To establish feasibility of studying the effects of yoga on function, pain, and quality of life in women with aromatase inhibitor–associated arthralgias

Women who had joint pain attributed to aromatase inhibitor treatment were studied. Patients met two times per week for eight weeks for yoga sessions taught by certified instructors. The protocol used was inspired by Iyengar yoga involving precise postures, and meditation using relaxation and combinations of static and active stretching and isometric and dyunamic strengthening. Sessions were done among groups of 5–10 women, and an abbreviated version was given for home practice during week 2. Patients were asked to perform home practice for 15 minutes three times per week on days when sessions did not take place. Study assessments were done at baseline and at the end of the program.

• The study reported on 10 female patients.
• Median patient age was 58 years, with a range of 50–71 years.
• All patients had breast cancer and joint pain,  were post-menopausal, and in active treatment with aromostase inhibitors.
• Of the sample, 90% were Caucasian, 40% were employed, and 60% had at least college-level education.

• Multisite
• Outpatient setting
• New Jersey

Patients were undergoing active antitumor treatment.

This was a quasi-experimental feasibility study.

• Functional reach and sit and reach testing
• Self-reported Patient Specific Functional Scale
• Functional Assessment of Cancer Therapy–Breast (FACT-B)
• Brief Pain Inventory (BPI)

Of the sample, 80% reported adherence to home practice as recommended. Participants had significant improvement in functional reach (p = 0.048) and sit and reach (p = 0.009). Participants experienced significant reduction in pain severity (3.9–2.79, p = 0.016).

The yoga protocol used here was associated with improvement in flexibility and pain associated with aromatase inhibitor–induced arthralgia.

• The study had a small sample, with less than 30 participants.
• The study had risk of bias due to no control group, no blinding, and no random assignment, and due to sample characteristics.
• Findings are not generalizable.
• Effects were related to arthralgia-related function and pain only.

There is limited evidence regarding interventions to reduce arthralgia pain in patients undergoing cancer treatment. This study shows that a yoga intervention is feasible and may provide some promising results. Further research in this area is warranted.

To evaluate the feasibility of tai chi to improve well-being for women with breast cancer treatment-associated arthralgia

Women met twice weekly for eight weeks for group tai chi under supervision. The program was a gentle, low-impact form of tai chi focusing on body awareness, deep breathing, and weight bearing. Women were provided written information for home practice. Participants were asked to complete journal entries after each tai chi session and home exercise, encouraged to maintain usual activities, and asked to refrain from other exercise during the study. Study measures were obtained at baseline and at the end of 8 weeks.

• N = 12  
• MEAN AGE = 59 years (range = 49-76 years)
• FEMALES: 100%
• KEY DISEASE CHARACTERISTICS: All had completed initial breast cancer treatment and were post-menopausal, currently free of disease, and on aromatase inhibitors.  
• OTHER KEY SAMPLE CHARACTERISTICS: 91.6% were white, one-third were employed full-time, and 58% had some college education.

• SITE: Single site
• SETTING TYPE: Outpatient   
• LOCATION: New Jersey

PHASE OF CARE: Late effects and survivorship

Single-group observational, mixed-method, feasibility

• Functional Assessment of Cancer Therapy–Breast (FACT-B)
• Brief Pain Inventory
• Hospital Anxiety and Depression Scale (HADS)
• Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-fatigue)
• Qualitative analysis of journal entries

There was a reduction in anxiety from 8.0 to 5.7 (p = .003) and in depression from 5.17 to 2.42 (p = .02). A positive, but non significant, reduction in fatigue and pain occurred. Themes from analysis of qualitative results were improved relaxation and reduced stress, an increase in undisturbed sleep, and perceived value from the group and instructor support. There were no adverse events.

Tai chi participation appears to be feasible for breast cancer survivors and may have positive effects on anxiety and other symptoms.

• Small sample (< 30)
• Risk of bias (no control group)
• Risk of bias (no blinding)
• Risk of bias (no random assignment)
• Risk of bias (no appropriate attentional control condition) 
• Risk of bias (sample characteristics)
• Other limitations/explanation: Adherence to sessions is not reported. Sample was limited to women on aromatase inhibitors experiencing pain from arthralgia.

