Introduction to ovarian cancer

Ovarian cancer is known as a fatal gynecologic cancer. It is not a distinct malignancy and consists of epithelial and non-epithelial kinds with subclasses of epithelial ovarian cancer depending upon grouping of structural and medical topographies [1]. Because of poor diagnosis, the identification of key factors for the prevention of ovarian cancer may have significant medical and public health implications. The stage of ovarian cancer during diagnosis plays a vital role that can be divided into various stages, viz., stages I, II, III, and IV. In stage I, cancer is totally confined to only ovary, while in later stages, it spreads outside the pelvis to other portions of the body through metastasis. Metastases occur either via lymphatics to nodes at the renal hilus or via blood vessels to the parenchyma of the liver or lung. Most often, lesser masses of cancer cells are shed by the ovary and embed on the surface of peritoneal, thus making several nodules. In case of ovarian cancer, contrasting to other cancers, no anatomical hindrance is present to widespread metastasis during the course of the peritoneal cavity. Tumor grafts block lymphatic vessels that pass through the diaphragm, thereby inhibiting the discharge of ascites fluid which outflows from the disordered tumor vessels in the presence of high levels of tumor-derived vascular endothelial growth factor A (VEGFA), which is known as a vascular permeability factor. Antibodies that deactivate VEGFA have lessened the buildup of ascites in animal models and in medical studies [2,3,4].

Ovarian cancer at initial stage normally exhibits very few exact symptoms. Thus, the 5-year survival level among more than 70% patients in advanced stage of diagnosis is less than 30%. If it is properly identified in stage I, ~ 90% of patients can be cured with the help of standard surgery and chemotherapy. But, only 25% are diagnosed in stage I because of the nonappearance of specific early warning signs [5]. When ovarian cancer is detected in advanced stage, the survival rate of patients is very poor, only 10–30% [6, 7].

Even though there are various precautionary methods being practiced globally, the overall ovarian cancer mortality rate is on the rise which is because of increasing obesity among populations and usage of the oral contraceptive pills. Therefore, there is an extreme necessity to identify it in early phases. There are several approaches used for its identification, mainly focusing on the presence of ascites followed by pelvic test to check the size of an ovary. Nonetheless, these techniques are expensive and laboratory-based and need time-consuming step-based analysis and skilled technicians. To address these issues, it is important to develop new and miniaturized sensing methodologies that would help the ovarian cancer–affected patients to get screened early and with ease. In this regard, the use of sensitive biosensors for recognition and continuous monitoring of ovarian cancer would be an excellent alternative.

A diversity of nanomaterials of various structures, chemical compositions with desirable surface properties, crystallographic orientations etc. had led to their widespread use in energy, biosensing applications [8,9,10,11,12,13]. The ability to tailor the structure and therefore the properties of nanomaterials provide brilliant projections towards fabrication of the novel biosensors and improve their performances. Biosensors can be categorized into various groups based on the mode of signal transduction. The objective of this review is to highlight current advancements in fabrication of different types of nanomaterial-based biosensors for ovarian cancer diagnosis. Thus, this review comprises of a brief overview of previously reported optical, electrical, electrochemical, micro-fluidic, paper-based, or flexible biosensing platforms towards determination of several ovarian cancer biomarkers.

Biomarkers for ovarian cancer

Biomarkers are the biomolecules which act as an index of a specific biological process or disease. Continuously monitoring the level of any biomarker in a particular biological media that could be serum, plasma, cellular fluid etc., often is critical in medical diagnosis. In this section, we briefly discuss about ovarian cancer–specific biomarkers.

Cancer antigen CA125 (MUC16)

CA125 (mucin 16), an important biomarker for checking the diagnosis, evolution, and management of ovarian cancer, is produced by the mesothelial cells and epithelial ovarian tumors. It is encoded by MUC16 gene and articulated as a protein of membrane bound at cells’ surface that carry out meta-plastic differentiation into a Mullerian-type epithelium or released in biological fluids in solvable form. The standard level of CA125 in human blood is typically < 35 units per milliliter (U mL−1), and beyond this level, it would cause the progression of ovarian cancer [14,15,16].

Human epididymis protein 4

Human epididymis protein 4 (HE4) is a vital biomarker for early identification of ovarian cancer and is also considered a primary sign of disease recurrence. It is firstly recognized in the epithelium of the distal epididymis and predicted to be a protease inhibitor involved in sperm maturation [17,18,19,20]. Nevertheless, the utilization of CA125 as a biomarker for initial recognition of ovarian cancer is strictly limited as its concentration is raised in only half of early-stage ovarian cancers and often in several benign gynecological diseases including ovarian cysts, endometriosis, and pelvic inflammation [20, 21]. Henceforth, quantification of HE4 with very low concentration in combination with CA125 is of prime significance and thus providing the paramount technique of differential diagnosis in ovarian cancer and other genital masses [22,23,24].

