Clinical Radiology (1990) 42, 166-169
Bone Scintigraphy in Nasopharyngeal Carcinoma F. X. SUNDRAM, E. T. CHUA, A. S. W. GOH, H. J. TOH, T. H. K H O R and E. J. C H U A
Department of Nuclear Medicine and Therapeutic Radiology Department, Singapore General Hospital One hundred and forty-three patients (Group 1) with histologically proven nasopharyngealcarcinoma (NPC) had bone scintigraphy with 99Tcm methylenediphosphonate (MDP) or dihydroxypropanediphospbonate (DPD) within 2 months of the initial diagnosis. A further 162 patients (Group 2) had bone scans during the course of follow-up if there were symptoms of bone pain or evidence of metastases at other sites. Twenty-three per cent (33/143) of the newly diagnosed NPC patients (Group 1) had evidence of bone metastases. Of these 143 patients, 101 were T0-T2, 16 were T3 and 25 were T4. Thirty-six patients had no neck nodes (NO), 44 were N1, 25 N2 and 38 N3. Of the 162 patients in Group 2, 96 (59%) had a positive bone scan. The commonest sites for bony metastases from NPC were the spine, ribs, pelvis and lower limbs in order of frequency. There is a highly significant association with the nodal stage but no association with the UICC T staging which is not adequate in nasopharyngeal carcinoma. In our part of the world, bone metastases from NPC are a common cause of an abnormal bone scan. Sundram, F.X., Chua, E.T., Goh, A.S.W., Toh, H.J., Khor, T.H. & Chua, E.J. (1990). Clinical Radiology 42, 166-169. Bone Scintigraphy in Nasopharyngeal Carcinoma
Nasopharyngeal carcinoma (NPC) is the fourth commonest cancer in the Singapore male, but in the age-group 15-34 it is the commonest male cancer (Lee et al., 1988). Previous reports have indicated relatively high incidence of distant metastases developing in patients with nasopharyngeal cancer (Tan and Oon 1967; O'Brien et al., 1971; Probert et al., 1974; Khor et al., 1978). Khor et al. (1978) quoted an incidence of bony metastasis of only 2% at presentation, but 50% developed such metastases during the course of the disease. However, they based the diagnosis of bone involvement on clinical examination and radiography, not scintigraphy. The present study used bone scintigraphy to determine: 1 the frequency of bone metastases at the time of initial diagnosis of NPC, for various stages of the disease; 2 the sites of bony involvement in patients with NPC and bone metastases; 3 the pattern of distant spread of NPC.
The bone scan was considered positive if there were two or more sites of abnormal tracer localization in the absence of history of previous trauma or degenerative bone disease. Scans showing solitary lesions, or focal areas of mild increase in uptake, were considered equiv0. cal. Those patients with equivocal scans had radiography to help elucidate the nature of the abnormality noted on bone scan. Due to non-specificity of bone scintigraphy it is possible that some lesions may have been misdiagnosed by the bone scan. Most patients were staged clinically before radiation therapy but in some cases computerized tomography (CT) of the post-nasal sPace was available. The primary tumour was staged using t h e U I C C T staging system. Nodal staging was according to Ho's system (Ho, 1982) dividing the neck into three regions- upper, mid and lower cervical. RESULTS
METHODS Patients referred to the Radiotherapy Department over a 6 year period for treatment of histologically proven NPC form Group 1 of the study population. Treatment was carried out by one of seven radiotherapists. All patients in this group had their bone scans performed within 2 months of the primary diagnosis. Group 2 comprised patients who underwent bone scintigraphy for symptoms developing during follow-up. These patients were distinct from the Group 1 patients. Bone scintigraphy was performed using an intravenous injection of 555-740 MBq (15-20 mCi) of 99TCm MDP or DPD, followed 3 hours later by spot views of the entire skeletal system. Patients were imaged on a large field of view (LFOV) gamma-camera and 400 000 to 1 million counts were accumulated, depending on the view. The images were stored on 8 × 10 film. Correspondence to: F. X. Sundram, Nuclear Medicine Department, Singapore General Hospital, Singapore.
