Vaccine Responses in Adult Hematopoietic Stem Cell Transplant Recipients: A Comprehensive Review

Simple Summary Patients who recently received a stem cell transplantation are at greater risk for infection due to impairment of their immune system. In order to prevent severe infection, these patients are vaccinated after their stem cell transplantation with childhood immunization vaccines. Timing of this vaccination is important in order to be effective and obtain proper immune response. Postponement of vaccination would lead to better immune response but would also cause longer-lasting risk of infection. This review describes available data on the timing of vaccination and its vaccine responses. Optimal timing of vaccination might require an individualized approach per patient. Abstract Consensus on timing of post-hematopoietic stem cell transplantation (HSCT) vaccination is currently lacking and is therefore assessed in this review. PubMed was searched systematically for articles concerning vaccination post-HSCT and included a basis in predefined criteria. To enable comparison, data were extracted and tables were constructed per vaccine, displaying vaccine response as either seroprotection or seroconversion for allogeneic HSCT (alloHSCT) and autologous HSCT (autoHSCT) separately. A total of 33 studies were included with 1914 patients in total: 1654 alloHSCT recipients and 260 autoHSCT recipients. In alloHSCT recipients, influenza vaccine at 7–48 months post-transplant resulted in responses of 10–97%. After 12 months post-transplant, responses were >45%. Pneumococcal vaccination 3–25 months post-transplant resulted in responses of 43–99%, with the response increasing with time. Diphtheria, tetanus, pertussis, poliomyelitis and Haemophilus influenzae type b at 6–17 months post-transplant: 26–100%. Meningococcal vaccination at 12 months post-transplant: 65%. Hepatitis B vaccine at 6–23 months post-transplant: 40–94%. Measles, mumps and rubella at 41–69 months post-transplant: 19–72%. In general, autoHSCT recipients obtained slightly higher responses compared with alloHSCT recipients. Conclusively, responses to childhood immunization vaccines post-HSCT are poor in comparison with healthy individuals. Therefore, evaluation of response might be indicated. Timing of revaccination is essential for optimal response. An individualized approach might be necessary for optimizing vaccine responses.


Introduction
Hematopoietic stem cell transplantation (HSCT) is a promising and often the only curative option for patients with hematological malignancies [1]. Post-HSCT, the immune system is temporarily suppressed due to prior conditioning and usage of immunosuppressive medication, resulting in infection-derived complications being a major cause of transplant-related mortality [2]. Furthermore, pre-HSCT established immunity against

Data Extraction and Data Synthesis
To compare studies on both timing of vaccination and vaccine responses, data were extracted and tables were constructed per vaccine, displaying vaccine response as either seroprotection (meaning an antibody titer above vaccine-specific cutoff value) or seroconversion (conversion from a vaccine-specific negative antibody titer to a positive titer). Response rates after complete vaccine schedule are shown, unless otherwise indicated. Results on alloHSCT and autoHSCT are reported in separate paragraphs and each vaccine is discussed separately. Results are summarized in a table for both groups separately. Studies on both alloHSCT and autoHSCT recipients are reported in both paragraphs if results were available for separate subgroups of the study population; if only data on the total study population were reported, the study is reported in only one paragraph.

Post-Hoc Analysis
Due to rapidly evolving research on COVID-19 vaccines in HSCT patients and their relevance for these patients at this moment, an analysis of these studies was added after completion of the original study.

Allogeneic Hematopoietic Stem Cell Transplantation
All results on alloHSCT patients are shown in Table 1. Start of vaccination is displayed in relation to moment of HSCT. N = number of vaccinated post-HSCT patients in the study population. * Response rates are for serotype A/H1N1 unless otherwise indicated. ** Respectively, serotypes A/H1N1, A/H3N2, B. *** 2 autologous stem cell recipients included in this percentage. **** 26.2% of study population < 18 years old. ***** Percentage is based on the whole study population consisting of 25% of patients < 18 years old. ****** 9.5% < 16 years old divided equally among both study arms; 101/158 patients received PPV23 vaccine. ******* 8 autologous stem cell recipients included in these percentages.

