Piloting delivery of PfSPZ vaccines for malaria through a cryogenic vaccine cold chain to travel and military medicine clinics

Abstract

Background

PfSPZ vaccines comprising Plasmodium falciparum (Pf) sporozoites (SPZ) have demonstrated > 90% protection against variant Pf malaria infections for at least 12 weeks; they are the only vaccines with the level of efficacy necessary to protect travellers. PfSPZ are eukaryotic cells stabilized by cryopreservation and distributed using a cryogenic (below −150 °C) cold chain. The Ebola vaccine and mRNA vaccines against SARS-CoV-2 pioneered uptake of vaccines requiring non-standard ultra-low temperature cold chains. The cryogenic cold chain using liquid nitrogen (LN₂) vapour phase (LNVP) cryoshippers, is simpler, more efficient than −80, −20 or 2–8 °C cold chains, and does not use electricity. This study was conducted to evaluate implementation and integration of a cryogenically distributed vaccine at travel and military immunization clinics.

Methods

We conducted sequential 28-day studies evaluating vaccine shipping, storage, maintenance and accession at two US military and two civilian travel health/immunization clinics. In each clinic, personnel were trained in equipment use, procurement and handling of LN₂, temperature monitoring and inventory record keeping by in-person or video instruction.

Results

Sites required 2–4 h/person for two persons to assimilate and develop the expertise to manage vaccine storage and LNVP operations. LN₂ for recharging cryoshippers was delivered every 1–2 weeks. Vaccine ordering, receipt, storage and inventory control was conducted effectively. Simulated single dose vaccine cryovial retrieval and thawing were performed successfully in different travel clinic settings. Continuous temperature monitoring at each site was maintained with only one short excursion above −150 °C (−145 °C) through shipping, use and reverse logistics. Staff, during and at study conclusion, provided feedback that has been incorporated into our models for cold chain logistics.

Conclusions

These studies demonstrated that the training in delivery, storage, administration and integration of PfSPZ vaccines can be successfully managed in different immunization clinic settings for travellers and military personnel.

Introduction

The malaria vaccine, PfSPZ Vaccine®, comprises attenuated sporozoites (SPZ) of Plasmodium falciparum (Pf). These eukaryotic cells must, unlike viruses and bacteria, be cryopreserved (below −136 °C¹) to retain viability and functionality as immunogens. Prior to the development of Ervebo, the Ebola vaccine,² and mRNA vaccines against SARS-CoV-2,³ that utilize ultralow temperature cold chains, the predominant opinion was that cryogenic technologies are difficult to implement for human vaccines. This misconception has endured despite below −150 °C cold chains being employed for several veterinary vaccines,^4,5 extensively by the huge livestock breeding industry,^6,7 for human reproductive medicine^8,9 and for the rapidly expanding field of CAR-T products¹⁰ and cellular therapies.¹¹

An efficacy of ≥90% lasting 12 weeks^12–14 or more, demonstrated for PfSPZ Vaccine in phase 1 and 2 clinical trials in naïve adults, meets the requirements for a travellers’ malaria vaccine. Real-life testing of this scenario is currently ongoing in Indonesia (unpublished). Following licensure it is anticipated that travel medicine clinics and military immunization centers will be among the first for uptake of PfSPZ vaccines. Maintaining a below −150 °C cold chain is accomplished using liquid nitrogen (LN₂) vapour phase (LNVP) cryoshippers (aka dry shippers, dry vapour shippers) that operate without the need for electricity and double as in-clinic vaccine storage units.

In contrast to personnel at our clinical trial sites who have been trained extensively in handling LN₂, LNVP cryoshippers, and the cryogenic vaccine vials, the four clinics in this study had no prior experience working with cryogenically stored and distributed vaccines or the use of LN₂. The CDC ‘tool kit’¹⁵ that covers vaccine delivery, receipt, storage, and administration, highlights the need to (a) train clinic personnel well and frequently, (b) employ reliable storage and temperature monitoring equipment, and (c) practice accurate vaccine inventory management.

