ≥1 past 24 hours, the computer automatically directed the patient to a set of questions on PI, and a binary screening question about BTP that read: ‘Have you had flare-ups of breakthrough cancer pain in the last 24 hours?’. The instruction given for this BTP screening question was: ‘Breakthrough cancer pain can be defined as a brief flare-up of pain. It can be a flare-up of the usual, steady pain you always experience (your baseline pain) OR it can be a pain that is different from your baseline pain’, formulated by the developers of the ABPAT.14 If patients confirmed having episodes of BTP, they were asked to complete the ABPAT.

      If the patients answered ‘no’ on the BTP screening question about flare ups of pain, they were presented a few questions from the ABPAT concerning the perceived cause of their pain, actions giving pain relief and time to onset of pain alleviation after medication. For these patients, the ABPAT questions were reworded using the word pain, not breakthrough cancer pain, to assess their experience with pain relief.

      Ethics approval was obtained at each site by the local study coordinator and written informed consent was obtained before any data were collected. Unique identification numbers were used to assure patient anonymity. The trial was registered in the ClinicalTrials.gov database (No. NCT00972634).

      Instruments

      The EORTC QLQ-C30, V.3.0,22 was used for QoL assessment. This cancer-specific 30-item self-report questionnaire evaluates the patient's last week, and consists of five functional scales: physical, role, cognitive, emotional and social, three symptom scales: fatigue, nausea/vomiting and pain, and six single items: dyspnoea, insomnia, appetite loss, constipation, diarrhoea and financial problems. These are scored from 1=not at all to 4=very much, while the last two items assess QoL and overall health on a scale from 1=very poor to 7=excellent. Combining these two gives a global health status/QoL score. All scores were transformed into continuous scales from 0 to 100, as described in the EORTC QLQ-C30 Scoring Manual.23 Although the clinical significance of the numeric rating scores is largely undefined, a mean score difference of 5–10 is usually regarded as a small but clinically noticeable change for the patients, a change between 10 and 20 as moderate and above 20 as a large clinical change.24 ,25

      The ABPAT was developed because of a need for validated assessment tools for BTP in cancer.14 ,26 It was intended for research but has also been recommended for clinical use.26 ,27 The initial validation study used a Delphi process with national and international expert panels, and input from patients with cancer with BTP.26 The ABPAT contains 15 questions with 3 having two items, resulting in a total of 18 questions. The questions cover the following domains: relationship to baseline pain, last time experienced, frequency, peak intensity of the BTP: 0=no pain, 10=worst possible pain and on an ordinal scale; 1=mild, 2=moderate, 3=severe, location (body map), pain quality with specified descriptors, time from onset to peak intensity, time from onset to end of episode, cause of BTP (triggers), predictability, general relief, relief from BTP medication, satisfaction with BTP medication, onset of pain relief and satisfaction with onset of pain relief.

      All questions were forward–backward translated into the primary language of the participating centre as necessary, followed by pilot testing in 5–10 patients.

      Statistical analyses

      Descriptive statistics were used to characterise the overall sample and the three pain groups: (1) no pain in the past 24 hours, (2) pain but no BTP (no BTP) and (3) BTP. Differences between the no BTP and BTP groups were evaluated for statistical significance, using logistic regression for categorical variables and analysis of variance (ANOVA) for continuous variables. Multiple logistic regression was performed in patients who reported pain ≥1 in the past 24 hours, to identify factors associated with BTP and the global health status/QoL scale, respectively. To ease the interpretation of the results, the BPI questions were dichotomised with scores from 7 to 10 representing severe pain, while scores from 1 to 3 on the 1–7 global health status/QoL scale represented poor QoL. The modelling approach consisted of first entering the potential confounders and variables with p values <0.15, which were identified from the univariate analyses, as a group. Backward selection was then used to eliminate variables that did not retain a p value <0.10 in the multivariate analysis. ORs with corresponding 95% CIs were used to report the results, adjusted for the other factors in the model. The sample size of 682 used in the subgroup analysis had 80% power to detect ORs as small as 1.35 at a 5% significance level. The accuracy of the regression models was assessed using the C-statistic, whereby values >0.7 are considered as having reasonable accuracy.28 Sample size calculations were not performed in this cross-sectional, descriptive study, as hypothesis testing was not a study objective. IBM SPSS Statistics V.22 (Armonk, New York, USA) was used for analyses. The significance level was set at p values <0.05.