Tai chi sessions are feasible for cancer survivors and may be of benefit. This combination of relaxation and exercise disciplines may be helpful and acceptable to some patients. Study findings here suggest that the supportive nature of instructor-led group sessions contributed to the positive results.

The study was conducted to investigate the clinical improvement observed in patients with advanced cancer and hypoactive delirium after the administration of methylphenidate hydrochloride (MPH).

First, a 10 mg test dose of MPH was given orally to all participants. If there were no distressing side effects, participants were given 10 mg of MPH twice daily at 8 am and 12 pm. Follow-up was daily for hospital inpatients and every three to four days for patients in the community. Doses of MPH were increased in 5 mg increments to reach the maximum tolerable dose for the patient’s satisfaction and the resolution of delirium.

Measurements were taken before delirium and delusions (T0), at baseline prior to the MPH treatment (T1), one hour after after the MPH dose (T2), and when a stable dose was achieved (T3). 

The Mini-Mental State Examination was used as an assessment tool on a daily basis for inpatients and every three to four days for outpatients.

• All participants had a diagnosis of advanced metastatic cancer. Participants had lung, colon, breast, prostate, cervical, or testicular cancer. 
• The total number of participants was 14.
• Participants were between the ages of 41–80 years.
• 64% of participants were male and 36% were female.

The study utilized a case series design for patients with advanced cancer and hypoactive delirium.

• The Mini-Mental State Examination (MMSE) measured global cognitive function.
• No specific information was provided on the instruments used to assess psychomotor retardation, sleep and drowsiness, absence of delusions, or absence of delirium.

All participants had a positive response to MPH that included increased alertness, partial-to-complete resolution of psychomotor retardation, normalization of slurred speech, and a marked increase in energy levels.

All 14 participants showed improvement in their cognitive function as documented by the MMSE. In 13 patients, the median MMSE score improved to 28 (mean = 27.8, standard deviation = 2.4, p = 0.02) compared with the median score one hour after the first dose of MPH. One patient died before reaching a stable dose of MPH.

The pretreatment MMSE median score was 21 (mean = 20.9, standard deviation = 4.9), which improved to a median of 27 (mean = 24.9, standard deviation = 4.7) after the first dose of MPH (p < 0.001).

MPH improved alertness and general cognitive function in a small sample of patients with advanced cancer. However, due to confounding issues with disease and treatment responses, more research is warranted to determine its effectiveness.

• The study had a small sample size.
• The study was limited to patients with advanced cancer.
• All participants received MPH, so there was no control group.
• The study had a limited measurement of cognitive function, as the MMSE is a global indicator with known practice effects.
• The study did not explain its assessment of psychomotor functioning or its determination of sleep and drowsiness.
• There were confounding issues with brain metastases in 29% of participants.
• There were confounding issues with pain medications in 57% of patients.
• There was no discussion regarding the clinical significance of the changes in MMSE scores.

To evaluate the degree to which a multi-component rehabilitation program improves symptom control and quality of life in patients with advanced cancer

The intervention was a 10-12 week program offered by a multidisciplinary team, consisting of nutritional counseling, a collaborative care plan based on patient goals, a palliative care physician specialist focused on symptom-related medical interventions, a pivot nurse for care coordination and case management, and an exercise component with semi-weekly exercise sessions with a physical therapist and a home exercise plan. Occupational therapy was provided and focused on self care, leisure, and productivity. Patients were assessed at baseline and during their final clinic visit at the end of the study.

• N = 131  
• MEAN AGE = 59.9 years (SD = 13.0 years)
• MALES: 50.4%, FEMALES: 49.6%
• KEY DISEASE CHARACTERISTICS: All had stage III and IV disease with a variety of primary tumor types including both hematologic and sold tumors.
• OTHER KEY SAMPLE CHARACTERISTICS: 38% were on current chemotherapy; most had ECOG performance status of 1 or 2.

SITE: Single site  

SETTING TYPE: Outpatient  

LOCATION: McGill University Cancer Center, Montreal, Canada

Quasi-experimental

Change in symptom severity was analyzed and Cohen’s d was used to calculate effect size. Severity of depression from ESAS declined (p <. 0001, d = 0.7); anorexia declined (p < .0001, d = .4);  pain declined (p < .0001, d = .4); physical and general fatigue declined (p < .0001, d = .7); mental fatigue declined (p < .0005, d = .4); and level of distress and difficulty coping declined (p < .0001).

