Recombinant zoster vaccine uptake and series completion among immunocompromised adults with cancer in the US

Abstract

Purpose

To estimate recombinant zoster vaccine (RZV) use among adults with cancer since the 2021 US Advisory Committee on Immunization Practices (ACIP) recommendation to vaccinate immunocompromised adults aged ≥ 19 years and gather insights into oncologist practices, perceptions, barriers, and facilitators to RZV vaccination among adults with cancer.

Methods

In this retrospective cohort study, RZV uptake (≥ 1 dose) and completion (2 doses) were assessed among RZV-naïve, immunocompromised patients aged ≥ 19 years with a solid tumor cancer or hematologic malignancy using IQVIA’s open medical and prescription claims. Patients were followed from the ACIP voting date (October 20, 2021) until June 30, 2023. RZV uptake, series completion, time to completion, and dosing schedule compliance were described. Generalized estimation equation models assessed factors associated with uptake. Fifteen oncologists were interviewed regarding their RZV practices, perceptions, barriers, and facilitators, and themes were generated based on responses.

Results

Among 388,923 and 277,314 RZV-naïve patients with a solid tumor cancer or hematologic malignancy, cumulative RZV uptake increased gradually, reaching 7.6% and 8.6%, respectively. Series completion rates at 6 and 12 months among those who initiated RZV were 63.2% and 70.9% in the solid tumor cancer cohort and 64.7% and 71.6% in the hematologic malignancy cohort. The odds of RZV uptake varied by cancer type and tended to increase with age and household income. Oncologists reported that they primarily serve an advisory role in vaccination due to workflow and financial challenges and competing patient needs, opting instead to delegate this responsibility to primary care providers and pharmacists.

Conclusion

RZV uptake among RZV-naïve immunocompromised patients with cancers has been slow and suboptimal since the 2021 ACIP recommendation, but series completion is likely once initiated.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12885-025-15530-w.

Prior presentation

1. Gatwood J, McGuiness C, Yasuda M, et al. Recombinant Zoster Vaccine Uptake among US Adults with Solid Tumor Cancers. Poster presented at NFID ACVR, 8–10 May 2024.

2. Gatwood J, McGuiness C, Yasuda M, et al. Recombinant Zoster Vaccine Uptake among US Adults with a Hematological Malignancy. Poster presented at NFID ACVR, 8–10 May 2024.

Context

Key objective

To examine the impact of the US Advisory Committee on Immunization Practices 2021 recommendation (i.e., to administer recombinant zoster vaccine [RZV] to immunocompromised adults aged ≥ 19 years) among adults with cancer.

Knowledge generated

Among adults aged ≥ 19 years with cancer who had not already received RZV prior to the Advisory Committee on Immunization Practices 2021 recommendation, RZV uptake was gradual and only reached 3.5–9.9% and 5.6–9.3% among patients with solid tumor cancers and hematologic malignancies, respectively, through June 2023. Among those who received an initial RZV dose, two-dose series completion rates were approximately 65% and 70% at 6 and 12 months, respectively.

Relevance

Immunocompromised adults with cancer are recommended to receive RZV due to an increased risk of herpes zoster. However, new uptake since the recommendation is suboptimal, hence the importance of vaccination in this population should be emphasized.

Introduction

Approximately 40% of Americans develop cancer during their lifetime and it has been estimated that over 17 million people were living with cancer in the United States (US) in 2021 [1]. Patients with cancer are at increased risk of herpes zoster (HZ), with incidence estimates of 12–14 per 1,000 person-years (PY) among adults with solid tumor cancers [2, 3] and 31–66 per 1,000 PY among adults with hematologic malignancies [2, 4], compared to approximately 3–12 per 1,000 PY (depending on age) in the general adult population [5]. The recombinant zoster vaccine (RZV) has been shown to be immunogenic and have an acceptable safety profile among patients with solid tumor cancer receiving immunosuppressive chemotherapy [6] and patients with a hematologic malignancy, in whom 87% efficacy was shown in a post-hoc analysis of phase 3 clinical trial data [4]. Based on these and other studies (including one in hematopoietic stem cell transplantation recipients) [7], the US Advisory Committee on Immunization Practices (ACIP) recommended RZV for the prevention of HZ in immunocompromised adults aged ≥ 19 years in 2021 [8], following its recommendation in immunocompetent adults aged ≥ 50 years in 2017 [9]. The 2021 recommendation [8] includes patients with cancer, in whom RZV should ideally be given before immunosuppressive treatments, chemotherapy, or radiotherapy are initiated [10]. If this is not possible, RZV can be given when the immune response is likely to be sufficiently robust [10]. Based on the ACIP recommendations, the 2022 National Comprehensive Cancer Network (NCCN) [11] and 2024 American Society of Clinical Oncology (ASCO) [12] guidelines also recommend that people with cancer aged ≥ 19 years receive RZV.

Estimates of RZV uptake among adults with cancer are limited. One study reported an RZV initiation rate of 59.3 per 1,000 PY among immunocompromised patients with cancer (vs. 48.1 per 1,000 PY among immunocompetent individuals) aged 50–64 years during 2018–2019 [13]. Another reported that approximately 9% of individuals aged ≥ 50 years with active malignancy received two RZV doses during May 2019 through May 2022, which was similar to the rate among all individuals aged ≥ 50 years [14].

To our knowledge, no studies have examined the change in RZV use among immunocompromised adults with cancer since the 2021 ACIP recommendation. Consequently, the objectives of this study were to estimate RZV uptake, series completion, and dosing schedule compliance among immunocompromised US adults with cancers since the 2021 ACIP recommendation, and to identify factors associated with RZV uptake and summarize vaccination practices, perceptions, barriers, and facilitators as described by oncologists.

