1.

Introduction

Malignant diseases are the second most common cause of death worldwide.¹ The incidence rate is expected to increase by 47% by 2040, reflecting both growth and aging of the population as well as changes in other risk factors associated with socioeconomic development.² In the combat of malignant diseases, there are presently three dominating and well-developed treatment modalities: surgery, chemotherapy, and radiation therapy, whereas other modalities such as immunotherapy^3,4 are receiving increased attention.

The main local treatment modality for solid tumors is surgical resection. If possible, the entire tumor with a safety margin is removed during such resection. It is, however, in many cases, impossible to ensure the removal of all malignant cells. In these cases, the surgical resection is supplemented with either systemic or other localized treatments to decrease the probability of tumor recurrence. The most prominent systemic treatment modality for malignant diseases is chemotherapy.⁵ Chemotherapy relies on the administration of cytotoxic drugs containing DNA-damaging agents. The agents inhibit, with some selectivity, proliferation of malignant cells. Another frequently utilized localized treatment modality is radiation therapy (RT). Approximately 50% of all cancer patients receive RT during their course of illness.⁵ In RT, malignant cells are exposed to high-energy radiation, either from externally or internally localized radiation sources. DNA within the exposed cells will be damaged by the high-energy radiation, thereby blocking the ability of these cells to divide and proliferate. With extensive damage to its DNA, the cell eventually dies. Although this is desirable for malignant cells, RT can also significantly impact the surrounding healthy tissues in close proximity to the tumor. Proton- and other hadron-based therapies, rely on the presence of a rather localized Bragg peak for energy deposition and provide a higher damage selectivity, but they are expensive and not widely available.⁶

Consider, for example, the range of treatment strategies for patients presenting with breast cancer. Treatment plans are tailored for patients based on the expression of the estrogen receptors (ER) and progesterone receptors (PR), HER2 protein, the size of the breast tumor, and degree of lymph node involvement.⁷ The first line treatment in the majority of cases remains surgical resection. In particular, for non-metastatic breast cancer, surgery is always recommended to eradicate all tumor cells in addition to sampling and/or removing the axillary lymph nodes with the aim of avoiding metastatic recurrence. Surgical resection is often supplemented by either chemo or hormone therapy. This additional treatment is selected based on whether the ER and PR are positive or negative and may be administered preoperatively (neoadjuvant treatment) and/or postoperatively (adjuvant treatment). Neoadjuvant therapies are provided in an increasing number of locally advanced cases to reduce both the tumor mass and the risk of spread prior to surgery.⁸ Furthermore, RT may also be used to treat a portion of, or the entire, tumor-involved breast (after lumpectomy), the axillary lymph nodes (when their involvement is considered a risk) and the chest wall (after mastectomy). Despite major advances in treatment management over the past decades, refractory diseases and recurrence remain potential problems, while psychological and physical side effects present a significant burden to patients.

In the effort to find treatment options with less frequent and less severe side effects, a majority of the emerging treatment modalities assessed for malignant diseases today are more targeted, localized therapies with minimal damage to healthy cells.^9–12 By having a better selectivity to the malignant cells in these schemes, it is also possible to locally increase the treatment dose and thereby more efficiently eradicate a higher fraction of the malignant cells. We find interstitial photodynamic therapy (IPDT) (i.e., intratumor light delivery) to be a promising modality among these localized treatment alternatives under clinical evaluation for some malignant tumor types.

