Locally recurrent nasopharyngeal carcinoma

Radiotherapy and Oncology 54 (2000) 135±142 www.elsevier.com/locate/radonline

Locally recurrent nasopharyngeal carcinoma Joseph Tung-Chieh Chang a,b,*, Lai-Chu See c, Chun-Ta Liao d, Shu-Hang Ng b,e, Cheng-Hsu Wang f, I-How Chen d, Ngan-Ming Tsang a,b, Chen-Kan Tseng a, Simon G. Tang a,b, Ji-Hong Hong a,b a

Department of Radiation Oncology, Chang Gung Memorial Hospital, Linkou, 33333 Taoyuan, Taiwan b Department of Medical Technology, Chang Gung University, Linkou, 33333 Taoyuan, Taiwan c Department of Public Health, Chang Gung University, Linkou, 33333 Taoyuan, Taiwan d Department of Otorhinolaryngology, Head and Neck Surgery, Chang Gung Memorial Hospital, Linkou, 33333 Taoyuan, Taiwan e Department of Radiology, Chang Gung Memorial Hospital, Linkou, 33333 Taoyuan, Taiwan f Department of Hematology and Oncology, Chang Gung Memorial Hospital, Linkou, 33333 Taoyuan, Taiwan Received 8 April 1999; received in revised form 22 October 1999; accepted 22 November 1999

Abstract Purpose: To assess the outcome of and determine prognostic factors for locally recurrent nasopharyngeal carcinoma (NPC) in patients treated with a second course of radiotherapy (RT). Materials and methods: From 1982 to 1995, 186 NPC patients, who had initially been treated in the Department of Radiation Oncology, Chang Gung Memorial Hospital-Linkou, developed local recurrence in the nasopharynx and were re-treated with RT ($20 Gy). The time from the initial RT to re-treatment ranged from 8 to 136 months (median: 23 months). All patients were treated with external RT and conformal radiotherapy was used in 35 patients after 1993. Fifteen received radiosurgery as a boost treatment. The RT dose at the nasopharyngeal tumor area ranged from 20 to 67.2 Gy (median 50 Gy). Eighty-two patients received one to eight courses of cisplatinbased chemotherapy in addition to RT. Results: The 1-, 3- and 5-year survival was 54.9, 22.1 and 12.4%, respectively. Patients whose tumor relapsed later than 2 years after the ®rst treatment had a better survival than those with earlier relapse (3-year survival: 30.1 vs. 10.8%; P ˆ 0:015), but the difference became insigni®cant in patients who received $ 50 Gy. Patients without evidence of intracranial invasion or cranial nerve palsy had better survival than those with such lesions (3-year survival: 30.9 vs. 3.7%; P ˆ 0:006). A re-treatment dose $50 Gy yielded better survival (3-year survival: 22.8 vs. 18.5%; P ˆ 0:003). Addition use of radiosurgery may improve survival. The use of chemotherapy did not improve survival. Conformal radiotherapy resulted in signi®cantly fewer severe complications than conventional RT. Conclusions: A repeat course of RT for locally recurrent NPC successfully prolongs survival in a signi®cant number of patients. Intracranial invasion and/or cranial nerve palsy and re-treatment dose affect the prognosis, with a dose of $50 Gy signi®cantly improving survival. Radiosurgery boost may also improve survival. Our preliminary data indicates that conformal radiotherapy may decrease the severity of radiation-induced complications. However; longer follow-up and larger sample size is necessary to document the ®ndings. Published by Elsevier Science Ireland Ltd. Keywords: Recurrent nasopharyngeal carcinoma; Radiotherapy; Radiosurgery; Chemotherapy

1. Introduction Nasopharyngeal carcinoma (NPC) is a common malignancy in Taiwan [16]. Radiotherapy is the major treatment modality, and the overall 5-year survival approaches 50± 70% [9,18,25,30]. The causes of treatment failure area either local recurrence or distant metastasis. The extent of the primary tumor is closely related to the degree of achievable local tumor control [9,20,30]. Primary tumor control * Corresponding author. 0167-8140/00/$ - see front matter. Published by Elsevier Science Ireland Ltd. PII: S 0167-814 0(99)00177-2

can be achieved in up to 80±90% of small tumors, but in only 50±60% of advanced tumors, despite total radiation doses are usually up to 70 Gy [25,27,29]. Recurrence of the primary tumor is one of the major causes of death from this disease, especially with locally advanced tumors. A major issue in discussions of the treatment of NPC is the management of these primary recurrent tumors. Re-treatment for local recurrent NPC with additional course of radiotherapy will induce high complications for high radiation dose. How to increase re-treatment effect but did not elevate too much complication is a major considera-

