Journal of Oncology Pharmacy Practice

Purpose. The aim of this prospective observational study was to investigate the patterns of treatment adherence to orally administered chemotherapy of patients being treated for cancer.

Methods. Patients were asked to participate in the survey during their visits to the study centers’ pharmacists or doctors to obtain their oral medication from April 2008 until May 2009. The data were collected using a self-reported anonymous 7-page long questionnaire, which contained questions about their demographic profile, disease and treatment characteristics, and side-effects and adherence information, both intentional and nonintentional.

Results. 99 Patients completed the questionnaire. Missing values ranged from 1% to 8%. Unintended nonadherence to therapy was reported by 19 patients. The most important factor correlating with unintended nonadherence was the patient’s belief regarding treatment effectiveness since only 16.7% of the patients believing that their treatment is effective reported nonadherence as opposed to 62.5% for those that did not believe that treatment is effective (p = 0.03). Intentional nonadherence was reported by 14 patients The most important factor correlating to intentional nonadherence was time since disease diagnosis, as nonadherence was reported by 33.3% of the patients having the disease less than 6 months, compared to 16.7% for those between 6 and 24 months and 8.3% for those between 2 and 5 years (p = 0.01).

Conclusion. Greek patients seem to have similar nonadherence pattern as in other countries. Confidence in treatment efficacy appeared as a significant adherence determinant.

Purpose and relevance. Chemotherapy-induced thrombocytopenia (CIT) can be a significant problem in patients with cancer, leading to numerous clinical complications. Understanding the types of patients at risk for these complications is essential to improve monitoring, counseling, and provide future targeted prophylaxis measures. Previous studies have limited prospective utility since they do not examine risk factors associated with complications from multi-agent regimens. This evaluation aims to identify the incidence and risk factors associated with clinical complications of CIT in patients receiving common chemotherapy regimens.

Methods. Retrospective evaluation of adult patients receiving first or second line regimens for the most common solid tumors associated with high rates (≥5%) of laboratory diagnosed thrombocytopenia. Patients were examined for clinically significant CIT (defined as platelet count <75,000 cells/µL as well as the presence of one of the following: bleeding, dose reduction/delay, platelet transfusion, or therapy cessation) and associated risk factors.

Results. About 254 patients receiving a total of 278 regimens were evaluated. The incidence of clinically significant CIT = 10.1%; complications were most common in patients receiving cisplatin/gemcitabine for bladder cancer (57%), or carboplatin/gemcitabine (29%) or cisplatin/etoposide (18%) for lung cancer. Bladder cancer (OR = 13.7 (2.89–64.7); p = 0.001) and concurrent or recent infection (OR = 3.8 (1.45–10.1); p = 0.007) was found to increase the risk of clinical complications while smoking was found to have a protective effect (OR = 0.17 (0.04–0.71)).

Conclusions. The incidence of clinically significant CIT is most commonly seen in patients using cisplatin/gemcitabine for bladder cancer, or carboplatin/gemcitabine or cisplatin/etoposide for lung cancer. Further evaluation of these patients is warranted.

In health care facilities, dermal contact and inhalation are considered to be the main routes of exposure to cytotoxic antineoplastic drugs (ADs). Hand-to-mouth contamination or accidental needle sticks as well as events due to inadequate disposal may also contribute to exposure. In order to measure the extent of contamination, biological and environmental monitoring are essential tools for routine testing. Moreover, reliable sampling and analytical procedures are required. During the last decade, several methods have been developed and validated. The appropriate analytical techniques were used to quantify even very low levels of some of the more commonly used ADs, such as cyclophosphamide, 5-fluoruracil, taxol, anthracyclines, and platinum-compounds. The main objective of this study is to assess the adherence to existing standards of practice through an effective monitoring program, including environmental and biological measurements. In seven hospitals located in Northern-Central Italy, periodic surveys were scheduled to verify continuing compliance with guidelines over a 5-year period. All biological samples were found to be below detection limits and a progressive, significant decrease in workplace contamination was observed. Our results confirm that a cost-effective monitoring regime, including fast and simple sample pre-treatment procedures, simultaneous determination of the analytes and their metabolites, validated procedures including uncertainty evaluation, and periodic surveys, is the adequate approach for the collection of reliable exposure data and hence for effective intervention.

