- Email: [email protected]
Hexaminolevulinate photodynamic diagnosis in non-muscle invasive bladder cancer: Experience of the BLUE group
Actas Urol Esp. 2011;35(8):439---445
Actas Urológicas Españolas www.elsevier.es/actasuro
ORIGINAL ARTICLE
Hexaminolevulinate photodynamic diagnosis in non-muscle invasive bladder cancer: Experience of the BLUE group夽 J.P. Burgués a,∗ , G. Conde b , J. Oliva c , J.M. Abascal d , I. Iborra e , M. Puertas f , F. Ordo˜ no g , Grupo BLUE (Blue Light Urologic Endoscopy) a
Servicio de Urología, Hospital Universitario Son Espases, Palma de Mallorca, Spain Servicio de Urología, Hospital Son Llàtzer, Palma de Mallorca, Spain c Servicio de Urología, Hospital Royo Villanova, Zaragoza, Spain d Servicio de Urología, Hospital Universitario Central de Asturias, Oviedo, Spain e Servicio de Urología, Instituto Valenciano de Oncología, Valencia, Spain f Servicio de Urología, Hospital Marina Baixa, La Vila Joiosa, Alicante, Spain g Servicio de Urología, Hospital Arnau de Vilanova, Valencia, Spain b
Received 25 February 2011; accepted 11 March 2011
KEYWORDS Photodynamic diagnosis; Hexaminolevulinate; Non-muscle invasive bladder cancer
Abstract Objectives: Photodynamic diagnosis (PDD) with hexaminolevulinate has been recently used to improve detection of non-muscle invasive bladder cancer. Our main purpose was to quantify the benefit of PDD versus conventional white light cystoscopy (WL) in our area. Materials and methods: Fluorescence-guided cystoscopy using hexaminolevulinate was performed at the time of the transurethral resection (TUR) in 305 patients from 7 Spanish hospitals. All lesions found with WL and PDD were numbered and recorded in an online database. Each lesion was sent separately for pathology analysis. Random biopsies were also obtained in 148 patients. Results: A total of 1659 lesions were biopsied: 522 were identified with PDD and WL, 237 only with PDD, 19 only with WL and 881 random biopsies. Of the 600 tumors, PDD detected 563, WL 441 and random biopsies 29 (20 CIS). The mean overdetection rate for PDD over WL was 31.9% for all types of lesions, but it was 209% for carcinoma in situ (CIS). Sensitivity was 93.8% for PDD and 78.2% for WL. Specificity was 81.5% for PDD and 90.5% for WL. In 23% of patients, PDD detected at least one more neoplastic lesion than with WL. Conclusions: Hexaminolevulinate fluorescence cystoscopy improves detection and resection of non-muscle invasive bladder cancer, especially of CIS. Sensitivity of PDD is higher than WL, but specificity is lower. In our study, random biopsies were able to detect some CIS not visible under PDD. © 2011 AEU. Published by Elsevier España, S.L. All rights reserved.
夽 Please cite this article as: Burgues JP, et al. Diagnóstico fotodinámico con hexaminolevulinato en el cancer vesical no músculo invasivo: experiencia del grupo BLUE. Actas Urol Esp. 2011;35:439---445. ∗ Corresponding author. E-mail address: [email protected] (J.P. Burgués).
2173-5786/$ – see front matter © 2011 AEU. Published by Elsevier España, S.L. All rights reserved.
440
PALABRAS CLAVE Diagnóstico fotodinámico; Hexaminolevulinato; Cáncer vesical no músculo invasivo
J.P. Burgués et al.
Diagnóstico fotodinámico con hexaminolevulinato en el cáncer vesical no músculo invasivo: experiencia del grupo BLUE Resumen Objetivos: El diagnóstico fotodinámico (DFD) con hexaminolevulinato se ha empezado a utilizar recientemente para mejorar la detección del cáncer vesical no músculo invasivo. Nuestro objetivo principal fue comparar el rendimiento diagnóstico de PDD frente a endoscopia con luz blanca convencional (LB) en nuestro medio. Material y métodos: Se realizó cistoscopia fluorescente con hexaminolevulinato en el momento de la RTU a 305 pacientes de 7 hospitales espa˜ noles. Todas las lesiones detectadas con LB y DFD fueron enumeradas y registradas en una base de datos online. Se analizó histopatológicamente cada lesión por separado. En 148 pacientes se tomaron además biopsias múltiples aleatorias (BMA). Resultados: Se biopsiaron un total de 1.659 lesiones: 522 identificadas con DFD y LB, 237 sólo con DFD, 19 sólo con LB y 881 BMA. De 600 neoplasias diagnosticadas DPD detectó 563, LB 441 y BMA 29 (20 CIS). La tasa media de sobredetección de DPD sobre LB fue del 31,9% globalmente, pero en el caso del CIS fue del 209%. La sensibilidad de DFD fue 93,8% y la de LB 78,2%. La especificidad de DFD fue 81,5% y la de LB 90,5%. En el 23% de los pacientes se detectó al menos una lesión neoplásica más con DFD que con LB. Conclusión: La RTU con hexaminolevulinato mejora el rendimiento diagnóstico y la calidad de la resección del cáncer vesical superficial, especialmente del CIS. La mayor sensibilidad de DFD es a costa de una menor especificidad. En nuestro estudio BMA rescató algunos falsos negativos de DPD para detectar CIS. © 2011 AEU. Publicado por Elsevier España, S.L. Todos los derechos reservados.
Introduction Non-muscle-invasive bladder cancer (NMIBC) represents around 80% of bladder tumors. It is estimated that the bladder tumor is the second most prevalent urological neoplasia in the western world, 139,500 new cases being diagnosed each year in Europe.1,2 Its clinical course is characterized by the succession of relapses, which require the performance of multiple endoscopic actions with diagnostic and therapeutic aim. This makes the NMIBC one of the neoplastic processes that consumes the most economic and sanitary resources, both at a material and human level. The main goal of surgical treatment is the eradication of neoplastic tissue and obtaining of tissue material to determine the degree of infiltration into the bladder wall. Several studies have shown that the quality of endoscopic resection of the NMIBC varies depending on the center, the surgeon’s experience, the size of the tumor, its multifocality, the visibility conditions, etc. This variability in the quality of the resection will determine a higher or lower rate of residual tumors, and a higher or lower false early relapse rate.3 But, in addition to resection as complete as possible, it is necessary to confirm or not the existence of carcinoma in situ (CIS) or other flat lesions of uncertain malignant potential (dysplasias), which carry a high recurrence rate and a high risk of progression. The problem is that often these flat lesions are not visible with conventional cystoscopy. In recent years, the use of photodynamic diagnosis (PDD) methods has spread, based on the use of photosensitizing agents, which instilled in the bladder are taken avidly by the neoplastic tissues. Once incised with a beam of blue light (280---440 nm), they emit a typical red fluorescence that evidences tumoral lesions that with conventional
white light (WL) would not be visible. Since the mid-90s, 5aminolevulinate (5-ALA) had been used as a photosensitizing agent, not very much spread in our area, but with extensive experience in the center of Europe.4---6 In the last 4 years, derivative of 5-ALA has started to be used, hexyl ester 5-ALA or hexaminolevulinate (HAL), which has better pharmacodynamic characteristics, greater lipid solubility, and a higher intensity of the light signal than 5-ALA with an equal tissue concentration.7 All this has meant that HAL is currently the agent of choice in photodynamic diagnosis of NMIBC, with a clear superiority in the detection of lesions compared with conventional white light.8---14 The main aim of this work is to determine in our area the diagnostic yield of photodynamic diagnosis (PDD) in non-muscle invasive bladder carcinoma, compared with conventional white light (WL). As secondary objectives, we intend to analyze the usefulness of PDD in the detection of CIS, as well as validate and compare the results with those of the published series.