Prostasin

Prostasin is generated at higher level in epithelial ovarian cancer and should be examined more as a biomarker either alone or in combination with CA125. As per our knowledge, very few reports on nanomaterial-based biosensors for recognition of this biomarker have not been published [16].

Mucin 1

Mucin 1 (MUC1), glycoprotein encoded by the MUC1 gene is articulated with high concentrations in all epithelial cell adenocarcinomas for the breast and ovarian cancer patients. The increasing expression of MUC1 results in large amounts of free protein found in the blood [25]. Thus, determination of MUC1 is of excessive significance in cancer identification.

Conventional diagnostic methods for diagnosis of ovarian cancer

Conventional diagnostic methods like mass spectrometry [26, 27], immunoassay [28], radioimmunoassay [29], enzyme-linked immunosorbent assay (ELISA) [30, 31], and polymerase chain reaction (PCR) [32, 33] have been used for biomarker determination. But these techniques involve intricate sample pre-treatment, necessity of employing professional personnel, expensive, sophisticated instrumentation, long data analysis time, cautious washing, and separation stages. Henceforth, it is essential to fabricate cost-effective, sensitive sensors with quick response time.

Biosensors for determination of ovarian cancer

Biosensing has major scientific significance in betterment of human life in terms of disease identification and treatment [34, 35]. This has directed the scientists to emphasis on developing simple biosensors with excellent analytical performances. A biosensor is a diagnostic device that translates the information about the existence of a chemical species (analyte) to an assessable indication. It comprises 2 main constituents: (a) receptor and (b) transducer. The receptor is a biomolecule, that distinguishes the analyte, and the transducer translates this determination incident into a useful signal like current, magnetoresistance, and absorbance. Widely used biorecognition elements are enzyme, antigen (Ag), antibody (Ab), or oligonucleotide [35, 36]. Immunosensors, one of the most significant classes of biosensors, have been broadly employed to sense protein biomarkers depending on Ag–Ab interaction with extraordinary sensitivity and particularly brilliant selectivity. It can be classified further into two types: (a) labeled and (b) label-free immunosensor. In label-free immunosensor, the target molecules are directly quantified from the antigen/antibody interaction wherein the labeled immunosensors employ a label or tag as well as an additional secondary Ab, namely, Ab2, to amplify the signal in addition to primary Ab, named Ab1 [37, 38].

Optical biosensors

Fluorescence-based biosensors

Jin et al. [39] demonstrated the photoluminescence (PL)-based biosensors using the aptamer (one class of artificial oligonucleotides) and 5-fluorouracil (5-Fu)-loaded Ag2S quantum dots (QDs). The surface of QDs was decorated with polyethylenimine, followed by combination with the aptamer/5-Fu complex to make Ag2S QDs/aptamer/5-Fu composite. In amalgamation of Ag2S QDs with aptamer/5-Fu complex, near-infrared (NIR) PL of QDs was reduced, ascribed to photo-induced electron transfer from QDs to 5-Fu (Fig. 1a(i)). The adding of CA125 made a recognizable NIR PL recovery. Hence, this composite was established as NIR PL turn-on probe for CA125, exhibiting a limit of detection of 0.07 ng mL−1 over the linear range of 0.1 to 106 ng mL−1 with high sensitivity and selectivity and a good corelation coefficient (R2) of 0.9961 (Fig. 1a(ii)). The practicability of this NIR PL probe for CA125 recognition was assessed by inspecting its performance in human serum, urine, and gastric juice.

Fig. 1
figure 1

a (i) Graphical representation of the fabrication routes of NIR PL sensor; (ii) corresponding calibration plot. Reprinted with the permission from [39]. b (i) Schematic depiction of the fabrication steps of the ECL-based HE4 immunosensor; (ii) calibration plot. Reprinted with the permission from [40]