Purpose. Hypertension is a common adverse effect of vascular endothelial growth factor (VEGF) signaling inhibitors, such as bevacizumab, with an incidence upwards of 35%. The management of bevacizumab-induced hypertension is important in order to avoid dose interruption/discontinuation and/or end organ damage. The efficacy of antihypertensive medications for this cause of hypertension has not been demonstrated. This study seeks to determine if antihypertensives are effective in treating anti-VEGF-induced hypertension from bevacizumab and determine which classes of antihypertensive agents are effective.

Methods. A retrospective review of all patients who received bevacizumab between January 2007 and September 2009 at two medical centers was conducted. Patients were included if they experienced new onset or exacerbation of preexisting hypertension, during bevacizumab treatment. Efficacy of antihypertensives was determined by recording a 28-day change in systolic blood pressure from the initiation or dose increase of individual antihypertensive medications. Secondary endpoints included an efficacy analysis of antihypertensive classes.

Results. Five-hundred thirteen patients were identified as receiving bevacizumab during the indicated time period. Fifty-seven patients met the full inclusion/exclusion criteria for analysis. The average systolic blood pressure declined by 23 mm Hg with 4 weeks of treatment (p < 0.0001). Each class had a statistically significant decline in systolic blood pressure of 15.5–57 mm Hg with the exception of diuretics and a group of miscellaneous antihypertensives.

Conclusions. This is the first data that demonstrates individual classes of antihypertensives are effective in bevacizumab-induced hypertension. Most antihypertensives were effective in reducing blood pressure, with the exception of diuretics and miscellaneous antihypertensives, which may be due to a limited sample size.

Background. NSC-631570 (Ukrain) is a semisynthetic derivative of the Chelidonium majus alcaloids and the alkylans thiotepa. It exerts a selective cytotoxic effect on tumor cells in vitro and in vivo and shows the ability to modulate immunocyte functions.

Purpose. The aim of our work was to carry out a comparative investigation of the effects of NSC-631570 alone or in combination with pathogen-associated molecules (PAM) on the growth of low- and high-metastasizing melanoma B16 in mice.

Methods. NSC-631570 was administered intravenously and PAM intramuscularly to tumor-bearing mice seven times every third day, starting from the second day after the transplantation of tumor cells. The effect of monotherapy and combined therapy on tumor growth was evaluated by the indices of tumor growth inhibition in experimental animals. Cell cycle distribution of cancer cells was determined by flow cytometry. TAP1 and TAP2 expression was evaluated by RT-PCR. The metabolic activity of phagocytes was determined by NBT-test, phagocytosis was tested by flow cytometry, and arginase activity was estimated by colorimetric determination of urea.

Results. Combined therapy and monotherapy with NSC-631570 resulted in significant inhibition of tumor growth in melanoma-bearing mice. Monotherapy with Ukrain was more effective in mice with high-metastasizing tumors. The therapeutic efficacy of NSC-631570 used in combination with PAM was more expressed in mice with low-metastasizing melanoma.

Conclusion. The effectiveness of monotherapy and combined therapy with NSC-631570 in the treatment of melanoma B16 depends on the biological properties of the tumor and the immune state of the organism.

Background. Erythropoiesis-stimulating agents (ESAs) provide symptom relief and decrease blood transfusion support among patients with chemotherapy-induced anemia. However, due to increased cardiovascular events associated with off-labeled usage of ESAs, the FDA incorporated black box warnings in 2007 to include the following key points: (a) ESAs should be used only to treat anemia due to concomitant chemotherapy of a noncurative intent and (b) target hemogloblin level should not exceed 12 g/dL. Thus, this study was designed to compare the prescribing of epoetin alfa at National Cancer Centre Singapore before and after FDA black box updates. The secondary objective of this study was to evaluate the appropriateness of efficacy and toxicity monitoring of epoetin alfa.

Methods. This was a retrospective, single-centered, drug utilization review. Patients who received at least one dose of epoetin alfa were included in this study. Utilization of epoetin alfa was segregated into two time periods: January 1, 2005 to October 15, 2007 (S1, Pre-safety advisories changes) and October 16, 2007 to December 10, 2009 (S2, Post-safety advisories changes).

