Lasers in Surgery and Medicine 41:612–621 (2009)

Ultrasound-Guided Photodynamic Therapy forDeep Seated Pathologies: Prospective Study

Waseem Jerjes, MSc (OMFS), PhD,1,2,3* Tahwinder Upile, FRCS,1,2 Zaid Hamdoon, MSc (OMFS),1,3

Farai Nhembe, MFDS,1 Rishi Bhandari, MFDS, MRCS,1 Sorcha Mackay, MFDS,1 Priya Shah, MFDS,1

Charles Alexander Mosse, MSc, PhD,4 Jocelyn AS Brookes, FRCR5 Simon Morley, FRCR,5 and

Colin Hopper, FDS, FRCS1,2,3

1UCLH Head and Neck Centre, London, United Kingdom2Department of Surgery, University College London Medical School, London, United Kingdom3Unit of Oral & Maxillofacial Surgery, Division of Maxillofacial, Diagnostic, Medical and Surgical Sciences, UCLEastman Dental Institute, London, United Kingdom4National Medical Laser Centre, London, United Kingdom5Department of Radiology, University College London Hospital, London, UK

Introduction: Interstitial photodynamic therapy remainsan attractive remedial option in minimally invasivesurgery. Our aim in this prospective study was to evaluatethe outcome following ultrasound-guided iPDT of deep-seated pathologies. Patients’ reports on quality of life withclinical and radiological evaluation were the main end pointparameters used to assess the outcome.Materials and Methods: Sixty-eight patients werereferred to the UCLH Head and Neck Centre for treatmentof various deep-seated pathologies involving the head andneck region, upper and lower limbs. All patients underwentinterstitial photodynamic therapy under general anaes-thesia, using 0.15 mg/kg mTHPC as the photosensitisingagent. Following treatment, patients were followed-up for amean of 7 months.Results: All three patients who presented with visualproblems reported improvement after treatment.Also, 14/17patients reported improvement of breathing. Improvementof swallowing was reported by 25/30 patients; while speak-ing improvement was evident in 16/22 patients and 33/40reported reduction in the disfigurement caused by theirpathology. All five patients with impeded limb functionreported some degree of improvement. Clinical assessmentshowed that half of the patients had ‘good response’ to thetreatment and a third reported ‘moderate response’ with twopatients being free of disease. Radiological assessmentcomparing imaging 6-week post-PDT to the baseline showedstable pathology with no change in size in 13 patients,minimal response in 18 patients, moderate response in23 patients and significant response in 11 patients.Conclusion: This study on 68 patients with deep-seatedpathologies undergoing interstitial photodynamic therapyprovided evidence that PDT can be the fourth modalityin the management of tissue disease. Lasers Surg. Med.41:612–621, 2009. � 2009 Wiley-Liss, Inc.

Key words: photodynamic therapy; ultrasound; headand neck; cancer treatment; tumour; vascular malforma-tion; hamartoma; interstitial

INTRODUCTION

Photodynamic therapy (PDT) remains an attractiveremedial option in oncology. In principle, light applicationto the target tissue leads to a non-thermal photochemicalreaction; this is usually initiated some hours following theadministration of the photosensitiser and leads to selectivedamage to the target tissue. The efficacy of the treatmentdepends on the type and the concentration of the photo-sensitiser, light dose, fluence rate, the availability ofoxygen and cellular localisation. Several photosensitisersare currently being tested for better specificity of targettissue [1–3].

Two generations of photosensitisers are currently beingused in oncological practice. Profimer sodium (Photofrin-first generation) is commonly used in Barrett’s high-gradedysplasia, cervical, gastric, oesophageal, endobronchialand papillary bladder cancers. While 5-aminolevulinic acid(5-ALA), a natural heme precursor, has been successfullyapplied in basal cell carcinoma, actinic keratosis and oraldysplasia. mTHPC (Foscan) a much more potent secondgeneration photosensitiser for cancer treatment, whencompared to Photofrin and ALA, is commonly used in headand neck cancer. Third generation photosensitisers (tinethyl etiopurpurin, mono-L-aspartyl chlorin e6, benzopor-phyrin derivative and lutetium texaphyrin) are already inclinical trials; initial results showed better tumour specif-icity and shorter generalised photosensitivity [1–3].

The response of lesions to PDT is critically based on thedosimetric profiles. It is advantageous to quantify thedistribution of the light fluence rate, the optical properties,the drug concentration and the tissue oxygenation forphotodynamic therapy (PDT). Zhu et al. developed anintegrated system to determine these quantities beforeand after PDT treatment using motorised probes. They

*Correspondence to: Waseem Jerjes, 256 Gray’s Inn Road,London WC1X 8LD, UK. E-mail: w.jerjes@ucl.ac.uk

Accepted 8 September 2009Published online 15 October 2009 in Wiley InterScience(www.interscience.wiley.com).DOI 10.1002/lsm.20853

� 2009 Wiley-Liss, Inc.

reported significant inter- and intra-prostatic variations inthe tissue optical properties and drug distribution, suggest-ing that a real-time dosimetry measurement and feedbacksystem for monitoring these values during treatmentshould be considered in future PDT studies [4]. This ismainly due to the complexity of PDT dosimetry due to thedynamic nature of the three essential components—light,photosensitiser and oxygen. Considerable progress hasbeen made in understanding the problem and in developinginstruments to measure all three, so that optimisation ofindividual PDT treatments is becoming a feasible target [5].

