Malaria Diagnosis using Paper-Based Immunoassay for Clinical Blood Sampling and Analysis by Miniature Mass Spectrometer

Abstract

In this work, we have developed a paper-based microfluidic device capable of remote biofluid collection followed by an analysis of the dried clinical samples using a miniature mass spectrometer. We have evaluated portable mass spectrometer as a possible surveillance platform by analyzing clinical malaria samples (whole blood) collected from Ghana. We synthesized pH-sensitive ionic probes and coupled them to monoclonal antibodies specific to Plasmodium falciparum histidine-rich protein 2 (PfHRP2) malaria antigen. We then used the antibody-ionic probe conjugates in a paper-based immunoassay to capture PfHRP2 antigen from untreated whole blood. After the immunoassay, the bound ionic probes were cleaved, and the released mass tags were analyzed through an on-chip paper spray mass spectrometry strategy. During process optimization, we determined the detection limit for PfHRP2 in untreated human serum to be 0.216 nmol/L when using the miniature mass spectrometer. This sensitivity is comparable to the World Health Organization’s suggested threshold of 0.227 nmol/L for PfHRP2, proving that our method will be applicable to diagnose symptomatic malaria infection (≥200 parasites per μL blood). The paper device can be stored at room temperature for at least 25 days without affecting clinical outcome, with each stored paper chip offering good repeatability and reproducibility (RSD = 4 - 12%). The stability and sensitivity of the developed paper-based immunoassay platform will allow miniature mass spectrometers to be used for point-of-care malaria detection as well as in large-scale surveillance screening to aid eradication programs.

Graphical Abstract

INTRODUCTION

Recent global efforts to control and prevent malaria have resulted in substantial declines in malaria cases and deaths^1-3. In 2016, this progress encouraged the formation of the E-2020 initiative by the World Health Organization (WHO), which identified 21 countries with the potential to achieve zero indigenous malaria cases by 2020¹. Of these, 2 countries (Algeria and Paraguay) are currently certified as malaria-free⁴. In February 2021, WHO admitted that the global progress in the malaria response has stalled, levelled off, and even reversed in some regions⁵. A Lancet Commission report calls for a worldwide strategic plan to eradicate malaria by 2050 and highlights surveillance (and better use of existing technologies) as one of the three pillars of successful malaria elimination^6,7.

The current study aims to develop a paper-based microfluidic device for remote sample collection followed by the analysis of protein antigens in the dried clinical samples using a miniature mass spectrometer for malaria diagnosis. We seek to evaluate mass spectrometry (MS) as a surveillance platform for malaria eradication programs. Our long-term goal is to take advantage of the high sensitivity of mass spectrometers to identify asymptomatic malaria infection, which occurs in approximately 80% of the population in sub-Saharan Africa⁸. We believe such MS-based technology in the malaria diagnostic process can provide three unique access points: (1) point-of-care (POC) application, (2) community-based surveillance detection to identify people with latent infection (i.e., asymptomatic patients) that serve as reservoirs for continuous transmission of the disease, and (3) field analysis in the case of an out-break (occurring every rainy season in endemic regions). Such capabilities become critical for National Malaria Control programs that conduct routine malaria surveillance from remote sentinel sites. Current MS methods are unable to operate under all the three levels because of requirements for extensive sample treatment and complexity of the instrument. In our approach, whole blood can be analyzed directly without sample pre-treatment, and we use simple miniature mass spectrometer for analysis with potential for YES/NO output.

