Faculty Peer Reviewed
History provides many examples of medical interventions that were intended for one use, but were ultimately found therapeutic for a wholly different purpose. A review of this week’s prominent medical journals finds a number of studies proposing new applications for established treatments, including the use of bariatric surgery for inducing the remission of diabetes, bone marrow cells to improve heart failure, aspirin for prevention of cancer metastasis, and highly active antiretroviral therapy (HAART) for preexposure prophylaxis in those most at risk for HIV infection.
Until now, only one randomized trial has suggested the superiority of gastric banding over usual medical care for the treatment of type 2 diabetes, and this trial involved patients with relatively less severe diabetes of shorter duration.[1] In the present STAMPEDE Trial,[2] Schauer et. al. of the Cleveland Clinic randomized 150 patients at that center to one of three treatment groups: intensive medical therapy alone, or medical therapy plus either Roux-en-y bypass or sleeve gastrectomy. Patients had a mean age of 48 years, mean BMI of 36 with a range of 27 to 43, and a mean duration of diabetes of 8 years, with 22% using insulin. 74% of patients were white. All patients were treated with aggressive medical therapy to achieve an A1c of less than 6%. The primary outcome was the proportion of patients with an A1c less than 6% at 12 months after randomization. This was achieved in 12% of the medical therapy group, 42% of the gastric bypass group and 37% of the sleeve gastrectomy group. The differences between medical therapy and either surgical option were highly statistically significant, but the difference between the two surgical groups was not significant. Notably, all patients in the gastric bypass group who achieved an A1c less than 6% did so without use of any medications or insulin, whereas 28% of patients in the sleeve gastrectomy group still required medical therapy.
Across all patients, mean A1c and fasting glucose were significantly lower in the surgical groups as compared the medical group at 12 months. Improvement in the surgical groups was seen at 3 months and sustained at one year. The average number of diabetes agents required for glycemic control increased in the medical therapy group but decreased in the two surgical groups, with 38% of the medical group still requiring insulin as compared 4% in the bypass group and 8% in the sleeve group. Although there was no difference in low-density lipoprotein (LDL) or mean blood pressure among groups, fewer medications were required in the surgical groups to achieve improved results. Improvement in glycemic control predated maximal weight loss, suggesting improvement in insulin resistance as the potential mechanism for improved glycemic control, which was supported by improvement in blood insulin levels in the surgical groups. Although there were no deaths among the three groups, additional surgical interventions were required in four of the surgically treated patients. The study was limited by short follow-up, a single center design at a tertiary academic center, and inadequate power and follow-up to show improvements in cardiovascular clinical outcomes.
Another New England Journal study by Mingrone et. al. compared conventional medical therapy versus gastric bypass or biliopancreatic diversion for inducing the remission of type 2 diabetes.[3] This randomized Italian study enrolled 60 patients with a BMI of 35 or more, at least a 5 year history of diabetes and an A1c greater than 7 to standard medical therapy with goal A1c less than 7, or standard medical therapy with gastric bypass or biliopancreatic diversion. The primary end point was remission of type 2 diabetes as defined by a fasting plasma glucose less than 100 or an A1c less than 6.5% for 1 year without medical therapy. The mean age of participants was 42, mean A1c was 8.6, mean BMI was 45, and the mean duration of diabetes was 6 years. At two years of follow-up, there were no remissions in the medical management group, whereas 75% of the patients in the bypass group and 95% of patients in the diversion group achieved remission, a statistically significant difference for each comparison. The average time to remission was 10 months in the bypass group and 5 months in the biliopancreatic diversion group. Age, sex, baseline BMI and A1c did not predict the ability to achieve remission. Among all patients, a significantly greater improvement in mean A1c was seen in each surgical group as compared the medical group. The lipid profile was most favorably improved by biliopancreatic diversion, except for HDL, which improved most after gastric bypass. At 2 years, total cholesterol normalized in 27% of patients in the medical therapy group as compared with all patients in the surgical groups. Blood pressure was improved in all groups, but the surgical groups required less medication at the end of follow up. Of note, there was again no correlation between the degree of weight loss and the likelihood of achieving remission amongst the surgically treated patients, suggesting that the ameliorating effects of bariatric surgery occur through means other than weight loss itself. The preoperative BMI also did not predict likelihood of remission. Limitations again included short duration of follow up, small sample size, single center design and inability to detect differences in clinical outcomes. There were no deaths in either group, but two surgical patients required re-operation.