Photodynamic therapy (PDT) is a minimally invasive treatment modality of diseased tissue structures that uses light-activated photosensitizers for localized tissue eradication. Following the pioneering research of Dougherty, Malik, Moan, Pottier, Kennedy, and van den Bergh (i.e., Refs. 13–20) among others, it has been approved for a variety of superficial tumor indications, such as, for instance, non-melanoma skin cancers. Many examples of studies utilizing the particularly promising sensitizer precursor δ-amino levulinic acid (ALA) can be found in the literature.^21–24 For larger and/or deeply localized tumors, treatment can be provided via interstitial light illumination using fiber optics. A severe challenge in such a treatment is delivering a sufficient treatment dosage to all malignant cells, while sparing the healthy tissue in the vicinity of the tumor. It is worth mentioning that, even without sophisticated light dosimetry, IPDT is more selective to malignant cells than RT due to the selectivity of the photosensitizer. Moreover, light is strongly attenuated in tissue and is thus confined to a relatively small volume surrounding the fiber optic delivering the treatment light. This yields a localized treatment but also provides a challenge of reaching all malignant cells with a sufficient light dose. A remedy is clearly to use multiple fiber optical probes.²⁵ Much of the research to improve IPDT has also been directed toward developing targeted photosensitizers.^26–29 Tumor therapy based on IPDT is yet to be fully accepted, awaiting full implementation of refinements, e.g., dosimetry to allow its intrinsically advantageous features to be fully exploited. Returning to the particular case of breast cancer management, IPDT may offer an effective minimally invasive treatment alternative with minimal scarring, thus being an attractive option when considering the increasing concerns of patients with respect to cosmetic results following treatment of low-risk tumor. In addition to breast cancer, a disease with strict treatment options that follows very rigorous protocols, there are several other relevant indications in which IPDT could prove beneficial to the patients, e.g., primary and recurrent prostate cancer.

The purpose of this perspective paper is to explore the potential role of IPDT in the management of malignant tumors and discuss its ability to reduce complications by increasing the selectivity, precision, and safety in tumor eradication. Figure 1 illustrates a perspective of the framework of IPDT that will be discussed. Moreover, adding to previous reviews of the field,^30–36 this paper also aims to provide insights into potential future directions for the scientific community as a stimulus for innovation with particular attention focused on the challenges in further improving the IPDT efficacy and tumor selectivity. This could include improved light delivery, dosimetry, and instrument miniaturization for better adaption to the clinical workflow.

Fig. 1

IPDT framework comprising four stages: initial profiling, pretreatment planning, treatment delivery and real-time feedback, and posttreatment outcome assessment.

2.

Applications and Challenges for Interstitial Photodynamic Therapy

The potential of IPDT to treat various solid tumors has been investigated in an ongoing process and has demonstrated varying levels of success. In this section, we will describe some of the most promising clinical indications in which advances in IPDT can provide clear benefit in terms of decreased morbidity and increased quality of life.

3.

Integration of Dosimetry in IPDT

As illustrated in Fig. 1, dosimetry in IPDT is of key importance due to the need to ensure that the tumor is completely treated, while at the same time avoiding over-treating surrounding organs at risk. The large variability in the outcome of some clinical studies on IPDT may be due to the failure to properly address dosimetry aspects. Since the clinically approved photosensitizers have limited or no cancer tumor specificity, light dose control and dosimetry are important parts of clinical IPDT design. Practical dosimetry approaches always start with medical images as input to dose planning and optical fiber placement. The methods to navigate the optical fibers to the desired positions in the tissue are closely related to those used in interventional radiology.

Because of the variation in tissue properties between individuals, in different parts of the treated tissue, and over time during the course of the treatment, monitoring of tissue parameters is crucial. At the scale relevant to PDT dose (millimeters), light attenuation is determined by the local absorption coefficient ${\mu }_a$ (${\mathrm{m}}^{-1}$), scattering coefficient ${\mu }_s$ (${\mathrm{m}}^{-1}$), and dimensionless scattering anisotropy factor $g$. In the diffuse light propagation regime, it is common to use the reduced scattering coefficient ${\mu }_{s^{\prime }}={\mu }_s(1-g)$. Often, these coefficients are combined into a single parameter, the effective attenuation coefficient ${\mu }_{\mathrm{eff}}={[3{\mu }_a({\mu }_a+{\mu }_{s^{\prime }})]}^{1/2}$. Estimation of these optical properties is based on optical measurements, performed interstitially or at the boundary of the tissue region of interest. This may be done using the same optical fibers as used for the therapeutic light delivery^30,110 or using a different set of light sources and detectors.^111,112 The optical properties are estimated by means of an inverse method in which the measurement data are fitted to a light propagation model.

