Evaluation of the surgical aspects of staging laparotomy for Hodgkin's disease in children

Evaluation of the Surgical Aspects of Staging Laparotomy for Hodgkin's Disease in Children By Toshihiro Muraji, Daniel M. Hays, Stuart E. Siegel, George Sleight, Jerry Finkelstein, Stephen A, Feig, and Darleen Powars Los Angeles, California 9 Experience with 72 children in which the type of staging laparotomy recommended by the Intergroup Hodgkin's Disease in Childhood Study (IHDCS) was employed ( 1 9 6 7 - 1 9 8 1 ) is reviewed. Laparotomy altered the stage in 3 5 % of these patients including advance in stage (I-II to Ill-IV) in 24 patients, and reduction in stage (111to II) in one patient. In adults, Stage III disease is divided into II1~ and III2 on the basis of the presence or absence of lower abdominal node involvement; and prognosis is significantly better in II1~. Nine patients from t w o additional institutions were included in a special study of Stage III disease. This included 22 children in II1~ and 11 children in III2. Although the children with Stage B (systemic symptoms) disease w e r e concentrated in III2, none of the measured difference b e t w e e n these t w o groups w e r e significant. No fatal postsplenectomy sepsis has been noted since the use of pneumococcal vaccine and prophylactic penicillin became standard. INDEX W O R D S : Hodgkin's disease; splenectomy; staging laparotomy.

HE I M P O R T A N C E of staging laparotomy

for children with Hodgkin's disease has T been recognized, and this procedure is regarded as an essential step in the evaluation of these patients in most institutions and cooperative group studies, l 5 Controversies remain regarding the justification for splenectomy in younger children, the feasibility and usefulness of partial splenectomy, and the extent of node sampling required for accuracy in evaluation. The purpose of this study is to review the results of 72 laparotomies with an attempt to establish the relative frequency of the involvement of different abdominal n o d e groups and organs in children with Hodgkin's disease. An attempt is also made to assess the prognostic implication of the subdivisions of Stage III (llIl and III2), now employed in adult series, 6"7 in pediatric patients. MATERIALS A N D METHODS Seventy-two laparotomies performed from 1967 through 1981 comprise the general study. These included all patients with Hodgkin's disease from three institutions in which the standard laparotomy which is recommended by the InterJournal of Pediatric Surgery, Vol. 17, No. 6 (December), 1982

group Hodgkin's disease in Childhood (Stage I-1l) Study (IHDCS) Committee, was performed. Patients included in this group were from the Children's Hospital of Los Angeles (51), the Los Angeles County-USC Medical Center (12), and Kaiser Permanente Hospital, Los Angeles (9). Nine additional patients with Stage III disease from two institutions were included in the special evaluation of Stage III disease in childhood. Negative laparotomies (or Stage IV patients) from these institutions were not evaluated. These were from the University of California, Los Angeles (7) and the Long Beach Memorial Hospital (2). In the general study (72 patients), the age at staging laparotomy ranged from I to 20 yr with a mean of I 1.3 yr. Clinical aspects of the course of these 72 patients are noted in Table 1. The length of post laparotomy observation in this group ranges from 4 mo to 15 yr (mean, 4.2 yr, and median, 4.1 yr). Except in two patients in which radiotherapy to extensive mediastinal involvement preceded surgery, laparotomy was carried out prior to therapy of any type. The recommended staging laparotomy included splenectomy, liver biopsies, upper abdominal lymph node (splenic hilar, superior periaortic, and porta hepatis) and lower abdominal lymph node (inferior periaortic, mesenteric, and iliac) biopsies. The technical aspect of this procedure has been described,s Excluded were patients in whom a less complete staging laparotomy was performed, with the exception of (A) those in which splenectomy was omitted to avoid postsplenectomy sepsis; and (B) those in which porta hepatis nodes were not biopsied prior to 1977. If attempts at biopsy in several node groups were unsuccessful, i.e., the specimens contained no lymphoid tissue, patients were still admitted to the series if four node groups were successfully biopsied, including at least one in the superior periaortic (above the pancreas) and one in the inferior periaortic areas. Clinical and pathological staging was performed according