Fluorescence resonance energy transfer (FRET) depends on the energy transfer between two fluorescent molecules, i.e., donor and acceptor [41, 42]. Hamd-Ghadareh et al. [43] developed FRET immunosensing of CA125 employing the carbon dots (CDs). The CDs functionalized with aptamer were used as a recognition probe wherein polyamidoamine (PAMAM)–dendrimers/gold nanoparticle (NP) composite was utilized for covalent attachment of CA125-Ab. This immunosensor displayed a limit of detection (0.5 fg mL−1) over the linear region of 1.0 fg mL−1–1.0 ng mL−1 of CA125 with good selectivity and sensitivity. This assay is suitable for the recognition of OVCAR-3 cancer cells in the region of 2500–20,000 cells with detection limit 4 cells/10 μL along with selective imaging of the cancer cells. Wang et al. [25] demonstrated the gold NP–CD composite as a fluorometric assay towards determination of MUC1 that showed good linearity over the MUC1 concentration range of 5.3–200 ng mL−1, a high detection limit of 5.3 ng mL−1 and good selectivity. But, salt was used for recovering the fluorescence of the CDs during sensing. Ma et al. [44] presented the fluorescence assay for MUC1 based on the aggregation of CDs with a size of ~ 2 nm that showed strong blue-green intrinsic fluorescence. MUC1 Abs and the MUC1 aptamer were covalently attached with the CDs. The sensing was performed through plotting the changes in fluorescence with the MUC1 concentrations over the range of 5–100 nM with a detection limit of 2 nM. The assay was also used to determine MUC1 in serum samples. But the synthesis procedure of the CDs was time-consuming, multiple step–based, and using acid-like concentrated sulfuric acid. Yang et al. [45] used a rolling mediated cascade amplification strategy to improve the visualization of MUC1 profiling on cell surfaces where the CdTe QDs were used as fluorescent labels. Li et al. [46] demonstrated a fluorometric assay for MUC1 based on signal amplification of the hybridization chain reaction, and the interaction between a luminescent ruthenium (II) complex and CdZnTeS QDs. This sensor exhibited a very low detection limit of 0.13 ng mL−1 in MUC1 concentration range of 0.2–100 ng mL−1. The method was also applied towards determination of MUC1 in spiked serum samples.

Surface plasmon resonance

Surface plasmon resonance (SPR) sensors employ surface plasmon waves to delineate alterations during interaction between the target molecule and bioreceptor on the sensor and they are sensitive towards the refractive-index change of their surrounding media [47, 48]. Suwansa-ard et al. [49] reported a N-ethyl-N0-(3dimethyl-amino-propyl-carbodiimide (EDC)-N-hydroxysuccinimide (NHS) functionalized gold surface-based SPR sensor for determination of CA125 where anti-CA125 was immobilized on gold surface using a self-assembled monolayer (SAM). The sensor showed detection limit of 0.1 U mL−1 over a narrow linear region of 0.1–40 U mL−1 that makes them not suitable for practical applications. Zhang et al. [50] presented a SPR sensor using gold nanorods. The concentrations of CA125 were determined by examining the level of gold nanorods’ aggregation triggered by CA125 Ab–Ag interactions. This sensor was able to detect 0.4 U mL−1 of CA125 over the linear region of 1–80 U mL−1 with good reproducibility. Yuan et al. [51] used silver (Ag) nanochip-based localized SPR (LSPR) sensor towards determination of HE4 where the nanochip was incubated in 11-mercaptoundecanoic acid solution to produce SAM on its surface, followed by activation of this layer with EDC-NHS. Afterwards, anti-HE4 solution was immobilized on the electrode’s surface and the HE4 was subjected for final LSPR measurements. The quantifiable recognition of HE4 was done by observing the Δλmax when the target HE4 molecules were bound to the NPs. This sensor showed low detection limit of 4 pM, decent selectivity, reproducibility, and durability over the linear range of 10–10,000 pM. The good correlation between LSPR and ELISA results in human serum ensures its suitability for real-time applications.

Chemiluminescence-based optical biosensors

Chemiluminescence (CL) employs energy resulting from chemical reaction that yields a luminescence emission of light because of the relaxation of the atoms from excited state to ground state. The reaction between the immobilized biomolecule marked with CL species and analyte results in generating light owing to this reaction [47]. Yang et al. [52] developed the 3D silica film for CL sensing of CA125 where the sensor displayed a wide range of 0.5–400 U mL−1 with good reproducibility and durability. In applicability tests with the serum samples, the relative errors were high (~ 8%). Al-Ogaidi et al. [53] used graphene QDs for recognition of CA125 which showed low limit of detection limit (0.05 U mL−1) over the linear range of 0.1–600 U mL−1. It is a well-substitute for a FRET-based assay as it removes the requirement of an additional excitation source.

Electrochemiluminescence-based optical biosensors

Electrochemiluminescence (ECL) is defined as a chemiluminescence activated by electrochemical method. Upon ECL reaction, the emitted light is identified in present of a required voltage [8]. Wang et al. [40] presented nickel–iron oxide (NiFe2O4) NT-based electrochemiluminescent sensor for determination of HE4. A polymeric material enhanced the ECL emission method of 2 luminophors via decomposition of hydrogen peroxide (H2O2). The hexagonal boron nitride nanosheets immobilized large amount of lucigenin, thereby retaining stable ECL emission of lucigenin (Fig. 1b(i)). It exhibited low detection limit of 3.3 fg mL−1 over the linear range of 10 fg mL−1–10 ng mL−1 of HE4 as illustrated in Fig. 1b(ii). Acceptably, the relative standard deviation (RSD) of DBAE and lucigenin were 0.18% and 0.48%, respectively, indicating its good stability. Babamiri et al. [54] used the PAMAM-QDs and PAMAM-sulfanilic acid-Ru(bpy)32+ for CA125 sensing where the CdTe/CdS QDs and Ru(bpy)32+ in presence of tripropylamine were acted as co-reactant for generation of ECL and Fe3O4–SiO2 was employed as a magnetic bead. This sensor showed low limit of detection (0.1 μU mL−1) over the linear region of 1 μU mL−1–1 U−1 of CA125. The results obtained with this immunosensor agree well with the results of ELISA as reference method, thus confirming its suitability towards practical applications. Tan et al. [55] described an ECL immunoassay for CA125 determination utilizing CdSe nanocrystals. This sensor displayed the limit of detection of 5 × 10−5 U mL−1 over the linear region of 10−4–1 U mL−1. Wu et al. [56] demonstrated an ECL sensor constructed on graphitic carbon nitride (g-C3N4)–based electrode for CA125 recognition. This sensor revealed a widespread linear range for CA125 (0.001–5 U mL−1) and limit of detection (0.4 mU mL−1). But its narrow detection range might not be suitable for the practical applications.

