In vivo native fluorescence spectroscopy and nicotinamide adinine dinucleotide/flavin adenine dinucleotide reduction and oxidation states of oral submucous fibrosis for chemopreventive drug monitoring

Sivabalan Sivabalan; C. Ponranjini Vedeswari; S. Jayachandran; D. Koteeswaran D.D.S.; C. Pravda D.D.S.; Prakashrao Aruna; Singaravelu Ganesan

doi:10.1117/1.3324771

1 January 2010 In vivo native fluorescence spectroscopy and nicotinamide adinine dinucleotide/flavin adenine dinucleotide reduction and oxidation states of oral submucous fibrosis for chemopreventive drug monitoring

Sivabalan Sivabalan, C. Ponranjini Vedeswari, S. Jayachandran, D. Koteeswaran D.D.S., C. Pravda D.D.S., Prakashrao Aruna, Singaravelu Ganesan

Author Affiliations +

Journal of Biomedical Optics, Vol. 15, Issue 1, 017010 (January 2010). https://doi.org/10.1117/1.3324771

Abstract

Native fluorescence spectroscopy has shown potential to characterize and diagnose oral malignancy. We aim at extending the native fluorescence spectroscopy technique to characterize normal and oral submucous fibrosis (OSF) patients under pre- and post-treated conditions, and verify whether this method could also be considered in the monitoring of therapeutic prognosis noninvasively. In this study, 28 normal subjects and 28 clinically proven cases of OSF in the age group of 20 to 40 years are diagnosed using native fluorescence spectroscopy. The OSF patients are given dexamethasone sodium phosphate and hyaluronidase twice a week for 6 weeks, and the therapeutic response is monitored using fluorescence spectroscopy. The fluorescence emission spectra of normal and OSF cases of both pre- and post-treated conditions are recorded in the wavelength region of 350 to 600 nm at an excitation wavelength of 330 nm. The statistical significance is verified using discriminant analysis. The oxidation-reduction ratio of the tissue is also calculated using the fluorescence emission intensities of flavin adenine dinucleotide and nicotinamide adinine dinucleotide at 530 and 440 nm, respectively, and they are compared with conventional physical clinical examinations. This study suggests that native fluorescence spectroscopy could also be extended to OSF diagnosis and therapeutic prognosis.

1. Introduction

Head and neck cancer is one of the most common cancers in the world, nearly two-thirds of which occur in developing countries. The overall survival rates for cancer are low in developing countries.¹ It is well documented that most invasive oral cancers arise from precancerous lesions such as leukoplakia, erythroplakia, and oral submucous fibrosis (OSF).² OSF is a high-risk precancerous condition with a reported prevalence ranging up to 0.4% in Indian rural populations³ The histopathological feature of OSF includes atrophy of oral epithelium with parakeratosis or hyperkeratosis, as well as marked collagen deposition and hyalinization in the lamina propria, submucous, and superficial muscle layer.⁴ Many reported that the chewing of areca nut and tobacco is a major possible cause for OSF in India, which includes genetic, viral, autoimmune, and carcinogenic agents.^{5, 6}

Although the accessibility of the oral cavity is easier than other sites, most patients present with advanced stages and require more aggressive treatment. This result yields very poor survival rate and higher morbidity,⁷ and hence there arises a need for early detection techniques. However, conventional screening and detection of oral cancer by visual inspection of the oral cavity with white light may be difficult to discriminate the early stage or oral cancer from nonspecific inflammation and irritation,⁸ In this regard, various optical spectroscopic techniques, such as fluorescence spectroscopy,^{9, 10} diffuse reflectance spectroscopy,^{11, 12} synchronous luminescence spectroscopy,¹³ and Raman spectroscopy,¹⁴ have been considered to characterize and diagnose oral neoplasia and imaging. Among various optical spectroscopic techniques, native fluorescence spectroscopy has been widely considered in the characterization of metabolic and pathological conditions of cells and tissues at the molecular level due to its sensitivity and availability of many complementary techniques.

