To determine the safety and efficacy of subcutaneous (SC) methylnaltrexone in opioid-induced constipation (OIC).

Double-Blind Phase

Patients were randomized to a single dose of study drug or placebo administered SC.  Groups were 0.15 mg/kg, 0.3 mg/kg, or placebo. Patients were randomly assigned in a 1:1:1 ratio to each study group. Baseline laxative regimens could be continued. Rescue laxatives (laxatives administered on an as needed [PRN] basis) were allowed, except within four hours before or after dose administration.

Open-Label Phase

This phase was 28 days with 1 dose per 24 hours PRN. The initial dose of 0.15 mg/kg could be decreased to 0.075 mg/kg or increased to 0.3 mg/kg, based on response.

Protocol Extension

Patients completing the open-label phase could enter a three-month extension. The initial dose was the same as in the open-label phase, with dosing adjusted to 0.075 mg/kg, 0.15 mg/kg, or 0.3 mg/kg by investigator discretion.

• Of the 154 patients initially randomized, the median age was 66 years, 81.2% had a cancer diagnosis, more than 65% had World Health Organization (WHO) performance status of 3 or 4, and 95% had baseline laxative use.
• Patients were included in the study if they had advanced illness with a life expectancy of one to six months, were on a stable opioid regimen at least three days prior to study entry, had no significant laxation 48 hours prior, and had stable vital signs.
• Patients were excluded if they had previously used methylnaltrexone, naltrexone, or naloxone; process suggestive of gastrointestinal obstruction; or nonopioid constipation, diverticular disease, fecal impaction, acute abdomen, or peritoneal catheter.

Double-Blind Phase

• The age range was 21 to 100 years, with similar median ages across all groups.
• The sample comprised 54.5% men and 45.5% women, with a similar distribution across groups.
• Sixty-seven percent had a WHO performance status of 3 or higher.
• Oral morphine equivalents were 207 mg/day in the 0.15-mg/kg group, 188 mg/day in the 0.3 mg/kg group, and 150 mg/day in the placebo group.
• No significant differences existed in age, constipation distress, or mean pain score across study groups.
• At baseline, 95% used laxatives, 83% used stimulants, 56% used osmotic laxatives, and 27% used stool softener.

Open-Label Phase

• Of the 147 patients who elected the open-label phase, 72 completed the phase.
• Of the remaining 75 patients, 36 withdrew and 39 died.

Extension Phase

• Twenty-seven patients elected extension.
• Nine patients completed extension.

• Patients were recruited from hospice and palliative care settings.
• 17 treatment sites

This was a double-blind, randomized, placebo-controlled, single-dose study, followed by an open-label phase, and then an open-label extension phase.

Efficacy/Primary Outcomes

• Percent of patients who defecated within four hours of dose
• Twenty-four–hour rescue-free laxation rates
• Median time to rescue-free laxation
• Global Clinical Impression of Change (GCIC) scale (seven-point rating scale from “much worse” to “much better”)
• Improvement in stool consistency (patient report, 1 = very hard to 6 = watery)
• Constipation distress (five-point scale from 1= none to 5 = very much)

Secondary Outcomes

• Pain, as measured on a 10-point scale from 0 (none) to 10 (worst possible)
• Modified Himmelsbach opioid withdrawal scale (composite summed score for seven symptoms, each scored on a four-point scale from 1 = none to 4 = severe)
• National Cancer Institute common toxicity criteria, version 2
• WHO performance status