Currently, malaria is diagnosed by three main methods: the gold standard light microscopy readings^9-11, rapid diagnostic tests (RDTs)¹², and polymerase chain reactions (PCR)^13,14. Today, all three methods provide the level of sensitivity required for low parasite densities in elimination programs¹⁵. Though light microscopy is considered a gold standard due to its ability to count parasites and determine the species from visualization, the method is technically challenging. It requires skilled personnel to measure the infected cells manually¹⁶. This labor-intensive process is subjective and prone to human errors¹⁶. Microscopy is applicable to only fresh blood samples, meaning that this technique is typically used to diagnose symptomatic malaria patients in hospital settings where blood is collected and analyzed immediately¹⁷. Temperature control (i.e., cold storage) is necessary if the liquid blood is to be analyzed later¹⁸. It is challenging to achieve temperature-controlled storage during field sampling at remote locations of developing countries. RDTs provide convenience, ease of use, and on-the-spot malaria detection away from laboratory settings. Although applicable at remote locations, RDT follows enzyme-linked immunosorbent assay (ELISA) techniques. Therefore, RDT results can be influenced by environmental factors such as temperature and humidity. These factors can drive degradation of assay/platform components and consequently affect result (colorimetric) interpretation¹⁹. It is important to note that the most useful and sensitive RDT BINAXNOW^™ cards can cost over $40 per test²⁰. This amount quickly escalates to a one-time user cost greater than $40K for routine surveillance studies that typically involve the analysis of over 1000 samples. Practically, this cost would triple if technical replicates needed to establish the variability of the protocol were to be performed. The absence of such replication studies increases false positive and negative outcomes with RDTs. Of these three methods, PCR enables malaria diagnosis using dried blood samples, aiding collection, storage, and transport of samples in large-scale surveillance testing of asymptomatic patients. However, significant sample preparation (for DNA extraction) and highly skilled personnel for PCR analysis limit its practical implementation in resource-limited settings²¹. Therefore, user-friendly platforms that can offer rapid and sensitive detection without sample pre-treatment are still needed to enable large-scale surveillance testing of asymptomatic malaria, even in resource-limited settings.

Recently, our laboratory introduced an innovative design for pH-sensitive ionic probes that allow direct MS analysis of immunoassays performed on ordinary paper substrates²². This new diagnostic platform provides prompt results with high sensitivity (from YES/NO qualitative and quantitative analysis) and allows user-friendly sample collection. The current report evaluates the performance of the platform using miniature mass spectrometer, which lowers the cost challenge and power consumption associated with traditional bench-top mass spectrometers. The miniature instrument can enable the implementation of the method in resource-limited settings.

EXPERIMENTAL SECTION

Two-Dimensional Paper Device

Two-dimensional (2D) wax-printed paper-based microfluidic devices was developed and implemented for the current study. Solid wax printing and laser cutting technologies were used to fabricate the 2D paper device on which immunoassay was performed. To enable on-surface immunoassay and on-chip MS analysis, two separate 2D wax-printed paper substrates were assembled as illustrated in Figure 1, comprising of reaction (top) and detection (bottom) layers, respectively. The immunoassay was performed in the test zones (the circular white areas in the top, reaction layer in Figure 1 by adding reagents manually, one step after another. This process utilized a wax-printed 2D paper substrate that is functionalized with aldehyde groups (prepared in-house). The techniques for aldehyde functionalization and wax-printing in the paper are described in the Supplementary Information (Figure S1).

Figure 1.

Schematic illustration of 2D wax-printed paper microfluidic device consisting of two paper layers: reaction layer and detection layer. Four circular test zones are created on the reaction layer for implementation of immunoassay. The detection layer served for on-chip paper spray ionization mass spectrometry (on-chip PSMS). The anti-PfHRP2 capture antibodies were covalently immobilized in test zones followed by blocking of vacant site using 1X Tris buffer saline (TBS). Thus, this bioactive paper chip would selectively capture PfHRP2 and enable the immobilization of the detection antibody-ionic probe bioconjugate in serial reagent additions for positive samples. Subsequently, on-chip PSMS was performed after cleaving the mass tag with the addition of cleaving solution. Abbreviation: cAb, capture antibody; PfHRP2, Plasmodium falciparum histidine-rich protein 2; dAb, detection antibody; TBS, tris buffer saline; on-chip PS, on-chip paper spray.