Materials and methods Patient recruitment Between December 2006 and December 2009, 305 patients belonging to 7 different urology departments of different public hospitals in the Spanish health network were recruited: Son Llatzer Hospital, Son Dureta Hospital (currently Son Espases), Central Hospital of Asturias, Arnau de Vilanova Hospital, Marina Baixa Hospital, Valencian Institute of Oncology, and Royo Villanova Hospital. These centers belong to the BLUE (Blue Light Urologic Endoscopy) group,
Hexaminolevulinate photodynamic diagnosis in non-muscle invasive bladder cancer created to bring together experience in the photodynamic diagnosis. All the patients had a diagnosis of suspected ultrasound or cystoscopic bladder tumor, or had a positive urine cytological study combined with the absence of ultrasound or cystoscopic findings. All were adults and, in the case of being women, the possibility of gestation was excluded. All the patients who had undergone instillations of chemotherapeutics or intravesical BCG on the previous 6 weeks were also excluded, as well as those with inflammatory disease of the bladder, bladder lithiasis, active bleeding, or history of pelvic radiotherapy. No patient with proven or suspicious hypersensitivity to hexaminolevulinate or any of its components took part in the study, and nor did patients with alterations of heme group metabolism or porphyrias.
Instillation of hexaminolevulinate Sixty minutes before the cystoscopy, we proceeded to intravesical instillation of 50 ml of hexaminolevulinate at a concentration of 1.7 mg/ml (HEXVIX® , General Electric Healthcare). The content was maintained instillated within the bladder for at least 60 min, and we proceeded to the carrying out of postural changes to ensure proper impregnation of the bladder walls.
White/blue light cystoscopy and biopsy taking Under general anesthesia or rachianesthesia, we proceeded to the bladder mapping under conventional white light (WL) illumination, drawing and listing the injuries in a bladder diagram. Later, the same mapping was conducted under blue light between 280 and 440 nm (PDD) wavelength. We proceeded to resection or biopsy of all the injuries, recording in each case if it was visible under WL, PDD, or both. Two special categories were also established, the ‘‘suspicious lesion’’ or pinkish (without clear red fluorescence) and the ‘‘unseen injury’’ (not visible in the first observation with WL, visible with PDD, and visible in a second observation with WL). In addition, in 148 patients multiple random biopsies (MRB) were taken.
Histological analysis Each lesion was sent separately for histological analysis. We agreed to use the 2002 UICC15 classification for the local stage, and the 1999 WHO16 classification for the tumor grade: papillary urothelial neoplasm of low malignant potential (PUNLMP), grade 1, grade 2, and grade 3 urothelial carcinoma. Non-neoplastic tissue was classified into 5 categories: healthy tissue, fibrous tissue, inflammatory tissue, hyperplasia, and papilloma.
Electronic database and statistical analysis We used an online database designed specifically for the study, in which the various researchers were recording the patients included, the characteristics of the lesions detected endoscopically, and the pathological result for
441
each lesion biopsied. Descriptive statistics of the data was performed using SPSS software® .
Results Analysis regarding patients A total of 305 endoscopic procedures were performed on 305 patients, 270 men, and 35 women. The mean age was 66.69 years, with an age range between 39 and 93. One hundred fifty-five patients had not had previous episodes of NMIBC resection, 92 had their first relapse, 27 the second one, 22 the third one, 4 the fourth one, 3 the fifth one, 1 the sixth one, and another one the tenth relapse. In 72 of the 305 patients (23.6%), at least one more lesion was diagnosed with PDD than with WL. The blue light was responsible for 16 new diagnoses of CIS. In 61 of the 148 patients in whom multiple random biopsies were performed, some neoplasias that had not been detected with PDD were detected with WL or multiple random biopsies (58.8% negative predictive value [NPV] of white light per patient). However, only 24 of these 148 patients were diagnosed with a neoplasia with WL or multiple random biopsies that had not been detected with PDD (83.7% negative predictive value [NPV] of blue light per patient). In no case were there adverse drug reactions. In two patients bladder filling was not tolerated for more than 20 min, and the instilled solution was expelled early. In both cases, there were PDD false negatives and CIS was diagnosed by multiple random biopsies.
Analysis regarding lesions In total, 1659 valid samples were obtained for histopathological examination. Of these, 522 were lesions identified both with PDD and WL (including 12 doubtful lesions with blue light), 19 lesions visible only with WL, and 237 lesions visible only with PDD (including three doubtful lesions with PDD, and 8 inadvertent lesions with WL). The other 881 samples were multiple random biopsies obtained in 148 patients. Of the 1659 biopsies taken, 600 corresponded to neoplastic lesions. Table 1 shows the distribution of neoplasias detected with WL, PDD, and multiple random biopsies according to tumor stage and degree of differentiation. The average rate of PDD oversensing versus WL was 31.9%, and 209% for the CIS in particular (Table 1). In this study, 15 lesions were classified as doubtful or pinkish, three of them visible with WL and 12 not visible with WL. Of these 15 lesions, only one corresponded histopathologically to neoplasia (pTaG1), while the remaining 14 were false positives. Also, 8 lesions were classified as inadvertent, i.e., initially not visible with WL, and visible with PDD, but in a second observation with WL, they became evident. These lesions corresponded to 5 pT1G3, a CIS, a low-grade dysplasia, and a hyperplasia. With regard to the 19 lesions only evidentiable with WL and not visible with PDD, these corresponded to a neoplastic process in 8 cases: 2 pT1G2, 2 pT1G3, 2 pT2aG3, a PUNLMP, and a low-grade dysplasia. Table 2 shows the global distribution of the biopsies based on whether or not they were visible lesions with WL or with PDD, and whether they corresponded or not to neoplastic
442 Table 1
J.P. Burgués et al. Distribution of lesions diagnosed with WL, PDD, and multiple random biopsies and PDD overdetection rate on WL.
Pathological anatomy
Total
Low-grade dysplasia 16 CIS 88 pTaG1 150 pTaG2 37 pTaG3 33 pT1G1 54 pT1G2 24 pT1G3 69 pT2aG3 22 pT2bG3 7 pT3G3 2 pT4G3 2 PUNLMP 46 pTxGx 50 Total neoplastic lesions 600 Total non-neoplastic lesions 1059 Total lesions 1659
Diagnosis with WL
Diagnosis with PDD
Diagnosis with multiple random biopsies
Additional lesions with PDD
Overdetection rate with PDD
6 22 117 33 26 45 24 62 22 4 2 2 32 44 441 100 541
8 68 149 37 33 54 22 66 20 7 2 2 45 50 563 196 759
7 20 1 ------1 --------------29 852 881
2 46 32 4 7 9 −2 4 −2 3 0 0 13 6 126 107 233
33.3% 209% 27.3% 12.1% 26.9% 20% −8.3% 6.4% −9.1% 75% 0% 0% 40.6% 13.6% 31.9% -----
lesions. There were 37 false negatives in total, that is, nonfluorescent lesions that proved to be malignant (2.2% of all the injuries). These false negatives corresponded to the 8 neoplasias mentioned in the previous paragraph, visible with WL, but not with PDD, and to the 29 tumor lesions diagnosed with multiple random biopsies. With WL, there were 159 false negatives (9.6% of all the lesions). 196 false positives were counted with PDD (25.8% of the 759 visible lesions) and 100 with WL (18.5% of the 541 visible lesions). The most common pathological anatomy of false positives, both of PDD and WL, was the subepithelial inflammatory infiltrate (Table 3). Of the 233 lesions detected additionally with blue light and not visible with white light, 126 were true positives (54.8%) and 107 false positives (45.1%). The most frequently found neoplastic additional lesion was CIS (46 lesions), followed by pTaG1 (32 lesions), and PUNLMP (13 lesions). From Table 2 we calculated the sensitivity, specificity, positive predictive value, and negative predictive value of PDD and WL for the detection of tumor lesions. The PDD overall sensitivity was 93.8%, while that of the WL was Table 2 Distribution of the 1659 lesions according to whether they were visible or not with WL/PDD and if they corresponded to tumor/non-tumor lesions.
78.2%. The PDD overall specificity was 81.5% and 90.5% WL. The positive predictive value (PPV) of PDD was 74.2% overall and 81.5% for WL. For lesions additionally detected with blue light, the PPV was only 54.8%. The NPV of PDD was 95.9% overall and 85.8% WL. We also calculated the sensitivity of PDD and WL for each of the types of neoplastic lesion (Table 4). For the detection of CIS, WL sensitivity was 25%, whereas that of the PDD was 77.2%.
Discussion The analysis of our series shows a clear superiority of PDD over WL in the diagnosis of neoplastic lesions regarding the bladder. The average rate of overdetection is 31.9%, superimposable or even better than the initially reported in the first studies8---14 (Table 5). If we stratify the sensitivity by tumor stage, in the diagnosis of tumors in pTa and pT1 stage, the sensitivity in our series is slightly higher than those reported historically, which is reinforced by a large sample size (Table 6). In the case of CIS, the relatively low performance of PDD (sensitivity 77.2%) compared with figures around 95% in the rest of studies is striking, especially when the capacity of oversensing with regard to the WL is 209%. The explanation
White light (WL)
Tumor histology Non-tumor histology
Visible
Non-visible
Table 3 Pathological anatomy of white light (WL) and blue light (PDD) false positives.