Results. A total of 171 patients were prescribed epoetin alfa at NCCS during the two time periods. However, only 139 patients were eligible for analysis, with 91 and 48 patients in S1 and S2 respectively. After safety advisory changes, there were more (18.2%) metastatic patients and fewer (19.1%) patients with cardiovascular co-morbidities who were prescribed epoetin alfa, the mean hemogloblin level when epoetin alfa was initiated was lowered by 0.46 g/dL, more (43%) dose adjustments were made for ‘excessive’ responders and more (40.7%) patients had fewer blood transfusions after epoetin alfa therapy (p < 0.05). However, blood pressure control, iron studies, and supplementation did not improve (p > 0.05).

Conclusion. This study suggested that oncologists have generally adopted the new ESA safety warnings and adjusted prescribing habits.

Background. The Centers for Medicare and Medicaid Services (CMS) issued a national coverage determination (NCD) in July 2007, which imposed restrictions on the reimbursement of ESAs for Medicare and Medicaid beneficiaries. Since a majority of our patients are Medicare or Medicaid beneficiaries, we changed our clinical practice regarding the use of erythropoiesis stimulating agents (ESAs) to coincide with the NCD’s reimbursement restriction.

Objective. To evaluate the number of transfusions in patients diagnosed with chemotherapy-induced anemia (CIA) receiving ESAs before and after the clinical practice was changed at the University of Illinois Medical Center (UIMC).

Methods. The medical records of all adult patients diagnosed with a nonmyeloid malignancy and CIA who received an ESA between July 2006 and June 2008 at the UIMC were evaluated. The patients were divided into two groups: patients in receipt of ESAs BEFORE (group 1) and AFTER (group 2). The number of transfusions, the response rates to chemotherapy and ESAs therapy, and overall survival were compared.

Results. Medical records for 110 patients were reviewed. More transfusions were given to patients AFTER we implemented the change in clinical practice (BEFORE 18 transfusions vs. AFTER 52 transfusions, p = 0.004). More patients responded to ESA therapy AFTER we implemented the change (67% vs. 83%, p = NS). The treatment response to chemotherapy and overall survival were similar between the two groups.

Conclusion. The primary goal of reducing the number of transfusions in patients with CIA by administering ESAs cannot be met when clinical practice coincides with the NCD.

Background. The World Health Organization classifies myelodysplastic syndromes (MDS) as a group of hematologic malignancies where at least one myeloid lineage contains ≥10% cells with myelodysplasia. Therapy-related MDS (t-MDS) may result from treatment with myelosuppressive chemotherapy and/or radiation. Myeloid growth factors, such as pegfilgrastim, have also been investigated as a possible risk factor for development of t-MDS. National guidelines and myeloid growth factor prescribing information recommend administering myeloid growth factors after a minimum of 24 h following myelosuppressive chemotherapy. Some evidence suggests that administering myeloid growth factors and myelosuppressive chemotherapy concurrently may increase the risk of developing t-MDS.

Objective. To evaluate the incidence of t-MDS in patients who received pegfilgrastim and chemotherapy on the same day compared to patients receiving pegfilgrastim at least 1 day following completion of their chemotherapy.

Methods. A retrospective chart review was performed to identify all patients who received myelosuppressive chemotherapy and pegfilgrastim on the same day between January 2003 and December 2007. Any initial diagnosis of t-MDS was clinically confirmed.

Results. A total of 227 patients received myelosuppressive chemotherapy and pegfilgrastim on the same day. Median time to follow-up was 3.47 years (IQR 1.15). Two patients had a confirmed diagnosis of t-MDS. The incidence of t-MDS in patients receiving concurrent myelosuppressive chemotherapy and pegfilgrastim in our study population is 0.88%.

Conclusion. Administering pegfilgrastim at least 1 day after myelosuppressive chemotherapy may help reduce the risk of t-MDS, but more studies with longer follow-up are needed to confirm this finding.

Summary. A serious adverse effect that can limit the utility of ifosfamide is neurotoxicity, known as ifosfamide-induced encephalopathy (IIE). Information regarding the usefulness of prophylactic administration of methylene blue, thiamine, and albumin to reduce the incidence of IIE is scarce. We present the results of a retrospective chart review evaluating the incidence of IIE in adult patients who received ifosfamide to treat sarcoma at the University of Washington within a 2-year period. Patients who received methylene blue, thiamine, and/or albumin were compared to patients who did not receive these prophylactic agents concurrently with ifosfamide. The primary objective was to evaluate if prophylaxis is associated with a reduced incidence of IIE. Identifying risk factors associated with IIE was a secondary objective. The cost of using prophylaxis with ifosfamide administration was reviewed.