Chemicals and Reagents

Whatman 1. Chromatography paper and Whatman gel blotting paper and Grade GB003 (20 x 20 cm) were purchased from Sigma Aldrich (St. Louis, MO). Water was prepared using a Milli-Q integral system with a resistivity of 18.2 MΩ-cm (Merck Millipore, Burlington, MA), a silver wire electrode (O.D. 1.5 mm) was purchased from Warner Instruments (Hamden, CT). Borosilicate capillaries (I.D. 0.86 mm) were purchased from Sutter Industries (Novato, CA). (3-Darboxypropyl)trimethylammoniumchloride, 1-Ethyl-3-(3′-dimethylaminopropyl)carbodiimidehydrochloride, 4-Dimethylamino-pyridine, acetylcholine, methacholine, acetonitrile (HPLC grade), phosphate buffered saline tablet (PBS; pH 7.4), Tris buffered saline (PBS, 10X), and Potassium periodate were purchased from Sigma Aldrich. Sodium bicarbonate and sodium carbonate were purchased from Fisher Scientific Co. (Hampton, NH, USA), 4-(2-Hydroxyethyl)phenyl isothiocyanate was purchased from Organix Inc (Woburn, MA, USA). Recombinant P. falciparum Histidine-rich protein II (His tag, ab227569) was purchased from Abcam, Inc (Cambridge, MS, USA), The anti-Malaria PfHRP2 IgG monoclonal antibody (ABMAL-0444, Clone 44) and the anti-Malaria PfHRP2 IgG monoclonal antibody (ABMAL-0445, Clone 45) were purchased from Arista Biologicals Inc. (Allentown, PA, USA) and used for capture antibody and detection antibody, respectively. Human serum was obtained from innovative research Inc. (Novi, MI, USA). Eppendorf tubes were purchased from Fisher Scientific Co. (Hampton, NH). Amicon Ultra 0.5 mL Centrifugal Filters, 100 Ka was purchased from Millipore Sigma (Burlington, MA, USA). Micro G-25 Spin Columns were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA).

Mass Spectrometry

LTQ linear ion trap mass spectrometer (Thermo Scientific, San Jose, CA, USA) and a ContinuityTM miniature mass spectrometer (ion trap from BaySpec Inc., San Jose, CA, USA) were used for acquiring mass spectrometry (MS) data. Two ionization platforms were employed: on-chip paper spray (on-chip PS) and nano-electrospray ionization (nano-ESI). On-chip PSMS was facilitated by using 4.5 kV direct current (DC) spray voltage and MeOH/H2O (80/20, v/v %) for spraying solvent. Nano-ESI was operated using 1.5 kV DC spray voltage applied to a silver electrode in contact with sample solution placed in pulled glass capillary. All experiments were performed in positive-ion mode. Tandem MS (MS/MS) with collisional-induced dissociation (CID) was performed for structural elucidation of analytes and identifying diagnostic ions used in quantification. Detailed experimental information is described in the Supplementary Information.

UV-VIS Spectrophotometer

SpectraMax QuickDrop Micro-Volume Spectrophotometer (Molecular Devices, San Jose, CA, USA) was used for confirming bioconjugation of detection antibody-probe. We used a nano-drop system of a spectrophotometer, which required 2 μL of samples for analysis.

Clinical Sample Preparation

Clinical samples were collected from volunteers from the Agona area within the Sekyere South District, Ghana. Para-site densities were determined by a microscopy-based detection method. The whole blood samples (250 μL) were stored at −80 °C. Each sample was aliquoted into 10 μL and the frozen whole blood sample was thawed at the time of analysis. The protocol for collecting clinical samples was approved by Institutional Review Boards from both research sites: Ohio State University (Study Number: 2020H0539) and Kwame Nkrumah University of Science and Technology (KNUST), Ghana (CHRPE/AP/332/19 and CHRPE/AP/377/20).

RESULTS AND DISCUSSIONS

Paper-Based Immunoassay and the Use of Ionic Probes

One of the pillars in malaria eradication program is the establishment of an efficient surveillance strategy. Although our long-term goal is to develop an automated²³ paper device for antigen capture from untreated blood, the current study used 2D wax-printed paper-based microfluidic device for process optimization.