The multi-component rehabilitation program provided here resulted in a significant improvement in multiple symptoms and a reduction in distress and difficulty coping.

• Risk of bias (no control group)
• Risk of bias (no blinding)
• Risk of bias (no random assignment) 
• Risk of bias (no appropriate attentional control condition)  
• Unintended interventions or applicable interventions not described that would influence results
• Subject withdrawals ≥ 10%
• Other limitations/explanation: 30% of patients withdrew from the study–the majority of these were due to death or disease progression. No information is provided about medication changes over the course of the study that might affect outcomes measured. The report states some different results in the body of the article versus tables provided.

A multi-component, multi-disciplinary rehabilitation and palliative care program can provide effective improvement of multiple symptoms in patients with advanced disease.

To compare antibiotic prophylaxis with placebo or no intervention or another antibiotic in patients with afebrile neutropenia

DATABASES USED: Cochrane Cancer Network Register of Trials (December 2004), Cochrane Library (Issue 4, 2004), EMBASE (January 1980–December 2004), and MEDLINE (January 1966–December 2004); the reference lists of all the articles also were searched.

FINAL NUMBER STUDIES INCLUDED = 95 RCTs

TOTAL PATIENTS INCLUDED IN REVIEW = 9,283 patients comparing prophylactic antibiotics with placebo, no intervention, or other prophylactic antibiotics

KEY SAMPLE CHARACTERISTICS: Sixty-four trials included only patients with hematologic malignancies, and nine trials consisted of more than 80% of patients with solid tumors. Twenty-seven studies included patients undergoing bone marrow transplantation (BMT).

Prophylactic antibiotics significantly decreased

• Overall mortality by 33% (although the effect was less robust in the well-designed trials)
• The risk of infection-related death by 42%
• Fever by 21%
• Clinically documented infections by 36%
• Microbiologically documented infections by 46%
• Gram-negative infections by 61%
• Gram-positive infections by 58%
• Bacteremia by 48%.

Fluoroquinolones, when compared with placebo or no intervention, decreased the risk of

• Overall mortality compared with placebo or no intervention by 48%
• Infection-related mortality by 62%
• Clinically documented infections by 47%
• Microbiologically documented infections by 50%
• Gram-negative infections by 74%
• Gram-positive infections by 71%
• Bacteremia by 36%.

The relative risk for adverse events was not statistically significant (relative risk 1.30 [confidence interval 0.61–2.76]). Comparatively, in trials comparing trimethoprim/sulfamethoxazole with placebo or no intervention, the corresponding estimates were statistically significant (relative risk 2.42 [confidence interval 1.35–4.36] and 3.63 [confidence interval 1.32–9.98], respectively). Moreover, in trials that compared fluoroquinolones with trimethoprim/sulfamethoxazole, less resistance developed to fluoroquinolones in the fluoroquinolone group than that developed to trimethoprim/sulfamethoxazole in the trimethoprim/sulfamethoxazole group (relative risk 0.45 [confidence interval 0.27–0.74]). When fluoroquinolones were compared with placebo, the number of fungal infection episodes did not statistically or significantly differ (relative risk 0.83 [confidence interval 0.56–1.22]).

Fluoroquinolone prophylaxis increased the risk of fluoroquinolone-resistant infections, but the increased risk was not statistically significant (relative risk 1.69 [confidence interval 0.73–3.92]).

The purpose of this meta analysis and sytematic review was to evaluate the effect of antibiotic prophylaxis on mortality and infection in neutropenic patients. In addition, subgroups of patients who may benefit the most were identified, and whether or not the effectiveness of different antibiotic regimens were similar was evaluated, as were the adverse effects of different regimens and the emergence of quinolone-resistant bacteria.

• Databases searched from 1996–2011 included MEDLINE, EMBASE, Cochrane databases of controlled trials and cancer network registry of trials, multiple relevant conference proceedings.
• Regarding keywords, appendices provided extensive detail of exact search terms used per database.
• Inclusion criteria included patients with cancer and neutropenia from chemotherapy or after bone marrow transplantation. Control groups received placebo, no intervention or an alternate intervention, quasi-randomized controlled trials, or randomized clinical trials (RCTs) for initial review through 2005. An update to 2011 included only RCTs. Exclusion criteria were not stated,.

