Cancer breakthrough pain characteristics and responses to treatment at a VA medical center

Pain 101 (2003) 55–64 www.elsevier.com/locate/pain

Cancer breakthrough pain characteristics and responses to treatment at a VA medical center q Shirley S. Hwang a,b,c,*, Victor T. Chang a,d, Basil Kasimis a,d a

Section of Hematology/Oncology, VA New Jersey Health Care System, 385 Tremont Avenue, East Orange, NJ 07018, USA b Patient Care Services, VA New Jersey Health Care System, East Orange, NJ 07018, USA c UMDNJ/ School of Nursing, Newark, NJ 07103, USA d UMDNJ/ New Jersey Medical School, Newark, NJ 07103, USA Received 5 March 2002; accepted 15 July 2002

Abstract The purpose of this study is to analyze cancer breakthrough pain (BP) characteristics and how BP responds to conventional cancer pain management. Seventy-four cancer pain patients with worst pain severity $4 out of 10 completed the Brief Pain Inventory (BPI), Memorial Symptom Assessment Scale-Short Form, Functional Assessment Cancer Therapy and Breakthrough Pain Questionnaires (BPQ) at an initial interview. Agency for Health Care Policy and Research (AHCPR) cancer pain management guidelines were followed. Pain syndromes and morphine equivalent daily dose (MEDD) orally were determined. One-week follow-up assessments were obtained in 66 patients with BPI and BPQ. The BP characteristics were similar at both time points. On day 1, 52 patients (70%) had BP, and the BP was unpredictable in 30 patients (58%). The median time to worst BP severity was 3 min. Patients with BP had significantly higher worst pain ðP , 0:001Þ. At week 1, the median MEDD doubled from 60 to 120 mg orally, and the number of patients who received adjuvant analgesics doubled from 31.1% (23 patients) on day 1 to 62.2% (41 patients). At week 1, 21 patients (32%) remained without BP, 21 patients (32%) were classified as BP responders and 24 patients (36%) were BP non-responders. The mean pain relief was similar for all three subgroups, i.e. around 80%. Compared to BP responders, BP non-responders had significantly higher worst pain ðP , 0:0001Þ, average pain ðP , 0:004Þ, and higher BPI interference parameters and shorter time to worst pain severity. The study confirmed the applicability of the BPQ to an US veteran population, and that pain management following the AHCPR guidelines is effective for a group of patients with cancer related BP. Underlying pain syndromes and the BP location may influence the response of BP to treatment. Patients with bone pain located in the spine, back, and pelvis may be at risk for resistant BP. q 2002 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. Keywords: Breakthrough pain; Pain interference; Pain syndromes

1. Introduction Breakthrough pain (BP) is a frequently encountered pain syndrome in cancer patients with an incidence rate of 50– 89% (Ferrell et al., 1999; Portenoy and Hagen, 1990; Portenoy et al., 1999; Zeppetella et al., 2000) and has been defined as ‘the transient exacerbation of pain occurring in a patient with otherwise stable, persistent pain.’ It is usually unpredictable and heterogeneous (Portenoy and Hagen, q The views expressed herein do not necessarily reflect the views of the Department of Veterans Affairs or the US Government.Preliminary results presented at the Annual Scientific Meeting of the American Pain Society in Atlanta, GA, USA. Hwang SS, Chang VT, Kasimis B. Cancer breakthrough pain and responses to treatment. American Pain Society, 19th Annual Scientific Meeting, Atlanta, GA. November, 2000, Abstract 625. * Corresponding author. Tel.: 11-973-395-7090. E-mail address: [email protected] (S.S. Hwang).

1990; Portenoy et al., 1999). In general, BP can be characterized into the following subtypes: incident pain (caused by patient movement), spontaneous pain or idiopathic pain (unrelated to patient action), and end-of-dose pain (occurring just prior to next scheduled dose of analgesics) (Portenoy and Hagen, 1990) and the treatment strategies are based upon the underlying pathophysiology. The presence of BP is usually associated with higher pain severity, greater pain-related functional impairment, psychological distress and decreased satisfaction with pain management (Portenoy et al., 1999; Zeppetella et al., 2000), and recognized as a poor prognostic factor in Edmonton Pain Staging System (Bruera et al., 1995). Mercandate et al. reported that only about 50% of advanced cancer patients with incidental pain had good pain control after treatment based upon the World Health Organization (WHO) analgesic ladder (Mercadante et al., 1992).

0304-3959/02/$20.00 q 2002 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. doi:10.1016/S0 304-3959(02)00 293-2