      Results

      Patient characteristics

      Of the 1070 patients who agreed to participate, 4 withdrew their informed consent and 15 had no data recorded because of technical failure. Among the 1051 eligible patients, 40 had missing data for the PI and ABPAT, and 33 others had missing data for ABPAT only. Thus, 978 (93%) patients were included in the analyses, ranging from 30 (Germany) to 491 patients (Norway) across the eight countries, and from 10 (Australian centre) to 165 (Norwegian centre) across centres.

      A total of 296 (30%) patients reported no pain in the past 24 hours, 393 (40%) had pain but no BTP and 289 (30%) had BTP. The proportion of patients within each group varied significantly across countries (p<0.001); from 2% in Italy to 56% in Austria reporting no pain, and from 15% to 46% reporting pain but no BTP in Canada and Italy, respectively. The highest percentage of patients with BTP per country was found in Canada: 75%; Italy: 53%; the UK: 37%, in contrast to 14% and 17% in Austria and Germany, respectively.

      Patient characteristics are presented by pain groups in table 1. The overall mean age was 62.2 years (SD 12.4, range 18–91) and 48.3% were women. The majority (65.5%) were married/cohabiting, and 30.2% had postsecondary education. The sample consisted of both, inpatients (57.0%) and outpatients (43.0%), and the majority (84.4%) had metastatic disease. The four most common diagnoses in all groups were gastrointestinal (26.6%), breast (17.0%), genitourinary (16.7%) and respiratory (16.7%) cancers. Comorbid conditions were similarly distributed across groups, with heart disease being the most prevalent (table 1).

      View this table:
      Table 1

      Patient characteristics by pain groups, N=978

      Several differences were observed between the groups (table 1). Patients under the age of 65 years, especially women, were more likely to report BTP. This is explained by the high proportion of breast cancer in this age-group. Patients with no pain were more often treated with chemotherapy than were patients in the other two groups. Significantly, more than two-thirds of the patients with BTP were functionally impaired (Karnofsky score <80) and ∼50% had bone metastases. Opioids were used by 68.2% of the patients with BTP, and another 19.6% used non-opioid pain medications. Drug use among patients with BTP was significantly higher than among patients who had pain but no BTP, notably opioids, laxatives and antiemetics.

      Pain and QoL

      Statistically significant differences were found for all PI scores at inclusion, with notably higher mean levels for patients with BTP compared with patients without BTP (table 2). Also, patients with BTP had significantly worse mean scores on 10 of the 15 functional and symptom scales of the EORTC QLQ-C30, than did patients who had pain but no BTP (table 3). Differences were in the range of 5–10 for nine outcomes, and exceeding 10 for pain.

      View this table:
      Table 2

      Mean pain intensity scores and pain timing categories for patients with and without breakthrough cancer pain (BTP)

      View this table:
      Table 3

      EORTC QLQ-C30 scores by no BTP versus BTP

      Characterisation of BTP

      A majority of patients (59.5%) reported BTP to be a brief flare up of their baseline pain. The median number of BTP episodes in the past 24 hours was 3 (range 1–30), with six patients reporting more than 12 episodes. Sixty-seven per cent claimed that this was their usual frequency of BTP episodes. Three most commonly identified triggers of BTP were walking, movement in bed and scheduled pain medication wearing off: 33%, 31% and 28%, respectively, while 27% reported no specific triggers of BTP. Patients described their BTP, using multiple descriptors in the ABPAT—most commonly as aching (42%) or stabbing (37%). The mean peak intensity score of the BTP was 7.1 (SD 2.1) on the 0–10 scale and 57% of patients rated BTP at its worst as being severe. The time from onset to peak PI was under 10 min in 42.9% of patients, 10 to 30 min in 25.0% and >30 min in 9.8%. The most common (51.3%) intervention for BTP was ‘as needed’ BTP medication, while 33.4% achieved relief through scheduled pain medication. Onset of relief occurred on average at 27.1 min (SD 18.7). Twelve per cent of patients with BTP reported that the time from onset to end of episode was <10 min, while 25.7% indicated that it took >30 min. Good relief from BTP medication was achieved in 72.5% of patients.

