Weekly Articles 4/11/2021

Disappointing HIV vaccine news, more MDR TB, and a possible HITting complication for one of the new COVID vaccines.

Gray GE, Bekker LG, Laher F, Malahleha M, Allen M, Moodie Z, Grunenberg N, Huang Y, Grove D, Prigmore B, Kee JJ, Benkeser D, Hural J, Innes C, Lazarus E, Meintjes G, Naicker N, Kalonji D, Nchabeleng M, Sebe M, Singh N, Kotze P, Kassim S, Dubula T, Naicker V, Brumskine W, Ncayiya CN, Ward AM, Garrett N, Kistnasami G, Gaffoor Z, Selepe P, Makhoba PB, Mathebula MP, Mda P, Adonis T, Mapetla KS, Modibedi B, Philip T, Kobane G, Bentley C, Ramirez S, Takuva S, Jones M, Sikhosana M, Atujuna M, Andrasik M, Hejazi NS, Puren A, Wiesner L, Phogat S, Diaz Granados C, Koutsoukos M, Van Der Meeren O, Barnett SW, Kanesa-Thasan N, Kublin JG, McElrath MJ, Gilbert PB, Janes H, Corey L; HVTN 702 Study Team. Vaccine Efficacy of ALVAC-HIV and Bivalent Subtype C gp120-MF59 in Adults. N Engl J Med. 2021 Mar 25;384(12):1089-1100. doi: 10.1056/NEJMoa2031499. PMID: 33761206; PMCID: PMC7888373.

This is a randomized, double-blind, placebo controlled trial involving 5407 patients between the ages of 18-35. It was designed to evaluate the efficacy of a recombinant canarypox vector containing subtype C envelope ALVAC-HIV and an MF59-adjuvanted subtype C bivalent glycoprotein 120 vaccine in the prevention of HIV 24 months after enrollment. Injections were done at months 0 and 1, and then at months 3, 6, 12, and 18. 138 infections occurred in the vaccine group (incidence rate of 3.4 per 100-person yeast, 95% CI 2.8 to 4.0) while 133 occurred in the placebo group (3.3 per 100 person-years, 95% CI 2.8 to 3.9) for a HR of 1.02 (95% CI 2.8 to 3.9). 

When stratified by sex and age, there was no statistical difference in the incidence of infection between either group:

There was also no differences in viral load, with mean log 10 viral load being 4.82 in the vaccine group compared to 4.64 in the placebo group. 

One of the things to note is that a high percentage of participants were women, which was by design (their goal was to get anywhere from 60-75% of participants being women), with women having a higher rate of HIV infection compared to men. This, along with the relatively high risk of STI diagnosis, may have contributed to the lack of efficacy. The authors also note the genetic diversity of the subtype C epidemic in sub-Saharan Africa as being an issue. Either way, quite disappointed in the results. 

DeSimone DC, Lahr BD, Anavekar NS, Sohail MR, Tleyjeh IM, Wilson WR, Baddour LM. Temporal Trends of Infective Endocarditis in Olmsted County, Minnesota, Between 1970 and 2018: A Population-Based Analysis. Open Forum Infect Dis. 2021 Jan 27;8(3):ofab038. doi: 10.1093/ofid/ofab038. PMID: 33728357; PMCID: PMC7944350.

This is a cohort study of patients with infective endocarditis in Olmsted County, Minnesota that spanned 1/1/1970 – 12/31/2018 that sought to examine the IE incidence. A subset analysis spanning the period of 1996-2018 was done to examine the temporal trends in IE due to viridans group streptococci to account for the AHA prophylaxis guidelines. 269 cases of endocarditis were identified, with the overall incidence (cases per 100,000 person years) being 7.9% (95% CI 7-8.9). A higher rate was seen in men (12.1) compared to women (4.5), and staph aureus (2.4, 95% CI 1.9-2.9) and VGS (0.9, 95% CI 0.6-1.20) were the most common bacteria here. Other interesting findings included an increase in the incidence of IE with increasing age (aIR 3.99, 95% CI 3.30-4.82), as well as increasing calendar year (aIR 1.39, 95% CI 1.03-1.86). Of the organisms listed, enterococcus was the one that increased in incidence (aIR 2.9, 95% CI 1.53-5.50). When looking at seasonality, there is a higher rate of IE in spring and winter, but the rate has been increasing rapidly during winter time in the past few years:

