Abstract
Objective
To develop an interferon γ (IFN-γ) recall assay to detect a T-cell response following vaccination because antibody is seldom detected in Δcps1-vaccinated dogs.
Methods
Peripheral blood mononuclear cells were collected from naturally infected dogs with a history of coccidioidomycosis or dogs from a nonendemic area. Two- to 5-month-old Beagles were vaccinated 2 to 4 times up to 28 days apart with live, avirulent Δcps1 and bled on study days 42, 68, or 90. Peripheral blood mononuclear cells were stimulated with Δcps1 and stained to detect CD4+ T-cell IFN-γ expression. The percentage of CD4+IFN-γ+ T cells compared to unstimulated baseline was calculated for each dog.
Results
All 5 naturally infected dogs with coccidioidomycosis had T-cell IFN-γ responses, and the 2 unexposed nonendemic dogs did not. Fourteen of 15 dogs vaccinated with Δcps1 had recall IFN-γ responses. Subcutaneously vaccinated dogs (n = 5) had a mean 1.85-fold increase in IFN-γ over baseline, whereas IM-vaccinated dogs (n = 10) were a mean of 6- to 10-fold higher. The percentage of CD4+IFN-γ+ cells was higher from restimulated cells compared to baseline in both SC- and IM-vaccinated dogs (SC twice [n = 5]: 95% CI, 0.001 to 0.543; IM twice [n = 5]: 95% CI, 0.9734 to 1.948; IM 4 times [n = 5]: 95% CI, 1.207 to 5.083).
Conclusions
The T-cell IFN-γ recall assay of peripheral blood mononuclear cells from vaccinated dogs detects a cell-mediated immune response to the Δcps1 vaccine.
Clinical Relevance
This assay has potential utility as a surrogate marker for vaccine-induced immunity.
Coccidioidomycosis is a systemic fungal infection endemic to the Southwestern US, particularly Arizona and California.1,2 The disease costs approximately $1.5 billion/year as a human health problem and greater than $80 million/year as a canine disease.3 Disease prevention in dogs and humans will benefit both species by a reduction in morbidity and death as well as a reduction in medical and veterinary financial burdens. A vaccine to prevent or mitigate the disease in both dogs and humans is in late-stage development for dogs and in early development for humans.
The canine vaccine candidate Δcps1 is a live, avirulent mutant strain of Coccidioides posadasii from which the carbamoyl phosphate synthetase 1 (CPS1) gene (160 kilobases) has been completely removed and replaced with a hygromycin B resistance cassette (hygromycin B 4-I phosphotransferase) for selection and ease of detection.4 The vaccine consists of purified, live arthroconidia (spores) of the avirulent mutant administered parenterally in a prime-boost regimen. The vaccine has been shown to provide significant protection from death and several log reductions in lung fungal burdens in experimental studies4,5 in both mice and dogs.
While infectious challenge with protection is the most definitive assessment of a vaccine’s efficacy, surrogate markers for immune responses to vaccination have practical, and sometimes regulatory, value in determining that vaccinees produced an immune response to a vaccine. A surrogate marker test is desirable for the evaluation of dogs and is expected to be even more important in early-phase human studies. We have previously reported that dogs vaccinated with up to 200,000 live spores of the Δcps1 vaccine produced undetectable or transient antibody responses using 2 different commercial serology tests,5 although they exhibited strong protection.5 An in-house ELISA utilizing a recombinant truncation of the immunodominant chitinase-1 antigen also failed to detect antibody responses in vaccinated dogs, though in later research on the archived sera, a nucleic acid programmable protein array did identify reactivity to chitininase-1 antigen in some of the vaccinated dogs by 4 to 6 weeks after primary vaccination.6
Like tuberculosis, coccidioidomycosis requires cell-mediated immunity for resistance. Tests to detect Mycobacterium tuberculosis infection in humans are based on the measurement of cellular responses, including interferon γ (IFN-γ) release assays (eg, the QuantiFERON test) related to T-cell recall activation (https://www.cdc.gov/tb/hcp/testing-diagnosis/index.html). Failing antibody detection after vaccination, a Coccidioides T-cell recall assay is a reasonable consideration for a surrogate marker test of host immune memory to Δcps1. In this study, we developed a CD4+ T-cell IFN-γ recall response assay of peripheral blood mononuclear cells (PBMC) stimulated with Δcps1 spores to use as a surrogate marker for a vaccine response in dogs. We assessed CD4+IFN-γ responses specifically because we can adoptively transfer protective immunity to mice with CD4+ but not CD8+ cells.3 Prior to testing vaccinated dogs, we determined that we could detect a T-cell recall response in naturally infected dogs from an endemic area and distinguish them from unvaccinated, uninfected dogs from a nonendemic area.
