Venous Thromboembolism and Weight Changes in Veteran Patients Using Megestrol Acetate As an Appetite Stimulant

Venous Thromboembolism and Weight Changes in Veteran Patients Using Megestrol Acetate as an Appetite Stimulant

Brandon LaMarr, PharmD, BCPS; and Russell Crawford, BPharm, BCOP

These study investigators sought to answer the question whether patients treated with megestrol acetate at a local VA health care system have a greater incidence of venous thromboembolism than that seen in the general population.

norexia and cachexia are occur in any patient, they are com- to thrombogenic processes associated associated with a variety of monly seen in patients with cancer with the disease and its treatments.9-11 diseases, including cancer, and AIDS. Cachexia in patients with The overall prevalence of VTE in pa- A acquired immunodeficiency cancer can be due to decreased ca- tients taking MA is unclear, as studies syndrome (AIDS), congestive heart loric intake (possibly due to causes have focused on single diseases such failure, chronic obstructive pulmo- such as gastrointestinal tumors or as cancer or AIDS. The MA prescrib- nary disease, liver disease, end-stage chemotherapy-induced nausea and ing information cautions against use renal disease, and endocrine abnor- vomiting), an increased metabolic in patients with a history of venous malities such as hyperthyroidism, state, or the production of proin- thromboembolic disease, but the adrenal insufficiency, and diabetes flammatory mediators such as inter- prevalence and degree of risk are not mellitus.1 Anorexia has also been leukin-1, interleukin-6, and tumor stated.12 For this reason, many of the shown to be associated with the necrosis factor-α. AIDS-related ca- studies evaluating the efficacy of MA aging process.2 Loss of appetite, lean chexia can be caused by a hyper- have excluded patients with a history muscle mass, and adipose tissue is metabolic state, secondary infection, of VTE. associated with physical weakness medications, or gastrointestinal dis- While the risk for VTE in patients as well as decreases in quality of life, turbances.7 with cancer is firmly established, the sense of well-being, and level of func- Many randomized trials have de- risk in patients who use MA is less tionality.3,4 In addition to health detri- termined the safety, efficacy, and ideal clear. Patients with cancer who use ments, the visual and social effects of dose of megestrol acetate (MA) when MA potentially share many of the physical wasting can be emotionally used for appetite stimulation. These same thrombogenic mechanisms, in- distressing to both patients and their studies have often been small and cluding increased levels of clotting family or caregivers.5 have produced inconclusive results factors as well as decreased levels of Increased metabolic needs and de- due to the short life expectancies of anticoagulant proteins. Oberhoff and creased appetite play important roles the study population, as well as the colleagues evaluated the effects of in disease-related wasting.6 Changing many confounding variables that MA on coagulation in patients with social conditions, psychiatric prob- exist among a generally very sick and gynecologic and breast cancers and lems, and use of medications may heterogeneous patient population. found no evidence of thrombogenic also contribute to changes in appe- Uncertainty regarding the optimal potential.13 Contrasting this, a study tite, weight, and nutritional status. dosing and possible thrombogenic ef- by Kropsky and colleagues found a While anorexia and cachexia may fects of MA remain. 6-fold increased incidence of deep The average annual incidence of vein thrombosis (DVT) in elderly pa- Dr. LaMarr is a clinical inpatient pharmacist at the venous thromboembolism (VTE) tients who live in nursing homes and Southern Arizona VA Health Care System in Tucson, in the general population is about take MA.14 Arizona. Mr. Crawford is a clinical pharmacist in he- 0.1%.8 Patients with cancer are at a The purpose of this study was to motology/oncology and PGY2 Oncology Residency Program Director also at the Southern Arizona VA 4 to 7 times greater risk of VTE than evaluate the safety and efficacy of MA Health Care System in Tucson, Arizona. patients without cancer, mainly due for patients in the Southern Arizona

