Mcpletterstoedapril07

Use of Bisphosphonates in Multiple Myeloma:
For patients achieving less than a very good partial response IMWG Response to Mayo Clinic Consensus Statement
and/or those with ongoing active bone disease, further BP use isrecommended. After 2 years, patients without active bone dis- To the Editor: The International Myeloma Working Group ease can discontinue BP use. Because no data indicating the (IMWG), which comprises 85 investigators specializing in the value of a reduced-frequency (and/or reduced-dose) schedule area of multiple myeloma, has reviewed and considered the are available, stopping BP treatment at 2 years is recommended, recent Mayo Clinic consensus statement for the use of rather than the Mayo recommendation to decrease BP use to a bisphosphonates (BPs) in multiple myeloma.1 Although the schedule of periodic treatment such as every 3 months.
IMWG is in general agreement with the Mayo consensus For patients with continued active bone disease after 2 statement, several important issues have been raised and are years of BP therapy, further BP use is recommended at the discussed subsequently. These comments are in response to discretion of the primary physician. Pamidronate or clodro- the recommendations of the Mayo group, which is normally nate is preferred for longer-term use (>2 years).
In patients who experience relapse with new bone disease, Recommendations for Using BP in Myeloma.
BP therapy, using pamidronate or clodronate if available, Starting BP Therapy. It is agreed that pretreatment dental
evaluation is important in patients who will be treated with Careful monitoring of renal function, including serum creatinine and periodic urinary protein measurements, is re- Bisphosphonate therapy is appropriate for patients with overt lytic bone disease on radiographs.
As noted by the Mayo Clinic team, osteonecrosis of the jaw Most investigators favor the use of other imaging studies in (ONJ) is the major new concern with long-term BP use.5-8 Use addition to radiographs to clarify the exact status of myeloma- of pamidronate or clodronate has the lowest risk, especially related bone disease as a basis for the decision to initiate BP with treatment duration of 1 or 2 years and the recommended therapy and to monitor bone disease serially. Magnetic reso- dental precautions. The risk of ONJ with pamidronate is 1% to nance imaging with gadolinium enhancement (confirmable 2% within the first 2 years of treatment, and the risk associated with computed tomography if necessary) and/or whole-body with clodronate is 0% to 0.5% (only rare cases have been computed tomography/positron emission tomography (if reported). The risk with zoledronic acid is approximately 2- available) can be used to identify focal bone destruction.3 fold higher than that with pamidronate.5-8 Oral nitrogen-con- Emphasis is placed on direct documentation of myeloma- taining BPs have been associated with ONJ5-8 and thus are not currently recommended as an alternative to clodronate.
The role of bone density testing is complex. Currently, it is Patients who develop ONJ should discontinue BP therapy.
not widely used in patients with myeloma. Nonetheless, those Choice of BP Therapy. There is nearly unanimous consen-
who do not use it concede and those who do use it emphasize its sus that pamidronate has equivalent efficacy compared to potential importance and utility both at baseline and for moni- other BPs plus a favorable toxicity profile. Thus, in general, toring. Bone loss, increasing age, female sex, and planned high- investigators favor pamidronate over zoledronic acid, as did dose corticosteroid use predict increased fracture risk and the potential need for oral or intravenous BP therapy. Other moni- Outside the United States, clodronate is a widely available toring tools include N-telopeptide and/or deoxypyridinoline oral alternative that is considered an equally safe option.9-11 measurements, which can indicate enhanced bone turnover.
Zoledronic acid is the primary choice of a few experts Even when all bone test results are inconclusive, some (13%), based on the convenience of the shorter infusion time investigators still recommend intravenous BP therapy for this but also because of the possible added clinical benefits of small subset of patients without definite bone disease but with reduction in bone events and/or improved survival compared documented active myeloma. However, BP use is not recom- with other BPs. Ongoing further studies of zoledronic acid are mended in patients with smoldering myeloma.
needed to assess schedules and infusion times, which may Duration of BP Therapy. We concur that BP use should no
reduce the risk of complications and/or enhance outcomes.