The total number of patients having bone scans at presentation (group 1) was 143. Of these patients, there were 104 males and 39 females. Group 2 contained 162 patients, 116 males and 46 females. Age Distribution
Table 1 shows the age distribution of the scan positive and scan negative patients in the two study groups. Table 1 - Bone scan results by age distribution for Groups 1 and 2 Group 1
,
Group 2 J
Age
+ ve scan
10-29 30-49 50-69 70-89
7 15 10 1
Total
33
ve scan
13 54 32 8 107 3
Equivocal
0 1 2 0
Total
Total
+ ve scan
-- ve scan
Equivocal
20 70 44 9
12 54 29 1
12 34 13 0
1 2 3 1
25 90 ~5 2
143
96
59
7
162
167
BONE SCINTIGRAPHY IN NPC Table 2 - Bone scan results by nodal status for Groups 1 and 2 Group 2
Group 1 + ve
-- ve
+ ve
Total
+ ve
-- ve
+__ve
Total
No neck nodes Upper neck nodes Mid neck nodes Lower neck nodes
2 5 8 18
33 38 17 19
1 1 0 1
36 44 25 38
22 37 16 21
24 27 3 5
2 3 1 1
48 67 20 27
Total
33
107
3
143
96
59
7
162
Nodal Stage
Other Metastases
Table 2 shows the bone scan results against nodal status of the patients. In Group 1, there was a highly significant association between bony metastases and nodal status (with P < 0.005). We did not attempt to correlate scan results with the nodal size, but bilateral node involvement was studied; this showed 14/57 (24%) incidence of positive bone scans in unilateral nodal disease as compared to 17/50 (34%) when nodes were bilateral (P < 0.01).
In the 33 patients with bone metastases at initial presentation, four also had lung secondaries and there was liver involvement in three. Thus 26 patients had metastases confined to bone alone. Ten patients in Group 1 developed extraosseous metastases which were detected at follow-up rather than at presentation. In Group 2 there were 30 patients with extraosseous metastases, but only five of them had no bone involvement. Of these 30 patients, 11 had lung secondaries, 12 had liver involvement and two had brain metastases, while five had nodal or skin involvement outside the head and neck regions. Three patients had bone~ liver and lung metastases while two had bone, lung and nodal involvement.
TumourStage Table 3 shows the distribution of UICC T staging against the bone scan results for patients in both groups.
BonePain
DISCUSSION
Table 4 shows the bone scan results in relation to bone pain for patients in both Group 1 and Group 2. It can be seen that approximately 2/3 of patients with bone pain have positive scans. There were three equivocal scans which when repeated on further follow up became positive.
Sex and Age
Table 3 - Bone scan results versos U I C C T staging for Groups 1 and 2 Group 1 +ve
--ve
T0,.T1, T2 21 T3 6 "f4 5 Unknown i Total
Group 2
33
+_re
Total
+ve
--re
-t-ve
Total
78 10 19 0
2 0 1 0
101 16 25 1
74 5 12 5
52 2 5 0
7 0 0 0
133 7 17 5
107
3
143
96
59
7
162
The sex ratio for both Group 1 and 2 patients is similar, i.e. male: female ratio of 2.7 : 1 for Group 1 patients and 2.5:1 for Group 2 patients. This corresponds to the national data (Lee et al., 1988). There is a survival advantage for female patients (Qin et al., 1988) although it is not known from that study whether the advantage takes into account other prognostic factors. The positive bone scan rate for the various age groups ranged from 13 to 29% for patients at first presentation. Two large studies (Huang, 1980; Qin et al., 1988) looking at prognostic factors indicated that younger patients tended to have a better prognosi s. On the other hand, in countries where NPC is rare, and is less often considered, younger patients tend to be misdiagnosed and are seen when the tumour is in the later stages (Bass et aI., 1985). From our study, for the age group below 30 years, there is a high positive bone scan rate of 7/20 or 35%, at time of initial diagnosis, with a poor prognosis probably related to late nodal stage or an aggressive form of the disease. Berry et al. (1980) have noted a poor prognosis for this age group. Over the age of 30, 26/123 or 21% had positive bone scans at, initial diagnosis.
