Rifaximin for sickle cell disease

Patients with sickle cell disease (SCD) suffer recurrent painful vaso-occlusive crisis (VOC), poor quality of life and a shortened lifespan. Although hydroxyurea and blood transfusion remain the mainstay of SCD therapy, the only cure currently available is allogeneic stem cell transplant (allo-SCT). However, many barriers prevent allo-SCT from being widely offered to adult SCD patients. The SCD patients most in need of allo-SCT transplant are also those at highest risk for transplant-related complications and mortality.1 Other therapeutic approaches are, therefore, much needed. Neutrophils have been implicated in regulating the disease course of VOC. The SCD patients with WBC > 15 × 109/L are more likely to develop stroke,2 acute chest syndrome,3 and premature death.4 Neutrophils in SCD also exhibit higher levels of activation molecules, for example, CD645 and CD11b/CD18,6 and their sera have elevated soluble CD62L.5, 7 Circulating aged neutrophils (CANs), a subset of neutrophils with high surface expression of CXCR4 and low CD62L, are also significantly elevated.7, 8 Both activated and aged neutrophils may be immobilized on endothelium, and form the nidus for sickled erythrocytes to adhere to, eliciting VOC. Several recent observations from our laboratory and others have supported frequent microbial and pathophysiologic changes in the intestines of SCD patients. These include enterocyte injury,7 altered microbial composition,9 increased permeability7 and bacterial overgrowth.10 We have previously proposed that increased translocation of intestinal bacteria/bacterial products into the systemic circulation is responsible for the increased number of activated neutrophils in SCD.11 Others have found intestinal microbiota to regulate CANs in SCD mice.8 Modulating intestinal microbial composition may, therefore, be a therapeutic option in SCD to reduce VOC, by decreasing both activated and aged neutrophils. We recently reported that rifaximin treatment reduced CANs in SCD patients treated in a clinical study (Clinicaltrials.gov: NCT03719729) funded by Bausch Health Companies, Inc.12 Rifaximin also shifted the intestinal microbiome in these patients towards increased abundance of Bacteroides.12 In this current paper, we report the clinical outcome on the same cohort of SCD patients accrued to the study, to determine whether the changes in the intestinal microbiome and reduction in the CANs were translated into clinical benefits of reduction in VOC. This study was approved by the Institutional Review Board of New York Medical College. The SCD patients were eligible for the study if they were 18 years or over, had a diagnosis of SCD (HbSS, HbSC, HbSβ-thalassemia), and had at least two painful crises needing inpatient or outpatient intravenous opioid analgesia (IOA), during the preceding 12 months. Patients may continue on the hydroxyurea or L-glutamine if they were already taking the medication long-term. Patients were only added to the study if their last episode of painful crisis was more than 4 weeks before starting rifaximin. Each patient received 550 mg of rifaximin twice a day for 6 months. The dose of rifaximin chosen was similar to that for preventing hepatic encephalopathy in patients with liver cirrhosis. Each patient was also given a self-reporting diary to record daily ingestion of the medication and associated side effects. All patients were followed up every 4 weeks, or earlier if they developed VOC. Each patient completed the FANLTC (Functional Analysis of Non-life-Threatening Conditions) questionnaire before, and 3 and 6 months after being started on rifaximin. The FANLTC questionnaire consists of four subjective blocks of questions covering physical, emotional, social/family, and functional well-being. Each question scores as follow: 0 = not at all; 1 = a little bit; 2 = somewhat; 3 = quite a bit; 4 = very much. Efficacy analyses were performed on data obtained from all the patients who completed 6 months of rifaximin. Changes in the number of VOC and days needing IOA during the study period, were compared to those expected for a six-month period, calculated from the average of the previous 12 months. Thirteen adult patients with SCD consented to the study. Their clinical characteristics are shown in Table S1. There were eight males and five females. Ten patients had HbSS, two HbSβ0thal, and one HbSC. Their median age was 29 years (range 24-56). The median number of VOC in the previous 12 months was 4.5 (range 2-13), and the number of days needing IOA was 25.5 days (range 8-198). In 12 of the 13 patients, the VOC was in the form of musculo-skeletal pain typical in distribution for the VOC that normally affected individual patients, and in one patient the VOC was predominantly in the form of recurrent priapism. Five of the patients were already taking hydroxyurea, and none were taking L-glutamine or on a transfusion program. Six patients were taking oral opioid at home for chronic pain. The study ran between November 2018 and early August 2019, covering the winter months when most of the patients reported having most frequent VOC. In all 12 evaluable patients, rifaximin was started during the winter months. Twelve of the 13 patients accrued to the study were evaluable for response. One patient (HbSC) did not return to be started on the rifaximin after consenting for the study. The remaining 12 patients completed 6 months of rifaximin. The mean self-reported medication compliance of the intended doses was 86% (range 50-100). The median number of VOC during the study period decreased from the expected 2.25 (range 1-6.5) per 6 months to one per 6 months (range 0-4) (P = .003) (Figure 1A). The one patient whose VOC was predominantly in the form of priapism did not respond. His self-reported medication compliance was 100%. The median decrease in days needing IOA over the six-month study-period was nine (range 1-55), and the median percent reduction was 