Future Oncology
Review
Adjuvant and neoadjuvant radiotherapy for bladder cancer: revisited Mohamed Saad Zaghloul Radiation Oncology Department, Children’s Cancer Hospital (57357), Egypt & National Cancer Institute, Cairo University, 1 Sekket El Emam, Sayeda Zainab, Cairo, Egypt n Tel.: +20 225 351 500 (ext. 1192) n Fax: +20 223 619 036 n
[email protected]
To date, radical cystectomy has continued to be the treatment of choice for muscle-invasive bladder cancer. It is associated with a 5-year disease-free survival rate ranging from 27–55% . This outcome is significantly worse when reporting upon locally advanced cases. The independent prognostic factors include: tumor stage, grade, pelvic nodal involvement and some other additional factors. Beside the higher reported incidence of distant metastasis, local recurrence either alone or combined with systemic relapse has been shown to be experienced by 23–50% of locally advanced patients – a rate that was much more frequent than previously believed. Nonrandomized trials of preoperative radiotherapy have suggested improved survival rates. However, only one out of the six randomized preoperative trials in the literature published in English has proved to be significant. On the other hand, the only randomized trial and most retrospective studies dealing with postoperative radiotherapy revealed a significant increase in disease-free survival. Late complications of post operative radiotherapy, contrary to former belief, were acceptable and generally depended upon the volume of the irradiated normal tissues and the radiotherapy techniques used. Most of these adjuvant or neoadjuvant reports were performed in the 1970s and 1980s using conventional radiation techniques. Modern radiotherapy, delivering higher doses to the tumor while saving a significant amount of the surrounding normal structure, has not been rigorously tested. However, these techniques have already succeeded in improving treatment end results in other pelvic tumors. Background
Bladder cancer represents a significant worldwide health problem with an estimated 356,370 new cases and 146,000 deaths reported globally for the year 2002 [1] . Although the majority of bladder cancers, present with disease confined to the superficial layers of the bladder wall, approximately 20–40% of the patients will present with or subsequently develop invasive cancer. Transitional cell carcinomas (TCC; also known as urothelial carcinoma) represented more than 90% of cystectomy specimens worldwide [2] . In areas where schistosomiasis is endemic, urothelial cancer represents approximately 50% of bladder cancers, while the other subtypes represents the remaining percentage [3] . In this article, a detailed analysis of the results of radical cystectomy, the treatment of choice for muscle-invasive bladder cancer and its recurrences is attempted. Factors affecting these recurrences and the different methods adopted to decrease them were systematically explored in the English language publications. Adjuvant and neoadjuvant radiotherapy were performed in the 1970s and 1980s of the last century without many convincing results. 10.2217/FON.10.82 © 2010 Future Medicine Ltd
However, modern radiotherapy, was established in the 1990s, in combination with surgery and/or chemotherapy was successful in improving the results of different pelvic tumors [4] . Exploration of the feasibility and applicability of modern radiotherapy as adjunctive to surgery is the main aim of this article. Radical cystectomy
The treatment of choice for muscle-invasive bladder cancer continued for decades to be radical cystectomy. The radical cystectomy procedure includes removal of the bladder, seminal vesicles and prostate, together with perivesical fat and peritoneal coverage, in addition to bilateral endopelvic lymphadenopathy in male patients. In females, it includes removal of the bladder, its perivesical fat and peritoneal covering, urethra, uterus, ovary and anterior wall of the vagina (anterior pelvic excenteration) [2,5,6] . Smith and Whitmore demonstrated that the most common locations for nodal metastases include the obturator (74%), external iliac (65%) and the internal iliac nodes (19%) [7] . The primary drainage starts from the external Future Oncol. (2010) 6(7), 1177–1191
Keywords adaptive radiotherapy radiotherapy n bladder cancer n IGRT n neoadjuvant radiotherapy n prognostic factor n radical cystectomy n radiotherapy technique n
n adjuvant
part of
ISSN 1479-6694
1177
Review