Allogeneic Hematopoietic Stem Cell Transplantation
All results on alloHSCT patients are shown in Table 1.

Influenza
The influenza vaccination in alloHSCT recipients has been widely studied: three RCT's and eight cohort studies were identified. Response rates varied from 10% to 97%. First, administration of single influenza vaccine is discussed. Several studies reported that the timing of vaccination influenced the subsequent vaccine responses. However, an RCT was conducted comparing early single-dosage vaccination to late double-dosage vaccination, which did not show any differences in response for both timing and dosages [20]. Results from observational studies are more conflicting: Avetisyan [12] reported no differences in response between early-vaccinated and late-vaccinated patients, whereas others showed improved responses upon extending the delay of the administration of the influenza vaccination [13,19]. For example, Issa [13] reported improvement of response for every additional year of interval between transplantation and vaccination (OR 1.79 (1.12-2.85)). This was supported by De Lavallade [19], who reported that the patients vaccinated 48 months post-transplant gained seroprotection, whereas those who were vaccinated median 6.5 months post-transplant did not (p = 0.015). In addition to timing, usage of immunosuppressive medication has an influence on vaccine response: use of rituximab during the year prior to vaccination negatively influenced response [13]. Of note, post-HSCT patients were considered to have a delayed vaccine response compared with healthy controls: Mariotti measured 54% response at 28 days post-vaccination versus 71% response at 90 days post-vaccination [15].
Several studies compared different types of influenza vaccines. Natori [21] performed an RCT comparing the adjuvant vaccine and non-adjuvant vaccine; they found no difference. An RCT on high dose (HD) influenza vaccine as compared with standard dose (SD) showed no benefit of the HD vaccine [22].
Multiple studies administered two influenza vaccines. De Lavallade [19] reported improvement in response from 46% after the first vaccine to 73% after the second. Dhédin [16] and Mohty [18] supported this benefit of a second vaccine, but Karras [20] and Roll [17] did not. These differences might be explained by GVHD rates and timing. GVHD rates in the study by Roll [17] were much higher (36.8%) compared with the studies that demonstrated benefit of booster vaccination (De Lavallade 0% [19]; Mohty 15.4% [18]). Additionally, timing of vaccination was earlier in the study by Roll [17]: approximately 15 months versus 36 and 30 months, respectively [18,19], which may have influenced booster efficacy.
In conclusion, the data suggest that influenza responses in alloHSCT recipients is increased by postponing vaccination and that booster vaccination might be effective, whereas usage of adjuvants or high vaccine dosages does not improve vaccine responses.

Pneumococcal
For pneumococcal vaccination in alloHSCT recipients, seven cohort studies and one RCT were included. There was huge variation in the number of vaccine dosages. Timing varied from 3 to 25 months post-transplant and responses varied from 43% to 99%.
To assess an optimal timing for vaccination, Cordonnier [29] performed an RCT to compare early (arm I: 3 months post-HSCT) to late (arm II: 9 months post-HSCT) vaccination using three conjugate vaccines (Prevenar PCV13 at 0, 1, 2 months) and one polysaccharide vaccine (Pneumovax PPV23 at 9 months). Responses were slightly better in arm II, which could possibly be explained by a higher rate of active GVHD in arm I (22.7% arm I versus 16.9% arm II). However, increase in response with increasing interval between HSCT and vaccination was supported by Langedijk [28]. Winkler [44], Meerveld-Eggink [41] and Shah [42] found comparable responses to Cordonnier [29] also using three conjugate vaccines (PCV7 or PCV13). Of note, Winkler [44] reported an improvement in measured vaccine response with time post-vaccination: measurement at 6 months post-vaccination showed seroconversion rates of 65-91% versus 55-67% at 1 month post-vaccination.
Usage of varying number of vaccines was studied. Usage of only one vaccine resulted in lower responses: 43-44.7% [26,40]. More recently, Cordonnier [27] evaluated immunogenicity of a fourth dose of PCV13 and concluded that this did not improve response. However, they did report better overall responses compared with their study in 2010. Some differences in study population might have played a role. The GVHD-rate was lower in 2015 (4.3% versus 16.9-22.7% in 2010) and conditioning included less myeloablative regimens (54.9% versus 79.6% in 2010).
Overall, best vaccine responses were obtained when three conjugate vaccines were administered. Vaccination initiated at 3 months post-HSCT provided adequate vaccine responses; however, postponing vaccination resulted in higher responses.