The delivery of light to initiate the photochemicalreaction depends on the tumour, size and location. Thetreatment of superficial pathologies is carried out throughsurface illumination. This is a very successful and con-trolled technique and the depth of treatment can reach upto 1 cm when using mTHPC as the photosensitiser. Thedifficulty arises when treating deep-seated pathologies.Special grids can be made and with preoperative imaging,needles can be inserted in the suspect tissue and fibres canbe fed through to deliver the light, then by using a ‘pull backtechnique of the needle and fibre’ all the treatment volume[1–3] is covered. Most recently, intra-operative imageguidance of the needles into the pathological mass hasenabled more accurate identification and illumination[3,6].

Recent study has suggested that 2D-ultrasound (2D-US)is a successful modality in guiding the deliver of light todeep-seated pathologies during interstitial photodynamictherapy (iPDT). The study by Jerjes et al. involved 23mTHPC US-iPDT treatments performed on 21 patientspresented with various pathologies. The results showedthat it was possible to clearly identify the needles (inc. thetip) during insertion in all 23 treatments and it was possibleto guide parallel needle insertions using ultrasound. Thestudy concluded that it was feasible to use 2D-US to guide‘real-time’ photodynamic therapy in deep-seated tumoursand other malformations and can augment the informationfrom other imaging modalities without affecting thepatient’s treatment outcome [7].

When the photosensitiser is initially activated, by light, itis expected that the photochemical reaction leads toapoptosis or direct tumour cell death, vascular shut-downand the subsequent inflammatory response. The two mostcommon side effects post-PDT include residual systemicphotosensitisation and pain. Severe skin burns have beenreported to occur up to 6 weeks postoperatively on directsun exposure. Most interestingly the injection site (intra-venous administration of the photosensitiser) is even moresensitive to sunlight and skin burn can occur if inappropri-ately covered. Minor concentrations of the photosensitiserin the skin may lead to oedema or even superficial skinnecrosis [1].

Pain following iPDT is generally mild to moderate inseverity. It commonly persisted for 2–4 weeks, declining inintensity as the healing (tissue regeneration) progresses.Short-term treatment for this is with non-steroidal anti-inflammatory drugs (NSAIDs) in addition to opiates; mostcommonly used are codeine phosphate, tramadol and

morphine; and for long-term and controlled release fen-tanyl and buprenorphine patches can be used. Thereporting of pain mirrored a temporary rise in white bloodcell levels, which followed the induction of tissue necrosisand the generation of acute and chronic inflammatoryresponses [1,3,7].

Advanced tumours and vascular malformations can betreated, (salvage or palliative), with the aim to reduce thetumour bulk to improve function, and also reduce bleedingand control pain. Difficulties arise when the tumour startsto metastasise through the lymphatic system. Photochem-ical reaction can be induced in single or multiple lymphnodes under image-guidance but this can be sometimesdifficult to perform when dealing with a complex lymphaticchain (i.e. the head and neck). Healing usually occurswith minimal scarring and the procedure can be repeatedwith little cumulative toxicity. There is sparing of tissuearchitecture, providing a matrix for regeneration of normaltissue.

In this prospective study, we evaluated the outcomefollowing ultrasound-guided iPDT (US-iPDT) of deep-seated pathologies. Patients’ reports on quality of life withclinical and radiological evaluation were the main param-eters used to assess the outcome.

Ethical Considerations

iPDT is commonly used in our department. All non-protocol applications were accompanied by multidiscipli-nary team recommendation, ethical approval and informedpatient consent. Many of the patients studied had end stagedisease and had few other treatment options.

MATERIALS AND METHODS

In this prospective study, 68 patients (34 males and 34females) were referred to the UCLH Head and Neck Centre,London for treatment of a range of deep-seated pathologiesrelated to the head and neck region, upper and lower limbs.The mean age was 50.7� 23.3 years (Min 0.5, Max 88) with88% being Caucasians. The most common primary sitesinvolved were the oropharyngeal region, face and limbs.Nearly half of the patients had carcinomas; other pathol-ogies include sarcomas and hamartomas (Table 1).

Disease staging was carried out and showed that nearlyall the malignant tumours were stage IV or of a high-gradenature; while all the hamartomatous malformations wereregional rather than local. Tumour volume was assessedusing preoperative magnetic resonance imaging andcomputed tomography, for some patients, when indicated.Patients (45.6%) were not otherwise suitable for conven-tional management, including palliative therapy, while theother 54.4% declined further treatment (Table 1).