Results. A total of 166 cycles were included. For the primary endpoint, more cycles in the prophylaxis group had patients with symptoms of IIE (21.1% in the prophylaxis group vs. 8.4% in the nonprophylaxis group); p = 0.026. The average number of risk factors per cycle was the same in each group; however, the type of risk factors differed.

Conclusions. Prophylaxis was not associated with a reduced incidence of IIE, and it does not appear to delay time to onset of symptoms at our institution. Results of this study provide no support for using methylene blue, thiamine, and/or albumin as routine prophylaxis when administering ifosfamide to adults with sarcoma.

Study objective. To determine the tolerability of rituximab, specifically cytokine release syndrome/acute infusion reactions (CRS), when it is administered before or after cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) chemotherapy in patients with non-Hodgkin’s lymphoma (NHL).

Methods. This study is a retrospective analysis of patients identified through pharmacy chemotherapy records. Inclusion criteria were diagnosis of NHL, first cycle of rituximab with CHOP or modified CHOP (mCHOP), treated between 1/1/04 and 6/30/09, age 18 years and greater, and inpatient status. Patients were excluded if their records/information were unavailable. Patients were divided into two groups based on practices observed at our institution: rituximab followed by CHOP (R-CHOP) or CHOP followed by rituximab (CHOP-R). Patient records were reviewed to determine demographic data, CRS, vital signs, evidence of chills/rigors, use of rescue medications, and rituximab infusion rates.

Results. One-hundred thirteen patients meeting the inclusion criteria were divided into two groups: R-CHOP (n = 29) and CHOP-R (n = 84). R-CHOP patients experienced numerically more CRS (65.5% vs. 42.9%, p = 0.0517) and significantly more chills/rigors (p = 0.0376). Maximum and minimum oxygen (O₂) saturations were significantly lower in the R-CHOP group (p = 0.0444 and 0.0165, respectively). Maximum temperature was significantly higher in the R-CHOP group (p = 0.0047). There was no difference between groups in use of rescue medications (p = 1). R-CHOP patients required significantly more rate reductions (p = 0.0431) than CHOP-R patients, although there was no difference in final tolerated rate between groups.

Conclusions. Patients with NHL who receive rituximab after CHOP experience significantly fewer chills/rigors, higher oxygen saturations, lower maximum temperatures, and fewer rate reductions than patients who receive rituximab before CHOP.

Background. Assessing the role of the pharmacist in an outpatient oncology clinic requires quantification of parameters that are important to the patient and the healthcare system. Patient evaluations not only serve as benchmarks for a process; but may also predict behavior such as utilization of care, continuity with the provider, and compliance.

Purpose. To identify the role of the patient–pharmacist relationship and enhance patient satisfaction with care, a survey tool was developed and utilized.

Methods. A 20-item, 2-page survey tool was developed and administered to oncology patients actively receiving treatment at the Cancer Therapy & Research Center (CTRC) in San Antonio, Texas. Following approval from the Investigational Review Board, survey distribution began on December 1, 2009 and continued through February 2010. Basic social and demographic data were collected. Satisfaction with pharmacy service was assessed with questions on a Likert ranking scale detailing time spent with pharmacists, knowledge of medication therapy, willingness to pay for clinical pharmacy services, etc. The primary outcome was the impact of the pharmacist–patient relationship assessed by examining the interaction between time spent with pharmacist, understanding of medications, and desire for future pharmacy counseling services.

Results. A total of 77 surveys were administered and collected. Of the patients surveyed, 86% stated that it is important for patients to discuss their treatment with a pharmacist and 76% requested pharmacy follow-up at future visits.

Conclusion. This cross-sectional survey reveals that patients are interested in visiting with a pharmacist regularly during chemotherapy treatment and patients may be willing to pay for pharmacy counseling services.

Introduction. Intravesical delivery of chemotherapy agents is used very commonly for the treatment of superficial bladder cancer. We recently completed a phase II study of intravesical gemcitabine in which an alkaline adjusted gemcitabine preparation was used initially, based on very early phase I studies. However, crystallization was noted in some of the pre-prepared syringes, which prompted us to investigate the conditions under which gemcitabine crystallized.