56

S.S. Hwang et al. / Pain 101 (2003) 55–64

In a multicenter, international survey of cancer pain syndromes in 1095 patients, a large difference was noted in the identification of BP by clinicians from different countries and the authors suggested that a standardized BP definition is needed (Caraceni and Portenoy, 1999). The Breakthrough Pain Questionnaire (BPQ) (Portenoy and Hagen, 1990) offers a way to standardize the identification of patients with BP. In a recent study of development of cancer pain prognostic scale from our group, consecutive patients with cancer related pain with worst pain severity equal to or greater than 4 out of 10 were asked to participate (Hwang et al., 2003). The Agency for Health Care Policy and Research (AHCPR) cancer pain management guidelines were followed (Agency for Health Care Policy and Research, 1994) and pain relief equal to or greater than 80% was selected as the ‘good pain control’ outcome measure. Patients were screened for BP with the BPQ at each visit. In this paper, we report exploratory analyses to describe the BP characteristics and response to conventional cancer pain management. 2. Materials and methods 2.1. Patient selection In this prospective, longitudinal study, patient recruitment began in January 1996 and ended in March 1997 at the VA New Jersey Health Care System at East Orange (VANJHCS). The study was approved by the VA NJHCS Institutional Review Board, and all patients signed informed consent before participating. A prospective convenience sample of 74 consecutive patients with cancer related pain (worst pain severity equal to or greater than 4 out of 10) were recruited from patients seen in the outpatient Hematology/Oncology clinic or from inpatients who had been admitted to the Hematology/Oncology service. Cancer-related pain was defined as pain caused by cancer or by treatment of cancer. Moribund or delirious patients who were not able to fill out the questionnaires were excluded. 2.2. Assessments Initial assessment included age, sex, primary site, and extent of disease. Each patient was asked to complete four assessment instruments: the Brief Pain Inventory (BPI) (Daut et al., 1983), Memorial Symptom Assessment Scale-Short Form (MSAS-SF) (Chang et al., 2000), Functional Assessment of Cancer Therapy (FACT-G) (Cella et al., 1993), and BPQ. Status of opioid use and adjuvant analgesic therapy (medications and radiation) was obtained. The number of pain sites was recorded. The primary and contributing pain pathophysiology and pain syndromes (Portenoy, 1992) associated with BP were assessed and recorded by the primary investigators (S.S.H. and V.T.C.). It was characterized as somatic, neuropathic or

visceral pains. Somatic pains were related to an etiology that involved somatic structures such as bone or muscle. Visceral pains were related to a lesion in a hollow or solid viscus. Neuropathic pains were related to a lesion which involved peripheral or central afferent neural pathways. Patients were screened for the presence of BP with the following questions: 1. Has the patient had surgery during the past week? (yes or no) If yes, then the patient was excluded from study. If no, then the following questions were asked. 2. Do you feel pain for more than half the time that you are awake? (yes or no) If the patient answered yes, the patient had background pain and the interview continued. 3. For most of the time that you have been feeling this pain in the last 24 h, is the pain mild, moderate, severe, or excruciating? If the patient answered severe or excruciating, then the patient had uncontrolled background pain and the interview would be terminated. 4. Do you also experience temporary flares of severe or excruciating pain? (yes or no) If yes, then the patient had BP and the interview would proceed with completion of the remaining questions on the BPQ to assess BP characteristics. A maximum of two BP sites for each patient were recorded. Based on the above screening process, we identified 52 patients with BP on day 1. After completion of the assessment, standard cancer pain management guidelines from the AHCPR for pain assessment, analgesic interventions, and management of neuropathic and bone pain were followed. Preferences for pharmacological management were implemented as described in the AHCPR Cancer Pain Management ((Agency for Health Care Policy and Research, 1994)). Briefly, these follow the WHO ladder, and recommend individualizing the patient’s regimen, opioids for severe cancer pain, an oral route as the preferred route, management of side effects, and other modalities of cancer pain therapies. All patients were placed on ‘around-the-clock’ and ‘rescue’ opioids and all opioid therapies were converted into oral morphine equivalent daily dose (MEDD) and recorded. The use of adjuvant analgesic therapies (medications and radiation therapy) also was recorded. Patients and family members received the standard AHCPR patient guidelines upon study participation. One-week follow-up assessments with BPI, and BPQ were obtained in 66 patients. Six patients died before the week 1 assessment and two patients refused further assessment; 24 patients had persistent BP at week 1. Patients whose BP had resolved at 1 week were defined as BP responders. 2.3. Instruments The BPI (Daut et al., 1983), is a validated and widely used tool to assess pain intensity with a numerical 0–10 scale,

S.S. Hwang et al. / Pain 101 (2003) 55–64

where 0 means ‘no pain at all’ and 10 means ‘the worst possible pain patient can imagine.’ Patients rate their worst, least, average, and immediate pain severity, pain relief ranging from 0 to 100%, and functional interference caused by pain in the areas of daily activity, mood, walking, sleeping, movement, enjoyment of life and relationship with others. The sum of answers to the interference questions, the total pain interference score, ranges from 0 to 70. The FACT-G (version 3) (Cella et al., 1993), is a validated, 28-item general patient-rated measure of Quality of Life (QOL) for cancer patients with any tumor type. Each item is scored from 0 to 4 and anchored from ‘not at all’ to ‘very much.’ There are five subscales: Functional Well Being (FWB) (seven items), Physical Well Being (PWB) (seven items), Social/Family Well Being (SFWB) (seven items), Relationship with MD (RMD) (two items), and Emotional Well Being (EWB) (five items), with total QOL scores ranging from 0 to 112. The MSAS-SF (Chang et al., 2000), is a validated patientrated instrument that includes patient assessment for symptom frequency or distress for 32 highly prevalent physical and psychological symptoms. Each symptom was scored from 0 to 4 ranging from ‘no symptom’ to ‘very much.’ MSAS-SF subscales include the Global Distress Index (GDI) (four psychological symptoms: worrying, feeling sad, feeling irritable, and feeling nervous; and six physical symptoms: lack of energy, pain, lack of appetite, feeling drowsy, constipation, dry mouth). The Physical Symptom Distress Score (PHYS) includes 12 prevalent symptoms: lack of energy, pain, lack of appetite, feeling drowsy, constipation, dry mouth, nausea, vomiting, change in taste, weight loss, feeling bloated, and dizziness. The Psychological Symptom Distress Score (PSYCH) includes six prevalent psychological symptoms: worrying, feeling sad, feeling nervous, difficulty sleeping, feeling irritable, and difficulty in concentrating. The number of symptoms (NS) is derived from screening for the presence of 32 symptoms at each interview. The BPQ (Portenoy and Hagen, 1990) contains four screening questions to identify the presence of BP and 12 questions related to characteristics of BP. Based on their definition, patients with BP need to experience background pain of mild to moderate intensity greater than half of the time while they are awake and also experience temporary flares of severe or excruciating pain. The BPQ was administered by a research assistant. If patients were identified to have BP, then the interview would proceed with 12 additional questions to assess BP location, temporal characteristics (frequency, onset, duration), relationship to fixed analgesic dose, precipitating factors, predictability, and responsiveness to treatment. 2.4. Statistical analysis The underlying pain syndromes causing BP and BP characteristics were tabulated at both initial and 1-week assess-