From the Divisions of Hematology/Oncology of the Departments of Pediatrics or Medicine of the Los Angeles Children's Hospital, the Los Angeles County-University of Southern California Medical Center, Harbor-University of California at Los Angeles Medical Center, the University of California at Los Angeles School of Medicine, and the Southern California Permanente Medical Group, all of Los, Angeles, California. Presented before the Thirteenth Annual Meeting of the American Pediatric Surgical Association, Phoenix, Arizona, May 29-June 1, 1982. Address reprint requests to Dr. D.M. Hays at the Division of Oncology-Hematology, Children's Hospital of Los Angeles, University of Southern California, P.O. Box 54700, Los Angeles, Calif. 90054. 9 1982 by Grune & Stratton, Inc. 0022-3468/82/1706~9022501.00/0

843

844

MURAJI ET AL.

Table 1. Staging Laparotomy in Childhood: Summary of Clinical and Pathological Aspects (72 Patients) Pathological Stage I

II

Ill

IV

Total

Sex M

18

8

17

2

45

F

8

12

7

0

27

Age <5yr

3

4

1

0

8

6 - 1 0 yr

7

4

8

2

21

11-15yr

15

9

14

0

38

1

3

1

0

5

25

15

15

1

56

1

5

9

1

16

12

10

11

1

34

8

7

9

0

24

1 6 - 2 1 yr Systemic Symptoms A(absent) B (present) Pathological classification Nodular sclerosis Mixed cellularity Lymphocyte predominance

Lymphocyte depletion Undetermined Total

5

3

4

1

13

0

0

0

0

0

1

0

0

0

1

26

20

24

2

72

to Ann Arbor Classification. 9 Histological types encountered were divided according to the Lukes Criteria. ~~ The 33 children with Stage II1 disease were further divided into llll, thoracic a n d / o r cervical disease with upper abdominal node involvement only; and II12, cervical/thoracic disease and lower, with or without upper, abdominal node involvement.

RESULTS

In 25 (34.7%) of the 72 patients in the general study, preoperative stage was altered by laparotomy, including advance in stage from I to III in 11 patients; from II to Ill in 11 patients; from I to IV in 1 patient; and from II to IV in 1 patient. In one patient, stage was reduced from Stage III to Stage II (Table 2). Splenectomy was performed in 66 of these 72 patients and revealed involvement in 19 (29%), (including 2 in Stage IV); no involvement in 44 Table 2. Alteration in Stage by Leparotomy Pathological Stage I

II

III

IV

Total

26 --. 26

-19 1

11

1

38

11

1

31

--

3

2

72

Clinical Stage

I II III IV Total

Change in S t a g e - - 3 4 . 7 % .

.

2 .

20

. 24

.

(67%); and equivocal involvement in 3. Twelve of 19 involved spleens (63%) were described as of normal appearance by gross inspection at surgery. Among 66 patients, in whom splenectomy was performed, there were 3 patients who were less than 6 yr of age. One of them, age 1 yr and 9 too, died due to overwhelming sepsis 2 mo after laparotomy in 1971. The other two patients have had no episodes of severe infection. The use of polyvalent pneumococcal vaccine and prophylactic penicillin or erythromycin therapy has been the recommended practice in all participating institutions since approximately February, 1978; but antibiotic blood and urine levels have not been obtained as a measure of compliance. Six patients did not undergo splenectomy at staging laparotomy because of the grossly normal appearance of the spleen and considerations relative to post-splenectomy sepsis. Their ages at the time of laparotomy were 2, 4, 4, 4, 5, and 10 yr. Three of these 6 patients subsequently relapsed, and two died. One patient underwent a second laparotomy with splenectomy at another hospital, 1 yr after the first. The spleen was involved at that time. Total nodal irradiation was given, and the patient has been without evidence of disease for 8 yr. One patient became symptomatic, was found to have a positive bone marrow 2 yr after laparotomy, and died due to progressive disease 2 yr later. The remaining patient, who was 10 yr of age at laparotomy, became symptomatic, and splenomegaly and a palpable left inguinal lymph node was noted 3 yr after the initial procedure. She underwent a second laparotomy with disease found in the spleen, periaortic and mesenteric nodes, as well as the left inguinal nodes. Splenectomy was performed and chemotherapy (COPP) instituted. Three years later she developed recurrence in the mediastinum and the right axilla. She died of disease 8 years after the first staging laparotomy. At least 3 of the 6 patients who did not undergo splenectomy did not have adequate splenic irradiation following laparotomy. Involvement of nodes from each specific abdominal group among the 33 patients with Stage Ill disease, are listed in Tables 3 and 4. Among five patients without splenic involvement, three had involvement of splenic hillar nodes; two, superior periaortic nodes; and one, a

