Br. J. Surg. 1990, Vol. 77, January, 93-96

H. Barr?*, N. Krasner?, P. B. Boulos*, P. Chatlanit* and S. G. Bown* ?GastrointestinalUnit, Walton Hospital, Liverpool. and The Department of Surgery, University College, London, UK Correspondence to: Mr H. Barr, Department of Surgery, The Rayne Institute,

University College London, 5 University Street, London WC1 E 6JJ. U K

Photodynamic therapy for colorectal cancer: a quantitative pilot study Ten patients with colorectal cancers unsuitable f o r operation were treated with endoscopic photodynamic therapy (PDT).The patients were assessed before treatment, and at 1 week and 1 month after treatment by colonoscopy with biopsy and endoluminal ultrasound examination. The depth of tumour was measured and the effect of PDT was quantified by measuring the reduction in tumour depth. All patients were sensitized with 2.5 mg k g - ' of haematoporphyrin derivative, 48 h before phototherapy. A standard treatment protocol of light exposure was used. Up to four parts of the tumour were treated with 50 J of red light (630nm) from a tuneable dye laser, through a flexible optical fibre passed through the colonoscope and inserted into the tumour. Two patients with small lesions are tumour-free 20 and 28 months after PDT. One treatment of an advanced tumour was complicated by a haemodynamically significant secondary haemorrhage. PDT may be most suitable f o r the treatment of small tumours or f o r small areas of persistent tumour where the bulk has been removed by alternative techniques. Keywords: Photodynamic therapy, colorectal cancer, laser, endoscopic ultrasound

In the last few years interest has developed in photodynamic therapy (PDT) as a method with potential for the selective destruction of malignant tumours. It is based on the systemic administration of photosensitizing drugs that are retained with some selectivity in tumours when compared with the normal tissue in which the tumour arose. When activated by light of the appropriate wavelength, usually from a laser, a cytotoxic effect which can lead to local necrosis occurs in the area exposed to the light. We have shown by careful adjustment of the photosensitizer concentration that it is possible to destroy experimental colon cancers without producing any damage to adjacent normal colon exposed to similar light doses'. This occurs because of photosensitizer photodegradation'. True selective tumour necrosis occurs to a depth of 1-2 mm. Greater depth of tumour necrosis is possible using higher doses of photosensitizer, but one must then either limit the light exposure to the tumour only, or be able to accept some necrosis in normal areas. Some necrosis in normal areas is acceptable as there is a biological advantage in the nature of the damage and healing process in the colon following PDT compared with other methods of local tissue destruction3. Photodynamic therapy of normal colon can cause full thickness necrosis without risk of perforation or a decrease in the mechanical strength of the bowel, and healing proceeds with regeneration of all layers of the bowel wall. In contrast, full thickness thermal necrosis of comparable extent is associated with a 22 per cent perforation rate and considerable weakening of the bowel wall4. Experimental tumours undergoing necrosis caused by PDT slough and the defect heals by regeneration of normal colon'. Following these experimental results, we have performed a pilot study on the efficacy of PDT in a series of ten patients with inoperable colorectal cancer. We did not attempt to achieve true selective tumour necrosis, but used larger doses of the photosensitizer to produce greater depth of necrosis. The light was directed principally to neoplastic areas, but adjacent normal areas were exposed to some light to ensure that adequate light doses reached the region where normal and neoplastic tissue met. Since our experimental studies had

0007-1323/90/010093-04$3.00

0 1990 Butterworth & Co (Publishers) Ltd

indicated safe healing following necrosis in such normal areas, this was considered to be justified. To date the vast majority of clinical studies of PDT have failed to measure the effect quantitatively, and treatment has been purely empirical. I t is important to be able to predict how much tumour necrosis will be produced with given doses of light and photosensitizer, and to match this to the extent of the tumour. We assessed the tumour response quantitatively by using endoluminal ultrasound, to allow predictive dosimetric calculations for future studies.

Patients and methods Photosensitizer and laser The photosensitizer used was haematoporphyrin derivative (HpD), synthesized and supplied by Paisley Biochemicals Ltd. (Paisley, UK). It contained 44 per cent dihaematoporphyrin ether/ester, considered to be the active component of HpD'. The patients were admitted to hospital and 2.5 mg kg- I HpD was administered as an intravenous infusion in normal saline over 30min6. Irradiation was performed using a continuous wave argon ion pumped dye laser tuned to deliver light at 630nm coupled to a 400pm quartz optical fibre (Living Technology Ltd., Glasgow, UK). Patients Ten patients (five men, five women) with a mean age of 73 (range 4490) years were treated; the patient details are shown in Tablel. Patient I had had a previous anterior resection and had developed a true anastomotic recurrence with no evidence of tumour invading from outside. Patient 8 had a large tumour in the descending colon (68cm from the renal verge). He was initially treated with high power external, and low power interstitial, Nd YAG laser therapy which reduced the polypoid carcinoma to a small area of residual intramural tumour. Further thermal laser therapy would have risked weakening the colonic wall and producing perforation, endosonography having shown that the tumour was invading the full thickness of the bowel wall. PDT was used to eradicate the residual tumour without risk of weakening the colonic wall. Tumour ussessmeni All tumours were biopsy proven adenocarcinomas which were predominantly exophytic. In four patients the tumour was completely

93

Photodynamic therapy for colorectal cancer: H. Barr et al.