57

ments (Tables 1 and 2). The t-test was performed between patients with and without BP based on the pain parameters, the MSAS-SF symptom distress subscales, and FACT-G

Table 1 Pain pathophysiology, primary pain syndromes and locations associated with breakthrough pain on day 1 and at week 1 Day 1 (N ¼ 52)

Week 1 (N ¼ 24)

N

N

%

%

Predominant pain pathophysiology Neuropathic pain 25 Somatic pain 22 Visceral pain 8 None 0

48 42 10 0

13 10 1 0

54 42 2 0

Contributing pain pathophysiology Neuropathic pain 9 Somatic pain 13 Visceral pain 1 None 28

18 25 2 55

3 8 0 13

13 33 0 54

10 2

42 8

0 1 0 0 1

0 4 0 0 4

Primary pain syndromes a Nociceptive (Somatic and visceral pains) Bone pain 14 27 21 Local tumor invasion (soft 11 tissue) Hepatic distension 3 6 Chest wall syndromes 2 4 Acute procedural pain 1 2 Acute pain with radiation 1 2 Other 3 6 Neuropathic pain Brachial plexopathy Epidural compression Cranial neuralgias Lumbosacral plexopathy Leptomeningeal metastases Cervical plexopathy Post radical dissection pain Stump pain Locations b Number of pain sites (median, range) Same as background pain Non-mid-line back Arm/shoulder Chest Abdomen Neck Leg Pelvis/hip Spine Anus Head Total

5 3 4 2 1 1 1 0

9 6 7 4 2 2 2 0

2 2 2 1 1 1 0 1

8 8 8 4 4 4 0 4

2

1–5

2

1–5

37 10 12 8 7 7 6 6 5 4 2 67

75 16 18 12 10 10 9 9 7 6 3 100

24 7 7 2 3 3 5 5 5 0 0 39

100 18 18 5 7 7 13 13 13 0 0 100

a Somatic pains: pain related to an etiology that involved somatic structures such as bone or muscle. Visceral pains: pain related to a lesion in a hollow or solid viscus. Neuropathic pains: pain related to a lesion involved peripheral or central afferent neural pathways. b Breakthrough pain locations, maximum two sites per patient were recorded.

58

S.S. Hwang et al. / Pain 101 (2003) 55–64

Table 2 Breakthrough pain characteristics Day 1 (N ¼ 52) N 1.

2.

3.

4.

5.

6.

7.

%

N

%

How many different types of temporary flare ups do you have? One 41 79 24 100 Two 11 21 0 0 When your flare becomes as bad as it gets, would you say that it is severe or excruciating? Yes 49 94 21 88 No 3 6 3 12 From the time that you first feel this pain start to flare until it gets to be as bad as it gets, how long does it take? Unpredictable 9 17 5 21 Less than 10 s 19 37 13 54 2–5 min 13 25 3 13 6–30 min 7 13 1 4 31–60 min 4 8 2 8 Median 3 min Less than 10 s When the pain flare comes on, does it start mildly and gradually get worse, or does it immediately start out severely? Mildly 20 38 9 38 Severely 32 62 12 50 Not sure 0 0 3 12 If the pain flare starts slowly and gradually gets worse, how long it takes from the time that you first start to feel the pain coming until it goes away? Median 15 min 15 min Range 1 min–2 h 10 min–1 h How many times a day do you get this specific kind of pain flare? Not sure 13 25 0 0 Median 5 times 4 times Range 1–30 times 1–50 times Does your pain flare usually come at a certain amount of time after you have taken your pain medication? No 32 62 14 58 Yes 10 21 7 29 Not sure 10 21 3 13

Precipitants 8. Do you know what causes your pain flare to come on? Movement 24 End of dose 9 Others 11 a No 10 Palliatives 9.

Week 1 (N ¼ 24)

Does anything help to lessen your painful episode? Medications 29 Others 14 c No 11

44 17 20 19

10 8 2b 4

42 33 8.3 16.7

54 26 20

20 2d 2

83 8.5 8.5

12 7 0 5

50 29 0 21

Predictability 10. How well can you predict whether your pain flare will occur? Never 30 57 Sometimes 10 19 Often 6 12 Almost always-Always 6 12

a Coughing in two; muscle spasm in one; sitting in one; swallowing in two; standing in one; night time in one; positioning in one; moving bowel in one; receiving radiation in one. b Feeling tired in one; standing in one. c Avoiding movement in four; repositioning in four; lying down in three; moving around in two; bed rest in one. d Avoiding movement in one; moving around in one.