      Of the 393 patients with no BTP who answered the modified ABPAT items, 28% attributed scheduled pain medication wearing off to their pain, while 18% and 16% said that pain was caused by sitting and coughing, respectively, relative to 21% and 16% in the BTP group. Lying down was the second best pain action for pain relief: 30%; while use of scheduled analgaesics alleviated the pain in 36.6%, with an average of 37 min before relief.

      Factors associated with BTP and the global health status/QoL scores

      Of the 682 patients with pain, 19.8% reported poor global health status/QoL. The PI questions were highly correlated with each other and therefore required preselection. All three questions that asked about ‘pain on the average’ had weak associations with both BTP and global health status/QoL, and were excluded from subsequent analyses. Among the three remaining questions, both BTP and global health status/QoL were most significantly associated with patients who reported their worst pain as severe ≥7–10, (OR=4.7, 95% CI 3.4 to 6.5, p value <0.001 and OR=2.6, 95% CI 1.8 to 3.8, p value <0.001, respectively). Two factors were significantly associated with both BTP and poor global health status/QoL (table 4): pain ‘at its worst’ in the last week categorised as severe, and taking medication for pain. Patients with severe pain had 4.1 times the odds of having BTP (p value <0.001) and 1.9 times the odds of having poor global health status/QoL (p value <0.008) than patients with less severe pain did. Functional impairment (Karnofsky score <80) remained significantly associated with poor global health status/QoL but was no longer significant in the multivariate analysis. Taking antidepressants and being an inpatient were only associated with poor global health status/QoL, and not with BTP.

      View this table:
      Table 4

      Associations between breakthrough cancer pain (BTP) and poor global health status/quality of life (GHS/QoL) score, multivariable models

      Discussion

      The present study confirms and expands on observations from earlier studies on BTP by surveying patients with advanced cancer from 17 centres in eight countries and involving four languages. The results document that BTP is sudden, severe and highly prevalent, and has a powerful and pervasive impact on virtually every dimension of QoL. The present study also adds to the body of research by identifying that severe pain ‘at its worst’ in the last week is a powerful predictor of both BTP and poor global health status/QoL.

      We found that the prevalence of BTP varied between countries, from 14% to 75%, in contrast to other studies.8 ,29 There may be different reasons for this, in particular, that the other studies were specifically designed to characterise BTP, while ours was not. The study by Davies et al8 employed a designed diagnostic algorithm for screening and characterising BTP in 1000 European patients in 13 countries. Furthermore, these patients were receiving care from specialist pain or PC teams, settings that have reported higher proportions of patients with BTP compared with general oncology wards.9 The fact that PC treatment was not an inclusion criterion in the present study, may have led to the variation in the results. In some countries, patients are referred to PC very close to end-of-life, when the symptom burden is high, whereas the PC approach with its systematic focus on symptom assessment is well integrated into mainstream oncology in other countries.

      The Bedard et al29 study also used a specifically developed registration tool for presence and management of BTP, and found no differences in the incidence of BTP when comparing data from 94 Canadian and 1000 European PC patients. On the other hand, an international survey published in 2004 demonstrated that physicians from English-speaking countries were significantly more likely to report BTP than were physicians from other countries.7 Although the prevalence of BTP in our study was highest in patients in Canada, and also high in the UK, no conclusion on this can be drawn, as we used patient-self-report for BTP. A recent study30 showed that initial disparities between patients' and clinicians’ perceptions of BTP were highest at the initial consultation, and subsided during a therapeutic relationship, primarily based on clinical considerations. Thus, the cross-sectional design in this study may have influenced the BTP variability, as the length of the treatment trajectory cannot be ascertained.