The patients’ main complaints were categorised into sixparameters: visual, breathing, swallowing and speakingproblems, as well as disfigurement and limb dysfunction.Three patients reported visual problems due to eithercompression of the eye by external swelling caused by thepathology or tumour invading the eye socket causingproptosis. Problems associated with pathological growthin the oropahryngeal/laryngeal region caused breathing

ULTRASOUND-GUIDED PHOTODYNAMIC THERAPY 613

problems in 17 patients, swallowing problems in 30patients and speaking (voice) problems in 22 patients.Disfigurement was reported by 40 patients and impairedlimb function was found in 5 patients. As there is no currentverified assessment of quality of life in patients undergoingphotodynamic therapy the patients were asked only toreport the nature of the complaint and not the severity(Table 2).

All patients were discussed at the UCLH Head and NeckMDT and it was agreed that the only available option, for 31of them, was to offer interstitial photodynamic therapyunder general anaesthesia, using mTHPC (Foscan) as thephotosensitising agent. The other 37 patients refusedalternative treatments. mTHPC is usually administeredat a dose of (0.15 mg/kg) intravenously into the midcubitalvein 96 hours prior to treatment. This would allow theagent to accumulate in the pathological area which wouldincrease efficacy. Patients are usually kept in a side room(with a dim light) to avoid systemic photosensitisation.

Intra-operatively, an ultrasound (EMP 1100 with high-resolution) probe was used to examine the pathologicaltissue (centre and periphery). The main aim here was todetermine tumour volume, depth, invasion of vascularstructures, hollow organs or hard tissue. This was followedby insertion of 18 Guage 70 mm long spinal needles underUS-guidance into the pathological tissue. Great care was

TABLE 1. Demographics of Study Group

No. (%)

Gender

Male 34 (50.0)

Female 34 (50.0)

Age (years)

Mean 50.7

Minimum–maximum 0.5–88

St. deviation 23.3

Race

Caucasian 60 (88.1)

Indian 4 (5.9)

Mediterranean 2 (3.0)

Middle-Eastern 2 (3.0)

Primary site

Oral cavity 10 (14.7)

Oropharynx 21 (30.9)

Larynx 1 (1.5)

Lower resp. tract 1 (1.5)

Mid face 19 (28.0)

Lower face 3 (4.4)

Neck 2 (3.0)

Upper limb 8 (11.8)

Lower limb 2 (2.9)

Glans penis 1 (1.5)

Secondary site

None 31 (29.2)

Oral cavity 21 (19.8)

Oropharynx 1 (0.9)

Nasopharynx 1 (0.9)

Hypopharynx 1 (0.9)

Larynx 2 (1.8)

Upper face 2 (1.9)

Mid face 2 (1.9)

Lower face 3 (2.7)

Neck 3 (2.7)

Upper limb 1 (0.9)

Tertiary site

None 41 (60.3)

Oral cavity 5 (7.4)

Oropharynx 1 (1.5)

Larynx 2 (3.0)

Lower face 2 (2.9)

Neck 15 (22.1)

Upper limb 2 (3.0)

Diagnosis

Squamous cell carcinoma 26 (38.2)

Adenoid cystic carcinoma 2 (2.9)

Adenocarcinoma 2 (2.9)

Acinic cell carcinoma 1 (1.5)

Mucoepidermoid carcinoma 1 (1.5)

Metastatic carcinoma 1 (1.5)

Chondrosarcoma 1 (1.5)

Osteosarcoma 2 (2.9)

Rhabdomyosarcoma 1 (1.5)

Fibromyxoid Sarcoma 1 (1.5)

Angiosarcoma 3 (4.4)

TABLE 1. (Continued)

No. (%)

Arteriovenous malformation 4 (5.9)

Haemangioma 7 (10.3)

Lymphangioma 9 (13.2)

Neurofibroma 1 (1.5)

Mixed hamartoma 2 (2.9)

Kimura disease 1 (1.5)

Pleomorphic adenoma 1 (1.5)

Malignant melanoma 1 (1.5)

Mantle cell lymphoma 1 (1.5)

Staging

Stage II 1 (1.5)

Stage IV 26 (38.2)

High-grade 15 (22.1)

Low-grade 1 (1.5)

Regional 25 (36.8)

Previous management

None 29 (42.6)

Surgeryþ radiotherapy 16 (23.5)

Surgeryþ chemorad 7 (10.3)

Chemorad 2 (2.9)

Radiotherapy 6 (8.8)

Surgery 5 (7.4)

Surgeryþ chemotherapy 1 (1.5)

Embolisation/sclerotherapy 1 (1.5)

CO2 laser 1 (1.5)

Treatment offered but rejected 37 (54.4)

Treatment not offered 31 (45.6)

614 JERJES ET AL.

taken to ensure that the needles are inserted parallel toeach other with 1 cm distance in between. If the treatmentwas close to a major blood vessel, a safety distance of 1 cmbetween the needle and the vessel was implemented toavoid any possible risk to promoting rupture in case thevessel wall contained tumour (Figs. 1 and 2).