Diphtheria, Tetanus, Pertussis (DTP) and Poliomyelitis
For diphtheria, tetanus, pertussis and poliomyelitis vaccination in alloHSCT recipients, four studies were included. Tetanus responses were 52-100% starting vaccination at least 6 months post-alloHSCT. Shah [42] and Meerveld-Eggink [41] reported better responses with respect to less acute GVHD and a prolonged interval between HSCT and vaccination. However, Conrad [43] and Winkler [44] showed higher response rates at 6-7.5 months post-transplantation, questioning the influence of interval between HSCT and vaccination. Winkler reported a worse seroconversion rate for polio, which is probably due to their already higher seroprotection rate pre-vaccination of 46% [44]. Conclusively, DTP and poliomyelitis vaccines might be administered at 6 months post-transplant.

Haemophilus influenzae Type b (Hib)
Hib vaccination in alloHSCT recipients was evaluated in five cohort studies. All studies showed adequate responses of 77-97%. On short-term immunity, no benefit was found for administration of a booster vaccine [40][41][42][43][44]. Vaccination might be initiated at 6 months post-transplantation.

Meningococcal
For meningococcal vaccination in alloHSCT recipients, two studies were identified, using meningococcal ACYW vaccine. Cheng [32] reported responses of 54-83% per serotype after one vaccine at 12 months post-HSCT.

Hepatitis B
On hepatitis B vaccination in alloHSCT recipients, four studies were included. Onozawa [33] started vaccination at 12 months post-transplantation, and Takahata [34] started vaccination after cessation of immunosuppressive therapy-on average 15 months post alloHSCT. Both studies found comparable response rates around 40% after three vaccines. Jaffe [35] started their vaccination later-namely, 23 months post-transplant-and achieved better responses using the same vaccination schedule. Conrad [43] started earlier, at 6 months post-transplant and found a response rate of 84%. However, 56% of their population was already seroprotected before vaccination, which makes comparison more difficult. Of note, Jaffe [35] concluded that GVHD was associated with a significantly poorer response.
Conclusively, hepatitis B vaccination achieved poorer responses compared with all other vaccinations and should not be administered before 12 months, as responses seem to improve with time post-transplantation.

Measles, Mumps and Rubella (MMR)
For MMR vaccination responses, two studies were included. Aoki [36] administered two doses of vaccine at median 69.3 months post-HSCT. All patients were seronegative before vaccination and seroconversion was found in only 19-30%. Evaluating their responses, chronic GVHD was negatively associated with seroconversion (OR 0.06; p = 0.023). This influence of GHVD is supported by Kawamura [37]. However, though vaccinating with only one dose and earlier post-transplantation (median 41.7 months post-HSCT), their seroconversion rates were remarkably better: 36-72%. This might partially be explained by different cutoff values for seroprotection.

Varicella Zoster Virus (VZV)
For VZV vaccination, one study was included with Shingrix vaccine. This study only studied 17 alloHSCT patients [38]. Their response rate was 18%, which they compared with an autologous population (n = 13; response 62%). The only influencing factor in their multivariable analysis was type of HSCT, thus allogeneic or autologous (OR = 0.08; p = 0.01).

Autologous Hematopoietic Stem Cell Transplantation
All results on autoHSCT patients are shown in Table 2. Start of vaccination is displayed in relation to moment of HSCT. N = number of vaccinated post-HSCT patients in the study population. * Response rates are for serotype A/H1N1 unless otherwise indicated. ** Autologous patients included: 17.6% [23], 29.5% [24]; results are on total study population. *** Only two autologous stem cell recipients included in this percentage. **** Respectively serotypes A/H1N1, AH3N2, B. ***** Allogeneic recipients median 34 months post-transplant, autologous recipients median 20 months post-transplant.