      A simple question about worst pain in the last week showed a powerful association with BTP and impaired QoL in the present study. Although a time frame of 1 week stood out as the strongest predictor, we strongly believe that high PI regardless of time window should prompt clinicians to perform a more in-depth pain assessment and re-evaluate the pain management, including the use of antitumour measures such as palliative radiotherapy. The measure of PI, supplemented by a multidimensional assessment, is the gold standard to guide pain management, in line with recommendations defining PI, pain relief and temporal pattern of pain as the most relevant outcomes in clinical practice.13 Associations between high PI and BTP have also been documented by others,8 ,31 ,32 and it is highly likely that the high pain levels also reflect BTP episodes. Also, the level of PI at the initial assessment has been reported as a significant predictor of the cancer pain complexity and time needed to obtain stable pain control,33 and is, as such, a key feature for pain classification.16

      BTP research has been challenged by lack of consensus on standard language and taxonomy.34 A review on pain classification from the EPCRC identified BTP as a subdomain included in three of the formal pain classification systems, but all used different terminology for BTP.35 In particular, the recommendations of the Science Committee of the Association for Palliative Medicine did not contain end-of-dose failure in BTP.2 In our study, aspects related to end-of-dose failure are noteworthy, as 28% of both patients with and without BTP attributed their pain to medication wearing off. This once again emphasises the need to perform regular assessments of PI, and to routinely screen for the presence of BTP and, concurrently, other aspects of the patient's experience, including QoL.36 ,37 For example, BTP in association with end-of-dose failure of opioids indicates that the cancer pain probably is undertreated and that BTP episodes could be reduced with more attention to dosing of scheduled opioids.

      In a previous multicountry survey of cancer pain patients,4 only one-third of patients experiencing BTP or inadequate pain relief were treated with additional analgaesic medication. Only 50% reported that they had good QoL, whether or not they had BTP, and patients commonly reported that pain interfered with function. In a second study, patients with cancer-related BTP were found to have greater interference of function by pain, lowered mood and decreased QoL.8 ,32 The present study confirms this by affirming the strong and negative impact of pain on functioning and QoL. As poorer QoL was reported more often by younger inpatients, more consideration may need to be given to these patients.

      A limitation to the EPCRC-CSA study is related to the cross-sectional study design, the inclusion of a convenience sample rather than consecutive patient enrolment, that most patients were recruited from university clinics and the length of the computerised assessment. The large proportion of patients with high Karnofsky performance scores may indicate that the frailest patients were not included, and lack of information about patients who were not approached or who declined participation may restrict the generalisability of the results. On the other hand, the large sample size, an overall compliance of 95%19 and the use of well-accepted assessment tools, increase the representativity. Further, our results show that BTP is a prevalent clinical problem across disease stages, level of functioning, patient characteristics and countries.

      The use of ABPAT to characterise BTP may be regarded as a limitation, as this tool lacked clinical validation at study start.3 Also, the dichotomous screening question had not been tested for its appropriateness. However, the clinical relevance of ABPAT has been demonstrated, as a later Italian study examined its validity in a national multicentre study in patients with advanced cancer.38 The ABPAT was well-accepted and comprehensive, albeit relatively long and, as such, less well suited for routine clinical work, as originally pointed out.3 ,14 Another Italian study developed a modified ABPAT-version to assess BTP characteristics through patient interviews.39 Finally, the newly developed Breakthrough Pain Assessment Tool (BAT) for BTP assessment and management40 concluded that the brevity may increase the applicability but reduce the comprehensiveness for additional validation studies.

      Suboptimal pain management is still frequently reported.41 ,42 A recent review, however, investigating temporal trends in cancer pain treatment by comparing studies published before and after 2007, concluded that the proportion of undertreated patients has decreased over time, but that still about 1/3 of patients do not receive adequate pain treatment.43 In our opinion, and because of the great heterogeneity in cancer-related BTP,8 ,10 this represents an obligation to improve the quality of pain assessment, with specific focus on PI and temporal variations, both of which are associated with impaired function and QoL. Despite the potential limitations, we believe our results provide directions on ways to improve pain assessment and management of BTP in patients with advanced cancer.

      Conclusion

      BTP remains a severe and highly prevalent problem in patients with advanced, incurable cancer. While many patients obtain relief with medication, a large proportion have prolonged BTP episodes despite analgaesics. The present study suggests that high PI levels should lead to in-depth pain assessment, in order to reveal BTP and other important pain characteristics. The seemingly high prevalence of end-of-dose failure suggests a need for professional as well as patient and carer education to optimise pain management and thereby QoL.