Dental Evaluation. Consensus exists concerning the need
We agree that the duration of BP therapy should be modi- for dental evaluation. The scope of baseline and serial evalua- fied on the basis of evidence of ongoing active bone disease.
tion depends on local details within the health care system as In patients who have achieved a complete response or a well as personal resources. Physicians should emphasize to very good partial response with transplantation and/or a novel their patients the need to avoid dental procedures such as therapy combination and have no active bone disease, BP extractions while they are receiving BP therapy.
therapy is not recommended beyond the first year. This rec- Conclusion. As noted recently,12 the preparation of good
ommendation is based on results from the Inter-Groupe guidelines is a complex process. We hope that the added details Francophone du Myelome trial, which showed no delay in provided herein will enhance the Mayo consensus statement.
time to onset of bone events with ongoing BP use.4 Brian G. M. Durie, MD, for the IMWG13Cedars-Sinai Outpatient Cancer Center 2007 Mayo Foundation for Medical Education and Research
Mayo Clin Proc. • April 2007;82(4):516-522 • www.mayoclinicproceedings.com For personal use. Mass reproduce only with permission from Mayo Clinic Pr Members of the International Myeloma Working Group.
9. McCloskey EV, Dunn JA, Kanis JA, MacLennan IC, Drayson MT.
Brian G. M. Durie, MD; Michel Attal, MD; Meral Beksac, MD; Long-term follow-up of a prospective, double-blind, placebo-controlled ran-domized trial of clodronate in multiple myeloma. Br J Haematol.
Andrew Belch, MD; William Bensinger, MD; Joan Bladé, MD; Mario Boccadoro, MD; Michele Cavo, MD; Raymond L.
10. Laakso M, Lahtinen R, Virkkunen P, Elomaa I, Finnish Leukaemia
Comenzo, MD; Meletios A. Dimopoulos, MD; Hermann Group. Subgroup and cost-benefit analysis of the Finnish multicentre trial ofclodronate in multiple myeloma. Br J Haematol. 1994;87:725-729.
Einsele, MD; Dorotea Fantl, MD; Gostä Gahrton, MD; Hartmut 11. Djulbegovic B, Wheatley K, Ross J, et al. Bisphosphonates in multiple
Goldschmidt, MD; Jean-Luc Harousseau, MD; Hiroyuki Hata, myeloma. Cochrane Database Syst Rev. 2002;3:CD003188.
MD; Joy Ho, MD; Vania Hungria, MD; Douglas Joshua, MD; 12. Steinbrook R. Guidance for guidelines. N Engl J Med. 2007;356:331-
Heinz Ludwig, MD; Giampaolo Merlini, MD; Angelina 13. Durie BGM, Harousseau J-L, Miguel JS, et al, International Myeloma
Rodriguez-Morales, MD; Antonio Palumbo, MD; Ray Powles, Working Group. International uniform response criteria for multiple myeloma.
MD; Donna E. Reece, MD; Anthony Reiman, MD; Jesús San Miguel, MD; Paul Richardson, MD; Orhan Sezer, MD;Kazayuki Shimizu, MD; Seema Singhal, MD; Pieter In reply: The Mayo Clinic Myeloma Group is pleased that the Sonneveld, MD; David Vesole, MD; Jan Westin, MD; Patrizia IMWG agrees in principle with most of our BP recommenda- Tosi, MD; Guido Tricot, MD; Ingemar Turesson, MD; Jeffrey tions. We believe that our groups concur on the most impor- tant aspects of our proposed guidelines. We agree that the 1. Lacy MQ, Dispenzieri A, Gertz MA, et al. Mayo Clinic consensus
duration of BP therapy should no longer be indefinite, that statement for the use of bisphosphonates in multiple myeloma. Mayo Clin pamidronate is the preferred drug, and that patients should have baseline and regular dental evaluations. We cannot 2. Tanvetyanon T, Stiff PJ. Management of the adverse effects associated
dispute the IMWG recommendations to include more forms with intravenous bisphosphonates. Ann Oncol. 2006 Jun;17:897-907. Epub2006 Mar 17.
of imaging studies to assist in the decision to initiate BP 3. Durie BG. The role of anatomic and functional staging in myeloma:
therapy. We have summarized in Table 1 the main aspects of description of Durie/Salmon PLUS staging system. Eur J Cancer. 2006 Jul; the Mayo Clinic recommendations and the response from the 4. Attal M, Harousseau JL, Leyvraz S, et al, Inter-Groupe Francophone du
Myelome. Maintenance therapy with thalidomide improves survival in patients Since our guidelines were published, new information has with multiple myeloma. Blood. 2006 Nov 15;108:3289-3294. Epub 2006 Jul 27.
become available that supports the conclusions made by us 5. Marx RE. Pamidronate (Aredia) and zoledronate (Zometa) induced avas-
cular necrosis of the jaws: a growing epidemic [letter]. J Oral Maxillofac Surg. and the IMWG. Zervas et al1 reviewed their experience with 303 patients with multiple myeloma and found that zoledronic 6. Durie BG, Katz M, Crowley J. Osteonecrosis of the jaw and bisphospho-
acid produced a 9.5-fold greater risk for developing ONJ nates [letter]. N Engl J Med. 2005;353:99-102.