The aim of this study was to demonstrate that introducing a cryopreserved vaccine requiring a cryogenic, below −150 °C, cold chain into active mid-sized travel and military medicine clinic operations, is feasible and we hypothesized that travel and military immunization clinics would be fully capable of developing the necessary expertise. Through this study, we also hoped to bring to the attention of the travel medicine community (1) an awareness of the currently underutilized cryogenic cold chain for vaccines and (2) development of PfSPZ vaccines against malaria.

Materials and methods

Study sites

1. Walter Reed National Military Medicine Center (WRNMMC), Bethesda, MD.

2. Madigan Army Medical Center, Joint Base Lewis-McChord, WA.

3. Traveler’s Medical Service (TMS), New York, NY.

4. Passport Health Silver Spring Travel Clinic (PH), Silver Spring, MD.

Outcomes

The primary study outcome was defined as demonstration that a functional cryogenic cold chain for PfSPZ vaccines could be established with the four travel and military medicine sites. Secondary outcomes were: demonstration of effective training, competency in the handling of LNVP cryoshippers, and storage temperature and inventory monitoring.

Training

Personnel at each site were trained on all equipment and operations either in person (WRNMMC, Madigan and PH) or with pre-recorded instructional videos and online meetings (TMS). To become familiar with all equipment, supplies and documentation, these were delivered to all sites in advance. Initial training required 2–4 h per person involving 2–4 staff at each site; these staff subsequently trained additional personnel at their site as needed. Effectiveness of training was monitored through records of LN₂ acquisition and recharging events, temperature monitoring and inventory logs. Personal protective equipment (PPE) for handling LN₂ was also provided and sites were trained in their use, and O₂ level monitors were provided if requested.

Cryovials and cryoshippers

Cryovials,¹⁶ 0.7 mL polypropylene with heat-annealed foil seal-protected septum, each containing a single dose of 20 μL media-CPA (cryoprotectant additive) as surrogate for PfSPZ vaccines were cryopreserved and stored below −150 °C. A box of cryovials in an 8 × 12 array of 96 cryovials per box, the standard packaging unit, was supported by a custom box holder with handle (except Madigan) allowing the rack to be easily elevated for individual cryovial retrieval directly from the cryoshipper.

Cryoshippers (Figure 1) (MVE Biological Solutions) are triple-walled Dewars with a vacuum between the outer and middle walls, and with a high surface area low density material between the middle and inner walls that holds the LN₂ refrigerant. The central chamber houses an internal stainless steel canister containing the custom box holder for one or more boxes of cryovials surrounded by protective padding and closed with a lid during transport secured by socket cap screws. The cryoshipper also has a protective casing used for transportation.

Figure 1.

The left panel shows the floor scale, 10 L Dewar and the cryoshipper (cream colour) inside its protective shipping case (blue). At the top left of the cryoshipper casing is the Libero temperature data logger (red) and to the right is the digital thermometer (yellow). The right panel shows the CryoPod Carrier as used at the Madigan site and to its right the VT-x1 vial thawing device.

These cryoshippers hold 10.2 L of LN₂ with nominal 0.7 L/day evaporation rate giving a static hold time of 14 days. Sites were required to recharge the cryoshipper with LN₂ on a weekly schedule and to assess LN₂ use by weight—cryoshippers weigh 15 kg empty and 22.2 kg fully charged. Cryoshipper weight was tracked using a floor scale (Angel USA Inc.). Sites obtained LN₂ in Dewars (10 or 25 L, Worthington Industries) through scheduled weekly or biweekly deliveries from Sanaria or a local vendor.

Additionally, at Madigan, a deployment center with potentially larger daily vaccinee throughput, a CryoPod carrier (Figure 1) (Azenta Life Sciences) was used for transportation of cryovials within the facility. The CryoPod is a 4 kg LNVP cold box with a handle and can accommodate two boxes of cryovials below −150 °C for up to 4 h while in active use. The CryoPod measures 320 mm × 330 mm, allowing it to be easily located on the benchtop and cryovials can be retrieved manually without use of forceps. CryoPods have built-in temperature recording and alarm functions, although for this project we included the additional Elpro temperature data logger devices indicated below (and shown in Figure 1).

