|Home | About | Journals | Submit | Contact Us | Français|
Defective immune reconstitution is a major barrier to successful hematopoietic cell transplantation (HCT), and has important implications in the pediatric population. There are many factors which affect immune recovery, including stem cell source and GVHD. Complete assessment of immune recovery, including T and B lymphocyte evaluation, innate immunity and response to neoantigens, may provide insight as to infection risk and optimal time for immunizations. The increasing use of cord blood grafts requires additional study regarding early reconstitution and impact upon survival. Immunization schedules may require modification based upon stem cell source and immune reconstitution, and this is of particular importance as many children have been incompletely immunized, or not at all, prior to school entry. Additional studies are needed in children post HCT to evaluate the impact of differing stem cell sources upon immune reconstitution, infectious risks and immunization responses.
Defective immune reconstitution is a major barrier to successful hematopoietic cell transplantation (HCT), regardless of graft orgin1, 2. Serious infections have been shown to account for a significant percentage (4–20%) of late deaths after HCT.3 Factors which impact immune recovery include graft manipulation, stem cell source and chronic graft vs host disease (GVHD). The tempo of immune reconstitution may affect another important pediatric issue--immunizations. Many pediatric patients may have been incompletely immunized, or not at all, prior to receiving chemotherapy and HCT, and the timing, type and response to immunizations may play an important role in preventing morbidity as they begin or reenter school.
In April 2011 the NCI/NHLBI along with the Pediatric Blood and Marrow Transplant Consortium (PBMTC) sponsored a consensus conference of international experts in clinical and biological research into late effects after HCT convened to review the state of the science of pediatric studies and identify key areas for future research. This manuscript will describe the conclusions shared at that conference relating to assessment of immune status after HCT, differences in immune status after different types of HCT, and approaches to immunization in order to effectively prevent selected viral and bacterial infections after HCT.
There is increasing use of unrelated cord blood (UCB) for pediatric patients who lack a matched related donor. Infection-related mortality (IRM) is the primary or secondary cause of death (with or without another major cause such as GvHD) in ≥ 50% of deaths after UCBT with the majority of them occurring in the first 100 days 4–7. Use of cord blood, a donor cell source composed of predominantly naive cells, raises unique challenges related to immune reconstitution and cell maturation that should be considered in the post-transplant period.
Investigators from the International Bone Marrow Transplant Registry (IBMTR) highlighted the unique features of infection incidence after UCBT. Outcome after transplantation was analyzed between recipients of either cord blood (n=150) or from marrow that was from HLA-matched (n=367) or mismatched donors (n=83)8. Infection related mortality (IRM) within 100 days after transplantation was significantly higher among recipients of mismatched cord blood than among recipients of either HLA-matched marrow or mismatched marrow (45%, 21%, and 24%, respectively; P=0.01). However, beyond day 100, the proportions of infection-related deaths were similar in the three groups. Importantly, at later time period there was a trend towards less serious infection in the UCBT group corroborating the findings of IBMTR report8. It appears that the profound immune deficit in the first 6 months post UCBT may be followed by significant improvements of protective immunity. This period of improving immunocompetence coincides with the time of thymic recovery.
Despite some experimental data that neonatal myeloid cells are hypofunctional 9, there is no evidence that post-UCBT the engrafting myeloid granulocytes or monocytes would have reduced function compromising their migratory, phagocytic, or bacterial killing capacity10. Despite the fairly rapid recovery of the innate immune system attaining normal values within the first months, T lymphocyte recovery is variable and numerically remains below age matched normal range for several months. In addition to lymphopenia there is co-existent functional compromise that may be explained by immunosuppressive medications and intrinsic factors such as antigen naivite and TH1/Tc1 immaturity. Investigators from Lyon and Marseille compared the tempo of lymphocyte recovery of 112 children who received UCBT with those (n=114)) who received unrelated bone marrow transplant (UBMT)11. Notably, median time for CD4+ T cell and NK cell recovery was not dissimilar in UCBT and UBMT recipients; however, CD8+ T cell recovery was delayed after UCBT with a median time to reach CD8+ T cells > 0.25 × 10e9/l almost 8 months for UCBT recipients, in contrast to approximately 3 months for pediatric BMT recipients (p< 0.001). Significant B lymphocyte recovery may not commence before ~8–10 weeks. However, the recovery of CD19+ lymphocytes may recover faster after UCBT compared to unrelated donor BMT11. In contrast, NK cell recovery is prompt in both adults and children attaining normal frequencies in blood by the first 2 months, similar to recipients of bone marrow11–13.