136

J.T.-C. Chang et al. / Radiotherapy and Oncology 54 (2000) 135±142

tion to treat the local recurrent NPC. There were a lot of new treatment modalities to approach the recurrent tumor. Conformal radiotherapy will reduce the radiation dose the normal tissue may reduce the complications. Radiosurgery has the bene®t of a rapid fall-off in radiation dose, increasing the radiation therapeutic ratio. The use of radiosurgery for recurrent NPC appears to have signi®cant potential [2,7]. Surgery for recurrent NPC has also been attempted with some success but the de®nite role remained uncertain [13,15,33]. Chemotherapy has been used for recurrent NPC, producing partial and complete responses in the range 60±70% [4,10,12]; however, long-term tumor control is rarely achieved. In this paper, we report our experience in treating recurrent NPC at Chang Gung Memorial Hospital, Linkou from 1982 to 1995. In particular, we tried to identify factors and the potential role of new treatment modalities for this disease.

2. Patients and methods 2.1. Patient characteristics From 1982 to 1995, 205 NPC patients initially treated in the Department of Radiation Oncology, Chang Gung Memorial Hospital, Linkou, had local failure in the nasopharynx and were re-treated with radiotherapy as the primary modality. We excluded the patients with combined relapse in neck and/or distant sites in this discussion. All patients had received .64.8 Gy in their initial course; details of the radiotherapy procedure have been described previously [3,29]. Recurrence was documented by biopsyproven tumor in the nasopharynx or by the onset of symptoms or signs coupled with new abnormalities CT scan or magnetic resonance imaging (MRI). Eleven patients had a MRI as a re-staging modality. Nineteen patients treated only palliatively with ,20 Gy were excluded from this study. Data for the remaining 186 patients were analyzed. The patients' characteristics and percentage were listed in Table 1. Ages ranged from 35 to 72 years (median: 47). One hundred and thirty-six patients were male and 50 female. All had good performance status (WHO scale ,2). The median time from last radiotherapy was 23 months (8±136 months). One hundred and ®ve patients had biopsyproven recurrence, and the remaining 81 were diagnosed clinically and with imaging. The criteria for diagnosis of recurrence without biopsy proof was based on progression of clinical symptoms and/or signs that compatible of new image ®ndings. The usual condition was the patients with tumor over deep submucosa, basilar skull or intracranial area adjacent to the critical structures that might have technique dif®culty or high morbidity probability for biopsy. All patients underwent complete blood count, blood chemistry, chest X-ray, bone scan, and head and neck CT scan or MRI to re-stage the tumor. CT or MRI images that were dif®cult

Table 1 Patients' characteristics and treatment modalities Number

Percentage

Total

186

100.0

Sex Male Female

136 50

73.1 26.9

Age #40 41±60 .60

12 113 28

24.2 60.8 15.1

Tissue proof Yes No

105 81

56.5 43.5

42 26 54 64

22.6 14.0 29.0 34.4

Radiosurgery No Yes

171 15

91.9 8.1

Chemotherapy No Yes

104 82

55.9 44.1

Re-radiotherapy duration (weeks) ,6 $6

123 63

66.1 33.9

Interval to recurrence (months) #24 .24

98 88

52.2 47.8

Retreatment radiation dose (Gy) ,50 $50

76 110

40.9 59.1

Recurrent T-stage rT1 rT2 rT3 rT4

to interpret were reviewed and discussed by a radiotherapist and a radiologist, retrospectively. Recurrent tumors were staged according to the 5th edition of UICC (International Union Against Cancer) and AJCC (American Joint Committee on Cancer) system [8]. Recurrence involved the nasopharynx only (rT1) in 42 patients, the nasopharynx with extension to the oropharynx or nasal cavity or parapharyngeal space (rT2) in 26, the bone, base of the skull or paranasal sinuses (rT3) in 54, the intracranial area or orbit and/or the presence of new or progressive cranial nerve palsy (rT4) in 64. 2.2. Treatment modalities All patients were treated by external radiotherapy as the major means of retreatment. Prior to 1993, the external radiotherapy was delivered by a 10 MV-photon beam with

