In this study, we demonstrate the capability of a custom-built fluorescence microendoscope to sensitively detect microscopic ovarian cancer (OvCa) tumour nodules and monitor treatment response in vivo.
This effort is motivated by a vital clinical need for improved reassessment imaging for this disease. The dismal survival rates for OvCa are largely because of the high proportion of patients diagnosed at a late stage, which is characterised by disseminated nodular studding throughout the peritoneal cavity and presents challenges for treatment (Jemal et al, 2008
; Cho and Shih, 2009
). Despite initial responses to standard tumour debulking and chemotherapy, many of these patients continue to harbour occult disseminated disease without immediate clinical evidence of tumour recurrence because of the microscopic size of residual nodules (Berek and Bast, 2003
). Second-look laparotomy allows the surgeon to check for persistent residual tumours, although even this highly invasive surgical reassessment suffers from a false-negative rate of up to 50% (Selman and Copeland, 1999
; DiSaia and Creasman, 2002
). These laparotomies are typically performed 6 months to 2 years after treatment, at which point, persistent disease, if present, has progressed to the point that is readily evident to the surgeon and the possibility of an earlier intervention has been precluded. Non-invasive diagnostic techniques, such as CA125 levels, ultrasound, computerised tomography, magnetic resonance imaging (MRI), and positron emission tomography (PET), which could in principle provide the benefit of more timely feedback, have all been shown to be less sensitive than reassessment surgeries (Pectasides et al, 1991
; Sugiyama et al, 1996
; Selman and Copeland, 1999
; Morrow, 2000
; Rose et al, 2001
; Tammela and Lele, 2004
). Hence, there is a vital un-met need for a new minimally invasive imaging approach with sufficient sensitivity and resolution to detect sub-millimetre OvCa nodules early in the treatment cycle, thus providing the basis for more timely interventions to mitigate recurrent disease.
To address the above requirements, several key features of optical imaging warrant consideration, including its inherently high resolution and the ability to achieve greater specificity by use of fluorescent contrast agents (Prasad, 2003
; Vo-Dinh, 2003
). Furthermore, technological advances in fibre optic fluorescence imaging, a modality that allows investigators to reach into body cavities through minimally invasive endoscopes, have considerably broadened the applications of in vivo
optical imaging (Gmitro and Aziz, 1993
; MacAulay et al, 2004
; Flusberg et al, 2005
). Although second-look laparoscopy by white light imaging has received significant attention as a minimally invasive procedure to detect OvCa recurrence (Davila and Estape, 2001
), it suffers from a high false-negative rate of 55% (Ozols et al, 1981
) and also fails to provide the requisite optical resolution and sensitivity to detect the smallest residual tumours.
An approach to OvCa detection that has shown promise is the use of photosensitisers (PS), traditionally used for photodynamic therapy (PDT), to generate tumour-selective fluorescence contrast. Although fluorescence imaging of non-therapeutic imaging agents has also shown impressive potential for OvCa detection (Sheth et al, 2009
), PS imaging offers a unique advantage. Upon wavelength-specific activation of the PS, fluorescence emission for diagnostic imaging and cytotoxic species (through intersystem crossing to the triplet state) for tissue destruction can both be produced. Previous studies that used PS (ALA-induced protoporphyrin IX and/or hexaminolaevulinate) for OvCa imaging were successful in detecting significantly more lesions than white light imaging (Chan et al, 2002
; Ludicke et al, 2003
; Loning et al, 2004
), with a ratio of fluorescence intensity of neoplastic to normal tissues of ~4 (Chan et al, 2002
; Ludicke et al, 2003
). The average size of the optically biopsied metastatic lesions was 1.0
mm (range: 0.3–2.5) compared with 1.5
mm (range: 0.5–2.9) with white light illumination (Chan et al, 2002
). However, in the absence of instrumentation for high-resolution in vivo
microscopy in these studies, the smallest occult nodules (<300μ
m) were not within the reported size range.
It is our thesis that if the demonstrated capability of PS fluorescence for detection of OvCa nodules is combined with high-resolution fibre-optic microscopy, it would allow for highly sensitive and minimally invasive detection of microscopic disease. To this end, we developed a custom fluorescence microendoscope () optimised for imaging the weak fluorescence emission from benzoporphyin-derivative monoacid ring A (BPD-MA) (delivered in its liposomal formulation known as verteporfin or Visudyne). BPD-MA is clinically approved as a therapeutic PDT agent and is used in our study to provide both diagnostic and therapeutic capabilities after a single administration. This inherent capability to integrate imaging and PDT treatment, which makes PS imaging particularly conducive to online monitoring of therapy response, was not used in the studies referenced above. It is also worth noting that the application of PDT to the treatment of OvCa has shown promise in clinical trials (Sindelar et al, 1991
; Hendren et al, 2001
; Hahn et al, 2006
), although further studies are warranted to establish optimal conditions for light and PS delivery in the clinic.
Schematic of the fibre-optic fluorescence microendoscope imaging system.
The goal of this study was to explore the potential of our fluorescence microendoscope to detect disseminated small volume OvCa tumour nodules and provide more accurate and timely treatment response information than current technologies. We present characterisation of our system in vitro
, using an OvCa 3D model system (I Rizvi*
, JP Celli*
, CL Evans, A Abu-Yousif, T Hasan. Adherent ovarian carcinoma cells migrate and assemble in vitro
into heterogeneous 3D micronodules with differential treatment response. Manuscript in preparation, to be submitted. *
Equal contribution), and in vivo,
using an established mouse model of disseminated OvCa (Molpus et al, 1996a
), both of which were previously developed in our laboratory. To further demonstrate the capability of our system to monitor treatment response, we conducted endoscopic assessment before and after a single PDT treatment. In this study, PDT serves as a therapy with a variable level of cytotoxicity to provide a suboptimal dosage conducive for monitoring therapeutic efficacy.
As proof of principle, this study illustrates the significant translational potential of fluorescence microendoscopy for detection of OvCa. Specifically, this method could provide the clinician with the capability to obtain more timely and accurate treatment response feedback, which could be used to optimise treatment parameters to better predict, and hence mitigate, adverse treatment outcomes. Moreover, the capability to detect sub-millimetre tumour nodules suggests the potential utility of fluorescence microendoscopy in early detection of OvCa, if combined with serum-based or other screening tests to identify candidates for endoscopic examination.