7. Bamias A, Kastritis E, Bamia C, et al. Osteonecrosis of the jaw in cancer
than pamidronate alone (P=.042) and a 4.5-fold greater risk after treatment with bisphosphonates: incidence and risk factors. J Clin Oncol. than subsequent use of pamidronate plus zoledronic acid (P=–.018). In the absence of randomized trials, their findings 8. Badros A, Weikel D, Salama A, et al. Osteonecrosis of the jaw in multiple
provide strong support for the preference of pamidronate over myeloma patients: clinical features and risk factors. J Clin Oncol. 2006;24:945-952.
TABLE 1. Comparison of Mayo Clinic Consensus Statement and IMWG Recommendations* All myeloma patients with lytic bone disease on In addition to radiographs, the IMWG recommends other plain radiographs; patients with osteopenia imaging studies including MRI with gadolinium or osteoporosis on bone mineral density studies enhancement, CT, and/or whole-body CT/PET todetermine the presence or absence of bone lesions thatmay benefit from BP prophylaxis BPs are not recommended except in the setting Discontinue if complete response or VGPR occurs and Decrease to every 3 mo if disease is active Continue BP therapy if less than a VGPR occurs and/or ongoing active bone disease is evident Discontinue BPs if no active bone disease is evidentIf active bone disease is present, further BP use is recommended at the discretion of the primary physician *BP = bisphosphonate; CT = computed tomography; IMWG = International Myeloma Working Group; MRI = magnetic resonance imaging; PET = positron emission tomography; VGPR = very good partial response.
Mayo Clin Proc. • April 2007;82(4):516-522 • www.mayoclinicproceedings.com For personal use. Mass reproduce only with permission from Mayo Clinic Pr Attal et al2 conducted a randomized trial evaluating the Testimony at FDA hearings raised concern that a BBW utility of thalidomide and pamidronate in patients with mul- could create unintended barriers to care if generalist physi- tiple myeloma who had completed tandem autologous stem cians became reluctant to prescribe antidepressants. This is cell transplantation. Seven hundred eighty patients were en- especially critical because visits to nonpsychiatric generalist rolled and randomly assigned to no therapy (arm A), (primary care) physicians account nationally for more than pamidronate alone (arm B), or pamidronate plus thalidomide (arm C). The 3-year postrandomization probability of event- The FDA panel’s decision-making process has been hotly free survival was 36% in arm A, 37% in arm B, and 52% in debated in the media and academic literature. Remarkably, not arm C (P<.009). The 4-year postdiagnosis probability of sur- one child or adolescent among the 4400 cases the FDA re- vival was 77% in arm A, 74% in arm B, and 87% in arm C viewed completed suicide. Furthermore, the FDA inveighed (P<.04). The proportion of patients who had skeletal events against “suicidality”—a vague clinical concept encompassing was 24% in arm A, 21% in arm B, and 18% in arm C (P=.40).
everything from passive death wishes to near-lethal suicide On the basis of these findings, it appears that, as proposed by attempts. Meanwhile, untreated depression is the single most us and by the IMWG, prolonged treatment with BPs does not widely recognized risk factor for suicide at all ages, even as decrease the incidence of bone events or improve survival.
depression in children is underrecognized and undertreated ina nation with woefully inadequate numbers of child and ado- lescent psychiatrists. Thus, family practitioners and pediatri- cians write most antidepressant prescriptions for young Americans. Large observational pharmacoepidemiological studies link increasing rates of selective serotonin reuptake inhibitor antidepressant prescriptions and decreased nationalsuicide rates.6,7 In contrast, decreased antidepressant prescrib- 1. Zervas K, Verrou E, Teleioudis Z, et al. Incidence, risk factors and
ing by primary care physicians could lead to increased suicide management of osteonecrosis of the jaw in patients with multiple myeloma: rates, an effect opposite the FDA’s intent.
a single-centre experience in 303 patients. Br J Haematol. 2006;134:620- Mayo Clinic Survey. The reasons for the 20% decrease in
prescriptions filled for children and adolescents by June 2005 2. Attal M, Harousseau JL, Leyvraz S, et al. Maintenance therapy with
thalidomide improves survival in patients with multiple myeloma. Blood.