      Acknowledgments

      The authors express their gratitude to Misha Eliasziw, Tufts University, Department of Public Health and Community Medicine, Boston, Massachusetts, USA; Sunita Ghosh, PhD, University of Alberta; and Neil A Hagen, Division of Palliative Medicine, University of Calgary, for assisting in the first versions of this manuscript.

      References

      1. ↵
        1. Portenoy RK,
        2. Hagen NA
        . Breakthrough pain: definition, prevalence and characteristics. Pain 1990;41:273–81. doi:10.1016/0304-3959(90)90004-W
        OpenUrlCrossRefPubMedWeb of Science
      2. ↵
        1. Davies AN,
        2. Dickman A,
        3. Reid C, et al
        . The management of cancer-related breakthrough pain: recommendations of a task group of the Science Committee of the Association for Palliative Medicine of Great Britain and Ireland. Eur J Pain 2009;13:331–8. doi:10.1016/j.ejpain.2008.06.014
        OpenUrlCrossRefPubMedWeb of Science
      3. ↵
        1. Haugen DF,
        2. Hjermstad MJ,
        3. Hagen NA, et al
        . Assessment and classification of cancer breakthrough pain. A systematic literature review. Pain 2010;149:476–82. doi:10.1016/j.pain.2010.02.035
        OpenUrlCrossRefPubMedWeb of Science
      4. ↵
        1. Breivik H,
        2. Cherny N,
        3. Collett B, et al
        . Cancer-related pain: a pan-European survey of prevalence, treatment, and patient attitudes. Ann Oncol 2009;20:1420–33. doi:10.1093/annonc/mdp001
        OpenUrlAbstract/FREE Full Text
      5. ↵
        1. Raj SX,
        2. Thronaes M,
        3. Brunelli C, et al
        . A cross-sectional study on prevalence of pain and breakthrough pain among an unselected group of outpatients in a tertiary cancer clinic. Support Care Cancer 2014;22:1965–71. doi:10.1007/s00520-014-2178-3
        OpenUrlPubMed
      6. ↵
        1. Portenoy RK,
        2. Bruns D,
        3. Shoemaker B, et al
        . Breakthrough pain in community-dwelling patients with cancer pain and noncancer pain, part 2: impact on function, mood, and quality of life. J Opioid Manag 2010;6:109–16. doi:10.5055/jom.2010.0010
        OpenUrlCrossRefPubMed
      7. ↵
        1. Caraceni A,
        2. Martini C,
        3. Zecca E, et al
        . Breakthrough pain characteristics and syndromes in patients with cancer pain. An international survey. Palliat Med 2004;18:177–83. doi:10.1191/0269216304pm890oa
        OpenUrlAbstract/FREE Full Text
      8. ↵
        1. Davies A,
        2. Buchanan A,
        3. Zeppetella G, et al
        . Breakthrough cancer pain: an observational study of 1000 European oncology patients. J Pain Symptom Manage 2013;46:619–28. doi:10.1016/j.jpainsymman.2012.12.009
        OpenUrlCrossRefPubMedWeb of Science
      9. ↵
        1. Greco MT,
        2. Corli O,
        3. Montanari M, et al.
        , Writing Protocol Committee; Cancer Pain Outcome Research Study Group (CPOR SG) Investigators. Epidemiology and pattern of care of breakthrough cancer pain in a longitudinal sample of cancer patients: results from the Cancer Pain Outcome Research Study Group. Clin J Pain 2011;27:9–18. doi:10.1097/AJP.0b013e3181edc250
        OpenUrlCrossRefPubMedWeb of Science
      10. ↵
        1. Deandrea S,
        2. Corli O,
        3. Consonni D, et al
        . Prevalence of breakthrough cancer pain: a systematic review and a pooled analysis of published literature. J Pain Symptom Manage 2014;47:57–76. doi:10.1016/j.jpainsymman.2013.02.015
        OpenUrlCrossRefPubMed
      11. ↵
        1. Caraceni A,
        2. Cherny N,
        3. Fainsinger R, et al
        . Pain measurement tools and methods in clinical research in palliative care: recommendations of an Expert Working Group of the European Association of Palliative Care. J Pain Symptom Manage 2002;23:239–55. doi:10.1016/S0885-3924(01)00409-2
        OpenUrlCrossRefPubMedWeb of Science
      12. ↵
        1. Daut RL,
        2. Cleeland CS,
        3. Flanery RC
        . Development of the Wisconsin Brief Pain Questionnaire to assess pain in cancer and other diseases. Pain 1983;17:197–210. doi:10.1016/0304-3959(83)90143-4