J.T.-C. Chang et al. / Radiotherapy and Oncology 54 (2000) 135±142

bilateral opposing ®elds to cover the recurrent tumor site and margins according to a customized block in 151 patients. Thereafter, we began using 3D conformal radiotherapy, which was administered to 35 of the study subjects. Customized beam eye view blocks in each portal and coplanar or non-coplanar gantry techniques were used. Generally, the conformal radiotherapy was delivered and try to make the previous radiation areas (such as temporal lobes, T-M joints and middle ears) less than 60% of prescribed dose as possible. We also limited the brain stem radiation dose be less than 50% of isodose line. Five fractions per week of 1.7±2.0 Gy per fraction were given. The total external radiation dose speci®ed at the nasopharyngeal tumor area ranged from 20 to 67.2 Gy (median 50 Gy). The radiation dose speci®cation followed the ICRU recommendations. Fifteen patients received radiosurgery as a boost treatment after external radiotherapy, administered by a linac-based system with 10 MV-photon and planned by commercial software. A single 8±15 Gy fraction was prescribed to the 80% isodose shell. The usual radiosurgery used two isocenters (range 1± 4). Radiation to the optic apparatus was limited to ,8 Gy and to the brain-stem, ,15 Gy. There were no strict selection criteria for the use of radiosurgery but for the caring physician preference. Nine patients' tumor belonged to rT1, one was rT2, two rT3 and three rT4 who received radio surgery as boost. In general, the radiosurgery dose will decrease to 8±10 Gy for larger tumor (.20 cm 3) but 10± 14 Gy for smaller tumors (,20 cm 3). The median volume for radiosurgery was 8.3 cm 3 (ranged from 1.8 to 30.2). The median total re-irradiation dose was 50 Gy, ranging from 25 to 72 Gy. Eighty-two patients received 1±8 courses of cisplatin-based chemotherapy. Only six patients received brachytherapy as part of salvage treatment. The use of large fraction size of brachytherapy was part of primary radiotherapy caused severe complications in our previous experience [3]. That is the reason we did not select the brachytherapy as salvage treatment modality. 2.3. Follow-up and statistics The majority of patients were followed regularly up to December 1996. All the patients had 12 months minimal follow-up. The follow-up time was up to 136 months and the median was 42 months. Their status was determined by clinical follow-up or by letters. The former involved regular clinical examinations and further work-up, e.g. CT scan or MRI if indicated. In the last 5 years of the study period, direct ®beroptic endoscopy was used routinely to evaluate the nasopharynx. Patients who expired within 2 years from the end of retreatment without another speci®c cause were regarded as having died of cancer. The cut-off point of patients' age was #40 or not for the previous study in our department showed that age #40 or not was a signi®cant prognostic factor in NPC patients [29]. The chi-square test was used to compare differences between groups. Survival rates were calculated using the Kaplan and Meier method

137

[17]. Prognostic factors in survivals were examined with a log-rank test univariately, and with a Cox proportional hazard model multivariately [11]. In general, P-values #0.05 were considered statistically signi®cant. 3. Results 3.1. Survival The 1-, 2- and 3-year survival was 54.9, 30.2 and 22.1%, respectively. Four patients were lost to follow-up. One hundred and forty patients had died by the end of the study. Twenty patients subsequently developed distant metastasis. The usual sites of metastasis were lung (9), bone (8), liver (4), brain (1) and soft tissue (1). Three patients had metastasis in two different sites. Two patients died of unrelated causes: one in a car accident and the other of a cerebrovascular accident. Death was the only outcome evaluated. In our experience, it is very dif®cult on CT scans done following salvage RT to differentiate residual tumor from post-radiation granulation [24]. Whether or not there is tumor recurrence in the deep submucosa is also dif®cult to determine on direct ®ber-optic scope examination. Therefore, we did not include diseasefree survival or local control rate in our analysis. 3.2. Prognostic factor Univariate analysis of prognostic factors was performed and is summarized in Table 2. We did not ®nd that age (#40 or .40) or sex (male or female) made any difference in survival. Patients with rT4 tumors had the worst survival. The overall survival is shown in Fig. 1. Patients whose tumor relapsed later than 2 years after ®rst treatment had better survival than those with earlier relapse (3-year survival: 30.1 vs. 10.8%; P ˆ 0:015). Re-treatment doses $50 Gy performed better than ,50 Gy (3-year survival: 22.8 vs. 18.5%; P ˆ 0:012). Radiosurgery improved 3-year survival (52.0 vs. 28.7%; P ˆ 0:036), but chemotherapy did not signi®cantly improve survival. Multivariate analysis was performed to analyze various prognostic factors. The endpoint was overall survival. The following parameters were included in the Cox regression analysis: age (#40 vs. .40), sex, T stage at recurrence, use of radiosurgery or not, use of chemotherapy or not, total reradiation interval (,6 weeks vs. .6 weeks), interval to recurrence (#2 years vs. .2 years) and radiation dose (,50 Gy vs. $50 Gy). A stepwise forward procedure was used to analyze the data. The results are summarized in Table 3. We further evaluated the 110 patients who received $50 Gy, since the RT dose, a factor signi®cantly affecting survival, can be controlled by the physician. Selection bias might be occurred in prescribing less radiation dose to patients with a poor prognosis. We used a chi-square test (dose ,50 Gy vs. $50 Gy) to compare the difference. The results