A summary of the reported methods, the importance of nanomaterials, and the recognition procedures utilized in optical sensing of ovarian cancer biomarkers is summarized in Table 1.

Table 1 List of nanomaterial-based optical biosensors towards determination of ovarian cancer biomarkers

Electrochemical biosensors

The electrochemical determination techniques have been found promising in biosensing applications because of various benefits including high sensitivity, specificity, low cost, low detection limit, good reliability, and easiness in handling [37, 57,58,59,60,61,62].

Raghav et al. [63] developed gold–Ag NP–based label-free detection of CA125. Direct immobilization of Ab on the surface of the electrode exhibited good linear response over 1–150 IU mL−1 (R2 = 0.994) and tolerable interference. This immunosensor showed sensitivity of 190 Ω IU−1 mL cm−2 which was because of high specific surface area for their core–shell structure. This immunosensor exhibited acceptable interference of 2–5% from serum components and retained 90% stability up to 20 days, thus displaying its suitability towards practical applications. Wang et al. [64] reported the TbFe-based combined metal−organic frameworks (MOF) to detect CA125 that showed a limit of detection of 58 μU mL−1 over a linear region of 100 μU mL−1–200 U mL−1 with excellent selectivity and stability, satisfactory reproducibility, and good applicability in serum. Johari-Ahar et al. [65] demonstrated an immunosensor for CA125 sensing where gold electrode was modified with mercaptopropionic acid and successively then conjugated with silica (SiO2)-coated gold NPs, CdSe QDs, and CA125 monoclonal antibody (mAb) (Fig. S1). The nanobiosensor displayed high stability and reproducibility and an ultra-low limit of detection (0.0016 U mL−1) over the region of 0–0.1 U mL−1. The applicability of the immunosensor was inspected by assessing CA125 concentration in the serum collected from the ovarian cancer patients.

Torati et al. [66] presented gold-based label free immunosensor towards CA125 recognition. The electrode was amine functionalized using cysteamine hydrochloride followed by activation with EDC-NHS. This immunosensor displayed linearity over narrow region ranging from 10 to 100 U mL−1 with high limit of detection of 5.5 U mL−1, thereby making it unsuitable for real-time applications. Zheng et al. [67] used the label-free Prussian Blue-Platinum (Pt)-NP-polyaniline composite–based bio-electrode for CA125 sensing which exhibited good linearity over the concentration range of 0.01–5000 U mL−1 and detection limit of 4.4 mU mL−1 with a sensitivity of 119.76 μA·(U/mL)−1 cm−2. Gasparotto et al. [68] fabricated the gold NP-zinc-oxide (ZnO)-based nanohybrid for label-free CA125 determination and displayed a limit of detection (2.5 ng μL−1) with high reproducibility, specificity, and notable durability. Gazze et al. [69] established a graphene–polyaniline-based biosensor for label-free recognition of CA125, and this electrode was then functionalized with anti-CA125 Ab by covalent cross-linking to polyaniline. This was able to detect CA125 as low as concentration of 0.923 ng μL−1 over the linear range of 0.92 pg μL−1–15.20 ng μL−1. Ravalli et al. [70] used the gold NP–modified electrode for label-free recognition of CA125 where the immunoassay was founded on SAM of gold NP–based electrodes followed by immobilization of monoclonal Ab anti-CA125. This sensing layer was also activated with EDC-NHS and able to detect CA125 concentrations in the range of 0–100 U mL−1 with high detection limit of 6.7 U mL−1. Jafari et al. [71] presented the Ag NP–graphene quantum dot (GQD) composite-modified label-free electrode which was able to detect 0.01 U mL−1 of CA125. Ren et al. [72] reported the ferrocenecarboxylic acid (FA), HCl-doped polyaniline, and chitosan hydrochloride composite as a substrate material for producing signal and enlarging specific surface area, and Ag–Co3O4 nanosheets were employed in this immunosensor for intensifying the antibody capacity. This biosensor showed detection limit of 0.25 pg mL−1 over the concentration range of 0.001–25 ng mL−1 with acceptable stability and reproducibility. Its good recovery percentage in the serum samples ensures suitability of this sensor in the clinical diagnosis.