• In the methylnaltrexone 0.15 mg/kg group, 61.7% had laxation within four hours, 64.4% reported improvement in constipation distress, and 58.7% reported improvement in GCIC.
• In the methylnaltrexone 0.3 mg/kg group, 58.2% had laxation within four hours, 63.5% had improvement in constipation distress, and 58.8% had improvement in GCIC.
• In the placebo group, 13.5% had laxation within four hours, 34% had improvement in constipation distress, and 21.6% had improvement in GCIC.
• Comparisons between each methylnaltrexone group and placebo were significant (p < 0.00001).
• In the open-label phase (n = 72), laxation within four hours was experienced by 55.8% in weeks 1 and 2, 61.7% in weeks 3 and 4, and 63.7% at more than 8 weeks.
• Nineteen patients had AEs. Three patients in the open-label phase reported serious AEs.
• A few patients experienced abdominal pain and watery stool.
• No changes were observed in pain or symptoms of opioid withdrawal four hours after dosing.
• In the open-label phase, the 0.3 mg/kg dose was associated with more abdominal pain than the 0.15 mg/kg dose.

SC methylnaltrexone is effective in the treatment of OIC and generally is well tolerated. No relationship exists between dose and laxation response, suggesting the optimal dose is 0.15 mg/kg.

• Patients who responded to SC dosing often defecated soon after drug administration, with 50% responding within 30 minutes. Average time to defecation or timing in all respondents to the medication were not described.  Timing of response may be important for planning patient care related to toileting.
• Three patients were reported as having severe AEs. When they occurred in the open-label phase of study and what percentage of cases were affected was unclear. Investigators noted patients were frail and rapid reversal of constipation in such cases can be associated with AEs. Data for AEs in control and study groups were not provided.
• What impact continued laxative use had on the results was unclear. Many patients were on multiple types of laxatives, as well as the study drug.
• This study reported on patients with a short life expectancy and advanced disease; therefore, findings may not be applicable to other populations.

The study involved three phases: the screening visit, an open-label dose-titration phase, and a double-blind treatment phase. Titration was based on each patient's previous opioid dose. The dose of fentanyl buccal tablet (FBT) was increased stepwise from 100 mcg to 200, 400, 600, or 800 mcg until investigators identified an effective dose. If investigators could determine an effective dose for a patient, the patient entered the next phase of the study. In the double-blind treatment phase, patients were randomly assigned to 1 of 18 double- blind dose sequences (seven FBTs of the dose identified as effective and three placebos) to treat 10 episodes of breakthrough pain (BTP). Ten tablets were labeled 1–10, and patients were to take them consecutively. The treatment sequence was randomly assigned by a statistician who had used a computer to generate it. Patients continued to use their ATC regimen. If no relief occurred within 30 minutes, they continued pretrial supplemental medication.

• Study enrollment consisted of 129 patients; 87 patients entered the double-blind phase. Of these, 86% completed the study and 90% of those were evaluable for efficacy.
• Of all patients, 41% had nociceptive pain, 17% had neuropathic pain, and 42% had mixed pain.

• Multisite
• Outpatient
• Thirty treatment centers in the United States, January 2005–September 2006

Randomized double-blind, placebo-controlled design

• An 11-point (0–10) scale, to measure pain intensity (PI)
• Sum of pain intensity differences (PID) for the first 60 minutes (SPID60)
• PIDs and measure of pain relief (PR), using a five-point Likert scale (0 = none, 4 = complete), assessed from five minutes through two hours
• Patient's assessment of total pain relief
• A five-point scale (0 = poor, 4 = excellent), to measure patient's rating of global medication performance
• Proportion of breakthrough pain episodes requiring supplemental medication
• Adverse events

Investigators were able to identify an effective dose of FBT for 70% of patients. The list that follows cites the single-tablet dose of FBT that the cited percentage of patients found adequate for the treatment of two consecutive breakthrough pain episodes: 100 mcg, 8%; 200 mcg, 12%; 400 mcg, 18%; 600 mcg, 28%; 800 mcg, 34%. SPID60 favored FBT versus placebo (p < 0.0001). PIDs and PR showed significant differences versus placebo at 10 minutes (p < 0.0001). Adverse events were typical (nausea, dizziness, fatigue) and were reported in 66% of participants during the study. Ten percent had adverse events related to FBT application site. Most adverse events occurred during dose titration and were mild and transitory. Difference in SPID60 of 3 was considered clinically relevant. Sample size of 70 was associated with a power of 90%.

