Glucocorticosteroids are widely recognized as markers for adrenal activity. Cortisol can reflect the short-term changes in the activity of the hypothalamic–pituitary–adrenocortical axis (HPA axis), making it a valuable surrogate marker for stress and glucose metabolism [56].

Many methods have been developed to detect unconjugated cortisol levels in various biological samples. However, the use of certain samples, such as urine, may not be recommended as these samples may contain many interfering substances thus complicating steroids extraction procedure [55]. In addition, the concentrations of corticosteroids in urine are low compared with those found in plasma, which makes the latter a better medium for measurement. Steroids have been measured by HPLC-FLD methods for various clinical purposes, such as investigating disease courses, pathogenesis, and metabolic processes [15].

The detection of corticosteroids in biological samples can be done directly depending on fluorescence signal enhancement upon the treatment of samples with the deproteinizing agent (ethanol) and sulfuric acid (Figure 1). A study to quantify corticosteroids in serum samples was performed using C₁₈ analytical column after treatment with ethanol and sulfuric acid [53]. Cortisol, testosterone and corticosterone emitted fluorescence, with LOD for cortisol = 0.3 pg/dL (S/N = 3). In another study, the same group measured cortisol in urine samples producing a LOD = 0.26 pg/dL (S/N = 3) [54]. Sudo et al. used post-column derivatization with sulfuric acid to measure cortisol and corticosterone in rat urine (LOD was 0.5 pmol for corticosterone at (S/N = 5)) [55]. They analyzed other corticoids, such as prednisolone, 6α-methylprednisolone, dexamethasone, and betamethasone, but with lower sensitivity. Moreover, ethyl acetate was used for extraction after sulfuric acid hydrolysis to prevent the acid from entering the HPLC system. It was noticed that the fluorescence intensities were dependent on the reactor temperature and sulfuric acid [55,77].

Since cortisol levels in hair might be used as a biomarker of chronic stress, a study developed a HPLC-FLD method to measure cortisol distribution in human hair samples. The method has achieved equal precision to mass spectrometry. Samples were prepared by pulverization and incubation in 0.1 M HCl, and extraction with ethyl acetate followed by derivatization with sulfuric acid. A detection limit of 1 pg/mg was achieved [56].

Derivatization with fluorophore-containing reagents was employed in many studies to detect and quantify corticosteroids in serum and urine samples. The derivatization process depends on functionalization of specific groups in corticosteroids structure (Figure 1). The primary alcohol functionality can be esterified through its reaction with either 2-(4-carboxyphenyl)-5,6-dimethylbenzimidazole (CDB) (Figure 1), 9-anthroyl nitrile (9-AN) (Figure 2), or 1-anthroyl nitrile (1-AN) (Figure 2). These derivatization processes proved to be selective as neither secondary alcohol, nor tertiary alcohol can react with fluorophore-containing reagents (Figure 2).

The derivatization through esterification of corticosteroid primary alcohol functionality with CDB was studied by Katayama et al. [78]. The results showed that the detector response to secondary alcohols was less than one fiftieth that of the primary alcohol. However, certain secondary alcohols in steroids, such as prednisolone acetate and testosterone, and tertiary alcohols, showed no reaction. Later, Katayama et al. used human plasma to detect eight corticosteroids by derivatization with CDB to their esters in acetonitrile. The esters were separated on a reversed-phase column (Zorbax ODS) with water:methanol (25:75, v/v) containing 5 mmol/L tetramethylammonium hydrogen sulfate as a mobile phase. The LOD ranged between 0.06 and 0.3 pg per 100 µL of plasma (S/N = 3) [58].

The use of 9-AN as a derivatizing agent for corticosteroids was reported in several studies. Neufled et al. used 9-AN to derivatize the primary alcohol in C₂₁ corticosteroids. The reaction was carried out at 45 °C for 2 h and the fluorescent derivatives were separated on silica stationary phase with a mixture of 2-propanol and hexane as a mobile phase in the gradient mode. The low temperature in derivatization prevented the thermal degradation of corticosteroids and avoided reaction with secondary hydroxyl groups [46,50].