We tested Plasmodium falciparum (Pf) malaria, which accounts for most severe malaria cases and specifically targeted histidine-rich protein 2 (HRP2)²⁴ biomarker. First, 2 μL of monoclonal capture antibodies (cAb) specific to Plasmodium falciparum histidine-rich protein 2 (PfHRP2) malaria antigen were covalently immobilized in the test zones. Next, the immobilization of cAb in the wax-printed 2D aldehyde-functionalized paper is followed by blocking vacant aldehyde sites with 1X Tris buffer saline (TBS) to prevent non-specific binding. This step finally yields a bioactive paper chip, which we used in subsequent immunoassays to selectively capture the malaria antigen PfHRP2 from complex biofluids. Note that this bioactive paper chip is prepared prior to any sampling experiment. Thus, the process of immobilizing the capture antibody in the paper does not count toward immunoassay time. To start immunoassay, the biofluid sample containing antigen (20 μL) is pipetted into the test zones for 15 minutes, after which the test zone was washed three times with 1X phosphate buffer saline (PBS). After the washing step, 5 μL solution of the detection antibody conjugated with our ionic probe was added to the test zone for another 15 minutes of reaction time. Lastly, a second washing step was performed with 1X PBS (three times). This completes the paper-based immunoassay and the 2D paper device on which PfHRP2 antigen has been captured can be analyzed immediately or stored at room temperature for later analysis by MS. In future, this test will involve a 3D microfluidic paper device within which reagent can be stored to eliminate manual pipetting.²³

For the production of pH-sensitive ionic probe, the current work developed and optimized a more straightforward synthetic method based on Steglich esterification²⁵ (Scheme 1).

Scheme 1.

Reaction scheme for synthesis of cleavable ionic probe by Steglich esterification. The final product (3) embodies three important parts: mass tag (blue), cleavable ester functional group (red), and conjugation unit (black). Number of carbons (n) between quaternary and carboxylic groups is 3 for current project although different n can be used.

This method avoids the challenging first activation step (i.e., preparing acyl chloride) used in our previous work that involved the dissolution of the polar starting reagent 1 in organic thionyl chloride solvent. We focused on the synthesis and application of 4-(4-isothiocyanatophenethoxy)-N, N, N-trimethyl-4-oxobutan-1-aminium chloride (ITBA, 3 (n = 3), Scheme 1) since our previous studies showed ITBA to offer more sensitive MS analysis than when using a shorter probe length (n = 1) in paper-based immunoassays. This cleavable ionic probe allows direct MS analysis (on-chip) of immunoassays performed on ordinary paper substrates. This capability is made possible because of the three unique features incorporated into the design of the ionic probe: (1) a mass tag for easy detection by electrospray, (2) a pH-sensitive ester group for release of the mass tag, and (3) a conjugate unit with isothiocyanate functional group for coupling of the whole ionic probe to detection antibody²². That is, we attached the cleavable ionic probe to anti-PfHRP2 antibodies through the reaction between lysine residues in the antibody. This bioconjugate of ionic probe serves as the detection antibody (dAb) in the immunoassay process. The pH-sensitive ionic probe and the reporter antibody conjugates were characterized and are fully described in the Supporting Information (Figures S2-4).

Analysis of Serum Samples and Assay Optimization

First, we used serum samples spiked with PfHRP2 at varying concentrations (0.01 – 100 nmol/L) for immunoassays performed on the 2D bioactive paper chip, following the immunoassay procedure described above (Figure 1). Control experiments followed a similar process except using a blank human serum that contains no PfHRP2 antigen. Note that it is only when the PfHRP2 antigen is present in the test zone of the paper chip that the detection antibody will be bound, immobilizing the ionic probe onto the paper substrate.