Dendritic cell and lymphocyte recovery may impact on survival post transplant. Between July 2005, and December 2007, 93 children with full donor chimerism following myeloablative conditioning (MAC) were longitudinally assessed at Duke University for reconstitution of DC and T cell subsets, B and NK cells by 4 color FACS. Measurements were taken at Day (D) 100, 180, 270 and 365 to determine thresholds associated with overall survival (OS). Parameters were dichotomized at the median (M) of each time point and hazard ratios estimated by Cox proportional hazards analysis. Patients were transplanted for non-malignant (N=63) and malignant diseases (N=30). Median age was 2.1 years, 58% male, 36% were 4/6, 42% were 5/6 and 23% were 6/6 HLA matched. OS at 2-years was 76%. At D100, lower percentage and absolute numbers of regulatory T cells (CD25+/CD62L+/CD4+Tregs) was associated with death (p=0.04 and p=0.02). A higher fraction of “activated” HLA-DR expressing CD8+ T cells predicted death at D100 with marginal significance at 180 and 270 while absolute numbers significant at D270. At D180, a period when thymic function may resume after MAC, recent thymic emigrants was associated with OS (RTE, CD45RA+/CD62L+ T cells), p=0.01 and p=0.04. Higher levels of ‘plasmacytoid’ CD123+ DC were consistently associated with better outcome after D100 (p<0.05). Major lymphocyte subsets, B, NK, CD3+, CD4+ T cells were NOT associated with superior outcome except at D180 when all lymphocytes were combined, exceeding an ALC >1180/ul (p=0.04). Figure 1A–C depicts a survival curve of patients who are below or above the median cut-off values for immune parameters at the designated time point; Fig 1A at Day 100, Fig 1B at Day 180, and Fig 1C at 1 year. These data demonstrate that different immune cell subsets will have unique kinetics and dynamics of recovery that confer shifting power in predicting outcome when tested at different time points within the first year. Nevertheless, by the second half of the first year, superior pDC and RTE recovery along with a lower activation state (%HLA-DR) are independently associated with better OS. These results should greatly facilitate the development of risk models for the early identification of those at highest risk of death and who may benefit from timely immunotherapy interventions
T cells play critical roles in host defense against viral and fungal pathogens. They are essential for directing appropriate antibody responses, and they assist in regulating the immune response and maintaining tolerance. Absent or defective T cell responses therefore have broad implications for infectious susceptibility and development of GVHD in the post-transplant period. Traditionally, assessment of the recovery of T cells after HCT has focused on relatively crude measures of numbers and function including simple enumeration of CD3, CD4, and CD8 T cells and in vitro proliferative responses to mitogens measured by incorporation of [H3]thymidine into the DNA of replicating cells or by CFSE dye-dilution studies using flow cytometry. Improved flow cytometry and molecular tools have now made a more detailed analysis of T cell recovery post-HCT possible. These tools can be applied to evaluate development and maturation of specific T cell subsets, assess thymic function and thymic T cell output, and to evaluate T cell receptor (TCR) repertoire diversity (Table I).
Abnormalities in development and maturation of both CD4 and CD8 T cells may be observed after HCT, particularly in the setting of cGVHD, which has been shown to cause skewing to a more activated, effector memory phenotype in both populations14, 15. Pairing of high-resolution immunophenotyping with measurement of TRECs can provide an estimate of thymic function and naive T cell output after HCT. Similar to effector T cells, reconstitution of regulatory T cells (Tregs) can be monitored after HCT by flow cytometry using an immunophenotyping panel that includes antibodies to CD4, CD25, CD127, and FOXP3.