remain unclear. Similarly unclear is whether a subsequent expansion of the upper age limit within the BBW will haveadditional impact on physicians’ prescribing practices. Spe-cifically, in December 2006, an FDA panel voted 6 to 2 to Impact of the FDA Black Box Warning on Physician
expand the warning to include adults through age 24,8 a group Antidepressant Prescribing and Practice Patterns:
at high risk for both initial episodes of depression and suicide, Opening Pandora’s Suicide Box
particularly in men. To gain insight into antidepressant druguse across all ages, we conducted an anonymous Web survey To the Editor: After a decade-long decline, annual suicide of physicians at the Mayo Clinic in Rochester, Minn, between rates in American children and adolescents increased in 2004.
November 2005 and January 2006. About half of the sample A report released in February 2007 described an 18% increase that responded prescribe for both adults and children and the in the suicide rate in persons aged 1 through 19 years between other half for adults only. Most questionnaire items could be 2003 and 2004.1 The incidence of suicide, the third-leading answered only by those actually prescribing antidepressants at cause of death in 15- to 19-year-old Americans, increased from 7.3 to 8.2 per 100,000 persons in 2004.
Of the 1433 doctoral-level health care professionals sur- This increase in suicides directly parallels the US Food and veyed, 344 (24%) responded, including 44 (64%) of 69 psy- Drug Administration (FDA) recommendation for caution in chiatry and psychology department members, 89 (32%) of 275 prescribing antidepressants to children and adolescents. A generalists, and 211 (19%) of 1089 specialists. Of the 344
March 2004 FDA advisory warned that antidepressants could respondents, 324 (94%) indicated that they were currently cause worsening depression and increased suicidality. Within treating patients. Two of the survey respondents did not an- a month, the number of antidepressant prescriptions filled for swer the question regarding age groups treated. Twenty-one children nationally declined 10%.2 In October 2004, after (7%) of the remaining 322 respondents treated exclusively public committee hearings held in September, the FDA issued children and adolescents, 152 (47%) treated only adults, and a black box warning (BBW),3 its most serious warning, for 149 (46%) treated both groups. Of the 324 respondents who children and adolescents (defined as ≤18 years, making no were currently treating patients, 247 (76%) indicated that they clear distinction between adolescents and children), on all prescribed antidepressants, including 37 psychiatrists, 79 gen- antidepressants marketed in the United States. Although the FDA encouraged prescribers to “balance this risk with the Of the 247 respondents who prescribed antidepressants, clinical need,”4 prescriptions filled for children and adoles- 233 (94%) had heard about the BBW, including all 37 psy- cents had decreased an additional 10% by June 2005,2 2 chiatrists and 78 (99%) of the 79 generalists vs 118 (90%) of the 131 specialists. Overall, 176 (76%) of the 233 physicians Mayo Clin Proc. • April 2007;82(4):516-522 • www.mayoclinicproceedings.com For personal use. Mass reproduce only with permission from Mayo Clinic Pr FIGURE 1. Impact of antidepressant prescribing behavior by patient age among physicians aware of the BBW (N-232).
Adol = adolescents and children.
who were aware of the BBW reported continuing to pre- to patients in any age group. No physician group in our survey scribe at the same rate, whereas 57 (24%) had decreased or increased prescribing of antidepressants.
stopped prescribing antidepressants. (One generalist did not In addition to rates of antidepressant prescribing, the sur- respond to the question specifically addressing age groups vey collected information on practice pattern response after treated, which resulted in a sample size of 232). Figure 1 the BBW across the 3 physician groups. The data included in shows the changes in prescribing patterns by physician Table 1 are from all respondents who prescribed antidepres- sants, whether or not they had heard of the BBW (n=247). All Decreased prescribing in response to the BBW was par- physician groups reported changes in practice patterns, with ticularly marked for generalists seeing children and adoles- 35% of psychiatrists, 12% of generalists, and 9% of specialists cents only, with 82% (9/11) indicating that they had decreased spending more time reviewing the treatment plan at the initial or stopped antidepressant prescribing vs only 28% (9/32) of meeting. In addition, 11% of psychiatrists, 15% of generalists, generalists treating exclusively adults and 18% (6/34) seeing and 4% of specialists saw or contacted patients sooner after both adults and children. Psychiatrists did not report notable the first appointment, and 4% of specialists said they would try decreases in prescribing—the 8% decline reported in psychia- counseling patients first before prescribing antidepressants. In trists treating both adults and children appears to be related to the largest practice pattern change, we found that a total of less willingness of patients or guardians to accept medication.
