2.1.

Diagnostic Problem

Iron deficiency can be detected by several methods,^7–11 which are invasive and require tissue or blood samples for laboratory analysis. Zinc protoporphyrin-IX (ZnPP), a metallo-porphyrin, is produced during heme biosynthesis when the supply of iron is limited and, therefore, is alternatively formed instead of heme by ferrochelatase (EC 4.99.1.1) from zinc ions and protoporphyrin-IX (PPIX), yet in a very low concentration, as shown in Fig. 1. Both, ZnPP and PPIX are located within erythrocytes.¹². For diagnostic purposes, the ZnPP/heme ratio is preferred over the absolute concentration of ZnPP (Ref. 12) as the ratio is independent of patient hematocrit. An elevated ZnPP/heme ratio most commonly indicates iron deficiency or lead exposure,^11–14 and a lowered ZnPP/heme ratio may be found in hereditary hemochromatosis.^12,15 The upper threshold for the ZnPP/heme ratio differs between studies and methods but is usually in the range of 40 to $80\mu \mathrm{mol}$ ZnPP/mol heme.^7,12,16 Especially in hospitalized patients, the specificity of the ZnPP/heme ratio for nutritional iron deficiency may be influenced by coexisting disorders, such as lead poisoning, anemia of chronic disease, or chronic inflammation.^7,17,18 In some circumstances, the ZnPP/heme ratio may serve as an index of chronic inflammation and can be used to monitor the effectiveness of treatment.¹⁷

Fig. 1

Synthesis of zinc protoporphyrin-IX (ZnPP) instead of heme in case of Fe deficiency.¹²

Routinely, ZnPP and PPIX concentrations can be measured by extraction and high-performance liquid chromatography (HPLC) separation^13,19 and detection by its fluorescence light emission upon blue light excitation. To calculate the ZnPP/heme ratio, additionally a routine Hb measurement is required. A low-cost and rapid method for determining the ZnPP/heme ratio is the use of a portable front-face fluorometer, the hematofluorometer,²⁰ requiring only a drop of (capillary or venous) blood, which directly measures the fluorescence light emitted by the erythrocyte ZnPP. In reasonable approximation, the signal is independent of the hematocrit²⁰ and is a direct measure for the ZnPP/heme ratio.^20,21 Due to its simplicity, the hematofluorometer is recommended as a screening device for targeted iron supplementation.²² However, the signal detected is influenced by background fluorescence of other blood constituents.^16,23–27 Potential elimination of background fluorescence entails further requirements, e.g., extended sample preparation time, additional laboratory equipment, and trained personnel. The application of this device is thus restricted if cost-effective measurements are needed, laboratory infrastructure is not available, or venipuncture is not feasible, e.g., in the case of point-of-care screening for iron deficiency under field conditions.²²

2.2.

Noninvasive Method

A method to measure the ZnPP/heme ratio independent of wthe background fluorescence is the dual-wavelength excitation method. This technique eliminates the autofluorescence background while retaining the porphyrin fluorescence emission.^1,2,28 Employing two laser diodes at 407 and 425 nm, it shows potential for field diagnosis while removing the need to wash the erythrocytes prior to ZnPP/heme ratio determination. In Fig. 2, the background-free ZnPP fluorescence signal measured from diluted whole blood is correlated with a reference HPLC measurement.

Fig. 2

Correlation of background-free ZnPP fluorescence ($y$ axis) to standard evaluation using high-performance liquid chromatography ($x$ axis).¹

In further investigations, the dual-wavelength excitation method¹ will be applied to noninvasive autofluorescence measurements to measure the faint erythrocyte ZnPP fluorescence noninvasively. The oral mucosa has been identified as a potential site to conduct these measurements, because the blood vessels are covered only by a thin, nonpigmented epithelial layer, such that light penetration of excitation light is not hindered. The capillary blood density is high, so that a sufficient amount of ZnPP fluorophores can be expected in the illuminated tissue volume. Still, background fluorescence is expected to be an even greater problem than for the whole-blood measurements. Among the main tissue fluorophores are collagen and elastin crosslinks,²⁹ whose fluorescence intensities are assumed to considerably exceed the ZnPP fluorescence signal remitted from tissue surfaces. Therefore, a method to efficiently reduce tissue background fluorescence also would be needed for successful noninvasive ZnPP/heme ratio quantitation. It was shown that the dual-wavelength excitation method eliminated, on average, 92% of the autofluorescence background for 20 subjects.²

In conclusion, these studies showed that the dual-wavelength excitation method successfully eliminates the autofluorescence background in whole blood while retaining the porphyrin fluorescence emission. Therefore, this approach allows for the construction of a simple, inexpensive point-of-care instrument quantifying the ZnPP/heme ratio from unwashed whole blood. For a future point-of-care instrument that quantifies the ZnPP/heme ratio noninvasively from the oral mucosa, dual-wavelength excitation can be used to largely eliminate the overwhelming tissue autofluorescence background to permit the quantitation of the faint ZnPP fluorescence signal.

3.1.