OCTOBER 2012 • FEDERAL PRACTITIONER • 31 Venous Thromboembolism

VA Health Care System (SAVAHCS). It was anticipated that data regarding 350 the safety and efficacy of MA for appetite stimulation generated by this 300 chart review would serve to optimize future use of MA in this veteran pop- 250 ulation. The primary objective of the 200 study was to determine the incidence of VTE in all patients using MA for 150 appetite stimulation. Secondary out- mg/d comes included evaluations of (1) the 100 effects of MA on weight, including rate of response to treatment (defined 50 by the percentage of patients main- 0 taining or gaining weight relative All patients Patients HIV/AIDS Patients with no can- to baseline) and average change in with cancer cer or HIV/AIDS weight; (2) the effects of MA on nu- Figure 1. Average daily MA dose by diagnosis. tritional status, including changes in MA = megestrol acetate. prealbumin and albumin levels while on treatment; (3) effects of MA dose dispensed at SAVAHCS for MA that ically effective, this dosage range was (low, medium, or high) on efficacy also contained the term appetite in categorized as high.12 By default, the (change in weight, rate of response) the directions to identify patients for range of doses > 100 mg and < 400 in all patients and among subgroups whom MA was dispensed and refilled mg was categorized as medium. by diagnosis (cancer, human immu- at least once. This method served to nodeficiency virus [HIV]-positive, ensure that an adequate trial period Statistics and noncancer/non-HIV-positive); of at least 30 days on MA was given. Statistical analyses on the secondary (4) comparison of VTE rates in all Searching retrospectively beginning endpoints assessed the effects of dose patients and among subgroups by di- September 1, 2008, a chart review on efficacy and VTE within diagnosis agnosis in patients treated with MA; was performed on the most recent categories as well as on comparisons and (5) effects of MA dose on VTE 100 patients treated with MA identi- of VTE incidence among diagnoses. incidence among patients with vary- fied in the prescription record search The chi-squared test was used for the ing doses. who met the previously mentioned efficacy analyses and compared the It was hypothesized that patients criteria. The reviews stopped once percentages of patients who main- treated with MA would have an in- 100 patients were identified. Patients tained or gained weight at different cidence of VTE greater than the ob- aged < 20 years or aged > 89 years dose levels with the different diagno- served rate in the general population were excluded. Patients who, on fur- ses. A one-way analysis of variance of 0.1% per year. It was also hypoth- ther review, were not using MA for was performed comparing the aver- esized that patients with cancer who appetite stimulation, were also ex- age changes in weight at different used MA would exhibit a higher rate cluded. dose levels with the different diagno- of VTE than those patients without In order to define the safest and ses as well as the incidences of VTE cancer. most effective doses of MA, the ef- among the different diagnoses. fects of low, medium, and high doses Methods of MA on VTE incidence and weight Results Prior to initiating this study, full in- maintenance were analyzed. Doses A total of 145 charts were reviewed stitutional review board approval < 100 mg are rarely studied, perhaps with 45 charts excluded. Reasons for was requested and obtained. A retro- because of their accepted lack of effi- exclusion included MA use for hot spective chart review was performed cacy.15 For this reason, doses < 100 mg flashes (37 charts), aged > 89 years (4 for patients using MA as an appetite were categorized as low. Since charts), gynecologic issues (3 charts), stimulant. Electronic medication re- product labeling states that doses and lack of any data relating to stud- cords were searched for prescriptions > 400 mg have been shown to be clin- ied endpoints (1 chart). Nine charts

32 • FEDERAL PRACTITIONER • OCTOBER 2012 Venous Thromboembolism

Table 1. Demographic data Table 2. Incidence of VTE Treatment courses (N) 111 Subgroup VTEs (N) Incidence Average age 70 years All patients 3 2.7% Male 96% Patients with cancer 3 7.3% Patients with cancer 37% Patients on 2 5.9% Patients with HIV/ 4% high-dose MA AIDS Patients on 1 1.7% Patients with no 59% low-dose MA cancer or HIV/AIDS MA = megestrol acetate; VTE = venous thromboembolism.