Methods

Overall study design

The study included quantitative and qualitative components. RZV uptake, series completion, and dosing schedule compliance were evaluated quantitatively using a retrospective design. The qualitative portion involved interviews of US oncologists to gain perspective on RZV practices, perceptions, barriers, and facilitators. The study complied with all applicable subject privacy laws, and the procedures for conducting the interviews were reviewed and approved by the Western-Copernicus Group Institutional Review Board (approval number 20241209).

Quantitative methods

The overarching quantitative study included patients with a variety of immunocompromising conditions [15], of which cancer was the most common and, as a result, became the focus of this substudy. This portion of the study used IQVIA’s open medical and prescription claims linked to OneKey healthcare provider profile data, with supplemental sociodemographic data from Experian ConsumerView^SM (Appendix Text A1, online only). The study start date (April 20, 2021) was 6 months before the ACIP recommendation voting date [8] (October 20, 2021 [index date]) and the end date was June 30, 2023 (Appendix Fig A1, online only). The 6 months up to and including the index date was used to identify immunocompromised patients with specific cancers (detailed below). Patients were followed from the index date until the earlier of a second RZV dose or the end of the data period (i.e., June 30, 2023 or the latest available data).

Inclusion Criteria. Patients had to meet the following criteria: (1) be aged ≥ 19 years at index; (2) have ≥ 1 pharmacy and ≥ 1 medical claim from stable pharmacy and provider sources (i.e., those that provided monthly data to IQVIA) during the study period; (3) have valid birth year, sex, region, and payer type data; (4) have no evidence of RZV during each patient’s claims history (all available to October 19, 2021); and (5) during 6-month baseline, have a claim related to (a) an included solid tumor cancer (bladder, breast, colorectal, endometrial, kidney, liver, lung, pancreatic, prostate, skin [melanoma], or thyroid) and evidence of immunocompromising medication use or (b) an included hematologic malignancy (acute lymphoblastic leukemia, acute myeloid leukemia, chronic lymphocytic leukemia, chronic myeloid leukemia, Hodgkin lymphoma, multiple myeloma, myelodysplastic syndrome, or non-Hodgkin lymphoma); and if they had myelodysplastic syndrome, evidence of immunocompromising medication use (Appendix Table A1, online only [cancer diagnosis codes] and Appendix Table A2, online only [immunocompromising medications]). Patients were assigned to one of these two mutually exclusive cohorts (solid tumor cancer and hematologic malignancy) based on the first observed included cancer.

Variables and Outcomes. Patient demographic data included age (at index), sex, payer type (based on most recent claim), census region, Charlson Comorbidity Index (CCI) score (excluding cohort condition [Dartmouth-Manitoba adaptation]), cancer type, pharmacy density, and Area Deprivation Index [16]. Among patients who were linkable to the supplemental sociodemographic consumer database, extracted data elements included race/ethnicity, education, and household income.

Key outcomes were RZV uptake (≥ 1 dose), 6- and 12-month series completion (among those who received a first dose ≥ 6 months before the end of the study), time to completion (among those who received 2 doses), and dosing schedule compliance (those who received a second dose 1–6 months after a first dose as per the ACIP recommendation [8] among those who received 2 doses). RZV was identified using National Drug Codes and Current Procedural Terminology codes.

Among patients who received ≥ 1 dose of RZV, their place of service (pharmacy or physician office) and co-administration with influenza, pneumococcal, or tetanus-diphtheria-pertussis (Tdap) vaccines was captured. Pharmacy store type and various physician characteristics (age [at December 31, 2022], census region, urban or rural area classification, Integrated Delivery Network classification, patients attributed to the provider, practice size, years since graduation [difference between graduation year and 2022], degree, and specialty) were also reported.

Statistical Analyses. Patient and provider characteristics were described as counts (%) or means (standard deviations [SD]) as appropriate. RZV uptake was reported as the percentage of patients who received ≥ 1 dose of RZV from October 21, 2021 through June 30, 2023 in the two cohorts and for each cancer subtype. Six- and 12-month series completion rates were reported as the percentages of patients who received a second dose within 6 and 12 months of the first dose among those who received a first dose by December 31, 2022. This cut-off date was chosen to allow ≥ 6 months of potential follow-up time, although the open claims data source did not allow the requirement of continuous follow-up so some patients may have had less follow-up time. Time to completion was estimated by Kaplan-Meier analysis and reported as the median (95% confidence interval [CI]) time between the first and second doses among those who received ≥ 1 dose by December 31, 2022. RZV dosing schedule compliance was reported as the percentage of patients who received their second dose 1–6 months after their first dose among those who received 2 doses.

Generalized estimation equation models assessed factors associated with RZV uptake among patients with available sociodemographic data. These models, which used binomial distributions and logit link functions, provided odds ratios (ORs) with 95% CIs. Clustering effects at the provider level were employed to account for correlations of the outcomes within the clusters. The outcome variable was receipt of ≥ 1 RZV dose (yes/no), with predictor variables including age, sex, payer type, cancer type, CCI score (excluding cohort condition), race/ethnicity, education, urban/rural setting, and household income.

Qualitative methods

Data Sources. Oncologists identified via IQVIA’s Research Network database were interviewed during two waves of 60-minute interviews during April 12–23 and June 5–12, 2024 (Appendix Text A2, online only). The interview guide was developed specifically for this study and its two main sections covered RZV vaccination practices/perceptions and barriers/facilitators to RZV vaccination (Appendix Text A3, online only [interview guide]).

