Viridans Group Streptococcal Bacteremia after Allogeneic Stem Cell Transplant with Post-Transplant Cyclophosphamide

BACKGROUND: Despite antibacterial prophylaxis, Viridans group streptococcal (VGS) bacteremia is well described during neutropenia after allogeneic hematopoietic stem cell transplant (alloHSCT). Post-transplant cyclophosphamide (PTCy) has been used at our center since 2017 as a graft-versus-host disease prevention strategy. Data are lacking regarding VGS bacteremia in PTCy alloHSCT.

METHODS: We conducted a retrospective chart review of VGS bacteremia incidences in 557 patients who underwent alloHSCT (excluding cord blood transplants (CBT)) at our center from Jan 2014 to May 2022. We assessed the incidence of VGS bacteremia at day 30, treatment-related mortality (TRM) at day 100, antibiotic sensitivity, and VGS bacteremia by prophylaxis. Logistic regression analysis was applied to identify the factors associated with VGS bacteremia.

RESULTS: One-hundred-fifty patients received PTCy alloHSCT, and 407 patients non-PTCy alloHSCT (non-PTCy) (Table). Patients who received PTCy were older (median age: 60 vs. 57 years, P<0.001) and more likely to be in non-remission (41.3 vs. 29%, P<0.01). PTCy alloHSCT included haploidentical (n=59), matched unrelated donor (MUD, n=56), mismatched unrelated donor (mMUD, n=31), matched related (n=3), and mismatched related HSCT (n=1), while non-PTCy alloHSCT included MUD (n=257), matched related donor (n=112), mMUD (n=29), mismatched related (n=8), and haploidentical transplant (n=1). Most patients received ciprofloxacin (n=356) as antibacterial prophylaxis, followed by levofloxacin (n=86) and others (n=115). PTCy alloHSCT had a longer duration of neutropenia (<500/µL) than non-PTCy (median 18 vs. 13 days, p<0.001).

Twenty-one patients developed VGS bacteremia (median 11 days, range 6-22): 16 in those who underwent PTCy alloHSCT and 5 in non-PTCy. The cumulative incidence of VGS bacteremia by day 30 post-alloHSCT was increased in PTCy than in non-PTCy (10.7% v. 1.2%, p<0.0001, Figure). The median day of onset of bacteremia was 11.5 (9-22) for PTCy and 7 (6-10) for non-PTCy. Multivariate analysis showed the use of PTCy and the incidence of neutrophil recovery >21 days increased the probability of VGS bacteremia (odds ratio 4.2, p<0.05 and 4.7, p<0.01, respectively). There was no significant increase in VGS bacteremia when assessing age, disease status, donor type, and antibacterial prophylaxis.

Since levofloxacin provides better VGS activity than ciprofloxacin, we compared the incidence of VGS bacteremia by prophylaxis in PTCy alloHSCT. The use of levofloxacin decreased the incidence of VGS (15.1% (n=73) v. 4.4% (n=45), p=0.08).

Among PTCy alloHSCT patients, those that experienced VGS bacteremia showed an increase in TRM compared to those without VGS bacteremia (19.3 vs. 1.5%, p<0.001). No TRM occurred in PTCy alloHSCT patients who received levofloxacin and developed VGS breakthrough bacteremia. Penicillin minimum inhibitory concentrations at the time of VGS bacteremia showed no impact on TRM (16.7% in intermediate/resistance (n=12) vs. 10.1% in susceptible (n=11), p=0.6).

CONCLUSIONS: To our knowledge, this is the first report of an association between PTCy and VGS bacteremia. A similar rate of VGS bacteremia has been reported in a cohort of alloHSCTs, but most cases were CBT, with prolonged post-HSCT neutropenia. Similarly, PTCy may increase the risk of VGS bacteremia by prolongation of neutropenia. However, even when controlling for this, PTCy use was significant. PTCy is also known to cause severe mucositis, which also may be an important contributor, as VGS is a common oral flora. We were unable to analyze the incidence of mucositis in our cohort.

Serious complications of VGS bacteremia in neutropenic patients (i.e., septic shock and adult respiratory distress syndrome) have been described with mortality rates of 6% to 30%. VGS-associated mortality was also seen in our cohort, suggesting that VGS breakthrough in the early period of alloHSCT warrants careful surveillance with interventions to combat morbidity and mortality. Based on the known antimicrobial spectrum of levofloxacin and our data, prophylaxis with levofloxacin (over other antibacterials) appears to be a better choice in PTCy alloHSCT patients to prevent VGS bacteremia. Further studies are necessary to optimize antibacterial prophylaxis.

Rakszawski:SeaGen: Consultancy, Speakers Bureau. Rybka:Merck: Research Funding; Spark Therapeutics: Membership on an entity's Board of Directors or advisory committees; Spark Therapeutics: Consultancy; Penn State Health: Current Employment. Claxton:Astellas: Research Funding.