Influenza
Data on influenza vaccination post-autoHSCT is scarce. Two studies reported only on autoHSCT recipients, of which one was an RCT by Villa [25]. This RCT compared single to double dosage vaccination at 1-year post-HSCT and reported no benefit from a second vaccine. Another study reported on autoHSCT and alloHSCT separately: Three autoHSCT recipients were included and vaccinated 2 years post-autoHSCT with response rates of 50-100% [14].
Gueller [23] and Engelhard [24] both reported on both autologous and allogeneic HSCT recipients. Respectively, 17.6% and 29.5% of the patients were autoHSCT recipients, with an average timing of vaccination of 20 months post-HSCT. Response rates were fairly low: 33-53% after one vaccine. Gueller reported benefit from booster vaccines improving response up to 91% [23], whereas Engelhard did not [24]. Assessing predictors for inadequate vaccine response, Gueller [23] concluded that usage of immunosuppressants caused failure in response: 88.9% of their non-responders used immunosuppressants versus 0% of their responders.
When comparing the autoHSCT population with the alloHSCT, some differences were reported. Yalçin [14] reported a better response in their autoHSCT recipients, and this is supported by Gueller [23]: they reported 71.9% of their seroconverters to be post-alloHSCT, whereas from their non-seroconverters 90% was post-alloHSCT.

Pneumococcal
Only two studies on pneumococcal vaccination in autoHSCT recipients were included, with responses of 58-78% with vaccination started at 6-12 months post-HSCT [45,46]. Conclusively, the timing of this vaccination seems to make no difference and might therefore be initiated at 6 months post-autoHSCT.

Diphtheria, Tetanus, Pertussis (DTP)
On diphtheria, tetanus and pertussis in autoHSCT patients, three studies were included. One study administering three dosages of DTP vaccine showed 100% response for tetanus, but all recipients appeared to already have protective antibody titers prevaccination [45]. The second study, also using three dosages, showed a seroconversion rate for tetanus of 60%, diphtheria 70% and pertussis 76%, with the majority of non-responders already being seroprotected pre-vaccination [46]. The final study vaccinated with a single DTP vaccine and reported low responses (7.1-32.1%). This study concluded that multiple vaccinations are required for adequate response [30]. Vaccination might thus be initiated at 6-12 months post-autoHSCT with preferably three vaccines to be administered.

Haemophilus influenzae Type b (Hib)
For Hib vaccination in autoHSCT recipients, three studies were included. Both studies by Van der Velden showed high responses for vaccination at 6 months post-HSCT: 94-100% [31,45]. Palazzo found a lower seroconversion of 71%, yet 24% of their study population already had protective titers pre-vaccination [46].

Meningococcal
Only one study on meningococcal vaccination in autoHSCT recipients was included. Cheng [32] reported comparable seroconversion rates for their alloHSCT and autoHSCT patients and suggested that, despite a response of 69-85% after one vaccine, a booster might be beneficial.

Hepatitis B
On hepatitis B vaccination in autoHSCT patients, one study was included: Palazzo reported 40% seroconversion after three vaccinations [46].

Varicella Zoster Virus (VZV)
For VZV vaccination, two studies were included with Shingrix vaccine. Both reported similar response rates, despite variety in moment of vaccination. Stadtmauer [39] included many more patients than Camargo [38] (n = 922 versus n = 13) and evaluated responses per age group with non-significant differences in response (age 18-49 years old, 57.7%; age > 49 years old, 71.4%). Conclusively, vaccination early post-autoHSCT might be feasible.