Table4 - Symptoms correlated with bone scan findings for Groups 1 and 2
Nodal Stage Group 1 Pain present
:+re scan 20 -re Scan 9 ~qUivocal scan 1 TOtal
30
Group 2 Pain absent
Total
Pain present
Pain absent
Total
13 98 2
33 107 3
86 32 7
10 27 0
96 59 7
113
143
125
37
162
Table 2 shows the bone scan results for the various nodal status for both groups of patients. The results correlated with the site of nodal involvement in the neck. At initial diagnosis only 5.5% of N O patients had a positive scan, whereas 47% of patients with lower neck nodes were scan positive. In both groups the difference in scan findings between different stages was found to be highly statistically significant ( P < 0.005).
168
CLINICAL RADIOLOGY
It would be reasonable to expect that patients with larger nodes would have a greater likelihood of metastases, and similarly with bilateral nodes the tumour bulk would be more and the likelihood of metastases increased. A higher incidence of positive scans was noted in patients with bilateral neck nodes but the site of nodes may be more important in determining the likelihood of metastases, since in our study 7/10 with unilateral lower neck nodes and 12/28 with bilateral lower neck nodes had positive bone scans. Studies discussing the prognosis in NPC have not generally taken into account the site of nodes. They were usually based on the staging system recommended by the UICC which takes into account fixity and bilateral neck involvement (Qin et al., 1988). A study by Wong et al. (1988) takes into account site, fixity and size of nodes. However, because of these multiple conditions, it is difficult to compare a prognostic factor such as site alone, unless multivariate analysis is performed. Tumour Stage
It was expected that there would be poor correlation between UICC tumour staging and bone scans because the T staging for TO, T1 and T2 is very difficult to determine clinically. As many as 51% of patients may be upgraded after a CT scan is performed (Zhen et al., 1989). The UICC T staging for T3 and T4 was easier to recognize clinically, and these represent more tumour bulk although some T4 patients may have small tumours with a tendency to invade nerves and cause cranial nerve palsies. Our results do not show any significant correlation between clinical T stage and the incidence of positive bone scans. Sites of Bone Metastases
The commonest sites of scintigraphic lesions were the spine, ribs, pelvis and lower limbs, corresponding to the sites of marrow production (Tan and Oon, 1967). Upper limb metastases were uncommon, being only 5% and 16% for Group 1 and 2 patients respectively. Tan and Oon (1967) using radiological skeletal surveys found a pretreatment positive bone metastasis rate of 4%, compared to 23 % in our present study, using bone scintigraphy. The post-treatment positive rate in their study was 13% compared to 59% in our study. It appears that use of bone scintigraphy has resulted in a higher yield of bone metastases in NPC. Bone Pain
The results in Table 4 suggest that it may not be worthwhile doing a bone scan at presentation for patients without bone pain as only 11% of them will have a positive scan. This of course is also dependent on the stage of disease, and bone scans may need to be done routinely in advanced nodal disease. As in breast cancers (Merrick, 1989) management of patients with nasopharyngeal carcinoma is little changed after a bone scan. The role of chemotherapy or other adjuvant treatment in nasopharyngeal carcinoma has not been conclusively proven to be of use and there is little or no treatment available for these asymptomatic bone lesions. Even with bone pain only 68% had a positive bone scan in both groups of patients.
A bone scan would definitely be indicated in patients who need restaging for local or nodal recurrence, or other sites of metastases.
Other Sites
Bone is the most frequent site for metastases in NPC. Only seven of the 33 patients with bone metastases at initial NPC diagnosis had other sites of involvement. On follow-up after treatment, however, 10 patients deveI. oped only non-bone sites of metastases, chiefly in the liver (6), lung (3) and lymph nodes below clavicle (1). A total of 33/143 or 23% of NPC patients were already known to have metastases at time ofinital diagnosis and referral for radiation therapy. In the second group, 30 of the 96 patients with positive bone scans had extraosseous metastases involving lung (11), liver (12), brain (2), skin, and extra-cervical lymph. nodes (5). The brain metastases were, however, difficult to differentiate from local recurrence in the primary site. The tumour in the brain may appear to be at a distance from the primary site and yet have spread by direct extension. Of the 59 bone scan negative patients there were five patients with only non-osseous sites of metastases. It would be necessary to maintain follow up on all the patients with a negative bone scan at presentation, to see if they subsequently develop metastases, and to correlate this with stage and bone symptoms.