In evaluating the 48 cases of VGS endocarditis, the application of the 2007 guidelines had no impact on the incidence of VSG IE. 

I think this is an interesting study, but does not really tell the whole story in terms of IE incidence and epidemiology, especially when looking only at Olmsted county. From my (limited) experience, it seems, though, that we mostly see Staph aureus and enterococci endocarditis, which seems to check out here. Would be interested in seeing a more comprehensive epidemiological study. Speaking of endocarditis…

Anna Bläckberg, Linn Falk, Karl Oldberg, Lars Olaison, Magnus Rasmussen, Infective Endocarditis Due to Corynebacterium Species: Clinical Features and Antibiotic Resistance, Open Forum Infectious Diseases, Volume 8, Issue 3, March 2021, ofab055, https://doi.org/10.1093/ofid/ofab055

This is a Swedish retrospective study evaluating cases of IE due to corynebacterium over a 10 year period and comparing them to Staph aureus, E faecalis, and Streptococci. A total of 4246 cases were included, of which 30 were due to corynebacterium, with the most common being C. striatum:

Patients with corynebacterium tended to be have prosthetic valve endocarditis (70%) and have a low risk of embolization (10%), though half required surgery and mortality was higher than that of enterococcus IE:

PVE due to corynebacteria, when compared to other causes of PVE, had a lower mortality rate and higher rate of surgery:

Most isolates were resistant to penicillin G (14/20), with 4/8 being resistant to rifampin, and all being sensitive to vancomycin (MIC ranging from 0.5 to 2). I have heard IE due to corynebacterium tends to involve prosthetic material, which is the case, and similar to Staph lugdunensis. I think it is helpful in that it may make us think twice about ignoring a Corynebacterium spp bacteremia in someone with a cardiac device. 

Okoli GN, Racovitan F, Abdulwahid T, Hyder SK, Lansbury L, Righolt CH, Mahmud SM, Nguyen-Van-Tam JS. Decline in Seasonal Influenza Vaccine Effectiveness With Vaccination Program Maturation: A Systematic Review and Meta-analysis. Open Forum Infect Dis. 2021 Feb 5;8(3):ofab069. doi: 10.1093/ofid/ofab069. PMID: 33738320; PMCID: PMC7953658.

This is a meta-analysis that evaluated the impact of repeat influenza vaccination on vaccine program effectiveness in outpatient settings after the 2009/2010 influenza pandemic. The vaccine program maturation was used as a proxy for population level repeated vaccination. Vaccine effectiveness was assessed against influenza A (H1N1)pdm09 and A(H3N2), and influenza B. Primary outcome was pooled influenza VE against all influenza across categories of vaccination program maturity. The program maturity (PM) is the number of years since the program inception, and these were grouped into categories, with Q2, Q3, and Q4. 72 studies were included. Overall, there was a trend towards lower vaccine effectiveness the longer the program was around, which occurred even among antigenically similar vaccines:

I have written in the past about the “original antigenic sin” and how subsequent vaccination may not necessarily prime the immune system against the most recent strain of a virus, but rather reinforce earlier antibodies already generated in the body. Whether this is what is going on remains to be seen, though I would not discourage obtaining the flu vaccine.