        OpenUrlCrossRefPubMedWeb of Science
      13. ↵
        1. Kaasa S,
        2. Apolone G,
        3. Klepstad P, et al
        . Expert conference on cancer pain assessment and classification, the need for international consensus: working proposals on international standards. BMJ Support Palliat Care 2011;1:281–7. doi:10.1136/bmjspcare-2011-000078
        OpenUrlAbstract/FREE Full Text
      14. ↵
        1. Hagen NA,
        2. Stiles C,
        3. Nekolaichuk C, et al
        . The Alberta Breakthrough Pain Assessment Tool for cancer patients: a validation study using a Delphi process and patient think-aloud interviews. J Pain Symptom Manage 2008;35:136–52.
        OpenUrlCrossRefPubMedWeb of Science
      15. ↵
        1. Hjermstad MJ,
        2. Gibbins J,
        3. Haugen DF, et al
        . Pain assessment tools in palliative care: an urgent need for consensus. Palliat Med 2008;22:895–903. doi:10.1177/0269216308095701
        OpenUrlAbstract/FREE Full Text
      16. ↵
        1. Knudsen AK,
        2. Brunelli C,
        3. Klepstad P, et al
        . Which domains should be included in a cancer pain classification system? Analyses of longitudinal data. Pain 2012;153:696–703. doi:10.1016/j.pain.2011.12.005
        OpenUrlCrossRefPubMed
      17. ↵
        1. Foley KM
        . Acute and chronic cancer pain syndromes. In: Doyle D, Hanks G, Cherny N, et al. eds. Oxford textbook of palliative medicine. 3rd edn. New York: Oxford University Press, 2004:298–316.
      18. ↵
        European Palliative Care Research Collaborative (EPCRC). http://www.epcrc.org (accessed 20 Feb 2015).
      19. ↵
        1. Hjermstad MJ,
        2. Lie HC,
        3. Caraceni A, et al
        . Computer-based symptom assessment is feasible in patients with advanced cancer: results from an international multicenter study, the EPCRC-CSA. J Pain Symptom Manage 2012;44:639–54. doi:10.1016/j.jpainsymman.2011.10.025
        OpenUrlCrossRefPubMedWeb of Science
      20. ↵
        1. Karnofsky D,
        2. Burchenal J
        . The clinical evaluation of chemotherapeutic agents in cancer. In: Macleod C, ed. Evaluation of chemotherapeutic agents. New York: Columbia University Press, 1949:199–205.
      21. ↵
        1. Folstein MF,
        2. Folstein SE,
        3. McHugh PR
        . “Mini-Mental State”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–98. doi:10.1016/0022-3956(75)90026-6
        OpenUrlCrossRefPubMedWeb of Science
      22. ↵
        1. Aaronson NK,
        2. Ahmedzai S,
        3. Bergman B, et al
        . The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993;85:365–76. doi:PMID:8080679
        OpenUrlAbstract/FREE Full Text
      23. ↵
        1. Fayers P,
        2. Aaronson N,
        3. Bjordal K, et al
        . EORTC QLQ-C30 Scoring Manual. Brussels: EORTC Study Group on Quality of Life, 1995.
      24. ↵
        1. Osoba D,
        2. Rodrigues G,
        3. Myles J, et al
        . Interpreting the significance of changes in health-related quality-of-life scores. J Clin Oncol 1998;16:139–44. doi:PMID:9440735
        OpenUrlAbstract/FREE Full Text
      25. ↵
        1. Cocks K,
        2. King MT,
        3. Velikova G, et al
        . Evidence-based guidelines for interpreting change scores for the European Organisation for the Research and Treatment of Cancer Quality of Life Questionnaire Core 30. Eur J Cancer 2012;48:1713–21. doi:10.1016/j.ejca.2012.02.059
        OpenUrlCrossRefPubMedWeb of Science
      26. ↵
        1. Hagen NA,
        2. Fisher K,
        3. Stiles C
        . Sublingual methadone for the management of cancer-related breakthrough pain: a pilot study. J Palliat Med 2007;10:331–7.
        OpenUrlCrossRefPubMedWeb of Science
      27. ↵
        1. William L,
        2. Macleod R
        . Management of breakthrough pain in patients with cancer. Drugs 2008;68:913–24. doi:PMID:18457459
        OpenUrlCrossRefPubMedWeb of Science
      28. ↵
        1. Hosmer D,
        2. Lemeshow S
        . Applied logistic regression. New York: John Wiley & Sons, 2000.
      29. ↵
        1. Bedard G,
        2. Davies A,
        3. McDonald R, et al