43% (23% + 20%) of specialists and 26% (17% + 9%) of Whether they treated adults only or adults and children, 25% generalists reported increasing their referrals to a psychiatrist of specialists decreased or stopped prescribing antidepressants or other mental health specialist for consultation.
TABLE 1. Most Common Practice Pattern Responses* After FDA Black Box Warning, by Physician Group (N=247)† More likely to refer patients to a mental More likely to refer patients to a mental health specialist Less likely to prescribe antidepressants and will refer personally after initial appointment (15%) patients to a psychiatrist for antidepressantprescriptions (20%) Spend more time explaining rationale for prescribing personally after initial appointment (11%) More likely to counsel patients themselves first before prescribing antidepressants (4%) and see patients
sooner or contact them personally after initial *Not all practice pattern changes are listed. The electronic survey allowed selection of only 1 practice pattern response, with all responses totaling 100%.
†FDA = Food and Drug Administration.
‡ An equal number of specialists chose these 2 responses.
Mayo Clin Proc. • April 2007;82(4):516-522 • www.mayoclinicproceedings.com For personal use. Mass reproduce only with permission from Mayo Clinic Pr Discussion. The FDA intended the BBW to improve physi-
BBW did not take effect until late 2004. However, as we cian-patient communication and encourage closer physician described previously, antidepressant prescriptions filled for monitoring. Although our research suggests that this expecta- children and adolescents had decreased by 10%2 in April tion was realistic, the FDA did not adequately take into ac- 2004, 1 month after the initial FDA public advisory.
count the implications of nonpsychiatric physicians ceasing We fear that the correlation seen between declining youth to prescribe. We found that large numbers of both generalists antidepressant prescription rates and increasing suicide rates and nonpsychiatric specialists who formerly would have could soon be seen in older populations, particularly with the written antidepressant prescriptions themselves are choosing FDA currently considering its panel’s recommendation to ex- instead to refer patients to psychiatrists or other mental health tend the BBW to young adults. Further research is needed to clarify why nonpsychiatrists have decreased antidepressant The FDA panel should have known from existing research prescribing, opting instead for referrals to limited psychiatry that FDA warnings are linked to decreased prescribing.9 Our resources already unable to handle the demand. If further findings underscore the impact of a BBW on physician pre- research confirms the BBW’s role in reduced antidepressant scribing and practice patterns across all age groups. Although prescribing and increased suicide rates, revising the BBW and the results of our survey showed that nonpsychiatrists were getting generalists and nonpsychiatric specialists to prescribe less likely to prescribe antidepressants to any age group, the these drugs again should become public health priorities.
decreased prescribing was particularly marked in children.
Despite the limitations of our pilot research, including low response rates, a possibly nonrepresentative academic medical center setting, and an inability to correlate physician self- report with actual prescribing, our study results suggest that the FDA’s BBW markedly altered physician prescribing and In addition, the FDA BBW mandates closer follow-up of patients taking antidepressants without addressing the stark Funding for this study was provided by the Mayo Clinic Depart- realities of the current mental health system,10 including (1) ment of Psychiatry and Psychology. insurance plans with poor or absent mental health coverage,(2) increased copayments for individual mental health visits, 1. Hamilton BE, Minino AM, Martin JA, Kochanek KD, Strobino DM,
Guyer B. Annual summary of vital statistics: 2005. Pediatrics. 2007;119:345- (3) strict limits on outpatient mental health visits and inpatient psychiatric care, and (4) limited access to mental health pro- 2. Rosack J. New data show declines in antidepressant prescribing.
viders. Current shortages—particularly of child and adoles- Psychiatric News [serial online]. 2005 Sep 2;40:1-7. Available at: http:// cent psychiatrists—are not expected to reverse in the near pn.psychiatryonline.org/cgi/content/full/40/17/1-a. Accessed March 7, 2007.
3. US Food and Drug Administration. FDA public health advisory:
future. Despite FDA expectations of increased medication suicidality in children and adolescents being treated with antidepressant medica- monitoring, patients cannot afford and physicians cannot pro- tions. October 15, 2004. Available at: www.fda.gov/cder/drug/antidepressants vide the recommended contacts, let alone the gold standard of /SSRIPHA200410.htm. Accessed March 7, 2007.