CONCLUSION In this series 23% of patients had bone metastasis at presentation with NPC, and the incidence of metastasis was related to clinical stage, in particular to nodal status. Earlier detection should therefore result in a higher cure rate with local treatment. The commonest sites of bone metastases in NPC are in spine, ribs and pelvis. The pattern of metastatic spread in this series is similar to that noted two decades ago in that the commonest sites of metastases are the bone, liver, lung and lymph-nodes below the clavicle. The use of the bone scan as a staging technique in nasopharyngeal carcinoma runs into the same controversies as in staging of other turnouts with a predilection for bone involvement (Merrick, 1989). The problem is more acute when there is no effective treatment presently available for asymptomatic cases. However, the presence of bone metastases at presentation may help select patients for whom primary radiotherapy would be pallia" tive and the treatment schedule for the primary sh0r" tened. At present it has been thought unwise to treat patients without symptoms as they may not manifest anY problem for many months or years. Some alternative treatment such as whole body or half body irradiation !~ combination with a cisplatinu/n based chemotherapeutlC regime may be more appropriate for these patients.
J Acknowledgements.The authors appreciate the technical service~iI Llm P " Hoon - H " T " Tan and Nazreen ' and the secreta" C. " ' " o th~ assistance of Jenny Ng. We would like to record our _~ratituge~LtN~:" medical staff of the Therapeutic Radiology Department tor cooperation in this, study. C.
BONE SCINTIGRAPHY IN NPC
gI~FERENCES 13°ss IS, Hailer, JO, Berdon, WE, Barlow, B, Carsen, G & Khakoo, Y (1985). Nasopharyngeal carcinoma: clinical and radiographic findilags in children. Radiology, 156, 651-654. BerrY,MP, Smith, CR, Brown, TC, Jenkin, DT & Rider, WD (1980). Nasopharyngeal carcinoma in the young. International Journal of Radiation Oncology, Biology, Physics, 6, 415-421. 14o,JHC (1982). Nasopharynx. In: Treatment of Cancer, ed. Halnan, KE, PP. 249-268. Chapman & Hail, London. 4uang, SC (1980). Nasopharyngeal cancer: a review of 1605 patients treated radically with Cobalt 60. International Journal of Radiation Oncology, Biology, Physics, 6, 401 407. Khor,TH, Tan, BC, Chua, EJ & Chia, KB (1978). Distant metastases in nasopharyngeal carcinoma. Clinical Radiology, 29, 27-30. Lee, HP, Day, NE & Shanmugaratnam, K (1988). Trends in Cancer Incidence in Singapore 1968-1982. IARC Scientific Publication No 9l, IARC, Lyon.
169
Merrick, MV (1989). Bone scintigraphy--an update. Clinical Radiology, 40, 231-232. O'Brien, PH, Carlson, R, Steubner, EA & Staley, CT (1971). Distant metastases in epidermoid cell carcinoma of the head and neck. Cancer, 27, 304-307. Probert, JC, Thompson, RW & Bagshaw, MA (1974). Patterns of spread of distant metastases in head and neck cancer. Cancer, 33, 127-133. Qin-DX, Hu YH, Yan, JH et al. (1988). Analysis of 1379 patients with nasopharyngeal carcinoma treated by radiation. Cancer, 61, 1117
1124.
Tan, BC & Oon, CL (1967). Bone metastases in carcinoma of the nasopharynx. Clinical Radiology, 18, 69-73. Wong, DC, Cai, WM, Hu, YH & Gu, XZ (1988). Long-term survival of 1035 cases of nasopharyngeal carcinoma. Cancer, 61, 2338-2341. Zhen, G, Zeng, Q, Wu, P & Yuan, C (1989). Computed tomography in the management of nasopharyngeal carcinoma. Clinical Radiology, 40, 25-29.