Table 3. Rates of response to treatment All patients Patients with Patients with Patients with no cancer HIV/AIDS cancer or HIV/AIDS Patients on 73% 71% 100% 73% high-dose MA Patients on 59% 50% N/A 64% medium-dose MA Patients on 65% 58% 100% 66% low-dose MA MA = megestrol acetate. documented more than 1 treatment 2). This resulted in a total overall When those patients who lost weight course with MA in the same patient incidence of VTE of 2.7%. All VTEs while on MA were excluded from with 2 charts documenting 3 sepa- occurred in patients with cancer lead- analysis, these effects became even rate MA courses. Each course was ing to an incidence of VTE of 7.3% more dramatic (Table 5). analyzed independently. In total, 111 in this subgroup. VTE occurred in 2 Albumin and prealbumin levels treatment courses with MA for ap- patients on high-dose MA and 1 pa- were not frequently measured in this petite stimulation were included in tient on low-dose MA resulting in an group. More than 1 albumin level the analysis. incidence of 5.9% for the high-dose was recorded in 37 patients while on Demographic data for the study group and 1.7% for the low-dose MA. Prealbumin levels were recorded population is presented in Table 1. group. No VTE occurred in patients more than 1 time during MA treat- The distribution of patients by MA with a prior history of VTE. ment in 16 patients. There was insuf- dosing category was as follows: 34 Ninety-nine patients had > 1 weight ficient data to perform analyses on patients (31%) received high doses, recorded while on MA. Sixty- these endpoints. 19 patients (17%) received medium six (67%) of these patients re- doses, and 58 patients (52%) responded to treatment. The average Discussion ceived low doses. change in weight was an increase of The rate of VTE observed in the SA- In all patients, the average daily 1.3 kg (range: - 14.0 kg to + 18.0 kg). VAHCS population using MA for ap- dose of MA was 216 mg/d. MA doses When patients who continued to lose petite stimulation was higher than were in the range of 20 mg/d to 800 weight while on MA were excluded, that of the population in general. mg/d. Patients with cancer received the average change in weight was an Since all VTE events occurred in pa- the highest average MA daily doses; increase of 4.0 kg. tients with cancer, it is not possible those without either cancer or HIV/ While not statistically significant, to compare the incidence of VTE AIDS received the lowest average MA high-MA doses resulted in higher with different diagnoses. The ob- doses (Figure 1). rates of response and more positive served rate of VTE in patients with Out of 111 total treatment courses weight changes than did lower doses cancer using MA was also high when with MA, 3 VTEs occurred (Table in all subgroups (Tables 3 and 4). compared with literature estimates.

OCTOBER 2012 • FEDERAL PRACTITIONER • 33 Venous Thromboembolism

Table 4. Change in weight (kg) All patientsa Patients with Patients with Patients with no cancer cancerb HIV/AIDS or HIV/AIDSc Patients on high-dose MA 2.1 (n = 30) 0.76 (n = 17) 4.0 (n = 2) 3.9 (n = 11) (≥ 400 mg/d) Patients on medium-dose 1.5 (n = 17) -0.5 (n = 6) N/A 2.6 (n = 11) MA (100-399 mg/d) Patients on low-dose MA 0.7 (n = 52) -1.66 (n = 12) 0.5 (n = 2) 1.5 (n = 38) (< 100 mg/d) aP = .48; bP = .60; cP = .35. MA = megestrol acetate.

Table 5. Change in weight (kg) in treatment responders All patientsa Patients with Patients with Patients with no cancer cancer b HIV/AIDS or HIV/AIDSc Patients on high-dose MA 5.4 5.1 4 6.1 Patients on medium-dose MA 5.0 3.0 N/A 5.9 Patients on low-dose MA 2.9 1.9 0.5 3.4 aP = .02; bP = .05; cP = .12. MA = megestrol acetate.