A four-channel 652 nm diode laser was used for illumi-nation. Bare polished tip laser light delivery fibres with acore diameter of 400mm were introduced through the spinalneedles into the tumour. The fibres were allowed toprotrude by 2–3 mm from the needle tip into tissue toensure maximal tissue illumination. Light was thendelivered from the fibres to the target tissue at 20 J/cm2

per site (200 seconds per treatment).Each bare tip fibre delivers an output power of �0.5 W.

Histopathological specimens acquired from previous stud-ies in our centre showed that illumination from 1 bare tipfibre results in a 1 cm in diameter necrosis. Hence fibres aredistant 1 cm away from each other to cover all the suspectarea. Any residual pathological tissue in between thenecrosis areas is also expected to die due to the lack of

TABLE 2. Patients Responses Pre- and Post-PDT,

Clinical and Radiological Assessments and

Demographics

No. (%)

Patient’s report

Visual problems 3 (4.4)

Improved (P< 0.001) 3 (4.4)

No change 0 (0.0)

Worse 0 (0.0)

Breathing problems 17 (25.0)

Improved (P< 0.001) 14 (20.5)

No change 2 (3.0)

Worse 1 (1.5)

Swallowing problems 30 (44.1)

Improved (P< 0.001) 25 (36.8)

No change 3 (4.4)

Worse 2 (2.9)

Speaking problems 22 (32.3)

Improved (P< 0.001) 16 (23.5)

No change 4 (5.9)

Worse 2 (2.9)

Disfigurement 40 (58.8)

Improved (P< 0.001) 33 (48.5)

No change 6 (8.8)

Worse 1 (1.5)

Limb function 5 (7.4)

Improved (P< 0.001) 5 (7.4)

No change 0 (0.0)

Worse 0 (0.0)

Clinical assessment

No response 1 (1.5)

Minimal response 11 (16.2)

Moderate response 20 (29.4)

Good response 34 (50.0)

Free of disease 2 (2.9)

Radiological assessment

No response—stable 13 (19.1)

No response—progressing 3 (4.4)

Minimal response 18 (26.5)

Moderate response 23 (33.8)

Significant response 11 (16.2)

Necrosis and peri-pathology

inflammation

68 (100.0)

PEG 9 (13.2)

Nasopharyngeal 4 (5.9)

Tracheostomy 10 (14.7)

NG feed 9 (13.2)

Round(s) of PDT

1 round 40 (58.8)

2 rounds 23 (33.8)

3 rounds 5 (7.4)

Complications

None 40 (58.8)

Emergency airway 2 (2.9)

Fistula 4 (5.9)

Infection 6 (8.8)

Skin necrosis 2 (2.9)

Skin burn 7 (10.3)

Peripheral oedema 1 (1.5)

Motor nerve injury 3 (4.4)

Severe haemorrhage/transfusion 1 (1.5)

Stroke 2 (2.9)

Follow-up (months)

Mean 6.9

Minimum–maximum 3–14

St. deviation 2.4

Fig. 1. Ultrasound-guided interstitial photodynamic therapy

of peri-carotid disease. Note parallel array needle insertion,

optical fibres feeding into the delivery needles and tissue

illumination.

TABLE 2. (Continued)

No. (%)

ULTRASOUND-GUIDED PHOTODYNAMIC THERAPY 615

blood supply. However, overlapping treatment fields areclinically insignificant since 200 seconds illumination of thearea is adequate to activate the non-thermal photochemicalprocess. Thick tumours are treated with 1 cm pull back ateach time to ensure illumination of the whole tumourvolume.

Diffuser fibres were used when treating vascular mal-formations along with the bare polished tip fibres (Figs. 3and 4). In our cases cylindrical diffusers were used andillumination with 652 nm light followed. The structure ofthe vascular malformations allows maximal tissue illumi-nation when using those types of fibres.

Iso-illumination treatment plan was carefully imple-mented and supervised by a senior physicist to ensureadequate light delivery to all suspect areas with minimaloverlapping between the fields of treatment using a gridsystem. No measurements were made with regard to thedistribution of the light fluence rate, the optical properties,the drug concentration, and the tissue oxygenation forphotodynamic therapy as these factors had already beenquantified in previous studies carried out in our centre.