QOL parameters for day 1 and week 1 (BPI parameters only) interviews (Tables 3 and 4). Based on the day 1 BP status and responsiveness to pain treatment, the patients were classified into three subgroups: patients without BP (no BP) (N ¼ 21, 31.8%), patients with

BP on day 1 and no BP at week 1 (BP responders) (N ¼ 21, 31.8%) and patients with persistent BP (BP non-responders) (N ¼ 24, 36.4%). The Wilcoxon match-paired sign-rank comparison was used to assess the differences in pain parameters and BPI functional interference between the two

S.S. Hwang et al. / Pain 101 (2003) 55–64

59

Table 3 Summary statistics of BPI pain, MSAS-SF and FACT-G parameters on day 1 All (N ¼ 74)

With breakthrough (N ¼ 52)

Without breakthrough (N ¼ 22)

P value a

Mean

SD

Mean

SD

Mean

SD

Pain parameters Worst pain Average pain Pain relief

8.3 6.2 40

1.9 1.9 30.7

8.7 6.2 41

1.5 1.9 28.8

7.4 6.1 36

2.3 2.4 37.5

0.001 0.3 0.7

Treatments MEDD (median, range) Adjuvant therapy (N, %)

60 23

0–420 31.1%

60 18

0–420 34.6%

30 5

0–240 21%

0.2

BPI interference Activity Mood Walking Work Relationship Sleep Enjoyment

5.9 5.6 4.1 4.3 3.8 5.5 34.3

4.9 3.9 3.9 4.5 4.2 4.0 22.3

5.6 5.6 4.8 4.6 4.2 4.8 35.3

4.0 4.1 3.9 4.5 4.0 3.9 22.7

6.8 5.3 2.5 3.2 2.7 5.9 31.1

4.2 3.5 3.5 4.4 3.6 4.1 20.5

0.5 0.6 0.05 0.2 0.2 0.3 0.5

MSAS-SF parameters Physical symptom Psychological symptom Global distress index Number of symptoms

1.23 1.16 1.45 12

0.72 0.97 0.76 5.6

1.22 1.21 1.48 13

0.71 1.02 0.76 6

1.24 1.02 1.38 11

0.76 0.89 0.78 6

0.9 0.5 0.6 0.4

FACT-G domains Physical well being Social/family well being Emotional well being Functional well being SUMQOL

16.8 21.5 15.8 12.3 73.5

6.6 5.9 4.2 6.3 18.2

16.1 21.6 16.1 12.2 73.1

3.9 5.9 6.4 6.4 18.3

18.5 22.8 14.8 11.2 74.4

5.9 5.1 4.5 6.1 18.2

0.2 0.4 0.2 0.3 0.8

a

t-Test between patients with breakthrough pain and without breakthrough pain.

Table 4 Summary statistics and t-test results of BPI pain parameters at week 1 All (N ¼ 66)

With breakthrough (N ¼ 24)

Without breakthrough (N ¼ 42)

P value a

Mean

SD

Mean

SD

Mean

SD

Pain parameters Worst pain Average pain Pain relief

6.0 3.4 75

2.7 2.6 22.5

8.2 4.7 76

1.8 2.5 23.9

4.7 2.6 78

2.5 2.4 18.1

0.0001 0.005 0.6

Treatments MEDD (median, range) Adjuvant therapy (N, %)

120 41

0–1080 62.2%

120 13

0–690 54.2%

90 28

0–1080 66.7%

0.8

BPI interference Activity Mood Walking Work Relationship Sleep Enjoyment Total interference

5.2 3.2 2.8 1.5 2.5 3.6 2.9 19.9

3.9 3.8 3.7 3.2 3.7 4.2 3.7 20.5

6.0 5.2 4.6 2.5 3.8 5.2 5.1 32.4

3.7 3.9 4.1 3.8 3.9 3.8 4.2 17.8

2.1 1.9 1.8 0.8 1.7 1.6 2.8 12.7

3.5 3.2 3.2 2.6 3.3 2.8 3.7 19.2

0.0001 0.001 0.005 0.05 0.05 0.0001 0.05 0.05

a

t-Test between patients with breakthrough pain and without breakthrough pain.

60

S.S. Hwang et al. / Pain 101 (2003) 55–64

Table 5 Effects of pain management on patients without breakthrough pain on day 1 or at week 1 Day 1 (N ¼ 21)

P value a

Week 1 (N ¼ 21)

Mean

SD

Mean

SD

Pain parameters Worst pain Average pain Pain relief

7.0 5.5 40

2.3 2.2 35.7

5.0 2.5 80

2.5 2.7 17.6

0.01 0.003 0.004

Treatments MEDD (median, range) Adjuvant therapy

30 6

0–366 28.5%

60 15

0–1080 71.4%

0.0008 0.005 b

BPI interference Activity Mood Walking Work Relationship Sleep Enjoyment Total interference

6.4 5.0 2.6 3.0 2.6 5.8 4.7 30.1

4.3 3.6 3.8 4.4 3.5 4.1 4.1 21.2

1.9 1.6 1.7 0.7 1.6 1.2 2.4 11.0

3.4 3.2 3.1 2.1 3.2 2.6 4.1 17.3

0.001 0.003 0.3 0.07 0.2 0.001 0.06 0.002

a b

Wilcoxon matched-pair signed-rank comparison. Chi-square test.

time points for each subgroup (Tables 5, 6, and 7). Data was analyzed with the STATA program v5.0 (College Station, TX, USA).