Measuring endogenous glucocorticoids and their metabolites is important as disturbances in the enzyme responsible for their metabolism can cause hypertension. One study described an HPLC-FLD method using 9-AN derivatization for the determination of cortisol, cortisone, and their tetrahydro- and allo-tetrahydro-metabolites in plasma and urine [45]. Following extraction with dichloromethane and SPE, 9-AN was used to derivatize the steroids in the samples to their fluorescent products. LOD (S/N = 3:1) achieved with this method was 3.0 ng/mL for all analytes. Similarly, Shibata et al. employed 9-AN for fluorescent derivatization and developed a method to investigate cortisone levels in plasma and urine samples of renal transplant patients who received prednisolone [47]. Cortisol, cortisone, prednisolone, prednisone, 6β-hydroxycortisol, and 6β-hydroxyprednisolone were derivatized to their fluorescent esters after extraction with ethyl acetate (Figure 2). The 6β-hydroxycortisone was used as an internal standard. LODs achieved for cortisol, cortisone, prednisolone and prednisone in plasma or urine were 0.1 ng/mL, while those for 6β-hydroxycortisol and 6β-hydroxyprednisolone in plasma or urine were 0.5 ng/mL.

Similar to 9-AN, fluorescent derivatization can be achieved using 1-AN. A study described the use of 1-AN to derivatize 18-oxygenated corticosteroids: 18-hydroxycortisol, 18-hydroxycortisone and 18-oxocortisol in human urine into their fluorescent 21-anthroyl esters [43] (Figure 3). A mixture of ether and dichloromethane was used to extract the steroids, and the anthroyl derivatives were enriched by SPE using a CN cartridge column. The LOD was 0.1 pmol (S/N = 5).

Fluorescent derivatization of corticosteroids can also be carried out using dansyl hydrazine targeting the carbonyl groups (Figure 7). In a study to determine corticosteroids in human plasma and urine samples were derivatized by dansyl hydrazine and quantified by HPLC-FLD, following extraction with methylene chloride. The linearity range of the method was between 0.5 and 60 ng of cortisol, proving it to be suitable for the routine analysis of cortisol in plasma and urine [71].

The detection of corticosteroids with ketolic groups in urine is key part in diagnostic procedure of Cushing’s syndrome. Ketol-containing corticosteroids, such as 17-hydroxycorticosteroid (breakdown product of cortisol), are usually excreted in urine as tetrahydro form making their detection using UV absorption at 240 nm not possible. To detect these steroids in urine, derivatization with amidine-containing compounds was used yielding fluorescent compounds [60] (Figure 5). Using this approach, the best sensitivity was achieved for cortisol.

The utilization of ketolic groups was also performed by Yamaguchi et al. to detect and quantify corticosteroids in urine after their conversion into fluorescent quinoxalines (Figure 6) [67]. They detected nineteen 21-hydroxycorticosteroids in human urine samples achieving LOD = 0.14–29.4 pmol/50 µL injection volume (S/N = 3). In another study by the same group, prednisolone and prednisone were quantified in plasma samples following similar derivatization procedure and analyzed by reversed-phase liquid chromatography with isocratic elution. The LOD of prednisolone and prednisone was 3 ng/mL in plasma (S/N = 3) [65].

Cholesterol quantification is commonly required in the analysis of biosamples. A study described a simple and sensitive method for the determination of cholesterol and phytosterols, such as β-sitosterol in biosamples (e.g., saliva and urine matrices) and food samples (cow and soybean milk) after derivatization with naproxen acyl chloride in toluene (Figure 7). The method employed a C₈ column with a mixture of methanol, isopropanol, and water, achieving a LOD of about 25 nM (S/N = 3). The analysis of cholesterol and sitosterol is usually time-consuming, but using the method, a relatively short period of time was required since solvent concentration, evaporation, and replacement steps were not necessary [79].

Free 7α-hydroxycholesterol (7-HC) levels in human serum were reported to be a good indicator for bile acid synthesis. An HPLC-FLD method was developed by Saisho et al. to quantify 7-HC levels in dog plasma, with the purpose of studying the effect of cholestyramine on plasma levels of 7-HC. 7-HC was converted to its fluorescence derivative, by two procedures, after being extracted and then purified [57]. The two derivatization reagents used were 1-AN and 7-methoxycoumarin-3-carbonyl azide (MC-CON₃). The MC-CON₃ derivatization resulted in higher fluorescence intensity compared to 1-AN route. 1-AN produced only a C-3 fluorescent derivative due to the bulky anthracene group and steric hindrance around the C-7 position. In comparison, MC-CON₃ yielded a double coumarin derivative at the C-3 and C-7 positions (Figure 8) producing a LOD of 4 pg (S/N = 5).