Fluorescence-Guided Resection

Several malignant tissues synthesize increased amounts of endogenous porphyrins after exposure to 5-aminolevulinic acid (5-ALA). It has been shown that C6 glioma cells, as a model for human malignant glioma, similarly synthesize porphyrins when exposed to 5-ALA and that selective synthesis occurs when C6 cells are inoculated into rat brains to form a tumor.³⁶ The kinetics of porphyrin fluorescence intensities in cultured C6 cells was investigated by flow cytometry. According to these in vitro and in vivo experiments, after exposure to 5-ALA, cultured C6 cells show a linear increase of PPIX fluorescence, which begins to plateau after 85 min. Marked fluorescence is also observed in solid and infiltrating experimental tumor. However, faint fluorescence also occurs in normal tissue. Based on these encouraging investigations, first, clinical applications could be envisioned, and subsequently, the benefit of fluorescent porphyrins that accumulate in malignant tissue after administration of a precursor (5-ALA) for labeling of malignant gliomas in patients could be confirmed.³⁷ For doing this intraoperatively, available clinical techniques and equipment from urological fluorescence diagnosis were adapted and transferred.^38,39 Hence, red porphyrin fluorescence was observed with a 455 nm long-pass filter upon excitation with violet-blue (375 to 440 nm) xenon light, and also quantitatively assessed by analysis of fluorescence spectra.⁴⁰ Fluorescing and nonfluorescing samples taken from the tumor perimeters were examined histologically. Normal brain tissue revealed no porphyrin fluorescence, whereas tumor tissue was distinguished by bright red fluorescence. For a total of 89 tissue biopsies, the sensitivity was 85% and the specificity was 100% for the detection of malignant tissue. For seven of nine patients, visible porphyrin fluorescence led to further resection of the tumor. Photobleaching caused a decay of the fluorescence intensity to 36% in 25 min during violet-blue light excitation and in 87 min during white light exposure. These observations suggested that 5-ALA induced porphyrin fluorescence may label malignant gliomas safely and accurately enough to enhance the completeness of tumor removal. Concurrent developments of neurosurgery-specific optical devices, aimed at improving such fluorescence-guided microsurgical resections of malignant gliomas using surgical microscopes, finally enabled uncomplicated and rapid recognition of the red tumor fluorescence and its borders to normal tissue, without interrupting the course of the surgery. Such systems appeared to constitute a useful tool for optimizing removal of malignant gliomas on a routine basis.^41,42 Hence, prospective clinical trials involving the fluorescence-guided resection (FGR) technique based on 5-ALA induced PPIX fluorescence were started.^37,43,44 This technique has meanwhile been evaluated in multicenter clinical trials, and it is nowadays established in a variety of neurosurgery hospitals.⁴³ So far, 5-ALA based FGR in neurosurgery is approved in Australia, Hong Kong, Israel, Taiwan, South Korea, and Japan.

However, even when employing FGR based on 5-ALA induced PPIX, which had proven to exhibit excellent sensitivity and specificity and to considerably facilitate “gross total tumor resections”^45–48 as shown in Fig. 3, one cannot expect the surgery to be curative, due to the infiltrative nature of the tumor growth.³⁷

Fig. 3

(a) Red fluorescence of 5-aminolevulinic acid induced protoporphyrin-IX during fluorescence-guided glioblastoma multiforme resection serves for contrast enhancement and allows demarcation of residual tumor tissue with millimeter resolution, thus offering surgeons a precise guidance during resection.⁴⁸ (b) Intraoperative white light view of the same situation.

3.2.

Photodynamic Therapy

Apart from the specific induction of tissue fluorescence, fluorophores such as PPIX may also cause the tissue to be photosensitized. PDT relies on the accumulation of significant amounts of such photosensitizing agents in the diseased tissue, which in combination with properly designed light exposure leads to phototoxic effects in the treated tissue. PDT is increasingly being used amongst health practitioners in combating a variety of diseases.⁴⁹ In the field of 5-ALA based PDT, a variety of clinical approaches are either under investigation or in clinical trials, which include the areas of dermatology, urology, brain, otorhinolaryngology, gynecology, and gastroenterology. In the following, the translation of basic scientific investigations to clinical application is sketched for the case of neurosurgery.

In vitro and in vivo investigations showed the potential of 5-ALA induced PPIX as photosensitizer for PDT in C6 glioma cells.^50–52 This suggests that the PPIX content in tumor tissue observed during FGR could also be exploited for PDT treatment of glioblastoma, both by surface irradiation of the surgical cavity and/or by stereotactically guided interstitial irradiation. These treatment modalities could be particularly helpful when clinical tumor removal by FGR had to be finished prematurely even though not all red or faintly fluorescing areas had been resected, e.g., because the affected tissue regions were part of eloquent areas. Intraoperative fluorescence spectroscopy showed higher sensitizer concentration in vital brain tumor versus the infiltration zone and in the infiltration zone versus adjacent normal brain, which contained very little PPIX.⁵³ Obviously tumor cells in the infiltration zone can be reached by PDT, which would otherwise be left behind untreated.

