Ga-67 Scan in Patients with lntrathoracic Esophageal Carcinoma Planned for Surgery

MAKOTO KONDO, MD, NOBUTOSHI ANDO, MD,' SHIGERU KOSUDA, MD, SHI-LONG LIAN, MD, ATSUSHI KUBO. MD. HIDEKAZU MASAKI, MD, SHOZO HASHIMOTO, MD, TAKETO TSUTSUI, MD, TERUO KAKEGAWA MD'

The authors evaluated Ga-67 scan in 68 patients with intrathoracic esophageal carcinoma initially planned for surgery. Of these, 59 patients were staged pathologically or surgically; their Ga-67 scan results were then compared with the results of pathologic or surgical investigations. Positive Ga-67 scan results correctly predicted the presence of extraesophageal spread and lymph node metastases. Of 38 clinical Stage I1 patients, 15 (39%) could be Stage I11 by the results of Ga-67 scan. However, two of four clinical Stage I patients (50%) and 35 of 38 clinical Stage I1 patients (92%) were eventually pathologic or surgical Stage 111. This high conversion rate and the high incidence of Stage 111 in esophageal carcinoma patients may not justify routine use of Ga-67 scan only for staging. Ga-67 scan, however, was useful for planning radiotherapy, because missing extraesophageal extension or gross metastases from the radiation fields were detected in six patients of 51 so treated. Ga-67 scan has its value in some patients with intrathoracic esophageal carcinoma.

Cancer 4 9 1031 -1 034, 1982.

EW REPORTS have discussed Ga-67 scan in esoph- F ageal squamous cell carcinioma, and most of them were case reports.'-' We found that positive scan pre- dicted the presence of extraesophageal spread of the primary tumor and lymph node: metastases with a few false positive results.' However, the presence of the pri- mary tumor itself has usually been established by esoph- agography and esophagoscopy before Ga-67 scan is con- sidered. If Ga-67 is to gain general acceptance for rou- tine use, it must prove to be valuable in these clinical settings. At our institution, radical surgery with or with- out preoperative irradiation has been the favored treat- ment for most patients. We can then compare clinical staging of esophageal carcinoma with pathologic or sur- gical staging in conjunction with the findings of Ga-67 scan. This study examines the value and limitations of Ga-67 scan in intrathoracic esophageal carcinoma pa- tients who were to be operated on.

Materials and Methods

Sixty-eight patients with in trathoracic esophageal squamous cell carcinoma were referred to the Depart-

From the Departments of Radiology a,nd *Surgery, Keio University

Supported in part by a grant from the Shiseikai Institute. Address for reprints: Makoto Kondo, MD, Department of Radiol-

ogy, Keio University School of Medicine, 35 Shinanomachi, Shinjuku- ku, Tokyo 160, Japan.

School of Medicine, Tokyo, Japan.

Accepted for publication February 6 , 1981.

ment of Surgery and were prepared for surgery at the first visit. All patients were diagnosed as having car- cinomas by esophagography and esophagoscopy wi th biopsy before Ga-67 scan. Physical examinations and chest x-ray were done. Liver and bone scans were not performed routinely and were not used for the current study.

Of the 68 patients, nine did not undergo subsequent surgical procedures and were excluded from the study: two refused surgery, and seven were inoperable on clin- ical evaluation and received irradiation alone. The stag- ing procedures in the remaining 59 patients are listed in Table 1. These patients were staged clinically, patho-

TABLE I . Staging Procedures in 59 Patients with EsoDhaeeal Carcinoma

Pathological procedures Radical surgery* Autopsy (within 2 mo.)

Surgical procedure Exploratory thoracotomyt Exploratory laparotomyt Mediastinoscopy $

41 6

2 8 3

TOTAL h o t

* Includes esophagectomy and lymph node dissection of the me- diastinum and upper abdomen; 33 received preoperative irradiation.

t One had both exploratory thoracotomy and laparotomly. $ In AJC system a clinical procedure, but in UICC system, 3 s u r -

gical procedure.' Thus, in this study, considered as ;I surgical pro- cedure.

