1.1.

Bladder

Each year there are 9000 cases of bladder cancer diagnosed in men and 3600 diagnosed in women in the United Kingdom. It is the fourth most common cancer in men and the eighth most common in women (the incidence is 8% in men and 3% in women¹). There has, however, been a reduction in the age standardized incidence since the 1980s. This is thought to be due to the reduction in smoking and the banning of aromatic amines in the 1980s, both of which are known risk factors for the development of bladder cancer.

The vast majority of bladder cancers are transitional cell carcinomas of the bladder, and the majority of these are superficial. Bladder cancer is usually diagnosed by means of a cystoscopy and biopsy under general anaesthetic. Once diagnosed, patients are then staged using the tumor-nodes-metastases (TNM) classification and entered into a surveillance program, if no further treatment is required. The surveillance program involves six monthly to yearly cystoscopies for at least ten years.

Unfortunately, some of the bladder tumors, especially carcinoma in situ and flat superficial tumors, can look just like cystitis, or normal bladder.

1.2.

Prostate Gland

The majority of men will suffer with symptoms related to diseases of the prostate gland at some point in their lives, with prostate cancer being the one of main concern. Adenocarcinoma of the prostate gland (CaP) is the most frequently diagnosed noncutaneous cancer affecting Western men.² Most (95%) of the cancers affecting the prostate gland are adenocarcinomas, the other 5% are made up of the following: transitional cell carcinoma, neuroendocrine carcinoma, sarcoma, and lymphoma. 40% of men aged between 60 and $70\text{years}$ have microscopic foci of well-differentiated CaP. It is the second leading cause of cancer-related death in men in Western Europe and North America,² with 3 to 5% of men dying of CaP. It is also a major cause of morbidity and healthcare-related costs with 10% of men diagnosed with CaP developing clinical disease.²

CaP is diagnosed by means of an abnormal digital rectal examination (DRE) and/or a raised level of serum prostate-specific antigen (PSA). Either of these will lead onto transrectal ultrasonography (TRUS) and biopsy. The biopsies are taken randomly, with targeting if a lesion is suspected on TRUS. Unfortunately, the biopsies have a high number of false negatives, and therefore a cancer can remain undiagnosed. These false negatives are mainly due to sampling error (CaP is present as a small focus within the prostate gland and the biopsy has missed it). It has been shown that even if a man has two sets of negative biopsies with a persistently raised PSA, he has a 10% risk of having a cancer found on the third biopsy.

PSA measurement also comes with its own problems, in that although it is highly sensitive, it is not very specific. PSA can also be raised in the following: benign prostate hyperplasia (BPH), prostatitis, acute urinary retention (AUR), and urinary tract infection (UTI), as well as following recent urethral instrumentation, catheterization, prostatic biopsy, and DRE.

The ideal would be to diagnose CaP at an early stage when the PSA is within the range of $2.5\text{to}10\mathrm{ng}∕\mathrm{ml}$ ; this would increase the chances of cure with radical treatment. The reality, given the low specificity of both the PSA, and the TRUS and biopsy, is that many of the cancers are diagnosed when a cure is not possible.

1.3.

Raman Spectroscopy

Raman spectroscopy is an optical technique that can act to interrogate biological tissues with chemical specificity. In doing so, it gives us an understanding of the changes in the molecular structure that is associated with disease. With this in mind, it was felt that Raman spectroscopy could be used to distinguish between pathologies within the bladder and the prostate gland. By the start of the 1990s, various groups were using Raman spectroscopy to distinguish between normal and neoplastic tissue. The first studies looked at differentiating between normal tissue and advanced cancers in the breast³ and gynecological organs.^{4, 5} As techniques were refined, interest moved to diagnosing neoplastic change at progressively earlier stages. To date, in vitro studies have also been undertaken to differentiate between different pathologies in a number of other tissues including colon,^{6, 7} esophagus,^{8, 9, 10} brain,¹¹ skin,^{12, 13} lung,¹⁴ and larynx.¹⁵

It has already been shown that Raman spectroscopy can be used to distinguish between different pathologies within the bladder and the prostate gland. This, however, has been exclusively in vitro.^{16, 17}

In view of previous studies, Raman spectroscopy is thought to have the potential for minimally invasive detection of malignancies and premalignancies within the bladder and the prostate gland. A problem with this is the inability to obtain spectra from significantly beneath the surface (more than of the order of $100\mu \mathrm{m}$ ).¹⁸ This is needed to ascertain whether there has been any local invasion of the disease in the case of the bladder, or to find a focus of adenocarcinoma in an otherwise benign prostate gland.

1.4.

Kerr Gating

A Raman signal from depths within tissues tends to be diminished by elastic scattering, and therefore the surface signal is usually significantly stronger. A possible way to overcome this is by temporal gating techniques such as picosecond Kerr gating. The Kerr-gating technique uses excitation with a picosecond pulsed laser, in combination with a fast temporal gating of the Raman scattered light. The scattered light is collected at various time delays following the laser pulse.

The ability of depth probing using Raman is determined by two stages: 1. the laser photons have to be able to migrate to a given depth within the tissue; and 2. the Raman photons that have been produced have to migrate back to the surface.

