From Storm to Suppression in Sepsis: Are Bands the Link?* Kirsten Neudoerffer Kangelaris, MD, MAS Eric J. Seeley, MD, FCCP Division of Hospital Medicine University of California San Francisco, CA eukocyte activation occurs early during sepsis and is a key com ponent o f the proinflam m atory response to infection. This response includes a “storm ” of cytokines released into the circulation and a cascade of im m unologic events directed at containing and killing pathogens (1). This early reaction to infection includes the mobilization of im m a ture granulocytes, or band forms, from the bone marrow. The presence of bands has long been identified as a sign of infec tion and remains an inflam m atory criterion in the definition of sepsis (2). However, identifying the exact m echanism s that lead to organ dysfunction and death in sepsis has rem ained elusive. In addition to a proinflam m atory response, evidence sug gests that sim ultaneous anti-inflam m atory processes lead to durable im m unosuppression and increased risk for sec ondary infections (3-5). Elevated levels of anti-inflam m a tory cytokines and a suppressed adaptive im m une response, including T-cell dysfunction and lym phopenia, have been widely observed (6-8). Moreover, the relationship between the proinflam m atory and anti-inflam m atory pathways in sepsis is poorly characterized, and w hether the cytokine storm is linked to the developm ent of im m unosuppression is not known. In this issue of Critical Care Medicine, Guerin et al (9) use multicolor flow cytometry to identify two leukocyte subsets, consistent with im m ature granulocytes, that predict clinical deterioration in septic adults. Furthermore, these data present intriguing support for the hypothesis that immature granulo cytes, classically linked to the inflammatory pathway in sepsis, may be directly involved in immunosuppression of the adaptive im m une response as well. The investigators evaluated 23 leukocyte subsets, includ ing im m ature and m ature granulocytes, dendritic cells,
L
*S e e a ls o p. 2007. K ey
W o rd s: flo w in fla m m a tio n ; s e p s is
c y to m e try ;
g ra n u lo c y te s ;
im m u n o s u p p re s s io n ;
Dr. K a n g e la ris ' in s titu tio n re c e iv e d g ra n t s u p p o rt fro m th e N a tio n a l H e a rt, L u n g, a nd B lo o d In s titu te (N H L B I) (1 K 2 3 H L 1 1 6 8 0 0 - 0 1 ; c a re e r d e v e lo p m e n t a w a rd ). Dr. S e e le y p ro v id e d e xp e rt te s tim o n y fo r M c K e e n (Le g a l C o n s u ltin g ) a n d re ce ive d s u p p o rt fo r a rtic le re se a rch fro m th e N a tio n a l In s titu te s o f H e a lth (N IH ) (N IH p a ys fo r salary). H is in s titu tio n re ce ive d g ra n t s u p p o rt fro m th e N IH - N H L B I ( H L 1 1 1 2 0 9 2 - 0 1 A 1 ; K a w a rd fo r salary). C o p y rig h t © 2 0 1 4 b y th e S o c ie ty o f C ritic a l C a re M e d ic in e a n d L ip p in c o tt W illia m s & W ilk in s
D O I: 1 0 .1 0 9 7 /C C M .0 0 0 0 0 0 0 0 0 0 0 0 0 3 7 8
Critical Care Medicine
m acrophages/monocytes, plasma cells, and T- and B-cell sub sets, using 10-color flow cytom etry in blood samples from 177 septic adults (2). Presurgical, noninfected patients and infected patients w ithout evidence of systemic inflam m atory response were included separately as controls. Compared with controls, dim inished expression of CD 16 (CD16dira) and CD10 (CD10d,m) on granulocytes was diagnostic of sep sis. W hen limited to just the sepsis cohort, CD16din7CD10dim emerged as the m ost predictive markers of clinical deteriora tion and 30-day mortality. In addition to identifying a subset of leukocytes that pre dicted clinical deterioration and mortality, the authors used multicolor flow sorting to identify and then isolate a subpopu lation of immature granulocytes that were CD14neg/CD24pos within the CD16dlmcompartment from seven patients with sep tic shock. These cells were then cocultured with activated autol ogous T cells in vitro and exhibited a dose-dependent cytotoxic effect on these T cells, suggesting that these cells might also lead to T-cell death in