B cell dysfunction, particularly the inability to generate functional antibody responses, is typically associated with susceptibility to recurrent sinopulmonary infections. Several studies have shown that under ideal circumstances, humoral immune responses to new antigens return as early as 6 months post transplant even though immune reconstitution is not yet complete16, 17. B cell development and maturation, however, can be significantly altered in the posttransplant period, especially in the setting of ongoing immune suppression or chronic GVHD. A small number of recent studies have shown that cGVHD is associated with marked B cell abnormalities including a general B cell lymphopenia and altered B cell maturation including a significant increase in transitional, immature, and naive B cell subsets with a concomitant decrease in mature, CD27+ memory B cells and class-switched memory B cells18–20. Under normal circumstances, B cell lymphopenia and absence of mature B cells is associated with high circulating levels of B cell activating factor (BAFF/BLyS/TALL-1), which plays important roles in stimulating proliferation and preventing apoptosis of B cells. Not surprisingly then, cGVHD is also associated with high circulating BAFF levels and in addition, a concomitant decrease in cell surface expression of the major BAFF receptor (BAFF-R) on the surface of B cells.
A growing understanding of innate and adaptive immune mechanisms combined with technical advancements that make assessment of these feasible, has made it possible to more effectively assess immune function after HCT. The most extensively studied of these mechanisms are the Toll-like receptors (TLRs), which allow cells to respond to a variety of molecules produced by pathogens including viruses, bacteria, and fungi In humans. The TLR proteins are encoded by 10 different genes, each of which encodes a protein capable of recognizing a different pathogen-derived product. Recent studies have evaluated whether single nucleotide polymorphisms (SNPs) in TLR proteins, some of which are known to affect TLR function, may be associated with susceptibility to specific pathogens in the post-transplant period. Interestingly, three studies have now suggested an association between TLR polymorphisms and invasive aspergillosis (IA) in patients undergoing HCT21–23. These studies suggest that alterations in TLR function either in non-hematopoietic cells (pulmonary epithelial cells, etc.) of the recipient or in hematopoietic cells from the donor alter susceptibility to specific pathogens in HCT recipients.
Measurement of the immune response to vaccination is an effective method to globally assess immune reconstitution because it requires effective antigen presentation, B cells capable of responding to antigen, and adequate T cell help. Unfortunately, routine vaccines are not ideal for evaluating immune reconstitution because these are likely to elicit recall responses from both donor and host cells, making them an ineffective measure of the immune system’s ability to respond to new challenges. In addition, they cannot be effectively used to evaluate immune responses in patients receiving supplemental immunoglobulin because of the presence of antibodies to routine antigens in the IgG preparations. Assessment of immunocompetency in the post-transplant period is therefore most effectively judged using a neoantigen that neither the donor nor recipient has been exposed to previously. Ideally, the neoantigen would also be one that the general population has not been broadly exposed to so that there would be little measurable antibody in pooled IgG preparations, allowing responses to be measured even in individuals who may be receiving ongoing IgG supplementation. Three most commonly used neoantigens that fit these criteria include Bacteriophage X174, Rabies vaccine, and Keyhole limpet hemocyanin (KLH).
Most articles addressing vaccination post HCT have focused on unmodified HLA matched sibling transplant recipients, with few studies including large numbers of unrelated HCT recipients or details about the extent or amount of immunosuppressive therapy in patients with chronic GVHD. There are no studies devoted exclusively to recipients of cord blood transplants, a population whose response to immunizations may differ significantly from other transplant groups due to the lack of transfer of memory T- and B-cells in the cord blood graft(s). All vaccine guidelines, including those published in 2009 24, acknowledge the variability in immune reconstitution following alternative donor transplants and the lack of studies assessing vaccine responses in these patient populations. Although vaccination at fixed times post HCT will likely be sufficient for highly immunogenic vaccines such as tetanus and polio, for others (hepatitis B, pneumococcus, meningococcus, and varicella) some studies have suggested that vaccination prior to the acquisition of critical populations of T and B cells may result in low response rates with inadequate durability of response. The 2009 influenza pandemic 25, recent increases in pertussis26 and drug resistant pneumococci 27, and outbreaks of measles 28 and mumps 29 in immunocompetent individuals highlight the need for effective revaccination of HCT patients. Vaccine preventable diseases such as those caused by pneumococcus 30, influenza31, herpes zoster 32 and Bordetella pertussis 26, 33 cause significant morbidity and to a lesser extent mortality even in immunocompetent individuals34. Many of these infections occur more frequently in HCT recipients and remain a significant cause of morbidity, re-hospitalization, and mortality after successful HCT35–38. Following autologous and allogeneic transplantation, patients lack immunity against pertussis and rapidly lose protection against pneumococcus and haemophilus influenza (reviewed in 24, 39). In the absence of revaccination, the majority of patients will become susceptible to measles, mumps, or rubella by 3–5 years post.