Pakchin et al. [73] reported the chitosan–gold NP–CNT–graphene oxide (GO) composite-based bio-electrode and lactate oxidase (Lox) as a label towards CA125 sensing. When a target Ag is linked with an associated Ab, a secondary Ab is used as label to generate the indication. To attain an extremely sensitive sandwich kind of immunosensor, a greater number of Abs on the electrode’s surface and label is essential. To increase the functional groups, chitosan, a highly abundant biopolymer with inherent functional groups on its surface, was utilized. But, it suffers from poor conductivity. To enhance direct electron transfer rate, the gold NP–CNT–GO composite was used for sensing along with chitosan. Figure S2 illustrates the graphical representation of fabrication stages of the immunosensor. The sensing performance was studied in terms of the H2O2 oxidation that presented low limit of detection (0.002 U mL−1) with excellent reproducibility, selectivity, and durability towards CA125 determination. The recovery results and its comparison of this immunosensor by the ELISA method confirmed its applicability for the real biological sample analysis with good accuracy and reliability. Taleat et al. [74] used a sandwich format based on poly-anthranilic acid-based electrode. This sandwich assay was then achieved through addition of anti-CA125 Ab2 labeled with gold NPs. The antibody-gold NPs captured onto the bioelectrode surface induced the deposition of Ag from its precursor solution. With the increasing concentration of CA125, more numbers of gold NPs were captured on the surface of the immunoelectrode to generate more numbers of Ag NPs. Over a narrow region of 5–25 U mL−1, the biosensor was able to detect 2 U mL−1 of CA125 with satisfactory sensitivity, selectivity, and reproducibility. Liang et al. [75] developed a sandwich-typed sensor based on pH reactive sensitivity amplifiable method for CA125 sensing. Firstly, the pH responsive polydopamine framework was prepared using zeolitic imidazolate framework as a template. This polydopamine–methylene blue composite was then covalently attached with Ab2 and utilized as a label. The gold-reduced graphene oxide (rGO) composite was functioned as substrate to capture target molecules, HCl was used for the annihilation of label and disassociation of Ag–Ab interaction layer, thus decreasing the impedance and enhancing the electron transfer efficacy. This immunosensor revealed a low limit of detection (0.336 μU mL−1) with acceptable reproducibility, sensitivity, and selectivity over the linear region of 0.0001–100 U mL−1. Wu et al. [76] reported the gold NPs and horseradish peroxidase (HRP) was used as a label. The current was decreased with the increasing concentrations of CA125 ranging from 0 to 30 U mL−1 with high limit of detection (1.73 U mL−1) with acceptable stability. A reduction of 10.3% in amperometric response in human serum after a 7-day storage confirms its incompatibility for routine clinical diagnosis.

Wu et al. [77] developed a carbon nanofiber for binding of CA125 and thionine was used as electron transfer mediator. The carbon nanofiber was activated and functionalized with EDC-NHS and HRP was employed as a label (Fig. S3). The labeled conjugate revealed a decent enzymatic activity towards electro-oxidation of thionine by H2O2. This sensor displayed satisfactory sensitivity, stability, reproducibility, and limit of detection of 1.8 U mL−1. Lu et al. [78] reported a labeled, gold NP–chitosan–titanium carbide (TiC)–based immunosensor for determination of HE4. The Ab1 namely, anti-HE4, was immobilized on the electrode surface. After immunoreaction, HE4, Ab2, and primer DNA were immobilized on the Ab2. This sensor exhibited low limit of detection (0.06 pM) over a linear concentration range of 3–300 pM of HE4 with good precision and regenerative ability.

Paimard et al. [79] presented an impedimetric assay for MUC1 using the core–shell nanofiber-multi-walled carbon nanotube (MWCNTs)–gold NP composite which was covalently modified with the MUC1-binding aptamer. MUC1 was sensed through the change of the resistance of the electrode surface. This sensor exhibited a high detection limit (2.7 nM) and good stability and selectivity over the narrow region of 5–115 nM of MUC1. The assay was successfully applied towards MUC1 determination in spiked serum samples with satisfactory recovery. Guo et al. [80] demonstrated the Ag nanocluster–based electrode for MUC1 sensing. This sensor exhibited a low detection limit of 0.5 nM over the wide MUC-1 concentration ranging from 1 to 500 nM. But they used the template (C12) during preparation of Ag nanoclusters.