vivo. Furthermore, interrogation of the clini cal dataset demonstrated that increased granulocyte CD16dimin septic patients was associated with T-cell lymphopenia. Nota bly, band forms were identified on manual differential among all septic CD16d™/CD10dimpatients, suggesting that these mark ers may identify the subpopulation of immature granulocytes that are commonly referred to as “bands.” The authenticity o f CD16dim/CD 10dim as a novel im m une m arker o f sepsis and disease progression is bolstered by com parisons with previously described im m unophenotypes in sepsis. N eutrophil expression o f CD64 is a well-charac terized diagnostic m arker of neonatal sepsis (10). Consis tent w ith these findings, CD64 expression was elevated in septic patients com pared with controls in the current study. CD64 positivity did not, however, predict clinical deteriora tion am ong septic patients. Decreased m onocyte expression of hum an leukocyte antigen (HLA) complex-DR, previously associated w ith secondary infections, shock, and m ortality in intensive care populations (11, 12), was associated with progression to septic shock and sepsis m ortality in this study. Interestingly, however, neither neutrophil CD64 expression nor m onocyte HLA-DR was associated with lym phopenia. The latter finding is notew orthy since HLA-DR expression, a m arker of m onocyte functionality, has been postulated as a potential biom arker o f sepsis-related im m unosuppres sion although the m echanism may be independent of T-cell suppression (12). Overall, these findings are consistent with p rio r studies and provide credence to the m ethodology of the current study. Thus, this study presents two m ain findings: 1) flow cytom etry measures of im m ature granulocytes may aid in the prognosis of patients presenting with sepsis and 2) these cells, know n m ediators o f the sepsis inflam m atory w w w .c c m jo u r n a l.o r g
2135
Editorials
response, may also be directly linked to sepsis-related immunosuppression. However, several im portant uncertainties remain in translating these results to the management of sepsis in clinical practice. First, this is a single-center, modest-sized cohort study. External validation of these findings will be necessary. Second, it is not clear that CD16d,n7CD10d,m cells are equivalent to bands. Although band forms were observed in all CD16din7CD10dlmpatients in this study, prior data suggest that the CD16dinVCD10dimcell surface signature may encompass a broader population of newly emerged granulocytes including segmented neutrophils (13). Reli able measures of bandemia must be correlated with the presence of granulocyte CD16dira/CD10dira and clinical out comes in septic populations. Third, whether flow cytom etry is practical for clinical use in sepsis remains to be seen. Im m unophenotyping has historically suffered from lack of standardization and reproducibility (14, 15). Standardized assays are being developed and should be applied in future research. Furthermore, although flow cytometry is increas ingly available in clinical laboratories, it will be difficult to use as a point-of-care test. In addition, further study will be required to determine whether flow cytometry can provide additive prognostic information over other accurate, poten tially more cost-effective measures of bandemia. Finally, despite intriguing in vitro evidence, it is not clear that CD14neg/CD24pos immature granulocytes cause T-cell death in vivo, and the association between CD16dim and T-cell lymphopenia could be due to separate distal effects inherent in a more severely ill population. A detailed mecha nistic approach determining the link between immature granulocytes and immunosuppression is warranted and, if established, could provide a potential therapeutic target in the future. We therefore return to our title: “From storm to suppres sion in sepsis: Are bands the link?” The answer is that we do not yet know, but the current work by Guerin et al (9) provides
2136
w w w .c c m jo u r n a l.o r g
intriguing leads for additional study to eventually answer this important question.
