Letters found that sTNF-R1 is independently as- ●●●●●●●●●●●●●●●●●●●●●●● travenous insulin and glucose (1) and dis- sociated with albuminuria in type 2 dia- References charged 7 days later in good condition. betic patients (4). To the best of our 1. Bierman EL: George Lyman Duff Memo- This report confirms that phenform- rial Lecture: Atherogenesis in diabetes. knowledge, however, it is not clear in-induced lactic acidosis (PLA) is still a Arterioscler Thromb 12:647–656, 1992 public health problem (1,2). To our whether serum homocysteine is associ- 2. Refsum H, Ueland M: Homocysteine and ated with TNF receptor in type 2 diabetic cardiovascular disease. Annu Rev Med 49: knowledge, phenformin is still used in patients. The aim of the present study was 31–62, 1998 Italy, China, and Brazil. In a Medline therefore to investigate the relationships 3. Rauchhaus M, Doehner W, Francis DP, search, we found 12 cases that occurred in between serum homocysteine and TNF Davos C, Kemp M, Liebenthal C, Nie- Italy between 1981 and 1998 (2). In two receptor in patients with type 2 diabetes. bauer J, Hooper J, Volk HD, Coats AJS, patients phenformin was even brought Fifty nonobese Japanese type 2 dia- Anker SD: Plasma cytokine parameters back into use soon after, thereby ques- and mortality in patients with chronic betic patients were studied. Their BMI, tioning the belief that PLA is adequately heart failure. Circulation 102:3060–3067, recognized (2). More importantly, ac- HbA , and serum creatinine were 22.6 Ϯ 1c 2000 cording to data by Intercontinental Mar- 0.3 kg/m2 (range 17.6–26.2), 7.8 Ϯ 4. Kawasaki Y, Taniguchi A, Fukuhishima keting Services (www.imshealth.com), 0.2% (5.5–12.3), and 0.70 Ϯ 0.02 mg/dl M, Nakai Y, Kuroe A, Ohya M, Nagasaka 838,000 preparations of phenformin and (0.46–0.98), respectively. They had not S, Yamada Y, Inagaki N, Seino Y: Soluble TNF receptors and albuminuria in non- a sulfonylurea have been sold in Italy be- been treated with insulin or any medica- tween January and October 2005. Be- tions known to alter homocysteine level. obese Japanese type 2 diabetic patients. Horm Metab Res 37:617–621, 2005 cause PLA occurs in 1 of 4,000 patients In conjunction with homocysteine, sys- (3) with a mortality rate of ϳ50%, these tolic and diastolic blood pressure, HbA1c, data raise worrying health care consider- ␣ glucose, lipids, serum creatinine, TNF- , Phenformin-Induced ations. In fact, diabetic patients often have sTNF-R1, and sTNF-R2 were measured Lactic Acidosis in an comorbid conditions known to favor after an overnight fast. Older Diabetic PLA. With univariate analysis, serum ho- Phenformin was removed from the mocysteine was positively correlated with Patient U.S. market in 1977, but, surprisingly, age (r ϭ 0.361, P ϭ 0.012), diabetes du- cases of patients who have been pre- ration (r ϭ 0.292, P ϭ 0.045), serum cre- A recurrent drama (phenformin scribed the drug abroad are continuously atinine (r ϭ 0.623, P Ͻ 0.001), sTNF-R1 and lactic acidosis) reported (1). Phenformin can also be ille- (r ϭ 0.415, P Ͻ 0.005), and sTNF-R2 gally obtained online or through mail ϭ Ͻ Editor’s note: The authors had the following (r 0.371, P 0.01). Other variables statement in their letter to me, with which I orders to replace metformin, which is including TNF-␣, however, were not as- agree, “Most physicians are aware of the risk of more costly. Furthermore, herbal medi- sociated with homocysteine. Multiple re- lactic acidosis in patients taking phenformin. cines containing phenformin are also gression analyses showed that serum However, this side effect is continuously ob- consumed in developed countries. In homocysteine was independently associ- served because phenformin is still used in Italy, February 2000, the Food and Drug Ad- ated with serum creatinine (F ϭ 20.1) and Brazil, and China. We believe that the publica- ministration recalled five Chinese herbal tion of our observation in an important journal sTNF-R1 (F ϭ 6.9), which explained medications containing phenformin (4), like Diabetes Care may help to prompt govern- while Health Canada is currently warning 49.3% of the variability of homocysteine. ments of these countries to ban phenformin, just Thus, TNF system activity may be respon- consumers not to take “Shortclean,” a like in the rest of the world. This is the only way phenformin-based Chinese “natural” sible for the evolution of atherosclerosis to prevent further cases of this avoidable, un- medicine (5). induced by homocysteine in nonobese acceptable and life-threatening complication.” Phenformin can always be replaced Japanese type 2 diabetic patients. 