Molecular imprinting technique

Molecular imprinting is a technique to generate precise active sites with high affinity towards target molecule in polymers via molecular template. Molecularly imprinted polymers (MIPs), also known as cross-linked polymers, are tailor-made materials. Three-dimensional cavities are produced inside polymer corresponding to the shape and size of the target molecule. The imprinted polymer cavities permit the target molecules to reside in the cavity space. Target molecules can also be utilized as a template for imprinting polymers. MIPs can be made easily and are stable, and unaffected to an extensive range of pH, temperature, and solvents. Deposition of MIPs on the surface of nanostructure-modified electrodes increases sensitivity for biomolecule sensing [81, 82].

Viswanathan et al. [83] presented the protein-imprinted polymer on 3D gold nanoelectrode for CA125 quantification wherein CA125 was firstly used as a template and upon its removal imprints for CA125 (Fig. 2a). The protein-imprinted sites on nanoelectrode are then employed for immunospecific internment of CA125 molecules, and the mass of bound on the electrode surface was noticed as a drop in the faradic current from the redox indicator. The sensor displayed decent linearity over the concentrations ranging from 0.5 to 400 U mL−1 with high limit of detection (0.5 U mL−1) (Fig. 2b). Replicated 5 independent analyses were directed to RSD of 3.1% only. In human serum samples, the recovery outcomes and its comparison of this immunosensor by the ELISA method confirmed its suitability for the real biological sample analysis.

Fig. 2
figure 2

a Schematic depiction of MI protein sensor. Step1: adsorption of CA125 on the electrodes’ surface. Step 2: electro-polymerization of phenol. Step 3: removal of template protein. Step 4: CA125 binding and signal generation. b Corresponding calibration plot. Inset of (b) is as follows: DPV results to CA 125. Reprinted with the permission from [83]

Büyüktiryaki et al. [84] testified the phosphoserine (PS)-imprinted CNT–Fe2O3 NP-polymer-based nanosensor for CA125 quantification where methacryloyl antipyrine europium(III) [(MAAP)2–Eu(III)] and methacryloyl antipyrine terbium (III) [(MAAP)2–Tb(III)] were utilized as monomers and PS was employed as template. This MIP sensor was able to detect 0.49 U mL−1 of CA125 with good sensitivity and reproducibility.

Paper-based biosensors

Paper-based biosensors have attracted significant interest because of their several benefits like inexpensiveness, disposability, biocompatibility, and abundance. Various approaches like UV photolithography, wax printing, and screen printing have been developed for the fabrication of paper-based biosensors [85,86,87].

Fan et al. [88] demonstrated the rGO–thionine–gold NP–modified paper-based immunosensor towards CA125 determination (Fig. S4). CA125 was detected through the reduction in the values of current upon binding of CA125 Ab and Ag. The paper-based immunosensor exhibited low limit of detection (0.01 U mL−1) over the linear region of 0.1–200 U mL−1 with great accuracy, reliability, reproducibility, and stability. In serum samples, the relative error was high nearly ~ 8.05%. Ge et al. [89] developed an array of addressable, chitosan–MWCNT composite-modified paper electrodes which were assembled on the crossing points of the row/column electrodes using a device holder used for quantification of CA125. One paper layer comprised of the sensing sites, the other paper layers for the printed counter electrode, and reference electrode. This sensor displayed low limit of detection (3.7 × 10−5 U mL−1) with outstanding linearity, durability, and reproducibility over the linear range of 1 × 10−4–100 U mL−1.

Bahavarnia et al. [90] fabricated a paper-based biosensor through hand writing of Ag/rGO nanoink on paper towards recognition of CA125 in human biofluid. Paper electrode was modified by cysteamine (CysA)-caped gold NPs and from the amine group with hydroxyl and carboxyl groups of Ag/rGO nanoink deposited on the surface of paper-based electrode. Then, anti-CA125 was immobilized on the electrode surface via gold NPs and CA125 interaction. The sensor displayed low limit of detection (0.78 U mL−1) over the region of 0.78–400 U mL−1 with poor stability.

Microfluidic-based lab-on-chip devices

Microfluidic-based lab-on-chip biosensor is one of the powerful miniaturized devices that eases the incorporation of manifold functionalities of one or more sensing platforms on a single system, consuming less sample volume. Microfluidic devices facilitate effective recognition of biomolecules at the expense of small volume of reagents and energy, thus paving a new avenue towards development of the point-of-care (POC) devices [91].