REFERENCES 1. A ngus D C, van der Poll T : Severe sepsis and se p tic shock. N Engl J Med 2 0 1 3 ;3 6 9 :8 4 0 - 8 5 1 2. Levy M M , Fink MP, M arshall JC, et al; S C C M /E S IC M /A C C P /A T S / SIS : 2001 S C C M /E S IC M /A C C P /A T S /S IS International S epsis D efinitions C onference. Crit Care M ed 2 0 0 3 ; 3 1 :1 2 5 0 - 1 2 5 6 3. Bone RC, G rodzin CJ, B alk RA: S epsis: A new hypothesis fo r pa tho genesis o f the disease process. C h e st 1 9 9 7 ; 1 1 2 :2 3 5 - 2 4 3 4. van der Poll T, O pal S M : H ost-pathogen interactions in sepsis. Lancet Infect Dis 2 0 0 8 ; 8 :3 2 - 4 3 5. H otchkiss RS, Karl IE: The pathophysiology and treatm ent o f sepsis. N Engl J Med 2 0 0 3 ; 3 4 8 :1 3 8 - 1 5 0 6. H otchkiss RS, S w anson PE, Freeman B D , et al: A p o p to tic cell death in patients w ith sepsis, shock, and m ultiple organ dysfunction. Crit Care Med 1 9 9 9 ; 2 7 :1 2 3 0 -1 2 5 1 7. H otchkiss RS, Tinsley KW, Sw anson PE, et al: S epsis-induced a p o p tosis causes progressive profound depletion o f B and C D 4 + T lym phocytes in humans. J Immunol 2 0 0 1 ; 1 6 6 :6 9 5 2 - 6 9 6 3 8. B oom er JS, To K, C hang KC, et al: Im m unosuppression in patients w ho die o f sepsis and m ultiple organ failure. JAMA 2 0 1 1 ; 3 0 6 :2 5 9 4 - 2 6 0 5 9. G uerin E, O rabona M, Raquil M-A, et al: C irculating Immature G ranulocytes W ith T-Cell Killing Functions P redict S epsis D eterioration. Crit Care Med 2 0 1 4 ; 4 2 :2 0 0 7 - 2 0 1 8 10. Bhandari V, W ang C, R inder C, et al: H em atologic profile o f sepsis in neonates: N eutrophil C D 6 4 as a d ia g n ostic marker. Pediatrics 2 0 0 8 ; 1 2 1 :1 2 9 - 1 3 4 11. Lukaszewicz AC, G rienay M, Resche-Rigon M, e ta l: M onocytic HLA-D R expression in intensive care patients: Interest for prognosis and sec ondary infection prediction. Crit Care Med 2 0 0 9 ; 3 7 :2 7 4 6 -2 7 5 2 12. Landelle C, Lepape A, Voirin N, et al: Low m onocyte human leukocyte antigen-D R is independently associated w ith nosocom ial infections after se p tic shock. Intensive Care Med 2 0 1 0 ; 3 6 :1 8 5 9 - 1 8 6 6 13. O rr Y, Taylor JM, Bannon PG, et al: C irculating C D 1 0 -/C D 1 6 lo w neu trop h ils provide a quantitative index o f active bone m arrow neutrophil release. B r J Haematol 2 0 0 5 ; 1 3 1 :5 0 8 - 5 1 9 14. Venet F, Lepape A, M onneret G : C linical review: Flow cytom etry per spectives in the IC U -F ro m diagnosis o f infection to m onitoring of injury-induced immune dysfunctions. Crit Care 2 0 1 1 ; 15:231 15. M aeckerH T, M cC o y JP, N ussenblatt R: Standardizing im m unopheno typing fo r the Human Im m unology Project. Nat Rev Immunol 201 2; 1 2 :1 9 1 -2 0 0
S eptem ber 2 0 1 4 - Volum e 4 2 • N um ber 9
Copyright of Critical Care Medicine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
L
*S e e a ls o p. 2007. K ey
W o rd s: flo w in fla m m a tio n ; s e p s is
c y to m e try ;
g ra n u lo c y te s ;
im m u n o s u p p re s s io n ;
Dr. K a n g e la ris ' in s titu tio n re c e iv e d g ra n t s u p p o rt fro m th e N a tio n a l H e a rt, L u n g, a nd B lo o d In s titu te (N H L B I) (1 K 2 3 H L 1 1 6 8 0 0 - 0 1 ; c a re e r d e v e lo p m e n t a w a rd ). Dr. S e e le y p ro v id e d e xp e rt te s tim o n y fo r M c K e e n (Le g a l C o n s u ltin g ) a n d re ce ive d s u p p o rt fo r a rtic le re se a rch fro m th e N a tio n a l In s titu te s o f H e a lth (N IH ) (N IH p a ys fo r salary). H is in s titu tio n re ce ive d g ra n t s u p p o rt fro m th e N IH - N H L B I ( H L 1 1 1 2 0 9 2 - 0 1 A 1 ; K a w a rd fo r salary). C o p y rig h t © 2 0 1 4 b y th e S o c ie ty o f C ritic a l C a re M e d ic in e a n d L ip p in c o tt W illia m s & W ilk in s
D O I: 1 0 .1 0 9 7 /C C M .0 0 0 0 0 0 0 0 0 0 0 0 0 3 7 8
Critical Care Medicine
m acrophages/monocytes, plasma cells, and T- and B-cell sub sets, using 10-color flow cytom etry in blood samples from 177 septic adults (2). Presurgical, noninfected patients and infected patients w ithout evidence of systemic inflam m atory response were included separately as controls. Compared with controls, dim inished expression of CD 16 (CD16dira) and CD10 (CD10d,m) on granulocytes was diagnostic of sep sis. W hen limited to just the sepsis cohort, CD16din7CD10dim emerged as the m ost predictive markers of clinical deteriora tion and 30-day mortality. In addition to identifying a subset of leukocytes that pre dicted clinical deterioration and mortality, the authors used multicolor flow sorting to