To initiate MS analysis, the ionic probe is cleaved by adding 2 μL of cleaving solution, which is 1.0 M NH₄OH/ACN (50/50, v/v %) solution containing internal standard (100 nmol/L, methacholine), to the test zone. The cleaved 3-carboxypropyltrimethylammonium chloride (CPTA, 1 in Scheme 1) species was analyzed by two ionization methods, nano-electrospray ionization (nano-ESI) and direct on-chip paper spray (on-chip PS). Optimization studies involving experimental parameters such as spray voltage, distances between emitter tip and MS inlet, ionization efficiency, and spray solvent were separately performed (Figures S5 and S6). Typical positive-ion mode paper spray MS/MS spectra for the cleaved probe and selected internal standard are shown in Figures 2A and B, respectively. Our studies showed direct on-chip PS to be more sensitive than nano-ESI, which is related to the high efficiency of transferring the cleaved probe from the paper substrate. We observed 98% transfer efficiency for the direct on-chip PS ionization upon a single application of 5 μL spray solvent (Figure 2C and D). On the contrary, transferring the 2 μL cleaving solution from the paper strip to the glass capillary of the nano-ESI emitter was about 50% efficient; a significant amount of the CPTA probe remained in the paper surface. This fact is reflected in the high sensitivity calculated for the on-chip PS ionization method (Table S1). Here, quantification was achieved through internal standard calibration, where the MS/MS signal from the cleaved probe (m/z 87) was compared with that from the internal standard (methacholine, m/z 101). Data derived from the complete assay covering the entire concentration range of the calibration curve is shown in Figures 2E and F for on-chip PS MS/MS analysis, performed on both miniature and bench-top ion trap mass spectrometers, respectively. Good linearity (R ≥ 0.99) was achieved within the concentration ranges of 0.1 – 2.5 nmol/L for both instruments (Figure S7). The limits of detection (LOD) were calculated as 0.216 nmol/L and 0.028 nmol/L for miniature and bench-top mass spectrometers, respectively. Additionally, quantitative studies were performed utilizing nano-ESI technique on both miniature and bench-top mass spectrometers, showing LOD 0.433 nmol/L and 0.374 nmol/L, respectively (Figure S8). The linear dynamic range for on-chip paper spray and nano-ESI recorded on the portable instrument are 0.1 - 2.5 nmol/L and 1.0 – 25 nmol/L, respectively. See similar data for the bench-top instrument in Table S1, as well as other analytical merits including linearity (R²) and standard deviations. The recommended LOD for clinical identification of PfHRP2 in RDT is 9.1 ng/mL²⁷ per WHO guidelines, which is equivalent to 0.227 nmol/L and 200 parasites per μL. The sensitivity obtained from coupling paper-based immunoassay with a miniature mass spectrometer is within this recommended range, although it is 10X lower than the sensitivity recorded from a standard bench-top mass spectrometer. This suggests that miniature mass spectrometers can be used to diagnose symptomatic malaria infection, considering that the vast majority of symptomatic malaria patients have more than 200 parasites per μL of blood.

Figure 2.

On-chip PSMS analysis of PfHRP2 spiked in human serum. Prior to analyte quantification, the structure of the mass tag (CPTA) and internal standard (methacholine) were characterized by MS/MS, which gave diagnostic fragment ions m/z 87 and 101, respectively (Figures 2A and B). The analysis of the cleaved mass tag in the immunoassay test zones was achieved by two ionization techniques, nano-ESI and on-chip paper spray. The efficiency of transferring the cleaved mass tag from the paper substrate to the proximal mass spectrometer was evaluated and we observed that on-chip PSMS to offered close to 98% efficiency upon a single addition of elution solvent compared to 50% for nESI (Figure 2C). This gives high signal for the on-chip PSMS method (Figure 2D). Using the on-chip paper spray ionization, calibration curves were generated using a miniature mass spectrometer (Figure 2E) and bench-top mass spectrometer (Figure 2F) within the range of 0.01 −100 nmol/L of spiked PfHRP2 human serum. The limit of detections (LOD) were calculated as 0.216 nmo/L and 0.029 nmol/L for miniature- and bench-top MS, respectively, followed by the equation, LOD = Mean _blank + 3.3 x SD_blank²⁶.

Device Stability, Before and After Immunoassay

As we aim to develop an approach toward three distinct levels of malaria detection, the stability of the device becomes essential for large-scale surveillance screening where samples must be collected and analyzed later. Two aspects of device stability were investigated: (i) Stability I, which involved room temperature storage of the bioactive paper strip containing the capture antibody before implementing immunoassay and (ii) Stability II, studied after the sandwich immunoassay has been completed and the PfHRP2 antigen has been captured onto the paper substrate. The first stability study will dictate the shelf life of our bioactive paper chip. The second stability study is also important because it indicates whether sample collection can be decoupled from the analysis steps. Such capability will enable few healthcare workers to be used for large-scale surveillance programs since flexible sampling requirement can be implemented easily without time restrictions related to signal decay or instability of assay reagents. These stability tests will also determine whether our paper-based immunoassay platform can be applied in direct-to-customer testing where mail-in services allow individual patients to order a medical test and perform home collection for themselves. Currently, majority of direct-to-customer tests are focused on DNA analysis²⁸. Enzyme-based assays do not offer this emerging test because of cold storage requirements, without which enzymatic activity is lost quickly^22,29.