For opioid-tolerant patients with chronic cancer pain and breakthrough pain, FBT was efficacious and well tolerated. FBT demonstrated rapid onset (with effect within 10 minutes), and the effect of FBT was sustained.

• Most common reasons for withdrawal were adverse effects, lack of efficacy, and withdrawal of consent; 11% of patients withdrew because of adverse effects; 6%, for lack of efficacy; and 6%, for withdrawal of consent.
• The use of an open-label dose-titration phase may have increased the chance of unblinding by sensitizing patients to the effects of FBT.
• The subgroup that entered the double-blind phase had already demonstrated a favorable response to FBT. Therefore, generalizing the results to the general population is difficult. However, this process is consistent with clinical practice. This study assessed effectiveness of FBT versus placebo, not versus supplemental medications.

To determine whether patients followed by a pivot nurse/nurse coordinator would have less symptom distress and less healthcare resource utilization as evidenced by fewer unscheduled clinic visits, fewer emergency room visits, and hospitalizations than a control group

Patients were randomly assigned to a pivot nurse in addition to usual care (experimental group) or usual care only. The pivot nurse (PNO) was a baccalaureate-prepared palliative care nurse with experience and additional training in cancer symptom management. The PNO met with patients and caregivers in the ambulatory setting to review understanding of the diagnosis, expected side effects of treatment, and resources available to the patient. Patients were taught ways to identify and cope with systems and offered education and support as needed. The PNO advocated for patients in interdisciplinary rounds, developed care plans and referrals, and provided support, information, coaching, etc., via follow-up telephone calls. Usual care included symptom assessment and teaching for management, but not in a formally coordinated manner. With usual care, patients did not necessarily see the same nurse at each appointment, and telephone follow-up was usually patient initiated. The study was conducted over a six-month period. Data were collected at each clinic visit, about every three weeks, for a maximum of eight measurements.

• The study reported on a total sample of 113 patients.
• Mean patient age in the control group was 60.5 years ± 11.1; mean patient age in the experimental group was 59.3 years ± 10.7.
• Patients were diagnosed with breast or lung cancer.
• The sample was 67%–71% female and 29%–33% male.
• Of the total sample, 67%–72% were married; 36% in the control group and 51% in the experimental group had stage III or IV cancer, and 19% in the control group and 23% in the experimental group had metastatic disease; 75%–83% were deemed as having adequate financial resources; and more than 70% in both groups had sufficient help at home.

• Outpatient setting
• Canada

Patients were undergoing the active treatment phase of care.

A randomized controlled trial design was used.

• McCorkle Symptom Distress Scale (SDS)
• Functional Assessment of Cancer Therapy Scale–General (FACT-G), version 4
• Brief Fatigue Inventory (BFI)
• Hospital records

There were no differences in SDS scores over time between groups. Over time, patients in the lung cancer groups had more distress than patients with breast cancer (p = 0.023). There were no significant differences between groups for BFI findings. Patients with lung cancer reported more fatigue (p = 0.002). There were no differences between groups in FACT scores. Significantly lower quality-of-life score were reported by patients with lung cancer (p = 0.0024). There were no differences between groups in healthcare resource utilization data included in this study. Over time, patients with breast cancer were less likely to have hospitalizations shorter than 72 hours than those with lung cancer (p = 0.001).

Care provided by a PNO did not result in any difference in symptom distress or healthcare resource utilization. Patients with lung cancer had higher symptom distress, had more fatigue, and used more healthcare resources, which is in concert with differences in the disease trajectories between lung and breast cancer.