Using native fluorescence spectroscopy, different pathological conditions of oral malignancies were better discriminated from normal subjects with very high statistical significance.^{15, 16, 17} Among various native fluorophores present in oral tissues, the emission contribution from tryptophan, collagen, flavin adenine dinucleotide (FAD) and nicotinamide adenine dinucleotide (NADH), and endogenous porphyrins were exploited in the discrimination of normal from oral malignancies. The fluorescence of collagen, elastin, and more generally proteins is due to the presence of aromatic amino acids that relate to the structural arrangement of cells and tissue. NADH, FAD, and endogenous porphyrin are related to metabolic processes or are in connection with the onset of pathological conditions.^{18, 19, 20} Further, it is also reported that the emission characteristics of these native fluorophores also vary with different anatomical sites of the oral cavity.^{21, 22} Recently, the relative change in the fluorescence emission intensity of FAD and NADH was also considered to study the oxidation-reduction state of cells, as the redox ratio is used to monitor the cellular metabolic rate under normal and neoplastic conditions.^{23, 24}

Although many extensive studies were made in the characterization and diagnosis of cancer using native fluorescence spectroscopy, to the best of our knowledge there are no reports of in vivo study on the monitoring of treatment prognosis of OSF. In this context, the focus of the study was on the in vivo characterization of healthy and OSF patients under pre- and post-treated conditions using native fluorescence spectroscopy. Attempts were also made to verify the statistical significance of the prior techniques in the discrimination of normal from OSF as well as pretreated OSF conditions from post-treated conditions.

2. Materials and Methods

2.1.

Patients

28 patients having a premalignant lesion condition of OSF from the Tamilnadu Government Dental College and Hospital Chennai, India, but not having any systemic diseases such as diabetes, hypertension, hepatic or renal diseases, or patients with any kind of allergy, were involved in this study. All patients had a chewing habit for more than three years and patients with grade IV and V oral submucous fibrosis were involved in this study. All patients had a mouth opening greater than $15 mm$ and an intolerance to spicy food, with fibrous bands in buccal mucosa. The clinical diagnoses were confirmed with the histopathological reports.

28 healthy volunteers with no clinically observable lesion and no habit of chewing areca (betel) nut or tobacco were used as normal subjects. All the patients and volunteers were in the age group of $20 to 40 years$ .

2.2.

Oral Subnucous Fibrosis Treatment

Patients were given an intralesional injection of a combination of dexamethasone sodium phosphate $(4 mg ∕ ml)$ and hyaluronidase (1500 IU) twice a week for $6 weeks$ . They were also given antioxidants and micronutrient supplementation and were taught muscle stretching exercises. All the treatments were implemented at the Tamilnadu Government Dental College and Hospital, Chennai, India.

Dexamethasone sodium phosphate injection $(4 mg ∕ ml)$ marketed by Prem Pharmaceuticals (Bangalore, India) was used in the study. Each milliliter of solution contains dexamethasone sodium phosphate IP equivalent to dexamethasone phosphate $4 mg$ , and methylparaben IP 0.18% w/v and prophylparaben IP 0.02% w/v as preservatives.

Hyaluronidase injection used in the study was hyanidase IP, manufactured by Shreya Life Sciences (Aurangabad, India). Each vial contains hyaluronidase as a freeze-dried powder, providing activity equivalent to 1500 IU.

At each visit, following the topical application of lignocaine 2%, 1500 IU of hyaluronidase was dissolved in $2.0 ml$ of dexamethasone sodium phosphate in a $2 - ml$ disposable syringe. The drugs were injected at multiple sites where the fibrotic bands were palpable submucosally by means of a gauge 24 needle, not more than $0.2 - ml$ solution per site.

The patients were clinically examined at every visit during the treatment period of $6 weeks$ by evaluating maximal mouth opening (MMO) and tongue protrusion (TP).

2.3.

Steady-State Fluorescence Measurements

The in vivo fluorescence was recorded using a Fluoromax-2 (ISA Jobin Yvon-Spex, Edison, New Jersey) spectrofluorometer. The excitation source ( $150 - W$ ozone-free xenon arc lamp) coupled to the monochromator delivers light to the sample spot at a desired wavelength, and the fluorescence emission from the tissues is collected by another monochromator connected to a photomultiplier tube (R928P, Hamamatzu, Shizuoka-Ken, Japan). The gratings in the excitation and emission monochromators have a groove density of $1200 grooves ∕ mm$ and are blazed at 330 and $500 nm$ , respectively. During fluorescence data acquisition, the excitation and emission slit widths are set at 5 and $8 nm$ , respectively, with an integration time of $0.5 s$ . The collected signal is transferred to a PC through an RS232 interface. The data were processed by the Windows-based data acquisition program Data Max powered by GRAMS/386. In the in vivo measurements, a bifurcated randomized fiber optic bundle of model 1950M is used with the fiber optic platform of model 1950F (see Fig. 1 ).