Human serum samples spiked with 100 nmol/L PfHRP2 were used for these stability studies. For the Stability I experiment, which evaluated the activity of the capture antibody, the bioactive paper chips were stored in a drawer at room temperature for up to 25 days. Then, on the analysis day, immunoassay was performed on the stored paper strips using 100 nmol/L of PfHRP2 in serum. After the immunoassay, the paper strip was analyzed immediately using on-chip paper spray MS. The results for this stability study are summarized in Figure 3A, where a stable positive signal was observed irrespective of how long the bioactive paper chip was stored in ambient air under room temperature conditions. The results for the second set of stability studies (Stability II) are provided in Figure 3B, in which the immunoassay was completed before initiating the storage process. The immunocomplex (including the malaria antigen and ionic probe), while still present on the paper surface were stored in ambient air for a maximum of 25 days before MS analysis. Here too, good stability was recorded for positive tests regardless of storage time. In both stability studies, the positive samples containing antigen showed approximately 5X higher signal intensity compared to control samples in which blank serum was used. In addition, good repeatability was observed from both stability studies, indicating relative standard deviation (RSD) values of 12 % and 4.14 %, respectively, for Stabilities studies I and II (Figure 3).

Figure 3.

Stability of stored paper chip studied under the ambient conditions to investigate (A) bioactivity of capture antibody immobilized in the paper before sample application (Stability I) and (B) test stability after immunoassay is completed (Stability II). Human serum spiked with 100 nmol/L PfHRP2 and 1% Bovine Albumin Serum in 1X PBS (1% PBSA) were used as positive and control samples, respectively. Error bars indicated standard deviations for six replicates. The capture antibody and the immuno-complex were found to be stable for over 25 days of storage in ambient air, with good relative standard deviation, 12 % and 4.14 % for stability I and II, respectively.

The technology described in this study will allow clinical studies performed in resource-poor environments to benefit from existing instrumentation anywhere else in the world. This capability is made possible because of the stable ionic probes (instead of enzymes) that mitigate the vulnerability to temperature changes, such as cool storage requirement. Although a full study focusing on temperature and humidity effects has not been performed yet, the design and chemical nature of the ionic probe makes it less prone to environmental stressors enabling the device to be stored under conditions not amendable to conventional methods. The 20 μL of biofluid sample requisite can be achieved through a less invasive microsampling approach such as finger prick, allowing at-risk patients such as infants to be included in surveillance studies. We note that immunoassay results can also be analyzed immediately after the test, offering opportunities for point-of-care applications.

Analysis of Clinical Malaria Samples

The study was designed to use clinical samples, collected from Ghana by our collaborators at Kwame Nkrumah University of Science and Technology in Kumasi, Ghana as a validation. Hundred patient blood samples were collected and shipped on ice, but most of the samples were lysed before arriving us in the United States. This incident represents a major challenge for shipping liquid biofluids. Our method will allow dried blood samples to be shipped at room temperature when the platform is fully developed. Regardless, we identified three malaria-positive and three malaria-negative whole blood samples that could be used in this initial study. Information for these six clinical samples are provided in Table S2. We sought to use these samples to test the ability of our paper-based immunoassay platform to 1) identify symptomatic malaria infections and to distinguish them from clinically certified negative samples and 2) statistically correlate signals derived from on-chip PS MS analysis to parasite density determined by light microscopy. We followed a similar immunoassay process as described earlier, using whole blood clinical samples.

Positive and negative samples showed distinctively different results (Figure 4). An order of magnitude higher signal was recorded for malaria-positive samples compared with the MS signal derived from the malaria-negative samples. This result indicates that coupling a bioactive paper chip with MS detection utilizing the cleavable ionic probe-based immunoassay strategy can diagnose symptomatic malaria infection.

Figure 4.

Clinical studies were performed using whole blood collected from malaria patients. Three samples were randomly picked from positive (GKN-0224, 0133, 1249) and negative (GKN-0198, 0172, 0173) sample pools for analysis. The whole blood (5 μL) was prepared from each sample and loaded onto the paper device. Distinctive results were observed from positive and negative samples as represented in Figure 4A, where five replicates were tested for each sample. Spearman statistical analysis (Figure 4B) was performed to compare our proposed MS-based platform to a gold standard, light microscopy technique. Good correlation value (r = 0.9411) was obtained.