While the design of this study did enrolled 271 patients and assessed 1 patient experienced DVT. not allow for differentiation between weight gain, body composition, ca- In 1999, Parnes and colleagues the thrombogenic effects of MA and loric intake, sense of well-being, ap- performed a randomized dose-find- those of cancer and its treatments, petite, and adverse effects (AEs) for ing study for MA in cancer patients the very high rate of VTE in these pa- placebo and MA doses of 100 mg, with cachexia.17 A total of 381 patients suggests that treatment with 400 mg, and 800 mg, respectively. tients receiving MA doses of 125 mg, MA may be contributory. MA use increased patient weight in 625 mg, or 1,250 mg correspond- The incidence of VTE in patients a dose-dependent manner (P < .001). ing to low-, medium-, and high-dose taking high-dose MA was nearly 3 Patients in the 800-mg group had groups, respectively, were followed. times higher than that of patients significantly (P < .001) greater final The study showed similar effects be- taking low doses. This may be due weight gain compared with base- tween the medium- and high-dose to greater thrombogenic potential in line than did patients in the placebo groups throughout the study. Pa- patients taking high-dose MA. An al- group. Lean body mass also increased tients in the medium- and high-dose ternative explanation for this result is in a dose-dependent manner. How- groups were not significantly more that it echoes the higher VTE rates in ever, differences in lean body mass likely to have a higher maximum cancer patients, since patients with were only statistically significant percentage weight gain, but were cancer represented the group using (P < .001) between the placebo and statistically less likely to lose weight. the highest doses of MA. 800-mg groups. A statistically signifi- The medium- and high-dose groups The results of this study suggest cant (P < .001) trend was observed were statistically more likely to cause that higher doses of MA were more correlating increased patient well-be- edema when compared with the low- effective. This observation is sup- ing scores with increased MA doses. dose group. VTE occurred in 9 pa- ported by previous studies. Von Patients treated with 800 mg of MA tients with no differences between Roenn and colleagues performed a also had significantly (P = .001) the 3 groups. double-blind, randomized, placebo- greater caloric intake than those Maltoni and colleagues performed controlled trial assessing the safety treated with placebo. No statistically a meta-analysis evaluating the effi- and efficacy of MA in patients with significant differences were found cacy of high-dose progestins (MA AIDS and cachexia.16 The study among the 4 groups in terms of AEs; and medroxyprogesterone acetate)