The position of the pulse oximeter was changed every30 minutes to avoid any skin burn or nail damage thatwould result from photochemical reaction by the red light(660 nm). Postoperatively, gradual light re-exposure andairway and pain control was the priority. Ten of ourpatients underwent elective tracheostomy prior to iPDTand four other had nasopharyngeal tube inserted. Thistechnique aimed to prevent any airway compromise whentreating tumours in the oropharyngeal/laryngeal region.For example, a nasopharyngeal tube is inserted intra-operatively and kept for 3 days postoperatively; intra-venous steroids are administered for 3 days to reduce localinflammatory responses. Patients were retreated with PDTwhen residual tumour is identified, 58.8% of this cohort had1 round of PDT, 33.8% had 2 rounds and 7.4% had 3 rounds.

Gradual light re-exposure at an incremental rate of 100lux per day was implemented. Every patient was instructedon the need to avoid direct sun exposure for up to 2 weeksafter injection and was given light exposure guidelines.

Fig. 2. Ultrasound image showing three level II metastatic

cervical lymph nodes (red arrows). Note needle insertion in the

node on the right hand side prior to PDT treatment (yellow

arrow).

Fig. 3. Interstitial photodynamic therapy of parotid disease

showing a left sided lesion. No shielding of surrounding normal

tissue is required when performing interstitial PDT as the

photochemical reaction is taking place at a deeper level.

Ulceration, dehiscence and cutaneous fistulas are unavoidable

when tumour is close to the surface. Parallel needle insertion is

recommended but sometimes difficult to achieve and US-

guidance can ensures that the inserted needles are inserted in

the exact predetermined area to ensure maximum tissue

response even when parallelism is difficult to achieve.

Fig. 4. Interstitial photodynamic therapy of vascular malfor-

mation of the right forearm.

616 JERJES ET AL.

Pain usually peaks at 48 hours postoperatively; paincontrol is applied according to UCLH post-PDT painprotocols; NSAIDs and opiates are usually supplied if notcontraindicated.

Patients were discharged from hospital care 3–7 dayspostoperatively. Five to six weeks postoperatively, MRI orCT scans were taken where appropriate to assess outcome.Patients were asked to report on the outcome of theirtherapy if there is any improvement, no change or worsen-ing of symptoms.

Postoperative clinical assessment was reported bythe treating clinician at the first post-PDT review (at�5-6 weeks). Postoperative radiological assessment wasreported by the same interventional radiologist. Radio-logical assessment parameters included: no response (nochange in pathology size), minimal response (reduce size by<25%), moderate response (reduce size by <50%) andsignificant response (reduce size by > 50%). Identificationof necrosis and peri-pathology inflammation was alsoreported (Fig. 5). The mean follow-up for those patientswas 6.9 months (Min 3, Max 14).

Statistical Analysis

This was performed using the SPSS 14 program(statistical package for social scientists) by an independentstatistician.

The patients’ data were entered onto proformas whichwere validated and checked by interval sampling. The

fields included a range of clinical, operative and radiologicalparameters related to the outcome following photodynamictherapy.

The results were cross tabulated and the Chi-squaredstatistic was used to test for differences in the incidence ofoutcome. Fisher’s exact test was used for the analysis ofcontingency tables and therefore to measure the P-value.Cross tabulation was also used to compare pathologyto clinical and radiological assessment and pathology tocomplications.

RESULTS

All patients who had visual problems reported improve-ment after treatment (P<0.001). Also 14/17 patientsreported improvement of breathing (P<0.001), with onepatient reported worsening of symptoms. Improvement ofswallowing was reported by 25/30 patients (P<0.001);while speaking improvement was evident in 16/22 patients(P<0.001) and 33/40 reported reduction in the disfigure-ment caused by their pathology (P<0.001). All patientsreported improved limb function (P<0.001) (Table 2).

Clinical assessment showed that half of the patients had‘good response’ to the treatment and a third reported‘moderate response’ with two patients being free of disease.Radiological assessment comparing imaging 5-6 weekspost-PDT to the baseline showed stable pathology withno change in size in 13 patients, minimal response(<25% reduction) in 18 patients, moderate response (<50%reduction) in 23 patients and significant response (> 50%reduction) in 11 patients (Figs. 6 and 7). Unfortunately,three patients showed no response with progressivetumour growth (squamous cell and adenoid cystic carcino-mas) radiologically (Table 2).

Interestingly, when comparing pathology to clinicaland radiological assessment we found that patients withsquamous cell carcinoma responded quite favourably to

Fig. 5. Post-PDT MRI image of cystic hygroma. The image

showing peri-pathology inflammation (red arrows), necrotic

focci (blue arrow) and breakdown in the structure of the

malformation (yellow arrows). This is the typical imaging

result expected after PDT.

Fig. 6. Pre- and post-PDT MRI of base of tongue squamous cell

carcinoma (mainly involving the right side). Six-month post-

PDT scan showing marked shrinkage in tumour volume and

bystander tissue regeneration and minimal scarring (red

arrows). Also improvement of airway (yellow demarcation)

can be seen.