138

J.T.-C. Chang et al. / Radiotherapy and Oncology 54 (2000) 135±142

Table 2 Overall survival after retreatment by prognostic factors Prognostic factor

1 Year

3 Years

Total

0.549

0.221

Sex Male Female

0.547 0.557

0.207 0.265

Age #40 .40

0.483 0.529

0.229 0.151

Recurrent T stage rT1 rT2 rT3 rT4

0.736 0.423 0.544 0.492

0.385 0.237 0.284 0.037

Radiosurgery No Yes

0.524 0.909

0.287 0.520

Chemotherapy No Yes

0.564 0.5176

0.211 0.232

Re-radiotherapy duration (weeks) ,6 $6

0.516 0.519

0.213 0.213

Interval to recurrence (months) #24 .24

0.449 0.595

0.108 0.301

Retreatment radiation dose (Gy) ,50 $50

0.370 0.626

0.185 0.228

P-value

0.319

0.543

0.003

0.036

0.962

0.015

0.012

are shown in Table 4. Patients receiving $50 Gy had RT for a longer period and were more likely to have received radiosurgery. There were no other differences between the two groups, including rT stage or relapse-free interval, indicat-

ing that there was no radiation dose selection bias between the two groups. The interval to recurrence became no more important in those patients who received retreatment radiation dose $50 Gy. Fifteen patients receiving radiosurgery as a boost had a better survival (52 vs. 20%) at 3 years, but the P-value was 0.09. In rT1-2 group, the patients who received radiosurgery had better 2-year survival than those who did not (50.0 vs. 41.4%); in rT3-4 group radiosurgery group also had better 3-year survival than no radiosurgery group (75.0 vs. 25.3%). The use of chemotherapy did not signi®cantly affect survival. These results were listed in Table 5. 3.3. Complications All patients experienced various degree of hearing impairment and trismus. Of 92 patients treated with conventional external RT, 13 patients were found to have brain necrosis on imaging studies; nine of these were symptomatic. This complication occurred in 11±32 months (median: 14 months). However, no brain necrosis occurred in patients treated with conformal RT after follow-up of 12 to 24 months (median: 16 months). Ten patients who developed hearing impairment required a hearing aid. Two had soft tissue necrosis. Severe trismus (mouth opening ,2 cm) was noted in nine patients. We de®ned severe complications as brain necrosis, hearing impairment needing hearing aid, soft tissue necrosis and severe trismus. Severe complications occurred 22.9% in the conventional RT group but in only 9.3% in the conformal RT group, a signi®cant difference (P ˆ 0:04). 4. Discussion Patients with recurrent nasopharyngeal carcinoma have a very poor prognosis if they are not treated. Yan et al. [34] reported only one 5-year survivor in a group of 276 patients with recurrent disease who received no further treatment. The 3- and 5-year survival in our subjects was 22.1 and Table 3 Multivariate analysis of prognostic factors Variables

Risk ratio

Recurrent T stage rT1 rT2 rT3 rT4

1 1.62 1.38 2.34

Retreatment radiotherapy dose (Gy) ,50 1 $50 0.59

Fig. 1. Overall survival of recurrent NPC patients.