Temperature monitoring

A Libero Te1-P data logger with 900 mm probe (Elpro-Buchs AG) monitored the cryoshipper chamber temperature during transportation, and the canister while at the clinic. Temperatures were logged at 5-min intervals, and thermal history PDF downloads via the USB connection were taken after receipt at the clinic, after each LN₂ recharge event, and at Sanaria after return of the cryoshipper. A digital thermometer (Fluke) with type-T thermocouple probe (Omega), was used to monitor cryoshipper temperature in real time and recorded the temperature at each cryovial retrieval for simulated immunization events.

Cryoshipper delivery and receipt

Cryoshippers containing surrogate vaccine cryovials were delivered to WRNMMC and PH by Sanaria, and by courier service to TMS and Madigan. Upon arrival, cryoshippers were checked for physical integrity and the data loggers’ thermal histories downloaded; findings were reported immediately to Sanaria. The cryoshipper was opened, the canister transferred into an insulated box containing LN₂ while the canister lid was removed. Padding used during transportation was removed and the box lid was unlocked and removed. Cryovial stock was verified against the inventory log and recorded. Following the first LN₂ recharge, the canister with cryovials was replaced into the cryoshipper, locating the probes for the thermometer and data logger inside the canister. The cryoshipper lid was replaced and the cryoshipper locked in a secure location with access limited to participating staff—locations included a stockroom (PH), pharmacy (TMS, Madigan) and adjacent research lab (WRNMMC). At the end of each study period, cryoshippers and all materials were repacked and returned the same way they were delivered.

Cryovial retrievals

On simulated vaccination days, the cryoshipper was placed adjacent to a working surface incorporating space for the VT-x1 vial thawer (Key Tech Inc; Figure 1), and a rack for thawed cryovials, diluent and syringes. Cryovials were retrieved using forceps directly from the cryoshipper, or at Madigan manually from the CryoPod. The CryoPod was used continuously for up to 2 h and recharged with LN₂ if the temperature rose to −170 °C.

Each cryovial retrieval was recorded. For sequential retrievals, the cryoshipper lid remained off but inverted—for ease of cryovial access and to suppress ‘fog’ formation inside the cryoshipper chamber. Each cryovial was thawed by inserting into the opening at the top of the vial thawer and pressing the frozen vial down until it clicked into place; after 30 s, the thawed vial pops up (like a toaster) and after removal of the foil seal is ready for immediate addition of diluent, mixing and syringe loading. The diluent addition to thawed cryovials and syringe loading did not form part of this study.

Results

Training

Teams of two to four persons were formed at each site, consisting of physicians, nurses and/or medical techs/corpsmen/medics. No team had prior experience with LNVP equipment or handling of LN₂. Training used Sanaria’s standard operating procedures for working with LN₂, use of PPE and attendant risks. All teams quickly became familiar with and competent in handling LN₂, monitoring its use and conducting the LN₂ recharges. Training time per person varied from 2 to 4 h (Table 1).

Table 1.

Summary of the training, equipment, couriers, LN₂ use and simulated immunizations for the four test sites

SUMMARY: equipment, activities and supplies	WRNMMC	Madigan	Traveller’s Medical	Passport Health
Training	In-person	Videos, in-person	Videos, zoom calls	In-person
	Test samples	Full test shipment	Full test shipment	Test samples
Estimated time commitment—training	2 h	4 h	4 h	2.5 h
Cryoshipper	MVE	MVE	MVE	MVE
Courier	Sanaria	World Courier	Biocair	Sanaria
LN2 Dewar	10 L	10 L	25 L	10 L
LN2 supply	Sanaria	Joint Base L-Mc	McKinney Welding	Sanaria
# LN2 deliveries	4	5	3	5
# LN2 recharges	4	8	5	4
Total LN2 used, 4 weeks	21.50 L	23.05 L	23.65 L	22.45 L
Est. weekly time commitment—LN2 recharging	10 min	20 min	10 min	10 min
CryoPod used	No	Yes	No	No
Total # simulated immunizations (4 weeks)	90	65	96	32
# simulated immunizations per clinic ‘session’	5–10	7–18	2–5	1–3
# sessions (4 weeks)	10	5	20	17
Temperature excursions above −150 °C	1: to −145 °C*	none	none	none