Methods. Gemcitabine was prepared in syringes in triplicate and conditions were varied with respect to pH, temperature, and duration. Samples were observed for up to 48 h for the development of crystallization. High-performance liquid chromatography analysis of gemcitabine concentrations was undertaken for all samples.

Results. Crystallization of gemcitabine was favored under conditions of bicarbonate treatment and lowering of temperature. However, the process was reversible, as demonstrated by recovery of gemcitabine concentrations in samples brought back to room temperature. Crystallization resulted in reduction of gemcitabine concentrations in the pre-prepared syringes.

Conclusions. Gemcitabine solutions may be associated with crystallization if the native pH is increased with the addition of sodium bicarbonate, and samples are stored in a cold environment.

Background. Bortezomib is a modified dipeptidyl boronic acid analogue used to treat multiple myeloma, mantle cell lymphoma, and, more recently, renal transplantation graft rejection. As per manufacturer recommendations, bortezomib is to be administered within 8 h of preparation or may be stored for up to 8 h in the vial or a syringe following reconstitution. Preserving unused reconstituted bortezomib beyond these 8 h may allow for cost savings. This study aims to examine the stability of unused reconstituted bortezomib when stored at 4°C for up to 15 days.

Methods. Using an LC-MS/MS assay, the concentration of reconstituted bortezomib was measured at predetermined time points following storage at 4°C in the manufacturer vial. Percent bortezomib remaining at a time point was calculated versus initial bortezomib concentration.

Results. The concentrations of bortezomib were found to be 51.93 ng/mL ± 4.60 after 1 day of storage, 57.40 ng/mL ± 4.77 after 8 days of storage, and 49.43 ng/mL ± 2.85 after 15 days of storage. The percent of bortezomib remaining was 110.53% and 95.19% after 8 days and 15 days, respectively.

Conclusion. Unused reconstituted bortezomib is stable for up to 15 days stored at 4°C in the original manufacturer vial. Such use of bortezomib may improve cost efficiency by reducing bortezomib waste.

Purpose. To determine the extent and severity of drug–drug interactions between anti-emetics and antipsychotics or antidepressants.

Summary. Oncology patients are often required to deal with chemotherapy-induced nausea and vomiting at the same time as psychosocial distress. A review of primary literature, as well as several drug interaction databases, was performed with anti-emetics used in The NCCN® 1.2010 Anti-emesis Guidelines (n = 11) and all currently US-marketed antidepressants or antipsychotics (n = 40).¹ The results from these databases were compiled into a single easy-to-use chart that portrays the severity of the interaction and brief recommendation.^2,3,4 In total, 197 drug–drug interactions out of a total of 440 possible combinations (44.8%) were discovered during the analysis.

Conclusions. Although most anti-emetics had several serious interactions with antidepressants or antipsychotics, palonosetron, and granisetron were found to have no significant interactions. The results can be used to avoid or limit drug interactions in the prescribing of new medications for the oncology patient.^1,2,3,4

Objective. To determine the necessary components and formatting of an intravenous chemotherapy label to maximize safe delivery and minimize errors.

Date sources. The MEDLINE and EMBASE databases (up to April 2009) were searched for relevant evidence. Reference lists from retained studies were then searched for additional trials. An environmental scan was also conducted to locate other published and unpublished sources of information.

Study selection. Relevant articles were selected and reviewed by one methodologist. Articles were selected for inclusion if they were published English language reports of Phases II or III randomized controlled trials, other comparative studies, single-arm studies, practice guidelines, or systematic reviews with or without meta-analyses, which related to the study question. MEDLINE and EMBASE searches yielded 685 potential studies of which 17 met the inclusion criteria. The environmental scan located one guideline. Three additional relevant studies were identified during the external review process. In total, 21 documents met the inclusion criteria.

Data extraction. Data were extracted by one methodologist. Quality of systematic reviews was assessed using the AMSTAR tool. All other studies were evaluated based on study characteristics applicable to the particular study design.

Data synthesis. The evidence collected and the consensus of expert opinion of Cancer Care Ontario’s Chemotherapy Labeling Panel form the basis of a series of recommendations for the generation of intravenous chemotherapy labels including formatting, required information, and order of information. These guidelines inform the efficient, effective, and safe administration of intravenous chemotherapy. Illustrative examples are provided.