4. US Food and Drug Administration. Class suicidality labeling language
for antidepressants. Available at: www.fda.gov/cder/drug/antidepressants In terms of increasing physician-patient communication, /PI_template.pdf. Accessed March 7, 2007.
our results indicate that the BBW may have selectively suc- 5. Stafford RS, MacDonald EA, Finkelstein SN. National patterns of medi-
ceeded in obtaining one of the FDA’s goals, providing earlier cation treatment for depression, 1987 to 2001. Prim Care Companion J Clin contact and more communication, which was noted in all physician groups and particularly in psychiatrists. As antide- 6. Olfson M, Shaffer D, Marcus SC, Greenberg T. Relationship between
antidepressant medication treatment and suicide in adolescents. Arch Gen pressant experts with nowhere to refer patients, more than a third of our psychiatrists reported talking to patients more, as 7. Gibbons RD, Hur K, Bhaumik DK, Mann JJ. The relationship between
did smaller proportions of both generalists and specialists. Of antidepressant medication use and rate of suicide. Arch Gen Psychiatry. concern in the current environment of limited psychiatric re- 8. Young D. Antidepressant black-box warning should include young adults,
sources, many generalists and specialists are choosing to no panel urges. ASHP Web site. January 15, 2007. Available at: www.ashp.org longer prescribe antidepressants themselves and instead refer /s_ashp/sec_news_article.asp?CID=167&DID=2024&id=18022. Accessed their patients to mental health specialists for care that may not be available in a timely fashion, if at all.
9. Lasser KE, Allen PD, Woolhandler SJ, Himmelstein DU, Wolfe SM, Bor
DH. Timing of new black box warnings and withdrawals for prescription Given that reduced prescribing by generalists in our survey medications. JAMA. 2002;287:2215-2220.
was most marked in children and adolescents, we fear that 10. US Department of Health and Human Services. Mental Health: A Report
decreased antidepressant prescribing due to BBW concerns of the Surgeon General. Chapter 6: Organizing and Financing Mental Health may underlie increasing youth suicide rates.1 Disconcertingly, Services. Available at: www.surgeongeneral.gov/library/mentalhealth/pdfs/c6.pdf. Accessed March 7, 2007.
a warning of increased “suicidality”—in the absence of any 11. March J, Silva S, Petrycki S, et al. Fluoxetine, cognitive-behavioral
actual suicides—could conceivably have driven increased sui- therapy, and their combination for adolescents with depression: treatment for cide rates already observed in American youths in 2004, the adolescents with depression study (TADS) randomized controlled trial. JAMA. year the BBW took effect. An argument can be made that the Mayo Clin Proc. • April 2007;82(4):516-522 • www.mayoclinicproceedings.com For personal use. Mass reproduce only with permission from Mayo Clinic Pr Sorafenib-Induced Pancreatitis
should be included in the differential diagnosis of patientswith abdominal pain who are receiving sorafenib. Recognition To the Editor: Sorafenib is a multikinase inhibitor recently of adverse effects from novel agents is essential to ensure approved by the US Food and Drug Administration for use appropriate diagnosis and treatment.
in the treatment of metastatic renal cell carcinoma. Thelong-term and toxic effects of sorafenib are currently unknown.
We thank Massimo Raimondo, MD, for performance and interpreta- We describe a patient who experienced sorafenib-induced tion of the endoscopic ultrasonographic studies in the patient.
Report of a Case. A 53-year-old woman had renal cell
carcinoma metastatic to the lungs and bones. Treatment with sorafenib, 400 mg twice a day, was initiated. Three weeks later, severe epigastric pain, nausea, and vomiting developed.
Her serum amylase level was 1361 U/L (reference range, 26- 102 U/L) with a lipase level of 99 U/L (10-73 U/L). Abdomi-nal computed tomography showed normal findings. Sorafenib 1. Wilhelm SM, Carter C, Tang L, et al. BAY 43-9006 exhibits broad spectrum
was discontinued, and the patient was hospitalized for treat- oral antitumor activity and targets the RAF/MEK/ERK pathway and receptortyrosine kinases involved in tumor progression and angiogenesis. Cancer Res.
ment of acute pancreatitis. At discharge 2 weeks later, she continued to have moderate pain requiring narcotics. Pancre- 2. Escudier B, Szczylik C, Eisen T, et al. Randomized phase III trial of the Raf
atitis was grade 2 according to the Common Terminology kinase and VEGFR inhibitor sorafenib (BAY 43-9006) in patients with advanced Criteria for Adverse Events (CTCAE).
renal cell carcinoma (RCC) [abstract]. J Clin Oncol. 2005;23(suppl):1093s.
Abstract 4510.