A higher proportion of patients in the intervention group had presence of metastases, and a slightly higher proportion had disease recurrence. It was not stated whether these differences were statistically significant, and these differences could have influenced results in terms of symptoms and resource usage needs. No information was provided in terms of any other chronic healthcare conditions that may have also influenced the results. Authors reported overall SDS results but did not identify actual symptoms experienced or differences at that level between groups. Clinically, different symptoms can be expected to yield different degrees of distress and needs for medical intervention and associated healthcare use. The authors reported a final sample of 113 and results in these, but provided baseline characteristics in 190 patients. One cannot evaluate actual final differences between groups in these characteristics. No power analysis was provided in order to determine if the sample size had sufficient power to detect differences. The study assumes that all practitioners in a given role should be expected to achieve standard results, while this may not be the case. The PNO in this study was not an advanced practice nurse, with associated advanced education. Nurses in the usual care group were highly experienced, and more than 75% were oncology certified. This factor may have caused there to be no substantial clinical difference in the actual nursing care provided for symptom management. The study only lasted six months—effect of better care coordination and symptom management may be more effective over a longer term. No information was provided regarding the time since diagnosis or the phase of care for patients studied, factors that could be expected to influence these aspects of care and patient needs.

This study did not provide any supportive findings for the role of a PNO as implemented in the study. Further research in this area needs to provide the ability to directly contrast this type of intervention with the nursing care provided in usual care. Research in the impact of various roles such as this, navigators, clinical nurse specialists, etc., need to provide better structure, consistency, and definition of these responsibilities and patient interactions. Inclusion of findings related to patient satisfaction with care in this type of research may be helpful.

To determine whether methylphenidate is useful for the management of fatigue in Chinese patients with cancer in the palliative care setting

Oral methylphenidate 5 mg daily; reassessed at day 8 and day 29 (if patient was still in study)

• N = 24  
• MEAN AGE = 64 years (SD = 47–81 years)
• MALES: 80%, FEMALES: 20%
• KEY DISEASE CHARACTERISTICS: Mostly gastrointestinal and lung cancers; patients receiving palliative or hospice care
• OTHER KEY SAMPLE CHARACTERISTICS: Fatigue score of at least 4 on 0–10 scale

• SETTING TYPE: Outpatient palliative care clinic or inpatient hospice
• LOCATION: Hong Kong, China

• PHASE OF CARE: Palliative
• APPLICATIONS: Palliative care

Prospective study

• Brief Fatigue Inventory (BFI)

For patients < 65 years old, scores were significantly lower at day 8 than at baseline but not at day 29. There were no significant differences at day 29 or in patients > 65 years old. Ten out of 24 patients stopped methylphenidate before day 8, with eight of the withdraws due to side effects of medication.

This study demonstrated no clinically significant effect for methylphenidate on cancer-related fatigue.

• Small sample (< 30)
• Risk of bias (no control group)
• Risk of bias (no blinding)
• Risk of bias (no random assignment)
• Subject withdrawals ≥ 10%
• Other limitations/explanation: There were a large number of refusals to participate in approached patients.

Methylphenidate is not recommended for the management of cancer-related fatigue.

The study reported the effects of tibolone 2.5mg daily on climacteric symptoms, vaginal dryness, and health-related quality of life in breast cancer survivors.

Patients were randomized one-to-one to tibolone 2.5 mg by mouth daily or 1 placebo pill by mouth daily, with mean duration of treatment of 2.75 years.

The study enrolled 3098 women, with 1556 on tibolone and 1542 on placebo. Combined treatment and placebo groups mean age was 52.7 (SD=7.3), range = 28-75.
    

• KEY DISEASE CHARACTERISTICS:  Non-metastatic breast cancer T1-3, N0-2, M0
• INCLUSION CRITERIA: Postmenopausal, younger than age 75 and treated in past 5 years.  At entry to study 6.5%were taking aromatase inhibitors, 66.6% taking tamoxifen, 4.3% taking GnRH analogues, 16.7% post oophorectomy.  Race and ethnicity are not reported.

This was a multi-site, multi-national study conducted in at least eight countries: Austria, Belgium, Germany, Spain, France, United Kingdom, Italy, The Netherlands.
 