Methods
Animals and blood collection
Client-owned dogs were bled with written owner consent (University of Arizona IACUC protocol No. 14-522). Dogs had a known history of coccidioidomycosis (infected; n = 5) or were from outside the endemic area with no known exposure (uninfected; n = 2). Blood was collected by jugular venipuncture into EDTA tubes. For dogs in the Δcps1 vaccine safety studies, bleeding was approved under Midwest Veterinary Services Animal Use Protocol #24032. Two- to 5-month-old Beagles (n = 5) were vaccinated twice by SC injection of 0.5 mL of Δcps1 vaccine spores in sterile water (100,000 viable spores/dose). Primary and booster vaccinations were administered on study days 0 and 28, and the dogs were bled on day 90 (SC X 2 group). Five dogs were vaccinated IM with 160,000 viable spores/dose of Δcps1 prepared as above on days 0 and 28 and bled on day 42 (IM X 2 group). Five dogs were vaccinated IM on study days 0 and 28 with 160,000 spores and twice more on days 42 and 56 with additional IM doses of 120,000 spores and bled on day 68 (IM X 4 group; Supplementary Table S1). Blood was collected from each dog once (7 to 15 mL) by jugular venipuncture into EDTA tubes at the convenience of the study personnel on the days indicated above.
Blood specimens not acquired locally (Tucson, Arizona) were transported overnight in insulated shipping containers with room-temperature gel packs to keep the temperature of the blood stable until processing. Local samples were held overnight at room temperature in an insulated box with room-temperature gel packs and processed the following day to recapitulate the handling of the shipped samples.
Peripheral blood mononuclear cell isolation and incubation
Blood was diluted 1:1 with sterile PBS and layered over a Ficoll-Paque PLUS (Cytiva Life Sciences, 17144003) gradient (density, 1.077 g/mL). Gradients were centrifuged at 1,000 X g for 30 minutes, and PBMC were collected from the interface. The remaining red cells were lysed with ammonium-chloride-potassium lysis buffer, and PBMC were washed in complete medium (RPMI 1640, 10% fetal bovine serum, 2 mM L-glutamine, 1 mM sodium pyruvate, 100 U/mL penicillin, and 0.1 mg/mL streptomycin). The viable cell concentration was determined by counting on ViCell (Becton Dickinson), and 1 X 106 cells/well were dispensed into 96-well culture plates. For Coccidioides antigen stimulation, cells were incubated with viable Δcps1 spores, ranging from 1 to 20 spores/cell, at 37 °C and 5% CO2 for 16 hours. Control wells for each dog were incubated with phorbol myristate acetate/ionomycin as a positive stimulation control and medium only for a negative control. Protein Transport Inhibitor Cocktail (eBiosciences) was added for the final 4 hours of incubation, per the manufacturer’s instructions, to allow accumulation of cytokines for intracellular staining. Stimulations were performed in triplicate where possible. Due to the number of PBMC available, only single positive and negative control wells were run from the dogs vaccinated SC X 2 and IM X 2; for the IM X 4 dogs, positive and negative controls were run in triplicate. For the SC X 2 dogs, we also had only single Δcps1-stimulated wells, whereas the IM X 2 and IM X 4 dogs had triplicate wells. Ten spores/cell (10 X 106 spores) produced the most consistent results across community and vaccinated dogs; data from the 10 spores/cell wells were used in the comparisons and statistical analyses (Supplementary Figure S1).