73-year-old man with diabetes pre- by metformin, which should not be asso- sented with upper-abdominal pain ciated with a higher risk of lactic acidosis 1 ATARU TANIGUCHI, MD and nausea. He also had a history of compared with nonbiguanide therapies 2 A MITSUO FUKUSHIMA, MD hypertension, a pace-maker implant, and (6). Despite most clinicians being aware 3 YOSHIKATSU NAKAI, MD peripheral arterial disease treated with of PLA, the only way for preventing fur- 1 MINAKO OHGUSHI, MD amputation of his left leg. His therapy in- ther cases is to forbid phenformin in 1 AKIRA KUROE, MD cluded ticlopidine, enalapril, omepra- countries where it is still used. 1 MICHIHIRO OHYA, MD zole, and 2 mg glibenclamide/30 mg 1 1,2 UTAKA EINO MD phenformin b.i.d. The patient was alert FILIPPO LUCA FIMOGNARI, MD Y S , 1 and cognitively intact. Blood pressure and RUGGERO PASTORELLI, MD 2 heart rate were 120/70 mmHg and 70 RAFFAELE ANTONELLI INCALZI, MD From the 1Division of Diabetes and Clinical Nutri- bpm, respectively. Radiographs of the tion, Kansai-Denryoku Hospital, Osaka, Japan; the From the 1Division of Internal Medicine, Leopoldo 2 chest and abdomen and an abdominal ul- Department of Health Informatics Research, Trans- Parodi-Delfino Hospital, ASL Roma G, Colleferro lational Research Informatics Center, Kobe, Japan; trasound study were normal. Laboratory (Rome), Italy; and 2University Campus Biomedico 3 and the Karasuma-Nakai Clinic, Kyoto, Japan. tests disclosed a severe lactic acidosis (pH of Rome, Rome, Italy. Address correspondence to Ataru Taniguchi, 6.8, pCO2 14.1 mmHg, pO2 108 mmHg, Address correspondence to Dr. Filippo L. Fimo- MD, Division of Diabetes and Clinical Nutrition, HCO 4.9 mmol/l, lactate 21 mmol/l, and gnari, Centro per la Salute dell’Anziano (CeSA), Uni- Kansai-Denryoku Hospital, 2-1-7 Fukushima, 3 versity Campus Biomedico of Rome, Via dei Fukushima-ku, Osaka City, Osaka 553-0003, Ja- anion gap 31 mmol/l). After phenformin Compositori 130, 00128, Rome, Italy. E-mail: pan. E-mail: [email protected]. discontinuation, the patient’s conditions filippo.fi[email protected]. © 2006 by the American Diabetes Association. rapidly improved. He was treated with in- © 2006 by the American Diabetes Association.

950 DIABETES CARE, VOLUME 29, NUMBER 4, APRIL 2006 Letters

concern. I was frustrated and angered by staff member about my diabetes and my Acknowledgments— We thank Dr. Antonio Muroni, Laboratori Guidotti SPA, Italy, for substitution of the sliding-scale for my treatment plan, and I always had a family providing data from Intercontinental Market- normal insulin regimen, especially as my member available to do this when I was ing Services. blood glucose spiraled out of control. It is unable to speak for myself. While this in- encouraging that this critical issue is re- formation was contained in my chart, it is ●●●●●●●●●●●●●●●●●●●●●●● ceiving increased attention. unreasonable to expect that everyone re- References My own bouts with surgeries at the members every detail all the time. These 1. Kumar A, Nugent K, Kalakunja A, Pirtle F: University of Pittsburgh Medical Center approaches kept my diabetes in the fore- Severe acidosis in a patient with type 2 this past winter highlight the benefits of front and dramatically enhanced each diabetes mellitus, hypertension, and renal focusing on inpatient diabetes manage- hospitalization. failure. Chest 123:1726–1729, 2003 ment. Not only did I demand a consult 2. Enia G, Garozzo M, Zoccali C: Lactic aci- with my endocrinologist and her staff to MICHAEL A. WEISS dosis induced by phenformin is still a develop and implement a treatment plan public health problem in Italy (Letter). From the American Diabetes Association, Alexan- for my hospital stay, I requested intrave- dria, Virginia. BMJ 315:1466–1467, 1997 Address correspondence to Michael A. Weiss, 58 3. Kreisberg R, Wood BC: Drug and chemi- nous insulin infusions during surgery and in the recovery room and intensive care Glen Ridge Ln., Pittsburgh, PA 15243. E-mail: cal-induced metabolic acidosis. Clin En- [email protected]. docrinol Metab 12:391–411, 1983 unit. The growing evidence supporting © 2006 by the American Diabetes Association. 