Eight weeks after sorafenib discontinuation, the patient’s 3. Nevaxar (sorafenib) [package insert]. West Haven, Conn: Bayer Pharma-
pain had improved, and her amylase and lipase levels were ceuticals Corporation; 2006. Available at: www.univgraph.com/bayer/inserts 863 U/L and 44 U/L, respectively. She was subsequently /nexavar.pdf. Accessed October 23, 2006.
4. Strumberg D, Richly H, Hilger RA, et al. Phase I clinical and pharmacoki-
referred to our medical center for further evaluation.
netic study of the novel Raf kinase and vascular endothelial growth factor Repeated abdominal computed tomography at our institu- receptor inhibitor BAY 43-9006 in patients with advanced refractory solid tion revealed a normal pancreas, liver, gallbladder, and bil- tumors. J Clin Oncol. 2005 Feb 10;23:965-972. Epub 2004 Dec 21.
iary ducts. Endoscopic ultrasonography showed a normal 5. Pierce GF, Mustoe TA, Lingelbach J, et al. Platelet-derived growth factor
and transforming growth factor-beta enhance tissue repair activities by unique gallbladder, common bile duct, and main pancreatic duct.
mechanisms. J Cell Biol. 1989;109:429-440.
Abnormal hyperechogenic foci were noted in the pancreas,along with strands and lobulation consistent with chronicpancreatitis.
Did a Flawed Study Design Affect the Results of the
The patient did not drink alcohol and had no prior hepato- VYTAL Study?
biliary disease, no family history of pancreatitis, no precedingviral illness, and normal serum triglyceride and calcium levels.
To the Editor: I am concerned that the design of the Vytorin Sorafenib was her only medication. The temporal association vs Atorvastatin in Patients With Type 2 Diabetes Mellitus with initiation of sorafenib and lack of other causes support and Hypercholesterolemia (VYTAL) Study, as outlined by the diagnosis of sorafenib-induced pancreatitis.
Goldberg et al,1 affected the results they reported. The prob- Discussion. Sorafenib is an oral multikinase inhibitor
lem is the large number of “participating centers” and “study with antiangiogenic and antiproliferative activity. It inhibits investigators” (147) and the small number of patients per cen- platelet-derived growth factor receptor, vascular endothelial ter. Given the number of patients randomized, each site sup- growth factor (VEGF) receptor, and the Ras/Raf/MEK path- plied a mean of only 8.4 patients or 1.66 patients per study way.1 Common adverse effects include diarrhea, rash, fatigue, and hypertension.2 Postmarketing data3 have noted both lipase Unclear in this report is why the participation of 147 cen- ters was needed to enroll 1229 patients. The patients being In one study, clinical pancreatitis was reported in 3 of 451 studied, individuals with type 2 diabetes and hypercholester- sorafenib-treated patients (1 had CTCAE grade 2, and 2 had olemia who were between 18 and 80 years of age and had CTCAE grade 4).3 Two of these patients subsequently re- hemoglobin A levels of 8.5% or less, are, unfortunately, sumed sorafenib.3 A phase 1 study of 69 sorafenib-treated common. The reason for so many centers and so few patients patients reported 3 cases of dose-independent grade 3 pancre- per center was not specified. Center selection itself could atitis. All 3 patients recovered after withdrawal of sorafenib.4 introduce bias depending on the selection criteria, but I will In our patient, the severity of symptoms and their incomplete confine my remarks primarily to the study design, which may resolution precluded reinitiation of sorafenib.
Our hypothesis is that the anti-VEGF effects of sorafenib Because the study used random assignment, some centers cause ischemia of pancreatic tissue, increasing the risk of would have no patients in some study arms, whereas others pancreatitis. Also, its continued action impairs the protective would have all their patients randomized to just 1 or 2 arms.
effects of VEGF and platelet-derived growth factor,5 increas- This design could lead to bias because of individual practice ing the severity of the resulting pancreatitis. Pancreatitis variations, especially in reference to the intensity of treatment Mayo Clin Proc. • April 2007;82(4):516-522 • www.mayoclinicproceedings.com For personal use. Mass reproduce only with permission from Mayo Clinic Pr of the comorbidity (diabetes). It is unlikely that nonpharma- dietary considerations, screening and enrollment criteria, ran- ceutical interventions such as dietary counseling, staff en- domization, and follow-up visits. Furthermore, patients were couragement, and follow-up visit variations were controlled in enrolled after a 4-week placebo run-in period before the study all 147 centers. Whether the centers were all primary care treatment was initiated. Should there have been between- facilities or included subspecialty care could have impacted site(s) treatment differences as suggested by Dr McMahon, we these other interventions because of variations in treat- would have expected greater variability in baseline character- ment regimens and goals for patients with diabetes and lipid istics and measurements to result during this run-in period, including baseline low-density lipoprotein cholesterol levels, Having few patients per center might also dilute patient which were found to be similar across treatment groups.