^*Storage conditions for PfSPZ Vaccine follow USP<1044>, which recommends an upper threshold temperature of −150 °C. However, −136 °C, generally accepted as the glass transition temperature of aqueous solutions, is indicated as the critical upper threshold for Cryopreservation Best Practices (ISBER BP5, in press)

Temperature monitoring

Data logger thermal histories were generated continuously at all sites (Figure S1). In addition, the temperature inside the cryoshipper (or the CryoPod at Madigan) indicated on the digital thermometer was recorded manually each time a vial was retrieved for a simulated immunization (Table S1). All measurements recorded at all sites indicated that the surrogate vaccine was maintained below the critical upper threshold temperature of −136 °C for the study duration from the time the cryoshippers departed Sanaria until their return. US Pharmacopoeia (USP < 1044>)¹⁷ recommends cryopreserved human clinical materials be maintained below −150 °C. Data logger PDF downloads confirmed only one transient temperature excursion to −145 °C occurred during the study (Table 1).

LN₂ procurement and recharging

All LN₂ recharges were performed without incident by two operators at each site: Similar LN₂ quantities were consumed by each team over the 28 days of storage and simulated use (range 21.5–23.65 L, average 22.66 L).

No problems with obtaining LN₂ were reported. Different arrangements for LN₂ delivery were made for each site. Sanaria delivered LN₂ directly to the clinics at WRNMMC and PH. Commercial suppliers delivered LN₂ to the Madigan clinic and to the ground floor loading dock at TMS. TMS is located on the ninth floor and two 25 L Dewars were used in rotation: when the distributor delivered a full Dewar this was exchanged for an empty Dewar which was returned full at the next delivery.

LN₂ recharging events occurred at ≤7 day intervals without incident. TMS reported that the swivel base provided was initially useful, however, the mouth of the Dewar was too low to conveniently dispense LN₂ directly into the cryoshipper during the second recharge event. No issues with O₂ levels were reported. Records for the LN₂ recharge events carried out at TMS are provided in Table 2.

Table 2.

Traveller’s Medical Service records for cryoshipper temperatures, weights of cryoshippers before and after LN₂ recharging and amounts of LN₂ added for each LN₂ recharge event day

Date	Temp of cryoshipper (°C)	Weight before (kg)	Weight after (kg)	LN2 added (L)
06 APR 2021	−194.5	19.75	21.10	1.35
13 APR 2021	−194.1	19.45	24.05	4.60
20 APR 2021	−191.4	18.35	24.05	5.70
27 APR 2021	−191.0	19.30	24.35	5.05
04 MAY 2021	−191.2	17.70	24.65	6.95
			total	23.65
			mean	5.58
			SD	1.02
Elapsed time	28 days	LN2/day	0.84	kg
			1.21	L

Documentation

Information key to the success of the project was (a) documentation of routine LN₂ recharges, (b) recording the LN₂ volume consumed by recharges, (c) recording the surrogate product temperature both by digital thermometer and data logger with periodic PDF downloads, and (d) recording surrogate vaccine vial retrievals and cryovial inventory log maintenance. From all documents, it was possible to construct how the storage, use and maintenance of the operations at each site were conducted and the effectiveness of the training provided.

Simulated vial retrieval and thawing for immunizations

After delivery, the cryoshippers, as LNVP storage units, were accessed periodically for cryovial removal to simulate immunization events. For three sites, cryovials were removed directly from the cryoshipper using curved forceps and transferred immediately to the VT-x1 vial thawer. At Madigan, the CryoPod was prepared by charging with LN₂ and, when stabilized below −170 °C, the whole box of cryovials was transferred from the cryoshipper. The CryoPod maintained a cryogenic environment for at least 2 h before it required a LN₂ recharge. Temperature records from the Madigan CryoPod are shown in Figure S2 for 3 days of simulated immunizations.