Mandal et al. [92] developed a microfluidic, capacitive, carboxyl CNT–based biosensor for determination of CA125 which was successfully incorporated with polydimethylsiloxane (PDMS) microfluidic channels. Only 5 μL of CA125 Ag solution was employed on the channel inlet. The Ag solution was then passed through this channel because of capillary effect and interacted with Ab that was exposed on top of the sensing device (Fig. 3a, b). The experiments were performed in two cases, (a) only phosphate-buffered saline (PBS) was passed through the microchannel and corresponding capacitive curve was considered “baseline” and (b) the capacitive measurements were executed while CA125 Ag solution was passed through. The Ag–Ab interaction yielded the change in the medium’s dielectric properties over the sensor surface. Ag–Ab conjugation capacitance of the carboxylic CNT-based microfluidic base device was higher than non-carboxylic CNT- and non-CNT-based biosensors (Fig. 3c) as unavailability of covalent attachment of the CNTs between CA125 Ab and SAM caused a loss of both CNTs and Ab underneath the shear flow state, thus providing less number of Ab present for the Ag–Ab conjugation and lowering the net capacitance. As the CNTs and Ab have higher dielectric permittivity, this microfluidic sensor exhibited remarkably greater capacitance than the non-CNT-based sensor.

Fig. 3
figure 3

a, b Schematic diagram of microfluidic, carboxyl CNT–based sensor for CA125 recognition. c Comparison in the capacitive performances of the carboxyl CNTs and non-carboxyl CNTs and non-CNT-based biosensor. Reprinted with the permission from [92]

Wang et al. [93] used a labeled, 3D microfluidic immunodevice for the determination of CA125. Molybdenum disulfide (MoS2) was acted as support for CA125 (Ab1) wherein the gold nanoflowers were used as the supporter of glucose, glucose oxidase (GOx), and CA125 secondary Ab (Ab2). The microfluidic system was made up using screen-printed electrodes on wax-patterned cellulose paper. GOx catalyzed glucose oxidation while MoS2 enhanced the reduction of H2O2. The immunosensor revealed great sensitivity, specificity, and low detection limit of 0.36 pg mL−1 with excellent linearity over the CA125 concentration range of 0.001–50 ng mL−1. This work paved a unique route towards the fabrication of POC devices founded on microfluidic paper-based analytical devices (μ-PADs). Wu et al. [94] presented the combination of polymerization-aided signal amplification with μ-PAD wherein the μ-PAD was fabricated using the photoresist-patterning as well as screen-printing methods. Glycidyl methacrylate was polymerized to offer epoxy groups for immobilization of label, HRP on the surface of GO–chitosan-modified paper electrode. CA125 was detected using the HRP-O-phenylenediamine-H2O2 electrochemical system which showed 0.05 ng mL−1 with good reproducibility, stability, and regeneration. The practicability was not shown. Nunna et al. [95] reported the gold NP–modified microfluidic biosensor for CA125 determination. The capacitive response of Ag–Ab conjugation on inter-digitated electrodes was boosted up by ~ 2.8-fold in static drop state and ~ 2.5-fold in microfluidic flow state with gold NP modification. Zhao et al. [96] demonstrated a microfluidic ExoSearch chip for multiplexed exosome recognition towards CA125 sensing that offers a continuous-flow strategy for quantifiable segregation and liberation of blood plasma exosomes in large range of preparation volumes (10 μL–10 mL) with good reproducibility (coefficient of variation, CV < 10%).

Chemiresistive or field effect transistor–based biosensors

Recognition of various target biomolecules using chemiresistive or field effect transistor (FET)–based sensors have gained wide attention among research communities as they involve a nominal or no sample preparation and make suitable for on-site monitoring because of quicker transduction mechanism. The transportability and entire analytical performances of the sensors can be enhanced by using nanostructures as attractive channel materials. Owing to small cross-sectional area, the nanostructures display no horizontal current shunting, thereby exhibiting excellent response.

Majd et al. [97] developed the flexible FET-type aptasensor for CA125 determination based on carboxylated MWCNT–rGO composite wherein poly(methyl methacrylate) (PMMA) was utilized as a substrate. RF sputtering method was used to deposit the source/drain gold electrodes while Pt was employed as gate electrode. A solution chamber containing 20 μL was employed for the measurements. The carboxylated MWCNTs were activated with EDC-NHS followed by immobilization of CA125 single strand DNA (ssDNA). The rGO surface was then modified with incubation of MWCNT-CA125 aptamer through π–π interaction. In the I–V curves (Fig. 4a), after deposition of MWCNTs modified with CA125 ssDNA on rGO channel, dI/dV increased significantly as MWCNTs with great specific surface area reduced the electrode resistivity whereas dI/dV decreased upon CA125 incubation. Figure 4 b illustrates the flexible liquid-ion-gated FET-based nanobiosensor. Under the gate voltages (Vg) ranging from − 1 to + 1 V at a fixed drain voltage (Vd = 0.2 V) in rGO FET (Fig. 4c), the drain current (Id) increased with the increasing Vg, signifying n-type characteristics. Figure 4 d displayed the transfer characteristics of the rGO FET nanosensor before and after incubation of 1.0 × 10−5 U mL−1 CA125. The current values of the rGO/MWCNTs/CA125 ssDNA aptasensor decreased because of interaction of CA125 ssDNA-wrapped MWCNTs with graphene and phosphate groups of the ssDNA are negatively charged gates, thus decreasing the current of the n-type channel. This aptasensor showed a good linear range for CA125 (1.0 × 10−9–1.0 U mL−1), a low limit of detection (5.0 × 10−10 U mL−1) with selectivity, stability, and flexibility as exhibited in Fig. 4e. The reproducibility of this FET biosensor was also evaluated through conducting experiments for 1 × 10−5 U/mL of CA125, and it revealed the response with RSD 3%. Good agreement between the ELISA method and this aptasensor ensures its applicability towards the determination of CA125 in real human serum samples.