The optical fibers used to deliver and measure the light may be bare-end or have some kind of diffuser at the tip. Bare-end fibers have the advantage, in comparison with extended diffusers, of the evaluation of optical properties being easier since light sources and detectors can be treated as points. It also potentially provides higher spatial resolution. On the other hand, cylindrical diffusers allow for using fewer optical fibers to cover the same treatment volume, and it has been recently shown that cylindrical diffusers can be used to evaluate the optical properties.^113,114

The measurements can be done in continuous wave (CW) mode or using pulsed ($\sim \mathrm{ps}$ short pulses) or frequency-modulated ($\sim 100\mathrm{MHz}$) light. Generally, modulated measurements may provide more information, which can give more accurate predictions, but the technology is more complicated and costly. In particular, using modulated light, it is possible to separate the effect of local inhomogeneities close to the optical fibers inside the tissue from the optical properties of the bulk medium.¹¹⁵ With CW data only, it may be difficult to achieve a similar distinction. Also by modulation frequency tagging, all measurements can be done simultaneously instead of sequentially.

Other tissue parameters of importance for PDT dosimetry are the photosensitizer concentration, tissue oxygen saturation, and singlet oxygen production. The photosensitizer concentration can be estimated by optical methods based on fluorescence emission.^{71,110,116–121} Fluorescence excitation by short wavelengths ($\sim 400$ to 600 nm) may be difficult to achieve in the interstitial setting because of the very limited light penetration length. Fluorescence excitation at wavelengths above 600 nm, on the other hand, can be utilized for photosensitizer agents that exhibit absorption in this wavelength region. Some model-based methods directly determine the concentration of the photosensitizer in the tissue based on fluorescence, which is hard to achieve since it requires absolute-calibrated detection and knowledge of fluorescence quantum yield as well as quenching effects in the tissue, in addition to knowledge of the optical properties. Tomographic reconstruction of the relative fluorescence signal throughout a tissue volume has been performed⁷¹ and provides a scheme for estimating spatial distribution of the photosensitizer within the prostate. In addition, tissue oxygen saturation can be determined by spectral fitting methods in the near-infrared region, where the differences in the absorption spectra of deoxyhemoglobin and oxyhemoglobin are significant.^{110,119,122,123}

Several approaches to conduct dosimetry plans to guide IPDT have been proposed.^{48,70,110,112,124} More generally, dosimetry models can be categorized into explicit dosimetry, implicit dosimetry, direct dosimetry, and biological response, with their respective advantages and constraints described broadly in the literature.^31,125–128 Table 2 summarizes some of the recent dosimetry systems proposed for clinical use by various groups, whereas Fig. 2 presents examples of systems with pretreatment planning and online dosimetry. All dosimetry methods presented in that table are based on a light dose threshold model, specifically determined in these cases, and depend on the photosensitizer, drug dose and its formulation, drug–light interval, and clinical indication. Here we will use one set of algorithms to illustrate the considerations needed to develop IPDT treatment planning and dose control, referred to as “Interactive Dosimetry by Sequential Evaluation” (IDOSE).⁷⁰ In the first step, specific to prostate cancer PDT, US images are acquired by TRUS. Based on the images, a computer model is generated with the key tissue types: prostate, urethra, rectum, and sphincters. In the next step, the positions of the optical fibers are calculated using a random-search optimization algorithm, similar to simulated annealing-type algorithms. The algorithm uses a light propagation model based on the diffusion equation and assumes that the prostate tissue is homogeneous with average optical properties.

Table 2

Comparison of some approaches and systems proposed for IPDT dosimetry.

Fig. 2

(a) Six-fiber IPDT system with diagnostic and treatment capability: top panels present diagnostic mode (left) and treatment mode (right) of operation; bottom panels depict the diagrams of fiber arrangement for the diagnostic (left) and therapeutic mode (right). Reproduced with permission from Ref. 33, courtesy of SPIE and OSA. (b) IPDT setup with real-time spectroscopic monitoring for brain tumors. Inset shows a photograph of the clinical setting during IPDT procedure. Reproduced with permission from Ref. 48, courtesy of Wiley. (c) IPDT prostate treatment: guidance with TOOGUIDE software (left), dedicated laser generator (middle), and preoperative view of vascular-targeted photodynamic therapy in action (right). Reproduced with permission from Ref. 131, courtesy of Springer Nature (based on CC BY license).