3. Results 3.1. Patient characteristics The patient characteristics have been reported in detail

elsewhere (Hwang et al., 2003). Of the 74 patients recruited, the median age was 63 years (range 40–82) and the median education level was 12th grade (range 6–18). There were 39 (53%) inpatients and 35 (47%) outpatients. BP was reported by 52 patients (70%) on day 1, and the most frequently reported pain syndromes associated with BP included bone pain (14 patients, 27%), local tumor invasion-soft tissue (11 patients, 21%), and brachial plexopathy (five patients, 9%). The reported pain locations for BP and background pain were the same in 37 patients (75%) and the

Table 6 Effects of pain management on breakthrough pain responders Day 1 (N ¼ 21)

Week 1 (N ¼ 21)

P value a

Mean

SD

Mean

SD

Pain parameters Worst pain Average pain Pain relief

8.2 6.4 34

1.9 1.8 26.9

4.8 2.9 74

2.6 2.4 21.5

0.0002 0.0002 0.0002

Treatments MEDD (median, range) Adjuvant therapy (N, %)

60 7

0–330 33.3%

180 14

0–720 66.7%

0.0003 0.03 b

BPI interference Activity Mood Walking Work Relationship Sleep Enjoyment Total interference

4.3 4.7 3.6 3.8 3.6 4.8 5.4 30.3

4.2 4.4 3.9 4.5 4.4 4.0 4.1 25.6

2.3 2.4 1.9 1.0 1.9 1.9 3.1 14.4

3.5 3.2 3.2 3.0 3.3 3.2 3.8 20.2

0.04 0.04 0.03 0.005 0.08 0.003 0.03 0.003

a b

Wilcoxon matched-pair signed-rank comparison. Chi-square test.

S.S. Hwang et al. / Pain 101 (2003) 55–64

61

Table 7 Effects of pain management on breakthrough pain non-responders (breakthrough pain on both day 1 and week 1) Day 1 (N ¼ 24)

Week 1 (N ¼ 24)

P value a

Mean

SD

Mean

SD

Pain parameters Worst pain Average pain Pain relief

8.9 6.5 39

1.1 1.7 29.8

8.0 4.7 73

1.9 2.5 17.0

0.003 0.003 , 0.0001

Treatments MEDD (median, range) Adjuvant therapy (N, %)

60 8

0–420 33.3%

120 13

0–690 54%

0.003 0.1 b

BPI interference Activity Mood Walking Work Relationship Sleep Enjoyment Total interference

6.9 6.5 5.8 5.9 4.7 5.3 5.9 41.0

3.2 3.5 3.5 4.1 3.3 3.9 3.9 18.2

6.0 5.2 4.6 2.5 3.8 5.2 5.1 32.4

3.7 3.9 4.1 3.8 3.9 3.8 4.2 17.8

0.3 0.1 0.2 0.01 0.4 0.9 0.3 0.02

a b

Wilcoxon matched-pair signed-rank comparison. Chi-square test.

most frequently reported locations were arm and shoulder (12 patients, 18%), non-mid-line back (ten patients, 16%), chest (eight patients, 12%), abdomen (seven patients, 10%), and neck (seven patients, 10%). One week after pain management began, 66 patients completed the week-1 interview, and persistent BP was reported by 24 patients (36%). The most frequently reported pain syndrome associated with BP at week 1 was bone pain (ten patients, 42%). All patients reported the same location for BP and background pain. The most frequently reported locations were non-mid-line back (seven patients, 13%), arm and shoulder (seven patients, 18%), leg (five patients, 13%), pelvis and hip (five patients, 13%) and spine (five patients, 13%). Details are summarized in Table 1.

3.2. BP pain characteristics for day 1 and week 1 The BP characteristics are listed in Table 2. Most results were similar for both time points. In summary, the majority of patients had one type of pain flare (79% on day 1 and 100% at week 1) and the flares were severe or excruciating (94% on day 1 and 88% at week 1). About 17% (day 1) to 21% (week 1) of patients were not able to predict how long it would take for BP to flare up to its worst severity. On day 1, the median patient rated time to worst severity was 3 min (range ,10 s to 60 min) and in 19 patients (37%), BP flared up to worst severity within 10 s. Among these 19 patients, eight patients had bone pain (bone pain or epidural compression). At week 1, the median time to the worst BP severity was less than 10 s (range from ,10 s to 60 min) and in 13 patients (54%), BP flared up to its worst severity in less than

10 s. Among these 13 patients, eight patients (the same eight patients as day 1) experienced bone pain. Among patients experiencing severe BP, rapid onset of the pain occurred in 62% of patients on day 1 and 50% at week 1. The duration of BP varied from 1 min to 2 h on day 1 and from 1 min to 1 h at week 1. In general, the duration of BP was short with a median of 15 min for both time points. The number of BP episodes ranged from 1 to 50 times per day with the median of five times per day on day 1 and four times at week 1. Only about 21% (day 1) to 29% patients (week 1) agreed that the BP occurred at a certain amount of time after taking pain medications. More than 80% of patients at both time points were able to identify the various precipitating factors, and movement was the most frequently reported precipitating factor, which accounted for about 40% of BP for both time points. End-of-dose BP was reported in 19% of patients for day 1 and in 33% of patients at week 1. About 54% of patients on day 1 and 83% of patients at week 1 stated that the pain medication palliated painful episodes, and half of the patients experienced unpredictable pain flares at both time points. 3.3. Comparison of pain parameters, MSAS-SF and FACT-G parameters between patients with and without BP on day 1 The summary scores and comparison between patients with and without BP based on the BPI pain parameters, MSAS-SF subscales and FACT-G QOL scores are listed in Table 3. For all populations, the mean worst pain severity was 8.3 (range 4–10) with average pain severity of 6.2 (range 3–10) and pain relief of 40% (range 0–100%). Analgesic therapy