Inclusion Criteria. Invitations with a link to a preliminary survey and consent were emailed to providers in the IQVIA Research Network database. Included participants were required to: (1) be licensed to practice in the US; (2) indicate hematology or oncology as their primary specialty; (3) treat patients with cancer; (4) work primarily in an outpatient setting; (5) respond to a brief questionnaire about personal and practice characteristics; (6) be willing to be recorded while responding to questions related to vaccination practices; and (7) be able to communicate in English. Those deemed eligible were contacted thereafter to arrange the interview. Recruitment continued until responses reached saturation according to the study team.

Respondent Characteristics. Provider characteristics collected prior to the interview included sex, age, race/ethnicity, board certification status, and number of years in practice. Practice characteristics included geographic details, healthcare organization size, academic and for-profit status, average daily patient load, and distribution of patients by race/ethnicity and insurance type.

Data Analysis. Respondent characteristics were presented descriptively. The transcribed interview data from all sessions were analyzed and thematically assessed by multiple researchers. Each theme and overall insights were provided, including representative quotes by theme. Ideas were classified as uncommon, common, and very common based on responses from 1 to 5, 6–10, and 11–15 oncologists (out of 15), respectively.

Results

Quantitative results

Patient Populations. In total, 388,923 and 277,314 patients met the full inclusion criteria, had one of the included cancers, and comprised the solid tumor cancer and hematologic malignancy cohorts, respectively (Appendix Fig A2, online only). The most common solid tumor cancers were breast (28.8%), lung (17.2%), and prostate (17.2%), while the most prevalent hematologic malignancies were non-Hodgkin lymphoma (42.2%), multiple myeloma (18.7%), and chronic lymphocytic leukemia (18.4%) (Table 1). Most patients were aged ≥ 50 years (89.3% in the solid tumor cancer cohort and 88.5% in the hematologic malignancy cohort, ranging from 44.0% to 97.5% by cancer type [Appendix Table A3, online only]). Overall, approximately half were male (43.6% and 52.0%, respectively) and most had a CCI score (excluding cohort condition) ≥ 2 (73.4% and 62.3%, respectively). Sociodemographic characteristics among patients who were linkable to the consumer database are detailed in Appendix Table A4, online only.

Table 1.

Patient characteristics

	Solid tumor cancers (n = 388,923)	Hematologic malignancies (n = 277,314)
Age, years, mean (SD)	66.4 (12.7)	67.5 (14.5)
Age, years, n (%)
19–29	2,661 (0.7)	6,902 (2.5)
30–39	10,335 (2.7)	9,140 (3.3)
40–49	28,560 (7.3)	15,942 (5.7)
50–59	61,982 (15.9)	33,976 (12.3)
60–64	50,771 (13.1)	28,995 (10.5)
≥65	234,614 (60.3)	182,359 (65.8)
Sex, n (%)
Male	169,591 (43.6)	144,302 (52.0)
Female	219,332 (56.4)	133,012 (48.0)
CCI score excluding cohort condition,a n (%)
0	66,486 (17.1)	69,688 (25.1)
1	37,112 (9.5)	34,956 (12.6)
≥2	285,325 (73.4)	172,670 (62.3)
Cancer type, n (%)
Breast	112,007 (28.8)	–
Lung	66,975 (17.2)	–
Prostate	66,778 (17.2)	–
Colorectal	47,917 (12.3)	–
Bladder	22,580 (5.8)	–
Kidney	16,482 (4.2)	–
Thyroid	13,267 (3.4)	–
Pancreatic	12,525 (3.2)	–
Skin (melanoma)	11,427 (2.9)	–
Endometrial	10,920 (2.8)	–
Liver	8,045 (2.1)	–
Non-Hodgkin lymphoma	–	117,018 (42.2)
Multiple myeloma	–	51,855 (18.7)
Chronic lymphocytic leukemia	–	51,116 (18.4)
Chronic myeloid leukemia	–	17,629 (6.4)
Hodgkin lymphoma	–	14,465 (5.2)
Acute myeloid leukemia	–	11,053 (4.0)
Myelodysplastic syndrome	–	9,358 (3.4)
Acute lymphoblastic leukemia	–	4,820 (1.7)
Payer type, n (%)
Commercial	230,432 (59.2)	149,734 (54.0)
Medicare	152,920 (39.3)	123,815 (44.6)
Medicaid	5,571 (1.4)	3,765 (1.4)
Patient census region, n (%)
South	160,599 (41.3)	110,567 (39.9)
Midwest	90,503 (23.3)	60,748 (21.9)
West	73,894 (19.0)	55,101 (19.9)
Northeast	63,927 (16.4)	50,898 (18.4)
Pharmacy density, n (%)
<1 pharmacy per 3,500 population	49,939 (12.8)	37,942 (13.7)
≥1 pharmacy per 3,500 population	338,984 (87.2)	239,372 (86.3)
ADI percentile, n (%)
1–25 (least disadvantaged)	66,552 (17.1)	53,708 (19.4)
26–50	99,359 (25.5)	74,051 (26.7)
51–75	169,248 (43.5)	115,629 (41.7)
76–100 (most disadvantaged)	53,764 (13.8)	33,926 (12.2)

^aDartmouth-Manitoba adaptation

Abbreviations ADI Area Deprivation Index, CCI Charlson Comorbidity Index, SD standard deviation

RZV Uptake. Between October 20, 2021 (ACIP vote) and June 30, 2023, cumulative RZV uptake among RZV-naïve patients was gradual and only reached 7.6% in the solid tumor cancer cohort, ranging from 3.5% (pancreatic) to 9.9% (prostate) (Fig. 1A). Over the same period, RZV uptake in the hematologic malignancy cohort was 8.6% and ranged from 5.6% (Hodgkin lymphoma and acute lymphoblastic leukemia) to 9.3% (chronic lymphocytic leukemia) (Fig. 1B).