Zheng H, He W, Jiao W, Xia H, Sun L, Wang S, Xiao J, Ou X, Zhao Y, Shen A. Molecular characterization of multidrug-resistant tuberculosis against levofloxacin, moxifloxacin, bedaquiline, linezolid, clofazimine, and delamanid in southwest of China. BMC Infect Dis. 2021 Apr 8;21(1):330. doi: 10.1186/s12879-021-06024-8. PMID: 33832459; PMCID: PMC8028109.

This was a cohort study out of China evaluating the susceptibility of MDR/RR TB for the higher generation quinolones, linezolid, bedaquiline, clofazimine, and delamanid. 339 strains were isolated, of which 88 were MDR. The MICs for the tested drugs are highlighted below:

The highest percentage of resistant strains were found in LFX/MXF, while the lowest was found in bedaquiline. Given that the quinolones have been out longer than bedaquiline, this is not surprising, but it is concerning there is a significant amount of resistance already brewing. With the 2019 WHO guidelines out advocating possibly using an all PO regimen that includes bedaquiline and quinolones along with linezolid, this may have us taking a step backwards towards therapy though the resistance for clofazimine and delamanid was still fairly low. 

Nestor, D., Andersson, H., Kihlberg, P. et al. Early prediction of blood stream infection in a prospectively collected cohort. BMC Infect Dis 21, 316 (2021). https://doi.org/10.1186/s12879-021-05990-3

This is a prospective cohort study of adult patients who presented to the ED and were found to be bacteremic. The purpose of this cohort was to see if the neutrophil/lymphocyte ratio (NLR) and/or the (not Ben) Shapiro score could be used to predict bacteremia in a cohort of patients presenting to the ED:

The recommended cutoff for the Shapiro score is >3, while for NLR is >12. 486 patients were included, of which 99 had a positive blood culture. 84 of these were clinically relevant. Patients with positive blood culture were more likely to have SIRS criteria (83% vs 64%), had a median Shapiro score of 3 (compared to 2), and had higher NLR (16 vs 8) as well as lactate (2.1 vs 1.6). The sensitivities and specificities are highlighted below, with both performing similarly:

Notably, combining both increased the specificity at the risk of sensitivity, meaning this was an effective way to rule out bacteremia, as 87% of patients with bacteremia had either a high Shapiro score or a high NLR. One of the interesting things is that gram negative infections tended to have higher NLR than gram positives:

I think the addition of NLR to a score prediction tool makes sense, though the Shaprio score accounts for risk factors for bacteremia. I do not think it is unreasonable to use these, though clinical gestalt + risk factors + NLR would be as good as calculating the Shapiro score. 

Hwang H, Kang H, Kwon YS, Jeon D, Shim TS, Yim JJ. Outcomes of multidrug-resistant tuberculosis treated with bedaquiline or delamanid. Clin Infect Dis. 2021 Apr 10:ciab304. doi: 10.1093/cid/ciab304. Epub ahead of print. PMID: 33837767.

This is a retrospective nationwide cohort study describing XDR or pre-XDR patients who received either Bedaquiline or Delamanid after request as part of their multidrug regimen in Korea. Patients were evaluated by a TB committee that decided who would get the drugs on a case-by-case basis. Primary outcome was treatment success.260 patients were included, with the bedaquiline group being more likely to have chronic renal or liver disease and cavitary lesions compared to the delamanid group. Negative sputum culture conversion was achieved in 111/120 (92.5%) of those who had positive sputum culture at baseline, which was not different between groups (93.3% vs 92% for bedaquiline and delamanid, respectively). The time to culture conversion was not statistically different either (45 days vs 54 days). Of the 260 patients, treatment success was 79.2%, with no statistical difference between the bedaquiline or delamanid groups (53 vs 68.8, p=0.327) 

Both drugs were safely tolerated, with no differences in total adverse events (15% in bedaquiline vs 17.7% in delamanid), and no difference in QTc changes in either, though 7 patients stopped delamanid compared to none of the bedaquiline group. It should be noted that  more patients in the delamanid group were treated with FQs (64% vs 52) and had a higher number of drugs administered during the study period (median 6 vs 5), though I do not think this affected the results as much. I think this is further evidence that both these drugs, when administered over 24 weeks, can have a positive impact on MDR TB and allows the use of an all-oral regimen that bypasses the need for a second-line injectable drug. I would like to have seen both drugs used in combination to have that data, since there is a rise in FQ resistance as seen above.