• PHASE OF CARE: Long term followup
• APPLICATIONS: Late effects & survivorship

The study was a multinational, multicenter, randomized, double-blind, parallel group, placebo-controlled trial.

Measurements and instruments included:

• Daily hot flash diary - frequency and severity - with calculated composite score    
• Climacteric symptoms form - hot flash frequency per day, change from baseline, maximum intensity of hot flashes, sweats, interference of flushes/sweats with normal life, palpitations, joint pain, vaginal dryness, incontinence
• Women’s Health Questionnaire - health-related quality of life,  subset of participants
   

Compared to placebo, tibolone resulted in a significantly greater reduction in:

1. Number of hot flashes per day at 12 and 104 weeks of treatment (p’s < 0.0001)weeks
2. Mean composite hot flash severity at week 12 (p < .0006)
3. Mean vaginal dryness score at week 104 (p < 0.0001)
4. WHQ scores in vasomotor, sexual and sleep at weeks 28, 52, 78, and 104 (p < .05)
5. WHQ scores of attraction at week 78 and mood at weeks 26, 52, and 78 (p’s < 0.05). 

There were interaction effects of tamoxifen and AI such that those using those therapies obtained less relief in hot flashes and climacteric symptoms with tibolone. 

Tibolone 2.5 mg orally daily was effective in alleviating menopausal symptoms in breast cancer survivors overall, but was less effective in tamoxifen users.

 A very small percentage of participants were on AIs or GnRH analogues at study entry.

Despite efficacy, the main trial report published in another journal indicated that tibolone increased the risk of breast cancer recurrence and is therefore contraindicated as a menopausal symptom therapy in breast cancer survivors.   Also, placebo effect was evident and persistent.

To investigate the effect of education and exercises on development and progression of lymphedema

Patients were informed about measures to prevent lymphedema and about exercises. They were given written material prepared by the investigators. No further specifics about the education or exercises is provided.

• The sample (N = 55) was comprised of 1.8% male and 98.2% female patients.
• All patients had breast cancer.
• Ninety-six percent of patients had modified radical mastectomy and 45% had adjuvant radiotherapy.

The study took place in an outpatient setting in Turkey.

The study used a prospective observational design.

Arm circumference was measured.

Authors compared the percent of patients with minimal to severe lymphedema between those who exercised and those who did not over a six-month period; however, only 10 patients were noted to not exercise and sample sizes used in analysis were extremely small per severity group. Some patients who had lymphedema at study entry were stated to have no lymphedema at week 6.

 Results are inconclusive given multiple limitations of the study.

• The study had a small sample size, with less than 100 participants.
• The study had risk of bias due to no blinding, no random assignment, and no control group.
• Measurement and methods are not well described, making the measurment validity and realiablity questionable.
• The study does not provide clear information about node dissection and radiotherapy in relation to lymphedema development, does not fully describe the intervention or any other standard care interventions (e.g., bandanging), and does not describe how severity of lymphedema was determined.
• There was only a single point at which arm circumference was measured, which is a questionable measurement method. 
• Time since surgery was highly varied with a range of 1–105 months.

Study findings are inconclusive regarding the effect of patient education and information to prevent or manage lymphedema in patients with breast cancer.  Findings provide minimal support for use of exercise because of study report limitations.

STUDY PURPOSE: To examine the effects of exercise on cancer-related lymphedema and associated symptoms, and to determine if wearing compression during exercise is needed for individuals with lymphedema

TYPE OF STUDY: Meta-analysis and systematic review

• FINAL NUMBER STUDIES INCLUDED = 21 (exercise) and 4 (compression and exercise) studies met inclusion criteria.
• TOTAL PATIENTS INCLUDED IN REVIEW = Not clearly described
• SAMPLE RANGE ACROSS STUDIES: Exercise studies (unclearly reported): five studies (sample sizes < 15), nine studies (sample sizes 15–30), four studies (sample sizes 31–60), and three studies (sample sizes > 61); studies with compression and exercise: four studies (sample size range: 18–31)
• KEY SAMPLE CHARACTERISTICS: Women with unilateral breast cancer-related lymphedema, male and female participants with lower limb lymphedema