441 100
159 959
Pathological anatomy
WL
PDD
Total
Hyperplasia Fibrosis Inflammation Normal mucosa Papillomas Total
7 3 58 26 6 100 (18.5%)
13 3 110 63 7 196 (25.8%)
20 6 168 89 13
Blue light (PDD)
Tumor histology Non-tumor histology
Fluorescent
Non-fluorescent
563 196
37 863
Hexaminolevulinate photodynamic diagnosis in non-muscle invasive bladder cancer
443
Table 4 Sensitivity of white light (WL) and blue light (PDD) stratified by stage (UICC 2002) and tumor grade (WHO 1999).
Table 6 Comparison of blue light sensitivity (PDD) for the diagnosis of pTa, pT1 and CIS in different series.
Pathological anatomy Number of lesions
WL sensitivity
PDD sensitivity
Authors (quote)
Low-grade dysplasia CIS pTaG1 pTa G2 pTa G3 pT1 G1 pT1 G2 pT1 G3 pT2a G3 pT2b G3 pT3 G3 pT4 G3 PUNLMP pTx Gx Total neoplasias
37.5% 25% 78% 89.1% 78.7% 83.3% 100% 89.8% 100% 57.1% 100% 100% 69.5% 88% 78.2%
50% 77.2% 99.3% 100% 100% 100% 91.6% 95.6% 90.9% 100% 100% 100% 97.8% 100% 93.8%
pTa Jocham et al. 20059 Loidl et al. 200511 Schmidbauer et al. 200412 BLUE group
16 88 150 37 33 54 24 69 22 7 2 2 46 50 600
is that almost 50% of our patients underwent multiple random biopsies and, thus, 20 additional CIS were diagnosed. In the other studies, the percentage of patients who underwent multiple random biopsies was lower. If we excluded these 20 CIS cases, the sensitivity in our series would be 100%; but this exclusion would be misleading, since in the analysis of results it has been outlined that there were 37 false negatives with PDD, 29 of them due to the diagnosis of neoplasia with the multiple random biopsy. Therefore, we can conclude that the high diagnostic ability of the blue light in the search for the CIS does not exclude the performance of multiple random biopsies when traditionally indicated. In our study, the use of PDD associated with multiple random biopsies led to the diagnosis of 100% of cases of CIS. We can also conclude that this diagnostic superiority of PDD against WL in the search for CIS is unquestionable, and it makes it its main indication. In this regard, we can mention the study by Fradet,17 which reflects that the incidence of CIS in the series of NMIBC is usually 5%, whereas with the use of PDD it reaches 17%. This fact reinforces the theory that the incidence of CIS is as high as the intensity of the effort in its search.18,19 From the demonstrated ability of oversensing of CIS derives the indication of the use of PDD with HAL in cases of positive cytology and absence of tumor findings in conventional cystoscopy (evidence level IIb and grade of recommendation B).20---22 Recently, the consensus Table 5 series.
pT1 Jocham et al. 20059 Loidl et al. 200511 Schmidbauer et al. 200412 BLUE group CIS Jocham et al. 20059 Loidl et al. 200511 Schmidbauer et al. 200412 BLUE group
n
PDD sensitivity
Rate of overdetection
143 88 376
137 (96%) 85 (96.6%) 365 (97%)
11% 5.6% 9%
220
219 (99.5%)
22.1%
26 46 82
25 (96%) 42 (91.3%) 79 (96%)
--10.8% ---
147
142 (96.5%)
18.1%
62 51 177
59 (95%) 47 (92.2%) 172 (97%)
27% 21.6% 39%
88
68 (77.2%)
209%
document on the use of PDD with HAL, published by a panel of European experts based on scientific evidence, seamlessly recommends the use of this technique in the diagnosis of CIS.23 When analyzing the diagnostic capacity of PDD in pTa and pT1 papillary neoformations, the results are not as spectacular as in the flat lesions, but at all times, the PDD is superior to WL. This superiority justifies its use in the resection of non-flat or papillary NMIBC, as an improvement in lesion detection and, therefore, in the quality of the resection, is obtained. In our series, the overdetection rate for Ta lesions was 22% and 18% for T1. With a more complete resection, it is logical that there is a decrease in the number of recurrences. To date, there are no long-term studies of PDD with HAL, there is only a multicenter study that in the short term (9 months) shows a 16% relative reduction in the recurrence rate compared to WL.24 There are long-term studies with 5-ALA that show a decrease of residual tumors, a longer time to the first relapse, and a longer time of disease-free survival with up to 8 years followup.25 Recent publications equate the results obtained with 5-ALA to those obtained with HAL.26 In our series, there were 72 patients (23% of the total) in which at least one additional
Sensitivity and specificity of white light (WL) and blue light (PDD) and global rate of overdetection in the different
Authors (quote)
n
Jichlinski et al. 20038 Jocham et al. 20059 Loidl et al. 200511 Schmidbauer et al. 200412 Grossman et al. 200714 BLUE group 2011
52 146 45 211 18 305
WL sensitivity 46% 77% 75.1 78% 84.5% 78.2%
PDD sensitivity 76% 96% 87.1% 96% 95% 93%
Global rate of overdetection 30% 19% 12% 19% 10.5% 31.9%
444 lesion was found with PDD that otherwise would not have been resected. Thus, fewer recurrences are expected than if we had used convetional WL. Today we can say that the literature supports the use of PDD in a selected majority of endovesical diagnostic and therapeutic interventions.20,22,23 In the consensus document of the European expert panel its use is recommended in every initial resection of NMIBC to ensure a more complete resection, as well as in patients with relapsed tumor that in its initial resections have not undergone PDD. On the other hand, due to lack of evidence, it is not recommended for use in a systematic way in the resection of recurrences if this technique has already been used in the first resection, except in the case of multifocal tumors or with aggressive clinical behavior.23 There is also evidence of medium and long-term reduced health spendings in series using 5-ALA. This is because the decrease in surgical events compensates beyond expectatives the cost of the photosensitizing substance.27,28 Based on the studies with 5-ALA predictive models have been made by applying the cost of HAL and assuming a behavior similar to 5-ALA.29 The relatively low specificity of PDD against WL is due to the high rate of false positives.30 In our series, the specificity of blue light was 81% and the rate of false positives was 25.9%; these values were well below those obtained with WL (89% specificity and 18.4% false positive rate). If we analyze other series,8,13 we obtain 74 and 63% specificity with PDD versus 93% and 74% with WL respectively. In our case, the study began when the participating urologists already had experience with the technique, and the performance of unnecessary biopsies has possibly been minimized. It is a proven fact that specificity can improve as the surgeon’s experience in the interpretation of fluorescence images increases. In any case, the alternative for the diagnosis of CIS is the multiple random biopsies, and with them the number of ‘unnecessary’ negative biopsies is much higher (in our series 861 negative biopsies of a total of 881). Finally, as in every diagnostic test, we want to evaluate their negative predictive value; i.e., the probability that when not seeing any positive lesion with blue light we are certain of the absence of neoplasia. This parameter can only be evaluated if the patient is subjected to multiple random biopsies as a control system. In the series of Jichlinski et al.8 the negative predictive value was 86% for PDD and 77% for WL. In the subgroup of patients in our series in which multiple random biopsies were performed, we obtain an 83.7% NPV for blue light and 58.8% for WL. These data support the usefulness of HAL in the monitoring and control of the post-BCG CIS-like lesions. If we also add the multiple random biopsies taking to the PDD, then the NPV is 100%. The relevance of this parameter lies in that a high NPV doubtlessly bids on bladder preservation with a greater safety margin. In summary, with the results of our multicenter study we can conclude that the use of PDD with HAL improves the detection of bladder neoplastic lesions, more so in the case of CIS, and to a lesser extent in Ta or T1 tumors. The detection of additional lesions with blue light allows for a better quality of NMIBC resection, an advantage which we believe must be assumed at the expense of a greater number of false positives. In our study, these false positives were less than in previous series, probably due to the greater experience of the urologists at the present time. Despite the relatively
J.P. Burgués et al. low diagnostic profitability of multiple random biopsies, it has been the key in the recue of false negatives with PDD. For this reason, when there is high suspicion of CIS (positive cytology with negative cystoscopy and normal study of the upper urinary tract) or in the post-BCG follow-up, we still support the use of multiple random biopsies associated with HAL fluorescence cystoscopy.