Furthermore, we used Spearman statistical analysis³⁰ to examine the degree of correlation between our method, which utilized paper-based immunoassay coupled to MS, and the light microscopy detection method. A high correlation level was obtained between the two detection techniques, representing r = 0.9411 (P ≤ 0.05), albeit small sample size. Our ongoing field studies will utilize a higher number of patient samples. The results of the six clinical sample analysis corroborate the assessment where parasite densities were much higher than the lowest recommended level (200 parasites per μL).²⁷ The excellent prediction of infection status, as compared to light microscopy analysis, indicate minimal non-specific binding in the bioactive paper chip and shows the high efficiency of our platform when the complexity of samples was increased from serum to untreated whole blood. We believe our diagnostic approach could become an effective alternative to light microscopy when applied in point-of-care diagnosis since it can alleviate challenges associated with human errors in data interpretation. The low cost per test (i.e., a whole community can use a single miniature mass spectrometer and for many other applications), high sensitivity, and the ability to perform technical replicates within short periods allow our device to have a competitive advantage over RDTs. Moreover, our ability to perform direct analysis from whole blood without sample pre-treatment has the potential to simplify complex mixture analysis when compared with PCR.

Another favorable feature includes the capacity to perform on-chip detection from only 20 μL of a sample. Collectively, these features will allow our approach to meet FDA CLIA waiver³¹ requirement and to implement in various biomedical applications. Thus, the lifetime cost (not cost per test) of the miniature mass spectrometer can be easily recovered by taking advantage of the versatility of the approach.

Some areas of improvement are identified for the proposed method. The 2D wax-printed microfluidic paper device is implemented using a manually step-by-step approach. In other words, although remote sample collection is possible (something necessary for surveillance studies), the paper-based immunoassay can still be tedious. We believe a 3D microfluidic paper device can provide an automatic process by which the immunoassay can be accomplished²³. The automated process is achievable because reagents can be stored dry within the confinements of the 3D microfluidic paper device. In that case, we envision a straightforward self-testing platform where the patient performs only two simple tasks: applying a drop of blood onto the paper device, followed by a wash step. The MS analysis will be performed after the delivery of device to a centralized facility. Moreover, we expect that a suitable amplification method can improve the sensitivity to detect asymptomatic malaria infection, with parasite density less than 200 per μL of blood.

CONCLUSION

We have demonstrated a mass spectrometry-based diagnostic platform utilizing a 2D paper microfluidic device and a miniature mass spectrometer. We successfully applied the platform to diagnose clinical malaria infection. The use of the ionic probe in the paper-based immunoassay offered improved stability and robustness over the conventional colorimetric-based assays such as those that utilize enzyme or fluorescence for signal transduction. In addition to the acceptable performance demonstrated on the miniature mass spectrometer, the bioactive paper chips can serve as a stable remote-sampling device with no need for cold storage. Moreover, the field-deployable feature of our miniature mass spectrometer (23 kg, utilizes ambient air for tandem MS, and no need for nebulizer or buffer gas during ionization and mass analysis) showcases the ability for onsite detection, not requiring laboratory settings as typically needed for bench-top mass spectrometers. Achieving better sensitivity than the suggested lowest threshold for RDT implied that our diagnostic platform has promise in detecting symptomatic malaria. Based on current sensitivity, our approach can offer opportunities for resource-limited setting in terms of (1) point-of-care testing, (2) analysis with high performance-to-cost ratio as related to the fact that the same miniature mass spectrometer can serve millions of people and be used for many other applications besides disease detection, (3) reducing power consumption and economic burdens in that miniature instrument can be turned off when not in use, unlike bench-top instruments which must run continuously to maintain performance, and (4) the stability of the paper chip, which can allow large-scale surveillance testing to be performed with ease and enable remote areas to have access to diagnosis of diseases without physical walk-ins. The method described can be adapted for other biomarkers (e.g., Plasmodium lactate dehydrogenase) to improve selectivity and species differentiation.