        . Breakthrough cancer pain: a comparison of surveys with European and Canadian patients. Support Care Cancer 2015;23:791–6. doi:10.1007/s00520-014-2426-6
        OpenUrlPubMed
      30. ↵
        1. Webber K,
        2. Davies AN,
        3. Cowie MR
        . Disparities between clinician and patient perception of breakthrough pain control. J Pain Symptom Manage 2016;51:933–7.e2. doi:10.1016/j.jpainsymman.2015.12.309
        OpenUrlPubMed
      31. ↵
        1. Hwang SS,
        2. Chang VT,
        3. Kasimis B
        . Cancer breakthrough pain characteristics and responses to treatment at a VA medical center. Pain 2003;101:55–64. doi:PMID:12507700
        OpenUrlCrossRefPubMedWeb of Science
      32. ↵
        1. Mercadante S,
        2. Lazzari M,
        3. Reale C, et al
        . Italian Oncological Pain Survey (IOPS): a multicentre Italian study of breakthrough pain performed in different settings. Clin J Pain 2015;31:214–21. doi:10.1097/AJP.0000000000000161
        OpenUrlPubMed
      33. ↵
        1. Fainsinger RL,
        2. Fairchild A,
        3. Nekolaichuk C, et al
        . Is pain intensity a predictor of the complexity of cancer pain management? J Clin Oncol 2009;27:585–90. doi:10.1200/JCO.2008.17.1660
        OpenUrlAbstract/FREE Full Text
      34. ↵
        1. Bennett MI
        . Cancer pain terminology: time to develop a taxonomy that promotes good clinical practice and allows research to progress. Pain 2010;149:426–7. doi:10.1016/j.pain.2010.01.013
        OpenUrlPubMedWeb of Science
      35. ↵
        1. Knudsen AK,
        2. Aass N,
        3. Fainsinger R, et al
        . Classification of pain in cancer patients—a systematic literature review. Palliat Med 2009;23:295–308. doi:10.1177/0269216309103125
        OpenUrlAbstract/FREE Full Text
      36. ↵
        1. Hjermstad MJ,
        2. Fayers PM,
        3. Caraceni A, et al
        . Studies comparing Numerical Rating Scales, Verbal Rating scales and Visual Analogue Scales for assessment of pain intensity: a systematic review. J Pain Symptom Manage 2011;41:1073–93. doi:10.1016/j.jpainsymman.2010.08.016
        OpenUrlCrossRefPubMedWeb of Science
      37. ↵
        1. Kaasa S,
        2. Loge JH,
        3. Fayers P, et al
        . Symptom assessment in palliative care: a need for international collaboration. J Clin Oncol 2008;26:3867–73. doi:10.1200/JCO.2007.15.8881
        OpenUrlAbstract/FREE Full Text
      38. ↵
        1. Sperlinga R,
        2. Campagna S,
        3. Berruti A, et al
        . Alberta Breakthrough Pain Assessment Tool: a validation multicentre study in cancer patients with breakthrough pain. Eur J Pain 2015;19:881–8. doi:10.1002/ejp.612
        OpenUrlPubMed
      39. ↵
        1. Caraceni A,
        2. Bertetto O,
        3. Labianca R, et al
        . Episodic (breakthrough) pain prevalence in a population of cancer pain patients. Comparison of clinical diagnoses with the QUDEI—Italian questionnaire for intense episodic pain. J Pain Symptom Manage 2012;43:833–41. doi:10.1016/j.jpainsymman.2011.05.018
        OpenUrlPubMed
      40. ↵
        1. Webber K,
        2. Davies AN,
        3. Zeppetella G, et al
        . Development and validation of the Breakthrough Pain Assessment Tool (BAT) in cancer patients. J Pain Symptom Manage 2014;48:619–31. doi:10.1016/j.jpainsymman.2013.10.026
        OpenUrlPubMed
      41. ↵
        1. Deandrea S,
        2. Montanari M,
        3. Moja L, et al
        . Prevalence of undertreatment in cancer pain. A review of published literature. Ann Oncol 2008;19:1985–91. doi:10.1093/annonc/mdn419
        OpenUrlAbstract/FREE Full Text
      42. ↵
        1. Foley KM
        . How well is cancer pain treated? Palliat Med 2011;25:398–401. doi:10.1177/0269216311400480
        OpenUrlAbstract/FREE Full Text
      43. ↵
        1. Greco MT,
        2. Roberto A,
        3. Corli O, et al
        . Quality of cancer pain management: an update of a systematic review of undertreatment of patients with cancer. J Clin Oncol 2014;32:4149–54. doi:10.1200/JCO.2014.56.0383
        OpenUrlAbstract/FREE Full Text
      View Abstract