62

S.S. Hwang et al. / Pain 101 (2003) 55–64

was used in 60 patients (81%) with the median MEDD of 60 mg (range 0–420 mg) orally. Adjuvant analgesic therapy was used in 23 patients (31.1%). The mean pain interference score was 34.5 (range 0–70). Patients with BP had significantly higher worst pain severity (8.7 vs. 7.4, P , 0:001) and interference with walking (4.6 vs. 3.2, P , 0:05). There were no significant differences in average pain severity, pain relief, other pain interference items (activity, mood, work, relationship with others, sleep and enjoyment of life) and total pain interference scores. There was no significant difference in the MSAS-SF parameters and FACT-G QOL parameters between patients with BP or without BP on day 1. 3.4. Comparison of pain parameters between patients with and without BP at week 1 For all populations, the mean worst pain severity was 6.0 (range 0–10) with average pain severity of 3.4 (range 0–10) and pain relief of 75% (range 10–100%). Opioids were used in 62 patients (94%) with the median MEDD of 120 mg (range 0–1080 mg). Adjuvant analgesic therapy was used in 41 patients (62.2%). The mean pain interference score was 19.9 (range 0–70). About 50% of BP was resolved by week 1 and BP was reported by 24 (36%) patients. Patients without BP had significantly lower worst pain (4.7 vs. 8.2, P , 0:0001), average pain severity (2.6 vs. 4.7, P , 0:0001) and lower BPI pain interference scores for each interference item and for total pain interference scores (P ranging from 0.05 to ,0.0001). There was no difference in pain relief (76 vs. 78%, P ¼ 0:6). These observations were different from the day-1 results (See Table 3). 3.5. Effects of pain management on BP Based on the day-1 BP status and its responses to cancer pain management, the patients were categorized into three groups (no BP, BP responders, and BP non-responders). The impact of cancer pain management on pain parameters was further assessed for each sub group. 3.5.1. No BP group There were significant improvements in worst pain severity (7.0 vs. 5.0, P ¼ 0:01), average pain severity (4.7 vs. 2.5, P ¼ 0:003) and pain relief (40 vs. 80%, P ¼ 0:004) by week 1. The median MEDD was doubled from 30 mg orally on day 1 to 60 mg orally at week 1 ðP ¼ 0:0008Þ. The use of adjuvant analgesic therapy was increased from six patients (28.5%) (NSAID in two, corticosteroids in two, tricyclic antidepressants in one) on day 1 to 15 patients (71.4%) (corticosteroids in seven, radiation in six, tricyclic antidepressants in five, NSAID in three, pamidronate in one) (x2 ¼ 7:71, P ¼ 0:005) at week 1. Seven patients received two adjuvant analgesic therapies. For pain interference items, there were significant differences in activity, mood, sleep, and total pain interference

scores (P ¼ 0:001; 0:003; 0:001; and 0:002, respectively) between day 1 and week 1. There were no significant differences in walking, work, sleep, and relationship with others. However, all of these parameters did illustrate slightly lower scores at week 1. 3.5.2. BP responders Twenty-one patients were classified as BP responders. The BP locations on day 1 included chest in six (21%), arm/shoulder in five (18%), abdomen in four (14%), anus in four (14%), non-mid-line back in three (11%), neck in two (7%), head in two (7%), pelvis/hip in one (4%) and leg in one (4%). There were significant differences in all the pain parameters, six pain interference items, and total pain interference score (Table 6). In summary, the mean worst pain was decreased from 8.2 to 4.8 ðP ¼ 0:0002Þ, mean average pain was decreased from 5.5 to 2.9 ðP ¼ 0:0002Þ, and the mean pain relief was increased from 34 to 74% ðP ¼ 0:0002Þ. The median MEDD was tripled from 60 mg orally on day 1 to 180 mg orally at week 1 ðP ¼ 0:0003Þ. The use of adjuvant analgesic therapy was increased from seven patients (33.3%) (radiation in three, NSAID in two, corticosteroids in one and tricyclic antidepressants in one) on day 1 to 14 patients (66.7%) (radiation in nine, corticosteroids in five, tricyclic antidepressants in five, anticonvulsants in one, and other in one) (x2 ¼ 4:67, P ¼ 0:03) at week 1. Seven patients received two adjuvant analgesic therapies. For pain interference scores, besides ‘relationship with others’ ðP ¼ 0:08Þ, there were significant differences in all other six interference items (activity, walking, work, mood, sleep and enjoyment of life) with the P value ranging from 0.04 to 0.005. The mean total pain interference score decreased from 30.3 on day 1 to 14.4 at week 1 ðP ¼ 0:003Þ. 3.5.3. BP non-responders Twenty-four patients were classified as BP non-responders. The BP locations on day 1 included non-mid-line back in seven (18%), arm/shoulder in seven (18%), neck in five (13%), pelvis/hip in five (13%), leg in five (13%), spine in five (13%), abdomen in three (7%) and chest in two (5%). Although the patients were classified as BP non-responders, significant improvements were observed in worst pain severity (8.9 vs. 8.0, P ¼ 0:003), average pain severity (6.4 vs. 4.6, P ¼ 0:003), and pain relief (38 vs. 73%, P ¼ 0:0001). The median MEDD was doubled from 60 mg on day 1 to 120 mg orally at week 1 ðP ¼ 0:003Þ. The use of adjuvant analgesic therapy was increased from eight patients (33.3%) (NSAID in two, corticosteroids in one, tricyclic antidepressants in two, and anticonvulsants in three) on day 1 to 13 patients (54%) (radiation in six, anticonvulsants in three, corticosteroids in three, tricylic antidepressants in two, NSAID in one and calcitonin in one) (x2 ¼ 2:12, P ¼ 0:1) at week 1. Three patients received two adjuvant therapies. For the pain interference items, there were significant