0008-543)3/82/0301 /I031 $0.70 0 American Cancer Society

1031

1032 CANCER March 1 1982 Vol. 49

TABLE 2. TNM Classification of the Intrathoracic Esophagus (AJC)

Primary tumor (T) TO No demonstrable tumor. TI A tumor 5 cm or less in esophageal length with no

obstruction, no circumferential involvement, and no extraesophageal spread.* A tumor more than 5 cm in esophageal length with no extraesophageal spread or a tumor of any size that obstructs or has circumferential involvement and no extraesophageal spread. Any tumor with extraesophageal spread.

(Clinical evaluation) Regional lymph nodes for the intrathoracic esophagus that are not ordinarily accessible for clinical evalution. (Surgical evaluation) No positive nodes. (Surgical evaluation) Positive nodes.

T2

T3

Nodal involvement (N) NX

NO N 1

Distant metastasis (M) MX Not assessed. MO No (known) distant metastasis.? MI Distant metastasis present.

* 1. Recurrent laryngeal, phrenic, or sympathetic nerve involve- ment. 2. Fistula formation. 3. Involvement of the tracheal or bron- chial tree. 4. Vena cava or azygos vein obstruction. 5. Malignant effusion.

7 Any cervical, supraclavicular, scalene, or abdominal lymph node is considered distant metastasis.

Stage grouping: Stage I-TI NO MO, TI NX MO Stage 11-T2 NO MO, T2 NX MO; Stage 111-Any T3, any N1, any MI.

logically, and/or surgically according to the American Joint Committee (AJC) Cancer Staging System (Table 2).6 The extent of extraesophageal invasion was clas- sified microscopically (in pathologic cases) or macro- scopically (in surgical cases) as a. (no invasion), a , (possible extension beyond the esophageal wall), a2 (definite extension beyond the esophageal wall), or a3 (invasion of neighboring structures). In three cases with preoperative irradiation, no viable cancer cells were found in the resected esophagus. In these cases, ex- traesophageal spread was estimated by the presence of scars of transwall penetrating ulcers. Lymph node me- tastases were also evaluated microscopically.

Anterior and posterior scans of the neck, chest, and upper abdomen were obtained 48 hours after intrave- nous administration of 2 mCi of Ga-67 citrate. Scans

were obtained with a Toshiba dual-probe rectilinear scanner with 12.7 cm (5 inches) NaI crystals, using medium-energy collimators (focal depth: 10 cm). Two of us read the scans without reference to the clinical data. The degree of Ga-67 accumulation was graded as 0 (negative), 1+ (uptake < liver) or 2+ (uptake 2 liver). Equivocal activity was interpreted as negative. Since Ga-67 readily accumulates in the bones while the mediastinum often shows diffusely increased activity, only an apparent localized increase in activity was in- terpreted as a positive result. It was shown that radia- tion doses more than 2000 rads (20 Gy) decreased up- take of Ga-67 by the primary tumor.' In the current study, all Ga-67 scans were obtained before reaching a radiation dose of 2000 rads.

Results

Ga-67 uptake in the primary regions was correlated with tumor extension in 52 patients who had investi- gation of the mediastinum pathologically or surgically (Table 3). In seven patients who had exploratory lap- arotomy only, such comparison was not possible. In these seven cases, there were two cases with 2+ uptake and one with 1+ uptake. From Table 3, 1+ and 2+ uptake apparently accompanied definite extension of the tumor beyond the esophageal wall or invasion of neighboring structures.

In Table 4, correlation between Ga-67 activity out- side of the primary sites and lymph node involvement is shown. In cases that had either exploratory laparot- omy or thoracotomy alone, the explored sites only were evaluated for Ga-67 uptake. Eight patients showed 2+ Ga-67 activity outside of the primary: two at the supra- clavicular regions, two at the paratracheal regions (Fig. l), and four at the upper abdomen (Fig. 2). All patients were confirmed to have lymph node metastases at each Ga-67 positive site. Those metastases were gross rather than microscopic and more than 3 cm in diameter in all cases. On the contrary, most of the lymph node metastases with 1+ or negative Ga-67 uptake had mi- croscopic lesions rather than gross ones. In five patients, bilateral hilar accumulation of Ga-67 was erroneously