identify and then isolate a subpopu lation of immature granulocytes that were CD14neg/CD24pos within the CD16dlmcompartment from seven patients with sep tic shock. These cells were then cocultured with activated autol ogous T cells in vitro and exhibited a dose-dependent cytotoxic effect on these T cells, suggesting that these cells might also lead to T-cell death in vivo. Furthermore, interrogation of the clini cal dataset demonstrated that increased granulocyte CD16dimin septic patients was associated with T-cell lymphopenia. Nota bly, band forms were identified on manual differential among all septic CD16d™/CD10dimpatients, suggesting that these mark ers may identify the subpopulation of immature granulocytes that are commonly referred to as “bands.” The authenticity o f CD16dim/CD 10dim as a novel im m une m arker o f sepsis and disease progression is bolstered by com parisons with previously described im m unophenotypes in sepsis. N eutrophil expression o f CD64 is a well-charac terized diagnostic m arker of neonatal sepsis (10). Consis tent w ith these findings, CD64 expression was elevated in septic patients com pared with controls in the current study. CD64 positivity did not, however, predict clinical deteriora tion am ong septic patients. Decreased m onocyte expression of hum an leukocyte antigen (HLA) complex-DR, previously associated w ith secondary infections, shock, and m ortality in intensive care populations (11, 12), was associated with progression to septic shock and sepsis m ortality in this study. Interestingly, however, neither neutrophil CD64 expression nor m onocyte HLA-DR was associated with lym phopenia. The latter finding is notew orthy since HLA-DR expression, a m arker of m onocyte functionality, has been postulated as a potential biom arker o f sepsis-related im m unosuppres sion although the m echanism may be independent of T-cell suppression (12). Overall, these findings are consistent with p rio r studies and provide credence to the m ethodology of the current study. Thus, this study presents two m ain findings: 1) flow cytom etry measures of im m ature granulocytes may aid in the prognosis of patients presenting with sepsis and 2) these cells, know n m ediators o f the sepsis inflam m atory w w w .c c m jo u r n a l.o r g
2135
Editorials
response, may also be directly linked to sepsis-related immunosuppression. However, several im portant uncertainties remain in translating these results to the management of sepsis in clinical practice. First, this is a single-center, modest-sized cohort study. External validation of these findings will be necessary. Second, it is not clear that CD16d,n7CD10d,m cells are equivalent to bands. Although band forms were observed in all CD16din7CD10dlmpatients in this study, prior data suggest that the CD16dinVCD10dimcell surface signature may encompass a broader population of newly emerged granulocytes including segmented neutrophils (13). Reli able measures of bandemia must be correlated with the presence of granulocyte CD16dira/CD10dira and clinical out comes in septic populations. Third, whether flow cytom etry is practical for clinical use in sepsis remains to be seen. Im m unophenotyping has historically suffered from lack of standardization and reproducibility (14, 15). Standardized assays are being developed and should be applied in future research. Furthermore, although flow cytometry is increas ingly available in clinical laboratories, it will be difficult to use as a point-of-care test. In addition, further study will be required to determine whether flow cytometry can provide additive prognostic information over other accurate, poten tially more cost-effective measures of bandemia. Finally, despite intriguing in vitro evidence, it is not clear that CD14neg/CD24pos immature granulocytes cause T-cell death in vivo, and the association between CD16dim and T-cell lymphopenia could be due to separate distal effects inherent in a more severely ill population. A detailed mecha nistic approach determining the link between immature granulocytes and immunosuppression is warranted and, if established, could provide a potential therapeutic target in the future. We therefore return to our title: “From storm to suppres sion in sepsis: Are bands the link?” The answer is that we do not yet know, but the current work by Guerin et al (9) provides
2136
w w w .c c m jo u r n a l.o r g
intriguing leads for additional study to eventually answer this important question.