The four clinics operated with normal client caseloads receiving and administering other vaccines. The simulated immunizations of PfSPZ Vaccine were organized around or within the clinic schedules minimizing any disruption while maximizing the allotted time to work with the surrogate cryopreserved vaccine. Sites found clustering retrieval and thawing events was most efficient, organizing retrievals in batches into morning or afternoon sessions. A retrieval schedule example is indicated for WRNMMC in Table S1. Although time measurement to complete various tasks was not a specific aim of this study, the median time to retrieve and thaw five cryovials at TMS was 12 min (range 9–15), consistent with the time to complete these manoeuvres in field trials of PfSPZ vaccines (unpublished). This time does not include addition of diluent to the vial, mixing and loading of the syringe which would add ~ 1.5 min to the preparation time. Injection is typically completed in 10–15 s.

User feedback

During the study and at its conclusion, participating clinic personnel provided comments and feedback. The main suggestions were to use a 10 L Dewar for LN2 recharging to make this a one-person operation and to have backup thawers and spare probes for the temperature monitors—these would be provided in any future supply kits. While none of the sites had prior experience working with LN₂, they found this to be less daunting as they gained familiarity with LN₂-related activities.

Discussion

PfSPZ vaccines cryopreserved in LNVP have been assessed in multiple phase 1 and 2 clinical trials in the US, three countries in Europe, six countries in Africa and in Indonesia. Because of their unprecedented protective efficacy in malaria-naïve adults in the US and in Germany,^13,18 one of the indications for PfSPZ vaccines is prevention of malaria in travellers. Phase 3 trials, to include real-life assessment of efficacy, are pending and if successful will then bring the use of PfSPZ vaccines into consideration by the travel medicine community. Since there are no licensed human vaccines maintained in LNVP. We undertook the pilot studies reported herein to determine how best to deliver vaccines to military and civilian travel immunization clinics. The results indicate that with effective training on the appropriate standard operating procedures, it should be straightforward for clinics to incorporate a PfSPZ vaccine into their routines.

Most human vaccines are distributed through a 2–8 °C cold chain. Exceptions include vaccines for smallpox and monkeypox (JYNNEOS) between −25 and − 15 °C,¹⁹) chickenpox (VARIVAX), shingles (ZOSTAVAX) and measles (MMR II) down to −50 °C,^20–22 the SARS-CoV-2 mRNA vaccines between −60 and − 20 °C (SPIKEVAX)²³ or − 90 and − 60 °C (COMIRNATY)²⁴ and the Ebola virus vaccine (ERVEBO, at −80 to −60 °C.²⁵ Distribution of cryopreserved products employing LN₂ or LNVP is well-established: Such cold chains are used for multiple (N = 11) veterinary vaccines,^4,5,26–29 semen and embryos by the large livestock reproductive industry,^30–32 semen, eggs and embryos for human reproductive use, and for a rapidly expanding repertoire of human cellular therapies and CAR-T products.^33–36

For human injectables, US Pharmacopoeia guidance recommends storage and transportation below −150 °C (USP<1044>¹⁷) giving a safety buffer of 14 °C below the generally accepted glass transition temperature (−136 °C: Tg) for aqueous materials. Below Tg samples are stable essentially indefinitely, while above Tg damage due to ice crystal formation and/or growth may occur. The LNVP cryogenic cold chain, therefore, has only an upper threshold temperature making it intrinsically easier to manage, and vaccines stored in LNVP theoretically have no expiration date. This study also demonstrates that vials could be accessed quickly from below −150 °C storage and only require 30 s to thaw, in contrast to other frozen vaccines which are generally thawed passively at room temperature and may require 30 min or longer to be ready for dilution and administration.