82% (range 20%-100%) (P = .008) (Figure 1B). Total number of days needing IOA was reduced from the expected 254 in 6 months to 98 during the study period. Laboratory analyses did not show any significant change in hemoglobin, white cell counts, serum lactate dehydrogenase, bilirubin, or haptoglobin due to rifaximin therapy. There was no significant change in the FANLTC scores in all four well-being blocks of questions when the patients were analyzed as a group. Subset analysis, however, found that functional well-being significantly improved at 6 months, for those who reported below average scores before rifaximin (mean score 11.4+/−1.7 to 14.6+/−2.19) (P = .02). The average score to the question “I am able to enjoy life” increased from 1.4+/−0.5 to 2.1+/−0.69 (P = .046). There was also improvement among those with below average scores for physical well-being. The average score to the question “I have pain” decreased from 3.56+/−0.72 to 2.3+/−1.6 (P = .022) at month three, and from 3.56+/−0.72 to 2.22+/−1.4 (P = .016) at month six. The average score to the question “I feel ill” also improved, decreased from 2.5+/−0.58 to 0.75+/−0.5 (P = .035) at months three and six. Rifaximin was well tolerated. Mild nausea was common in the first 1-2 weeks of rifaximin therapy and occurred in eight of the 12 patients. Other adverse events included self-limiting diarrhea (n = 1) (negative for pathogens by culture, Clostridium difficile by PCR, and gastrointestinal multiplex PCR) which occurred during the first month of therapy, and polydipsia (n = 2). The results of this study, therefore, suggest that rifaximin may potentially benefit patients, as measured by a decrease in the number of VOCs and days needing IOA, in patients with SCD. These clinical benefits were translated into improvement in the QOL, as measured by FANLTC, for those who experienced below average quality of life prior to the study. Reducing the intestinal microbial load using antibiotics has been found beneficial in SCD mice. Mice with SCD acquired by transplantation when treated with an oral combination of ampicillin, neomycin, vancomycin and metronidazole exhibited lower number of CANs, and were protected against tumor necrosis factor-α-induced fatal VOC.8 The identical cocktail of oral antibiotics, when given to the Townes SCD mice reduced intestinal microbial load and SCD-related osteopenia.10 Although the beneficial effects of the combination antibiotics on SCD are likely related to reduction in the intestinal microbial density, because two of the four antibiotics are absorbed into the systemic circulation, any systemic effects of the antibiotics in modifying SCD complications cannot be totally ruled out. Furthermore, while the findings in these two studies were interesting, the approach is clearly not directly translational. It is highly impractical to administer the combination of the four antibiotics long-term to SCD patients. In our study, more than 90% of those consented remained in the study for 6 months. This high retention rate is likely due to two factors. Besides telephone calls to the patients, until the patients answered the call, on the morning of their follow-up appointments, text messages were also sent to the patients. The clinical benefits, in terms of less frequent painful VOC and improved QOL, experienced by the patients may also contribute to the high patient retention rate. We have chosen to test rifaximin in this study for three reasons. First, rifaximin has an established safety profile for long-term use in patients with liver cirrhosis to prevent hepatic encephalopathy.13 Second, the risks for the development of antibiotic-associated Clostridium difficile infection induced by long-term use of antibiotics is likely low, since rifaximin also has activities against Clostridium difficile. Third, being a minimally absorbed antibiotic, rifaximin provided the opportunity to dissect the effect of reducing intestinal microbial load in SCD, from any systemic effect an antibiotic might also have on the disease process. The results of our study using rifaximin, therefore, support the local effects of an antibiotics in the intestine in modifying the course in SCD. The mechanisms responsible for the beneficial effects of rifaximin in reducing VOC remains to be determined. It is likely that rifaximin works via decreasing the intestinal microbial load, and hence reducing translocation of intestinal bacteria/bacterial products into the systemic circulation to activate and age neutrophils, as reported by us previously that CANs in these individuals dropped significantly while taking the rifaximin during the study period.12 Since rifaximin has an anti-inflammatory property,14 the role contributed by a reduction in the inflammatory process associated with enterocyte injury7 and bacteria/bacterial products translocated across the intestinal barrier as a result of improvement in the integrity of the intestinal barrier cannot be totally ruled out. In summary, this is the first human study showing the potential efficacy of microbial modulation on VOC in patients with SCD. Intestinal microbial modulation with rifaximin is also safe and well-tolerated. The major limitations in our work is the small sample size and the single-center single-arm nature of the study. Furthermore, any placebo effects of the intervention on the QOL of the patients cannot be totally ruled out. The intense follow up of the patients may also be responsible in part for the observed clinical improvement. A multicenter placebo-controlled randomized trial is, therefore, warranted to confirm our findings. We would like to thank the following for their critiques and advice during the preparation of the manuscript: Dr. Peter Gillette at State University of New York Downstate Medical School and King's County Medical Center, Dr. Howard Franklin at Salix Pharmaceuticals, and Dr. Robert Israel at Bausch Health Companies, Inc. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.