Staining and flow cytometry
Cells were first stained with antibodies for surface markers (anti-CD5 [clone YKIX322.3; phycoerythrin; ThermoFisher, 12-5050-42], anti-CD4 [clone YKIX302.9; Pacific Blue; BioRad, MCA1038PB], and anti-CD8 [clone YCATE55.9; Alexa Flour 700; BioRad, MCA1039A700]). Following fixation and permeabilization, cells were stained for IFN-γ (clone CC302; Alexa Fluor 647; BioRad, MCA1783A647). Data were collected using a BD Fortessa flow cytometer (BD Bioscience) and analyzed using FlowJo software, version 10.10.0 (FlowJo). Cells were gated for lymphocytes, then T cells, but the T cells were detected with Cluster of differentiation (CD)5 rather than CD3 because the CD3 receptor is downregulated on canine T cells after activation, but CD5 is not.7,8 Gating then separated CD4+ versus CD8+ T cells, and the final analysis was the percentage of CD4+ cells producing IFN-γ (Figure 1). The baseline CD4+IFN-γ+ cells from unstimulated wells for each dog were compared to the Δcps1-stimulated counts.
Canine peripheral blood mononuclear cells were separated and dispensed into 96-well plates (1 X 106 cells/well) with 1 to 20 viable spores of Δcps1/cell and incubated for 24 hours, the last 4 hours with protein inhibitor cocktail. After staining with canine antibodies for Cluster of differentiation (CD)5, CD4, CD8, and IFN-γ, cells were gated for mononuclear cells by size exclusion, then CD5+ cells. CD4+ cells were sorted from CD8+ cells, and final detection was for the percentage of CD4+ T cells producing IFN-γ. CPS1 = Carbamoyl phosphate synthetase 1. IFN-γ = Interferon γ. Comp-Alexa-A = Alexafluor 700 area. Comp-APCA = Allophycocyanin area. Comp-Pacific = Pacific Blue area. FSC = Forward scatter. SSC = Side scatter.
Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.25.04.0119
Statistical analysis
The T-cell IFN-γ recall response for each dog was assessed by comparing unstimulated cells to Δcps1-stimulated cells with 10 spores/cell. The mean fold change of the percentage of CD4+IFN-γ+ cells was calculated for each dog, and then a group mean was determined arithmetically. The percentage of CD4+IFN-γ+ cells compared to baseline was analyzed for each group of vaccinated dogs by a 2-tailed t test following a test for normality on each group of dogs. Ninety-five percent CIs and P values were calculated. Statistics were performed using Prism, version 10.4.0 (GraphPad Software LLC).
Results
Community dogs
Client-owned dogs consisted of 3 dogs with resolved clinical coccidioidomycosis illness (2 pulmonary and 1 disseminated disease) or were asymptomatic but persistently seropositive by antibody gel immunodiffusion at commercial laboratories over a period of > 2 years (n = 2) and 2 dogs from an East Coast state that had never been in the endemic area. The nonendemic dogs, which were presumed to be naïve with no travel history to the Western US, produced essentially no IFN-γ after stimulation with Δcps1 spores, whereas recovered and asymptomatic dogs had 2.6- to 16.8-fold increases above unstimulated control wells (Figure 2). Both infected and nonendemic dogs produced IFN-γ to phorbol myristate acetate/ionomycin (data not shown because of graph scale). This demonstrated that the assay is able to detect recall T-cell responses in naturally infected dogs while unexposed dogs did not respond (Supplementary Table S2).
Interferon γ produced by CD4+ T cells of naturally infected and unexposed client-owned dogs stimulated with Δcps1 spores (10 spores/cell). A—Percentage of CD4+ T cells producing IFN-γ after stimulation with Δcps1 (red) compared to the medium-only wells (NS) for each dog. B—Data expressed as fold change from background (medium only) showing that the fold change over background in the infected dogs ranged from 2.56 to 16.79, whereas the unexposed dogs did not respond.
Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.25.04.0119
Dogs vaccinated with Δcps1 SC
Five dogs (SC X 2) were vaccinated twice SC on study days 0 and 28 with 100,000 spores and bled on study day 90 (62 days after vaccination) as described above. Four of the 5 dogs (80%) had a 1.57- to 3.31-fold increase in the percentage of CD4+IFN-γ+ cells, and one dog did not make increased IFN-γ to Δcps1 spore stimulation. Including the nonresponder, the SC X 2 group had a mean 1.85-fold increase in IFN-γ (Figure 3). Analysis of the percentage of CD4+IFN-γ+ cells over baseline barely reached statistical significance (P = .049; 2-tailed t test) and had a 95% CI of 0.001 to 0.543 (Figure 4).