CDC provides guidance for vaccine distribution, storage and administration through its ‘Tool Kit’ and Best Practices publications¹⁵ including recommendations for refrigerators (2 to 8 °C), freezers (−15 to −50 °C), and ultralow freezers (−60 to −90 °C). We have designed the LNVP cold chain to parallel the recommendations and best practices found in these guidelines, with the major difference being management of the cold chain with LN₂ is independent of electricity. A model of the LNVP cold chain in comparison to a standard 2–8 °C cold chain for vaccine distribution in the Expanded Program on Immunization found the costs per vaccine regimen were comparable.³⁷

The cryoshippers used in this study are standard equipment for Sanaria’s clinical trials (38 to date) using cryopreserved PfSPZ products and for immunizations at contracted independent study sites. They have maintained the below −150 °C cold chain throughout transport, including to remote sites in Africa and Asia, and double as temporary in-clinic vaccine storage units, as in this study, often for months at a time. Cryoshipper payload capacity for this study was 480 single dose cryovials in latched boxes of 96, although only one box was distributed to each site. Static hold-time for this model cryoshipper is ~ 14 days. Cryoshipper in-clinic hold time is extended, as in this study, by recharge events where LN₂ refrigerant is replenished regularly, usually weekly. This distribution and on-site storage model is most suitable for mid-sized travel clinics; adjustments to the model would be anticipated for use with smaller or larger clinics.

Alternative delivery and storage models include (1) the replacement and (2) off-load models—which have yet to be tested. In the replacement model, the on-site cryoshipper is returned to the distribution hub for LN₂-recharge and product reload, and a fully-loaded fully-LN₂-charged second cryoshipper is simultaneously delivered to the clinic. Alternatively, a replacement fully-LN₂-charged cryoshipper is delivered, and the payload transferred from the first into this second cryoshipper and the empty cryoshipper is returned to the distribution hub. A modification to the model, using small cryoshippers or cryoshippers with extended hold times of up to one month, would suit smaller clinics and those in more remote locations. In the off-load model, the cryoshipper payload is transferred into a larger static LNVP cryobank at the site and vaccine cryovials retrieved as needed: the cryoshipper is returned to base empty. This model is suited for larger immunization sites with high client throughput (such as Madigan) and where CryoPods can transfer vaccine from pharmacy LNVP storage to the immunization clinic.

Key to success at each site was training. Personnel quickly became competent in equipment use and procedures. Documentation related to cryovial inventory maintenance was not always complete—this was a study sponsor requirement that had no consequences (there was no vaccine to waste or lose)—if the impetus for record keeping had come from the site, we believe compliance would be self-correcting.

Challenges identified by the sites included the following: (1) space required for the equipment; (2) fog generated by moist ambient temperature room air entering the cryoshipper and affecting visibility during vial retrieval; and (3) manipulation of vials with insulated gloves and long forceps. The cryoshippers used were ‘high capacity’ with a footprint diameter of 0.56 m when in their protective shipping case (Figure 1), comparable to the expected footprint of a small pharmacy or laboratory grade refrigerator or freezer. Ancillary equipment (LN₂ Dewar, floor scale, other small items) may occupy a 0.3 m × 0.4 m space. The type of equipment (under development) that will be used in travel clinics after vaccine roll out will be smaller and more compact than the items used for this pilot study and tailored to the distribution model adopted (see above). Importance of the other issues declined as operators became more familiar and practiced and were eliminated with CryoPod introduction at Madigan. An entire box (96 cryovials) could be transferred into the CryoPod, transported to the point of care and maintained for up to 2 h of immunizations. CryoPods (~0.32 m × 0.33 m) fit comfortably on a benchtop, easing clinical space constraints, and their use precluded fog generation. Vials are easily retrieved manually from the CryoPod without forceps or heavy gloves.

Other observations and suggested improvements include use of a 10 L Dewar for LN₂ recharging vs the 25 L Dewar with swivel base (used at TMS): the 10 L unit is easier to handle so that LN₂ recharging can be performed by one person. The 25 L Dewar is heavier, more cumbersome when full, but was selected because at the New York site 25 L of LN₂ was the minimum quantity available from the supplier. Selecting a weekly 10 L delivery from an alternate supply company is an option.