119 total references were retrieved. Cochrane Handbook for Systematic Reviews methods were used to evaluate and commend on the literature used.

• 109 total studies were evaluated.
• 13,579 cases were reviewed.
• Key characteristics: patients with cancer—sites not specified—but most involved patients with leukemia. Some studies sampled febrile episodes rather than cases.

Active antitumor treatment

Antibiotic prophylaxis resulted in significant reduction in risk of mortality across 46 trials analyzed (RR = 0.66, 95% confidence interval [CI] [0.55, 0.79], p < 0.00001). The greatest effect was with quinolones, although differences between regimens was not statistically significant. The effect was larger for trials in which prophylaxis was begun at the onset of neutropenia. An advantage was seen for all quinolones except for norfloxacin. Antibiotic prophylaxis significantly reduced infection-related mortality (RR = 0.61, 95% CI [0.48, 0.77], p = 0.04), decreased occurrence of fever, documented infection, and occurrence of bacteremia. Quinolones and TMP-SMZ were both associated with side effects that were mostly diarrhea and nausea. TMP-SMZ was associated with drug resistant bacteria cultures (RR = 2.42, 95% CI [1.35, 4.36]).  With quinolones, no significant differences were noted between study groups compared to placebo or other interventions. Addition of gram-positive coverage did not show any apparent benefits in terms of mortality.

Findings support use of quinolones as prophylaxis of choice since they reduced risk of death compared to placebo or not intervention and were generally associated with fewer side effects and less resistant bacterial cultures in treated patients. Levofloxacin or ciprofloxacin are recommended.

• All cause mortality was only available for 47 of the studies. 
• Length of follow-up in studies may have been too short to fully detect emergence of resistant bacteria. 
• Most studies were limited to patients with hematologic cancers. 
• Applicability to patients with solid tumor types requires further study.

Prophylactic quinolone antibiotic therapy is recommended for patients with hematologic cancers and those who are likely to develop neutropenia. Additional research is needed to better define patients with solid tumors that may benefit from antibiotic prophylaxis. In most studies, prophylaxis was begun when chemotherapy was initiated, rather than when neutropenia occurred.  Prophylaxis should be accompanied by surveillance to monitor quinolone-resistant gram-negative bacteria and other resistant organisms.

To compare antibiotic prophylaxis with placebo, no intervention, or another antibiotic to prevent bacterial infections in patients with afebrile neutropenia

DATABASES USED: Electronic searches on the Cochrane Cancer Network Register of Trials (2005), Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library Issue 4, 2005), MEDLINE (1966–2005), EMBASE (1980–2005), and abstracts of conference proceedings; references of identified studies; the first author of each included trial was contacted.

FINAL NUMBER STUDIES INCLUDED = 101

TOTAL PATIENTS INCLUDED IN REVIEW: 12,599

KEY SAMPLE CHARACTERISTICS: RCTs or quasi-RCTs performed from 1973–2005; patients with cancer and neutropenia as a result of chemotherapy or bone marrow transplantation

Antibiotic prophylaxis significantly decreased the risk of death when compared with placebo or no intervention (RR 0.66 [95%CI 0.55 to 0.79]). The authors estimated the number needed to treat in order to prevent one death from all causes as 50 (95% CI 34 to 268). Prophylaxis with any antibiotic resulted in a significant decrease in the risk of infection-related death (RR 0.59 [95% CI 0.47 to 0.75]) and in the occurrence of fever (RR 0.77 [95% CI 0.74 to 0.81]). Quinolone prophylaxis reduced the risk for all-cause mortality (RR 0.52 [95% CI, 0.37 to 0.74] and the risk of infection-related mortality (RR 0.49 [95% CI 0.31 to 0.77]). 