Fig. 1.

Cumulative RZV uptake since the 2021 ACIP recommendation among RZV-naïve immunocompromised US adults with (A) solid tumor cancers or (B) hematologic malignancies^a

^aPopulation sizes can be found in Table 1

Abbreviations: ACIP, Advisory Committee on Immunization Practices; RZV, recombinant zoster vaccine; US, United States

Cumulative RZV uptake was considerably higher among those aged ≥ 50 vs. 19–49 years at index in the solid tumor cancer cohort (8.3% vs. 1.6%) and the hematologic malignancies cohort (9.4% vs. 2.4%) (Appendix Table A5, online only). Among patients aged 19–49 years, uptake ranged from 0.8% (liver) to 4.6% (multiple myeloma).

Patients in the solid tumor cancer and hematologic malignancy cohorts who received RZV mainly did so in a pharmacy (86.7% and 84.0%, respectively) or a physician office (12.2% and 14.8%), with few receiving one dose in each setting (1.2% and 1.1%). Place of service of each dose are detailed in Appendix Table A6 (online only). Almost a quarter of patients were co-administered RZV with influenza, pneumococcal, and/or Tdap vaccine (23.8% and 24.1% in the solid tumor cancer and hematologic malignancy cohorts, respectively [Appendix Table A6, online only]). Of the doses received in pharmacies, most were in big box/chain drugstores (70.1%), with 6.9% in independent drugstores, and 23.0% in other pharmacy types across both cohorts. Among doses received in physician offices, the most common associated specialties were family medicine and internal medicine (Appendix Table A7, online only). Notably, only 0.4% and 3.3% of RZV doses received were given by oncologists to solid tumor cancer or hematologic malignancy patients, respectively.

RZV Series Completion and Dosing Schedule Compliance. Among patients with solid tumor cancers who received a first dose of RZV, 63.2% received a second dose within 6 months (57.9–66.8% across tumor types) (Fig. 2A). This increased to 70.9% at 12 months (63.4–74.9% across tumor types). Among patients with a hematologic malignancy who received a first dose of RZV, 64.7% received a second dose within 6 months (59.7–66.8% across cancer types) (Fig. 2B), and this increased to 71.6% at 12 months (64.9%–74.0% across cancer types).

Fig. 2.

RZV series completion among immunocompromised adult patients with (A) solid tumor cancers or (B) hematologic malignancies^a

^aAmong adults who received a first RZV dose by December 31, 2022 (i.e., numbers in brackets)

Abbreviation: RZV, recombinant zoster vaccine

RZV series completion rates were slightly higher among those aged ≥ 50 vs. 19–49 years in the solid tumor cancer cohort (6-month: 63.2% vs. 59.7%; 12-month: 70.9% vs. 67.8%) and the hematologic malignancies cohort (6-month: 64.8% vs. 60.3%; 12-month: 71.7% vs. 67.3%) (Appendix Table A5, online only). Completion rates by age group and cancer type are detailed in Appendix Table A5 (online only).

Among patients who received ≥ 1 dose by December 31, 2022, the median (95% CI) times to the second dose were 4.2 (4.1–4.2) and 4.0 (3.9–4.0) months in the solid tumor cancer and hematologic malignancy cohorts, respectively (Appendix Table A8, online only). Overall, 88.9% and 90.3% of patients in the solid tumor cancer and hematologic malignancy cohorts who received 2 doses were compliant (i.e., received the second dose of RZV within 1–6 months of the first dose) (see Fig. 3 for rates by cancer type).

Fig. 3.

Compliance^a with the recommended RZV series schedule among immunocompromised adult patients with (A) solid tumor cancers or (B) hematologic malignancies

^aSecond dose within 1–6 months of the first dose among patients who received 2 RZV doses

Abbreviation: RZV, recombinant zoster vaccine

Predictors of Uptake. Factors associated with increased odds of RZV uptake across both cancer cohorts included older age, higher household income, lower Area Deprivation Index percentile, non-Hispanic Asian/other or Hispanic (compared to non-Hispanic White), and female sex (Table 2). Within the solid tumor cancer cohort, patients with prostate, thyroid, breast, skin (melanoma), or kidney tumors had notably higher odds of RZV uptake, while those with pancreatic, lung, liver, or colorectal tumors had the lowest odds. Among those with a hematologic malignancy, patients with non-Hodgkin lymphoma, chronic lymphocytic leukemia, or multiple myeloma had the highest odds of RZV uptake, while those with myelodysplastic syndrome had the lowest odds.

Table 2.

Odds of RZV uptake (first dose) among immunocompromised adult patients with cancer