Conflict of interest The authors declare that they have no conflict of interest.
References 1. Ferlay J, Parkin DM, Steliarova-Foucher E. Estimates of cancer incidence and mortality in Europe in 2008. Eur J Cancer. 2010;46:765---81. 2. Babjuk M, Oosterlinck W, Sylvester R, Kaasinen E, Böhle A, Palou J, et al. Guidelines on TaT1 (non-muscle invasive) bladder cancer. EAU Guidelines. In: Edition presented at the 25th EAU Annual Congress. 2010. 3. Brausi M, Collette L, Kurth K, van der Meijden AP, Oosterlinck W, Witjes JA, et al. Variability in the recurrence rate at first follow-up cystoscopy after TUR in stage Ta T1 transitional cell carcinoma of the bladder: a combined analysis of seven EORTC studies. Eur Urol. 2002;41:523---31. 4. Jichlinski P, Forrer M, Mizeret J, Glanzmann T, Braichotte D, Wagnieres G, et al. Clinical evaluation of a method for detecting superficial surgical transitional cell carcinoma of the bladder by light-induced fluorescence of protoporphyrin IX following the topical application of 5-aminolevulinic acid: preliminary results. Lasers Surg Med. 1997;20:402---8. 5. Koenig F, McGovern FJ, Larne R, Enquist H, Schomacker KT, Deutsch TF. Diagnosis of bladder carcinoma using protoporphyrin IX fluorescence induced by 5-aminolaevulinic acid. BJU Int. 1999;83:129---35. 6. Kriegmair M, Baumgartner R, Knuchel R, Stepp H, Hofstadter F, Hofstetter A. Detection of early bladder cancer by 5aminolevulinic acid induced porphyrin fluorescence. J Urol. 1996;155:105---9 [discussion 109---110]. 7. Lange N, Jichlinski P, Zellweger M, Forrer M, Marti A, Guillou L, et al. Photodetection of early human bladder cancer based on the fluorescence of 5-aminolaevulinic acid hexylester-induced protoporphyrin IX: a pilot study. Br J Cancer. 1999;80:185---93. 8. Jichlinski P, Guillou L, Karlsen SJ, Malmstrom PU, Jocham D, Brennhovd B, et al. Hexyl aminolevulinate fluorescence cystoscopy: new diagnostic tool for photodiagnosis of superficial bladder cancer----a multicenter study. J Urol. 2003;170:226---9. 9. Jocham D, Witjes F, Wagner S, Zeylemaker B, van Moorselaar J, Grimm MO, et al. Improved detection and treatment of bladder cancer using hexaminolevulinate imaging: a prospective, phase III multicenter study. J Urol. 2005;174:862---6 [discussion 866]. 10. Witjes JA, Moonen PM, van der Heijden AG. Comparison of hexaminolevulinate based flexible and rigid fluorescence cystoscopy with rigid white light cystoscopy in bladder cancer: results of a prospective Phase II study. Eur Urol. 2005;47:319---22. 11. Loidl W, Schmidbauer J, Susani M, Marberger M. Flexible cystoscopy assisted by hexaminolevulinate induced fluorescence: a new approach for bladder cancer detection and surveillance? Eur Urol. 2005;47:323---6. 12. Schmidbauer J, Witjes F, Schmeller N, Donat R, Susani M, Marberger M. Improved detection of urothelial carcinoma in situ with hexaminolevulinate fluorescence cystoscopy. J Urol. 2004;171:135---8.
Hexaminolevulinate photodynamic diagnosis in non-muscle invasive bladder cancer 13. Durek C, Wagner S, Zeylemaker B, Van Moorselaar J, Witjes F, Grimm MO, et al. The significance of hexyl 5-aminolevulinate hydrochloride based fluorescence cystoscopy in treatment decisions. Results of a prospective phase 3 multicenter study. Eur Urol. 2004. 14. Grossman HB, Gomella L, Fradet Y, Morales A, Presti J, Ritenour C, et al. A phase III, multicenter comparison of hexaminolevulinate fluorescence cystoscopy and white light cystoscopy for the detection of superficial papillary lesions in patients with bladder cancer. J Urol. 2007;178:62---7. 15. Sobin DH, Wittekind CH. TNM classification of malignant tumours. 6th ed New York: Wiley-Liss; 2002. 16. Mostofi FK, Davis CJ, Sesterhenn IA. Histological typing of urinary bladder tumours. In: International histological classification of bladder tumours. N◦ 10. World Health Organization. Berlin: Springer; 1999. 17. Fradet Y, Grossman HB, Gomella L, Lerner S, Cookson M, Albala D, et al. A comparison of hexaminolevulinate fluorescence cystoscopy and white light cystoscopy for the detection of carcinoma in situ in patients with bladder cancer: a phase III, multicenter study. J Urol. 2007;178:68---73 [discussion 73]. 18. Karl A, Tritschler S, Stanislaus P, Gratzke C, Tilki D, Strittmatter F, et al. Positive urine cytology but negative whitelight cystoscopy: an indication for fluorescence cystoscopy? BJU Int. 2009;103:484---7. 19. Ray ER, Chatterton K, Khan MS, Thomas K, Chandra A, O’Brien TS. Hexylaminolaevulinate ‘blue light’ fluorescence cystoscopy in the investigation of clinically unconfirmed positive urine cytology. BJU Int. 2009;103:1363---7. 20. Oliva Encina J, Rioja Sanz C. Photodynamic diagnosis (PDD) in non-muscle invasive bladder cancer. Literature review. Actas Urol Esp. 2009;33:965---75. 21. Jocham D, Stepp H, Waidelich R. Photodynamic diagnosis in urology: state-of-the-art. Eur Urol. 2008;53:1138---48. 22. Kausch I, Sommerauer M, Montorsi F, Stenzl A, Jacqmin D, Jichlinski P, et al. Photodynamic diagnosis in nonmuscle-invasive bladder cancer: a systematic review and
23.
24.
25.
26.
27.
28.
29.
30.
445
cumulative analysis of prospective studies. Eur Urol. 2009;57: 595---606. Witjes JA, Redorta JP, Jacqmin D, Sofras F, Malmstrom PU, Riedl C, et al. Hexaminolevulinate-guided fluorescence cystoscopy in the diagnosis and follow-up of patients with non-muscle-invasive bladder cancer: review of the evidence and recommendations. Eur Urol. 2010. Stenzl A, Burger M, Fradet Y, Mynderse LA, Soloway MS, Witjes JA, et al. Hexaminolevulinate guided fluorescence cystoscopy reduces recurrence in patients with nonmuscle invasive bladder cancer. J Urol. 2010;184:1907---13. Denzinger S, Burger M, Walter B, Knuechel R, Roessler W, Wieland WF, et al. Clinically relevant reduction in risk of recurrence of superficial bladder cancer using 5-aminolevulinic acid-induced fluorescence diagnosis: 8-year results of prospective randomized study. Urology. 2007;69:675---9. Burger M, Stief CG, Zaak D, Stenzl A, Wieland WF, Jocham D, et al. Hexaminolevulinate is equal to 5-aminolevulinic acid concerning residual tumor and recurrence rate following photodynamic diagnostic assisted transurethral resection of bladder tumors. Urology. 2009;74:1282---6. Burger M, Zaak D, Stief CG, Filbeck T, Wieland WF, Roessler W, et al. Photodynamic diagnostics and noninvasive bladder cancer: is it cost-effective in long-term application? A Germany-based cost analysis. Eur Urol. 2007;52:142---7. Malmstrom PU, Hedelin H, Thomas YK, Thompson GJ, Durrant H, Furniss J. Fluorescence-guided transurethral resection of bladder cancer using hexaminolevulinate: analysis of health economic impact in Sweden. Scand J Urol Nephrol. 2009;43:192---8. Sievert KD, Zyczynski T, Sweet A, Rößler DW, Stenzl A. Hexvix fluorescence cystoscopy for non-invasive bladder cancer management: an economic model of the impact on German healthcare costs. Value Health. 2007;10:PCN 40. Brausi MA. Arguments against the use of fluorescence for the diagnosis of non-muscle-invasive bladder tumours (NMIBT). Eur Urol. 2008;7:430---3.