Interval to recurrence (months) #24 1 .24 0.63

95% Con®dence interval

P-value

0.003 0.91±2.90 0.84±2.27 1.44±3.79 0.010 0.42±0.83 0.008 0.44±0.89

J.T.-C. Chang et al. / Radiotherapy and Oncology 54 (2000) 135±142 Table 4 The distribution of different re-treatment radiation dose group (,50 or $50 Gy) Variable

Total radiation dose

P-value

,50 Gy n ˆ 76

$50 Gy n ˆ 110

Sex Male Female

58 (76.3%) 18 (23.7%)

78 (70.9%) 32 (29.1%)

Age #40 .40

21 (27.6%) 55 (72.4%)

24 (21.8%) 86 (78.2%)

Recurrent T stage rT1 rT2 rT3 rT4

16 11 24 25

Radiosurgery No Yes

76 (100%) 0 (0.0%)

95 (86.4%) 15 (13.6%)

Chemotherapy No Yes

49 (64.5%) 27 (35.6%)

55 (50.0%) 55 (50.0%)

Re-radiotherapy duration (weeks) ,6 63 (82.9%) $6 13 (17.1%)

60 (54.6%) 50 (45.5%)

Interval to recurrence (months) #24 39 (51.3%) .24 37 (48.7%)

59 (53.6%) 51 (46.4%)

0.414

0.363

0.913 (21.1%) (14.5%) (31.6%) (32.9%)

26 15 30 39

(23.6%) (13.6%) (27.3%) (35.5%) 0.001

0.051

,0.000

0.755

12.4%. Most of our patients presented with advanced recurrence; almost 60% had rT3 or rT4 lesions, despite regular follow-up after treatment of their primary disease. Our data show that stage at which recurrence is diagnosed is the most signi®cant prognostic factor. The 2-year survival for patients with rT1 lesions was 49%, but for rT4 lesions, it was only 11%. Previous studies of recurrent nasopharyngeal cancer [5,19,21,22,26,31] have demonstrated a difference in survival depending on tumor stage. All the studies, however, were based on UICC 1992 staging [8] or Ho's 1978 staging [14]. Both these systems include tumors with basilar skull erosion, intracranial invasion and/or cranial nerve palsy in the same stage. Using a more recent staging system that split these two groups, we found that even with basilar skull involvement (rT3), 2-year survival was 39.0%, while more advanced disease (rT4) had a signi®cantly poorer 2-year survival of only 11%. CT scans often result in upgrading the primary lesion from early to late stage. Yu et al. [35] reported that 36% of patients with no clinical or plain X-ray evidence of basilar skull lesions did show skull involvement on CT scan. In our previous study, we showed both CT and MRI [23] are able to delineate the lesions of the base of the

139

skull but the latter is better choice to ®nd the minimal lesion. MRI had 20% of chance to up-grade stage from T2 to T3 or T3 to T4 according the new edition of staging system. Chua et al. [5] showed that tumor volume is important factor in determining primary tumor control. Since ours is a retrospective study, we did not have adequate data to analyze the effect of 3-dimensional tumor volume. However, to our best knowledge, it is not unusual to detect even a very small volume early rT3 tumor recurrence, especially with modern CT scans using a bone window imaging technique or even MRI. This may explain why we did not ®nd a worse prognosis for rT3 lesions as compared with rT2 tumors. We propose that lesions involving the basilar skull erosion should be differentiated from other advanced stage tumors when evaluating tumor control or survival. The new 1997 edition of UICC/AJCC staging system is a better system to use to compare the tumor extension in advanced stage tumor. The new staging system whether can tell the difference in different stage especially for large volume of T2 lesion to small volume of T3 is still waited for further studies to con®rm the new staging system. Time to relapse after initial RT is an important prognostic factor. Wang [30] has reported the 5-year survival for patients relapsing at .24 months was 66 vs. 13% for the relapse at #24 months, a signi®cant difference. Other groups have found survival differences at other cut-off points (Teo et al. [31], 18 months; Chua et al. [6], 12 months). We found a signi®cant difference at 24 months for our patients overall. This difference was lost, however, when we looked only at patients who received $50 Gy. This may suggest that a higher RT dose may compensate for other poor prognostic factors. Wang [32] suggested that more radiation dose is crucial for salvage in recurrent disease. Lee et al. [19] also demonstrated the bene®t of a higher dose; whether the recurrent lesions were early or advanced, a more radiation dose conferred better results. Teo et al. [31] studied the use of a higher re-radiation dose for patients with recurrence after radical RT. The complications were severe and the results were poor but they still achieved a 30% 5-year survival for early stage recurrence. We found that survival in advanced local recurrent NPC is poor but that higher radiation doses did yield better survival. The radiation dose $50 Gy did signi®cantly improve the survival. Buatti et al. [2] proposed the radiosurgery as a boost for recurrent nasopharyngeal carcinoma, noting that radiosurgery has a dose distribution advantage and fewer complications. They proposed the radiosurgery boost for #4 cm tumor after 40±50 Gy of external radiotherapy. Cmelak et al. [7] treated the tumor size up to 50 cm 3 and the local tumor control reached nearly 60% for recurrent NPC. We did not set volume limitation but the larger volume (.20 cm 3) would use the less dose to decrease the complication probability. Fifteen patients in our study who received radiosurgery as boost after 50 Gy of external RT in our series tended to have better survival, although the results