Table 1 Details of’ (lie patienrs and tumours Patient

Reason for inoperability

Tumour site (cm from anus) ~

~~

1

9

10

Recent myocardial infarction. Recurrence after anterior resection

1&11.5

Metastases, debility Metastases Refused surgery Cardiovascular disease Cardiorespiratory disease Metastases, cardiac disease Cardiomyopathy, general debility. Residual tumour after Nd YAG laser therapy Cerebrovascular disease. general debility

8-12 8-13 13-17 18-21 &I 1 3-6 68-70

Metastases

2 1-24

All patients were advised to avoid direct sunlight for 1 month after injection of HpD, and to wear a wide brimmed hat and gloves if going out in the sunlight. If following PDT there were still areas of tumour remaining and the treatment had failed to provide adequate symptomatic relief, the patients were treated with the Nd YAG laser’.

Results Patient 1, who had a small recurrent anastomotic tumour (maximum depth 6mm, length 15mm) lOcm from the anal margin, was treated at three sites. There was necrotic tumour present I week after treatment and by 4 weeks there was no evidence of tumour either at endoscopic examination and biopsy, or at endosonography. This patient is disease-free 28 months after treatment. Figure 2 shows a photomicrograph of the tumour prior to treatment and Figure3 is a biopsy from

3G32

ColonoscoDe

circumferential. The longitudinal extent of the tumour was assessed by sigmoidoscopy. All tumours were 5cm in length or less. If greater than this, eradication with P D T was considered to be difficult, and the dangers of secondary haemorrhage significant’. Palliative treatment was then considered to be most appropriately performed with the Nd YAG laser. The maximum depth of invasion of the tumour was measured using the rectal ultrasound endoprobe type 1846 (Bruel and Kjaer, Naerum, Denmark) with either a 5.5MHz or 7.0MHz 9 0 transducer (focal lengths of l 4 c m and 2-5 cm respectively). The method of examination was as described by Beynon et The examination was recorded on video tape and the position of the depth measurements was recorded to ensure accurate follow-up. By taking the area of maximum depth of tumour before and after treatment. it was possible to calculate the amount of tumour removed. Ultrasound examination was not possible in one patient with a stenotic tumour. The two tumours. 30 and 68 cm from the anal margin, were inaccessible to the rigid rectal probe so assessment was made using the Olympus flexible endoscopic ultrasound (GIF-UM2. Key Med, Southend. UK). Measurements were taken immediately before treatment and 1 week. 3 weeks and 4 weeks after treatment. Merliod of’ treatnienr Before phototherapy the patients received a rectal washout. Those with colonic tumours were given a bowel preparation with sodium picosulphate and magnesium citrate (Picolax ‘, Nordic Pharmaceuticals Ltd., Feltham. UK). All were treated 48 h after H p D injection. the time considered to give the best therapeutic ratio between tumour and normal tissue6. The procedure was performed through a colonoscope. Sedation with 4-10 mg midazolam (Hypnovel ’, Roche Products Ltd., Welwyn Garden City, U K ) was only given if the patient was anxious or found the long exposure times uncomfortable. The laser was set to deliver 5G100mW from the tip of the optical fibre. thus avoiding thermal damage3. The end of the optical fibre was cleaved to ensure a circular output. and the outer cladding was removed for I mm from the fibre tip. It was then passed down the instrumentation channel of the colonoscope and inserted I mm into the tumour (i.e. up to the point where the cladding had been removed). This interstitial method of light delivery was found to provide the best means of illumination. Surface irradiation of rectal tumours was dificult as it was hard to hold the fibre in a constant position during bowel movements; also mucus, blood or faeces were likely to cover the tumour and reduce the useful delivered light dose. The exposure time used at each treatment point was 5 W 1 0 0 0 s . giving an energy of 50J. It must be emphasized that the choice of treatment parameters was empirical, although our data from experimental studies indicated that these parameters should produce safe tumour necrosis. Reports of previous human studies give no indication of the precise extent of necrosis that might be expected from each treatment point. It was one of the aims of this study to quantify this. In patients with circumferential tumours up to four areas of the tumour were treated. the same light dose being delivered to each. In these tumours the area where the maximum depth of invasion had been measured was treated, and other treatment sites were in separate tumour nodules. such that the light fields just overlapped (Figure 1 ). In large polyploid tumours the four sites of treatment were again in the bulky part of the tumour.