Bone Scintigraphy in Nasopharyngeal Carcinoma F. X. SUNDRAM, E. T. CHUA, A. S. W. GOH, H. J. TOH, T. H. K H O R and E. J. C H U A
Department of Nuclear Medicine and Therapeutic Radiology Department, Singapore General Hospital One hundred and forty-three patients (Group 1) with histologically proven nasopharyngealcarcinoma (NPC) had bone scintigraphy with 99Tcm methylenediphosphonate (MDP) or dihydroxypropanediphospbonate (DPD) within 2 months of the initial diagnosis. A further 162 patients (Group 2) had bone scans during the course of follow-up if there were symptoms of bone pain or evidence of metastases at other sites. Twenty-three per cent (33/143) of the newly diagnosed NPC patients (Group 1) had evidence of bone metastases. Of these 143 patients, 101 were T0-T2, 16 were T3 and 25 were T4. Thirty-six patients had no neck nodes (NO), 44 were N1, 25 N2 and 38 N3. Of the 162 patients in Group 2, 96 (59%) had a positive bone scan. The commonest sites for bony metastases from NPC were the spine, ribs, pelvis and lower limbs in order of frequency. There is a highly significant association with the nodal stage but no association with the UICC T staging which is not adequate in nasopharyngeal carcinoma. In our part of the world, bone metastases from NPC are a common cause of an abnormal bone scan. Sundram, F.X., Chua, E.T., Goh, A.S.W., Toh, H.J., Khor, T.H. & Chua, E.J. (1990). Clinical Radiology 42, 166-169. Bone Scintigraphy in Nasopharyngeal Carcinoma
Nasopharyngeal carcinoma (NPC) is the fourth commonest cancer in the Singapore male, but in the age-group 15-34 it is the commonest male cancer (Lee et al., 1988). Previous reports have indicated relatively high incidence of distant metastases developing in patients with nasopharyngeal cancer (Tan and Oon 1967; O'Brien et al., 1971; Probert et al., 1974; Khor et al., 1978). Khor et al. (1978) quoted an incidence of bony metastasis of only 2% at presentation, but 50% developed such metastases during the course of the disease. However, they based the diagnosis of bone involvement on clinical examination and radiography, not scintigraphy. The present study used bone scintigraphy to determine: 1 the frequency of bone metastases at the time of initial diagnosis of NPC, for various stages of the disease; 2 the sites of bony involvement in patients with NPC and bone metastases; 3 the pattern of distant spread of NPC.
The bone scan was considered positive if there were two or more sites of abnormal tracer localization in the absence of history of previous trauma or degenerative bone disease. Scans showing solitary lesions, or focal areas of mild increase in uptake, were considered equiv0. cal. Those patients with equivocal scans had radiography to help elucidate the nature of the abnormality noted on bone scan. Due to non-specificity of bone scintigraphy it is possible that some lesions may have been misdiagnosed by the bone scan. Most patients were staged clinically before radiation therapy but in some cases computerized tomography (CT) of the post-nasal sPace was available. The primary tumour was staged using t h e U I C C T staging system. Nodal staging was according to Ho's system (Ho, 1982) dividing the neck into three regions- upper, mid and lower cervical. RESULTS
METHODS Patients referred to the Radiotherapy Department over a 6 year period for treatment of histologically proven NPC form Group 1 of the study population. Treatment was carried out by one of seven radiotherapists. All patients in this group had their bone scans performed within 2 months of the primary diagnosis. Group 2 comprised patients who underwent bone scintigraphy for symptoms developing during follow-up. These patients were distinct from the Group 1 patients. Bone scintigraphy was performed using an intravenous injection of 555-740 MBq (15-20 mCi) of 99TCm MDP or DPD, followed 3 hours later by spot views of the entire skeletal system. Patients were imaged on a large field of view (LFOV) gamma-camera and 400 000 to 1 million counts were accumulated, depending on the view. The images were stored on 8 × 10 film. Correspondence to: F. X. Sundram, Nuclear Medicine Department, Singapore General Hospital, Singapore.