While PPIX based PDT is well-established for the treatment of actinic keratosis and basal cell carcinoma,⁵⁴ there are only occasional clinical reports about its application for GBM treatment.⁵⁵ Even in these cases, the photosensitizer Photofrin® has mostly been used in addition, obviously because the surgeons were not trusting in a sufficient phototoxic potential of the accumulated PPIX from 5-ALA alone. A variety of different photosensitizers have meanwhile been investigated for intracranial PDT⁵⁶, ⁵⁷including metronomic PDT.^58,59 Published clinical experience with PDT for GBM treatment relying solely on 5-ALA induced PPIX is limited.^60–64 In these trials and individual treatment attempts, PDT was applied by interstitial placement of radial diffusers while relying on the same or only slightly increased 5-ALA dosage as usually used for FGR.

Successful PDT requires homogeneous irradiation and light detection for dosimetry purposes. Irradiation devices for focal PDT of the brain cavity after FGR of the tumor tissue had been developed,⁶⁵ and the accumulation of PPIX in the brain tumor and adjacent tissue had been investigated to improve the PDT effect⁶¹ before interstitial PDT (iPDT) could be applied clinically for the first time.⁶⁶

Limited knowledge about the light, temperature, and photosensitizer distribution within the target volume initially hampered the clinical application of iPDT of gliomas. Monte Carlo (MC) simulations of fluence rate and heat transport resulted in an improved three-dimensional (3-D) treatment planning, which allowed to assess and define the treatment volume more accurately and to optimize the position of the light diffusers within the lesion.⁶⁰ Optical needle endoscopy was implemented for safe and precise stereotactically guided biopsy sampling in neurosurgery, which may also provide an innovative means to further optimize and individualize the iPDT treatment in the future.⁶⁷

Overall, stereotactic iPDT in combination with treatment planning could be shown to be a safe and feasible treatment modality.⁶⁶ These single-case treatments were extended to also include on-line monitoring of PPIX fluorescence and photobleaching kinetics, which seems important as dramatically different PPIX concentration levels and photobleaching kinetics have been observed. Such data were assessed and analyzed in order to employ them for real-time treatment monitoring and as early prognostic markers for the PDT response of individual patients. With regards to the PPIX concentration, it could be shown that necrotic regions typically located in the center of a GBM tumor are characterized by significantly lower PPIX levels than the outer regions consisting of vital tumor tissue. As indicated by this example, the implementation of fluorescence spectroscopy during iPDT could become a promising tool for individualized treatment concepts.^61,68,69 The evaluation of such spectroscopic data obtained from interfiber measurements of fluorescence and transmission during clinical stereotactic iPDT showed that the intratumoral PPIX concentration in glioblastoma exhibits pronounced inter- and intratumoral variations, which are directly linked to likewise variable levels of fluorescence intensity.⁶⁴ A high intratumoral PPIX concentration, associated with strong fluorescence intensity and complete photobleaching in the course of an iPDT treatment, also seems to be associated with a favorable treatment outcome. A typical intraoperative situation during an iPDT treatment with real-time monitoring of PPIX fluorescence intensity and photobleaching is shown in Fig. 4. The monitoring procedure turned out to be feasible and might be suitable for early treatment prognosis of iPDT. Furthermore, an individualization of treatment strategy and treatment parameters based on this information appears to bear a potential to further improve the clinical outcomes.^64,70 Improving all these techniques and the interaction between highly motivated partners may improve the clinical situation for treating GBM in neurosurgery for the benefit of the patients to prolong symptom-free survival with the highest degree of quality of life.⁶³

Fig. 4

Clinical stereotactic interstitial photodynamic therapy using several fibers ending in cylindrical diffusors for therapeutic light application (e.g., 635 nm). The system is capable of interfiber detection of fluorescence (e.g., 705 nm) and transmission (e.g., 635 nm) for on-line monitoring during treatment.^64,70

4.1.

Challenge in the Clinical Management of Breast Cancer

BC remains the most common oncological disease for women in North America⁷¹ and worldwide.^72,73 The challenges BC presents for health care systems and the affected individuals are different in high-income countries where BC screening and therapy are well-established compared to low- and middle-income countries where, particularly for the latter, mammographic screening remains a bottleneck leaving women often nondiagnosed.

In high-income countries, the combination of high participation in mammography screening programs⁷² in combination with advanced therapeutic options have led to a high five-year survival of $>90\%$, for BC patients.⁷¹ In particular, the advanced treatment options comprising surgery, chemotherapy (prior to surgery and postsurgery), and radiation therapy resulted in statistically equal five-year survival times for a nonscreened and a mammographic screened population. This opens the doors to different interpretation concerning the need for mammographic screening. While lowering screening compliance will result in more late-state tumors, it will also reduce overdiagnosis, thus reducing stress and unnecessary secondary testing, including invasive biopsies in false-positive women. Conversely, the treatment of late-stage BC will increase costs $\sim 20$ to 30 times⁷⁴ more than that of stage I/II BC. Additionally, survival statistics beyond five years are not available with adequate repeats to generate final recommendations about BC screening’s efficacy. Conversely, while 40- to 50-year-old women have an overall low incidence of BC,⁷² their incidence rates are increasing at the highest rate particularly in countries undergoing a lifestyle change, which is well-documented in South Korea⁷⁵ and also seen in countries such as Mexico and Egypt, although with less-solid data. It is in particular this age group that benefits of the most from early detection of BC as they would face potentially the most life years lost. Hence, improving the selection of women entering a BC screening program and adjusting the screening frequency based on a personalized risk assessment will lead to a better utilization of available screening resources in low- and middle-income countries and hence enable detection of predominantly early-stage BC, thus simultaneously reducing the overall cost burden to these health care systems.