In this way Raman spectra from differing depths within the tissue will emerge at different times, thereby making their separation feasible.¹⁹

There have been studies utilizing the Kerr-gating technique to improve on Raman spectra obtained from bone interiors,²⁰ as well as test experiments demonstrating the depth resolving power of the Raman Kerr-gating concept on artificially prepared samples.¹⁹ This study is designed to expand on this work by studying the facility of Kerr-gated Raman spectroscopy to measure biochemical data from depths of a few millimeters in soft tissue. To make certain that a signal has been detected from beneath the tissue layer, specimens with strong and distinct Raman signals have been used. These were also selected for their relevance in urology.

2.1.

Subjects

We obtained approval from the Gloucestershire Research Ethics Committee to obtain samples for experimentation using Raman spectroscopy. The samples used for this study were taken following fully informed consent.

2.2.

Tissue Collection and Preparation

Prostate samples were obtained by taking an extra core at prostate biopsy procedures, or a chip at trans-urethral resection of the prostate (TURP). The prostate sample used in this study was assessed histopathologically to be benign prostatic hyperplasia. Bladder samples were collected at cystoscopic procedures including TURP (normal biopsies) and during trans-urethral resection of bladder tumor (TURBT). The bladder sample used was histologically normal and included the urothelial and subsurface layers. The samples were then snap-frozen in liquid nitrogen and transferred to an $-80°\mathrm{C}$ freezer for storage.

At the time of the experiment, the tissue was passively warmed to room temperature and placed either on a cell containing urea, or on one containing uric acid. Urea and uric acid were chosen because they can be found within the bladder and prostate gland, and may also form crystals. Urea and uric acid are also very strongly Raman active, and were chosen to allow a demonstration of probing of Raman spectra at depths through the tissue samples.

2.3.

Instrumentation

The Kerr-gating system is based on the high throughput $4\text{-}\mathrm{ps}$ optical Kerr shutter that has been described in previous publications by Matousek ^{21, 22} The Kerr gate is made up of two crossed polarizers ( $41\times 41\mathrm{mm}$ , Glan Taylor polarizers) and a Kerr medium consisting of a $2\text{-}\mathrm{mm}$ optical cell that is filled with $\mathrm{C}{\mathrm{S}}_2$ (see Fig. 1 ).

Fig. 1

A schematic diagram and photograph showing the Kerr-gated Raman system at the Rutherford Appleton Laboratory (RAL).

When the gate is closed, the light from the sample is blocked as the polarizers are in cross orientation. When the gate is open, the light collected from the sample has its polarization rotated to allow it to pass through the cross polarizer. The gate is opened by a short $1\text{-}\mathrm{ps}$ gating pulse at $800\mathrm{nm}$ . This bypasses the polarizers and creates a transient anisotropy within the Kerr medium, this then acts to polarize the light from the sample.

The spectra were collected using a conventional Raman spectrometer (Spex, Triplemate). To prevent residual elastically scattered light from the $488\text{-}\mathrm{nm}$ probe laser entering the spectrometer, a Kaiser holographic notch filter was used. To also prevent the residual $800\text{-}\mathrm{nm}$ gating beam scatter from entering, a saturated copper sulphate solution in a $1\text{-}\mathrm{cm}$ -thick optical cell was placed in front of the spectrometer slit.

The Raman scattered light was collected in $180\text{-}\mathrm{deg}$ geometry using a lens with f number of 2. A water heat-exchanger cooled, deep depletion near-infrared (NIR) CCD (Andor Technology, Belfast, UK, DU420BR-DD) was used to record the Raman spectra from the tissue.

The probe wavelength used was $488\mathrm{nm}$ , and the pulse duration was $1\mathrm{ps}$ . The pulse energy at the sample was $5\mu \mathrm{J}$ $\left(1\mathrm{kHz}\right)$ corresponding to $5\mathrm{mW}$ of average power. The beam was focused down to a $300\text{-}\mu \mathrm{m}\text{-}\mathrm{diam}$ spot.

To demonstrate the ability to obtain Raman spectra from deep layers, we performed two experiments: the first involved placing a bladder sample urothelium upward onto a uric acid cell, and the second involved placing the prostate sample onto a urea cell. Both tissue samples were $1\text{to}2\mathrm{mm}$ thick. The cells were made from UV-grade quartz $100\mu \mathrm{m}$ thick. The spectra were obtained for $20\mathrm{s}$ and 5 accumulations at varying distances through the tissue, i.e., at various time delays following the laser probe excitation pulse.

A simple approximation as to the distance traveled by the incident photons in the tissue can be made (the refractive index of the tissue was taken to be 1.4). In a time of $1\mathrm{ps}$ , the photons will have traversed approximately $0.2\mathrm{mm}$ of the tissue. This will not necessarily be in a straight line, as the tissue is highly scattering. By taking the scattering coefficient ${\mu }_{\mathrm{s}}$ to be $48{\mathrm{mm}}^{-1}$ and the absorption coefficient ${\mu }_{\mathrm{a}}$ to be $1.9{\mathrm{mm}}^{-1}$ (for epithelial tissue at $577\mathrm{nm}$ ),²³ an optical depth (OD) can be calculated from