The total number of patients having bone scans at presentation (group 1) was 143. Of these patients, there were 104 males and 39 females. Group 2 contained 162 patients, 116 males and 46 females. Age Distribution
Table 1 shows the age distribution of the scan positive and scan negative patients in the two study groups. Table 1 - Bone scan results by age distribution for Groups 1 and 2 Group 1
,
Group 2 J
Age
+ ve scan
10-29 30-49 50-69 70-89
7 15 10 1
Total
33
ve scan
13 54 32 8 107 3
Equivocal
0 1 2 0
Total
Total
+ ve scan
-- ve scan
Equivocal
20 70 44 9
12 54 29 1
12 34 13 0
1 2 3 1
25 90 ~5 2
143
96
59
7
162
167
BONE SCINTIGRAPHY IN NPC Table 2 - Bone scan results by nodal status for Groups 1 and 2 Group 2
Group 1 + ve
-- ve
+ ve
Total
+ ve
-- ve
+__ve
Total
No neck nodes Upper neck nodes Mid neck nodes Lower neck nodes
2 5 8 18
33 38 17 19
1 1 0 1
36 44 25 38
22 37 16 21
24 27 3 5
2 3 1 1
48 67 20 27
Total
33
107
3
143
96
59
7
162
Nodal Stage
Other Metastases
Table 2 shows the bone scan results against nodal status of the patients. In Group 1, there was a highly significant association between bony metastases and nodal status (with P < 0.005). We did not attempt to correlate scan results with the nodal size, but bilateral node involvement was studied; this showed 14/57 (24%) incidence of positive bone scans in unilateral nodal disease as compared to 17/50 (34%) when nodes were bilateral (P < 0.01).
In the 33 patients with bone metastases at initial presentation, four also had lung secondaries and there was liver involvement in three. Thus 26 patients had metastases confined to bone alone. Ten patients in Group 1 developed extraosseous metastases which were detected at follow-up rather than at presentation. In Group 2 there were 30 patients with extraosseous metastases, but only five of them had no bone involvement. Of these 30 patients, 11 had lung secondaries, 12 had liver involvement and two had brain metastases, while five had nodal or skin involvement outside the head and neck regions. Three patients had bone~ liver and lung metastases while two had bone, lung and nodal involvement.
TumourStage Table 3 shows the distribution of UICC T staging against the bone scan results for patients in both groups.
BonePain
DISCUSSION
Table 4 shows the bone scan results in relation to bone pain for patients in both Group 1 and Group 2. It can be seen that approximately 2/3 of patients with bone pain have positive scans. There were three equivocal scans which when repeated on further follow up became positive.
Sex and Age
Table 3 - Bone scan results versos U I C C T staging for Groups 1 and 2 Group 1 +ve
--ve
T0,.T1, T2 21 T3 6 "f4 5 Unknown i Total
Group 2
33
+_re
Total
+ve
--re
-t-ve
Total
78 10 19 0
2 0 1 0
101 16 25 1
74 5 12 5
52 2 5 0
7 0 0 0
133 7 17 5
107
3
143
96
59
7
162
The sex ratio for both Group 1 and 2 patients is similar, i.e. male: female ratio of 2.7 : 1 for Group 1 patients and 2.5:1 for Group 2 patients. This corresponds to the national data (Lee et al., 1988). There is a survival advantage for female patients (Qin et al., 1988) although it is not known from that study whether the advantage takes into account other prognostic factors. The positive bone scan rate for the various age groups ranged from 13 to 29% for patients at first presentation. Two large studies (Huang, 1980; Qin et al., 1988) looking at prognostic factors indicated that younger patients tended to have a better prognosi s. On the other hand, in countries where NPC is rare, and is less often considered, younger patients tend to be misdiagnosed and are seen when the tumour is in the later stages (Bass et aI., 1985). From our study, for the age group below 30 years, there is a high positive bone scan rate of 7/20 or 35%, at time of initial diagnosis, with a poor prognosis probably related to late nodal stage or an aggressive form of the disease. Berry et al. (1980) have noted a poor prognosis for this age group. Over the age of 30, 26/123 or 21% had positive bone scans at, initial diagnosis.