TABLE 3. Correlation between Extent of Invasion and Ga-67 Activity

Degree of activity

0 I + 2+ (Negative) (Uptake < liver) (Uptake B liver) Total

a, (No invasion) 6 0 0 6 (12%) a , (Possible extension beyond the esophageal wall) 7 0 0 7 (13%) a2 (Definite extension beyond the esophageal wall) 11 7 5 23 (44%) a, (Invasion of neighboring structures) 6 3 7 16 (31%)

Total 30 (58%) 10 (19%) 12 (23%) 52 (100%)

No. 5 G,4-67 SCAN IN ESOPHAGEAL CA - Kondo et al. 1033

attributed to lymph node metastasis, but no lesions were found; in two, the mediastinum had false positive ac- cumulation of Ga-67.

All 59 patients were staged by AJC Cancer Staging System clinically. There were four Stage I, 38 Stage 11, and 17 Stage I11 cases. Taking into account that Ga-67 uptake had the modest sensitivity but the high specificity to detect extraesophageal spread and/or lymph node metastases, which constitute Stage 111, we studied the conversion from low stage to higher stage when Ga-67 findings were also used as clinical staging tools. For this purpose, 1+ and :2+ uptake was consid- ered positive for extraesophageal extension, and only 2+ uptake was used for lymph node metastases. Table 5 showed the incidence of Ga-67 uptake inside of the primary sites according to clinical stages determined before Ga-67 scan was done. In 38 clinical Stage I1 patients, 13 (34%) could be upstaged to Stage I11 by the Ga-67 uptake findings inside of the primary regions. Other two clinical Stage I1 cases (5%) could also be Stage I11 based on Ga-67 accumulation in lymph node metastases (Table 6). In four clinical Stage I patients, no abnormal findings were detected by Ga-67 scan.

Conversion from clinical stages to pathologic or sur- gical stages is shown in Table 7. Two of four clinical Stage I cases (50%) and 35 of 38 clinical Stage I1 cases (92%) were converted to pathologic or surgical Stage 111. None of clinical Stage I1 or I11 cases was down- staged by pathologic or surgical evaluation. There were 54 pathologic or surgical Stage 111 cases of 59. Of the 59 patients, 42 were T3 cases (71%), 24 N1 (41%), and 28 M1 (47%) after pathologic or surgical evaluation.

A total of 51 patients received radiotherapy, and ra- diation fields were examined with Ga-67 scan findings for appropriate coverage of the gross diseases. Missing extraesophageal extention from the radiation fields was detected in two patients, and missing gross lymph node metastasis in four (Figs. 1 and 2).

Discussion

Ga-67 uptake in patients with esophageal carcinoma apparently indicated the presence of extraesophageal extension of the primary tumor or the presence of lymph node metastases. Care was taken in this study that no

TABLE 4. Correlation betweeri Scan and Lymph Node Involvement

Lymph node involvement

Scan Positive Negative

2+ 8 I + 6 0 23

TOTAL 37

0 7

15 22

TABLE 5. Ga-67 Uptake Inside of the Primary Sites according to Clinical Stages

Ga-67 activity

0 1 + 2+

Stage I 4 0 0 Stage I1 25 7* 6 Stage 111 5 4 87'

TOTAL 34 11 14

* One had exploratory laparotomy only. t Two had exploratory laparotomy only, but clinically had recurrent

laryngeal nerve palsy.

TABLE 6. Ga-67 Uptake Outside of the Primary Sites according to Clinical Stages

Ga-67 activity

0 and I + 2+

Stage I Stage I1 Stage 111

TOTAL

4 36 1 1 51

equivocal accumulation was considered positive; that was probably the reason of the high specificity to detect extraesophageal extension. Although Ga-67 uptake in the liver varies, especially in patients with cirrhosis, we think that unequivocal accumulation in the medias- tinum could easily be identified without reference to the

TABLE 7. Conversion from Clinical to Pathologic or Surgical Stages

Clinical Pathologic or surgical

Staee T N M No. Staae T N M No.