Variable	Solid tumor cancers (n = 240,122)		Hematologic malignancies (n = 174,994)
	OR (95% CI)	P	OR (95% CI)	P
Age, years (ref ≥ 65 years)	–	–	–	–
19–29	0.17 (0.11–0.25)	< 0.0001	0.22 (0.17–0.28)	< 0.0001
30–39	0.08 (0.06–0.10)	< 0.0001	0.18 (0.14–0.22)	< 0.0001
40–49	0.20 (0.18–0.23)	< 0.0001	0.34 (0.30–0.38)	< 0.0001
50–59	0.93 (0.89–0.98)	0.0025	1.14 (1.08–1.20)	< 0.0001
60–64	1.10 (1.05–1.15)	< 0.0001	1.24 (1.18–1.31)	< 0.0001
Sex (ref male)	–	–	–	–
Female	1.14 (1.09–1.19)	< 0.0001	1.05 (1.02–1.09)	0.0016
Race/ethnicity (ref non-Hispanic White)	–	–	–	–
Non-Hispanic Asian/other	1.26 (1.15–1.39)	< 0.0001	1.35 (1.21–1.50)	< 0.0001
Hispanic	1.15 (1.08–1.22)	< 0.0001	1.16 (1.09–1.24)	< 0.0001
Non-Hispanic African American	1.02 (0.96–1.09)	0.45	1.04 (0.97–1.12)	0.23
Unknown	0.99 (0.94–1.03)	0.54	1.05 (1.00–1.10)	0.069
Solid tumor cancer type (ref no)	–	–	–	–
Prostate	1.49 (1.42–1.56)	< 0.0001	–	–
Thyroid	1.46 (1.35–1.58)	< 0.0001	–	–
Breast	1.26 (1.21–1.31)	< 0.0001	–	–
Skin (melanoma)	1.26 (1.16–1.36)	< 0.0001	–	–
Kidney	1.18 (1.09–1.26)	< 0.0001	–	–
Endometrial	1.11 (1.02–1.21)	0.021	–	–
Bladder	1.09 (1.02–1.16)	0.011	–	–
Colorectal	0.84 (0.80–0.89)	< 0.0001	–	–
Liver	0.73 (0.64–0.84)	< 0.0001	–	–
Lung	0.70 (0.67–0.74)	< 0.0001	–	–
Pancreatic	0.48 (0.42–0.53)	< 0.0001	–	–
Hematologic malignancy type (ref no)	–	–	–	–
Acute lymphoblastic leukemia	–	–	1.17 (0.99–1.38)	0.063
Non-Hodgkin lymphoma	–	–	1.14 (1.09–1.19)	< 0.0001
Chronic lymphocytic leukemia	–	–	1.10 (1.05–1.16)	0.0001
Multiple myeloma	–	–	1.08 (1.03–1.14)	0.0036
Chronic myeloid leukemia	–	–	0.97 (0.89–1.04)	0.38
Hodgkin lymphoma	–	–	0.94 (0.86–1.04)	0.22
Acute myeloid leukemia	–	–	0.92 (0.83–1.01)	0.090
Myelodysplastic syndrome	–	–	0.75 (0.68–0.84)	< 0.0001
CCI score,a (ref 0)	–	–	–	–
1	1.01 (0.95–1.06)	0.85	0.90 (0.85–0.95)	0.0002
≥2	0.75 (0.72–0.78)	< 0.0001	0.88 (0.85–0.91)	< 0.0001
Patient payer type (ref Commercial)	–	–	–	–
Medicare	0.95 (0.92–0.98)	0.0026	0.96 (0.93–1.00)	0.042
Other	0.85 (0.71–1.01)	0.068	0.91 (0.76–1.09)	0.32
Pharmacy density (ref < 1 per 3,500)	–	–	–	–
≥1 per 3,500	0.91 (0.87–0.95)	< 0.0001	0.92 (0.87–0.96)	0.0003
ADI (ref 1–25 [least disadvantaged)	–	–	–	–
26–50	0.93 (0.89–0.97)	0.0014	0.90 (0.86–0.95)	< 0.0001
51–75	0.87 (0.84–0.91)	< 0.0001	0.87 (0.83–0.91)	< 0.0001
76–100 (most disadvantaged)	0.78 (0.74–0.83)	< 0.0001	0.70 (0.66–0.75)	< 0.0001
Education (ref completed high school)	–	–	–	–
Attended vocational/technical school	0.96 (0.86–1.07)	0.45	0.97 (0.86–1.10)	0.63
Completed college	1.10 (1.06–1.14)	< 0.0001	1.10 (1.06–1.14)	< 0.0001
Completed graduate school	1.20 (1.14–1.26)	< 0.0001	1.19 (1.12–1.26)	< 0.0001
Unknown	1.04 (0.97–1.11)	0.27	1.01 (0.94–1.09)	0.72
Household income (ref <$25,000)	–	–	–	–
$25,000–49,999	1.04 (0.98–1.10)	0.24	1.04 (0.96–1.12)	0.36
$50,000–74,999	1.10 (1.03–1.17)	0.0029	1.10 (1.03–1.19)	0.0058
$75,000–99,999	1.14 (1.08–1.22)	< 0.0001	1.14 (1.07–1.23)	0.0002
$100,000–149,999	1.26 (1.19–1.34)	< 0.0001	1.28 (1.19–1.37)	< 0.0001
≥$150,000	1.37 (1.28–1.46)	< 0.0001	1.46 (1.36–1.58)	< 0.0001
Missing/unknown	0.90 (0.21–3.84)	0.88	0.74 (0.18–3.09)	0.68

^aDartmouth-Manitoba adaptation, excluding hematologic malignancy or solid tumor cancer for the solid tumor cancer and hematologic malignancy models, respectively

Abbreviations ADI Area Deprivation Index, CCI Charlson Comorbidity Index, CI confidence interval, OR odds ratio, RZV recombinant zoster vaccine

Qualitative results

The median (interquartile range) age of the 14 hematologists/oncologists and 1 medical oncologist included in the study was 50 (48–59) years and 80.0% were male. The mean (SD) number of patients seen daily was 24 (8) and they reported their patients were of diverse ethnicity and had mixed payer types (Appendix Table A9, online only).

Following thematic analysis of the transcribed interviews, three themes related to barriers to RZV vaccination emerged, each with multiple subthemes: (1) financial concerns; (2) lack of formal processes; and (3) insufficient awareness/education. Representative quotes supporting each theme are listed in Table 3 by subtheme.

Table 3.