S.S. Hwang et al. / Pain 101 (2003) 55–64

differences in the ‘work’ item (5.9 vs. 2.5, P ¼ 0:001) and in total pain interference score (41 vs. 32, P ¼ 0:02) between day 1 and week 1. Lower interference scores were seen in the areas of activity, mood, walking, relationship with others, sleep and enjoyment of life at week 1. Details are summarized in Table 7.

4. Discussion In this paper, we report cancer BP characteristics and responses to conventional analgesic therapy at a VA Medical Center. The BPQ was first used to define BP characteristics in patients referred to a pain service (Portenoy and Hagen, 1990). A confirmation study with a larger inpatient population was reported in 1999 (Portenoy et al., 1999). Both studies were conducted at NCI designated cancer centers. Zeppetella et al. used the BPQ in cancer patients admitted to a hospice and reported similar BP characteristics with a higher incidence of BP (89%) (Zeppetella et al., 2000). We have used the BPQ to study veteran cancer patients at a VA medical center. This population reflects the national VA population, which differs from the general American population as it experiences a higher mortality rate (Fisher and Welch, 1995), reflects the lower 10% of the socioeconomic strata (Harris et al., 1989), and has poor health status scores (Kazis et al., 1998). The results support the BPQ’s generalizability to different patient populations. On day 1, the incidence of BP (70%) was similar to the previous reports; 36% (17 patients) of BP was caused by bone pain (bone pain and epidural compression) and 21% (11 patients) by local tumor invasion – soft tissue. Patients with BP had higher worst pain (8.7 vs. 7.4, P ¼ 0:001) and interference in walking (4.8 vs. 2.5, P ¼ 0:04) than patients without BP. Regardless of BP status on day 1, all patients had a high average pain severity with a mean score of 6.2 and low mean pain relief of 40%, which suggests that the majority of patients’ background pain on day 1 was under treated. These patients were still able to recognize the presence of BP, despite having significant underlying pain. Further supporting the larger effect of background pain are the similar results of the BPI pain interference score, MSAS-SF symptom distress and FACT-G QOL results between patients with and without BP. These results are of interest as others have found that patients with uncontrolled background pain can also experience temporary severe exacerbations (Petzke et al., 1999; Zeppetella et al., 2000). The effectiveness of conventional analgesics and adjuvant analgesics in palliating BP has not been reported. At week 1, about 50% of patients with BP reported a positive response to opioids and adjuvant analgesic treatment and the incidence rate of BP was decreased to 36%. This may be the first paper to suggest that conventional analgesic therapies not only can improve the background pain but also can reduce the BP prevalence.

63

The underlying pain syndromes and the BP location may influence response to treatment. Our results show that the BP caused by local tumor invasion – soft tissue improved the most [from 11 patients (21%) down to two patients (8%)]. In contrast, the BP caused by bone metastasis with association of movement (incidental pain) may be the most difficult to palliate. The incidence of BP related to bone pain (bone pain and epidural compression) increased from 33% on day 1 (17 out of 52 patients) to 50% (12 out of 24 patients) at week 1. The presence of incidental pain has been reported to reduce responsiveness to analgesic therapy (Banning et al., 1991; Bruera et al., 1995; Mercadante et al., 1992). The high incidence of bone pain syndromes is reflected in the high incidence of BP in spine, neck, leg, pelvis, arm/shoulder and non-mid-line back at week 1 and might also contribute to the ‘instant’ flare ups (less than 10 s to its worst severity) associated with movement. The BP associated with neuropathic pain was seen in 34 patients (64%) on day 1, and in 16 patients (67%) at week 1. Although incidence was still high, more than 50% of neuropathic pain was resolved by week 1. The utilization of adjuvant analgesic therapies such as corticosteroids, tricyclic antidepressants, and anticonvulsants, in conjunction with opioid therapy, might have significantly influenced the neuropathic BP responses. Further studies to confirm these results are needed. At week 1, patients with persistent BP not only had higher worst pain severity, but also had higher average pain severity, higher pain interference scores in six areas and higher total pain interference scores. These results were different from the day 1 results. Subgroup matched-paired signedrank analysis suggested that BP responders (Table 6) had the greatest improvement in pain severity (more than 3 points for worst pain and average pain), followed by the no BP group (Table 5) (2 points for worst pain and 3 points for average pain). The BP non-responders (Table 7) also showed significant improvement in pain severity with 1 point for worst pain and 2 points for average pain. Although the BP non-responders also showed significant improvement in total pain interference scores (9 points); the improvement was relatively small in comparison to the BP responders (16 points) and the no BP group (19 points). All three of these subgroups demonstrated a similar significant improvement in pain relief. This suggests that nonresponding BP may be a marker for poorly controlled pain syndromes. However, the presence of BP by itself was not a predictor of pain relief in our population (Hwang et al., 2002). These observations raise the question of how measures of BP relief are related to overall pain relief. Because of the small number of patients, the results of these subgroup analyses must be interpreted with caution, and studied in larger prospective trials. Treatment of incidental pain remains a challenge due to its fast onset, short duration and spontaneous characteristics. All the patients in this trial were placed on ‘around-the clock’ or ‘long acting’ opioids and ‘as needed’ or ‘rescue’