Table4 - Symptoms correlated with bone scan findings for Groups 1 and 2
Nodal Stage Group 1 Pain present
:+re scan 20 -re Scan 9 ~qUivocal scan 1 TOtal
30
Group 2 Pain absent
Total
Pain present
Pain absent
Total
13 98 2
33 107 3
86 32 7
10 27 0
96 59 7
113
143
125
37
162
Table 2 shows the bone scan results for the various nodal status for both groups of patients. The results correlated with the site of nodal involvement in the neck. At initial diagnosis only 5.5% of N O patients had a positive scan, whereas 47% of patients with lower neck nodes were scan positive. In both groups the difference in scan findings between different stages was found to be highly statistically significant ( P < 0.005).
168
CLINICAL RADIOLOGY
It would be reasonable to expect that patients with larger nodes would have a greater likelihood of metastases, and similarly with bilateral nodes the tumour bulk would be more and the likelihood of metastases increased. A higher incidence of positive scans was noted in patients with bilateral neck nodes but the site of nodes may be more important in determining the likelihood of metastases, since in our study 7/10 with unilateral lower neck nodes and 12/28 with bilateral lower neck nodes had positive bone scans. Studies discussing the prognosis in NPC have not generally taken into account the site of nodes. They were usually based on the staging system recommended by the UICC which takes into account fixity and bilateral neck involvement (Qin et al., 1988). A study by Wong et al. (1988) takes into account site, fixity and size of nodes. However, because of these multiple conditions, it is difficult to compare a prognostic factor such as site alone, unless multivariate analysis is performed. Tumour Stage
It was expected that there would be poor correlation between UICC tumour staging and bone scans because the T staging for TO, T1 and T2 is very difficult to determine clinically. As many as 51% of patients may be upgraded after a CT scan is performed (Zhen et al., 1989). The UICC T staging for T3 and T4 was easier to recognize clinically, and these represent more tumour bulk although some T4 patients may have small tumours with a tendency to invade nerves and cause cranial nerve palsies. Our results do not show any significant correlation between clinical T stage and the incidence of positive bone scans. Sites of Bone Metastases
The commonest sites of scintigraphic lesions were the spine, ribs, pelvis and lower limbs, corresponding to the sites of marrow production (Tan and Oon, 1967). Upper limb metastases were uncommon, being only 5% and 16% for Group 1 and 2 patients respectively. Tan and Oon (1967) using radiological skeletal surveys found a pretreatment positive bone metastasis rate of 4%, compared to 23 % in our present study, using bone scintigraphy. The post-treatment positive rate in their study was 13% compared to 59% in our study. It appears that use of bone scintigraphy has resulted in a higher yield of bone metastases in NPC. Bone Pain
The results in Table 4 suggest that it may not be worthwhile doing a bone scan at presentation for patients without bone pain as only 11% of them will have a positive scan. This of course is also dependent on the stage of disease, and bone scans may need to be done routinely in advanced nodal disease. As in breast cancers (Merrick, 1989) management of patients with nasopharyngeal carcinoma is little changed after a bone scan. The role of chemotherapy or other adjuvant treatment in nasopharyngeal carcinoma has not been conclusively proven to be of use and there is little or no treatment available for these asymptomatic bone lesions. Even with bone pain only 68% had a positive bone scan in both groups of patients.
A bone scan would definitely be indicated in patients who need restaging for local or nodal recurrence, or other sites of metastases.
Other Sites
Bone is the most frequent site for metastases in NPC. Only seven of the 33 patients with bone metastases at initial NPC diagnosis had other sites of involvement. On follow-up after treatment, however, 10 patients deveI. oped only non-bone sites of metastases, chiefly in the liver (6), lung (3) and lymph nodes below clavicle (1). A total of 33/143 or 23% of NPC patients were already known to have metastases at time ofinital diagnosis and referral for radiation therapy. In the second group, 30 of the 96 patients with positive bone scans had extraosseous metastases involving lung (11), liver (12), brain (2), skin, and extra-cervical lymph. nodes (5). The brain metastases were, however, difficult to differentiate from local recurrence in the primary site. The tumour in the brain may appear to be at a distance from the primary site and yet have spread by direct extension. Of the 59 bone scan negative patients there were five patients with only non-osseous sites of metastases. It would be necessary to maintain follow up on all the patients with a negative bone scan at presentation, to see if they subsequently develop metastases, and to correlate this with stage and bone symptoms.