Themes, subthemes, and quotes describing barriers to HZ vaccination

Themes	Subthemes	Quotes
Financial concerns	Cost of stocking and storing the vaccine	• “The vaccine is easily available at any drug store or pharmacy. Walgreens, CVS – they all have the vaccines. Most of the PCP offices do not stock shingles either, because it is not practical. From my experience, most of the PCPs will tell the patients to go to Walgreens or CVS.”
	Complex reimbursement and challenges	• “If the patient cannot afford it, I will go above and beyond to get them the drug free from the company. I am a champion for my patient when he’s on treatment. I don’t care about money or anything like that.”
Lack of formal processes	Lack of centralized tracking of vaccination history leading to incomplete information	• “Whenever you do a vaccine, it goes to the state registry and then eventually pulls down to the EMR. There have been some instances where I say you still haven’t gotten your vaccine, and they say I got it last month. It will take some time to show up.” • “Although there is a demand by the government that everybody has EMR systems, there is no interaction between them for us to get any records from an outside practice. The right and left hands don’t know what they are doing with regards to vaccination records.” • “If there were direct feedback between major pharmacies and our Epic system. If it auto updated that the patient got it at CVS, that would make a big difference. Benefit is that there is none of that back and forth and less risk of something being lost to follow-up.” • “They need to be reminded, they need to be in clinic in order to get it. Our EMR is not paired with the PCP’s EMR, so it’s not always easy for them to see previous reconciles, who had what…” • “The barriers are not stocking it in the office.” • “…we don’t have any vaccine reminders. It’s not built in our EMR system at all. You try to get that in the very first one or two office visits. Talk to them about their vaccine history to encourage why they want to do it at all.”
	Lack of time/ priority	• “We barely have enough manpower to run chemo and give people their chemo. We don’t have manpower for vaccines. I would refer them to PCP and these days it is even easier to go through a pharmacy.” • “We stock chemo. I have enough on my plate to deal with. I cannot track everybody’s vaccination and what is taken. I think it’s too much in our work, the load that we see patients. I cannot track everybody’s vaccine. That’s why we don’t stock. I don’t think it should be my job.” • “This must be a little bit part of the primary doctors. I strongly, and I will insist, it should not be part of the oncologists, except certain circumstances when you have post-transplant or severely immuno-compromised, you must address that, but not to track. Our part is to guide the treatment and offer it, but not track if you took the second dose or not. This is not my job.” • “We do handle the transplant piece, but I don’t know if we can handle routine immunizations for everybody on Velcade. That might be a lot.” • “…it is usually given through the PCP, and I can order it. Occasionally it’s given in the cancer center, but it’s not our main focus, just given the workload we have. I do communicate with whomever referred that patient as far as what vaccines I would recommend for that patient…” • “We have the mammograms and colonoscopy and the A1C, but the shingles vaccine is not benchmarked.” • “Then we talk to them about checking with their family doctor. You should get this, this, this vaccination. They can go to Walgreens or CVS or wherever they want to get it if they want to do that… Hopefully, they will get it done but we are not policing them for it.”
Insufficient education/ awareness	HCP awareness of effective use of technology	• “We use Epic. Patients have MyChart… I have never seen it on the patient’s side but I think these alerts also show up for the patients…I know they are because they ask me am I due for this vaccine? Am I due for colonoscopy?”
	Incomplete HCP knowledge about RZV	• “…if someone is on high dose corticosteroid for a while, cytotoxic agent, certain people on rituximab, I try to avoid the vaccine. I would wait until they’re done… because I think it’s contraindicated.” • “The only thing is if somebody is going for bone marrow transfer. If someone has myeloma, they are going to have bone marrow transfer. These people will need all their vaccines post transplantation.” • “The concern I have is if they are immunocompromised, they are on cytotoxic chemo with myelosuppression, they may not get the benefit of it. That is my concern.” • “I’m very supportive of the drug rep when they come here. Of course, the drug rep, his job is to promote the drug, and he wants to sell it. But in the end, it’s my choice and my decision. But it’s good to discuss it with them and see what he has to do, or she say, and I learn from them.” • “To promote this with a flyer from a pharmaceutical company coming to the clinic and educate very quickly the providers, the patients. They can give presentation or a lunch and presentations. Talk about the shingles vaccine or any vaccination and keep us updated of the current guidelines once per year.” • “Education is the key for them, not just the patients but for the oncologists. That is important. It is not as well known. It has been around for a few years as far as I know so maybe more knowledge. I’ve never met a rep for this vaccine. If they go to more PCP, but it might be helpful to speak to oncologists as well or whoever has immunosuppressed populations.”
	Lack of patient information/ misinformation	• “Patients will say I have had shingles before, and I don’t want to go through that. Am I at risk for getting shingles again during chemo? Then we talk about it. Yes, they are because during chemotherapy their immune system would be more compromised. Then the next question is should I get the vaccine and when is it appropriate for me to get the vaccine. The patient typically initiates that conversation.” • “If we put a flyer in the cancer center in the lobby or in exam room, especially in exam room, people will have downtime waiting. If they see the flyer, that would remind them and they will ask the questions. That will help both the patient and the HCP with that conversation. If I get encountered, the patient usually already has cancer diagnosis. Our cancer center, most of the patients are Medicare age. They should already have gotten those vaccines. If their PCP has already talked about it…It would make my life easier. [Helpful if PCP was on top of this.] Right. Education, there is a lot of misinformation on the internet.” • “There is more hesitancy about vaccines and fewer people are getting vaccines since COVID. [What about with shingles? ] thankfully many folks don’t have specific concerns about the shingles vaccine. In general, patients just say they refuse all vaccinations including shingles.” • [How do you make hesitancy less of a challenge? ] “Overall education.”