64

S.S. Hwang et al. / Pain 101 (2003) 55–64

opioids based on the AHCPR guidelines. However, due to the study design, all opioid dosages were converted to 24-h oral MEDD and recorded. We were not able to trace how the rescue doses were taken and how they were related to BP responses. Even the short acting rescue opioids used orally in the study require 45–60 min to work, and they may not be appropriate to palliate the fast-onset incidental pain. Even intravenous rescue doses would require 15 min to work and may not be rapid enough for the fast onset incidental pain. A number of alternatives can be considered for patients with resistant BP. These include intravenous or subcutaneous patient controlled analgesia, oral transmucosal fentanyl citrate (OTFCw), and spinal analgesia. This study was conducted prior to the availability of OTFCw (Farrar et al., 1998). Studies such as ours, while small, may help identify patients at risk for resistant BP, and who would be good candidates for additional therapy. There are some limitations in our study. First, the study was conducted at a VA Medical Center and the results may not be generalizable to all advanced cancer patients. Second, the conclusions were drawn from a small sample size. Further studies to include both genders in community settings with large sample sizes can effectively address these limitations. The study confirmed the heterogeneous and unpredictable characteristics of BP. We were able to decrease the BP incidence rate from 70 to 36% by following the AHCPR cancer pain management guidelines. Patients with bone pain located in the spine, back, and pelvis did not respond as well to conventional analgesic treatment. Patients with persistent BP had brief pains and showed significantly higher pain severity (worst pain and average pain) and interference scores, even with the achievement of acceptable pain relief. Tailored pain management strategies with newly developed, fast acting medications may be helpful in treating BP. Acknowledgements Ms Chris Corpion assisted with the interviews. References Agency for Health Care Policy and Research. Management of cancer pain. Clinical Practice Guideline No. 94-0592. Rockville, MD: Agency for

Health Care policy and Research, US Department of Health and Human Service, Public Health Service, March: 1994. Banning A, Sjogren P, Henriksen H. Treatment outcomes in a multidisciplinary cancer pain clinic. Pain 1991;47:129–134. Bruera E, Scholler T, Wenk R, MacEachern T, Marcelino S, Hanson J, Suarez-Almazor M. A prospective multi-center assessment of the Edmonton staging system for cancer pain. J Pain Symptom Manage 1995;10:348–355. Caraceni A, Portenoy RK. A working group of the IASP Task Force on Cancer Pain. An International survey of cancer pain characteristics and syndromes. Pain 1999;82:263–274. Cella DF, Tulsky DS, Gray G, Sarafian B, Linn E, Bonomi A, Silberman M, Yellen SB, Winicour P, Brannon J, et al. The Functional Assessment of Cancer Therapy Scale: development and validation of the general measure. J Clin Oncol 1993;11:570–579. Chang VT, Hwang SS, Feuerman M, Kasimis B. The Memorial Symptom Assessment Scale Short Form. Validity and reliability. Cancer 2000;89:1162–1171. Daut RL, Cleeland CS, Flanery RC. Development of the Wisconsin Brief Pain Questionnaire to assess pain in cancer and other disease. Pain 1983;17:197–210. Farrar JT, Cleary J, Rauck R, Busch M, Nordbrock E. Oral transmuscosal fentanyl citrate: randomized, double-blinded, placebo-controlled trial for treatment of breakthrough pain in cancer patients. J Natl Cancer Inst 1998;90:611–616. Ferrell BR, Juarez G, Borneman T. Use of routine and breakthrough analgesia in home care. Oncol Nursing Forum 1999;26:1655–1661. Fisher ES, Welch HG. The future of the Department of Veterans Affairs health care system. J Am Med Assoc 1995;273:651–655. Harris RE, Hebert JR, Wynder EL. Cancer risk in male veterans utilizing the Veterans Administration medical system. Cancer 1989;64:1160– 1168. Hwang SS, Chang VT, Fairclough DL, Kasimis B. Development of a cancer pain prognostic scale. J Pain Symptom Manage 2003 (in press). Kazis LE, Miller DR, Clark J, Skinner K, Lee A, Rogers W, Spiro 3rd A, Payne S, Fincke G, Selim A, Linzer M. Health-related quality of life in patients served by the Department of Veterans Affairs. Arch Intern Med 1998;158:26–632. Mercadante S, Maddaloni S, Roccella S, Salvaggio L. Predictive factors in advanced cancer pain treated only by analgesics. Pain 1992;50:151– 155. Petzke F, Radbruch L, Zech D, Loick G, Grond S. Temporal presentation of chronic pain: transitory pains on admission to a multidisciplinary pain clinic. J Pain Symptom Manage 1999;17:391–401. Portenoy RK. Cancer pain: pathophysiology and syndromes. Lancet 1992;339:1026–1036. Portenoy PK, Hagen N. Breakthrough pain: definition, prevalence and characteristics. Pain 1990;41:273–281. Portenoy RK, Payne D, Jacobsen P. Breakthrough pain: characteristics and impact in patients with cancer pain. Pain 1999;21:129–134. Zeppetella G, O’Doherty CA, Collins S. Prevalence and characteristics of breakthrough pain in cancer patients admitted to a hospice. J Pain Symptom Manage 2000;20(2):87–92.