PHASE OF CARE: Late effects and survivorship
 
APPLICATIONS: Palliative care

Although exercise did not result in a worsening of lymphedema and associated symptoms of the affected limb, no statistically significant effect of exercise on lymphedema or related symptoms existed. Subgroup analyses for exercise mode (aerobic, resistance, mixed, and other) and intervention duration (> 12 weeks or 12 weeks) showed consistent results, that is, no effect on lymphedema or associated symptoms. Too few studies existed to conduct a meta-analysis for evaluating the effect of compression during regular exercise.

Exercise appeared to have no effect on lymphedema and associated symptoms in individuals with secondary limb lymphedema. However, the findings indicate that individuals with secondary limb lymphedema can safely perform exercise without experiencing a worsening of lymphedema or related symptoms. Insufficient evidence exists to support or disprove the current clinical recommendation to wear compression garments during regular exercise. In addition, injuries from exercises do occur, which were reported in several studies, yet the review did not mention the injuries.

• No quality evaluation
• Low sample sizes
• The majority of study samples had stable, mild, unilateral breast cancer-related lymphedema and were likely to be physically active before study participation.

Given the health benefits of exercise on the overall quality of life of survivors, the findings from this review and meta-analysis suggest that nurses can educate cancer survivors with limb lymphedema to conduct progressive/supervised regular exercise, which likely will not worsen lymphedema or associated symptoms. However, injury does occur with exercise, so reporting it is important, but the review did not mention potential injury with exercise.

To demonstrate improvement in constipation in individuals on prolonged-release (PR) oxycodone and PR naloxone compared with individuals receiving single-agent PR oxycodone.

Prerandomization comprised a run-in phase for conversion and titration of prestudy pain medication regimen to the PR oxycodone and bisacodyl laxative regimen. In the double-blind phase, patients were randomized to receive one of the following for 12 weeks: PR oxycodone/PR naloxone in a 2:1 ratio and PR oxycodone placebo, or PR oxycodone alone and PR oxycodone/PR naloxone placebo. All patients completing the double-blind phase were eligible to enter a 52-week extension phase and receive PR oxycodone/PR naloxone.

• The study reported on a sample of 277 patients aged 18 years or older.
• Patients had a documented history of the following:
  • Moderate-to-severe chronic noncancer pain (eg, musculoskeletal, neuropathic, visceral), needing 24-hour coverage with an oxycodone equivalent of 20 to 50 mg/day
  • Constipation caused or worsened by an opioid
  • Expectation of benefit from the World Health Organization (WHO) step III opioid therapy treatment for the duration of the study.

• Multi-site
• Outpatient
• 4 European countries

This was a randomized, double-blind, parallel-group, phase III study.

• Bowel Function Index (BFI)
• Patient Assessment of Opioid Induced Constipation (PACOI)
• Patient Assessment of Constipation Symptom Questionnaire (PAC-SYM)
• Pain Intensity Scale
• Brief Pain Inventory (Short Form) (BPI-SF)

• At week 4, the mean BFI score for the PR oxycodone/PR naloxone group decreased by 26.9 points. In contrast, the mean BFI score for the PR oxycodone group decreased by 9.4 points. The difference between those scores was significantly in favor of the PR oxycodone/PR naloxone group (p < 0.0001).
• At week 4, the mean PACOI score in the PR oxycodone/PR naloxone group decreased to 6.4% compared to 9.4% in the PR oxycodone group, showing statistically significant improvement (p < 0.0001).
• During the first four weeks, significantly fewer patients in the PR oxycodone/PR naloxone group took bisacodyl compared with the PR oxycodone group (p < 0.0001).
• No statistically significant difference existed between the two groups for efficacy of pain control.
• Mean pain intensity scores remained stable throughout the study.
• Twelve patients (three in the PR oxycodone/PR naloxone group and 9 in the PR oxycodone group) experienced serious adverse events, including gastrointestinal bleeding, headache, and cerebrovascular accident.