140

J.T.-C. Chang et al. / Radiotherapy and Oncology 54 (2000) 135±142

Table 5 Overall survival after retreatment by prognostic factors in patients group who received more than 50 Gy of retreatment radiation dose Variable

Number

1-Year survival

3-Year survival

Sex Male Female

78 32

0.614 0.654

0.203 0.283

Age #40 .40

24 86

0.622 0.628

0.215 0.232

Recurrent T stage rT1 rT2 rT3 rT4

26 15 30 39

0.830 0.467 0.650 0.549

0.378 0.207 0.342 0.121

Radiosurgery No Yes

95 15

0.592 0.909

0.200 0.520

Chemotherapy No Yes

55 55

0.621 0.633

0.225 0.228

Re-radiotherapy duration (weeks) ,6 $6

60 50

0.718 0.513

0.251 0.197

Interval to recurrence (months) #24 .24

59 51

0.591 0.668

0.148 0.290

did not reach statistical signi®cance. It tended to have selection bias for more percentage of radiosurgery group patients belonged to rT1-2. However, it also had survival difference tendency after dividing patients who received $50 Gy into two groups: rT1-2 and rT3-4. Further studies with the power to detect a difference may be warranted. It also needs a longer follow-up to understand the real complications of radiosurgery to the recurrent NPC patients It is likely that the complication rate is underestimated in this study, especially the incidence of brain complications before CT or MRI was routinely used in follow-up. The lower incidence of complications we saw in patients treated with conformal radiotherapy and radiosurgery may have been a function of a shorter follow-up. Certainly patients whose life is prolonged after radiotherapy have a signi®cant incidence of late complications. However, our preliminary data does suggest that radiosurgery may improve survival and conformal radiotherapy may decrease complications by Kaplan±Meier function model [17]. The use of combined radiosurgery as a boost to improve the tumor control and 3D conformal radiotherapy to decrease complications may be the next step in attempting salvage patients with recurrent NPC. Quality of life and cost-bene®t analyses are issues worthy of study to de®ne the value of advanced treatment techniques for these patients.

P-value 0.467

0.165

0.021

0.088

0.904

0.134

0.302

Surgery for recurrent NPC has also been attempted with some success [13,15,33]. Teo et al. [31] reported that nasopharyngectomy resulted in better survival than high dose reradiation in early recurrence, although selection bias could not be excluded. By contrast, Chua et al. [6] failed to demonstrate a de®nite difference between RT and surgery. While rapid debulking of the tumor may be achieved surgically, this is dif®cult to accomplish without signi®cant complications. A complete resection of most lesions is impossible; hence, despite an aggressive and hazardous surgical approach, the need for sterilizing doses of RT remains. Hsu et al. [15] reported eight of 24 patients had positive margins on permanent pathology section after operation. There were 70% of patients (14 of 20) who received postoperative radiotherapy as part of salvage because of doubtful or unclear resection margins in the study of Teo et al. [31]. Combining surgery and re-irradiation may increase morbidity and adversely affect the quality of life. We suggest that surgery is likely to be of bene®t only in a highly select group of patients with recurrent NPC. The ef®cacy of chemotherapy for recurrent disease, either as the sole treatment or in combination with RT is uncertain. Choo and Tank [4] had reported that more aggressive chemotherapy agents could get tumor response rate to 70% but the median survival was only 7 months. Gebbia