I TI NX MO 4 I T1 NO MO 2

111 TI NI M I I T3 NO MO 1

11 T2 NX MO 38 11 T2 NO MO 3

111 T2 NO M I 3 T2 NI MO 3 T2 N I MI 1 T2 NX MI* 3 T3 NO MO 1 1 T3 NO MI 1 T3 N I MO 7 T3 NI M I 6

111 T2 NX MI 6 111 T3 NO M I I T3 NX M I * 3 T3 NI MI 2

T3 NX MO LO T2 N1 M l t 1 T3 NX M1 1 T3 NO MO 4$

T3 NO MI 3 T3 N1 MI 3

* Exploratory laparotomy alone. t Recurrent laryngeal nerve palsy was due to lymph node metastasis. $One was NX because of exploratory laparotomy alone.

1034 CANCER March 1 1982 V O l . 49

FIG. 1. A clinical T2NXMO and pathologic T2NlMO patient with carcinoma at the midthoracic portion. A closed mark is a radiation field. Arrows show a paratracheal lymph node metastasis, which was not known before Ga-67 scan. The primary and metastatic tumors show the different degree of Ga-67 uptake. Ga-67 uptake in the pri- mary region is equivocal.

FIG. 2. A clinical T2NXMO case with carcinoma extending from the intrathoracic portion to the cervical portion on esophagography. By Ga-67 scan, the main tumor may be located in the cervical portion. An arrow head shows an upper abdominal lymph node metastasis. A closed mark indicates a radiation field. These findings were confirmed at operation.

liver uptake, and that predicting the presence of ex- traesophageal extension was possible. Without referring to the liver uptake, one might be confused about whether uptake outside of the primary sites was 1+ or 2+, thus changing the sensitivity and specificity. How- ever, in our series, all 2+ uptake outside the primary sites was so unequivocal that it could not be mistaken (Figs. 1 and 2). False positives for lymph node metas- tases were mostly caused by nonpathologic Ga-67 ac- cumulation to the hili of the lung. Small or microscopic lymph node metastases produced many false negatives, and it is unlikely that tumors less than a few centimeters in the midline structures produce positive Ga-67 scan.

The proportion of the Stage I11 cases in the current series was very high; almost all patients were Stage I11 (54 of 59). Since seven patients had exploratory lapa- rotomy only, the true incidence of N1 cases was prob- ably higher than reported here. It is unknown whether or not it is judicious to include any T3, any N1, and any M1 under the heading of Stage 111. However, uni- formly fatal outcomes of esophageal carcinoma (about 10% five-year survival in most series) may substantiate this staging system.

It may be inadvisable to use Ga-67 scan only for predicting upstage conversion, since the incidence of pathologic or surgical Stage I11 cases is so high. Ga-67 scan is expensive and burdens limited medical resources. Therefore, Ga-67 scan should only inform the treatment of choice. From a radiotherapeutic point of view, Ga- 67 scan was useful in some cases, although few, to change radiation fields. Since the patients received sur- gery and/or radiation, it is unclear whether or not cases with or without positive Ga-67 scan had the same prog- nosis or the same degree of extraesophageal extension. Possibly, the clinical T2 cases having positive Ga-67 scan inside of the primary sites (1 5 of 44) might be more aggressive both anatomically and biologically than cases without it, and surgery without preoperative radiotherapy may prove that.

REFERENCES

I . Kondo M, Hashimoto S, Kubo A, Kakegawa T, Ando N. 67Ga scanning in the evaluation of esophageal carcinoma. Radiology 1979;

2. Lavender JP, Lowe J, Barker JR, Burn JI, Chaudhri MA. Gal- lium 67 citrate scanning in neoplastic and inflammatory lesions. Br J Radio1 1971; 44:361-366.

3. Ito Y, Okuyama S, Awano T, Takahashi K, Sat0 T, Kanno 1. Diagnostic evaluation of 67Ga scanning of lung cancer and other dis- eases. Radiology 1971; 101:355-362.

4. Langhammer H, Glaubitt G, Grebe SF, et al. 67Ga for tumor scanning. J Nucl Med 1972; 13:25-30.

5. Higashi T, Nakayama Y, Murata A, et al. Clinical evaluation of 67Ga-citrate scanning. J Nucl Med 1972; 13:196-201.

6. American Joint Committee for Cancer Staging and End-Results Reporting. Manual for Staging of Cancer. 1978; 65-70.

7. International Union Against Cancer. TNM Classification of Malignant Tumours, 3rd ed. 1978; 57-62.

131:723-726.