Zaghloul
and internal iliac, as well as the obturator sites; secondary drainage is from the common iliac sites; and tertiary drainage from the trigone and posterior bladder wall is to the presacral nodes [8] . Vazina et al. reported 24% of patients (42 out of 176) demonstrating pathologic lymph node metastases. The most common sites of nodal metastases were the external iliac and hypogastric/obturator regions [8] . In addition, 5% of patients had presacral nodal involvement and 9% had nodal disease above the common iliac bifurcation. The extrapelvic node involvement was invariably associated with endopelvic nodal disease [9] . Nagele et al. also demonstrated that in N1 patients (single LN, 2 cm at greatest diameter), metastatic disease is almost exclusively localized within the pelvis, and skipped lesions are absent in N2 patients (single LN 2–5 cm in diameter) [10] . Skipped lesions in the form of metastasis above the aortic bifurcation with negative endopelvic nodal findings were not detected in any extended lymphadenectomy series [11–13] . However, the question of whether the number of removed nodes increases the likelihood of detecting occult nodal metastases remains a controversial issue. Herr et al. indicated that the probability of detecting lymph node metastases would be clearly correlated with the total number of removed nodes [12] . On the other hand, Abdel-Latif et al., [11] and Leissner et al. [13] were not able to confirm such a correlation. Many studies reported favorable results (>60% achieved 5‑year survival) for patients with pathologically organ-confined disease [2,5,6,14–16] .
However, the results were significantly worse when reporting upon locally advanced tumors (pT3N0, pT4aN0 or with pelvic nodal involvement). These worse outcomes were experienced by both schistosoma-associated and nonassociated bladder cancer patients, when compared stage by stage [2,5,6,14–16] . Although it is believed that aberrant differentiation (other than urothelial) leads to even worse results, many authors reported similar results when comparing the same stage in urothelial (TCC) and nonurothelial (squamous or adenocarcinoma) cases (Table 1) . Rogers et al. reported a 5‑year progression-free survival of 60 (± 2%) after radical cystectomy for TCC and 55 (± 11%) for squamous cell carcinoma (SCC) [17] . This difference was statistically insignificant. Patients with TCC or SCC had statistically significant higher progression-free survival rates than non-TCC or non-SCC patients, including those having adenocarcinoma. It is worth noting that most of the publications concerned with primary bladder adenocarcinoma, including that of Rogers et al., reported a small number of patients with short-term follow-up. When reporting on considerable numbers of patients, adenocarcinoma proved to have similar 5‑year disease-free end results [18–20] . Almost similar results were reached by Lughezzani et al. using 17 Surveillances, Epidemiology and End Results (SEER) registries [21] . The only difference of statistical importance was that adenocarcinoma patients underwent radial cystectomy (RC) at more advanced disease stages than their UC counterparts. However, another recent study using a similar SEER database demonstrated
Table 1. 5-year survival rates for patients treated with radical cystectomy in large radical cystectomy series. Study (year)
Patients (n)
Type
P2a
P2b
P3
P4
LN+
Ghoneim et al. (1997) Cheng et al. (2000)† Stein et al. (2001) ‡ Medersbacher et al. (2003) Nishiyama et al. (2004) Takahashi et al. (2004) Zaghloul et al. (2006) § Dhar et al. (2006) Zaghloul et al. (2007) § Ghoneim et al. (2008) Niu et al. (2008) Monoharan et al. (2009)
1026 218 1054 507 1113 466 192 385 216 2720 356 432
TCC, SCC and adenocarcinoma TCC TCC TCC TCC, SCC and adenocarcinoma TCC Adenocarcinoma TCC Adenocarcinoma TCC, SCC and adenocarcinoma TCC, SCC and adenocarcinoma TCC
66 50 81 62 84 74 100
47 50 68 62 69 74 47
60
31 28 47 40 59 47 40 19 40 40 22 43
19 17 44 49 43 38 44 NM 30 29 0 17
23 11 35 26 35 30 31 9 31 27 8 22
[116]
Ho et al. (2009)
148
TCC
62
59
9
31
[117]
63 51 75 73
53 44
Ref. [5] [30] [2] [14] [15] [16] [19] [66] [20] [6] [115]
Values lying between columns P2a and P2b apply to both. † 10‑year all-cause survival. ‡ 48% of patients received neoadjuvant or adjuvant radiotherapy and/or chemotherapy. § 24% of patients received postoperative radiotherapy. LN+: Lymph nodal involvement; NM: Not mentioned; SCC: Squamous cell carcinoma; TCC: Transitional cell carcinoma.