REFERENCES 1. A ngus D C, van der Poll T : Severe sepsis and se p tic shock. N Engl J Med 2 0 1 3 ;3 6 9 :8 4 0 - 8 5 1 2. Levy M M , Fink MP, M arshall JC, et al; S C C M /E S IC M /A C C P /A T S / SIS : 2001 S C C M /E S IC M /A C C P /A T S /S IS International S epsis D efinitions C onference. Crit Care M ed 2 0 0 3 ; 3 1 :1 2 5 0 - 1 2 5 6 3. Bone RC, G rodzin CJ, B alk RA: S epsis: A new hypothesis fo r pa tho genesis o f the disease process. C h e st 1 9 9 7 ; 1 1 2 :2 3 5 - 2 4 3 4. van der Poll T, O pal S M : H ost-pathogen interactions in sepsis. Lancet Infect Dis 2 0 0 8 ; 8 :3 2 - 4 3 5. H otchkiss RS, Karl IE: The pathophysiology and treatm ent o f sepsis. N Engl J Med 2 0 0 3 ; 3 4 8 :1 3 8 - 1 5 0 6. H otchkiss RS, S w anson PE, Freeman B D , et al: A p o p to tic cell death in patients w ith sepsis, shock, and m ultiple organ dysfunction. Crit Care Med 1 9 9 9 ; 2 7 :1 2 3 0 -1 2 5 1 7. H otchkiss RS, Tinsley KW, Sw anson PE, et al: S epsis-induced a p o p tosis causes progressive profound depletion o f B and C D 4 + T lym phocytes in humans. J Immunol 2 0 0 1 ; 1 6 6 :6 9 5 2 - 6 9 6 3 8. B oom er JS, To K, C hang KC, et al: Im m unosuppression in patients w ho die o f sepsis and m ultiple organ failure. JAMA 2 0 1 1 ; 3 0 6 :2 5 9 4 - 2 6 0 5 9. G uerin E, O rabona M, Raquil M-A, et al: C irculating Immature G ranulocytes W ith T-Cell Killing Functions P redict S epsis D eterioration. Crit Care Med 2 0 1 4 ; 4 2 :2 0 0 7 - 2 0 1 8 10. Bhandari V, W ang C, R inder C, et al: H em atologic profile o f sepsis in neonates: N eutrophil C D 6 4 as a d ia g n ostic marker. Pediatrics 2 0 0 8 ; 1 2 1 :1 2 9 - 1 3 4 11. Lukaszewicz AC, G rienay M, Resche-Rigon M, e ta l: M onocytic HLA-D R expression in intensive care patients: Interest for prognosis and sec ondary infection prediction. Crit Care Med 2 0 0 9 ; 3 7 :2 7 4 6 -2 7 5 2 12. Landelle C, Lepape A, Voirin N, et al: Low m onocyte human leukocyte antigen-D R is independently associated w ith nosocom ial infections after se p tic shock. Intensive Care Med 2 0 1 0 ; 3 6 :1 8 5 9 - 1 8 6 6 13. O rr Y, Taylor JM, Bannon PG, et al: C irculating C D 1 0 -/C D 1 6 lo w neu trop h ils provide a quantitative index o f active bone m arrow neutrophil release. B r J Haematol 2 0 0 5 ; 1 3 1 :5 0 8 - 5 1 9 14. Venet F, Lepape A, M onneret G : C linical review: Flow cytom etry per spectives in the IC U -F ro m diagnosis o f infection to m onitoring of injury-induced immune dysfunctions. Crit Care 2 0 1 1 ; 15:231 15. M aeckerH T, M cC o y JP, N ussenblatt R: Standardizing im m unopheno typing fo r the Human Im m unology Project. Nat Rev Immunol 201 2; 1 2 :1 9 1 -2 0 0
S eptem ber 2 0 1 4 - Volum e 4 2 • N um ber 9
Copyright of Critical Care Medicine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