PR oxycodone/PR naloxone demonstrated superiority in the management of constipation in patients with chronic noncancer pain without compromising analgesia.

The study only included patients with noncancer pain.

PR oxycodone/PR naloxone may be effective in the management of constipation without compromising pain control for patients with chronic pain. However, the study did not include patients with cancer or patients receiving doses of oxycodone equivalent higher than 50 mg/day. Additional studies are warranted with higher doses of opioids and the inclusion of patients with cancer.

STUDY PURPOSE: To evaluate the effectiveness of benzodiazepines in relieving dyspnea in individuals with advanced disease; in addition, to compare the effectiveness of different benzodiazepines and different dosages, routes of administration, side effects, as well as a comparison of effectiveness in various diseases

TYPE OF STUDY: Meta-analysis and systematic review

• FINAL NUMBER STUDIES INCLUDED =  8 studies in qualitative synthesis and 7 studies in quantitative synthesis
• TOTAL PATIENTS INCLUDED IN REVIEW = 214 with advanced cancer and COPD
• SAMPLE RANGE ACROSS STUDIES: 5–101
• KEY SAMPLE CHARACTERISTICS: Adults suffering from dyspnea related to advanced malignant and nonmalignant diseases. Advanced diseases meeting inclusion criteria were cancer, COPD GOLD stage III or IV, HF NYHA class III or IV, MND, and IPF; however, all relevant studies included only individuals with cancer and COPD. The interventions included the use of any benzodiazepines at any dose, frequency, route, duration, or route.

PHASE OF CARE: Multiple phases of care
 
APPLICATIONS: Elder care, palliative care

There is currently insufficient evidence to recommend the use of benzodiazepines for the prevention or relief of dyspnea in individuals with cancer and COPD. The adverse effects of somnolence is more prevalent with benzodiazepines than placebo; however, somnolence is more prevalent when treating dyspnea with morphine compared to benzodiazepines. Results must be interpreted with caution because of limited quality and high heterogeneity in the studies evaluated.

• No quality evaluation
• High heterogeneity
• Low sample sizes
• Heterogeneity in the benzodiazepine group, disease group, control group, and others

Additional high quality studies are needed to fully evaluate the impact of benzodiazepines on dyspnea. Treatment of dyspnea with benzodiazepines does have side effects with potentially no benefit. Given uncertain benefits of treating dyspnea with benzodiazepines, interventions for management of dyspnea should include nonpharmacological approaches as first-line when appropriate. Assessment of response to benzodiazepines administered to treat dyspnea should include knowledge of potential benefits and potential burdens of the medication and their impact on overall quality of life; for example, drowsiness may be an acceptable side effect for some but not others.

The primary aim of the study was to determine the efficacy of benzodiazepines for the relief of breathlessness in patients with advanced disease. The secondary aim was to determine the efficacy of different benzodiazepines, different doses of benzodiazepines, different routes of benzodiazepines, adverse effects of benzodiazepines, and the efficacy in different groups for the relief of breathlessness .

Databases searched were the Cochrane Pain, Palliative, and Supportive Care Trials Register (September 2009), Cochrane Central Register of Controlled Trials (Central) in the Cochrane Library (September 2009), Cochrane Database of Systematic Reviews in the Cochrane Library (September 2009), Database of Abstracts or Reviews of Effectiveness (September 2009), MEDLINE (1950–2009), EMBASE (1980–1989 and 2009), CINAHL (1980–1989 and 2009), PsycINFO (1806–1809 and 2009), American College Physicians Journal Club (September 2009), Health Technology Assessment (September 2009), NHS Economic Evaluation Database (September 2009), Database of Halley Stewart Library (St Christopher’s Hospice) (September 2009), International Pharmaceutical Abstracts (1970–1979 and 2009), and Iowa Drug Information System (1966–1969 and 2009).