CONCLUSION In this series 23% of patients had bone metastasis at presentation with NPC, and the incidence of metastasis was related to clinical stage, in particular to nodal status. Earlier detection should therefore result in a higher cure rate with local treatment. The commonest sites of bone metastases in NPC are in spine, ribs and pelvis. The pattern of metastatic spread in this series is similar to that noted two decades ago in that the commonest sites of metastases are the bone, liver, lung and lymph-nodes below the clavicle. The use of the bone scan as a staging technique in nasopharyngeal carcinoma runs into the same controversies as in staging of other turnouts with a predilection for bone involvement (Merrick, 1989). The problem is more acute when there is no effective treatment presently available for asymptomatic cases. However, the presence of bone metastases at presentation may help select patients for whom primary radiotherapy would be pallia" tive and the treatment schedule for the primary sh0r" tened. At present it has been thought unwise to treat patients without symptoms as they may not manifest anY problem for many months or years. Some alternative treatment such as whole body or half body irradiation !~ combination with a cisplatinu/n based chemotherapeutlC regime may be more appropriate for these patients.
J Acknowledgements.The authors appreciate the technical service~iI Llm P " Hoon - H " T " Tan and Nazreen ' and the secreta" C. " ' " o th~ assistance of Jenny Ng. We would like to record our _~ratituge~LtN~:" medical staff of the Therapeutic Radiology Department tor cooperation in this, study. C.
BONE SCINTIGRAPHY IN NPC
gI~FERENCES 13°ss IS, Hailer, JO, Berdon, WE, Barlow, B, Carsen, G & Khakoo, Y (1985). Nasopharyngeal carcinoma: clinical and radiographic findilags in children. Radiology, 156, 651-654. BerrY,MP, Smith, CR, Brown, TC, Jenkin, DT & Rider, WD (1980). Nasopharyngeal carcinoma in the young. International Journal of Radiation Oncology, Biology, Physics, 6, 415-421. 14o,JHC (1982). Nasopharynx. In: Treatment of Cancer, ed. Halnan, KE, PP. 249-268. Chapman & Hail, London. 4uang, SC (1980). Nasopharyngeal cancer: a review of 1605 patients treated radically with Cobalt 60. International Journal of Radiation Oncology, Biology, Physics, 6, 401 407. Khor,TH, Tan, BC, Chua, EJ & Chia, KB (1978). Distant metastases in nasopharyngeal carcinoma. Clinical Radiology, 29, 27-30. Lee, HP, Day, NE & Shanmugaratnam, K (1988). Trends in Cancer Incidence in Singapore 1968-1982. IARC Scientific Publication No 9l, IARC, Lyon.
169
Merrick, MV (1989). Bone scintigraphy--an update. Clinical Radiology, 40, 231-232. O'Brien, PH, Carlson, R, Steubner, EA & Staley, CT (1971). Distant metastases in epidermoid cell carcinoma of the head and neck. Cancer, 27, 304-307. Probert, JC, Thompson, RW & Bagshaw, MA (1974). Patterns of spread of distant metastases in head and neck cancer. Cancer, 33, 127-133. Qin-DX, Hu YH, Yan, JH et al. (1988). Analysis of 1379 patients with nasopharyngeal carcinoma treated by radiation. Cancer, 61, 1117
1124.
Tan, BC & Oon, CL (1967). Bone metastases in carcinoma of the nasopharynx. Clinical Radiology, 18, 69-73. Wong, DC, Cai, WM, Hu, YH & Gu, XZ (1988). Long-term survival of 1035 cases of nasopharyngeal carcinoma. Cancer, 61, 2338-2341. Zhen, G, Zeng, Q, Wu, P & Yuan, C (1989). Computed tomography in the management of nasopharyngeal carcinoma. Clinical Radiology, 40, 25-29.