Corsini Campioli C, Go JR, Abu Saleh O, Challener D, Yetmar Z, Osmon DR. Antistaphylococcal Penicillin vs Cefazolin for the Treatment of Methicillin-Susceptible Staphylococcus aureus Spinal Epidural Abscesses. Open Forum Infect Dis. 2021 Feb 16;8(3):ofab071. doi: 10.1093/ofid/ofab071. PMID: 33738321; PMCID: PMC7953649.

Nafcillin vs cefazolin. The ultimate question when it comes to MSSA infections. This was a single-center retrospective study of adult patients with MSSA spinal epidural abscess that compared patients who got an antistaphylococcal penicillin (nafcillin or oxacillin) with cefazolin. 79 cases were included, 45 of which got cefazolin while 34 got either nafcillin/oxacillin. There were some differences in the groups, with cefazolin patients having a higher prevalence of steroid use, hypertension, and malignancy and ASPs having a higher prevalence of peptic ulcer disease, diabetes, and PVD. Further, the ASP group had a higher percentage of patients admitted to the ICU (29.4% vs 11%). All but one patient underwent aspiration or surgical debridement, and 26 patients had either OM or diskitis. There was no statistical difference in hospital length of stay, antibiotic failure at week 6 or 12, or mortality at 30 days or 90 days, though the median time to death was longer in the cefazolin group:

I’ll say it again; the concern with cefazolin is the inoculum effect, where a high density of bacteria leads to activation of certain blaA/C genes that de-activate the beta-lactam component of cefazolin. One tip off here is if MSSA is resistant to clindamycin, which is a surrogate marker for this gene. The literature recently has suggested that in the vast majority of cases, this is not a problem, even with serious infections such as IE and osteomyelitis. While retrospective and single center, I think this adds a bit to that literature. 

Sara Belga, Clayton MacDonald, Diana Chiang, Dima Kabbani, Soroush Shojai, Juan G Abraldes, Carlos Cervera, Donor Graft Cytomegalovirus Serostatus and the Risk of Arterial and Venous Thrombotic Events in Seronegative Recipients After Non-Thoracic Solid Organ Transplantation, Clinical Infectious Diseases, Volume 72, Issue 5, 1 March 2021, Pages 845–852, https://doi.org/10.1093/cid/ciaa125

This is a retrospective study of patients transplanted in a university hospital in Canada that evaluate the effect of CMV exposure at transplantation on post-transplant thromboembolic events in a CMV-negative populations. All recipients were CMV seronegative and they were divided based on the serostatus of the donor; either D+/R- (high risk) or D-/R- (low risk). Mis-match patients received valgan or ganciclovir prophylaxis for 3 months in liver/pancreas transplant and 6 months in multivisceral or kidney transplants. 392 solid organ transplant (SOTs) recipients were evaluated, of which 46% were CMV mismatches. Other than age (mismatch cohort was a bit oldre), both populations were fairly well balanced.  The incidence of post-transplant thromboembolic events was higher in the mismatch group compared to the seronegative group, 19% vs 10% (p=0.008), with similar findings occurring 30 days after surgery (to account for surgery related events), 18% vs 8% (p=0.002): 

19 patients in the mismatch group had CMV infection with 11 preceding the embolic event. In multivariate analysis, CMV mismatch, age, pre-transplant thromboembolic events, and liver transplant were associated with higher risk of embolic events at 5 years:

Furthermore, mortality was higher in the mismatch group at 60 months (89.6% vs 81.2% survival rates in seronegative and mismatch, respectively) with age adjusted regression analysis revealing that D+/R- status had an aHR of 1.71 (95% CI 1-2.9) for mortality. Post-transplant thromboembolic events had a negative impact on mortality, with an aHR of 2.6 (95% CI 1.5-4.6), suggesting a stronger effect, independent of CMV status. 