Photonics-based tools for BC detection are still required particularly for premenopausal women and women with high mammographic tissue density as dense glandular and connective tissue hinders the detection of small lesions in the breast. While high-risk BrCa I/II gene mutation carrier are typically imaged every six months, alternatingly with MRI or ultrasound (US), other high-risk women, particularly those with a strong family history of BC or on long-term immune suppression therapy for an unrelated disease, are not given the same considerations. Nonionizing low-cost imaging-based screening is highly desirable for this population. The requirements for these imaging modalities are providing high contrast between glandular and, in particular, malignant tissue, being mostly independent of connective and fatty tissue. Low cost is desirable for implementation of this technology as standard BC screening technology, particularly in higher-mid- to lower-high-income countries with pending implementation of a national BC screening program, is an urgent task as these countries also currently have the largest gains in life expectancy and, hence, overall increase in BC incidence. Here again if the situation from Korea and Japan repeats itself, the increase will be disproportional in women $<50\text{years}$ of age who are commonly not captured by x-ray based BC screening. It is noteworthy that this is an economically limited environment, although it encompasses up to $1/4$ of the female world population. The needs of these women pertaining to BC screening and early detection are unlikely to be met with current x-ray based technology including tomosynthesis as this is aimed primarily on the highest-income countries; see Table 1 (modified from Ref. 76).

Table 1

Breast cancer screening programs in 26 ICSN countries in 2012.

A field where an adoption of photonics-based imaging tools is anticipated and highly likely is for neoadjuvant chemotherapy outcome prediction, where the therapy is limited to a very short time span due to the commonly advanced nature of the breast-invading tumor. The physician needs confirmation that the chosen chemotherapeutics are effective in shrinking the tumor volume or affecting its metabolism already after one or two cycles. As spatial resolution is secondary and the overall tumor response is desired, a low spatial resolution technique yet nevertheless with high contrast to changing oxygen consumption or vascularity can suffice.

Similarly to diagnostic technologies, therapeutic approaches show a significant qualitative difference between high- and middle-income countries on the one hand and low-income countries on the other hand. It is particularly evident in the latter group where the changing population age-pyramid coincides with rapidly changing environmental exposure. Middle-income countries try to emulate high income countries in their approach to treat advanced BC, which comprises neoadjuvant chemotherapy, surgery, intensity modulated radiation therapy, and chemotherapy with tyrosine or aromatase inhibitors. These treatment approaches pose a tremendous strain on the health care systems of middle-income countries, and they are generally not affordable for low-income countries. Hence, the majority of the women in low-income countries are not offered therapy, which is often further enhanced by a stigma with which these women are associated due to a BC diagnosis.

Hence, there are plenty of opportunities for novel enabling technologies to fundamentally change the clinical management of BC in high-income countries as well as low- and middle-income countries, so the technologies introduced into these markets will be different.

4.2.

Optical Technologies Aimed at Improving BC Risk Assessment, Diagnostics, and Response Prediction

Various demographic- and lifestyle-based BC risk assessment tools have been developed, such as the Gail breast cancer risk assessment model or familiar risk models,^77,78 and shown to be of utility for some screening decision making. However, their predictive power or odds ratio hovers below 2, and hence, they are not of utility to adjust entry and frequency of standard screening programs for the entire female population. An additional significant impediment of these risk assessment techniques is that some required predictors are not available until women are of standard mammographic screening age, missing the population of the sub-40- to 50-year-old women.

Risk assessment based on physical risk factors, analogous to blood pressure measurements for cardiovascular, cancer, and other diseases, does not face these limitations. To this effect, there is ample research showing that mammographic breast density (MBD) is one of the strongest risk factors, reaching odds ratios of up to 6 for two-dimensional mammographic projections when evaluating the top 25th percentile versus the lowest 10th percentile.⁷⁹ For 3-D assessment of the mammographic density, odds ratios up to 10 are anticipated, which suggest a causal relationship between MBD and BC.⁸⁰ When considering personalized oncology, the use of MBD has not progressed beyond recommendations to reduce the screening frequency for women with low MBD, so MBD is not suitable as a criterion for entry into a standard screening program. MBD cannot overcome the limited access to BC screening programs in low- and middle-income countries and as such does not overcome one of the major bottlenecks in BC management worldwide.