Antibiotic prophylaxis resulted in a significant decrease in the occurrence of clinically documented infection (RR 0.66 [95% CI 0.61 to 0.73]), microbiologically documented infection (RR 0.53 [95% CI 0.48 to 0.58]), microbiologically documented gram-negative infection (RR 0.38 [95% CI 0.32 to 0.45]), microbiologically documented gram-positive infection (RR 0.44 [95% CI 0.38 to 0.51]), and bacteremia (RR 0.52 [95% CI 0.47 to 0.59]. Quinolone prophylaxis reduced the risk of bacteremia (RR 0.58 [95% CI 0.50 to 0.68]. When compared to placebo or no intervention, all prophylactic antibiotics caused more side effects (RR 1.59 [95% CI 1.37 to 1.85]. There was no statistically significant difference in the number of episodes of fungal infection when prophylactic antibiotics were compared to placebo (RR 1.07 [95% CI 0.83 to 1.37, 38 studies, 2,682 participants]).

When compared to placebo, patients given quinolones and sulfamethoxazole/trimethoprim (SMZ-TMP) were found to be at increased risk of harboring bacilli resistant to the specific drug than patients receiving placebo (RR 1.47 [95% CI 1.08 to 2.01]). For quinolones, the RR was 1.18 (95% CI 0.81 to 1.70) and for SMZ-TMP, 2.42 (95% CI 1.35 to 4.36). When quinolones were compared to SMZ-TMP, the following were significantly reduced: microbiologically documented infections (RR 0.72 [95%CI 0.6 to 0.86]) (comparison 5.2), gram-negative infections (RR 0.21 [95% CI 0.13 to 0.36]) (comparison 6.2), gram-negative bacteremia (RR 0.35 [95% CI 0.21 to 0.59]), and side effects (RR 0.74 [95%CI 0.63 to 0.87]). The addition of antibiotic against gram-positive infection to quinolones resulted in a significant decrease in the number of bacteremic episodes (RR 0.72 [95%CI 0.57 to 0.92], gram-positive infections (RR 0.49 [95% CI 0.37 to 0.64], and gram-positive bacteremia (RR 0.61 [95% CI 0.45 to 0.83]), and also in more side effects.

• Most studies were limited to patients with hematologic cancer (mostly leukemia).
• Most were conducted on hospitalized patients.
• Information on all-cause mortality could not be obtained for all of the studies.
• Many studies were dated.

Acupuncture-like transelectrical nerve stimulation (AL-TENS) with low-frequency, high-intensity stimulation using acupuncture points for emesis and analgesia was delivered by a nurse practitioner in five consecutive daily treatments. The study was divided into three groups:  AL-TENS, standard care, and standard care plus placebo.

• The study included 15 adults with multiple cancer diagnoses admitted for symptom control for pain or nausea and vomiting.
• Complete data were collected for 13 participants.
• Age ranged from 38 to 74 years.
• Of the participants, 14 were female, all were Caucasian, and three were in the end-of-life phase of care.

• Hospice
• United Kingdom

The study was a pilot study and a double-blind, randomized, controlled trial.

• European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30) fatigue subscale at baseline and day six (posttreatment)
• Pain
• Nausea and vomiting
• Quality of life (QOL)

• High baseline levels of electrical resistance were observed.
• Fatigue decreased compared to the controls.

No significant differences were observed, but the study was underpowered and groups were not equivalent in symptoms at baseline.

• The very small sample led to a lack of power and inability to control covariates.
• Groups were not equivalent in baseline symptoms.
• The power estimates for QOL seemed inaccurate.
• No information was provided related to patient tolerance of treatment or adverse events.

Nurses should be trained in the use of AL-TENS and identification of acupuncture points. Future trials focused on fatigue are recommended.

To determine the pharmacokinetics, safety, and efficacy of two dosing regimens of APF530

There were two separate studies reported in this paper. The first study included 45 patients and used three escalating dosing schedules of 250 mg, 500 mg, or 750 mg. The second study included 35 patients with two dosing schedules of 250 or 500 mg. Safety and efficacy were reported. Drug levels were measured from predose to 168 hours after administration. Doses were given via subcutaneous injection in the abdomen prior to chemotherapy. All patients also received dexamethasone.

• N = 80
• MEAN AGE = 64 years (SD = 12.5 years [trial 1]), 55.7 years (SD = 8.7 years [trial 2])
• MALES: 40% (trial 1), FEMALES: 60% (trial 1), 100% (trial 2)
• KEY DISEASE CHARACTERISTICS: Ovarian cancer, breast cancer, lung cancer, lymphoma, leukemia, endometrial cancer, cervical cancer, vulvar cancer, colorectal cancer, bladder cancer, thymoma, and myeloma
• OTHER KEY SAMPLE CHARACTERISTICS: Chemotherapy regimens included carboplatin and combinations of cyclophosphamide-anthracycline, cyclophosphamide combinations, irinotecan, topotecan, cisplatin combinations, anthracycline, and gemcitabine-vinorelbine, among others.