STAGING LAPAROTOMY FOR HODGKIN'S DISEASE

845

Table 3. E x t e n t of Abdominal Involvement in 33 Patients W i t h Hodgkin's Disease Stage III Total Number Biopsied or Excised*

Number Number With Positive Equivocal (%) Histologyt

Superior Abdominal Spleen Splenic hilar nodes Superior periaortic nodes Porta hepatis nodes

32 19

24 (75) 12 (63)

3

25 9

11 (44) 5 (55)

1

Inferior abdominal Inferior periaortic nodes Mesenteric nodes lilac nodes

24 24 20

5 (21) 4 (17) 5 (25)

1 1

*Patients with either single or multiple nodes taken from each node group. tAdditional patients with abnormal histology but without unequivocal features of Hodgkin's disease. Tables 3, 4, and 5 include patients from the unselected group (72) patients, and in addition nine selected patients with Stage III disease from two additional institutions (see Materials and Methods).

porta hepatis node. No patient in this group with uninvolved spleens had involvement of lower abdominal lymph node groups (Table 4). Three patients had equivocal splenic involvement, and in one of these definite abdominal involvement was confined to an iliac node. Among 33 patients with Stage Ill disease (Table 5), 22 patients were identified at IlI~, and 11 as II12. Five out of 11 patients with Stage 1II2, and 6 out of 22 patients in Stage III~ were symptomatic (Stage B). Three patients in Stage III~ had recurrence of disease which was controlled, while none of the patients with Stage 1112 had recurrence during the period of observation. However, two patients with Stage 1112 expired,

Table 5. Characteristics of Patients in Stage III Subtypes Stage III1-1112 Total No. M/F

III1

1112

22

11

15/7

8/3

Clinical stage I II III

10 10 2

5 5 1

Systemic symptoms A B

15 6

6 5

Histology Nodular sclerosis Mixed cellularity Lymphocyte predominance

11 10 11

5 3 3

Relapse/death

3/0

0/2

one secondary to severe bone marrow depression 15 mo after initiation of therapy; and the other with acute myelogenous leukemia, 4 yr after laparotomy. Nine of 72 patients in the general study expired (12.5%). All deaths occurred in patients admitted to the study prior to 1977. The causes of death, and histological types are listed in Table 6. Four patients have hypothyroidism, probably secondary to irradiation, and have required exogenous thyroid. One patient has amenorrhea secondary to pelvic irradiation and chemotherapy. One patient developed intestinal obstruction requiring lysis of adhesions (no resection) 4 wk after the staging laparotomy, and a second had repeated episodes of obstruction requiring intestinal intubation. No patient treated with pneumococcal vaccine and prophylactic penicillin or erythromycin has had a known septic episode.

Table 4. Splenic Involvement and Upper Abdominal Node I n v o l v e m e n t - - ( S t a g e III)

Spleen Positive (24 patients)

Splenic hilar nodes Superior periaortic nodes Porta hepatis nodes

Spleen Negative (5 patients)

Splenic hilar nodes Superior periaortic nodes Porta hepatis nodes

Positive

Negative

Not Biopsied

11 (73%)* 8 (44%) 4 (57%)

4 9 3

9 6/11" 17

1 3 2

1 0 2

3 2 1

*Percent of total node groups biopsied in this location which were positive. tAbnormal, but without unequivocal features of Hodgkin's disease. Among 24 patients with positive spleens five had lower abdominal node involvement. No patient with a negative spleen had lower abdominal node involvement.

846

MURAJI ET AL.

Table 6. Deaths in the Series Age at

Histological Type

Years from Laparotomyto Death

Diagnosis

Sex

Pathological Stage

1.

4

F

IA

NS

4

2.

10

F

IA

NS

8 0

Cause of Death Progressive Progressive Sepsis Progressive Progressive Progressive

disease disease

3.

1

M

IIA

LP

4.

10

F

IIA

NS

1

5.

14

F

I IA

MC

1

6.

14

F

liB

NS

15

7.

7

M

IliA

MC

2

8.

10

F

IliA

NS

4

Acute myelogenous leukemia

9.

10

M

IVB

NS

5

Progressive disease

DISCUSSION

This series of 72 patients has the same general characteristics noted in previous reports, 2'3'4'11 including a male preponderance and identification of nodular sclerosis as the most common histologic subtype. In this series, the mortality was the same in the asymptomatic (A) as in the symptomatic (B) patients (12% in both cases). Staging laparotomy altered the clinical stage in 35% of the patients in this series, which is the same as in the previous report 4 from the same institutions in 1976. Alterations in stage resulting from laparotomy in this study continue to be from Clinical Stage I-II to Pathological Stage III, rather than the reverse which has been reported in some groupsJ z These differences between institutional series reflect the clinical interpretation of minimal enlargement of spleen and liver. The importance of biopsy of all of the standard node groups within the abdomen was illustrated in this series, as in two patients the basis for placing the patient in Stage III was involvement of a single node from a node group not routinely biopsied in some studies (mesenteric and iliac). The particularly frequent involvement of the porta hepatis nodes is also significant, as this node group was rarely included in the sampling prior to 1977. A collaborative study of Children's Cancer Study Group reported by Chilcote ~3 in 1976 showed a 10% of incidence of fulminant septicemia in 200 children, and half of them died. In this series, one patient died from postsplenectomy overwhelming sepsis in 1971. Among six patients who did not undergo splenectomy in the present series, three had recurrence, and two died. Both biopsy of splenic perihilar (or other

disease disease disease Bone marrow depression

superior abdominal) lymph nodes and also partial splenectomy ]4 have been suggested as procedures which could be carried out to eliminate the need for total splenectomy. The accuracy of these procedures in predicting splenic involvement would n o t appear to be sufficient to warrant their routine use. ]5 The status of the splenic hilar nodes did not necessarily reflect splenic involvement, in this series (Table 4). Polyvalent pneumococcal vaccine with prophylactic penicillin or erythromycin has been recommended following laparotomy for the past 3-4 yr in all institutions in the study. No patient treated by this protocol has experienced an episode of sepsis. Ammann et alJ 6 showed that antibody response to pneumococcal vaccine was not altered by splenectomy, per se. Patients who are > 2 yr of age are expected to have an adequate response, in contrast to younger patients, and 71/72 in this series were beyond infancy. However, as pneumococcal vaccine is less effective in terms of antibody response when it is given to patients who have been previously treated for Hodgkin's disease with chemotherapy, radiotherapy 17 or both, patients should be immunized soon after diagnosis and prior to institution of therapy. ]7,~8 The substages of Stage III (III~ and III2), based on extent of abdominal node involvement, effect prognosis significantly in adult series; and consequently treatment for substage IIIl has been modified6"7in some centers, and is similar to regimens for Stage II, i.e., radiotherapy only with salvage chemotherapy when required. No differences between IIIl and III2 were noted in this series, but the number of patients with Stage III in this study was not large enough to make statistical analysis of the differences in substage possible. Lymphangiography is unquestionably an aid

STAGING LAPAROTOMY FOR HODGKIN'S DISEASE

847

in identifying involvement of iliac and periaortic lymph nodes. Technical difficulty with this procedure varies inversely with patient age, and it was not used consistently in the institutions in this study. In general, those patients that had lymphangiograms were treated identically to those in which the study was not performed,

except for those patients in which there were positive or equivocally involved nodes, which were removed for study at the time of laparotomy. The effects of a previous lymphangiogram in increasing the accuracy of the staging laparotomy could not be evaluated from analysis of patients in this study.