1178
Future Oncol. (2010) 6(7)
future science group
Adjuvant & neoadjuvant radiotherapy for bladder cancer
that SCC was also more aggressive than urothelial cancer after adjusting for common prognostic factors, such as stage and grade [22] . The authors concluded that SCC was an independent predictor of mortality among patients with stage III and IV disease, and among patients with stage I and II disease who did not undergo cystectomy as part of their initial treatment regimen. Therefore, squamous histology was not associated with increased mortality among patients with Stage I and II tumors when treated with cystectomy. Recently, Ploeg et al. studied all invasive bladder cancer cases treated in The Netherlands during the period of 1995 and 2006 exclusive (28,807 patients) [23] . They reached the conclusions that the relative survival of muscleinvasive adenocarcinoma patients equals that of muscle-invasive urothelial carcinoma patients. For stage II and III disease, these patients have better outcomes [23] . Muscle-invasive squamous cell carcinoma patients show worse survival regardless of stage. Therefore, one can cautiously conclude that RC treatment end results were not affected by patients’ sex, age, tumor histology (TCC, SCC or adenocarcinoma) or etiology, but were affected by other prognostic factors. These data emphasized the extreme need to compare the results head to head, and to test all possible variables before reaching a conclusion. Therefore, the need for prospective comparative study is still warranted. Prognostic factors after radical cystectomy
Most authors considered pathological stage as the most important prognostic factor after RC. Some authors classify invasive tumors into organ-confined (including pT2a and pT2b) and extravesical (pT3 and pT4). Others determine the post-surgical T stage as the prognostic determinant [24] . Pelvic nodal involvement was also identified as a very strong independent prognostic factor. Nodal involvement led to 5‑year survival rates ranging between 8 and 35%, much lower than with uninvolved nodes (Table 1) . It was found that the extent of this nodal involvement determined the survival rate, since the 5‑year survival rate decreased from 44 to 27% and 0% in N1, N2 and N3 patients (metastasis in lymph node[s] more than 5 cm in greatest diameter), respectively [25] . Furthermore, it was shown in several prospective and retrospective studies that both the extent of lymphadenectomy and number of dissected nodes determined the survival rates, even in negative node patients [25,26] . Moreover, it was proven that the ratio of involved future science group
Review
to uninvolved dissected nodes determined the survival rate. Recent studies suggest that lymph node density, the ratio of positive lymph nodes to the total number of lymph nodes removed, is superior to TNM pathologic nodal staging and to the absolute number of positive lymph nodes in predicting disease-specific survival and recurrence-free survival [27] . This is evident in single institutional studies, as well as in SEER studies [28] . With a large number of patients in SEER study, Wright et al. showed that there was a continuous negative effect of the lymph node density with no cutoff point after which the mortality plateaued [29] . Few studies have included tumor size in the analysis of prognostic factors affecting bladder cancer end results. Cheng et al. showed that tumor size had a significant effect in determining the distant metastasis-free, cancer-specific and all-cause survival, both in the univariate and multivariate analyses [30] . However, Pollack et al. found that size had no influence on distant metastasis [31] . The cutoff size remained, logistically and practically, an important point to be determined. The histological grade was shown to be a significant prognostic effect in many studies [5,32–34] , yet some other studies failed to prove its effect [35–37] . Similarly, lympho vascular invasion (LVI) had a controversial effect as an independent prognostic factor in different studies [36,38,39] . On the other hand, Lotan et al. showed that LVI has its prognostic effect restricted to the node-negative patients [40] . The presence of LVI correlated significantly with increasing tumor stage (p = 0.002) and grade (p