• SITE: Multi-site    
• SETTING TYPE: Outpatient  
• LOCATION: Gabrail Cancer Center in Canton, OH; St. Louise Regional Hospital in Gilroy, CA; Department of Gynecologic Oncology at the University of Medical Science in Poznan, Poland

• PHASE OF CARE: Active antitumor treatment
• APPLICATIONS: Elder care and palliative care

Prospective

• Plasma concentrations were measured.  
• Safety was assessed by vital signs, physical examinations, and clinical laboratory tests.  
• Twelve lead electrocardiograms were completed at screenings.  
• Symptoms were assessed with patient diaries.  
• Effectiveness was measured using diaries, information about the use of rescue medications, the number of emetic episodes, the number of retching episodes, and the number of nausea episodes for a seven-day period after the administration of medications.  
• Noncompartmental methods and descriptive statistics were used.

Both studies met the primary objective by defining pharmacokinetics. Adverse events did not appear to be dose-related. Most were mild to moderate and were unrelated to the study drug. Injection site reactions were low and were not associated with dosing, and 17.7% of erythema was reported in the 250 mg arm. No erythema was reported in the 750 mg arm. The plasma concentrations of granisetron were maintained for seven days with a single dose of the drug. Preliminary data demonstrated another option for the prevention of acute and delayed chemotherapy-induced nausea and vomiting. Patients treated with APF530 at 250 or 500 mg obtained complete response 83% of the time in the acute-onset and delayed-onset phases. Complete control was obtained in 76%. Nausea was controlled almost as well as emesis. Nausea reports were mostly mild.

Granisetron exposure was maintained for seven days with a single dose of subcutaneous AFP530. Mild injection site irritation was noted. Nausea was mild, and nausea and vomiting were controlled in the acute and delayed phases.

• Small sample (< 100)
• Risk of bias (no control group)
• Risk of bias (no blinding)
• Risk of bias (no random assignment) 
• Risk of bias (no appropriate attentional control condition)  
• Findings not generalizable

This could be another option for treating chemotherapy-induced nausea and vomiting, but it is possible that this treatment causes unnecessary discomfort when oral and transdermal approaches are available. This is very preliminary data, and the study did not compare this treatment to standard care. Additional research to determine the usefulness of this drug for chemotherapy-induced nausea and vomiting is needed.

PURPOSE: To review absorbable hemostatic agents including pharmacology, clinical efficacy, adverse events and toxicities, dosage and administration, and safety issues

Discusses nine different agents and the different composition of each (e.g., porcine or bovine gelatin, bovine collagen or oxidized cellulose). The two newest agents (approved as U.S. Food and Drug Administration devices, not drugs) are FloSeal^® and CoStasis^®, and these products include bovine thrombin.

• Absorbable agents provide hemostasis via contact activation (which initiates clotting) and/or promotion of platelet aggregation.
• Absorbable hemostatic agents are indicated for use during surgery when bleeding can not be controlled by conventional methods such as pressure.
• Emphasis that topical thrombin is for TOPICAL application only. Misadministration (e.g., IV) can be fatal.

• Literature on these agents is mostly case reports. There are few well-controlled trials that actually have compared agents and/or the efficacy of such agents specifically in instances of bleeding associated with malignancy. The summary of trials and reports is presented. Only 2 of the 14 trials had patients with cancer in the sample (patients with hepatocellular carcinoma undergoing elective hepatic resection and patients with gynecologic cancer undergoing exploratory laparotomy).
• Allergic reactions to thrombin can occur, as can development of antibodies to bovine components of the product.
• No standardized dosing regimens exist; the minimal amount of product needed to achieve hemostasis should be used. Follow product labeling, especially for newer agents such as FloSeal (Baxter) and CoStasis (Cohesion Technologies). A summary table for dosage and administration per each agent is provided.
• Good clinical data are lacking evaluating efficacy of these agents, especially for off-label use (e.g., in nasal packing for uncontrolled epistaxis in patients with thrombocytopenia).