J.T.-C. Chang et al. / Radiotherapy and Oncology 54 (2000) 135±142

et al. [11] reported that use of cisplatin-based chemotherapy could reach about 65% of tumor response in recurrent and/ or metastatic NPC but the mean survival is only about 11 months. These results might suggest that the recurrent NPC be high responsive to chemotherapy but survival with chemotherapy alone is poor. It must be emphasized, however, that in our series, there was selection bias in that most patients given chemotherapy had other adverse prognostic factors, and chemotherapy was given either before reirradiation because of a more advanced disease, or after the retreatment with RT because of poor response. Our ®ndings do not support the routine use of chemotherapy. AlSarraf et al. [1] have shown an emerging role for concomitant chemoirradiation as the primary treatment of local and regionally advanced NPC. It is uncertain whether it is applicable to retreatment of purely local recurrences. Concomitant chemotherapy may have a role in enhancing the sensitivity of otherwise radioresistent cancer cells in recurrent disease. It is disturbing that, despite regular follow-up, many of our patients had recurrent tumors that were not detected until they were quite advanced. Sham et al. [28] had reported the value of clinical follow-up and frequent use of ®broscope examination. He also advocated the use of cross-sectional imaging techniques and the monitoring tumor markers as supplement. MRI [24] gives better soft tissue delineation and has the ability to differentiate between tumor recurrence and mature postradiation ®brosis. Routine use of MRI to follow post-radiotherapy NPC patients may afford earlier detection of recurrence and may therefore improve salvage rates. In recent years, we have used MRI instead of CT scan for routine follow-up. In conclusion, RT for locally recurrent NPC treated can yield signi®cant survival. Intracranial invasion and/or cranial nerve palsy suggests a poor prognosis, but this is not necessarily true of basilar skull involvement only. Doses $ 50 Gy are necessary to achieve better survival. Addition of radiosurgery boost may improve survival, while 3D conformal radiotherapy may decrease severe radiation complications. Combination therapy with 3D conformal RT, concomitant chemotherapy and a radiosurgery boost may be the next modalities to be tested in local recurrent NPC. References [1] Al-Sarraf M, LeBlanc M, Gin PGS, et al. Chemoradiotherapy versus radiotherapy in patients with advanced nasopharyngeal cancer: phase III randomized intergroup study 0099. J. Clin. Oncol. 1998;16:1310± 1317. [2] Buatti JM, Friedman WA, Bova FJ, Mendenhall WM. Linac radiosurgery for locally recurrent nasopharyngeal carcinoma: rationale and technique. Head Neck 1995;17:14±19. [3] Chang JT, See LC, Tang SG, Lee SP, Wang CC, Hong JH. The role of brachytherapy in early stage nasopharyngeal carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 1996;36:1019±1024. [4] Choo R, Tannock I. Chemotherapy for recurrent or metastatic carci-

[5]

[6]

[7] [8] [9]

[10] [11] [12] [13] [14]

[15] [16] [17] [18] [19] [20] [21]

[22] [23] [24] [25] [26] [27]

141

noma of the nasopharynx. A review of the Princess Margaret Hospital experience. Cancer 1991;68:2120±2124. Chua DTT, Sham JST, Kwong DLW, et al. Volumetric analysis of tumor extent in nasopharyngeal carcinoma and correlation with treatment outcome. Int. J. Radiat. Oncol. Biol. Phys. 1997;39:711± 719. Chua DTT, Sham JST, Kwong DLW, Wei WI, Au GKH, Choy D. Locally recurrent nasopharyngeal carcinoma: treatment results for patients with computed tomography assessment. J. Radiat. Oncol. Biol. Phys. 1998;41:379±386. Cmelak AJ, Cox RS, Adler JR, Fee WE, Gof®net DR. Radiosurgery for skull base malignancies and nasopharyngeal carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 1997;37:997±1003. Cooper JS, Cohen R, Steven RE. A comparison of staging system for nasopharyngeal carcinoma. Cancer 1998;83:213±219. Cooper JS, Scott C, Marcial V, et al. The relationship of nasopharyngeal carcinomas and second independent malignancies based on the radiation therapy oncology group experience. Cancer 1991;67:1673± 1677. Decker DA, Drelichman A, Al-Sarraf M, Crissman J, Reed ML. Chemotherapy for nasopharyngeal carcinoma: a 10-year experience. Cancer 1983;52:602±605. Elisa TL. Statistical methods for survival data analysis. Belmont, CA: Lifetime Learning Publications, 1980. Gebbia V, Zerillo G, Restivo G, et al. Chemotherapeutic treatment of recurrent and/or metastatic nasopharyngeal carcinoma: a retrospective analysis of 40 cases. Br. J. Cancer 1993;68:191±194. Hasselt CAV. The role of operations in the treatment of nasopharyngeal carcinoma. Ear Nose Throat J. 1990;69:279±281. Ho JHC. Stage classi®cation of nasopharyngeal carcinoma: a review. In: De-The G, Eto Y, editors. Nasopharyngeal carcinoma: etiology and control, International Agency for Research on Cancer. Lyon: IARC Scienti®c, 1978. pp. 99±113. Hsu MM, Ko JY, Sheen TS, Chang YL. Salvage surgery for recurrent nasopharyngeal carcinoma. Arch. Otolaryngol. Head Neck Surg. 1997;123:305±309. Hsu MM, Tsu SM. Nasopharyngeal carcinoma in Taiwan: clinical manifestations and results of therapy. Cancer 1983;52:362±368. Kaplan E, Meier P. Non-parametric estimations from incomplete observations. J. Am. Stat. Assoc. 1958;453:457±480. Lee AWM, Chan DKK, Fowler; JF, et al. Effect of time, dose and fractionation on local control of nasopharyngeal carcinoma. Radiother Oncol 1995;36:24±31. Lee AWM, Foo W, Law SCK, et al. Reirradiation for recurrent nasopharyngeal carcinoma: factors affecting the therapeutic ratio and ways for improvement. Int. J. Radiat. Oncol. Biol. Phys. 1997;38:43±52. Lee AWM, Foo W, Poon YF, et al. Staging of nasopharyngeal carcinoma: evaluation of N-staging by Ho and UICC/AJCC systems. Clin. Oncol. 1998;8:146±154. Lee AWM, Law SCK, Foo W, et al. Retrospective analysis of patients with nasopharyngeal carcinoma treated during 1976±1985: survival after local recurrence. Int. J. Radiat. Oncol. Biol. Phys. 1998;26:773± 782. McNeese MD, Fletcher GH. Retreatment of recurrent nasopharyngeal carcinoma. Radiology 1981;138:191±193. Ng SH, Chang TC, Ko SF, et al. Nasopharyngeal carcinoma: MRI and CT assessment. Neuroradiology 1997;39:741±746. Ng SH, Wan YL, Ko SF, Chang JT. MRI of nasopharyngeal carcinoma with emphasis on relationship to radiotherapy. J. Mag. Reson. Imag. 1998;8:327±336. Perez CA, Devineni VR, Marcial-Vega V, Marks JE, Simpson JR, Kucik N. Carcinoma of the nasopharynx: factors affecting prognosis. Int. J. Radiat. Oncol. Biol. Phys. 1992;23:271±280. Pryzant RM, Wendt CD, Deldos L, Peters LJ. Re-treatment of nasopharyngeal carcinoma in 53 patients. Int. J. Radiat. Oncol. Biol. Phys. 1992;22:941±947. Sanguineti G, Geara FB, Garden AS, et al. Carcinoma of the naso-