When the fibers have been put in place, but before PDT light delivery starts, a series of measurements is performed to collect the value of the light attenuation between each mutual pair of fibers. These measurements are fed into an inverse algorithm that fits the data to a light propagation model based on the diffusion equation, with the aim of determining the local attenuation coefficient ${\mu }_{\mathrm{eff}}$ around each fiber. The approach is feasible since the same set of fibers can be used for diagnostics as well as treatment. As a matter of fact, this is also the case regarding oxygenation and sensitizer distribution assessment.^25,30

A second optimization algorithm is used to calculate the light dose to be delivered from each fiber. The algorithm is based on Cimmino’s method. Briefly, the dose given to each tissue voxel in the tissue model is calculated using a diffusion model, by summing the contribution from all fibers. The values of ${\mu }_{\mathrm{eff}}$ (for each voxel) from the previous step are used as input to the diffusion model. The emitted light dose from each fiber represents the optimization parameters. Each tissue type is assigned a dose threshold value, with the aim of reaching at least the threshold dose in the target region, while avoiding reaching the threshold dose in all other tissue types. Each tissue type can also be assigned a weight coefficient depending on its importance. This leads to a set of inequalities that can be solved by Cimmino’s method. The implementation uses the block-action method described by Censor et al.¹³² The method is iterative and converges toward a solution even if all constraints are not met.

During the PDT session, light delivery is interrupted at intervals to acquire new measurements, which are again evaluated by the inverse fitting algorithm. This results in updated values of ${\mu }_{\mathrm{eff}}$ in case changes have occurred in the tissue due to, for example, variations in blood flow. The ${\mu }_{\mathrm{eff}}$ values are used to calculate new light doses for each fiber, and an updated dose plan is presented to the user for approval. This cycle is repeated until all fibers have delivered their full dose. The Cimmino algorithm does not allow for straightforward implementation of constraints based on dose-volume histograms (DVHs), but DVHs are used to evaluate the performance of the dose planning. An example of a dose plan from a clinical case is shown in Fig. 3.

Fig. 3

Illustration of dose plan for focal prostate PDT, in which the right posterior side of the prostate is targeted using seven bare fibers as sources. The dose unit is $\mathrm{J}/{\mathrm{cm}}^2$. The upper chart shows the DVH with the dose threshold indicated at $20\mathrm{J}/{\mathrm{cm}}^2$. The images show the isodose curves overlaid on US images and segmentations of prostate (in red), urethra (in yellow), and rectum (in brown). The lower right chart shows the total light illumination time durations per fiber.

IPDT dosimetry is an essential tool in optimal treatment management, ensuring that the appropriate light dose is delivered without over- or under-treatment. Although biological processes governing the treatment response are complex and may seem challenging to control in real time, several approaches show great promise for predictable and reliable dosimetry protocols. In addition, for certain indications (e.g., palliative care in HNC), extremely precise dosimetry may not be necessary to fulfilling the clinical objective.

4.

Photosensitizers

Photosensitizers are key elements in the PDT approach to tumor management. The ideal photosensitizers must be chemically stable and systemically non-toxic, accumulate in high concentrations in the tumor tissue as compared with surrounding tissue, have high absorption at wavelengths sufficient for deep tissue penetration, have high ROS generation yield, and be cleared from the system rapidly after treatment.²⁹ Despite a huge research effort, to date only a relatively small number of photosensitizers have received approval. Photofrin, the first FDA-approved photosensitizers based on an HpD, paved the way for PDT as an alternative cancer treatment in clinics. This first-generation photosensitizer remains widely used in many countries in the treatment of various cancers, despite its low selectivity, limited absorption, and long-term photosensitivity.¹³³ However, this has led to the development of second-generation photosensitizers such as chlorins, bacteriochlorins, phthalocyanines, and other porphyrin derivatives in an effort to mitigate these drawbacks. Several second-generation photosensitizers have been approved in certain countries for the treatment of different cancer types such as ALA for skin and brain (EU and North America); temoporfin for HNC, bile duct, and lung (EU); padeliporfin potassium (Tookad) for prostate (EU and Mexico); and talaporfin sodium (Laserphyrin) for lung and brain (Japan), to name a few. The treatment outcomes obtained with these photosensitizers have demonstrated reduced side effects, improved selectivity to the target tissue, and deeper penetration depths thanks to longer excitation wavelengths.¹³⁴