Search strategies for the 14 databases included variations of the following keywords: dyspnea, breathing, breathless, shortness of breath, breathing difficult, and breathing labour paired with benzodiazepine, anxiety agents, and a long list of specific benzodiazepine agents.

Randomized controlled trials and controlled clinical trials assessing the effect of benzodiazepines in relieving breathlessness in patients with advanced stages of cancer, chronic obstructive pulmonary disease (COPD), chronic heart failure (CHF), motor neuron disease (MND), and idiopathic pulmonary fibrosis (IPF) were included.

Studies using all drugs in the pharmacologic class called benzodiazepines at any dose, frequency, duration, and through any route for the relief of breathlessness compared with placebo or active control were included.

Studies were excluded if they

• Were not controlled or randomized trials
• Included participants with acute or chronic asthma, pneumonia, or other potentially curable diseases.

A total of 1,309 references were reviewed initially from the databases, which were narrowed to 31 articles for closer evaluation. The final evaluable seven studies included seven randomized controlled trials, five crossovers, and two parallel designs, four with COPD and three with cancer. All studies were initially assessed for quality using the Review Manager (RevMan) and secondarily evaluated using “The Edwards Method Score,” and articles were graded for inclusion in data analysis or the meta-analysis if high quality. Two studies used alprazolam, one study used diazepam, two studies used midazolam with morphine, one study used lorazepam, and one study used clorazepare.

• Of the 200 participants analyzed, 52 had COPD and 148 had cancer.
• All studies were small, ranging from 5-29 participants, except for two studies (Navigante, 2006 with 101 participants and Naviagante, unpublished with 63 participants).
• Three studies had power calculations, but only two reached adequate numbers for power.
• Three studies included patients with cancer.
• Most stuides used the VAS or NRS scale for assessing breathlessness.
• Most studies measured breathlessness at rest, and three studies measured breathlessness with exercise.
• Anxiety was measured in two studies, depression in two studies, walking tests in three studies, and benzodiazepine adverse effects or attrition in all studies.

Only six of the seven studies were included in meta-analysis, and the other was included in general data. Other measured outcomes of the studies included anxiety, depression, adverse effects of benzodiazepines, functional exercise capacity, quality of life, and study attrition. Overall, the analysis (four studies) and meta-analysis (three studies) with 52/47 participants showed no significant effect of three different benzodiazepines in relief of breathlessness in patients with advanced COPD. The three studies of patients with cancer included in analysis included two with morphine control and one with placebo control. One morphine-controlled study showed no significant effect of midazolam as compared to morphine, and one showed a slightly better improvement of breathlessness in patients receiving midazolam. Although overall no effect of benzodiazepines could be demonstrated, this meta-analysis should be interpreted with caution given the hetereogeneity and design differences of these studies. Pooling of placebo-controlled and morphine-controlled data showed no significant effect of benzodiazepines on breathlessness at rest. Four of seven studies measured anxiety with different scales, and none demonstrated anxiety alterations from baseline or as compared to a control group. Three studies examined depression and did not show differences between the intervention and placebo groups.

When considering all studies, no enhanced effectiveness for management of breathlessness was noted with use of benzodiazepines either at rest or with breakthrough dyspnea for patients with COPD or cancer. When excluded studies with lesser research strength of evidence were compared with stronger evidence, these conclusions were affirmed.

Although overall no effect of benzodiazepines could be demonstrated, this meta-analysis should be interpreted with caution given the hetereogeneity and design differences of these studies.

The authors recommend larger studies with more participants, inclusion of more patients with other known etiologies of breathlessness (e.g., CHF, MND), treatment of breakthrough dyspnea, and use of benzodiazepines in patients with breathlessness during panic attacks.