Fig. 4
figure 4

a I–V responses of with and without MWCNTs modified with CA125 ssDNA and after CA125 incubation. b The flexible FET-based nanobiosensor. c The Id–Vg result of the rGO FET aptasensor at Vd = 0.2 V. d Responses of the rGO FET nanosensor with and without MWCNTs modified with CA125 ssDNA and after CA125 incubation. e Calibration plot. Reprinted with the permission from [97]

Bangar et al. [98] demonstrated the polypyrrole (Ppy)-based chemiresistive immunosensor for CA125 determination. Glutaraldehyde and EDC were covalently functionalized on the surface of single Ppy nanowire with CA125 Ab to fabricate a nanoimmunosensor for CA125 recognition. This sensor exhibited good sensitivity with relatively high limit of detection (1 U mL−1) of CA125 over the linear region until 1000 U mL−1.

Wang et al. [99] reported the field-effect enzymatic detection (FEED)–type HRP-labeled CA125 sensor using glutaraldehyde functionalized CNTs. When the conventional FET-based biosensor uses the FET to offer a reference current, this FEED method provides a voltage-controlled current magnification, generated using quantum tunneling of electrons, depending upon the distance of electrons traveled via tunneling and the potential energy profile. This sensor was able to detect 0.9 U mL−1 of CA125 which was high. Its applicability in real serum samples was not checked.

Giant magnetoresistive biosensor

Giant magnetoresistive (GMR) technique has been developed for biomolecular recognition based on determination of magnetically labeled targets as it exhibits several advantages like good sensitivity, compatibility with the integrated circuits (ICs), and potentiality for quantifying concentrations of manifold biomarkers all at once(viz., multiplex) in a transportable device [100]. Multiplexing is easy in GMR biosensing since the magnetic dipole fields remain confined to a small area near each sensor’s surface and thus avoiding interference with the neighboring sensors.

Klein et al. [101] reported a multiplexed GMR biosensor array prototype to quantify CA125, HE4 wherein magneto-resistance signals were examined by a nearly balanced Wheatstone bridge circuit. Every single sensor was printed with the preferred antibodies that act as binding sites for the biomarkers in the multiplex assess towards concurrent measurements. The sensor displayed ultra-low limit of detection (3.7 U/mL, 7.4 pg/mL) towards sensing of CA125 and HE4, respectively.

A summary of the reported electrochemical and other methods employed in biosensing of ovarian cancer biomarkers is summarized in Table 2.

Table 2 List of nanomaterials based biosensors towards determination of ovarian cancer biomarkers

Conclusion and future perspectives

Timely recognition and prevention of ovarian cancer are the ultimate principles to achieve success in the entire cancer treatment process. Firstly, so far, a wide variety of biosensors have been reported in the literature towards determination of only CA125. Very few reports are available on recognition of other important biomarkers like HE4, prostasin. Therefore, substantial research needs to be done towards fabrication of biosensors for determination of HE4, prostasin. Secondly, at present, the ELISA is the only existing commercial recognition kit for ovarian cancer biomarker sensing that exhibits narrow detection range and low sensitivity. To eliminate these issues, the nanomaterial-based biosensors are advantageous as compared with the ELISA as they have the ability to be attached with aptasensors, thereby showing better sensitivity and selectivity. Thirdly, even if the optical biosensors resulted in favorable analytical consequences, most of them are involved in intricate, time-consuming arrangement which frequently failed to deliver a straightforward determination way of biomarkers. To tackle this, magnetoresistive, electrical, or electrochemical sensors exhibit enormous potential with their several benefits. Finally, still very few reports are presented on portable sensors to be employed outside the laboratory for recognition of ovarian cancer biomarkers. To commercialize the biosensors, incorporation of this biosensing platform into device level is essential that is still in early stages. To address it, current research towards fabrication of economical paper-based disposable biosensors or flexible FET type, GMR, or microfluidic-based lab-on-chip devices for determination of ovarian cancer biomarkers displays huge prospective.

In this regard, most of the recent pertinent articles were reviewed and corresponding analytical performances with sensing mechanisms for different biosensors were assessed. To conclude, exciting progresses are anticipated to take place in the near future for scaled-down sensing methodologies and testing kits that would pave a new avenue for ovarian cancer–affected patients to check their own health at ease.