Despite the improvements enabled by the second generation of photosensitizers, several challenges remain to be solved by the next generation of sensitizing agents. First and foremost, new strategies to further increase the selectivity to the tumor cells to improve efficacy and reduce systemic cytotoxicity should be pursued. Recently, emerging third-generation photosensitizers refer to modified second-generation photosensitizers that aim at targeting strategies such as antibody-conjugated photosensitizer binding to receptors over-expressed in tumor cells (active targeting) and photosensitizer-loaded nanocarriers (passive targeting), assisting in the delivery of photosensitizers to the tumor and increasing selectivity versus normal tissue. This should result in a lower photosensitizer dosage, increased allowed light intensity, and fewer unwanted side effects. Several clinically approved monoclonal antibodies have become an appealing option for targeted delivery of anticancer drugs owing to their high target specificity and affinity. The first antibody–photosensitizer conjugate (water-soluble silicon–phthalocyanine derivative IRDye700DX conjugated to Cetuximab) targeting epidermal growth factor receptors has received early conditional marketing approval in Japan for the treatment of advanced HNC¹³⁵ while conducting a global phase 3 multi-center clinical trial (NCT03769506). This delivery strategy is highly versatile as different antibodies can be used to target IR700 to other antigens, which has been demonstrated in preclinical models for breast, brain, prostate, and oral cancers. On the other hand, the fact that most effective photosensitizers tend to be insoluble means that encapsulation in nanodrug carriers could improve their performance. Different organic (e.g., lipids and peptides) and inorganic (e.g., metallic, silica, and quantum dot) NPs have proven successful in in vitro and in vivo models.¹³⁶ However, the mechanism by which NPs enter solid tumors is more complex than previously thought and the enhanced permeability and retention effect is significantly more pronounced in the small animal xenograft models compared with tumors growing in humans.¹³⁷

The second significant challenge, particularly in the case of solid tumors, is elevated intratumoral pressure and tumor hypoxia (oxygen pressure of $<10\mathrm{mm}\mathrm{Hg}$). Novel ways of overcoming heterogeneous drug uptake resulting from increased interstitial fluid pressure may be important for increasing the efficacy of PDT in certain cases.^138–140 The therapeutic efficacy of conventional PDT relies predominantly on the type II mechanism, which requires the presence of three inseparable elements: photosensitizer, light, and oxygen. However, in solid tumors, the already low oxygen concentration is further diminished by the PDT process, leading to low efficacy of the treatment. Therefore, for these applications, oxygen-independent PDT would be beneficial. It has been shown that intermittent light delivery (fractional PDT) allowing for replenishment of cellular oxygen and oxygen delivery strategies can improve treatment outcomes.^141,142 Recently, another strategy utilizing type I PDT, which activates free radicals, appeared to be a direct way of overcoming hypoxia limitations.^143,144

Development of novel photosensitizers is an active area of research, and the real clinical influence of newer NP-based systems is anticipated in the next decade. With many existing photosensitizers being evaluated in clinical trials and novel strategies to mitigate their potential drawbacks for applications in solid tumors, IPDT has the potential to gain wider clinical acceptance for the treatment of deep-seated tumors in the near future.

5.

Future Opportunities

The general perspective and potential of IPDT in the management of solid tumors will now be summarized through a subjective SWOT analysis in comparison with other local treatment options for well-selected clinical indications. The main strength of IPDT stems from its minimally invasive nature and the fact that it can prove to potentially be an effective treatment with fewer side effects. The treatment neither targets the nucleus nor the collagen network in the tissue, meaning that it is non-mutagenic and can therefore be repeated as many times as necessary without impairing the integrity of the tissue structure. The tissue also typically heals well after treatment as typically the collagen and lipid structures are not affected and hence scars are only minute. This offers an attractive solution for indications requiring a good cosmetic effect such as breast cancer or HNC. The relatively shallow light penetration can be seen as both a strength and a limitation. Light, and thereby the treatment response, will be confined to a rather small region in the vicinity of the treatment fiber tip, facilitating a selective treatment opportunity. This forced confinement proves useful when it comes to tumors in close vicinity of critical structures such as prostate or HNC. At the same time, the short light penetration is a challenge when treating large volumes, calling for the use of using multiple optical treatment fibers in such cases.