34 • FEDERAL PRACTITIONER • OCTOBER 2012 Venous Thromboembolism

in cancer patients with anorexia- to analyze different disease popula- pharmacologic therapy to patients. cachexia.18 Fifteen studies qualified tions independently, overall results for analysis, which included 2,102 must be understood in the proper References 1.Thomas DR. Anorexia: Aetiology, epidemiology, patients. Analysis of weight gain context. Additionally, the design of and management in older people. Drugs Aging. included 11 of the 15 studies and this study did not allow for differen- 2009;26(7):557-570. 2. Serra Prat M, Fernández X, Ribó L, Palomera E, Pap- showed that patients on a high-dose tiation between the thrombogenic ef- iol M, Serra P. Loss of appetite in elderly people in progestin treatment were more than fects of cancer and its treatments and the community and its relationship with functional capacity. Med Clin (Barc). 2008;130(14):531-533. twice as likely to gain weight as were any potential thrombogenic effects of 3. Keys A, Brozek J, Henschel A, Mickelesen O, Taylor those patients on placebo (OR = 2.66; MA. The study also did not allow for HL. The Biology of Human Starvation (Volumes 1 and 2). Minneapolis, MN: University of Minnesota Press; 95% CI, 1.80%-3.92%). differentiation between the type of 1950. When analysis was performed weight that was gained (ie, lean body 4.Andreyev HJ, Norman AR, Oates J, Cunning- ham D. Why do patients with weight loss have a on just the patients who responded mass, adipose tissue, or edema). worse outcome when undergoing chemotherapy to MA, significant (P = .02) positive In conclusion, higher doses of for gastrointestinal malignancies? Eur J Cancer. 1998;34(4):503-509. trends of weight gain were observed. MA seem to generate more weight 5. Poole K, Froggatt K. Loss of weight and loss of ap- These results support previously gain and higher rates of response to petite in advanced cancer: A problem for the patient, the carer, or the health professional? Palliat Med. published studies that showed an in- treatment. The VTE rates in patients 2002;16(6):499-506. creasing dose-response relationship with cancer using MA as an appetite 6. Argilés JM, Alvarez B, López-Soriano FJ. The metabolic basis of cancer cachexia. Med Res Rev. in patients treated with MA. How- stimulant were considerable. While 1997;17(5):477-498. ever, there was also a large portion MA seems to be effective for appetite 7. Grunfeld C. What causes wasting in AIDS? N Engl J Med. 1995;333(2):123-124. of the study sample that did not re- stimulation, a thorough risk assess- 8. Heit JA. The epidemiology of venous thromboembo- spond to MA for unknown reasons. ment for VTE should be performed lism in the community. Arterioscler Thromb Vasc Biol. 2008;28(3):370-372. Identification of factors that lead in patients with cancer prior to initi- 9. Blom JW, Vanderschoot JP, Oostindiër MJ, Osanto to response to treatment with me- ating treatment with MA for appetite S, van der Meer FJ, Rosendaal FR. Incidence of venous thrombosis in a large cohort of 66,329 cancer dium- or high-dose MA for appetite stimulation. patients: Results of a record linkage study. J Thromb stimulation is a potential area of fur- Haemost. 2006;4(3):529-535. 10. Blom JW, Doggen CJ, Osanto S, Rosendaal FR. ther research. Until these factors are Acknowledgment Malignancies, prothrombotic mutations, and the risk known, it may be prudent to try al- This manuscript was prepared and re- of venous thrombosis. JAMA. 2005;293(6):715-722. 11. Heit JA, Silverstein MD, Mohr DN, Petterson TM, ternative appetite stimulants in pa- search was conducted with resources O’Fallon WM, Melton LJ 3rd. Risk factors for deep tients not exhibiting a response to and the use of facilities at the Southern vein thrombosis and pulmonary embolism: A population-based case-control study. Arch Intern Med. treatment with MA after receiving an Arizona VA Health Care System in Tuc- 2000;160(6):809-815. adequate trial. son, Arizona. 12. Megace [package insert]. Princeton, NJ: Bristol-My- ers Squibb Company; 2012. This study was limited in that 13. Oberhoff C, Hoffmann O, Winkler UH, Schindler its retrospective nature did not per- Author disclosures AE. Hemostatic effects of high-dose megestrol acetate therapy in patients with advanced gynecologi- mit conclusions to be attributed to The authors report no actual or poten- cal cancer. Gynecol Endocrinol. 2001;15(5):341-348. direct causes and effects. Addition- tial conflicts of interest with regard to 14. Kropsky B, Shi Y, Cherniack EP. Incidence of deep- venous thrombosis in nursing home residents using ally, in order to allow an adequate this article. megestrol acetate. J Am Med Dir Assoc. 2003;4(5): trial of MA, patient selection may 255-256. 15. Berenstein G, Ortiz Z. Megestrol acetate for treatment have created a preference for pa- Disclaimer of anorexia-cachexia syndrome. Cochrane Database tients in whom MA was more effec- The opinions expressed herein are those Syst Rev. 2005;(2):CD004310. 16. Von Roenn JH, Armstrong D, Kotler DP, et al. Meges- tive thereby introducing a sampling of the authors and do not necessarily trol acetate in patients with AIDS-related cachexia. bias. The study population consisted reflect those of Federal Practitioner, Ann Intern Med. 1994;121(6):393-399. 17. Parnes HL, Conaway M, Aisner J, et al. Megestrol mostly of elderly males, which lim- Quadrant HealthCom Inc., the U.S. acetate for the treatment of cachexia in patients with ited the generalizability of the results. Government, or any of its agencies. advanced lung or colorectal cancers. Cancer Ther. 1999;2:75-82. The study also attempted to compare This article may discuss unlabeled or 18. Maltoni M, Nanni O, Scarpi E, Rossi D, Serra P, Ama- safety and efficacy of MA between pa- investigational use of certain drugs. dori D. High-dose progestins for the treatment of cancer anorexia-cachexia syndrome: A systematic tient populations with different dis- Please review complete prescribing in- review of randomised clinical trials. Ann Oncol. ease processes (eg, cancer vs AIDS), formation for specific drugs or drug 2001;12(3):289-300. which created a potentially hetero- combinations—including indications, geneous population in terms of dis- contraindications, warnings, and ad- ease. While the study was designed verse effects—before administering

OCTOBER 2012 • FEDERAL PRACTITIONER • 35