Actas Urológicas Españolas www.elsevier.es/actasuro
ORIGINAL ARTICLE
Hexaminolevulinate photodynamic diagnosis in non-muscle invasive bladder cancer: Experience of the BLUE group夽 J.P. Burgués a,∗ , G. Conde b , J. Oliva c , J.M. Abascal d , I. Iborra e , M. Puertas f , F. Ordo˜ no g , Grupo BLUE (Blue Light Urologic Endoscopy) a
Servicio de Urología, Hospital Universitario Son Espases, Palma de Mallorca, Spain Servicio de Urología, Hospital Son Llàtzer, Palma de Mallorca, Spain c Servicio de Urología, Hospital Royo Villanova, Zaragoza, Spain d Servicio de Urología, Hospital Universitario Central de Asturias, Oviedo, Spain e Servicio de Urología, Instituto Valenciano de Oncología, Valencia, Spain f Servicio de Urología, Hospital Marina Baixa, La Vila Joiosa, Alicante, Spain g Servicio de Urología, Hospital Arnau de Vilanova, Valencia, Spain b
Received 25 February 2011; accepted 11 March 2011
KEYWORDS Photodynamic diagnosis; Hexaminolevulinate; Non-muscle invasive bladder cancer
Abstract Objectives: Photodynamic diagnosis (PDD) with hexaminolevulinate has been recently used to improve detection of non-muscle invasive bladder cancer. Our main purpose was to quantify the benefit of PDD versus conventional white light cystoscopy (WL) in our area. Materials and methods: Fluorescence-guided cystoscopy using hexaminolevulinate was performed at the time of the transurethral resection (TUR) in 305 patients from 7 Spanish hospitals. All lesions found with WL and PDD were numbered and recorded in an online database. Each lesion was sent separately for pathology analysis. Random biopsies were also obtained in 148 patients. Results: A total of 1659 lesions were biopsied: 522 were identified with PDD and WL, 237 only with PDD, 19 only with WL and 881 random biopsies. Of the 600 tumors, PDD detected 563, WL 441 and random biopsies 29 (20 CIS). The mean overdetection rate for PDD over WL was 31.9% for all types of lesions, but it was 209% for carcinoma in situ (CIS). Sensitivity was 93.8% for PDD and 78.2% for WL. Specificity was 81.5% for PDD and 90.5% for WL. In 23% of patients, PDD detected at least one more neoplastic lesion than with WL. Conclusions: Hexaminolevulinate fluorescence cystoscopy improves detection and resection of non-muscle invasive bladder cancer, especially of CIS. Sensitivity of PDD is higher than WL, but specificity is lower. In our study, random biopsies were able to detect some CIS not visible under PDD. © 2011 AEU. Published by Elsevier España, S.L. All rights reserved.
夽 Please cite this article as: Burgues JP, et al. Diagnóstico fotodinámico con hexaminolevulinato en el cancer vesical no músculo invasivo: experiencia del grupo BLUE. Actas Urol Esp. 2011;35:439---445. ∗ Corresponding author. E-mail address: [email protected] (J.P. Burgués).
2173-5786/$ – see front matter © 2011 AEU. Published by Elsevier España, S.L. All rights reserved.
440
PALABRAS CLAVE Diagnóstico fotodinámico; Hexaminolevulinato; Cáncer vesical no músculo invasivo
J.P. Burgués et al.
Diagnóstico fotodinámico con hexaminolevulinato en el cáncer vesical no músculo invasivo: experiencia del grupo BLUE Resumen Objetivos: El diagnóstico fotodinámico (DFD) con hexaminolevulinato se ha empezado a utilizar recientemente para mejorar la detección del cáncer vesical no músculo invasivo. Nuestro objetivo principal fue comparar el rendimiento diagnóstico de PDD frente a endoscopia con luz blanca convencional (LB) en nuestro medio. Material y métodos: Se realizó cistoscopia fluorescente con hexaminolevulinato en el momento de la RTU a 305 pacientes de 7 hospitales espa˜ noles. Todas las lesiones detectadas con LB y DFD fueron enumeradas y registradas en una base de datos online. Se analizó histopatológicamente cada lesión por separado. En 148 pacientes se tomaron además biopsias múltiples aleatorias (BMA). Resultados: Se biopsiaron un total de 1.659 lesiones: 522 identificadas con DFD y LB, 237 sólo con DFD, 19 sólo con LB y 881 BMA. De 600 neoplasias diagnosticadas DPD detectó 563, LB 441 y BMA 29 (20 CIS). La tasa media de sobredetección de DPD sobre LB fue del 31,9% globalmente, pero en el caso del CIS fue del 209%. La sensibilidad de DFD fue 93,8% y la de LB 78,2%. La especificidad de DFD fue 81,5% y la de LB 90,5%. En el 23% de los pacientes se detectó al menos una lesión neoplásica más con DFD que con LB. Conclusión: La RTU con hexaminolevulinato mejora el rendimiento diagnóstico y la calidad de la resección del cáncer vesical superficial, especialmente del CIS. La mayor sensibilidad de DFD es a costa de una menor especificidad. En nuestro estudio BMA rescató algunos falsos negativos de DPD para detectar CIS. © 2011 AEU. Publicado por Elsevier España, S.L. Todos los derechos reservados.
Introduction Non-muscle-invasive bladder cancer (NMIBC) represents around 80% of bladder tumors. It is estimated that the bladder tumor is the second most prevalent urological neoplasia in the western world, 139,500 new cases being diagnosed each year in Europe.1,2 Its clinical course is characterized by the succession of relapses, which require the performance of multiple endoscopic actions with diagnostic and therapeutic aim. This makes the NMIBC one of the neoplastic processes that consumes the most economic and sanitary resources, both at a material and human level. The main goal of surgical treatment is the eradication of neoplastic tissue and obtaining of tissue material to determine the degree of infiltration into the bladder wall. Several studies have shown that the quality of endoscopic resection of the NMIBC varies depending on the center, the surgeon’s experience, the size of the tumor, its multifocality, the visibility conditions, etc. This variability in the quality of the resection will determine a higher or lower rate of residual tumors, and a higher or lower false early relapse rate.3 But, in addition to resection as complete as possible, it is necessary to confirm or not the existence of carcinoma in situ (CIS) or other flat lesions of uncertain malignant potential (dysplasias), which carry a high recurrence rate and a high risk of progression. The problem is that often these flat lesions are not visible with conventional cystoscopy. In recent years, the use of photodynamic diagnosis (PDD) methods has spread, based on the use of photosensitizing agents, which instilled in the bladder are taken avidly by the neoplastic tissues. Once incised with a beam of blue light (280---440 nm), they emit a typical red fluorescence that evidences tumoral lesions that with conventional
white light (WL) would not be visible. Since the mid-90s, 5aminolevulinate (5-ALA) had been used as a photosensitizing agent, not very much spread in our area, but with extensive experience in the center of Europe.4---6 In the last 4 years, derivative of 5-ALA has started to be used, hexyl ester 5-ALA or hexaminolevulinate (HAL), which has better pharmacodynamic characteristics, greater lipid solubility, and a higher intensity of the light signal than 5-ALA with an equal tissue concentration.7 All this has meant that HAL is currently the agent of choice in photodynamic diagnosis of NMIBC, with a clear superiority in the detection of lesions compared with conventional white light.8---14 The main aim of this work is to determine in our area the diagnostic yield of photodynamic diagnosis (PDD) in non-muscle invasive bladder carcinoma, compared with conventional white light (WL). As secondary objectives, we intend to analyze the usefulness of PDD in the detection of CIS, as well as validate and compare the results with those of the published series.
Materials and methods Patient recruitment Between December 2006 and December 2009, 305 patients belonging to 7 different urology departments of different public hospitals in the Spanish health network were recruited: Son Llatzer Hospital, Son Dureta Hospital (currently Son Espases), Central Hospital of Asturias, Arnau de Vilanova Hospital, Marina Baixa Hospital, Valencian Institute of Oncology, and Royo Villanova Hospital. These centers belong to the BLUE (Blue Light Urologic Endoscopy) group,
Hexaminolevulinate photodynamic diagnosis in non-muscle invasive bladder cancer created to bring together experience in the photodynamic diagnosis. All the patients had a diagnosis of suspected ultrasound or cystoscopic bladder tumor, or had a positive urine cytological study combined with the absence of ultrasound or cystoscopic findings. All were adults and, in the case of being women, the possibility of gestation was excluded. All the patients who had undergone instillations of chemotherapeutics or intravesical BCG on the previous 6 weeks were also excluded, as well as those with inflammatory disease of the bladder, bladder lithiasis, active bleeding, or history of pelvic radiotherapy. No patient with proven or suspicious hypersensitivity to hexaminolevulinate or any of its components took part in the study, and nor did patients with alterations of heme group metabolism or porphyrias.