The utility of bone marrow cells (BMCs) has been hotly pursued for the treatment of many hematologic diseases for some time. An interesting article in this week’s Journal of the American Medical Association sought to further assess the utility of transendocardial delivery of bone marrow mononuclear cells for the treatment of chronic ischemic heart disease and left ventricular dysfunction.[4] A previous phase I study by the same group had shown the safety of this method but was insufficiently powered to assess its efficacy. The present study, FOCUS-CCTRN, was a phase 2 randomized double-blinded placebo controlled trial enrolling patients with clinically stable coronary disease, an LVEF less than 45% and NYHA class II-III heart failure with perfusion defects on SPECT imaging but no anatomy suitable to revascularization. 92 patients were randomized in a 2:1 ratio to transendocardial delivery of either 90 to 100cc of BMC from the iliac crest or a placebo sham procedure. BMCs were delivered to left ventricular regions identified as viable by electromechanical sampling. The mean age was 63, LVEF was 31%, and 95% of patients were male. The primary outcomes were left ventricular end-systolic volume (LVESV), maximal oxygen consumption as measured by the Naughton treadmill method and defect size on SPECT at 6 months after delivery as compared to baseline. At its conclusion, the study showed no difference in any of the primary outcomes, or in NYHA class, need for medications for angina or BNP. However, at six months, there was significant improvement in LVEF and stroke volume in the BMC group as compared the placebo group, although the improvements were small (1.4% as compared -1.3%, 2.7mL as compared 5.8mL, respectively). Interestingly, the authors found that improvement in LVEF correlated with the proportion of CD34 and CD133 cells in the delivered sample. These cells are known to be involved in cell survival and give rise to vascular and endothelial progenitor cells. The study was limited by sample size, especially given that power calculations were based on likely unrealistic estimates of improvement (LVESV of 27mL, SPECT reversibility of 10 percentage points). Based on the data the authors provided, there were no adverse events clearly associated with BMC therapy.
The benefits of aspirin in the primary and secondary prevention of cardiovascular disease have been well studied, and the rich data associated with these studies has now made it possible to study the effect of aspirin on other disease processes. In the current issue of the Lancet, Rothwell et. al. conduct a meta analysis evaluating the effect of aspirin on the risk of cancer metastasis.[5] These authors noted that in other recent studies on aspirin and cancer, the effect of aspirin on cancer mortality was greater than the effect on cancer incidence and also that the mortality reduction occurred too quickly to be due to primary prevention alone. As a result, the group hypothesized that aspirin may also affect the metastatic progression of established occult cancers at the time of study enrollment, which they endeavored to show in this trial. Rothwell analyzed the 5 major randomized trials of aspirin versus placebo control for the primary prevention of vascular events in the United Kingdom. Through detailed chart review and communication with the primary authors, data were extracted for all cases of incident cancer during the trials, including the date of diagnosis of solid cancer, site of primary, staging and treatment, presence of metastasis at diagnosis and follow-up and mortality due to cancer or other causes. Hematologic cancers and primary brain cancer were excluded. Among the five trials included, there were 17,285 participants, 1,101 incident cancers and 563 deaths due to cancer. Definitive determination of metastatic disease could not be made in 20% of the studied cases. Aspirin therapy was found to reduce the incidence of any new cancer during the trial (OR 0.88 p=0.04) but to have an even greater effect on the risk of death due to incident cancer (OR 0.77 p=0.002). In support of the author’s hypothesis, aspirin was found to reduce the risk of cancer with definite metastatic disease at the time of diagnosis (HR 0.64 p=0.001) and conversely increase the risk of cancer with local disease only. There was no predilection to reduce the risk of metastasis by any particular metastatic site. More detailed analysis showed that the protective effect of aspirin on metastatic cancer at diagnosis was significant for colon cancer, with trends towards significance in other adenocarcinomas, but without a significant effect or trend in non-adenocarcinomas. In patients with adenocarcinoma who did not have metastatic disease at diagnosis, the risk of later metastasis was reduced by aspirin (HR 0.45, p=0.0009) with the effect most pronounced in colon cancer patients. Overall survival was higher in aspirin groups as compared controls for death due to any cause and death due to cancer. The authors hypothesize that aspirin’s protective effect is mediated through platelets, which are thought to play a facilitating role in hematogenous cancer metastases. Study limitations include the 20% of cancers identified during the study where the status of metastatic disease was uncertain. One may also wonder whether patients on aspirin were more likely to have cancer detected earlier. For example, might aspirin have predisposed to bleeding events and thus diagnostic imaging which could have uncovered cancer earlier?