ULTRASOUND-GUIDED PHOTODYNAMIC THERAPY 617

US-iPDT, with only 2/26 patient having progressive diseaseafter PDT. Also, vascular malformations (haemangiomasand lymphangiomas) remained either stable or reduced insize following the treatment (Table 3). Nearly 60% of ourpatients reported no problems following PDT. Unfortu-nately, due to the extended duration of skin photosensiti-sation following treatment, skin burn (mostly 2nd degree)was reported by seven of our patients; while two patientshad skin necrosis caused by the treating subcutaneouspathologies very close to the surface (Table 2).

Patients who suffered from tumours in the oropahryngeal/laryngeal regions developed postoperative complications,not related to iPDT, including the need for emergencyairway in two patients, stroke in two patients, orocutaenousfistulas in four patients and the need for intra-operativetransfusion (4 units) in one patient (Table 4).

Motor nerve injuries were reported by three patients (twomarginal mandibular nerves and one facial nerve) whichcould be due to direct trauma caused by needle insertionduring the treatment. Two of those patients recoveredcompletely. Six patients reported postoperative infectionsand were treated with anti-microbials and one patientreported peripheral oedema (left hand) following treatmentof vascular malformation of face, neck and tongue wastreated with steroids (dexamethasone) and achieved fullrecovery (Table 2). When comparing pathology to compli-cations post-PDT, we note that the 2/3 of patients (6/9) who

had skin burn and necrosis (Fig. 8) presented with vascularmalformations.

DISCUSSION

PDT is currently being used in the management ofseveral cancers including head and neck, skin, brain, lung,pancreas, intra-peritoneal cavity, breast and prostate. Alsowas proven to be very successful in treating vascularmalformations. The growing body of evidence regardingits efficacy suggests that it will have a leading role ininterventional oncology, especially with the development ofimage-guided iPDT [1–3].

It is fundamental to the success of photodynamic therapythat the depth of PDT necrosis is greater than the depth ofeach individual tumour or malformation. Interstitialphotodynamic therapy represents the optimal solution forthis problem by treating a target volume. Ultrasound cansimply guide the optical fibres to the appropriate diseasevolume (tumour or vascular malformation) which can be co-registered with a treatment volume allowing accurateassessment of delivered treatment doses [2,3,7]. Ultra-sound also allows the guidance of delivery apparatus awayfrom vital structures that may not have been otherwisepalpable, that is arteries and nerves. It also ensures theparallel placement of delivery needles enabling improveddosage administration profiles since the surface orientationof the needles is not a reliable measure of their deepercourse which may have been distorted by changes in tissueconsistency or architecture (bone, tendons).

Vaidya et al. [8] performed contrast-enhanced MRI-guided PDT with a bifunctional polymer conjugate con-taining both a magnetic resonance imaging contrast agentand a photosensitiser in athymic nude mice bearing MDA-MB-231 human breast carcinoma xenografts. The polymerconjugates preferentially accumulated in the solid tumourdue to the hyperpermeability of the tumour vasculature,resulting in significant tumour enhancement for accuratetumour detection and localisation by MRI. Significanttherapeutic response was observed for PDT with thebifunctional conjugate as compared to the control. Yusufet al. used endoscopic ultrasound-guided photodynamic(EUS-PDT) therapy with the photosensitising agent por-fimer sodium to ablate pancreatic tissue in six swine. Thetail of the pancreas was located with EUS and was used toguide the placement of a light catheter. The pancreatic tailwas exposed to 10, 15, or 20 minutes of laser light (689 nm).Localised tissue necrosis within the pancreatic tail (range6.6–30.5 mm in diameter) was seen in all animals. Thediameter of the necrotic tissue was directly related tothe dose of light. No postprocedural complications wereobserved [9].

A phase I-II study was conducted to assess the safety andefficacy of interstitial image-guided-iPDT for patients withpersistent or recurrent head and neck cancer unsuitable forfurther treatment with surgery, radiotherapy or chemo-therapy, recruited for ‘last hope’ salvage treatment. Forty-five patients were sensitised with Foscan prior to CT orMRI-guided light delivery from fibres inserted directly into

Fig. 7. 3D-arteriogram of the lower leg (below knee

cross section) showing haemangioma post-iPDT. Image ‘A’

showing extensive haemangioma of the same area; while

image ‘B’ showing major resolution post-PDT.

618 JERJES ET AL.

the target tumour. The results showed that nine patientsachieved a complete response and five are alive and free ofdisease 10–60 months later [10]. Jager et al. [11] reportedthe use of an open interventional MR system to guideinterstitial PDT for advanced head and neck tumours andfound this technique to be accurate and safe. Betz et al. [3]reported a series of 11 patients with lymphatic and venousmalformations treated with interstitial PDT. The resultsshowed that in all cases there was a significant reduction in

the volume of abnormal tissue without damage to theoverlying skin. The best results were obtained withlymphatic malformations, especially for those that hadnot undergone previous surgery.

