Transfusions, Necrotizing Enterocolitis (NEC), Cytomegalovirus (CMV) and more.. Oh my. Updates on Neonatal Transfusion Medicine

Transcription

1 Transfusions, Necrotizing Enterocolitis (NEC), Cytomegalovirus (CMV) and more.. Oh my Updates on Neonatal Transfusion Medicine Cassandra D. Josephson, MD Professor, Pathology and Pediatrics Emory University School of Medicine Medical Director, CHOA Blood, Tissue, and Apheresis Services Disclosures Immucor, LLC consultant Octapharma consultant Biomet Zimmer consultant Sysmex unrestricted research grant funding Medtronics unrestricted research grant funding 1

2 Objective Upon completion of this presentation, participants should be able to: Examine risk estimates of transfusion transmission of cytomegalovirus (TT CMV), review strategies to prevent TT CMV, discuss necrotizing enterocolitis (NEC) in the context of CMV, anemia, and red blood cell (RBC) transfusion in very low birth weight (VLBW) infants. Overview Examine transfusion rates in VLBW infants Frequency of RBC, platelet, plasma transfusions Timing of transfusion Discuss TT CMV birth cohort study Examine TT CMV and maternal breast milk acquired CMV risk estimate results Review risk reduction strategies for TT CMV, leukoreduction and CMV seronegative blood Examine findings of NEC in the context of CMV infection Review the association between anemia and NEC Explore RBC transfusion as surrogate vs potential risk factor NEC Review potential mechanisms for association of anemia, RBC transfusion, and NEC 2

3 Examine transfusion rates in VLBW infants 5 Transfusion rates in VLBW infants VLBW infants are a highly transfused population with ~ 57 67% of infants receiving RBC transfusion. Limited data on the rates of blood product utilization and distribution over time in this population exists. Estimated rates and timing of blood product utilization may guide the design of transfusion studies. Data may inform individualized discussions with parents on the probability of their VLBW infant receiving a blood product transfusion. Keir et al. Transfusion Ekhaguere et al. Pediatr Res

4 Transfusion Rates Study: Methods for TT-CMV VLBW Infant Birth Cohort Setting: Secondary analysis of a prospective, multicenter observational cohort study from in 5 hospitals (private and academic) in Atlanta, GA. Sample: Infants, BW 1500 g, not transfused before admission were included, followed until 90 d, discharge, or death. Prospectively Recorded: products transfused and timing Blood product utilization rates, with 95% confidence intervals, were estimated per 100 days, for RBC, FFP, PLT, CRYO transfusions. Centers used CPDA 1, leukoreduced RBCs and apheresis platelets. No standard transfusion protocol across all sites. Patel et al. AABB 2018 National Meeting Results Overall transfusion rates 598 VLBW infants, 2223 blood product transfusions were recorded during 37,369 infant days of follow-up. Any blood product transfusion occurred in 58% of infants. Blood product Frequency of transfusion, n (%) Transfusion rate per 100 infant days [95% CI] RBC 330 (55%) 4.43 [4.22, 4.65] PLT 108 (18%) 1.10 [1.00, 1.21] FFP 52 (9%) 0.40 [0.34, 0.47] Cryo 7 (1%) 0.02 [0.01, 0.04] Any 344 (58%) 5.95 [5.71, 6.20] For full cohort (n=598) 4

5 Results Gestational age-specific rates Gestational age Any product transfusion 8.4* 23 wk (n=13) 100% 24 wk (n=51) 100% 7.2* Incidence rate ratios 25 wk (n=61) 97% 26 wk (n=66) 80% 27 wk (n=84) 68% 28 wk (n=70) 49% 29 wk (n=98) 36% 4.3* 3.8* * P < * Ref 30 wk (n=63) 37% wk (n=92) 21% Patel et al. AABB 2018 National Meeting Results Birthweight-specific rates Birthweight Any product transfusion * P < <750 g (n=121) 750 to <1000 g (n=162) 98% 73% Incidence rate ratios [95% CI] 7.3* [ ] 3.2* [ ] >1000 g (n=315) 34% Ref Patel et al. AABB 2018 National Meeting 5