Instillation of hexaminolevulinate Sixty minutes before the cystoscopy, we proceeded to intravesical instillation of 50 ml of hexaminolevulinate at a concentration of 1.7 mg/ml (HEXVIX® , General Electric Healthcare). The content was maintained instillated within the bladder for at least 60 min, and we proceeded to the carrying out of postural changes to ensure proper impregnation of the bladder walls.
White/blue light cystoscopy and biopsy taking Under general anesthesia or rachianesthesia, we proceeded to the bladder mapping under conventional white light (WL) illumination, drawing and listing the injuries in a bladder diagram. Later, the same mapping was conducted under blue light between 280 and 440 nm (PDD) wavelength. We proceeded to resection or biopsy of all the injuries, recording in each case if it was visible under WL, PDD, or both. Two special categories were also established, the ‘‘suspicious lesion’’ or pinkish (without clear red fluorescence) and the ‘‘unseen injury’’ (not visible in the first observation with WL, visible with PDD, and visible in a second observation with WL). In addition, in 148 patients multiple random biopsies (MRB) were taken.
Histological analysis Each lesion was sent separately for histological analysis. We agreed to use the 2002 UICC15 classification for the local stage, and the 1999 WHO16 classification for the tumor grade: papillary urothelial neoplasm of low malignant potential (PUNLMP), grade 1, grade 2, and grade 3 urothelial carcinoma. Non-neoplastic tissue was classified into 5 categories: healthy tissue, fibrous tissue, inflammatory tissue, hyperplasia, and papilloma.
Electronic database and statistical analysis We used an online database designed specifically for the study, in which the various researchers were recording the patients included, the characteristics of the lesions detected endoscopically, and the pathological result for
441
each lesion biopsied. Descriptive statistics of the data was performed using SPSS software® .
Results Analysis regarding patients A total of 305 endoscopic procedures were performed on 305 patients, 270 men, and 35 women. The mean age was 66.69 years, with an age range between 39 and 93. One hundred fifty-five patients had not had previous episodes of NMIBC resection, 92 had their first relapse, 27 the second one, 22 the third one, 4 the fourth one, 3 the fifth one, 1 the sixth one, and another one the tenth relapse. In 72 of the 305 patients (23.6%), at least one more lesion was diagnosed with PDD than with WL. The blue light was responsible for 16 new diagnoses of CIS. In 61 of the 148 patients in whom multiple random biopsies were performed, some neoplasias that had not been detected with PDD were detected with WL or multiple random biopsies (58.8% negative predictive value [NPV] of white light per patient). However, only 24 of these 148 patients were diagnosed with a neoplasia with WL or multiple random biopsies that had not been detected with PDD (83.7% negative predictive value [NPV] of blue light per patient). In no case were there adverse drug reactions. In two patients bladder filling was not tolerated for more than 20 min, and the instilled solution was expelled early. In both cases, there were PDD false negatives and CIS was diagnosed by multiple random biopsies.
Analysis regarding lesions In total, 1659 valid samples were obtained for histopathological examination. Of these, 522 were lesions identified both with PDD and WL (including 12 doubtful lesions with blue light), 19 lesions visible only with WL, and 237 lesions visible only with PDD (including three doubtful lesions with PDD, and 8 inadvertent lesions with WL). The other 881 samples were multiple random biopsies obtained in 148 patients. Of the 1659 biopsies taken, 600 corresponded to neoplastic lesions. Table 1 shows the distribution of neoplasias detected with WL, PDD, and multiple random biopsies according to tumor stage and degree of differentiation. The average rate of PDD oversensing versus WL was 31.9%, and 209% for the CIS in particular (Table 1). In this study, 15 lesions were classified as doubtful or pinkish, three of them visible with WL and 12 not visible with WL. Of these 15 lesions, only one corresponded histopathologically to neoplasia (pTaG1), while the remaining 14 were false positives. Also, 8 lesions were classified as inadvertent, i.e., initially not visible with WL, and visible with PDD, but in a second observation with WL, they became evident. These lesions corresponded to 5 pT1G3, a CIS, a low-grade dysplasia, and a hyperplasia. With regard to the 19 lesions only evidentiable with WL and not visible with PDD, these corresponded to a neoplastic process in 8 cases: 2 pT1G2, 2 pT1G3, 2 pT2aG3, a PUNLMP, and a low-grade dysplasia. Table 2 shows the global distribution of the biopsies based on whether or not they were visible lesions with WL or with PDD, and whether they corresponded or not to neoplastic
442 Table 1
J.P. Burgués et al. Distribution of lesions diagnosed with WL, PDD, and multiple random biopsies and PDD overdetection rate on WL.
Pathological anatomy
Total
Low-grade dysplasia 16 CIS 88 pTaG1 150 pTaG2 37 pTaG3 33 pT1G1 54 pT1G2 24 pT1G3 69 pT2aG3 22 pT2bG3 7 pT3G3 2 pT4G3 2 PUNLMP 46 pTxGx 50 Total neoplastic lesions 600 Total non-neoplastic lesions 1059 Total lesions 1659
Diagnosis with WL
Diagnosis with PDD
Diagnosis with multiple random biopsies
Additional lesions with PDD
Overdetection rate with PDD
6 22 117 33 26 45 24 62 22 4 2 2 32 44 441 100 541
8 68 149 37 33 54 22 66 20 7 2 2 45 50 563 196 759
7 20 1 ------1 --------------29 852 881
2 46 32 4 7 9 −2 4 −2 3 0 0 13 6 126 107 233
33.3% 209% 27.3% 12.1% 26.9% 20% −8.3% 6.4% −9.1% 75% 0% 0% 40.6% 13.6% 31.9% -----
lesions. There were 37 false negatives in total, that is, nonfluorescent lesions that proved to be malignant (2.2% of all the injuries). These false negatives corresponded to the 8 neoplasias mentioned in the previous paragraph, visible with WL, but not with PDD, and to the 29 tumor lesions diagnosed with multiple random biopsies. With WL, there were 159 false negatives (9.6% of all the lesions). 196 false positives were counted with PDD (25.8% of the 759 visible lesions) and 100 with WL (18.5% of the 541 visible lesions). The most common pathological anatomy of false positives, both of PDD and WL, was the subepithelial inflammatory infiltrate (Table 3). Of the 233 lesions detected additionally with blue light and not visible with white light, 126 were true positives (54.8%) and 107 false positives (45.1%). The most frequently found neoplastic additional lesion was CIS (46 lesions), followed by pTaG1 (32 lesions), and PUNLMP (13 lesions). From Table 2 we calculated the sensitivity, specificity, positive predictive value, and negative predictive value of PDD and WL for the detection of tumor lesions. The PDD overall sensitivity was 93.8%, while that of the WL was Table 2 Distribution of the 1659 lesions according to whether they were visible or not with WL/PDD and if they corresponded to tumor/non-tumor lesions.
78.2%. The PDD overall specificity was 81.5% and 90.5% WL. The positive predictive value (PPV) of PDD was 74.2% overall and 81.5% for WL. For lesions additionally detected with blue light, the PPV was only 54.8%. The NPV of PDD was 95.9% overall and 85.8% WL. We also calculated the sensitivity of PDD and WL for each of the types of neoplastic lesion (Table 4). For the detection of CIS, WL sensitivity was 25%, whereas that of the PDD was 77.2%.
Discussion The analysis of our series shows a clear superiority of PDD over WL in the diagnosis of neoplastic lesions regarding the bladder. The average rate of overdetection is 31.9%, superimposable or even better than the initially reported in the first studies8---14 (Table 5). If we stratify the sensitivity by tumor stage, in the diagnosis of tumors in pTa and pT1 stage, the sensitivity in our series is slightly higher than those reported historically, which is reinforced by a large sample size (Table 6). In the case of CIS, the relatively low performance of PDD (sensitivity 77.2%) compared with figures around 95% in the rest of studies is striking, especially when the capacity of oversensing with regard to the WL is 209%. The explanation
White light (WL)
Tumor histology Non-tumor histology
Visible
Non-visible
Table 3 Pathological anatomy of white light (WL) and blue light (PDD) false positives.