Detection of steroid hormones and their metabolites are important in diagnosis of metabolic diseases. Several HPLC-FLD methods were described for the detection of steroid hormones. A study developed a method for monitoring progesterone and 17-hydroxyprogesterone in the serum from pregnant women. The quantification employed fluorescent derivatization using 4,4-difluoro-5,7-dimethyl-4-bora-3a,4adiaza-s-indacene-3-propionohydrazide (BODIPY^TM FL hydrazide) [71] (Figure 9). The derivatization was carried out in ethanol at room temperature (about 22 °C) for 15 h. This derivatization method was reported to be 50 times faster than that using dansyl hydrazine. The LODs for progesterone, 17-hydroxyprogesterone, dehydroepiandrosterone, androstenedione, testosterone and 17-methyltestosterone were in the range of 550–3700 fmol per 10 µL injection (S/N = 5).

The derivatization reagent 1-AN was used by Shimada et al. to determine the endogenous steroid pregnenolone in Wistar and Sprague–Dawley rat brain samples [46]. The samples were homogenized in isotonic saline then deproteinized with methanol, before subjected to sequential steps of extraction and derivatization. The samples were derivatized with 1-AN (Figure 10) and the excess reagent was removed by purification, carried out on two successive silica gel columns. The 3β-hydroxy-16-methylpregna-5,16-dien-20-one was used as an internal standard.

Several steroids in animal tissue were detected in another study using ultraviolet, fluorometric, and electrochemical detectors [80]. The study demonstrated the selectivity of FLD in determination of estradiol by exploiting its fluorescence emission, although it was eluted with the same fraction containing other steroids and their respective metabolites. Extraction with methanol allowed the separation of acidic corticosteroids (diethylstilbestrol, estradiol, zeranol/zearalenone, and their metabolites) from their neutral anabolic counterparts (testosterone, trenbolone and progesterone).

Urine is the best source for the estimation of estrogen concentrations, since they are primarily excreted renally as glucuronides and sulfates. Previous studies used urine samples for the determination of estrogen levels in their studies, albeit with different extraction methods [42,81]. Mao et al. used chemical hydrolysis in methanol and concentrated hydrochloric acid at 70 °C for 1 h to release the conjugation before SPE sample preparation, whereas Kumar et al. used a fabric phase sorptive extraction (FBSE) procedure, which offered shorter sample preparation times and 400 times higher sorbent loading. This method achieved a lower detection limit and analysis time, proving to be greener and more economical. The LODs for β-estradiol, 17α-ethinylestradiol, and bisphenol A were 20 pg/mL, 36 pg/mL, and 42 pg/mL, respectively.

In another study, Mao et al. used p-nitrobenzoyl chloride at 25 °C as a derivatization reagent, which can easily react with the hydroxyl and phenolic hydroxyl groups of organic chemicals (Figure 11) without the need for a catalyst [42]. Moreover, 4-nonylphenol, bisphenol A (BPA), 17α-ethinylestradiol, and three endogenic estrogens, including 17α-estradiol, 17β-estradiol, and estriol, were determined in urine samples collected from 20 healthy volunteers. Samples were hydrolyzed with HCl and subjected to SPE purification. Separation was performed on a C₁₈ column with gradient elution using acetonitrile and water as a mobile phase. The LODs of the method were 2.7 µg/L for BPA and 17β-estradiol, 2.9 µg/L for 4-nonylphenol, 4.6 µg/L for 17α-estradiol and 17α-ethinylestradiol, and 8.3 µg/L for estriol.

A recent study examined the determination of estrone (E1), 17β-estradiol (E2), and estriol (E3) and their conjugated metabolites levels in cow and river buffalo meat [32]. Samples were extracted with methanol, enzymatically deconjugated and purified by C₁₈ SPE before they are analyzed by HPLC-FLD. The effect of temperature on the studied steroid concentrations was also investigated and it was shown that heating processes was not able to significantly affect the level of phenolic estrogens in meat.

Estrogens in human urine samples were determined by HPLC-FLD following vortex-assisted dispersive liquid–liquid microextraction (VA-DLLME). Fine droplets of nonanoic acid floating on the top of sample solution were used to extract the estrogens using vortex-mix to assist the dispersion of the extraction solvent into the aqueous sample. Derivatization was not required in this work as the high extraction efficiency of DLLME improved the sensitivity and shortened the analysis time. LOD values were 0.01 ng/ml for E3, 0.01 ng/mL for βE2, and 0.06 ng/mL for E1, respectively [82].