Multi-disciplinary team working is essential to theefficient modern management of deep-seated diseaseespecially when end-stage or locally advanced. The conceptof disease treatment has evolved from just the two-dimensional static image that is captured by conventional

TABLE 3. Clinical and Radiological Assessment Per Pathology

Dx vs. clinical assessment

No response,

no.

Minimal

response, no.

Moderate

response, no.

Good

response, no.

Free of

disease, no.

Squamous cell carcinoma 1 4 6 14 1

Adenoid cystic carcinoma 0 0 1 1 0

Adenocarcinoma 0 1 1 0 0

Acinic cell carcinoma 0 0 1 0 0

Mucoepidermoid carcinoma 0 0 0 0 1

Metastatic carcinoma 0 0 0 1 0

Chondrosarcoma 0 0 0 1 0

Osteosarcoma 0 1 0 1 0

Rhabdomyosarcoma 0 0 1 0 0

Fibromyxoid sarcoma 0 0 0 1 0

Angiosarcoma 0 0 1 2 0

Arteriovenous malformation 0 0 3 1 0

Haemangioma 0 1 2 4 0

Lymphangioma 0 2 2 5 0

Neurofibroma 0 0 0 1 0

Mixed hamartoma 0 1 1 0 0

Kimura disease 0 0 0 1 0

Pleomorphic adenoma 0 1 0 0 0

Malignant melanoma 0 0 1 0 0

Mantle cell lymphoma 0 0 0 1 0

Dx vs. radiological assessment No

response—

stable, no.

No

response—

progressing, no.

Minimal

response,

no.

Moderate

response,

no.

Significant

response,

no.

Squamous cell carcinoma 4 2 9 9 2

Adenoid cystic carcinoma 1 1 0 0 0

Adenocarcinoma 1 0 0 0 1

Acinic cell carcinoma 0 0 0 1 0

Mucoepidermoid carcinoma 0 0 0 1 0

Metastatic carcinoma 0 0 0 1 0

Chondrosarcoma 0 0 1 0 0

Osteosarcoma 1 0 0 1 0

Rhabdomyosarcoma 0 0 0 1 0

Fibromyxoid sarcoma 0 0 0 1 0

Angiosarcoma 0 0 1 2 0

Arteriovenous malformation 3 0 1 0 0

Haemangioma 0 0 1 3 3

Lymphangioma 2 0 2 0 5

Neurofibroma 0 0 0 1 0

Mixed hamartoma 1 0 1 0 0

Kimura disease 0 0 1 0 0

Pleomorphic adenoma 0 0 0 1 0

Malignant melanoma 0 0 1 0 0

Mantle cell lymphoma 0 0 0 1 0

ULTRASOUND-GUIDED PHOTODYNAMIC THERAPY 619

TA

BL

E4.

Po

st-P

DT

Co

mp

lica

tio

ns

Per

Pa

tho

log

y

Non

e

Em

ergen

cy

air

way

Fis

tula

Infe

ctio

n

Sk

in

nec

rosi

s

Sk

in

bu

rn

Per

iph

eral

oed

ema

Mot

orn

erve

inju

ry

Sev

ere

haem

orrh

age/

tran

sfu

sion

Str

oke

Squ

am

ous

cell

carc

inom

a21

22

10

00

00

0

Ad

enoi

dcy

stic

carc

inom

a1

00

00

10

00

0

Ad

enoc

arc

inom

a1

00

00

00

00

1

Aci

nic

cell

carc

inom

a1

00

00

00

00

0

Mu

coep

ider

moi

dca

rcin

oma

10

00

00

00

00

Met

ast

ati

cca

rcin

oma

00

00

00

00

10

Ch

ond

rosa

rcom

a0

01

00

00

00

0

Ost

eosa

rcom

a1

00

01

00

00

0

Rh

abd

omyos

arc

oma

10

00

00

00

00

Fib

rom

yxoi

dsa

rcom

a1

00

00

00

00

0

An

gio

sarc

oma

00

02

01

00

00

Art

erio

ven

ous

malf

orm

ati

on4

00

00

00

00

0

Haem

an

gio

ma

20

00

03

10

01

Lym

ph

an

gio

ma

30

12

01

02

00

Neu

rofi

bro

ma

00

00

01

00

00

Mix

edh

am

art

oma

10

00

10

00

00

Kim

ura

dis

ease

10

00

00

00

00

Ple

omor

ph

icad

enom

a0

00

00

00

10

0

Mali

gn

an

tm

elan

oma

00

01

00

00

00

Man

tle

cell

lym

ph

oma

10

00

00

00

00

620 JERJES ET AL.

radiological methods. Important considerations such as thedisease margin or host/tumour interface are a significantfactor in eventual outcome. We now consider a tumourvolume and slightly larger treatment volume to ensure amore logical and complete iPDT treatment. This is aided byuse of computer modelling, a needle grid (to ensure paralleliso-doses of illumination), timed illumination of fibres.Whilst this is relatively simple to conceptualise in analmost 2D skin model, it is significantly more complex indeep-seated lesions where 3D imaging in the form of‘real time’ ultrasound is necessary to help avoid vitalstructures during needle placement under image guidance.