142

[28] [29] [30] [31]

J.T.-C. Chang et al. / Radiotherapy and Oncology 54 (2000) 135±142 pharynx treated by radiotherapy alone: determinants of local and regional control. Int. J. Radiat. Oncol. Biol. Phys. 1997;37:985±996. Sham JST, Choy D, Wei WI, Yau CC. Value of clinical follow-up for local nasopharyngeal carcinoma relapse. Head Neck 1992;14:208±217. Tang SGJ, Lin FJ, Chen MS, Liaw CC, Leung WM, Hong JH. Prognostic factors of nasopharyngeal carcinoma: a multivariate analysis. Int. J. Radiat. Oncol. Biol. Phys. 1990;19:1143±1149. Teo P, Tsao SY, Shiu W, et al. A clinical study of 407 cases of nasopharyngeal carcinoma in Hong Kong. Int. J. Radiat. Oncol. Biol. Phys. 1989;17:515±530. Teo PML, Kwan WJ, Chan ATC, Lee WY, King WWK, Mol CO. How successful is high-dose (.60 Gy) reirradiation using mainly external beams in salvaging local failures of nasopharyngeal carcinoma? Int. J. Radiat. Oncol. Biol. Phys. 1998;40:897±913.

[32] Wang CC. Re-irradiation of recurrent nasopharyngeal carcinoma ± treatment techniques and results. Int. J. Radiat. Oncol. Biol. Phys. 1987;13:953±956. [33] Wei WI, Ho CM, Yuen PW, Fung CF, Sham JST, Lam KH. Maxillary swing approach for resection of tumors in and around the nasopharynx. Arch. Otolaryngol. Head Neck Surg. 1995;121:638±642. [34] Yan JH, Hu YH, Gu XZ. Radiation therapy of recurrent nasopharyngeal carcinoma-report on 219 patients. Acta. Radiol. Oncol. 1983;22:23±28. [35] Yu XH, Xu GZ, Huang YR, Hu YH, Su XG, Gu XZ. Value of computed tomography in staging the primary lesion (T-staging) of nasopharyngeal carcinoma (NPC): an analysis of 54 patients with special reference to parapharyngeal space. Int. J. Radiat. Oncol. Biol. Phys. 1985;11:2143±2147.