Post-Hoc Analysis: COVID-19
For studies on COVID-19 vaccines in post-HSCT patients, a separate search was conducted on 22 November 2021 using the original search strategy (Supplementary Table  S1) with the addition of MeSH term 'COVID-19' or Title/abstract-term 'covid*' or 'corona*' through the Boolean operator 'AND'. This search resulted in 77 results. These results were

Discussion
This study intended to provide an overview of current knowledge on the optimal timing and schedule of post-HSCT vaccination. A total of 35 studies with varying vaccination schedules were included: 24 on alloHSCT, 5 on autoHSCT and 6 addressing both. In general, results are very heterogenic, and integrated studies are scarce.

Allogeneic Hematopoietic Stem Cell Transplantation
Drawing a firm conclusion based on the current literature is impossible due to the heterogeneity of the studies. Nevertheless, a few definitive points can be taken: 1. Many post-HSCT patients respond inadequately to vaccination, and responses are remarkably low compared with the healthy population. Thus, evaluating responses post-vaccination might be indicated. Suggested antibody titers for this evaluation in clinical practice are provided in Supplementary Table S5, including suggested cutoff values. 2. Most studies showed improved vaccine responses when delaying revaccination for at least 6 months post-transplantation and thereby allowing better immunological recovery [8,9]. However, postponing vaccination increases risk of infection: the estimated incidence of invasive pneumococcal disease (IPD) is 347 infections per 100,000 alloHSCT recipients compared with only 7 per 100,000 persons in the general population [28]. Timing of revaccination is a balance between immune recovery versus infection risk. Insights in individual immune recovery would help to assess the risk of failure. The factors that influence the immune system and the extent of influence are unknown.
For specific vaccines, some comments can be made in alloHSCT recipients: 1. Influenza vaccine studies were extremely heterogeneic. Responses seemed to improve with postponement of vaccination. Most studies initiated their vaccination at least 12 months post-transplantation, and therefore, no proper recommendation can be made on earlier vaccination. However, in case of a pandemic such as we experienced with COVID-19, a less adequate response to vaccination, and thus poorer seroprotection, would be preferable over no seroprotection [58]. Therefore, one might then consider vaccinating all patients that are at least 3 months post-HSCT and clinically stable. Earlier post-HSCT and vaccinating family members with close contact to the patient might be considered [4]. 2. Pneumococcal vaccines might already be effective at 3-6 months post-transplant [27,29]. Earliest administration was three months post-transplant with response > 68%. However, responses did increase when vaccination was postponed: vaccination at 7-12 months post-transplant resulted in 69-99% responses [28,41,42,44]. 3. Diphtheria, tetanus, pertussis, poliomyelitis and Hib were earliest administered at 6 months post-transplant. Responses for tetanus varied widely (52-100% [41][42][43][44]); for Hib responses were slightly more comparable between studies: 77-97% [40,44]. Vaccination on DTP, poliomyelitis and Hib might be started at 6 months post-transplantation. 4. Hepatitis B vaccine was administered 6-23 months post-HSCT with varying responses: 40-84% [33][34][35]43]. In the general population, responses to hepatitis B vaccine are also quite low compared with other vaccines [47,59]. Therefore, it is advised not to initiate vaccination before 12 months post-HSCT. 5. Vaccine responses to MMR remain suboptimal multiple years post-HSCT, and vaccination must be considered carefully per individual. As the vaccine is live-attenuated, the clinical condition of the individual patient is very important. Vaccination must be postponed in case of active GVHD and/or usage of immunosuppressants. 6. VZV vaccination has not yet been studied widely, and thus, no proper recommendation on timing can be made. However, with the recombinant subunit vaccine (Shingrix ® ), potential severe side effects of the live-attenuated VZV vaccine are no longer an issue, and vaccination can be considered earlier post-HSCT. Finally, COVID-19 vaccine responses are poor compared with the healthy population. However, vaccination should be considered shortly post-HSCT, as any chance for response must be taken in times of a pandemic such as COVID-19.