Conclusions

These pilot studies have assisted understanding of how PfSPZ vaccines can be transitioned from clinical trials into mid-sized immunization clinics serving travellers and military personnel. The cryogenic cold chain was effective and efficient: all shipments of surrogate vaccine were successfully sent, delivered, received, monitored, maintained in local storage, accessed for ‘immunization events,’ repacked for return and reverse-shipped. Only one non-critical temperature excursion occurred and all LN₂ recharging events were handled appropriately. The equipment functioned well, and several constructive recommendations were made by the sites to improve ergonomics and efficiency. These recommendations are being incorporated into plans for scale up and the eventual commercial launch of PfSPZ vaccines for use.

Funding

This study was supported by the National Institute of Allergy and Infectious Diseases at the National Institutes of Health SB1AI077262 (for studies at Traveler’s Medical Service, New York and Passport Health, Silver Spring) and the Department of Defense contract W81XWH18C0326 (for studies at Walter Reed National Military Medical Center, Bethesda, and Joint Base Lewis-McCord Madigan).

Author Contributions

Eric R. James, L.W. Preston Church, Stephen L. Hoffman, Thomas L. Richie (Conceptualization); Eric R. James, Stephen L. Hoffman (Funding acquisition); Eric R. James, L.W. Preston Church, Brian D. Robertson, Patrick W. Hickey, David J. Schwartz, Theresa D. Asare, Macie L. Jones, Rebecca W. Acosta, Alberto M. Acosta, Mark Noble, Thomas Henkel (Data curation); Eric R. James, L.W. Preston Church, Brian D. Robertson, Patrick W. Hickey, David J. Schwartz, Theresa D. Asare, Macie L. Jones, Rebecca W. Acosta, Alberto M. Acosta, Mark Noble, Thomas Henkel (Formal analysis); Eric R. James, L.W. Preston Church, Stephen L. Hoffman, Brian D. Robertson, Patrick W. Hickey, David J. Schwartz, Patrick T. Logan, Theresa D. Asare, Macie L. Jones, Jeannie L. Bay, Austin K. Roschel, Jacqueline L. Pfeiffer, Rebecca W. Acosta, Ethan Schiavi, Alberto M. Acosta, Mark Noble, Thomas Henkel, Cebrina Young (Investigation); Eric R. James, L.W. Preston Church, Stephen L. Hoffman, Brian D. Robertson, Patrick W. Hickey, David J. Schwartz, Patrick T. Logan, Theresa D. Asare, Macie L. Jones, Jeannie L. Bay, Austin K. Roschel, Jacqueline L. Pfeiffer, Rebecca W. Acosta, Ethan Schiavi, Alberto M. Acosta, Mark Noble, Thomas Henkel, Cebrina Young (Methodology); Eric R. James, L.W. Preston Church, Stephen L. Hoffman (Project administration); Eric R. James, Stephen L. Hoffman, Brian D. Robertson, Patrick W. Hickey, David J. Schwartz, Patrick T. Logan, Theresa D. Asare, Macie L. Jones, Jeannie L. Bay, Austin K. Roschel, Jacqueline L. Pfeiffer, Rebecca W. Acosta, Ethan Schiavi, Alberto M. Acosta, Mark Noble, Thomas Henkel, Cebrina Young (Resources); Eric R. James, L.W. Preston Church, Stephen L. Hoffman, Thomas L. Richie (Writing—original draft); Eric R. James, L.W. Preston Church, Stephen L. Hoffman, Thomas L. Richie, Brian D. Robertson, Patrick W. Hickey, David J. Schwartz, Patrick T. Logan, Theresa D. Asare, Macie L. Jones, Jeannie L. Bay, Austin K. Roschel, Jacqueline L. Pfeiffer, Rebecca W. Acosta, Ethan Schiavi, Alberto M. Acosta, Mark Noble, Thomas Henkel (Writing—review and editing).

Conflict of interest: None.