441 100
159 959
Pathological anatomy
WL
PDD
Total
Hyperplasia Fibrosis Inflammation Normal mucosa Papillomas Total
7 3 58 26 6 100 (18.5%)
13 3 110 63 7 196 (25.8%)
20 6 168 89 13
Blue light (PDD)
Tumor histology Non-tumor histology
Fluorescent
Non-fluorescent
563 196
37 863
Hexaminolevulinate photodynamic diagnosis in non-muscle invasive bladder cancer
443
Table 4 Sensitivity of white light (WL) and blue light (PDD) stratified by stage (UICC 2002) and tumor grade (WHO 1999).
Table 6 Comparison of blue light sensitivity (PDD) for the diagnosis of pTa, pT1 and CIS in different series.
Pathological anatomy Number of lesions
WL sensitivity
PDD sensitivity
Authors (quote)
Low-grade dysplasia CIS pTaG1 pTa G2 pTa G3 pT1 G1 pT1 G2 pT1 G3 pT2a G3 pT2b G3 pT3 G3 pT4 G3 PUNLMP pTx Gx Total neoplasias
37.5% 25% 78% 89.1% 78.7% 83.3% 100% 89.8% 100% 57.1% 100% 100% 69.5% 88% 78.2%
50% 77.2% 99.3% 100% 100% 100% 91.6% 95.6% 90.9% 100% 100% 100% 97.8% 100% 93.8%
pTa Jocham et al. 20059 Loidl et al. 200511 Schmidbauer et al. 200412 BLUE group
16 88 150 37 33 54 24 69 22 7 2 2 46 50 600
is that almost 50% of our patients underwent multiple random biopsies and, thus, 20 additional CIS were diagnosed. In the other studies, the percentage of patients who underwent multiple random biopsies was lower. If we excluded these 20 CIS cases, the sensitivity in our series would be 100%; but this exclusion would be misleading, since in the analysis of results it has been outlined that there were 37 false negatives with PDD, 29 of them due to the diagnosis of neoplasia with the multiple random biopsy. Therefore, we can conclude that the high diagnostic ability of the blue light in the search for the CIS does not exclude the performance of multiple random biopsies when traditionally indicated. In our study, the use of PDD associated with multiple random biopsies led to the diagnosis of 100% of cases of CIS. We can also conclude that this diagnostic superiority of PDD against WL in the search for CIS is unquestionable, and it makes it its main indication. In this regard, we can mention the study by Fradet,17 which reflects that the incidence of CIS in the series of NMIBC is usually 5%, whereas with the use of PDD it reaches 17%. This fact reinforces the theory that the incidence of CIS is as high as the intensity of the effort in its search.18,19 From the demonstrated ability of oversensing of CIS derives the indication of the use of PDD with HAL in cases of positive cytology and absence of tumor findings in conventional cystoscopy (evidence level IIb and grade of recommendation B).20---22 Recently, the consensus Table 5 series.
pT1 Jocham et al. 20059 Loidl et al. 200511 Schmidbauer et al. 200412 BLUE group CIS Jocham et al. 20059 Loidl et al. 200511 Schmidbauer et al. 200412 BLUE group
n
PDD sensitivity
Rate of overdetection
143 88 376
137 (96%) 85 (96.6%) 365 (97%)
11% 5.6% 9%
220
219 (99.5%)
22.1%
26 46 82
25 (96%) 42 (91.3%) 79 (96%)
--10.8% ---
147
142 (96.5%)
18.1%
62 51 177
59 (95%) 47 (92.2%) 172 (97%)
27% 21.6% 39%
88
68 (77.2%)
209%
document on the use of PDD with HAL, published by a panel of European experts based on scientific evidence, seamlessly recommends the use of this technique in the diagnosis of CIS.23 When analyzing the diagnostic capacity of PDD in pTa and pT1 papillary neoformations, the results are not as spectacular as in the flat lesions, but at all times, the PDD is superior to WL. This superiority justifies its use in the resection of non-flat or papillary NMIBC, as an improvement in lesion detection and, therefore, in the quality of the resection, is obtained. In our series, the overdetection rate for Ta lesions was 22% and 18% for T1. With a more complete resection, it is logical that there is a decrease in the number of recurrences. To date, there are no long-term studies of PDD with HAL, there is only a multicenter study that in the short term (9 months) shows a 16% relative reduction in the recurrence rate compared to WL.24 There are long-term studies with 5-ALA that show a decrease of residual tumors, a longer time to the first relapse, and a longer time of disease-free survival with up to 8 years followup.25 Recent publications equate the results obtained with 5-ALA to those obtained with HAL.26 In our series, there were 72 patients (23% of the total) in which at least one additional
Sensitivity and specificity of white light (WL) and blue light (PDD) and global rate of overdetection in the different
Authors (quote)
n
Jichlinski et al. 20038 Jocham et al. 20059 Loidl et al. 200511 Schmidbauer et al. 200412 Grossman et al. 200714 BLUE group 2011
52 146 45 211 18 305
WL sensitivity 46% 77% 75.1 78% 84.5% 78.2%
PDD sensitivity 76% 96% 87.1% 96% 95% 93%
Global rate of overdetection 30% 19% 12% 19% 10.5% 31.9%
444 lesion was found with PDD that otherwise would not have been resected. Thus, fewer recurrences are expected than if we had used convetional WL. Today we can say that the literature supports the use of PDD in a selected majority of endovesical diagnostic and therapeutic interventions.20,22,23 In the consensus document of the European expert panel its use is recommended in every initial resection of NMIBC to ensure a more complete resection, as well as in patients with relapsed tumor that in its initial resections have not undergone PDD. On the other hand, due to lack of evidence, it is not recommended for use in a systematic way in the resection of recurrences if this technique has already been used in the first resection, except in the case of multifocal tumors or with aggressive clinical behavior.23 There is also evidence of medium and long-term reduced health spendings in series using 5-ALA. This is because the decrease in surgical events compensates beyond expectatives the cost of the photosensitizing substance.27,28 Based on the studies with 5-ALA predictive models have been made by applying the cost of HAL and assuming a behavior similar to 5-ALA.29 The relatively low specificity of PDD against WL is due to the high rate of false positives.30 In our series, the specificity of blue light was 81% and the rate of false positives was 25.9%; these values were well below those obtained with WL (89% specificity and 18.4% false positive rate). If we analyze other series,8,13 we obtain 74 and 63% specificity with PDD versus 93% and 74% with WL respectively. In our case, the study began when the participating urologists already had experience with the technique, and the performance of unnecessary biopsies has possibly been minimized. It is a proven fact that specificity can improve as the surgeon’s experience in the interpretation of fluorescence images increases. In any case, the alternative for the diagnosis of CIS is the multiple random biopsies, and with them the number of ‘unnecessary’ negative biopsies is much higher (in our series 861 negative biopsies of a total of 881). Finally, as in every diagnostic test, we want to evaluate their negative predictive value; i.e., the probability that when not seeing any positive lesion with blue light we are certain of the absence of neoplasia. This parameter can only be evaluated if the patient is subjected to multiple random biopsies as a control system. In the series of Jichlinski et al.8 the negative predictive value was 86% for PDD and 77% for WL. In the subgroup of patients in our series in which multiple random biopsies were performed, we obtain an 83.7% NPV for blue light and 58.8% for WL. These data support the usefulness of HAL in the monitoring and control of the post-BCG CIS-like lesions. If we also add the multiple random biopsies taking to the PDD, then the NPV is 100%. The relevance of this parameter lies in that a high NPV doubtlessly bids on bladder preservation with a greater safety margin. In summary, with the results of our multicenter study we can conclude that the use of PDD with HAL improves the detection of bladder neoplastic lesions, more so in the case of CIS, and to a lesser extent in Ta or T1 tumors. The detection of additional lesions with blue light allows for a better quality of NMIBC resection, an advantage which we believe must be assumed at the expense of a greater number of false positives. In our study, these false positives were less than in previous series, probably due to the greater experience of the urologists at the present time. Despite the relatively
J.P. Burgués et al. low diagnostic profitability of multiple random biopsies, it has been the key in the recue of false negatives with PDD. For this reason, when there is high suspicion of CIS (positive cytology with negative cystoscopy and normal study of the upper urinary tract) or in the post-BCG follow-up, we still support the use of multiple random biopsies associated with HAL fluorescence cystoscopy.