The abuse of anabolic steroids has gained worldwide concerns, as frequent high doses can irreversibly affect the endocrine system, mineral metabolism, and may result in hepatic carcinomas [26]. Therefore, several committees have prohibited or imposed strict rules on their use for illegal purposes, such as increasing performance activity for athletes and stimulating meat production in cattle and poultry. Amin et al. used micellar chromatography with the detergent solution as the mobile phase to detect testosterone and bolasterone in human urine samples [26]. The method made use of fluorescence by means of energy transfer from the aromatic carbonyls in the anabolic steroids to the terbium ion in micellar media. No sample preparation was required and urine samples were injected directly onto the HPLC column. Excitation of terbium by means of energy transfer from steroids resulted in 183-fold fluorescence enhancement. The detection limits were 10 ng/200 µL injection volume for testosterone and 2 ng/200 µL for bolasterone.

The analysis and detection of steroid blood levels is vital for the investigation of food safety and effect on health [42,83]. Many studies have reported the presence of synthetic steroids in our daily food, both from plant and animal sources [61,84,85,86]. BPA, an industrial chemical, was found in biological fluids because of its ability to leach into food or liquids or through dental sealants into patient’s saliva [32,42,61,62,63]. One of the reasons behind their existence is that farmers increase their profit by using endocrine disruptive chemicals (EDCs) to support the feed conversion and growth rate in animals. Substances with hormonal actions are prohibited in the European community for use in animals intended for meat production due to their possible toxic effects on public health [83].

When bound to human estrogen receptors, they can stimulate the transcriptional activity of various estrogen receptor subtypes. The increase in levels of estrogen may be linked to the possibility of cancer occurrence among meat consumers [85].

In addition, studies have proved the ability of BPA to cross the placenta and blood brain barrier, in turn affecting the endocrine organs in animals and humans. This includes increasing prostate size, decreasing the number of produced sperms, and causing early puberty in females, in addition to effects on sexual differentiation [62,87]. Therefore, several HPLC-FLD methods were developed and used for quantification of EDC’s.

A sensitive HPLC method was developed for quantification of BPA and eight compounds of alkylphenol [59]. The alkylphenols assessed were; 4-sec-butylphenol, 2-tert-butylphenol, 3-tert-butylphenol, 4-tert-butylphenol, 4-n-pentylphenol, 4-tert-pentylphenol, 4-n-hexylphenol, and 4-n-heptylphenol. Derivatization was done using 2-(4-carboxyphenyl)-5,6-dimethylbenzimidazole (CDB) at 40 °C for 60 min. A C₁₈ stationary phase was used for separation. The reported detection limits were in the range of 0.1–10.0 pg/mL. The method was applied to determination of bisphenol A in mother and infant rat serum samples.

The fluorescent reagent 4-(4,5-diphenyl-1H-imidazol-2-yl) benzoyl chloride (DIB-Cl) (Figure 12) was used by Sun et al. to detect EDC’s. To the evaporated sample residue, DIB-Cl suspension in acetonitrile and triethylamine in acetonitrile were added and reacted at room temperature [63].

The study determined BPA in rat brain samples using HPLC-FLD coupled with a microdialysis [63]. A microdialysis probe was inserted into the hypothalamus of rat brains and artificial cerebrospinal fluid was used for perfusion. After the administration of a single intravenous or oral dose of BPA, concentrations were monitored in brain and plasma for 8 h. The obtained data proved that BPA could penetrate the blood brain barrier. The LOD of BPA was 0.3 ppb in 60 µL brain microdialysate at (S/N = 3).

Kuroda et al. used a similar method for determination of BPA in human blood serum and ascitic fluid samples. Samples were extracted by LLE using chloroform [62]. The LOD of BPA for both samples was 0.04 ppb at (S/N = 3). Human breast milk was also used to investigate the presence of BPA employing DIB-Cl as a fluorescent derivatizing agent [61]. Two steps of LLE were applied, and two C₁₈ columns were used to separate DIB-BPA from the endogenous material in breast milk. The detection limit in 23 samples of healthy lactating women was 0.11 ng/mL at (S/N = 3).

The chromatographic conditions, sample preparation and detection limits of the HPLC-FLD methods that were employed in the steroid studies reviewed in this article are summarized and presented in Table 1.