Since 1995, efforts have been made to immunize HCT recipients against vaccine preventable diseases in a standardized fashion. Both the European Group for Blood and Marrow Transplantation (EBMT) 40 and the Centers for Disease Control (CDC) 41 have published guidelines which differ in the number of recommended doses of tetanus, polio, and Haemophilus influenza vaccines, as well as the time to initiate re-vaccination 24, 39. In 2009 24, an international committee of infectious disease and transplant physicians met under the auspices of the Center for International Blood and Marrow Transplant Research (CIBMTR), to develop an updated, unified international set of vaccine guidelines for autologous and allogeneic transplant recipients. These guidelines differ from prior guidelines primarily by 1) inclusion of the 7-valent protein conjugated pneumococcal vaccine (PCV7) (3 monthly doses) in all patients starting at 3–6 months post HCT, followed by PPV23 in patients without chronic GVHD 2) use of the live varicella vaccine in select patient groups starting at 24 months post HCT, and 3) optional use of vaccines licensed since 2005 such as the tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis vaccine (Tdap) for adolescents and adults, the recombinant human papilloma vaccine 42, and the protein conjugated meningococcal vaccine 43. An important limitation of both current and past HCT vaccine guidelines is the recommendation that all patients should be vaccinated at fixed times post HCT. This does not take into account the variable kinetics of immune reconstitution on the basis of age, donor, stem cell source 44–48 or use of monoclonal antibodies such as anti-CD20 49. In addition, guidelines can quickly become outdated due to licensing of new vaccines 50, 51, resulting in changes in recommendations from the Advisory Committee on Immunization Practices (ACIP) to better protect immunocompetent patients.
Multiple studies have demonstrated the superior response of patients re-immunized with 3 doses of tetanus toxoid, inactivated polio vaccine (IPV), and/or the protein conjugated Haemophilus influenza vaccine (HIB) compared to those given a single vaccine 52–59 as well as the benefit of donor immunization prior to stem cell harvest in the case of HIB, diphtheria, and tetanus toxoid 57–59. In addition, published reports have shown the poor or transient responses of HCT recipients to pure polyscaccharide vaccines such as PPV23 and MSV4 59–64 compared to the protein conjugated polysaccharide vaccines (ie HIB, Prevnar 7)65, 66, 67.
The results of a randomized, prospective multi-center trial performed in 158 patients (aged 5–65 years of age) was reported by Cordonnier et al in 200968. All patients received an allogeneic HCT and myeloablative conditioning. Approximately 65% of patients in both groups had no acute GVHD and limited or no chronic GVHD was seen in 56% and 33% of patients, respectively. The study demonstrated similar responses in both groups when tested one month after the third PCV7 dose (79% vs 82%, p=0.64), although the percentage of patients who still had positive titers to all 7 seroytpes at 24 months post HCT was significantly lower in patients vaccinated early vs later post HCT (59% vs 83%, p=0.013).. As Prevnar7 is no longer manufactured, a multi-center, prospective international trial of a series of three PCV13 immunizations followed by a booster Prevnar13 and the PPV23 in adults and children is currently underway.
Vaccination against hepatitis B is often mandatory for entry to school and required for certain jobs 69. In a study of 267 allogeneic transplant recipients immunized with rHBV following acquisition of pre-set milestones of immune competence, 64% seroconverted 70 including 73% of 99 children, and 59% of 168 adults (p = 0.02). In multivariate analyses, response was adversely affected by age >18 years (p<0.01) and history of prior chronic GVHD (p<0.0001) but not by donor type, use of T cell depletion, adoptive immunotherapy, or post transplant rituximab 70. Evaluation of serial titers in 98 patients demonstrated that 80 remained seropositive five years after their last vaccine. As the time to develop normal numbers of CD27+IgD+ memory B cells and CD4 helper cells varies among different allogeneic HCT populations 71, surrogate markers such as these may be instrumental in determining the need and timing of booster immunizations to maintain durable protection.