Conflict of interest The authors declare that they have no conflict of interest.
References 1. Ferlay J, Parkin DM, Steliarova-Foucher E. Estimates of cancer incidence and mortality in Europe in 2008. Eur J Cancer. 2010;46:765---81. 2. Babjuk M, Oosterlinck W, Sylvester R, Kaasinen E, Böhle A, Palou J, et al. Guidelines on TaT1 (non-muscle invasive) bladder cancer. EAU Guidelines. In: Edition presented at the 25th EAU Annual Congress. 2010. 3. Brausi M, Collette L, Kurth K, van der Meijden AP, Oosterlinck W, Witjes JA, et al. Variability in the recurrence rate at first follow-up cystoscopy after TUR in stage Ta T1 transitional cell carcinoma of the bladder: a combined analysis of seven EORTC studies. Eur Urol. 2002;41:523---31. 4. Jichlinski P, Forrer M, Mizeret J, Glanzmann T, Braichotte D, Wagnieres G, et al. Clinical evaluation of a method for detecting superficial surgical transitional cell carcinoma of the bladder by light-induced fluorescence of protoporphyrin IX following the topical application of 5-aminolevulinic acid: preliminary results. Lasers Surg Med. 1997;20:402---8. 5. Koenig F, McGovern FJ, Larne R, Enquist H, Schomacker KT, Deutsch TF. Diagnosis of bladder carcinoma using protoporphyrin IX fluorescence induced by 5-aminolaevulinic acid. BJU Int. 1999;83:129---35. 6. Kriegmair M, Baumgartner R, Knuchel R, Stepp H, Hofstadter F, Hofstetter A. Detection of early bladder cancer by 5aminolevulinic acid induced porphyrin fluorescence. J Urol. 1996;155:105---9 [discussion 109---110]. 7. Lange N, Jichlinski P, Zellweger M, Forrer M, Marti A, Guillou L, et al. Photodetection of early human bladder cancer based on the fluorescence of 5-aminolaevulinic acid hexylester-induced protoporphyrin IX: a pilot study. Br J Cancer. 1999;80:185---93. 8. Jichlinski P, Guillou L, Karlsen SJ, Malmstrom PU, Jocham D, Brennhovd B, et al. Hexyl aminolevulinate fluorescence cystoscopy: new diagnostic tool for photodiagnosis of superficial bladder cancer----a multicenter study. J Urol. 2003;170:226---9. 9. Jocham D, Witjes F, Wagner S, Zeylemaker B, van Moorselaar J, Grimm MO, et al. Improved detection and treatment of bladder cancer using hexaminolevulinate imaging: a prospective, phase III multicenter study. J Urol. 2005;174:862---6 [discussion 866]. 10. Witjes JA, Moonen PM, van der Heijden AG. Comparison of hexaminolevulinate based flexible and rigid fluorescence cystoscopy with rigid white light cystoscopy in bladder cancer: results of a prospective Phase II study. Eur Urol. 2005;47:319---22. 11. Loidl W, Schmidbauer J, Susani M, Marberger M. Flexible cystoscopy assisted by hexaminolevulinate induced fluorescence: a new approach for bladder cancer detection and surveillance? Eur Urol. 2005;47:323---6. 12. Schmidbauer J, Witjes F, Schmeller N, Donat R, Susani M, Marberger M. Improved detection of urothelial carcinoma in situ with hexaminolevulinate fluorescence cystoscopy. J Urol. 2004;171:135---8.
Hexaminolevulinate photodynamic diagnosis in non-muscle invasive bladder cancer 13. Durek C, Wagner S, Zeylemaker B, Van Moorselaar J, Witjes F, Grimm MO, et al. The significance of hexyl 5-aminolevulinate hydrochloride based fluorescence cystoscopy in treatment decisions. Results of a prospective phase 3 multicenter study. Eur Urol. 2004. 14. Grossman HB, Gomella L, Fradet Y, Morales A, Presti J, Ritenour C, et al. A phase III, multicenter comparison of hexaminolevulinate fluorescence cystoscopy and white light cystoscopy for the detection of superficial papillary lesions in patients with bladder cancer. J Urol. 2007;178:62---7. 15. Sobin DH, Wittekind CH. TNM classification of malignant tumours. 6th ed New York: Wiley-Liss; 2002. 16. Mostofi FK, Davis CJ, Sesterhenn IA. Histological typing of urinary bladder tumours. In: International histological classification of bladder tumours. N◦ 10. World Health Organization. Berlin: Springer; 1999. 17. Fradet Y, Grossman HB, Gomella L, Lerner S, Cookson M, Albala D, et al. A comparison of hexaminolevulinate fluorescence cystoscopy and white light cystoscopy for the detection of carcinoma in situ in patients with bladder cancer: a phase III, multicenter study. J Urol. 2007;178:68---73 [discussion 73]. 18. Karl A, Tritschler S, Stanislaus P, Gratzke C, Tilki D, Strittmatter F, et al. Positive urine cytology but negative whitelight cystoscopy: an indication for fluorescence cystoscopy? BJU Int. 2009;103:484---7. 19. Ray ER, Chatterton K, Khan MS, Thomas K, Chandra A, O’Brien TS. Hexylaminolaevulinate ‘blue light’ fluorescence cystoscopy in the investigation of clinically unconfirmed positive urine cytology. BJU Int. 2009;103:1363---7. 20. Oliva Encina J, Rioja Sanz C. Photodynamic diagnosis (PDD) in non-muscle invasive bladder cancer. Literature review. Actas Urol Esp. 2009;33:965---75. 21. Jocham D, Stepp H, Waidelich R. Photodynamic diagnosis in urology: state-of-the-art. Eur Urol. 2008;53:1138---48. 22. Kausch I, Sommerauer M, Montorsi F, Stenzl A, Jacqmin D, Jichlinski P, et al. Photodynamic diagnosis in nonmuscle-invasive bladder cancer: a systematic review and
23.
24.
25.
26.
27.
28.
29.
30.
445
cumulative analysis of prospective studies. Eur Urol. 2009;57: 595---606. Witjes JA, Redorta JP, Jacqmin D, Sofras F, Malmstrom PU, Riedl C, et al. Hexaminolevulinate-guided fluorescence cystoscopy in the diagnosis and follow-up of patients with non-muscle-invasive bladder cancer: review of the evidence and recommendations. Eur Urol. 2010. Stenzl A, Burger M, Fradet Y, Mynderse LA, Soloway MS, Witjes JA, et al. Hexaminolevulinate guided fluorescence cystoscopy reduces recurrence in patients with nonmuscle invasive bladder cancer. J Urol. 2010;184:1907---13. Denzinger S, Burger M, Walter B, Knuechel R, Roessler W, Wieland WF, et al. Clinically relevant reduction in risk of recurrence of superficial bladder cancer using 5-aminolevulinic acid-induced fluorescence diagnosis: 8-year results of prospective randomized study. Urology. 2007;69:675---9. Burger M, Stief CG, Zaak D, Stenzl A, Wieland WF, Jocham D, et al. Hexaminolevulinate is equal to 5-aminolevulinic acid concerning residual tumor and recurrence rate following photodynamic diagnostic assisted transurethral resection of bladder tumors. Urology. 2009;74:1282---6. Burger M, Zaak D, Stief CG, Filbeck T, Wieland WF, Roessler W, et al. Photodynamic diagnostics and noninvasive bladder cancer: is it cost-effective in long-term application? A Germany-based cost analysis. Eur Urol. 2007;52:142---7. Malmstrom PU, Hedelin H, Thomas YK, Thompson GJ, Durrant H, Furniss J. Fluorescence-guided transurethral resection of bladder cancer using hexaminolevulinate: analysis of health economic impact in Sweden. Scand J Urol Nephrol. 2009;43:192---8. Sievert KD, Zyczynski T, Sweet A, Rößler DW, Stenzl A. Hexvix fluorescence cystoscopy for non-invasive bladder cancer management: an economic model of the impact on German healthcare costs. Value Health. 2007;10:PCN 40. Brausi MA. Arguments against the use of fluorescence for the diagnosis of non-muscle-invasive bladder tumours (NMIBT). Eur Urol. 2008;7:430---3.