Autologous Hematopoietic Stem Cell Transplantation
For vaccination of autologous HSCT patients, fewer data are available. Although guidelines [3][4][5][6][7] recommend vaccination post-autoHSCT, data on vaccinating this population are scarce. One can postulate that responses post-autoHSCT might be better because of maintenance of pre-transplant-obtained immunity in post-autoHSCT recipients [60]. For post-autoHSCT patients, revaccination is not always general practice despite it be-ing recommended in current guidelines, similar to guidelines for post-alloHSCT patients. Data on vaccination of post-autoHSCT patients are scarce; however, responses seem to be slightly better in this group compared with post-alloHSCT patients. A more tailored approach might therefore be indicated. However, one could question whether that would be cost-effective.
An important factor to consider for this more individual approach in autoHSCT patients would be the original indication of the transplantation. Additional immunosuppressive treatment post-autoHSCT might decrease vaccine responses [61]. For example, rituximab maintenance in mantle cell lymphoma likely has a significant suppressive effect due to continuous B cell depletion. Likewise, lenalidomide maintenance post-autoHSCT for multiple myeloma patients could have an impact on vaccine responses. On the one hand lenalidomide can induce leukopenia, but it is also associated with immune modularity properties resulting, for example, in increased incidence of GVHD [62]. The current evidence of Palazzo et al. [46] does however not suggest a significant immunosuppressive effect of lenalidomide on vaccine response. Any further supporting evidence with respect to live attenuated vaccines in the setting of lenalidomide would be helpful. Of note, some guidelines still suggest avoiding these live vaccines whilst on lenalidomide maintenance post-autoHSCT beyond two years. However, interestingly, recommendations for non-autoHSCT patients on similar medication are lacking. Nevertheless, we recommend being cautious with live attenuated vaccines in patients with multiple myeloma because not only medication but also the disease itself as well as aging contribute to the risk of vaccine-related disease.

Overall Discussion
Some general recommendations can be made: 1. Physicians may need to take into account the current state of immunosuppression and associated expectations for immune recovery to determine the precise timing and the estimated success of vaccination post-HSCT. 2. Future studies with standardized definitions, especially on vaccine response (seroconversion/seroprotection), are necessary to increase generalizability and relevancy of results. 3. More insight in predictors of vaccine response, such as immune reconstitution, GVHDstatus and immunosuppressive therapy [61], might help to design an optimized and probably more individualized vaccination schedule. For example, rituximab is known to interfere with immune recovery for 6-12 months after cessation [63] and was reported to negatively influence vaccine response [13]. Therefore, postponing vaccination is necessary in these specific patients. 4. Finally, though evaluating vaccine responses post-vaccination is highly recommended, it is important to note that antibody titers might still increase with time due to possibly delayed immunization in the post-HSCT population [15,44].
The literature search was performed extensively and considered to be as complete as possible. Accuracy was assured by studying each vaccine separately and also by displaying both seroprotection and seroconversion rates to allow optimal comparison among data.
To ensure the quality of this study, great effort was made to maintain proper validity and reliability throughout all included studies. However, there were some limitations. First, only antibody titers were evaluated as a measure of vaccine response, which might underestimate vaccine effectiveness: patients might still benefit from vaccination despite low antibody responses [64]. Furthermore, only the effect of the first vaccination post-HSCT was evaluated, and no conclusion can therefore be drawn on effectiveness of booster vaccination. Finally, as only childhood immunizations were studied, additional vaccines might be indicated regionally, such as the hepatitis A or yellow fever vaccine.

Conclusions and Recommendations
In conclusion, vaccine responses to the childhood immunization vaccines in post-HSCT patients are lower as compared with healthy individuals, and therefore, measurement of response might be indicated. Furthermore, timing of vaccination is essential to reach optimal responses. Considering the influence of patient factors such as GVHD and usage of immunosuppressive drugs, an individualized approach might be necessary to optimize vaccine responses. However, guidance on how to initiate this individualized approach is currently lacking and further studies are needed.
Based on the included studies, general recommendations on timing of the childhood immunization vaccines post-alloHSCT are made and shown in Table 4. Due to overall lower response rates compared with healthy populations [47], assessment of response after vaccination is recommended. For autoHSCT recipients, no recommendations could be formed based on our review due to the scarcity of data.    Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest.