Investigation of Retroperitoneal Lymph Nodes in Hodgkin’s Disease

For staging of IIodgkin’s disease 45 patients underwent lymphography and abdominal computed tomography (CT); in 31 of them ultrasound examination was also performed. In cases of retroperitoneal lymph node enlargement lymphography and CT are diagnostically equivalent, whereas the poorer resolution of ultrasound reduces the reliability of this method when the lymph nodes are only slightly enlarged.

tion of ultrasound reduces the reliability of this method when the lymph nodes are only slightly enlarged. Theoretically with lymphography also small lymphomatous lesions in normal-sized lymph nodes should be possible to demonstrate, but there is a clear tendency to overdiagnosis. CT is therefore recommended as the initial method and this examination is considered sufficient both in clearly negative and clearly positive cases. In doubtful cases and in those where CT has revealed solitary slightly enlarged lymph nodes, the examination should be supplemented with ultrasonic scanning and lymphography.

I N TROD UCT ION
In Hodgkin's disease it is important for the planning of appropriate therapy and for the prognostic evaluation to determine the extent of the disease. Until a few years ago lymphography was the method of choice for examination of retroperitoneal lymph nodes. The advent of computed tomography (CT) and ultrasonography has brought new possibilities of investigating the retroperitoneal space.
Opinions Siemens Somatom DR 2 unit was used, with a scanning time of 4.5 seconds and a slice thickness of 8 m m . In the pelvis the distance between the slices was 20 mm and from the aortic bifurcation in the cranial direction it was 1 0 m m .
Routinely, the patients were given 300 ml of a 2.5% solution of Gastrografino (370 mg I / m l ) orally one hour before the examination and a further 300 ml immediately before the s t a r t of i t . Lymph nodes with a diameter of more than 10-15 m m , depending upon the level in the abdomen ( 1 4 ) were regarded a s pathologically enlarged.
Lymphography was performed with injection of 6 to 8 ml Lipiodol Ultra Fluid@ (Guerbet) into a lymph vessel on the dorsum of each foot. Films were taken in a . p . and oblique projections, after the injection of contrast medium and 24 hours later. Diagnostic criteria for Hodgkin's disease described by Abrams (1) and Davidson & Clarke (8) were used in the evaluation of the lymphographic findings.
Static B-mode ultrasound scanning with a Philips Sono Diagnost B apparat u s , and dynamic scanning with an ATL sector scanner (Mark 111, 3 PJI-Iz) were performed; during the last years only dynamic ultrasound was used. Lesions of low echogenicity around the great vessels in t h e retroperitoneal region were considered to be pathologic lymph nodes.

RESULTS
In 30 of the 45 examined patients both the lymphographic and the CT findings were considered normal ( Table 1); 22 of these 30 patients also underwent ultrasound examination (Table 2 ) , i n all cases with a negative result.
Six of these patients were laparotomized and in 5 of them all lymph node biopsies were negative. One of these patients, however, had lyrnphornatous changes in the spleen. In one patient who underwent laparotomy 2 1 months after lyrnphography and CT, all lymph node biopsies were positive; the infiltrated lymph nodes, however, were of normal size. A further 13 patients in this group have not received any therapy to the abdomen, and of these, 11 have been free from recurrence for 1 to 5 years.  In the remaining 15 patients the outcome of one or more of the examinations was positive (Tables 1 and 2 ) . In 5 of these the lymphographic diagnosis was a pathologic structural pattern in normal-sized lymph nodes. The findings at C T were negative in all these 5 patients. Four patients were examined with ultrasound with a negative result. Two of the 5 patients were laparotomized and nothing abnormal was found. In the remaining 3 patients the findings lacked histopathologic confirmation. Two of them have received abdominal radiation therapy and one patient, who has not been given such treatment, has been free from recurrence, but so far the observation period is only one year.
In 2 other patients both lymphography and ultrasound revealed solitary pathoIogic, clearly enlarged lymph nodes in the pelvis, whereas C T was negative. Nor can these lymph nodes, with diameters of 2-3 cm, be identified retrospectively on the CT scans.
In 2 further patients lymphography gave normal results, while the findings at CT were considered to be pathologic. Neither of these patients was examined by ultrasound. In one of them CT revealed enlarged lymph nodes outside the lymphographic area and in the other one the CT diagnosis was probably wrong, a s this patient did not receive any therapy to abdominal lymph nodes and is still free from recurrence after 34 years.
In 6 patients both CT and Iymphography gave positive results. As a rule the lymph nodes in these cases were only slightly to moderately enlarged ( Fig.1) and only in one case were conglomerates of lymph nodes found. Three patients in this group underwent ultrasound examination; in one of them the findings were positive and corresponded to those at both CT and lymphography. In the other two patients the ultrasound result was negative. When the lymphograms were compared with computed tomograms, it was noted that at CT slightly to moderately enlarged lymph nodes were mast easily demonstrated in the left para-aortic chain and were more difficult to identify in the para-caval chain.

DISCUSSION
In Hodgkin's disease, the diagnostic method which gives the best information about the extent of the disease in the abdomen is staging laparotomy with splenectomy. This operation has been questioned, however, as it does not improve the prognosis with certainty and as splenectomy implies a risk of fatal pneumococcal sepsis ( 2 , 9 ) , therefore in several hospitals, including our own, it is not performed. The demands upon the radiologic investigation have therefore increased further. To the left of the aorta there is a slight enlarged lymph node (arrow). Accuracy rates for C T and ultrasonography in the evaluation of retroperitoneal lymph nodes in Hodgkin's disease have been reported ( 3 , 4 , 6 , 7 , 1 1 , 121, but these series are definitely smaller than those dealing with lymphography on the same subject ( 1 5 ) . The non-invasive methods have several advantages, however, over lymphography, being easy to carry out and well accepted b y the patient. Moreoever, in contrast to lymphography, with both of these methods all lymph node groups in the abdomen as well as liver and spleen may be examined. The nodes must, however, be enlarged, as neither of the methods permits an evaluation of the internal lymph node structures.
The possibility of demonstrating lymphomatous infiltrates in iliac lymph nodes by CT has not been investigated to any great extent. L A C K N E R et coll.
(12) reported a sensitivity of 81% and a specificity of 90%, but the histopathologic verification was incomplete. In our experience the diagnostic reliability was lower. In two patients in the present material, clearly enlarged lymph nodes were not detected at CT, whereas both ultrasound and lymphography demonstrated these nodes. The less good results of CT in the pelvis may have been due partly to the fact that our examinations were performed with an older equipment with relatively low geometric resolution, and the results with modern CT equipment seem to be better.
One drawback with ultrasonic scanning of the retroperitoneal space is that the examinations are often unsuccessful because of interference by intestinal gas (16). This is largely attributable, however, to the use of the static B-mode scanning technique. When a dynamic sector scanner is used, interfering gas-containing intestines can often be displaced and the abdomen compressed with the transducer, so that the entire retroperitoneal space can be evaluated. The outcome of an ultrasound examination is largely dependent, however, upon the proficiency of the examiner. Moreover, the poorer resolu-tion of the ultrasound method implies that the sensitivity at examination of para-aortic lymph nodes is Iowcr than that of CT. Consequently, ultrasound i s insufficient as the only examination, but i s of great value as an adjunct to CT.
W e recommend CT as the initial method, as this i s sufficient with clearly positive findings and also with clearly negative findings. CT should be sufficient since the diagnostic reliability of Iymphography in discrete changes in normal-sized nodes is low (13). In cases where CT reveals solitary slightly enlarged lymph nodes, ultrasound examination and lymphography are also performed. The criteria for a positive diagnosis at Iymphography then have to be stringent, in order to avoid a large number of false positive results.