4. Aschwanden C: Herbs for health, but how the value of infusions is overwhelming. safe are they? Bull World Health Organ 79: It is sometimes easy, however, to ●●●●●●●●●●●●●●●●●●●●●●● 691–692, 2001 overlook another effective tool for in- References 5. Health Canada warns consumers not to hospital diabetes management— the pa- 1. Peterson AA, Charney P, Rennert NJ: take the Chinese medicine Shortclean due tient. Undoubtedly, the most helpful step Eliminating inpatient sliding-scale insu- to potential health risk [article online], for me was continuing to manage my own lin: a reeducation project with medical 2005. Available from http://www. insulin pump therapy while in the hospi- house staff (Letter). Diabetes Care medicalnewstoday.com/medicalnews. 12:2987, 2005 php?newsidϭ34038#. Accessed 23 tal. Clearly, every patient demonstrating proficiency, whether using a pump or 2. Baldwin D, Villanueva G, McNutt R: November 2005 Eliminating inpatient sliding-scale insu- 6. Salpeter S, Greyber E, Pasternak G, Sal- multiple insulin injections, should be en- couraged to continue self-management lin: a reeducation project with medical peter E: Risk of fatal and nonfatal lactic house staff (Letter). Diabetes Care 12: acidosis with metformin use in type 2 di- on the nursing floor. For me, this allevi- 2987, 2005 abetes mellitus: systematic review and ated the anxieties often felt by patients meta-analysis. Arch Intern Med 163: when their diabetes management routines 2594–2602, 2003 have been disrupted. And the results were Acute Neuropathic phenomenal. My blood glucose stayed Joint Disease: A within normal ranges almost the entire Medical Emergency? COMMENTS AND time! With the help and oversight of my consulting endocrinologist and certified RESPONSES diabetes educators, self-management pre- Response to Tan et al. sented few difficulties for me and relieved Patient Self- the surgical staff of this additional respon- e read with some interest the sibility. Many of the nurses and other hos- commentary by Tan et al. (1) on Management of pital staff were actually curious to learn W the management of the Charcot Insulin Doses in the more about insulin therapy and, particu- foot in diabetes. While we agree entirely larly, pump therapy. that this condition should be ranked as a Hospital I kept my own glucose monitor and a medical emergency, because failure to act This letter may seem “far out,” but in my expe- supply of strips with me, as well as re- quickly can lead to irreversible adverse rience and that of some of my colleagues, with placement batteries and other pump sup- consequences, we do not agree that the select patients, usually type 1’s, patient self- plies. While staff would routinely check evidence is available to support uncritical management (with physician oversight) yields my blood glucose levels, the timing was use of bisphosphonates. The only blinded better glycemic results (and less patient and somewhat irregular and did not always trials conducted so far did not demon- physician anxiety) than if insulin dosing is left correlate with meals. I carefully recorded strate any overt improvement in long- to the vagaries of the busy floor staff. Getting monitor readings, food intake, and insu- term prognosis (2,3). There is much the hospital administration to allow this is often lin dosing to review with my consulting suggestive evidence to favor the consider- the biggest challenge. diabetes specialists. I also maintained a ation of bisphosphonate use, but it is not eading the exchange of letters in the cache of glucose tablets and fruit juice to currently accepted by all authorities that December 2005 issue regarding the treat inevitable lows. Although these were this therapy is essential. R management of inpatient hypergly- available on the hospital floor, I was con- A number of other treatments also de- cemia made me reflect on personal expe- cerned about getting the immediate atten- serve consideration (4,5). For example, riences as a hospital patient (1,2). In past tion of the nursing staff during a sudden intranasal calcitonin and tumor necrosis years, as surgeons and cardiologists hypoglycemic episode. Nevertheless, I re- factor-␣ antagonists may prove useful, al- tended to their more immediate tasks, my ported every incident and the actions though the efficacy of both has yet to be diabetes was often relegated to a second- taken for entry into my medical records. established in controlled trials. In the ab- ary and sometimes seemingly nonexistent In addition, I reminded every hospital sence of evidence to support the use of

DIABETES CARE, VOLUME 29, NUMBER 4, APRIL 2006 951