The majority of children will become seronegative for measles, mumps, and rubella by three years post HCT. King et al demonstrated that 68% of children given MMR at a median of 48 months post HCT responded to all three attenuated viruses 72. Also, there are limited data on the immunogenicity and durability of response to MMR following alternative donor HCT or response to mumps following any type of HCT 73. Of 113 children immunized with MMR at Memorial Sloan-Kettering Cancer Center following an HLA matched sibling (n=60) or unrelated (n=38) HCT, the response was 70% following a single MMR, including 42 recipients of an HLA matched related HCT and 27 recipients of an unrelated HCT (p=NS). There were no significant adverse reactions, even in non-responders 73.
There is widespread use of influenza vaccines, but there is limited data on response of HCT recipients to the killed influenza vaccine, particularly the H1N1 strain., Studies have shown a relatively poor response 74, 75. Although immunization with the inactivated influenza vaccine has been shown by Machado et al. to decrease the risk of influenza when administered >6 months post HCT (2 of 19 vaccinated vs 12 of 24 unvaccinated) 74, the majority of serious infections occur earlier post HCT, when vaccine response to influenza is very poor.
Since licensing of the live attenuated varicella vaccine (LAVV) in 1995, an increasing number of children, adolescents, and young adults undergoing HCT or donating stem cells have been protected against varicella by one or two doses of the LAVV. As the LAVV generally induces lower levels of antibody titers than natural infection, with titers that decrease over time post vaccination, two LAVV are recommended for VZ seronegative individuals76, 77. Only a limited number of studies have evaluated the LAVV in children post HCT. In the study by Sauerbrei, 15 children (median age 18 months) received LAVV following an autologous (n=7) or allogeneic (n=8) HCT78. Patients had a circulating lymphocyte count of >1000/ul, IgG >500 mg/dL, and a positive skin test to a recall antigen such as Candida. Studies have documented safety and seroconversion in >60% of evaluable patients 79, 80.
Protein-conjugated meningococcal vaccines. In 2010 81, a quadravalent protein conjugated meningococcal vaccine was licensed in the United States. Unlike the pure polysaccharide vaccine, this vaccines elicits a T cell dependent B cell response yielding the potential to induce long-term memory. There have been small studies showing either poor response, or rapid decrease in response following a single immunization 82.
Currently there are two vaccines containing tetanus, reduced diphtheria and acellular pertussis toxoids (Tdap) approved for individuals 10–55 years of age. These vaccines contain similar amounts of tetanus and diphtheria toxoid, but different amounts of pertussis toxoid (PT) (2.5 vs 8 mcg/dose). Response to a single TDAP is poor83, 84 in allogeneic or autologous HCT recipients likely due in part to the small amount of PT contained in the vaccine and the limited numbers of memory specific T and B cells transferred from the graft capable of a pertussis response.
For most patients, T cell reconstitution after transplant can be adequately monitored using a combination of lymphocyte subset analysis, “high content” immunophenotyping, and evaluation of T cell proliferative capacity to mitogens. In patients who demonstrate poor T cell reconstitution, persistent susceptibility to viral or fungal infections, or significant autoimmunity/GVHD, additional testing may be of benefit to more completely assess thymic T cell output, Treg numbers, T cell function, and TCR diversity.
As discussed, B cell responses to neoantigen return approximately 6 months after transplant, we therefore suggest that this is an appropriate time to begin to follow B cell reconstitution and maturation by “high-content” B cell immunophenotyping (including cell surface BAFF-R expression). Over time, a normal maturation of B cells from immature to mature naive then to a mature memory phenotype (CD27+) should be observed. Persistently poor B cell development with an absence of memory B cells, decreased cell surface BAFF-R levels, and high serum BAFF may be indicative of cGVHD8–10. In addition, these studies can help to identify those patients who may need further evaluation by neoantigen immunization and/or ongoing treatment with supplemental immunoglobulin.
Very valuable to assess functional immunity in patients who have evidence of poor or incomplete immune reconstitution (based on immunophenotyping) or who have chronic GVHD. Rabies vaccine is currently available clinically but Bacteriophage and KLH are available only on study protocols. Patients who demonstrate poor responses to neoantigen challenge should be considered for ongoing antimicrobial prophylaxis and immunoglobulin supplementation (IVIg, SCIg, etc.).