HAART has extended the lives of millions of patients living with HIV/AIDS. The recent iPrEx study suggests that prophylactic pre-exposure HAART with daily tenofovir-emtricitabine (PrEP) may reduce the risk of HIV transmission in adherent high-risk MSM populations by up to 73%.[6] Many questions about PrEP remain, however, including those concerning the cost-effectiveness of this intervention and its potential for encouraging high risk behavior among MSM. A recent study in Annals by Jussola uses a complex mathematical model to address the cost-effectiveness of PrEP.[7] The authors adapted a deterministic dynamic compartmental model of HIV transmission and progression. Although the mathematical details of the model are beyond our scope, model parameters included HIV prevalence, assumptions about sexual behavior and different risk populations within the MSM population, condom use, screening, infectivity based on treatment or lack of treatment, effectiveness of PrEP, costs of testing, counseling and treatment, and costs of HIV diagnosis, ART and PrEP, among others. The authors concluded that initiating PrEP in 20% of the undifferentiated-risk MSM population would save 60 thousand infections at 20 years of follow up, and would cost $172,091 per quality adjusted life year (QALY) gained. This was felt not cost effective. However, imitating PrEP in the highest risk 20% of the MSM would cost about $50,000 per quality adjusted life year (QALY), which compares favorably to other prophylactic interventions. However, the total cost would still be nearly $4 billion per year, which is high, particularly in view of competing interventions for HIV/AIDS prevention and treatment. As the authors note, the effectiveness of PrEP for high risk MSM will depend on clinician’s ability to reach those most at risk, and the adherence of this population with PrEP, which may be difficult, as this group tends to have limited contact with the medical system, a high incidence of substance abuse, and is most subject to stigma. The authors note that if PrEP is shown to be effective in pre-exposure use only (as compared with daily use), or if the cost of HAART declines, then cost effectiveness will improve further.
Many physiologic processes have effects on the pathology of multiple organ systems. The case of aspirin provides useful insight into this observation, where platelets clearly play a role in thrombosis and cardiovascular events, but may also play a role in cancer metastasis. In all the cases described above, the promise of known therapies for unknown uses requires further insight into the mechanisms of action of the interventions themselves.
Dr. Matthew Ingham is a 2nd year resident at NYU Langone Medical Center
Peer reviewed by Danise Schiliro, MD, contributing editor, Clinical Correlations
Image courtesy of Wikimedia Commons
References:
1. Dixon JB. Adjustable gastric banding and conventional therapy for type 2 diabetes: a randomized controlled trial. JAMA 2008: 299; 316-323. http://jama.ama-assn.org/content/299/3/316.short
2. Schauer, PR. Bariatric Surgery versus Intensive Medical Therapy in Obese Patients with Diabetes. NEJM. 366: 17. 1567-1576. http://www.nejm.org/doi/full/10.1056/NEJMoa1200225
3. Mingrone G. Bariatric Surgery versus Conventional Medical Therapy for Type 2 Diabetes. NEJM. 366: 17. 1577-1585. http://www.nejm.org/doi/full/10.1056/NEJMoa1200111
4. Perini EC. Effect of Transendocardial Delivery of Autologous Bone Marrow Mononuclear Cells on Functional Capacity, Left Ventricular Function, and Perfusion in Chronic Heart Failure. JAMA. 307: 16. 1717-1726. http://jama.ama-assn.org/content/early/2012/03/21/jama.2012.418.abstract
5. Rothwell P. Effect of daily aspirin on risk of cancer metastasis: a study of incident cancers during randomized controlled trials. Lancet. 379: 1591-601. http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(12)60209-8/abstract
6. Grant RM. Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. NEJM. 363: 2587-99. http://www.nejm.org/doi/full/10.1056/NEJMoa1011205
7. Jussola J. The cost-effectiveness of pre-exposure prophylaxis for HIV prevention in the United States in men who have sex with men. Annals of Internal Medicine. 156. 541-550. http://www.annals.org/content/156/8/541.abstract