Background. The John Marsh Cancer Center is an ambulatory oncology clinic located in Hagerstown, MD. In 2004, a clinical oncology pharmacist was hired to help manage therapies and control cost. The objective was to evaluate clinical interventions, consultations and cost savings by the clinical oncology pharmacist.

Methods. This was a retrospective descriptive analysis of clinical interventions by the clinical oncology pharmacist from September 4, 2004 to October 27, 2006. Interventions were categorized as either drug-related or consultative. Drug-related interventions included medication reconciliation, dosing, and adverse effect management and prevention. Consultations incorporated drug information questions, patient visits, and patient education sessions. Information was extracted from an online documentation program linked to medical charts.

Results. A total of 583 clinical interventions were documented among 199 patients. Average time spent per intervention was 10 minutes. Drug-related and consultative interventions accounted for 35% and 65%, respectively. Included among the drug-related interventions were adverse events (131), medication reconciliation (52) and dosing (22). Consultation services consisted of patient education (143), patient visits (124) and drug information (25). The on-site pharmacist saved $210,000 by admixing chemotherapy. Patient and colleague surveys evaluated pharmacist services with positive ratings of 95% and 98%, respectively.

Conclusion. Analysis of clinical interventions, cost-savings, and feedback from patients and colleagues confirmed beneficial services provided by a clinical pharmacist in this outpatient oncology center.

We describe a patient who experienced immediate onset severe thrombocytopenia as part of a hypersensitivity reaction to oxaliplatin, following substantial relatively uncomplicated previous exposure to the drug. We review other reports of acute hematological toxicity from oxaliplatin. We draw attention to the important need for vigilance when patients develop unusual acute symptoms during or shortly after oxaliplatin infusion.

Treating two active malignancies concurrently can be exceedingly difficult. Complications can occur from the different treatment regimens, especially if they share common targets, and the progressing diseases can make managing treatment side-effects even more challenging. We report a case of a patient with coexisting CML and mRCC who progressed on multiple lines of mRCC therapy while experiencing significant dose limiting side-effects.

Pregnancy occurring after multimodal therapy in a woman with breast cancer with a 1-year follow-up period is a relatively rare condition and has been defined as pregnancy-associated breast cancer. A patient can become pregnant after chemotherapy for breast cancer while she is on tamoxifen. However, the effects of tamoxifen on fetus and on the course of the pregnancy are still unknown. Here, we present a 39-year-old woman treated with chemotherapy and radiotherapy for bilateral breast cancer, and who became pregnant while taking tamoxifen.

Mucinous breast carcinoma (MBC) is a rare histological type of breast cancer and rarely associated with advanced disease. We report a case that had MBC with an isolated adrenal metastasis which was removed by laparoscopic adrenelectomy. This case is unique due to the unexpected metastasis of pure mucinous carcinoma developed after 4 years of hormone therapy.

Context. Erlotinib, an epidermal growth factor receptor tyrosine kinase inhibitor, exhibits a drug interaction with proton pump inhibitors (PPIs) and histamine-2 receptor antagonists (H2RAs). The manufacturer recommends avoidance of the combination however, the extent of the drug interaction is not clearly understood.

Evidence acquisition. A literature search was performed and the pharmacokinetics and pharmacology of acid-reducing agents were reviewed.

Results. Acid-reducing agents reduce the solubility, and subsequent absorption, of erlotinib by raising gastric pH. Our literature search was unable to identify any published studies or case reports that address this issue. Until more information is available, the clinical relevance of this interaction, and whether it actually leads to failure of therapy, is unknown. PPIs would all be expected to exhibit a similar effect on erlotinib. Of the H2RAs, co-administration appears to have a greater impact than administering them separately. Ranitidine, famotidine, and nizatidine should likely have similar effects on erlotinib absorption. Cimetidine has a shorter duration of action, but should be used with caution because of its effects on cytochrome P450 3A4, a pathway also utilized by erlotinib. Antacids are not expected to have a significant effect.

Conclusions. The clinical relevance of this drug interaction is unknown. Until more information is available, decision making regarding this interaction should be on a patient-by-patient basis. The indication for acid-reducing therapy should be reevaluated and stopping therapy or changing therapy can be considered. However, there may be occasions where any benefit of such actions will be exceeded by its risks.