Fig. 1

Schematic diagram showing the in vivo fluorescence measurement setup.

During fluorescence measurements, the common end of the fiber assembly is kept perpendicular to the tissue surface, whereas the excitation and collection ends of the fiber assembly are connected to the excitation and emission monochromators, respectively. As the fluorescence emission from the tissues is highly anisotropic and they are emitted in all directions, more fibers are used at the collection end of the fiber assembly to improve collection efficiency. Further, to collect the maximum fluorescence signal over the solid angle of its emission, the common end of the fiber assembly is designed such that there is always a distance of $3 mm$ between the tissue surface and the fiber tip during the measurements. Before recording the spectra, all the volunteers and OSF patients were asked to rinse their mouth for $1 \min$ with saline solution to eliminate the influence of consumed food. The probe is placed at the site where fibrotic bands are palpable, as suggested by the physician, and for each site two spectra were recorded.

2.4.

Reduction and Oxidation Ratio

The reduction and oxidation ratio was computed using the relation

redox ratio = \frac{{FAD}_{intensity}}{{FAD}_{intensity} + {NADH}_{intensity}},

where FAD intensity and NADH intensity are the emission intensity of FAD and NADH at 530 and

440 nm

, respectively.²⁵

2.5.

Data Processing

Normalization

Each fluorescence spectrum was normalized by dividing the fluorescence intensity at each emission wavelength by the peak emission intensity of the spectrum. Normalizing a fluorescence spectrum removes absolute intensity information, and the main advantage of utilizing a normalized spectrum is that fluorescence intensity does not need to be recorded in calibrated units.^{26, 27}

Introducing ratio variables

To quantify the results and to estimate the potential of the present technique, 12 ratio variables were introduced. Emission wavelengths $390 \pm 10$ , $450 \pm 10$ , and $530 \pm 10 nm$ were taken by considering the peak emission wavelength of collagen, NADH, and FAD, respectively. The mean and standard error values of all the ratio variables were calculated for each group of experimental subjects. A two-tailed student’s t-test was performed to determine the level of significance ( $p$ value) with which each ratio variable discriminates diseased subjects with respect to normal subjects.

The following three different discriminant analysis were performed to:

1. discriminate between the OSF patients and the normal subjects $(D_{1})$
2. discriminate between pre- and post-treated OSF conditions $(D_{2})$
3. find the discrimination factor for normal, pre-, and post-treated conditions $(D_{3})$

In the present work, the discriminant analysis was performed using SPSS (SPSS Incorporated, Chicago, Illinois). This analysis used a partial F-test (F to enter 3.84 and F to remove 2.71) and a step-wise method to sequentially incorporate the set of 12 variables into a Fisher linear discriminant function. Step-wise analysis based on maximizing an F ratio was computed from the Mahalanobis distance (measure of how much a case’s values on the independent variables differ from the average of all cases. A large Mahalanobis distance identifies a case as having extreme values on one or more of the independent variables) between the groups. Further, we use leave-one-out cross-validation. In this procedure, discriminant scores of one particular sample were eliminated, and discriminant analysis was used to form a classification algorithm using the remaining samples. The resulting algorithm was then used to classify the excluded case. This process was repeated for each sample.

3. Results

The native fluorescence emission spectra from healthy oral mucosa, OSF under both pre- and post-treated conditions, were measured at $330 - nm$ excitation $(λ_{em} = 350 to 600 nm)$ . The individual case and clinical details of OSF patients such as age, duration, and frequency per day of areca net chewing, clinical grading, MMO, and TP before and after treatment are given in Table 1 .

Table 1

Clinical report of the individuals before and after treatment who were involved in the study (BT is before treatment and AT is after treatment).