94

Diugrunt of the niethodoftreatment of’u IonRitudinalspreadina runiour n / / h e c d o n , /lirough the colonoscope with /he laser,fihre inserted iriro /lie runtour nodules. Three minintully overlapping lighr ,fields are

Figure 1

sliowi

Figure 2 Pltoromicrogruph nf biopsy ,from colon of’ patient I (udmoc~crrcinornu); hueniuro.u~~liti und eosin sruin, niugn~fica~ion x 11)O

Br. J. Surg., Vol. 77, No. 1, January1990

Photodynamic therapy for colorectal cancer: H. Barr et al.

Discussion The best results in this series were obtained when the tumour was small and could be totally eradicated. Of the two patients in whom the tumour was completely destroyed, one had a small true anastomotic recurrence and the other had a small area of residual tumour after the bulk had been removed with the Nd Y A G laser. These patients illustrate the important biological advantages of PDT. First, the risk of perforation is small even if full thickness necrosis is produced to eradicate the t ~ m o u r ~Second, , ~ . areas of PDT necrosis heal predominantly by regeneration with little scarring3. In contrast perforation and severe scarring resulting in stricture formation have occured after Nd Y A G laser therapy". The major disadvantage of P D T is that the amount of turnour destruction is limited by the penetration of 630nm light". Improved light delivery, such as cylindrical rather than point source irradiation, multiple fibre irradiation, and precise placement of the fibres may increase the amount of tumour removed. However, other methods of local tumour destruction with high power, and low power interstitial, N d Y A G laser hyperthermia will produce larger amounts of tumour necrosis since the infrared beam (1064 nm) penetrates further into tissue' ' . These thermal methods are also unlikely to produce haemorrhage. One of our patients had a haemodynamically significant haemorrhage 8 days after PDT. This complication was first reported in a series from the Mayo clinic. Ten patients with bronchial carcinomas were treated with PDT. Two patients with large advanced tumours had major intrabronchial haemorrhage 6 and 11 days after PDT,

Table 2 Maximunt deprh of invasion before photodvnamic therapy and I week ufier treatment Figure 3 Photoniicrograph ofbiopsy,fromrhe colon ofpatient I showing colonic mucosal regeneration at site of adenocarcinoma; haematosvlin and rosin stain, magnijkation x 100

the area 1 month after treatment showing that the colonic mucosa has completely regenerated. Patient 8 had his residual tumour completely eradicated, and he is disease-free 20 months after PDT. A further patient was thought, at colonoscopy, to have tumour completely eradicated 4 weeks after PDT although technical problems prevented satisfactory ultrasound examination. Four months later an intraluminal nodule was noticed a little distance from the site of previous treatment and was found to be a recurrent adenocarcinoma. This recurrence was in an area near the treated tumour, which had not been exposed to the light during treatment. In eight patients it was possible to measure the depth of tumour invasion before and 1 week after treatment. The results are shown in Table 2 . The symptomatic improvement at the 1 month follow-up in all patients is shown in Table3. The two patients who had small tumours completely eradicated became asymptomatic. Six patients have died from malignant disease 1, 2 , 2, 6, 12 and 15 months after treatment. One patient died from cardiorespiratory disease, 8 months after PDT, and one patient is alive receiving Nd Y A G laser therapy (16 months). Two patients developed complications following PDT. Patient 7, with a large tumour, had a haemodynamically significant haemorrhage from the residual rectal cancer, causing hypotension, and required transfusion with 2 units of blood, 8 days after PDT. The bleeding stopped spontaneously. On examination there had been extensive tumour necrosis exposing friable tumour vessels, that bled easily on contact. We have not treated any more large tumours because of the risk of haemorrhage. One other patient developed mild erythema on one arm when she left it exposed in bright sunlight for 2 h, 4 days after photosensitization.