Given the increased use of unrelated and cord blood donors and of conditioning regimens of varying intensity over the last 10 years, it remains unclear whether vaccination according to fixed times post HCT will protect the majority of HCT recipients. All vaccine guidelines, including those published in 2009 24, acknowledge the variability in immune reconstitution following alternative donor transplants and the lack of studies assessing vaccine responses in these patient populations. Although vaccination at fixed times post HCT will likely be sufficient for highly immunogenic vaccines such as tetanus and polio, for others such as hepatitis B, pneumococcus, meningococcus, and varicella, some studies have suggested that vaccination prior to the acquisition of critical populations of T- and B- cells may hinder response rates and the durability of response. Current recommendations to immunize with live attenuated viral vaccines at 24 months are somewhat arbitrary, particularly since studies have shown earlier vaccination is safe. Earlier vaccination with live vaccines would be particularly useful in children returning to school during the first transplant year who are vulnerable to these infections.
Current vaccine guidelines therefore recommend yearly inactivated influenza vaccination in patients ≥6 months post HCT, strongly advocate immunization of household members and caretakers, and stress prompt evaluation of patients with respiratory symptoms so that early anti-viral therapy can be instituted in infected patients 24.
In 2011, the ACIP recommended a two dose series of conjugated meningococcal vaccine in individuals with functional or congenital asplenia, complement deficiency, or HIV infected individuals 83. Consideration of a similar two dose series administered two months apart should be evaluated in HCT recipients.
The 2009 CIBMTR guidelines recommended a series of Tdap containing 8 ug of pertussis toxoid/dose 24. The effectiveness of this strategy is currently being tested.
The increasing use of cord blood as an alternative HSC source requires additional research into the kinetics of immune reconstitution and its impact upon outcome. In addition, it is unknown as to whether cord blood outcomes may be predicted based upon varying subsets of immune recovery.
Several studies have suggested the value of assessing specific parameters of immune reconstitution as biomarkers for cGVHD and predictors of outcome. Recent studies have suggested B cell immunophenotyping and serum BAFF levels may be particularly informative. These studies should be extended to include more patients in order to validate these findings in a broader patient population. In addition, the increasing availability of Next generation sequencing technologies could be incorporated to allow the further study of polymorphic variants of Toll-like receptors (TLR) and other immunity-related proteins in transplant patients. Identified variants may be useful to stratify risk and possibly identify patients who require specific therapies (i.e. prolonged antifungal prophylaxis, etc.) in the post-transplant period.
In an ideal setting, neoantigen could be used to evaluate the reconstitution of functional immunity in every patient after transplant. Unfortunately, the cost (sometimes not covered by insurance) and the clinical inaccessibility of common neoantigens impedes their regular use. A prospective study of neoantigen responses in a wide variety of HCT scenarios should be a high priority to evaluate the clinical utility of this approach for identifying patients at risk of serious late infections and to determine how these tools may be most effectively used and made more broadly available.
Future vaccine trials should ideally include parallel assessments of in vitro parameters of immune reconstitution to determine whether surrogate markers of immune reconstitution can predict vaccine response. In vitro correlates of vaccine responses might allow earlier revaccination of patients with the requisite T and B cell populations and prevent premature vaccination and/or risk in patients unable to respond. Studies assessing the recombinant human papilloma virus vaccine, particularly in children with increased risk such as those with Fanconi anemia or certain primary immunodeficiency diseases are warranted as are trials evaluating earlier vaccination of children without GVHD against varicella, measles, mumps, and rubella. Current recommendations to immunize with live attenuated viral vaccines at 24 months are somewhat arbitrary, particularly since studies have shown earlier vaccination is safe. This would be particularly useful in children returning to school during the first transplant year who are vulnerable to these infections.
Funding for this work was made possible in part by the following National Institute of Health grants: 1R13CA159788-01 (MP, KSB), U01HL069254 (MP), R01 CA112530-05 (KSB), 5RO1 CA132110 (PS), and 5R01HL091749 (PS). The views expressed in this manuscript do not reflect the official policies of the Department of Health and Human Services; nor does mention by trade names, commercial practices, or organizations imply endorsement by the U.S. Government. Further support was provided by a generous grant from the St. Baldrick’s Foundation and the Lance Armstrong Foundation, as well as the following pharmaceutical companies: Genzyme, Otsuka America Pharmaceutical, Inc., and Sigma-Tau Pharmaceuticals, Inc. The content is solely the responsibility of the authors and does not necessarily represent the official views of those that provided funding.
Financial Disclosures: None
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.