It has been known that CMV infection has negative impact on graft survival and overall mortality, and seems that thromboembolic events, due to chronic inflammation, may be a part of that trend. While this was mostly seen on the abdominal organs, I anticipate that heart and lung would have similar effects. I would also be interested if the type of immunosuppression/induction has any impact on these outcomes (sounds like this may make a good blog topic…). Of course, this also begs the question of whether mismatch patients should be prophylactically anticoagulated if they meet certain criteria. Too many questions, so little answers. 

Ranjbar, K., Moghadami, M., Mirahmadizadeh, A. et al. Methylprednisolone or dexamethasone, which one is superior corticosteroid in the treatment of hospitalized COVID-19 patients: a triple-blinded randomized controlled trial. BMC Infect Dis 21, 337 (2021). https://doi.org/10.1186/s12879-021-06045-3

I think this study proves my hypothesis that dexamethasone 6mg is probably not enough for a lot of COVID patients. This was a randomized controlled trial evaluating methylprednisone 2mg/kg with dexamethasone 6mg over a 10 day period. Patients were randomized in a 1:1 fashion with primary endpoint being all-cause mortality at 28 days and clinical status after 5 and 10 days by the 9-point WHO ordinal scale. 86 patients were enrolled, with the methylprednisolone group showing a significant reduction in the 9 point ordinal scale by day 10:

Further, mortality in the dexa group was 37.5% compared to 18.6% in the methylprednisolone group, though this did not reach statistical significance (p=0.076).  Hospital length of stay was also shorter in the methylpred group, 7.43 d vs 10.52 (p=0.015) as was need for mechanical ventilation (18.2% in the methylpred group vs 38.1% in the dexa group). While at first glance this makes methylprednisone a better choice, keep in mind the doses were high. In a 70kg person, this would equate to 140mg of methylpred, compared to the dexa equivalent of methylpred which was 32mg. I think dosing was skewing the results, big time. If anything, this would suggest that in those who are sicker, using a higher dose of dexamethasone may not be unreasonable. 

Greinacher A, Thiele T, Warkentin TE, Weisser K, Kyrle PA, Eichinger S. Thrombotic Thrombocytopenia after ChAdOx1 nCov-19 Vaccination. N Engl J Med. 2021 Apr 9. doi: 10.1056/NEJMoa2104840. Epub ahead of print. PMID: 33835769.

This is a case series of 11 patients who had recently received the Astra-Zeneca vaccine. All patients had thrombocytopenia and venous thrombosis, with 9/11 having cerebral venous thrombosis and 5 having signs of DIC. All 7 patients who were tested for PF4-dependent platelet activation assay were positive:

Platelet activation in the first 4 patients with PF4 positivity demonstrated weak-to-moderate reactivity, which was inhibited by heparin (the shorter the reaction time, the stronger the platelet-activating levels):

This was also seen in a subset of 24 patients that tested positive for PF4-heparin ELISA, with most being inhibited by LMWH, suggesting a similar pathophysiology to that of HIT

This was also seen in the ELISA results of all 28 patients which showed inhibition with heparin exposure:

The clinical features resemble that of HIT, with serum showing strong reactivity on the PFT-heparin ELISA and inhibition when exposed to heparin, which is suggestive of this. It is unclear the trigger for this, as things other than heparin can kickstart this cascade. Adenovirus was noted to be able to bind and activate platelets, however the amount in the vaccine would not be enough to trigger such a response. Despite this, something is triggering the activation of the PF4-platelet complex and leading to this presentation, hopefully we will get more info in the coming weeks.