Br. J. Surg., Vol. 77, No. 1, January 1990

Maximum depth of Maximum depth of Depth of tumour before PDT tumour after PDT tumour removed Patient (mm) (mm) (mm 1 1 2 3 4 5 6 I 8

6 16 12 31 26 20 35 4

0 11

5 28 20 17 20 0

PDT, photodynamic therapy

Table 3 Symptomatic response of parienrs treated with photodwamic tlierapy Pat ien t

I 2

3

Symptoms (11 referral

Symptoms afier PDT

Very occasional rectal hleeding Constipation and overflow diarrhoea. Abdominal colic (obstructing tumour)

No bleeding

Constipation and bleeding (obstructing tuntour) Diarrhoea 3-4 times a day

Intermittent diarrhoea Bowels open hourly Tenesmus I

8 9 10

Severe constipation (obstructing tumour) Rectal bleeding Discharge and bleeding Rectal bleeding

Bonels open daily. No colic

Constipation improved

N o improvement Bowel habit more regular Very little improvement. Tenesmus not improved Bowel action more regular No bleeding No improvement No improvement

95

Photodynamic therapy for colorectal cancer: H. Barr et al.

and both patients died7. This report in conjunction with o u r experience prompts us to suggest that incomplete treatment of large cancers with P D T may be hazardous. Delayed haemorrhage may occur when the necrosed tumour sloughs, exposing a raw surface of malignant tissue with its fragile tumour vessels. The purpose of using endoluminal ultrasound was twofold: first, to measure accurately the depth of the tumour prior to treatment and to avoid treating tumours that were large and could not be eradicated using PDT (the patient who bled had a stenotic tumour that could not be assessed by endoluminal ultrasound); and second to calculate the depth of necrosis produced from a standard photodynamic dose (photosensitizer dose x light dose), thus allowing dosimetric calculations and examination of current dosimetric theory*. If assumptions on tissue optical properties and photosensitizer concentrations are made from the available literature”, the data presented in Table 2 are consistent with current dosimetric theory2. In conclusion, P D T has promise for the treatment of small tumours of the gastrointestinal tract or for small residual areas of tumour after the bulk has been removed by other techniques (e.g. Nd YAG laser therapy, polypectomy or after limited surgical resection). A combination approach proved particularly effective in one of our patients. In its present form this technique should be used with caution for the treatment of advanced cancers, especially when other techniques such as Nd YAG laser therapy are effective.

Acknowledgements Mr H. Barr is a Wellcome Trust Surgical Fellow and D r S . G . Brown is supported by the Imperial Cancer Research Fund. The patients were treated at the Gastrointestinal Unit, Walton Hospital, Liverpool.

96

References Barr H, Tralau CJ, Lewin M, Clark CG, Boulos PB, Bown SG. Selective destruction of experimental colon cancer using photodynamic therapy. Br J Surg 1988; 75: 61 1-12. 2. Potter WR, Mang TS, Dougherty TJ. The theory of photodynamic dosimetry: consequences of photodestruction of sensitizer. Photochem Photobiol 1987; 46: 97-101. 3. Barr H, Tralau CJ, MacRobert AJ et al. Photodynamic therapy with phthalocyanine sensitisation in the normal rat colon. Br J Cancer 1987; 55: 389-95. 4. Barr H, Tralau CJ, Boulos PB, MacRobert AJ, Tilly R, Bown SG. The contrasting mechanisms of colonic photodynamic damage between photodynamic therapy and thermal injury. Photochem Photobiol 1987; 46: 795-800. 5. Dougherty TJ, Potter WR, Weishaupt KR. The structure of the active component of hematoporphyrin derivative.In: Doiron DR and Comer CJ, eds. Porphyrin Localization and Treatment of Tumours. New York: Alan R. Liss, 1984: 301-14. 6. Dougherty TJ. Photosensitizers: therapy and detection of malignant tumors. Photochem Photobiol 1987; 45: 879-89. 7. Cortese DA, Kinsey JH. Endoscopic management of lung cancer with hematoporphyrin derivative phototherapy. Mayo Clin Proc 1982; 57: 543-7. 8. Beynon J, Foy DMA, Roe AM, Temple LN, Mortensen NJMcC. Endoluminal ultrasound in the assessment of local invasion in rectal cancer. Br J Surg 1986; 73: 4747. 9. Bown SG, Barr H, Matthewson K et a/. Endoscopic treatment of inoperable colorectal cancers with the Nd YAG laser. Br J Surg 1986; 73: 949-52. 10. Krasner N, Barr H, Skidmore C, Morris AI. Palliative laser therapy for malignant dysphagia. Gut 1987; 28: 792-8. 11. Svaasand LO. Thermal and optical dosimetry for photoradiation therapy of malignant tumors. In: Andreoni A and Cubeddu R, eds. Porphyrins in Tumor Phototherapy. New York: Plenum Press, 1984: 261-79. 12. Profio AE, Doiron DR. Transport of light in tissue in photodynamic therapy. Phoiochem Photobiol 1987; 46: 591-9. I.

Paper accepted 11 July 1989

Br. J. Surg., Vol. 77, No. 1, January 1990

Photodynamic therapy for colorectal cancer: a quantitative pilot study.

Ten patients with colorectal cancers unsuitable for operation were treated with endoscopic photodynamic therapy (PDT). The patients were assessed befo...
1016KB Sizes 0 Downloads 0 Views