Mean fold change in the percentage of CD4+IFN-γ+ T cells of dogs vaccinated with live, avirulent Δcps1 vaccine spores. A—Five dogs were vaccinated twice SC on study days 0 and 28 and bled on day 90 (SC X 2). Four of 5 dogs produced an IFN-γ recall response, and one failed to respond. The mean fold change was 1.85 including the nonresponder and 2.09 without it. B—Five dogs were vaccinated IM on study days 0 and 28 and bled on study day 42 (IM X 2). All dogs responded, and the mean fold change was 6.36. C—Five dogs were vaccinated IM on study days 0, 28, 43, and 56 and bled on study day 68 (IM X 4). The mean fold change for dogs vaccinated 4 times was 10.03. The solid horizontal line is 1 (no fold increase) on each graph. There was only a single stimulated well for the SC X 2 group and triplicates for others.
Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.25.04.0119
CD4+ T-cell IFN-γ recall responses in NS and Δcps1-stimulated (CPS1) peripheral blood mononuclear cells from dogs vaccinated SC (A) or IM (B). Dogs were vaccinated SC twice (SC X 2) and IM either twice (IM X 2) or 4 times (IM X 4). The plots show the mean and 95% CI. There was a statistically significant increase in the percentage of CD4+IFN-γ+ cells in all 3 groups of dogs, though it was borderline (P = .049) for the SC X 2 group. The 95% CI in the SC X 2 group, with a nonresponsive outlier, overlaps with baseline responses. The IM route of vaccination in this study produced higher recall IFN-γ responses than SC, though other factors may have contributed to this observation. *P = .05. ****P = .0001; unpaired t test.
Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.25.04.0119
Dogs vaccinated with Δcps1 IM
Ten dogs were vaccinated on study days 0 and 28 with 160,000 spores of Δcps1 and were divided into 2 groups: 5 dogs vaccinated twice (IM X 2) and 5 dogs vaccinated 4 times (IM X 4). The IM X 2 dogs were bled on study day 42 (14 days after vaccination). The IM X 4 dogs were vaccinated twice more on study days 42 and 56 with 120,000 spores IM and bled on study day 68 (12 days after vaccination). All 10 of the dogs vaccinated IM and bled 14 or 12 days later, respectively, had recall responses to Δcps1 ranging from 4.7- to 9-fold increases in the percentage of CD4+IFN-γ+ T cells for the IM X 2 dogs and 3- to 20-fold for the IM X 4 dogs (Figure 3). Figure 4 shows that CD4+IFN-γ+ T cells were significantly higher compared to baseline after restimulation with Δcps1 in dogs vaccinated at least twice IM (IM X 2 95% CI, 0.9734 to 1.948; IM X 4 95% CI, 1.207 to 5.083; P < .0001 for both). This indicates that we can detect an immune response to the vaccine by measuring T-cell IFN-γ recall responses in dogs.
Discussion
The most common surrogate markers of vaccine immunity in infectious diseases are serum antibodies.9 However, for the Δcps1 live, avirulent coccidioidomycosis vaccine, we have previously published that dogs vaccinated with doses of the vaccine that induced protection against infection did not generate consistent antibody responses detectable by either a commercial enzyme immunoassay or commercial immunodiffusion tests.5,10 Since host immune control requires T helper 1 and T helper 17 cellular immunity to resolve disease,11,12 and since the dogs did not make antibodies to the vaccine, we sought a cell-mediated surrogate marker test. The IFN-γ recall assay described here successfully detected an immune response to the Δcps1 vaccine in 14 of 15 dogs. A significant limitation of this assessment is that there were no unvaccinated study dogs and no prevaccination samples from the study dogs for comparison, although 2 uninfected, nonendemic community dogs failed to make IFN-γ when stimulated with Coccidioides antigen.