Number	Age/Sex	Areca quidchewing		Durationofsymptoms(inmonths)	Sites ofinvolvement	Clinicalgrading	Maximalmouthopening(in mm)		Tongueprotrusion(in mm)
Number	Age/Sex	Durationin months	Frequencyper day	Durationofsymptoms(inmonths)	Sites ofinvolvement	Clinicalgrading	BT	AT	BT	AT
1	38/F	276	2	60	4	V	17	26	15	20
2	26/M	180	15	6	4	IV	22	30	30	30
3	20/M	180	10	3	6	V	16	28	15	25
4	40/M	240	12	24	6	IV	17	32	30	35
5	27/M	120	10	36	5	IV	19	29	30	30
6	26/M	360	15	3	5	V	20	32	20	25
7	36/M	600	6	12	4	V	16	28	20	30
8	29/M	240	3	12	4	IV	18	31	30	30
9	20/M	120	4	3	6	IV	20	31	30	25
10	28/M	240	3	1	4	V	20	32	15	20
11	40/F	120	3	8	5	V	16	29	15	30
12	38/M	180	8	12	5	V	19	30	15	25
13	35/M	120	25	3	4	IV	17	28	30	30
14	36/M	300	5	12	3	IV	16	29	25	30
15	40/M	240	2	12	6	IV	18	29	25	25
16	40/M	120	3	12	5	V	21	31	10	25
17	38/M	600	6	18	4	IV	16	26	25	35
18	30/M	720	5	4	4	V	16	24	20	25
19	39/M	960	5	12	3	V	18	30	15	20
20	24/M	240	5	18	3	IV	20	31	30	30
21	28/F	220	10	15	5	V	16	28	15	25
22	37/M	200	15	12	6	V	17	32	30	35
23	42/M	280	4	10	8	V	17	31	15	25
24	30/M	320	2	18	4	IV	15	26	25	35
25	21/M	600	12	6	4	IV	16	30	15	30
26	27/M	720	2	12	3	IV	19	30	15	25
27	39/F	960	3	18	6	V	19	26	15	20
28	40/M	320	7	8	5	V	22	32	30	32

3.1.

In Vivo Fluorescence Spectroscopy

The averaged fluorescence emission spectra of healthy subjects and patients with OSF under pre- and post-treated conditions at $330 - nm$ excitation are shown in Fig. 2 . The fluorescence emission spectra of normal and OSF (under pre- and post-treated) show two peaks centered at 390 and $465 nm$ , with a well in-between the two peaks at $420 nm$ . The emission intensity at $390 nm$ for OSF cases is higher than that of the emission at $465 nm$ . On the other hand, the emission peak at $390 nm$ for normal oral mucosa is considerably less than that at $460 nm$ . The overall fluorescence intensity of OSF cases is higher than that of normal subjects. Further, it is observed that the emission intensity of OSF at $390 nm$ is around 1.25 times higher than that of normal mucosa. The fluorescence emission spectra of OSF patients were also measured after a $7 -week$ period of the post-treatment as discussed in Sec. 2. The fluorescence emission spectra of post-treated OSF cases have spectral signatures similar to that of normal oral mucosa, i.e., the emission intensity at $390 nm$ is less than that at $465 nm$ . However, it is noted that the overall fluorescence intensity is comparatively less than that of the normal subjects.

Fig. 2

In vivo fluorescence spectra of normal (- - -), pretreated (—), and posttreated (—●—) OSF cases and normal in the wavelength region $350 to 650 nm$ at an excitation of $330 nm$ . The spectra were measured at every $1 - nm$ interval with an integration time of $0.5 \sec$ . The excitation and emission slit widths were kept at 5 and $8 nm$ , respectively.

Figure 3 shows the averaged normalized emission spectra of normal, pre-treated, and post-treated OSF cases. From Fig. 3, it is clearly observed that the emission intensity at $390 nm$ for normal and post-treated OSF cases is less than that at $465 nm$ . Both the normal and post-treated cases exhibit similar spectral signatures.

Fig. 3

Averaged normalized in vivo fluorescence spectra of pretreated (—) and post-treated (—●—) OSF cases and normal (- - -) in the wavelength region $350 to 650 nm$ at an excitation of $330 nm$ . The spectra were measured at every $1 nm$ interval with an integration time of $0.5 \sec$ . The excitation and emission slit widths were kept at 5 and $8 nm$ , respectively.