      Footnotes

      • Collaborators The European Palliative Care Research Collaborative (2006–2010) was funded by the European Commission's Sixth Framework Programme (contract number LSHC-CT-2006-037777), with the overall aim to improve treatment of pain, depression and fatigue through translation research. Core scientific group/work package leaders: Stein Kaasa (project coordinator), Frank Skorpen, Marianne Jensen Hjermstad and Jon Håvard Loge, Norwegian University of Science and Technology (NTNU); Geoffrey Hanks, University of Bristol; Augusto Caraceni and Franco De Conno, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan; Irene Higginson, King's College London; Florian Strasser, Cantonal Hospital St Gallen; Lukas Radbruch, RWTH Aachen University; Kenneth Fearon, University of Edinburgh; Hellmut Samonigg, Medical University of Graz; Ketil Bø, Trollhetta AS, Norway; Irene Rech-Weichselbraun, Bender MedSystems GmbH, Austria; Odd Erik Gundersen, Verdande Technology AS, Norway. Scientific advisory group: Neil Aaronson, The Netherlands Cancer Institute; Vickie Baracos and Robin Fainsinger, University of Alberta; Patrick C Stone, St George's University of London; Mari Lloyd-Williams, University of Liverpool. Project management: Stein Kaasa, Ola Dale and Dagny F Haugen, NTNU. The palliative care centre at the Fondazione Istituto Nazionale dei Tumori (Italy) was partially funded by the Floriani Foundation of Milan.

      • Contributors As the main author and principal investigator of the EPCRC-CSA study, MJH is the main contributor and as such responsible for the overall content. All the coauthors contributed to the study planning, conduct and reporting of results, as stated below. MJH contributed to study planning, writing initial study protocol, overall study conduct, manuscript planning, drafting, statistical analyses, discussions, discussions about manuscript, writing of manuscript and its submission. SK contributed to study planning, writing initial study protocol, manuscript planning, and discussions about manuscript. AC contributed to study planning, writing initial study protocol, patient recruitment, manuscript planning, and discussions about manuscript. JHL contributed to study planning, writing initial study protocol, manuscript planning, and discussions about manuscript. TP contributed to planning analyses, performing the statistics, general discussions about manuscript, and writing of manuscript. DFH contributed to study planning, writing initial study protocol, patient recruitment, manuscript planning, and discussions about manuscript. NA contributed to study planning, writing initial study protocol, patient recruitment, manuscript planning, and discussions about manuscript.

      • Funding European Commission's Sixth Framework Programme (contract number LSHC-CT-2006-037777).

      • Competing interests None declared.

      • Ethics approval In Norway: Regional Committee for Medical and Research Ethics HSØ; regional committees in all other participating countries.

      • Provenance and peer review Not commissioned; externally peer reviewed.