In theatre we regularly have a specialist medicalphysicist to help to conform to our iso-illumination treat-ment plan reducing unnecessary treatment and side effectsfrom photo-activation. Our results showed that the treat-ment is well tolerated and effective especially for end-stagesquamous cell carcinoma and vascular lesions. The greatutility of this treatment modality is not only in itsrepeatability (unlike radiotherapy) but in the accuracy oftreatment (to avoid unwanted bystander tissue damageagain unlike radiotherapy).

Intuitively it is reasonable to expect image-guided treat-ment to be better than the ‘blind’ insertion illuminationfibres. This is borne in the outcomes of this study whichshowed that where there is a response to PDT (either byreducing tumour or vascular lesional growth), image-

guided treatment is efficacious. The minimally invasivenature of fibre delivered photo-activated treatment couldallow this treatment modality to be used in many otherspecialities and for many other pathologies ranging fromadjunctive tumour treatment to the continuum of vascularlesions.

In this study, various pathologies were treated includingcarcinomas, sarcomas and vascular malformations. It isexpected that response and outcome to vary betweendifferent pathologies which might depend on the structure,volume and severity of the pathology. Also it is worth toacknowledge that response to PDT varies among patientssuffering from the same grade of disease. It is worth toclarify that the criteria for the clinically assessed outcomedoes not provide a complete objective assessment ofthe outcome. Further objective assessment of the outcomecould be carried out using approved quality of lifequestionnaires (i.e. apply modifications to the Universityof Washington Quality of Life Questionnaire to be appliedon patients undergoing photodynamic therapy).

In conclusion, the growing body of evidence regarding theefficacy of interstitial photodynamic therapy suggests thatit may have a leading role in minimally invasive surgeryand interventional oncology, especially with the develop-ment of image-guided iPDT.

REFERENCES

1. Hopper C. Photodynamic therapy: A clinical reality in thetreatment of cancer. Lancet Oncol 2000;1:212–219.

2. Jerjes W, Upile T, Betz CS, El Maaytah M, Abbas S, WrightA, Hopper C. The application of photodynamic therapy in thehead and neck. Dent Update 2007;34(8):478–480, 483–484,486.

3. Betz CS, Jager HR, Brookes JA, Richards R, Leunig A,Hopper C. Interstitial photodynamic therapy for a symptom-targeted treatment of complex vascular malformations in thehead and neck region. Lasers Surg Med 2007;39(7):571–582.

4. Zhu TC, Finlay JC, Hahn SM. Determination of thedistribution of light, optical properties, drug concentration,and tissue oxygenation in-vivo in human prostate duringmotexafin lutetium-mediated photodynamic therapy. J Pho-tochem Photobiol B 2005;79(3):231–241.

5. Wilson BC, Patterson MS. The physics, biophysics andtechnology of photodynamic therapy. Phys Med Biol 2008;53(9):R61–R109.

6. Hopper C, Kubler A, Lewis H, Tan IB, Putnam G. mTHPC-mediated photodynamic therapy for early oral squamous cellcarcinoma. Int J Cancer 2004;111(1):138–146.

7. Jerjes W, Upile T, Vincent A, Abbas S, Shah P, Mosse CA,McCarthy E, El-Maaytah M, Topping W, Morley S, Hopper C.Management of deep-seated malformations with photody-namic therapy: A new guiding imaging modality. Lasers MedSci 2009;(24)5:769–75.

8. Vaidya A, Sun Y, Ke T, Jeong EK, Lu ZR. Contrast enhancedMRI-guided photodynamic therapy for site-specific cancertreatment. Magn Reson Med 2006;56(4):761–767.

9. Yusuf TE, Matthes K, Brugge WR. EUS-guided photo-dynamic therapy with verteporfin for ablation of normalpancreatic tissue: A pilot study in a porcine model (withvideo). Gastrointest Endosc 2008;67(6):957–961.

10. Lou PJ, Jager HR, Jones L, Theodossy T, Bown SG, HopperC. Interstitial photodynamic therapy as salvage treatmentfor recurrent head and neck cancer. Br J Cancer 2004;91(3):441–446.

11. Jager HR, Taylor MN, Theodossy T, Hopper C. MR imaging-guided interstitial photodynamic laser therapy for advancedhead and neck tumors. Am J Neuroradiol 2005;26(5):1193–1200.

Fig. 8. Healing of 3rd degree skin burn following inadvertent

sun exposure. This is the arm where the photosensitiser is

administered. Patients are given pre- and post-treatment

instructions (including written) regarding light exposure and

complication avoidance and reporting.

ULTRASOUND-GUIDED PHOTODYNAMIC THERAPY 621