Weaknesses of the IPDT modality include both the elaborate treatment mechanism complicating its planning and optimization and potential side effects of the treatment, which while milder than those of other therapy modalities are still present. The dependence of the treatment outcome on many parameters can be viewed as both a complication and an opportunity as it suggests the need for sophisticated treatment optimization and planning. This implies also the requirement for specifically developed instrumentation^25,30 and expertise.^70,124 The major remaining side effects of IPDT involve possible pain during and after treatment, treatment response deviating from the plan due to undetected bleeding in connection with placing the optical fibers, and adverse reactions to the photosensitizer given, the latter clearly being a factor pertaining to all types of medication. The mitigation of these side effects is appropriate treatment protocols with suitable dose and type of photosensitizer and the use of multiple fibers to minimize the influence of any bleeding around the fiber tip. Repositioning of a fiber once bleeding is detected (and it can be readily detected in an integrated diagnostic/treatment IPDT system) can be employed to mitigate a reduced dose.

The opportunities, however, for IPDT are very high in handling unmet clinical needs. The complexity of the treatment mechanisms yields many opportunities to optimize and personalize the treatment via advanced dosimetry and targeting photosensitizers. This, together with the few and mild side effects, yields an attractive treatment option for a number of clinical applications as described above. Furthermore, the treatment also provides additional opportunities to utilize biomarkers indicative of PDT outcome when deciding the treatment strategy or monitoring and optimizing the outcome, which still remains a largely unexplored path. This can then aid in selecting IPDT when the modality is well suited and allow for individualisation of the treatment plan depending on prior information obtained from blood samples or tumor biopsies. For example, it has been suggested that ABCG2, FECH, and HO‐1 could serve as indicators of treatment outcome for ALA-PDT of brain tumors, while in HNC, STAT3 could be used as a molecular marker of cumulative photoreaction therapy monitoring.^77,145 In addition, the use of multiple fibers, when arranged for an integrated combination of treatment, and monitoring of important data such as light flux, oxygenation, and sensitizer distributions provides excellent opportunities for optimized treatments, as discussed.^25,30,70 Furthermore, a special feature of PDT and IPDT is their adaptability for a realistic and low-cost tumor treatment modality in countries with very low resources and lack of operating rooms and facilities for ionizing radiation.^80,146

What then may be a threat or impair a favorable utilization of IPDT for the management of solid malignant tumors? Some of the current challenges faced by IPDT are outlined in Table 3. It could possibly be that the approach would not achieve the expected convincing results regarding treatment outcomes in certain conducted studies. The reason for this may be that full control of important treatment parameters could not be achieved, in view of the fact that the modality, at least in certain cases, would require full control of light dose, sensitizer, and oxygen availability. Suboptimized treatment delivery would obviously result in suboptimal treatment outcomes. Clinical trials with equipment allowing for full control of relevant parameters without causing extra trouble, neither for the patient nor for the doctor, would then lead the way to improved results. It is important to correctly plan studies and translate the results from one study to another to help optimize a study for a different indication, based on prior experience. Clearly, the development of other novel treatment modalities with good outcomes might also challenge IPDT. Such techniques might possibly include immunomodulated therapy. One could then consider IPDT to be part of a combined treatment protocol for tumor debulking.

Table 3

Summary of the present challenges and needs of IPDT to be included in clinical guidelines for the treatment of certain indications.

6.

Conclusions

The potential clinical benefits of IPDT have been clearly demonstrated over the years. However, the modality is still at an early clinical stage with clear opportunities for further development. Treatment selectivity and the low number of adverse side effects present IPDT in a favorable light in comparison with standard modalities used to treat solid tumors. In our perspective, this treatment once fully exploited in clinical settings will extend its reach beyond indications investigated in current clinical trials. Following present approvals of first clinical indications and further considerations regarding optimized dosimetry and suitable photosensitizers fulfilling needs for each individual oncological case, IPDT will gain entry into the treatment guidelines and wider acceptance among the clinical community. Advances in photosensitizing agents, light delivery systems, and treatment planning schemes offer numerous opportunities for this technique to become a robust, standard first-line therapy, either alone or in combination with other treatments, for a variety of malignant diseases.