The risk assessment for a prescreening technology to be of utility in BC management in high-income countries needs to provide extremely high sensitivity comparable to mammography for the equivalent MBD and good specificity, as this is only a prescreening technology and the screening entry and frequency are to be personalized. In settings with limited access to standard screening technology, the main obligations are to identify the appropriate fraction of the population to advance to mammographic screening, thus optimizing the infrastructures for the women at highest risk. If the instrument is affordable, it can be shown that such a program is financially self-sustaining. For example, reducing the stage III/IV tumor incidence in countries with population of $>100$ million from the mid-30% to $<15\%$, entirely attainable when the available infrastructure approximates the OECD guidelines of 12 mammographs per 1 million BC cases, treatment costs can be reduced by over 300 to 500 million USD/year depending on the actual population size and the female life expectancy. Optical prescreening (Fig. 5) would be built on the intrinsic breast tissue optical properties using either absorption⁸¹ or fluorescence properties⁸² or both,⁸³ or on optical coherence spectroscopy.⁸⁴

Fig. 5

(a) Principal optical components for transmission measurements (left) and setup with four quadrants identified on the right breast (right). The small insert shows the tissue during measurements. (b) Optical density (OD) per physical optode separation (or relative optical density) in units of 1/cm for women with high (left) and low (right) mammographic breast density (MBD), respectively. (c) Chemometrix (principal component analysis) score based clustering of high MBD and low MBD.

Here, in particular, the work of the Turino^85,86 and Toronto^81,87–89 groups have demonstrated the ability to identify women with known physical risk factors, such as MBD^79,90,91 or biological event linked to the development of breast cancer, such as glandular atrophy during permenopause and postmenopause.⁹² While it is not clear at this time what fraction of the BC can be attributed to these different risk factors, however, the differences in the breast composition are accessible optically. Contrast is provided by the wavelength-dependent absorption spectra of predominantly water, lipid, hemoglobins, and collagen as well as by changes in the light scattering coefficients. The risk to develop BC does not correlate with particular tissue substructures, and hence spatial resolution is not required. The Toronto group favors steady-state spectrally resolved diffuse reflectance measurements, using either only chemometrix analysis, such as principal component analysis, for disease classification^93,94 or quantitative tissue chromophore extraction building on prior work by Farrell et al.⁹⁵ to first extract the spectral absorption and scattering coefficient followed by least-square fitting to derive the chromophore concentrations. The Milano group favored the use of time domain measurements to extract chromophore concentrations, allowing an analytical determination of the spectrally resolved light transport parameters preferable in establishing relationships of optical properties and known BC risk factors and future incidence of BC.

The ability to identify women with high MBD has been demonstrated for a screening aged population with sensitivity and specificity $>0.9$ when combining with menopausal status and body mass index by the Toronto group^96–98 and a $p<0.0001$ to differentiate a BIRADS 4 scores indicative for high MBD versus lower BIRADS scores by the Milano group.^85,86 Strong correlations have also been demonstrated for other risk factors, such as parity, age, and menopausal status.^87,89

For immediate application in low- and middle-income countries, robust and low-cost solutions, possibly based on cw measurements, are preferable as long as they satisfy the sensitivity and specificity requirements in identifying the subpopulation at highest risk to benefit from the screening infrastructure. High sensitivity and specificity $>0.8$, sufficient to be subsequently screened by butting national screening programs, can be achieved only using relative spectra shape, and thus, absolute instrument calibration is not required, facilitating the use of these instruments in resource-limited environments.

Optical tools are also being investigated to predict treatment response, in particular, for neoadjuvant chemotherapy, during the initial work therapeutic session to determine if a particular chosen treatment regime has the desired effect toward shrinking of the tumor to render it amendable to therapy. In particular, frequency domain spectroscopic scanning or tomographic approaches are being developed by various groups.^99–107 However, frequency domain optical tomography has demonstrated only limited utility as a screening tool due to its limited spatial resolution,¹⁰⁸ even when using a large number of source and detector pairs. Using added information, such as spatial information about the distribution of fatty versus glandular tissue from clinical imaging and knowledge of the chromophores’ absorption spectra demonstrated,^109–111 does not significantly improve the resulting spatial resolution in order to detect early-stage cancers at a rate comparable to current clinical imaging technologies even in premenopausal women. While some of the technology is comparable to that described previously for risk assessment, the number of source-detector pairs is commonly larger to achieve some spatial localization of the contrast. As vascular normalization is a primary goal of neoadjuvant chemotherapy, pruning of the aberrant vascular tree will modify the total hemoglobin concentration as well as the oxygen saturation in the affected tissue volumes.^106,112 During initial clinical studies, it was demonstrated that correct response prediction was achieved and that the hemoglobin contrast to normal tissues exceeds a ratio of 2.

Photoacoustic imaging, as for example developed by the Twente group with their photoacoustic mammography,¹¹³ is promising as a screening alternative, particularly for premenopausal women where tumors are masked by high mammographic density as shown in a comparative study versus MRI.¹¹⁴ Contrast is relying on the angiogenesis associated with tumor development for the selective absorption contrast. As absorption is mostly independent of the structures, providing high MBD photoacoustic mammography is also applicable in premenopausal women. In a recent small clinical study, Kitai et al. demonstrated the ability to detect all cases of ductal carcinoma in situ and most tumors that underwent prior neoadjuvant chemotherapy.¹¹⁵

The lower technical complexity of photoacoustic imaging over diffuse optical tomography makes it possibly the preferred technology independent of the available health care resources.