3.2.

Statistical Analysis

To quantify and verify the statistical significance of the observed spectral differences between normal and OSF patients, as well as the spectral differences between pre- and post-treated OSF patients, 12 statistically significant ratio variables were introduced. The mean values of these ratio variables with standard deviation for the previous groups are given in Table 2 . In comparison of normal and pretreated OSF cases, the $P$ value for most of the ratio parameters is highly significant $(p = 0.001)$ , except for the variables $V_{11}$ and $V_{12}$ , whose $p$ values are found to be 0.1. Similarly, in comparison of pretreated cases with post-treated OSF cases, the ratio variables are highly significant, with $p$ value less than 0.0001. From this, it is observed that all the ratio values of normal and post-treated conditions of OSF are almost the same, and they differ from pretreated OSF cases with excellent statistical significance. The ratio variables included for discriminant analysis are given in Table 2.

Table 2

The ratio parameters used for statistical analysis with mean±SD after passing the student t-test.

Variables	Parameters	Normal	Pretreated	Post-treated
$V_{1}$	$I_{385} ∕ I_{550}$	$1.869 \pm 0.053$	$2.377 \pm 0.115$	$1.778 \pm 0.048$
$V_{2}$	$I_{387} ∕ I_{550}$	$1.868 \pm 0.053$	$2.382 \pm 0.116$	$1.778 \pm 0.049$
$V_{3}$	$I_{390} ∕ I_{550}$	$1.868 \pm 0.053$	$2.379 \pm 0.116$	$1.789 \pm 0.050$
$V_{4}$	$I_{385} ∕ I_{468}$	$0.959 \pm 0.010$	$1.161 \pm 0.024$	$0.915 \pm 0.022$
$V_{5}$	$I_{387} ∕ I_{468}$	$0.958 \pm 0.009$	$1.164 \pm 0.024$	$0.915 \pm 0.022$
$V_{6}$	$I_{390} ∕ I_{468}$	$0.958 \pm 0.009$	$1.162 \pm 0.024$	$0.919 \pm 0.021$
$V_{7}$	$I_{385} ∕ I_{335}$	$1.091 \pm 0.022$	$1.375 \pm 0.057$	$1.075 \pm 0.038$
$V_{8}$	$I_{387} ∕ I_{355}$	$1.090 \pm 0.022$	$1.379 \pm 0.058$	$1.076 \pm 0.039$
$V_{9}$	$I_{390} ∕ I_{355}$	$1.090 \pm 0.022$	$1.377 \pm 0.059$	$1.083 \pm 0.040$
$V_{10}$	$I_{387} ∕ I_{430}$	$1.025 \pm 0.009$	$1.132 \pm 0.014$	$0.962 \pm 0.018$
$V_{11}$	$I_{530} ∕ I_{450}$	$0.602 \pm 0.011$	$0.572 \pm 0.014$	$0.601 \pm 0.013$
$V_{12}$	$I_{450} ∕ I_{530}$	$1.676 \pm 0.033$	$1.779 \pm 0.046$	$1.685 \pm 0.038$

These ratio variables were subjected to the step-wise multiple linear discriminant analysis as follows.

3.2.1.

Discriminant Analysis 1 (Discrimination of Oral Submucous Filorosis from Normal)

The step-wise multiple linear discriminant analysis was performed between the 28 pretreated and 28 normal subjects, resulting in the following expression for a canonical discriminant function (DF):

D_{1} = 8.946 (V_{1}) - 26.464 (V_{5}) - 3.891 (V_{7}) + 26.253 (V_{11}) - 1.641 .

Of these 12 input variables, four variables turned out to be significant and were included in the linear discriminant function. Figure 4 shows the scatter plot of the discriminant score, i.e., the value of the discriminant function for normal and OSF individuals. $D_{1}$ shows that OSF cases are discriminated with a sensitivity of 89% using the present techniques. In this discriminant analysis, out of 28 OSF cases, three cases were misclassified as normal. However, all the normal groups were classified correctly with a specificity of 100%. In this analysis, it is also obtained that 95% of the original grouped cases and 93% of the cross-validated grouped cases were correctly classified. From the figure it is observed that the value of $D_{1}$ is greater than zero for normal oral mucosa, and $D_{1}$ is less than zero for pretreated conditions (Fig. 4). The classification results of this discriminant analysis are given in Table 3 .