Abbreviation: chemo chemotherapy, COVID Coronavirus disease 2019, EMR electronic medical records, HCP healthcare provider, HZ herpes zoster, PCP primary care provider, rep representative, RZV recombinant zoster vaccine

Financial Concerns. Related to the cost of vaccinating their patients, oncologists’ comments resulted in two subthemes: (1) cost of stocking/storing the vaccine and (2) complex reimbursement and challenges. Some providers said they had opted to write prescriptions for RZV to facilitate the vaccination process given the perception that the vaccine is easily accessible at area pharmacies. Respondents generally indicated they were not involved with reimbursement and billing for vaccination, but those who reported stocking RZV indicated that reimbursement for vaccinating their patients, particularly their younger patients, was complex and often not sufficient.

Lack of Formal Processes. In general, those interviewed indicated that their priority was cancer care and viewed vaccination as the responsibility of primary care providers; however, they were willing to play a vaccine-related role and were strong advocates for vaccination. Interviewees indicated that logistical challenges impacted their ability to vaccinate eligible patients, both in terms of available patient data (i.e., a lack of centralized tracking of vaccination history leading to incomplete information [very common]) and their typical workflow (i.e., lack of time/priority [very common]). Specifically, a lack of integration with data from primary care, pharmacies, and state registries contributed to incomplete vaccination information, requiring oncologists to rely on patient recall to make recommendations based on need. To address this, they recommended incorporating structured fields and reminders or alerts into electronic health records and clinical workflow to assist in making vaccine-related recommendations and facilitate improved communication between different healthcare providers. However, beyond this, a lack of time during encounters was cited as a challenge as respondents indicated that vaccine conversations are not often planned for given their primary task of addressing the underlying conditions for which the patients were referred to for care.

Insufficient Education/Awareness. The third theme centered on a need for a clearer understanding of the indications and appropriate use of RZV in patients with cancer, and how technology may address these shortfalls. Subthemes were provider awareness of effective use of technology (very common), incomplete provider knowledge about RZV (common), and lack of patient information/misinformation (uncommon). For instance, oncologists noted that they could benefit from a better understanding of the indications and use of RZV in their patient populations, support for which could come from manufacturers. Several specialists mentioned stronger vaccine recommendations for patients with pancreatic cancer, as their treatment may involve surgery on the spleen. Additionally, those interviewed reported lacking awareness of where vaccination histories were documented and the extent to which patients may have access to such information in online portals. Similarly, for patients, improved efforts to educate on vaccines are needed among those with cancer, specifically on the benefits of vaccination prior to and during periods of active oncolytic therapy. Providers suggested increasing the amount and type of educational materials (e.g., ACIP recommendations, vaccine benefits and risks) available to patients in the clinic (e.g., posters, pamphlets) could assist them in facilitating more proactive, vaccine-related discussions, managing any misinformation, and reinforcing the need for preventive care during patient visits.

Discussion

Results of this study (summarized in a graphical abstract) suggest that uptake of RZV since the October 2021 ACIP recommendation [8] among patients who had not already been vaccinated was gradual and only ranged from 3.5% to 9.9% across included cancer types by June 2023, with rates of only 0.8% to 4.6% among patients aged 19–49 years. While this is suboptimal, it should be noted that patients who had received RZV before the 2021 recommendation [8] (most likely based on the 2017 recommendation for those aged ≥ 50 years) [9] were excluded from the analysis (approximately 17%; Appendix Fig A2, online only). Therefore, the numbers reported herein are a measure of new uptake since the recommendation rather than an estimate of the overall RZV uptake in patients with cancer.

There were notable differences in new RZV uptake by cancer type, which ranged from 3.5% among patients with pancreatic cancer to 9.9% among those with prostate cancer. Interestingly, patients with pancreatic cancer had the lowest RZV uptake, before and after adjustment, which is at odds with providers’ feedback as several of the interviewed specialists mentioned stronger vaccine recommendations for patients with pancreatic cancer, as their treatment may involve surgery on the spleen. This could be because patients aged ≥ 50 years with pancreatic cancer may have been more likely to receive RZV before receiving immunosuppressive treatments even before the 2021 recommendation and were therefore excluded from the study population.

In the solid tumor cancer cohort, the odds of RZV receipt in multivariable analysis were aligned with the raw data; however, in the hematologic malignancy group, patients with acute lymphoblastic leukemia had the second lowest proportion of RZV uptake but the highest odds of RZV receipt in multivariable analysis. This could be because results were adjusted for age, due to underlying differences among those with linked consumer data, or a reflection of the influence of social determinants of health data included in the model.

Many of the predictors of RZV uptake in these cancer cohorts are aligned with previous studies of uptake of RZV or any HZ vaccine (i.e., RZV or zoster vaccine live, which was available in the US from 2006 [17] to 2020 [8]). These included older age [13, 18–23], female sex [13, 18, 21, 23], having commercial insurance [18, 24], lower social disadvantage [18, 21], higher household income [23–25], and higher education level [21, 23, 24]. However, some predictors of RZV uptake in the current study were surprising, such as Hispanic ethnicity (significantly higher vs. non-Hispanic White) and non-Hispanic African American race (similar to non-Hispanic White), both of which have been associated with considerably lower uptake of HZ vaccines in other studies [18–25]. The current study also found higher RZV uptake among non-Hispanic Asian/other (vs. non-Hispanic White) individuals, which aligns with one previous RZV study [18], but other HZ vaccine studies have reported higher or similar uptake among non-Hispanic White (vs. non-Hispanic Asian) adults [19, 20, 22, 24, 25].