6 Distribution of transfusions over 28 d Patel et al. AABB 2018 National Meeting Transfusion (%) Distribution of transfusions over 90 d Transfusion (%) 1656 RBC Tx for 330 infants. 411 PLT Tx for 108 infants. 142 FFP Tx for 52 infants. 8 Cryo Tx for 7 infants. Patel et al. AABB 2018 National Meeting 6

7 Conclusions RBC and PLT transfusion rates decrease with increasing GA and BW, while no similar trends were observed for FFP or Cryo. A large proportion of blood product transfusions occur in the first postnatal week. PLT and FFP transfusions demonstrate a bimodal distribution with a second peak in the third week of life. Patel et al. AABB 2018 National Meeting Discuss TT-CMV Birth Cohort Study 1. Examine Study Results 2. Review Risk Reduction Strategies in for blood transfusion 7

8 Cytomegalovirus (CMV) CMV is a β herpesvirus, (HHV 5) Establishes a lifelong infection in the host. During the chronic phase of CMV infection the virus becomes latent and reactivates intermittently. Viral excretion during the chronic phase can be related to reactivation of latent virus (in monocytes) or superinfection with new CMV strains. Causes of reactivation: immunosuppression, pregnancy, allergies Prevention: handwashing to prevent spread Shedding of CMV a. Urine b. Blood c. Breast milk d. Saliva e. Cervical secretions f. Semen Huygens A, et al. Front Immunol. 2014;552. Beltran PMJ and Cristea IM. Expert Rev Proteomics, 2014;11(6):697 Jahan M. J Med Micro. 2010;4:39. Effects of CMV infection CMV can target multiple cell types Latent CMV harbors in monocytes (circulating in the blood stream) can be free floating virus in the plasma Congenital CMV infection can cause intracerebral calcifications chorioretinitis hepatitis neurodevelopmental impairment Plosa et al. Peds in Review hearing loss 8

9 Postnatal CMV infection Postnatal acquired CMV infection in VLBW infants can cause a sepsis like syndrome Most infants have no overt symptoms The two major sources of postnatal CMV infection (onset after 2 weeks of life) are: Breast milk Blood transfusions Source: CDC Risk and sources of postnatal CMV Josephson et al. JAMA Pediatrics

10 Study Flow Diagram Assessed for eligibility (n=1455) Met enrollment criteria (n=1102) Enrolled (n=541 infants) (n=462 mothers) Followed for 90 days or until hospital discharge or death (n=539) 3 infants transferred to non-study hospital n=415 n=499 n=2 Did not meet inclusion criteria: - birthweight (n=327) - age 5d (n=26) Met exclusion criteria: - Not expected to survive >7d (n=56) - Congenital anomaly (n=11) - Transfusion before screening (n=2) Declined consent Incorrectly enrolled: - did not meet enrollment criteria Josephson et al. JAMA Pediatrics 2014 Surveillance Testing for CMV infection Mothers N= % 462/ d 50% 55/ d d d 90 d or d/c For initially CMV seronegative mothers VLBW infants N= % 538/539 74% 379/511 75% 331/443 74% 224/301 93% 494/ % of infants evaluated for CMV by nucleic acid testing at least 2 times and 98% 3 or more times Josephson et al. JAMA Pediatrics

11 Testing for sources of CMV infection All Cellular Blood Products Leukoreduced CMV sero-negative 93% (882/954) units tested for CMV 92% (878/954) units tested for residual WBCs Maternal Birth Breast Milk 73% 229/312 Week1 Week3 Week 4 Week 5+ 69% 208/303 57% 153/268 60% 150/252 91% of milk tested at least 1 time for breast feeding mothers and 69% tested at least twice Josephson et al. JAMA Pediatrics 2014 Baseline characteristics of cohort Maternal Age mean years ± SD 29.4 ± 6.5 CMV seropositive 352 (76.2) Cesarean delivery 349 (75.5) Indications for premature delivery Isolated spontaneous labor 138 (29.8) Premature rupture of membranes (<37 weeks) 128 (27.7) Pregnancy-associated hypertension 109 (23.6) Fetal distress / Poor biophysical profile 67 (14.5) Bleeding complication 44 (9.5) Gestational age mean weeks ± SD 27.8 ± 2.6 Birth weight mean grams ± SD 1011 ± 273 N=462 (mothers); 539 (infants) Josephson et al. JAMA Pediatrics

12 Incidence of postnatal CMV infection Characteristic Proportion Infected Cumulative Incidence 95% CI CMV infection 29/ % % - infants born to CMV (-) mothers - Infants born to CMV (+) mothers 0/ % % 29/ % % - CMV disease or mortality 5/ % % Josephson et al. JAMA Pediatrics 2014 Source of postnatal CMV infection Source Proportion infected Cumulative Incidence 95% CI Breast milk-transmitted 27 / % % Transfusion-transmitted 0 / % % Unknown source 1 / % % Josephson et al. JAMA Pediatrics

13 Breast milk CMV shedding Factor Frequency Cumulative Incidence 95% CI CMV in breastmilk in CMV seropositive mothers CMV in breastmilk in CMV seronegative mothers 189 / % 70% - 80% 0 / 81 0% 0% - 5% Josephson et al. JAMA Pediatrics 2014 Risk factors for postnatal CMV infection (univariate) CMV Risk Factor CSHR 95% CI P Birth Weight (per 100g increase) Leukopenia at birth (WBC <5000) Late onset sepsis Breast milk feeding days (per 7 day increase) Isolated spontaneous labor Chorioamnionitis Premature rupture of membranes ROM > 18 hour Breast milk CMV viral load (per 1 log 10 IU/ml) < Female gender SNAP (per 1 unit increase) Receipt of antenatal steroids Josephson et al. JAMA Pediatrics

14 Risk factors for postnatal CMV infection (multivariate) CMV Risk Factor CSHR 95% CI P Premature rupture of membranes Breast milk CMV viral load (per 1 log 10 IU/ml increase) < Josephson et al. JAMA Pediatrics 2014 Breast milk CMV viral load and infection Patel RM et al. prelim accept Arch Dis Child

15 CMV transmission by maximum breast milk CMV viral load Maximum breast milk CMV viral load Probability of CMV transmission (95% CI) 1,000 IU/ml 9% (5% - 14%) 10,000 IU/ml 17% (11% - 25%) 100,000 IU/ml 30% (20% - 42%) Epidemiology of breastmilk & CMV Hamprecht K et al. Lancet

16 Examine findings of NEC in the context of CMV infection Background on NEC NEC is a leading cause of mortality in VLBW infants NEC is characterized by intestinal inflammation and necrosis although the exact pathogenesis is unknown Approximately 7% of VLBW infants (< 1500 grams) develop NEC, with a 15 30% mortality rate Neu and Walker, NEJM 2011 Lin and Stoll, Lancet 2006 Patel et al. N Eng J Med

17 Postnatal CMV Infection Outcomes: Death and NEC One infant died of pneumonia following NEC Had a max viral load of 13,000 IU/ml 2 other infants died of NEC with viral loads at death of 8,000 IU/ml and 4,000 IU/ml 2 surviving infants that developed CMV disease: a) brain calcifications; b) Sepsis like syndrome Treated with ganciclovir and valganciclovir with clinical improvement All of the above infants were fed frozen and thawed breast milk 5 week old, ELBW infant developed acquired pcmv from maternal breast milk and developed CMV sepsis syndrome and clinical evidence of NEC. Serial dried blood spots from DOL 4 21 revealed CMV DNAemia present for several weeks before onset of severe CMV disease. First being noted on DOL 18, increasing to 4.8 log 10 genomes/ml on DOL 21, ~ 8 days before onset of abdominal distention and 15 days before onset of CMV sepsis syndrome and NEC

18 35 Postnatal CMV and NEC Among 2,132 VLBW infants, 7% evaluated for postnatal CMV infection and 27 (19%) positive NEC incidence among postnatal CMV infants 21% vs. 10% in uninfected/untested infants (P=0.11) Mukhopadhyay et al. Am J Perinatol

19 CMV immediate-early antigen (IEA) or CMV late antigen (LA) ~70% of NEC tissue specimens N=61 infants (70 specimens) Omarsdottir et al. J Clin Virol

20 In a cohort of 17 infants with NEC who had suitable samples, no CMV detected by blood PCR Skeath et al. Acta Paediatrica Findings from our studies 20

21 Postnatal CMV and NEC 77% Patel RM et al. prelim accept Arch Dis Child 2019 Association between postnatal CMV and NEC Univariable Models: Risk Factors for NEC and Mortality using Competing-risks Cox Regression Models (n=596) a NEC Mortality Risk Factor n (for NEC) N CSHR (95% CI) P CSHR (95% CI) P A. Univariable analysis postnatal CMV TDC (early) b,c (1.08, 22.59) f postnatal CMV TDC (midpoint) b,d (0.54, 9.2) f Birth weight (per 100g increase) (0.66, 0.86) < (0.42, 0.64) <0.001 Log 10 maximum CMV viral load in breast milk in first 14 days of life (per 1 Log e increase) (1.14, 1.64) < (0.54, 1.19) 0.28 Log e maximum CMV viral load in breast milk in first 14 days of life, excluding infants never fed (per 1 Log e increase) e Log e maximum CMV viral load in breast milk in first 14 days of life among infants fed CMV+ breastmilk (per 1 Log e increase) e (1.09, 1.59) (0.56, 1.23) (1.02, 1.78) (0.49, 1.48)

22 Association between postnatal CMV and NEC Multivariable Models: Risk Factors for NEC and Mortality using Competing-risks Cox Regression Models (n=596) a NEC Mortality Risk Factor CSHR (95% CI) P CSHR (95% CI) P B1. Multivariable (N=596) postnatal CMV TDC (midpoint) b,d 2.81 (0.73, 10.88) f Birth weight (per 100g increase) 0.75 (0.65, 0.86) < (0.41,0.65) <.0001 B2 Multivariable (N=596) postnatal CMV TDC (early) b,c 6.02 (1.28, 28.41) f Birth weight (per 100g increase) 0.75 (0.65, 0.86) < (0.41,0.65) <.0001 B3. Multivariable (N=596) postnatal CMV TDC (early) b,c 3.94 (0.77, 20.03) f Birth weight (per 100g increase) 0.76 (0.67, 0.88) (0.38,0.62) <.0001 Log 10 maximum CMV viral load in breast milk in first 14 days of life (per 1 Log 10 increase) 1.26 (1.05,1.52) (0.48,1.03) Patel RM et al. prelim accept Arch Dis Child 2019 Association between postnatal CMV and NEC NEC incidence 18% among infants with pcmv infection, vs 7% among infants without infection Multivariable analysis, pcmv infection associated with NEC using early estimates of onset (models B2), but the association diminished using a midpoint estimate of onset (model B1) Max breastmilk CMV viral load in first 14 d of life was associated with NEC, similar findings among enterally fed infants and those born to mothers with CMV positive breastmilk. In multivariable analyses, controlling for BW and pcmv infection, the max breast milk CMV viral load in first 14 d of life was independently associated with NEC (cause specific hazard ratio per 1 Log 10 increase 1.26; 95% CI ) (model B3). Patel RM et al. prelim accept Arch Dis Child

23 Wright CJ. Permar SR. J Pediatr Conclusions Postnatal CMV infection occurs in 7 13% of VLBW infants born to CMV seropositive mothers. Freezing of breastmilk decreases but does not prevent transmission. Further studies needed to understand the potential effects of CMV exposure from breast milk and NEC. Currently, there is not enough data to guide prevention strategies or routine testing of VLBW infants. 23

24 Anemia, Red Blood Cell Transfusion and NEC NEC-related deaths by postnatal age Percent of all deaths caused by NEC Postnatal age Patel RM, et al. N Engl J Med

25 Pre-1960s NEC described 1987: Outbreak of NEC associated with RBC transfusion 1978: Bell staging 2010: Josephson et al. Christensen et al. 2011: Singh et al. Paul et al. El-Dib et al. Blau et al. 2012: Mohamed & Shah. 2006: Mally et al. 2014: Derenzio et al : Hebert et al. NEJM 2005: Iowa trial 2006: PINT trial 2019 TOP trial 2014: Sharma et al. Wallenstein et al. (no association) Liberal vs. restrictive trials: meta-analysis of approaches and NEC Kirpalani H, Zupancic JA. Semin Perinatol

26 Transfusion and NEC: Meta-analysis Mohamed and Shah. Pediatrics Is it the RBC transfusion or anemia? Singh R et al. J Perinatol

27 Primary objective: To test the hypothesis that the risk of NEC is greater in very low birth weight infants exposed to red cell transfusion compared to non-transfused very low birth weight infants Secondary objective: To determine if exposure to severe anemia (hemoglobin 8 g/dl) is an independent risk factor for NEC in very low birth weight infants Patel RM et al. JAMA Multivariate Analysis Anemia, NEC and RBC transfusion Primary outcome: RBC transfusion in a given week was not significantly related to the rate of NEC Patel RM et al. JAMA

28 Multivariate Analysis Anemia, NEC and RBC transfusion Secondary Outcome: Based on evaluation of 4565 longitudinal measurements of Hb (median 7 per infant), the rate of NEC was significantly increased among VLBW infants with severe anemia in a given week compared with those who did not have severe anemia. Patel RM et al. JAMA Risk factors among RBC transfused infants N=319 with 44 NEC events Patel RM et al. JAMA

29 Temporality between RBC transfusion, severity of anemia and NEC 1430 RBC txs occurred and 53% of infants received 1 or more txs 44 infants that developed NEC and 33 (75%) received a RBC tx before dx The overall risk of developing NEC within 48 hours after RBC transfusion was 0.49% (7 events following 1,430 RBC transfusions). 7(15.9%) received a tx within 48 hrs of dx Updated meta-analysis of red cell transfusion and NEC Remember: Observational studies can suggest an association between exposure and outcome but cannot prove causality. Hay S et al. Semin Perinatol

30 = increase in RBC NECrelated deaths in < 1000 gm ELBWs Pre-1960s NEC described 1987: Outbreak of NEC associated with RBC transfusion 1978: Bell staging 2010: Josephson et al. Christensen et al. 2011: Singh et al. Paul et al. El-Dib et al. Blau et al. 2012: Mohamed & Shah. 2006: Mally et al. 2014: Derenzio et al : Hebert et al. NEJM 2005: Iowa trial 2006: PINT trial 2019 TOP trial 2014: Sharma et al. Wallenstein et al. (no association) Transfusion of prematures (TOP) Goal is to determine among extremely low birth weight infants if a restrictive vs liberal RBC transfusion threshold results in a reduction in death or neurodevelopmental impairment at months of corrected age. NEC is a secondary outcome. Remember: Evidence from well-designed RCTs is ultimately needed to resolve the uncertainty around optimal RBC transfusion practices which may prevent NEC. 30

31 Transfusion of Prematures (TOP) Trial Enrollment complete (n=1824) and follow-up ongoing ClinicalTrials.gov Identifier: NCT Candidate Mechanisms for RBC Transfusion and NEC Anemia Decreased oxygen delivery to intestines Degree of Prematurity Enteral Feeding Illness severity Necrotizing Enterocolitis Increased gut permeability Increased inflammation Prolonged RBC Storage Prolonged Irradiation Storage Time Decreased nitric oxide (NO) Decreased Vasoconstriction oxygen delivery to intestines 31

32 Enrolled as of : 138 LBW infants NSH: 84 Grady: 37 Midtown: 17 New Grant: PPG with Boston Children s Hospital Martha Sola-Visner, MD (PI) C Josephson (PI), MD, Sean Stowell, MD PhD, Ravi Patel, MD, MSc, and Patti Denning, MD 32

33 Acknowledgements Thanks to the families who volunteered to participate in our studies Ravi Patel, MD, MSc Angela M. Caliendo, MD; PhD, John Roback, MD, PhD, Sean Stowell, MD, PhD, Connie Arthur, PhD, Martha Sola-Visner, MD Deborah Abdul-Ali, BS; Jessica Ingersoll, MS, Natia Saakadze Doris Igwe,BS; Shieghla Barclay, MT; Natia Saakadze, BS Kirk A. Easley, MapStat, Andrea Knezevic, MS, Neeta Shenvi, MS, Marjorie Howard, MS Jane Skvarich, BSN, MS Janna Benston, RN Katrina Grier, RN, Renee Cook, BSN, Hailey Butler, BSN Michael T. Hinkes, MD, Sarah Keene, MD, Patricia Denning, MD, George Bugg, MD, Andi Shane, MD, MPH, MSc Blood Bank Technologists at all participating hospitals Funding: NHLBI P01 HL086773, K23 HL Questions??? 33

34 What is the current guidance in the United States on breastfeed and risk of CMV? Current recommendations in US There are no recommendations against breastfeeding by mothers who are CMV-seropositive. However, premature infants (born <30 weeks gestational age and <1500g) who acquire CMV from breast milk may be at risk of developing a late-onset sepsis-like syndrome. The potential benefits of human milk versus the risk of CMV transmission should be considered when making a decision about breastfeeding of very low birth weight infants (birth weight <1500 g) by mothers known to be CMV-seropositive. Centers for Disease Control and Prevention (2016) Source: 34

35 AAP Policy Statement Decisions about breastfeeding of very low birth weight infants (birth weight < 1500 g) by mothers known to be CMV-seropositive should be made with consideration of the potential benefits of human milk versus the risk of CMV transmission. American Academy of Pediatrics Section on Breastfeeding (2005) Updated AAP Policy Statement There is no contraindication to breastfeeding for a fullterm infant whose mother is seropositive for cytomegalovirus. There is a possibility that CMV acquired from mother s milk may be associated with a late-onset sepsis-like syndrome The value of routinely feeding human milk from seropositive mothers to preterm infants outweighs the risks of clinical disease, especially because no long-term neurodevelopmental abnormalities have been reported. American Academy of Pediatrics Section on Breastfeeding (2012) 35

36 Incidence of postnatal CMV infection in VLBW infants in the US Outcome Risk among infants fed untreated breast milk (95% CI) Number of cases based on 2008 US data Postnatal CMV infection 19% (11-32%) 2000 Lanzieri TM et al. Pediatrics 2013 How can we prevent breastmilk transmission? 36

37 Current recommendations in US Freezing and pasteurization of breast milk can decrease the risk of transmission; however, freezing does not eliminate the risk of transmission. CDC (2016) Freezing of milk reduces but does not eliminate CMV. AAP (2012) Source: Eidelman et al. Section of Breastfeeding. AAP. Pediatrics Estimates of incidence of postnatal CMV infection in VLBW infants Outcome Infants fed untreated breast milk (95% CI) Postnatal CMV infection 19% (11-32%) - Infants fed frozen milk 13% (7-24%) Postnatal CMV sepsis-like syndrome 4% (2-7%) - Infants fed frozen milk 5% (2-12%) Lanzieri TM et al. Pediatrics

38 Pasteurization Short (5 sec) pasteurization Long (30 min) Holder pasteurization at 63 C Goelz et al. Pediatr Res Irradiation Lloyd et al. PLOS One

39 Antiviral treatment No guidance currently available on when to treat postnatal CMV infection Medications include ganciclovir and valganciclovir Mukhopadhyay et al. Am J Perinatol