Fig. 4

Scatter plot showing the distribution of $D_{1}$ for normal (●) and pretreated (○) oral submucous fibrosis.

Table 3

Classification results of the discriminant analysis done across the different groups. Bold values represent the specificity/sensitivity value of the corresponding groups. The total percentage of correctly classified cases gives the average of the percentage of predicted groups.

Discriminantanalysis	Cases	Group	Percentage ofpredicted group			Total percentageof correctlyclassified cases
Discriminantanalysis	Cases	Group	1	2	3	Total percentageof correctlyclassified cases
$D_{1}$	Original	1.00	89.3	10.7	—	94.6
		2.00	0.0	100.0	—
	Cross-validated	1.00	89.3	10.7	—	92.9
		2.00	3.6	96.4	—
$D_{2}$	Original	1.00	85.7	14.3	—	89.3
		2.00	7.1	92.9	—
	Cross-validated	1.00	85.7	14.3	—	89.3
		2.00	7.1	92.9	—
$D_{3}$	Original	1.00	71.4	25.0	3.6	69.0
		2.00	0.0	71.4	28.6
		3.00	7.1	28.6	64.3
	Cross-validated	1.00	71.4	25.0	3.6	69.0
		2.00	0.0	71.4	28.6
		3.00	7.1	28.6	64.3

3.2.2.

Discriminant Analysis 2 (Discrimination of Pretreated Oral Submucous Fibrosis from Post-Treated Oral Submucous Fibrosis)

The step-wise multiple linear discriminant analysis was also performed across the whole set of 28 pre- and post-treated OSF cases, resulting in the following expression for a canonical discrimination function:

D_{2} = 8.308 (V_{4}) - 8.626 .

Of these 12 variables, only one variable turned out to be significant. From these 12 input variables, $V_{4}$ $(I_{385} ∕ I_{460})$ turned out to be significant and was included in the linear discriminant function $D_{2}$ . Figure 5 shows the scatter plot of the discriminant score, i.e., the values of discriminant function for pre- and post-treated OSF patients. In this analysis, the pre- and post-treated conditions are correctly classified with 86 and 93%, respectively. From the scattered plot, it is observed that $D_{2}$ is greater than zero for pretreated cases, and $D_{2}$ is less than zero for post-treated cases.

Fig. 5

Scatter plot showing the distribution of $D_{1}$ for pretreated (○) and post-treated (▼) oral submucous fibrosis.

3.2.3.

Discriminant Analysis 3

A third discriminant analysis was carried out across normal, pretreated OSF and post-treated OSF groups. Among 12 variables, only four variables— $V_{2}$ , $V_{6}$ , $V_{8}$ , and $V_{12}$ —were retained in this analysis. Resulting canonical discriminant functions are given by

D_{3 a} = 263.934 (V_{2}) - 390.631 (V_{6}) + 10.59 (V_{8}) + 1499.009 (V_{12}) - 522.763,

D_{3 b} = 291.072 (V_{2}) - 470.908 (V_{6}) - 1.214 (V_{8}) + 1578.650 (V_{12}) - 527.741 .

Figure 6 shows the scatter plot of $D_{3 a}$ versus $D_{3 b}$ , the corresponding group centroids and the standard deviation from the group centroids. The classification results are shown in Table 3. From $D_{3}$ analysis, pre- and post-treated OSF, and normal mucosa were compared, and it was found that all the pretreated OSF cases showed values of $D_{3 a}$ , which is greater than zero, and the normal and post-treated OSF showed $D_{3 a}$ , which is less than zero (Fig. 6). Further, it was found that the discriminant factor $D_{3}$ for normal and post-treated cases grouped together, and they were well discriminated from the pretreated OSF cases. This clearly indicates that there is a significant change in the oral cavity due to treatment.

Fig. 6

Scatter plot showing the distribution of $D_{3}$ versus $D_{4}$ for normal, pretreated, and post-treated oral submucous fibrosis.