To summarize opportunities in the management of BC, the field of risk assessment or prescreening has significant potential particularly for younger women at risk. Here photonics-based diagnostics may complement US- or MRI-based assessment and/or preselection of women at risk of developing or harbouring BC in a resource-limited environment. Photoacoustic imaging can also become a valuable tool for BC detection, whereby monitoring of neoadjuvant chemotherapy by diffuse optical tomography should be considered whenever this therapy is offered.

4.3.

Photonics-Based Therapeutic Solution

As mentioned previously, there is indirect evidence that the current decrease in BC-related mortality in high-income countries is predominantly due to improved therapeutic efficacy and the present move towards personalized cancer medicine. The number of targets for BC is constantly increasing ranging from tyrosine and protein kinase inhibitors, epigenetic regulations, and nanomedicine with several of these approaches being introduced particularly in high-income countries. However, independent of the various therapies offered to the patient, surgical removal of the primary tumor is the standard of care, and its efficacy is limited by the need to demonstrate tumor-free resection margins. In a simplification to the use of FGR of brain tumors described previously, in BC, significant wider resection margins are acceptable,¹¹⁶ reducing the need for quantitative assessment of fluorescence. In general, the aim is a move towards near-infrared fluorescence in order to capture nests of infiltrating tumors several millimeters below the resection cavity surface.^117,118 Clinical trials for indocyanine green (ICG) fluorescence are ongoing, so primary data have not been published¹¹⁹ to date.

A second clinical application for fluorescence guidance is the intraoperative detection of sentinel lymph nodes using ICG as contrast medium^120,121 with or without active targeting, or intensely staining blue dyes. The first published multicenter clinical trial¹²² demonstrated an equal detection ability compared to radiolabeling or blue dyes. This was also confirmed by a recent meta-analysis¹²³ suggesting equal performance between fluorescence and radiolabel detection.

An alternative to surgical removal of the primary tumor was evaluated using either PDT^124,125 or photothermal applications, such as with interstitial laser photocoagulation^32,126–128 and interstitial laser hyperthermia.^129,130 Particularly the photothermal ablation models are currently not being researched as it becomes increasingly evident that complete surgical resection or ablation of the primary tumor will lead to high five-year survival rates, but without an immune effect introduced, a long-term survival is not guaranteed.¹³¹ More recent research is focusing on the therapy of metastatic BC particularly with spinal and bone involvement.¹³²

An interesting photon generation solution was proposed by Batista and Liang using solar irradiation, which could potentially have utility in extreme resource-limited environments when the primary tumor is to be destroyed in situ.¹³³

In summary, particularly supporting surgical lumpectomy by fluorescence-guided resection and detection of tumor infiltrated lymph nodes currently appear to be the most promising avenues for photonics solutions in the management of BC. Removal of the primary tumor is currently still best achieved with surgical resection followed by the various chemotherapies and radiation therapies aimed at treating the remaining micro metastasis and preferably also inducing the desired immune response.

5.1.

Endonasal Laser Treatment

After topical anesthesia [e.g., 4% tetracaine and 0.5% xylometazoline solution ($1:1$), 10 to 15 min) photo- or video-documentation via a rigid endoscope should first be performed. Prior to introduction of the laser light application system, laser safety precautions are mandatory. Conveniently, laser light should be applied in noncontact mode using a flexible silica bare fiber (core diameter: 400 to $600\mu \mathrm{m}$) guided via a device for precise endonasal fiber guidance.¹⁴⁰ Laser parameters setting need to be adjusted with respect to the laser emission wavelength [e.g., 8 to 12 W for 940 nm,^141,142 4 to 5 W for 1470 nm,^141–143 2 to 4 W for 1940 nm.¹⁴⁴ So far, diode lasers emitting at 900 to 1000 nm are in clinical use. Maneuvers for energy application itself should be performed via guiding the fiber from the posterior to the anterior free edge of the inferior turbinate under endoscopic control until adequate blanching of the tissue is obtained as judged by the operating surgeon. In cases where the head of the inferior turbinate appeared to be especially prominent, only some single laser spots were directed onto the head of the turbinate. Postoperatively, nasal cavities were treated with antibiotic and steroid-containing ointment (e.g., Jellin-Neomycin®: 0.25 g fluocinolone acetonide / 4.25 g neomycin sulfate). Patients received prescriptions for nasal ointments and nasal decongestants.

A typical outpatient laser-assisted inferior turbinate reduction of the hyperplastic inferior turbinate using a Tm:fiber laser emitting at 1940 nm at 3 W using a fiber guidance system is shown in Fig. 6, prior to, immediately after, and two months after treatment.¹⁴⁴

Fig. 6

Outpatient therapy for laser-assisted inferior turbinate reduction of a hyperplastic inferior turbinate using a Tm:fiber laser emitting at 1940 nm at 3 W using a fiber guidance system.¹⁴⁴ (a) Endonasal situation prior to laser energy application without decongestion showing the turbinate hindering continuous airflow, (b) immediately after laser energy application coagulation could be observed, and (c) two months after laser treatment the reduction of the turbinate is obvious (without decongestion) and continuous airflow is possible.