Interferon γ is highly associated with protective anti-Coccidioides host responses.11,13 Researchers have shown that recall production of IFN-γ after Coccidioides-specific stimulation is indicative of a controlling host response in humans, both in recovered patients and asymptomatically infected people.14,15 This study preliminarily demonstrates that Δcps1-vaccinated and naturally infected dogs both develop a T-cell IFN-γ recall response to the live, avirulent Δcps1 vaccine spores, and with current technology, this is a practical method of assaying for vaccine response in a research setting at least. Restimulation of human PBMC with coccidioidal antigens has employed soluble preparations of spherule/endospore culture filtrates.14,16,17 We found that the culture filtrates produced unpredictable results in canine PBMC, but both live and dead spores of the avirulent vaccine strain stimulated canine cells predictably at a concentration of 10 spores/cell; we used viable spores because we have ready access to them, and they produce slightly higher results. We have also stimulated human coccidioidomycosis patient PBMC and shown CD4+ T-cell IFN-γ recall responses (DA Powell, PhD, University of Arizona, 2024-2025, unpublished). We are confident that this IFN-γ recall assay using Δcps1 spores will be easily adapted as a surrogate marker test for vaccinated humans in phase I studies.
Though the T-cell IFN-γ recall assay is highly promising for a surrogate marker test of the Δcps1 vaccine in dogs, this study has limitations. First, it suffers from small sample sizes. The SC group dogs, which had one nonresponder as well as lower IFN-γ responses than the IM vaccinated dogs, demonstrate the large effect of this small sample size, reducing confidence in the observation. Besides the route of administration, the time from last vaccination to bleeding was very different for the SC and IM dogs because of one-time convenience sampling, and this might have also affected the magnitude of responses, though it was unlikely to have affected which dogs did or did not respond. We also had insufficient cell numbers in some of the groups, leaving us without replicate wells for some conditions, especially in the SC X 2 vaccinated group. The nonresponder in this group had a high unstimulated background but no specific response to the antigen, and we had no information about the clinical condition of the dog and no opportunity for repeat sampling. As alluded to above, we were unable to test the repeatability or durability of the recall responses due to the lack of longitudinal sampling. The range of times since the detection of illness or positive serology in the community dogs suggests that recall is as durable in dogs as it is in people,18 but a powered, preferably longitudinal, community study or vaccinated dog study would need to be performed to confidently discuss any relationship between time and recall response. Finally, we had only 2 nonendemic, uninfected community dogs to demonstrate nonresponses, which limited our ability to determine if there is overlap between unexposed dogs and infected dogs with low positive responses.
Future work includes testing serial PBMC samples in a final vaccine-efficacy study in dogs, wherein we will determine whether the detection of an IFN-γ recall response after vaccination is predictive of protection from the infection challenge and whether it correlates with quantitative disease measures, such as lung fungal burden, lymph node fungal burden, and radiographic abnormalities.5 We will also be able to test PBMC from unvaccinated control dogs for comparison.
In conclusion, we showed that we can differentiate known, naturally Coccidioides-infected dogs living in an endemic area from uninfected dogs in a nonendemic area with this T-cell IFN-γ recall response assay. While this is a small observational cohort, the results were clearcut and showed the feasibility of the assay as a surrogate marker test for Δcps1 vaccination in dogs. The robust responses of most of the vaccinated dogs confirm that this assay can indeed detect Δcps1 immune responses, making it an excellent candidate for a surrogate marker assay not just in dogs but also in people as the vaccine progresses into human development. This project is an excellent example of One Health work.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org.
Acknowledgments
The authors thank Hennessy Research Associates (Lenexa, KS) and Midwest Veterinary Services (Manhattan, KS) for collecting the vaccinated dog blood samples. They thank the community dog owners for allowing them to collect blood from their dogs for this study. They also thank the American Kennel Club Canine Health Foundation for their support of the foundational community dog study of T cells in naturally infected dogs that led to the development of the surrogate vaccine marker assay.
Disclosures
Lisa Shubitz is a coinventor of the Δcps1 live, avirulent vaccine (patent No. WO2014164843-A1). The remaining authors have nothing to disclose.
No AI-assisted technologies were used in the composition of this manuscript.
Funding
This work was funded by the National Institute of Allergy and Infectious Diseases grant No. R01-AI-132140, the American Kennel Club Canine Health Foundation grant No. 2408, and donations to the Valley Fever Center for Excellence directed to companion animal research.