Among patients with solid tumors or hematologic cancers who received a first RZV dose, approximately 57.9–66.8% and 63.4–74.9% received a second dose within 6 months and 12 months, respectively. These completion rates are lower than estimates from previous studies in the general population aged ≥ 50 years, where 6- and 12-month completion rates of 67–72% and 78–83%, respectively, have been reported [26–29]. This could be because it was not possible to impose a minimum follow-up time requirement using the open databases in the current study whereas most previous studies only included patients with 6 or 12 months of available follow-up [26–28].

Key findings from the provider interviews suggest that, while oncologists advocate for vaccination with RZV, multiple barriers may hinder vaccine uptake. Respondents in the current study generally only view their roles in vaccination as advisory due to time constraints and competing patient needs, as has been reported in a recent survey of US specialists including oncologists [30]. This is likely a leading reason for RZV being overwhelmingly administered by pharmacists and primary care providers as reflected in the accompanying quantitative results. Importantly, several potential facilitators of HZ vaccination were also identified, including protocol changes (e.g., nurse involvement, consistent documentation, patient follow-up), increased time with patients, adoption of a proactive approach towards vaccination, better use of electronic health records, and provider and patient education.

Limitations

The quantitative part of the study relied on the availability and accuracy of the pharmacy and medical records, which were primarily collected for billing and reimbursement purposes. Data that were not recorded, miscoded, or inadequate could have introduced bias. Open data does not contain patient enrolment dates, so some patient encounters may not have been captured. Hence, some patients could have already received RZV and RZV uptake and series completion may have been underestimated. Medical claims were only sourced from professional fee claims (not hospital fee claims), but as few patients receive RZV in a hospital, this should have had a minimal impact.

Cancer and immunocompromising medication status was only assessed during the 6 months before the ACIP recommendation. As patients are recommended to receive RZV before starting immunosuppressive treatments [10], patients included in this study may have been less likely to receive RZV than patients who had been diagnosed with cancer but not yet started treatment. It is also possible that patients may not have continued to be immunocompromised during the study period. Also, only one cancer claim (and one immunocompromising medication claim for certain cancers) was required, so there is the potential for misclassification. Moreover, this study did not include patients who would have met the eligibility criteria after the study’s index date. It should also be noted that the index date was the day of the ACIP vote (October 20, 2021), but the ACIP [8], NCCN [11], and ASCO [12] recommendations were not published until January 2022, March 2022, and May 2024, respectively. It would also have taken time for the new recommendation to be adopted, but Fig. 1 indicates no discernible difference in uptake around these publication dates. However, the completion rates are likely underestimated as it was not possible to impose a minimum follow-up time.

The current results are not necessarily generalizable because (1) the analysis only included patients with specific types of cancer; (2) the cohorts mainly included commercially insured and Medicare patients; (3) the study population was less racially diverse than the overall US population [31], and (4) the sociodemographic characteristics may also have differed from the overall population.

Related to the provider interview portion, limitations include recall bias and reduced reliability due to the small sample size. Although the aim was to gain insight from oncologists that were representative of practices across the country, the experience and opinions of those who agreed to participate may be different from those of US oncologists in general.

Conclusions

This study suggests that RZV uptake among RZV-naïve immunocompromised patients with cancer since the 2021 ACIP recommendations has been slow and gradual during 20 months of follow-up, indicating that it can take considerable time for recommendations to be fully adopted. However, among those who received 1 dose, series completion (2 doses) was likely and usually achieved within the recommended timeframe. Given the increased risk of HZ among patients with cancer, reinforcing the importance of HZ prevention (e.g., by administering RZV as per ACIP [8], NCCN [11], and ASCO [12] recommendations) in this patient population should be emphasized. Insights from provider interviews can be used to develop future interventions, improve clinical workflow, and make changes in policies to address the unmet vaccination needs of patients with cancer.

Abbreviations

ACIP

Advisory Committee on Immunization Practices

ASCO

American Society of Clinical Oncology

CCI

Charlson Comorbidity Index

CI

confidence interval

HZ

herpes zoster

NCCN

National Comprehensive Cancer Network

ORs

odds ratios

PY

person-years

RZV

recombinant zoster vaccine

SD

standard deviation

Tdap

tetanus-diphtheria-pertussis

US

United States

Authors’ contributions

**Conception and design: ** Justin Gatwood, Catherine McGuiness, Marie Yasuda, Chi-Chang Chen, Vasudha Gupta, and Nikita Stempniewicz.**Collection and assembly of data: ** Catherine McGuiness, Vasudha Gupta, and Marie Yasuda.**Data analysis and interpretation: ** Justin Gatwood, Catherine McGuiness, Vasudha Gupta, Marie Yasuda, Chi-Chang Chen, and Nikita Stempniewicz.**Final approval of manuscript: ** All authors.**Accountable for all aspects of the work: ** Justin Gatwood.

Funding

GSK funded this study (GSK study identifier: VEO-000679) and was involved in all stages of study conduct, including analysis of the data. GSK also took charge of all costs associated with the development and publication of this manuscript.

Data availability

The data that support the findings of this study are available from IQVIA but restrictions apply to the availability of these data, which were used under license for the current study and are not publicly available. However, data are available from the authors upon reasonable request and with permission of IQVIA.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

JG and NS are employed by, and hold financial equities in, GSK. JG also reports receipt of grants from Merck & Co. and AstraZeneca, consulting fees from Merck & Co. and Janssen, and conference travel reimbursement from Genentech outside the submitted work during the past 3 years. CM, VG, MY and CCC declare being employees of IQVIA which received funding from GSK to conduct the current study. IQVIA also conducts research projects funded by and in collaboration with the pharmaceutical industry. CM also holds financial equities in Pfizer. The authors declare no other financial and non-financial relationships and activities.