5.2.

Clinical outcome of Endonasal Laser Treatment Studies

Investigations and clinical trials show the safety and efficacy of laser treatment for volume reduction of hyperplastic turbinates in single cases as well as in prospective, randomized, and blind studies. The macroscopically visible tissue effect depends on the wavelength used. Observations by the operating surgeon on the basis of tissue whitening and tissue reduction confirm that 1940 and 1470 nm irradiation was about equivalent to the effects of the commonly used 940 nm laser system, yet it required a reduced irradiance, a significantly shorter treatment time, and less total energy.^141–143 These treatment procedures could be performed as an outpatient procedure under local anesthesia, and therefore, the patient acceptance and satisfaction were exceptionally high. The overall pain sensation was very moderate, with a trend towards less intraoperative pain using the longer wavelengths.¹⁴¹ Neither minor nor major complications could be witnessed in these studies during operation as well as postoperatively.^145–147 The main symptoms due to hyperplastic inferior turbinate (nasal congestion and nasal obstruction during exertion) could be significantly improved. Correspondingly, a significant symptom improvement was also shown in the validated assessment tool SNOT 20 GAV (“need to blow nose”).^141,144

Especially the long-term outcome seems to be the critical issue with the laser treatments of the turbinates.^{139,148–153} Moreover, there is currently no clear consensus or gold standard in the literature indicating the most optimal technique for turbinate reduction.^{139,154–157} This lack of a gold standard renders an appropriate evaluation or a comparison of novel techniques challenging. Nevertheless, laser-assisted and radiofrequency-assisted reduction of hyperplastic turbinates seem to be standing out as methods that can be applied under local anesthesia providing minimal morbidity.¹³⁹ For these reasons, the 940 nm diode laser was used for more than a decade for this indication and is regarded as the standard laser application.^{139,145,146,152}

With regard to lowering the applied irradiation, medical devices emitting at 1940 nm should be preferred, but these are rarely available in hospitals. As the 1470 nm diode laser systems are more widespread, these systems offer a highly efficient alternative to conventional diode laser systems in treatment of nasal obstruction due to hyperplastic nasal turbinates. In therapy-refractory rhinitis medicamentosa, outpatient diode laser inferior turbinate reduction of hyperplastic inferior turbinate represents a highly effective, safe, and well-tolerated treatment option that provides long-lasting recovery by markedly improving nasal airflow and stopping addiction to nasal decongestants.¹⁵⁸ It had also been shown that rhinomanometry with topical decongestion has a high predictive value for the objective outcome of laser-assisted turbinoplasty.¹⁵⁹

In conclusion, laser surgery of inferior turbinates can be performed as an outpatient procedure under local anesthesia. Due to a minimally invasive and controllable coagulation and ablation of soft tissue, almost no complications or bleedings were observed during the operation or postoperatively. Depending on the chosen parameters (power, energy) and the application modalities, laser treatment of hyperplastic inferior nasal turbinates achieved comparable or better results than most of the conventional techniques for turbinate surgery, like conchotomy, electrocautery, cryotherapy, chemical cauterization, and vidian neurectomy. Laser treatment can be considered a useful, cost-effective, and time-saving procedure for the reduction of hyperplastic inferior nasal turbinates. Short operation time, good clinical outcome, and minor side effects compared to other surgical methods provide an excellent clinical response of the patients.

6.1.

Endovenous Light Application

The endovenous laser treatment relies on the transformation of luminous energy into heat due to absorption. This process depends on the wavelength-dependent optical properties of the tissue and can be investigated by MC simulations.^182–185 Thus, the endoluminal application of laser energy implies the necessity of controlling a variety of parameters all together influencing the alteration produced on the vein wall. Variations in the laser wavelength, power settings, and irradiance result in different temperature levels and thermal alterations up to perforation.^{182,183,186–188} As blood is the primary medium around the laser fiber tip, it influences the mechanism and the alteration process as well, especially in cases where carbonization is induced.^{162,164,181,189}

Initially, laser energy was applied by using bare fibers emitting coaxially in the vessel lumen. The approach was the development of a specific radial emitting fiber to deliver the energy in direction to the vessels wall.^{167,168,176,190} In dependency of the used wavelength, the transmission through the existing thin blood layer around the fiber tip differs. As shown in Fig. 8, irradiation pattern showed maximum intensity deflected in an angle of $\sim 70\mathrm{deg}$ without any axial irradiance transmission. The measured transmission efficiency of such device was 94 to 97%. In comparison to the bare fiber technique, the irradiance (if contact to the tissue is assumed) can be reduced by a factor of 7 to reach irradiance values just below the ablation threshold of tissue.^168,176

Fig. 8

Radially emitting fiber and its irradiation characteristics.

In tests of the radial fiber technique on an ex vivo vein model¹⁶⁷ using heparinised blood containing veins, a shrinkage in length, a thickening of the wall, and the increased rigidity assessed by digital inspection could be achieved perfectly without any perforation.^{167,168,176,190} Additional investigation of the wavelength dependency of this treatment also showed that using a laser emitting at 980 nm an output power of ${\mathrm{P}}_{980}=(20\pm 2)\mathrm{W}$ is needed to achieve the desired macroscopic tissue alteration. In contrast, for a 1470 nm emitting diode laser, an output power of ${\mathrm{P}}_{1470}=6$ to 8 W is only necessary to achieve the same macroscopic results. On inspection of the surgically opened lumen of the vein, charred blood clots could be observed in the case of 980 nm irradiation, whereas in case of 1470 nm irradiation, a clean white coagulated vein intima surface was observed.^167,168,190 Further investigation also showed wavelength-dependent discrepancies.^186,191 These effects are clearly related to the wavelength-dependent optical properties of vein tissue and blood¹⁸⁹ and were confirmed by MC simulations.^182–185

Heat induction ($T=85°\mathrm{C}$ for 30 s) of the vein tissue samples showed a swelling of the sample concomitant with shrinkage in length. Additionally, the reddish vein color changed to whitish color of denatured tissue. The feeling sensation changed from flexible, smooth, and elastic to rigid and “macaroni-al-dente-like.” Vein tensile experiments showed native veins are elastic and can be stretched with low tensile power up to rupture, while cooked veins are inelastic and high tension powers are necessary for rupture. Both factors may explain patients’ description of having stretch discomfort after ELT.^167,168,190

Technologies such as the 360 deg radial fiber in combination with 1470 nm laser light^168,192 look promising as a means to induce safe, reliable, and reproducible tissue alteration for the ELT. By means of these optimizations, ELT treatment is getting closer to the goal of standardizing an effective method for the treatment of varicose veins. In a variety of investigations, disadvantages of previous ELT application techniques could be shown.^168,187,188 The introduction of the more effective wavelength and the new radial procedure has been established in clinical use since 2009.^193,194 First clinical studies show a clinical benefit.^195–200 Today, long-term follow-ups confirm the persistent effectiveness and safe occlusion of the veins.²⁰¹ Based on the reduction of undesirable side effects and the accelerated convalescence, endovenous treatment methods became the treatment option of first choice for insufficient veins in some countries.^202,203 Despite these improvements, some minor effects like carbonization and adhesion of the fiber to the vessels wall could still be observed during the clinical procedure. Therefore, implementation of feedback technologies may further assist standardization of the procedure.

6.2.

On-Line Monitoring During ELT

Although endoluminal techniques are medically approved and the clinical outcome of endoluminal treatments are accepted by physicians, due to meta-analysis of a large cohort studies,^204–210 as well as by the patients there are still requests for further improvement as adjacent structures should be prevented from temperature-related changes. Currently, real-time monitoring of physical, physiological, and tissue conditions is not available. Feedback information may have the potential of controlling the treatment or supplying immediate hints for the success of the treatment.

Tissue effects that could be optically detected in principle are shrinkage of the lumen, white light remission of the vessel wall, autofluorescence of the vessels wall, and temperature on the tissue or at the fiber tip. Investigations were performed to develop systems for endoluminal in vivo on-line monitoring of such parameter, during laser energy application and within the irradiation field.^176,211

MC simulation of optical detection of the shrinkage effect looks promising only for small vessel calibres. Using the radially emitted light of wavelengths between 600 nm (pilot beam) and 1500 nm (therapeutic wavelength), which is reflected by the vessels wall and then detected by the same fiber, showed that the moving vessel wall can only be detected when the distance between the fiber and tissue is $<2\mathrm{mm}$ in case of being filled with pure water and $<1\mathrm{mm}$ in case of the presence of blood. As native and coagulated vein tissues differ in their optical properties, the white light remission spectrum changes its shape during the denaturation process accordingly. Unfortunately, in the presence of blood, white light remission measurements as well as in situ measurements of autofluorescence are challenging. Finally, a temperature measuring system based on the analysis of the temperature-dependent fluorescence of a ruby crystal is developed.²¹¹ This sensor can be manufactured such that it is inert and biocompatible. It was tested to be useable in a high electromagnetic field, such as within the laser light irradiation field. Temperatures ranging from 20 to 200°C could be measured with an accuracy of ±2°C. Clinically adapted ex vivo experiments in a blood filled vein showed accurate measurements when the sensor tip is positioned in the vein parallel to and directly within the radially-emitting therapeutic fiber.²¹¹

In conclusion, the suggested technical feedback improvements are not yet clinically available. As visual control of the immediate tissue effect, such as lumen shrinkage or vein wall thickening, extra corporal UA and photoacoustic techniques are suggested but used from the skin surface. Additionally, the control of the pullback velocity of the treatment fiber and the irradiation parameters may yield improved light dosimetry. Implementation of a local endoluminal temperature may yield an improved reliable and successful treatment for the benefit of the patient.