ORCID
Christine D. Butkiewicz https://orcid.org/0000-0001-5019-7567
References
- 1.↑
Cole GT, Kirkland TN, Sun SH. An immunoreactive, water-soluble conidial wall fraction of Coccidioides immitis. Infect Immun. 1987;55(3):657–667. doi:10.1128/iai.55.3.657-667.1987
- 2.↑
McCotter OZ, Benedict K, Engelthaler DM, et al. Update on the epidemiology of coccidioidomycosis in the United States. Med Mycol. 2019;57(suppl 1):S30–S40. doi:10.1093/mmy/myy095
- 3.↑
Galgiani JN, Shubitz LF, Orbach MJ, et al. Vaccines to prevent coccidioidomycosis: a gene-deletion mutant of Coccidioides Posadasii as a viable candidate for human trials. J Fungi (Basel). 2022;8(8):838. doi:10.3390/jof8080838
- 4.↑
Narra HP, Shubitz LF, Mandel MA, et al. A Coccidioides posadasii CPS1 deletion mutant is avirulent and protects mice from lethal infection. Infect Immun. 2016;84(10):3007–3016. doi:10.1128/IAI.00633-16
- 5.↑
Shubitz LF, Robb EJ, Powell DA, et al. Δcps1 vaccine protects dogs against experimentally induced coccidioidomycosis. Vaccine. 2021;39(47):6894–6901. doi:10.1016/j.vaccine.2021.10.029
- 6.↑
Koehler MA, Song L, Grill FJ, et al. Discovery of a unique set of dog-seroreactive Coccidioides proteins using nucleic acid programmable protein array. J Fungi (Basel). 2024;10(5):307. doi:10.3390/jof10050307
- 7.↑
Pantelyushin S, Ranninger E, Bettschart-Wolfensberger R, Vom Berg J. OMIP-065: dog immunophenotyping and T-cell activity evaluation with a 14-color panel. Cytometry A. 2020;97(10):1024–1027. doi:10.1002/cyto.a.24168
- 8.↑
Chow L, Wheat W, Ramirez D, Impastato R, Dow S. Direct comparison of canine and human immune responses using transcriptomic and functional analyses. Sci Rep. 2024;14(1):2207. doi:10.1038/s41598-023-50340-9
- 9.↑
Hogrefe WR. Biomarkers and assessment of vaccine responses. Biomarkers. 2005;10(suppl 1):50–57. doi:10.1080/13547500500216629
- 10.↑
Pappagianis D, Zimmer BL. Serology of coccidioidomycosis. Clin Microbiol Rev. 1990;3(3):247–268. doi:10.1128/CMR.3.3.247
- 11.↑
Ward RA, Thompson GR, Villani AC, et al. The known unknowns of the immune response to Coccidioides. J Fungi. 2021;7(5):377. doi:10.3390/jof7050377
- 12.↑
Hung CY, Hsu AP, Holland SM, Fierer J. A review of innate and adaptive immunity to coccidioidomycosis. Med Mycol. 2019;57(suppl 1):S85–S92. doi:10.1093/mmy/myy146
- 13.↑
Kirkland TN, Hung CY, Shubitz LF, Beyhan S, Fierer J. The host response to coccidioidomycosis. J Fungi. 2024;10(3):173. doi:10.3390/jof10030173
- 14.↑
Ampel NM, Nelson DK, Chavez S, et al. Preliminary evaluation of whole-blood gamma interferon release for clinical assessment of cellular immunity in patients with active coccidioidomycosis. Clin Diagn Lab Immunol. 2005;12(6):700–704. doi:10.1128/CDLI.12.6.700-704.2005
- 15.↑
Ampel NM, Christian L. In vitro modulation of proliferation and cytokine production by human peripheral blood mononuclear cells from subjects with various forms of coccidioidomycosis. Infect Immun. 1997;65(11):4483–4487. doi:10.1128/iai.65.11.4483-4487.1997
- 16.↑
Ampel NM, Kramer LA, Li L, et al. In vitro whole-blood analysis of cellular immunity in patients with active coccidioidomycosis by using the antigen preparation T27K. Clin Diagn Lab Immunol. 2002;9(5):1039–1043. doi:10.1128/CDLI.9.5.1039-1043.2002
- 17.↑
Ampel NM, Nesbit LA, Nguyen CT, et al. Cytokine profiles from antigen-stimulated whole-blood samples among patients with pulmonary or nonmeningeal disseminated coccidioidomycosis. Clin Vaccine Immunol. 2015;22(8):917–922. doi:10.1128/CVI.00280-15