anonymity, leading to unintended effects on compliance and To be eligible for entry into the trial, patients were not other factors. Including a large number of sites can also lead to allowed to receive lipid-lowering drug therapy for 6 to 8 difficulty in controlling the introduction of investigator biases.
weeks before enrollment. This requirement provided adequate For example, prerandomization bias may have occurred be- time for wash-out of any lingering lipid-altering effects. Thus, cause it is unclear how patients were selected at each center.
the results of the study are representative of the expected The authors state that they were “randomized…after washout response to the initiation of therapy with ezetimibe/simva- and placebo run-in periods,” but how they were initially se- statin or atorvastatin in the study population. Moreover, the lected or if there were any selection criteria is unclear. Without lipid-altering effects of ezetimibe/simvastatin and atorvastatin a study-wide protocol to prevent enrollment bias toward fail- demonstrated in this trial are similar to the results shown in ures of current lipid-lowering treatment, there could be a previous trials in different patient populations.1,3,4 positive trend toward one study arm. Knowing the number of We agree that using a study design that minimizes bias is patients whose hypercholesterolemia was controlled or not important, and we believe that the protocol used in the controlled with what specific medications at selection is im- portant if this study portends to evaluate first-line treatmentoptions, as stated in the conclusion.
Thus, it is possible that, by designing a study with such a large number of centers and resultant small number of patients per center, the efforts for randomness, double blinding, and clinical usefulness may have been defeated.
David E. McMahon, MDMount Carmel Medical Center 1. Goldberg RB, Guyton JR, Mazzone T, et al. Ezetimibe/simvastatin vs
atorvastatin in patients with type 2 diabetes mellitus and hypercholesterolemia: the VYTAL Study. Mayo Clin Proc. 2006;81:1579-1588.
In reply: We appreciate Dr McMahon’s comments regarding the potential for study bias and error in the VYTAL Study.
The study design and recruitment procedure in this trial were similar to those used in other relatively large shorter-term efficacy studies, including the Vytorin Versus Atorvastatin (VYVA) Study, which compared the safety and efficacy of ezetimibe/simvastatin vs atorvastatin in 1902 patients ran- domized to treatment at 216 sites,1 and a recent study thatcompared ezetimibe/simvastatin vs rosuvastatin in 2959 pa- 1. Ballantyne CM, Abate N, Yuan Z, King TR, Palmisano J. Dose-compari-
tients with hypercholesterolemia who were randomized to son study of the combination of ezetimibe and simvastatin (Vytorin) versusatorvastatin in patients with hypercholesterolemia: the Vytorin Versus treatment at 214 centers.2 Although recruitment from fewer Atorvastatin (VYVA) Study [published correction appears in Am Heart J. 2005; sites is possible, the potential for site-related bias would be 149:882]. Am Heart J. 2005;149:464-473.
greater. In the VYTAL Study, patients were randomized by 2. Catapano AL, Davidson MH, Ballantyne CM, et al. Lipid-altering efficacy
of the ezetimibe/simvastatin single tablet versus rosuvastatin in hypercholes- allocation concealment using an interactive voice response terolemic patients. Curr Med Res Opin. 2006;22:2041-2053. system to ensure an even distribution of patients among 3. Ballantyne CM, Blazing MA, King TR, Brady WE, Palmisano J. Efficacy
treatment groups. This even distribution is confirmed by the and safety of ezetimibe co-administered with simvastatin compared with ator- fact that the baseline characteristics among the treatment vastatin in adults with hypercholesterolemia. Am J Cardiol. 2004:93:1487-1494.
4. McKenney J, Ballantyne CM, Feldman TA, et al. LDL-C goal attainment
To avoid individual site variations, all the investigators with ezetimibe plus simvastatin coadministration vs atorvastatin or simvastatin followed a standard protocol that defined the procedure for monotherapy in patients at high risk of CHD. MedGenMed. 2005;7:3.
Mayo Clin Proc. • April 2007;82(4):516-522 • www.mayoclinicproceedings.com For personal use. Mass reproduce only with permission from Mayo Clinic Pr

Source: http://myeloma.ca/docs/bisphosphonates_letter_to_the_editor_mayo_proceedings.pdf