REFERENCES

1. Rosenstock JG, D'Angio G J, Kiesewetter WB: The incidence of complications following staging laparotomy for Hodgkin's Disease in children. Am J Roentgenol Rad Ther Nucl Med 120:531-535, 1974 2. Filler RM, Jaffe N, Cassady JR, et al: Experience with clinical and operative staging of Hodgkin's disease in children. J Pediatr Surg 10:321-328, 1975 3. Donaldson SS, Glatstein E, Rosenberg SA, et al: Pediatric Hodgkin's disease II. Cancer 37:2436-2447, 1976 4. Cohen IT, Higgins GR, Powars DR, et al: Staging laparotomy for Hodgkin's disease in children. Arch Surg 112:948-951, 1977 5. Dearth JC, Gilchrist GS, Bergert EO Jr, et al: Management of stage I to II1 Hodgkin's disease in children. J Pediatr 96:829-836, 1980 6. Stein RS, Hilborn RM, Flexner JM, et al: Anatomical substages of stage III Hodgkin's disease. Implications for staging, therapy, and experimental design. Cancer 42:429436, 1978 7. Stein RS, Golomb HM, Diggs CH, et al: Anatomic substage of stage II1-A Hodgkin's disease: A collaborative study. Ann Intern Med 92:159-165, 1980 8. Hays DM, Karon M, lsaacs H, et al: Technique and results of staging laparotomy in childhood. Arch Surg 106:507-512, 1973 9. Carbone PP, Kaplan HS, Musshoff K, et al: Report of the committee on Hodgkin's disease staging classification. Cancer Res 31:1860-1861, 1971.

10. Lukes R J, Butler J J, Hicks EB: Natural history of Hodgkin's disease as related to its pathologic picture. Cancer 19:317-344, 1966 11. Poppema S, Lennert K: Hodgkin's disease in childhood. Histopathological classification in addition to age and sex. Cancer 45:1443-1447, 1980 12. Glatstein E, Guernsey JM, Rosenberg SA, et al: The value of laparotomy and splenectomy in the staging of Hodgkin's disease. Cancer 24:709-718, 1969 13. Chilcote RR, Baehner RE, Hammond GD, et al: Septicemia and meningitis in children splenectomized for Hodgkin's disease. N Engl J Med 295:793-800, 1976 14. Boles ET Jr, Haase GM, Hamoudi AB: Partial spleneetomy for Hodgkin's disease: An alternative approach. J Pediatr Surg 13:581-584, 1978 15. Dearth JC, Gilchrist GS, Telander RL: Partial splenectomy for staging Hodgkin's disease: Risk of false-negative results. N Engl J Med 299:345-346, 1978 16. Ammann A J, Addiego J, Wara DW: Polyvalent pneumococcal polysaccharide immunization of patients with sickle cell anemia and patients with splenectomy. N Engl J Med 297:897-900, 1977 17. Levine AM, Overturf GP, Field RF, et al: Use and efficacy of pneumococcal vaccine in patients with Hodgkin's disease. Blood 54:1171-1175, 1979 18. Addiego JE Jr, Ammann AJ, Schiffman G, et al: Response to pneumococcal polysaceharide vaccine in patients with untreated Hodgkin's disease. Lancet 2:450--452, 1980

Discussion T. Boles (Columbus, Ohio): This excellent study confirms the value of laparotomy in the staging of Hodgkin's disease. The study also emphasizes the importance of the complete staging laparotomy that the IHDS protocol has emphasized. In two instances a single positive lymph node was the only evidence of Hodgkin's disease in the abdomen. The only issue is the necessity for a total splenectomy. The lymph nodes and liver are only biopsied, whereas the spleen is removed, subjecting the patient to the hazard of post splenectomy sepsis. There are a number of factors important in this issue. Splenectomy is a diagnostic not a therapeutic proce-

dure. The spleen is positive in most series in about 35% of staging laparotomies, so that clearly a large number of normal as well as a smaller number of involved spleens are inevitably removed. Of those spleens which are involved, the best data indicates that in about 10%, the disease would be missed if a partial rather than total splenectomy were performed. If this is transposed to the entire number, not just the 30-odd-percent in which the spleen is involved, the risk of understaging by doing a partial rather than total splenectomy is around 2%-3%. The risk of post splenectomy sepsis has been markedly reduced by active immunization against

848

pneumonococcus. The risk of such sepsis in adults and at long time intervals following splenectomy has been documented by Robinette and Fraumeni in a study in Lancet about 5 yr ago, and by Stendage and Goss in the current issue of the American Journal of Surgery. The current preventive measures may indeed essentially eliminate post splenectomy sepsis as a risk of this procedure, but at the present time we do not have the final answer. I would like to ask the authors concerning the cases with Stage III disease in which the spleen was positive. How many of these cases showed positive lymph node biopsies when all biopsied lymph node groups are concerned? D. Johnson (Salt Lake City, Utah): The major concern over this whole business of staging has been removal of the spleen. Dr. Muraji stated that since they have used prophylaxis with pneumococcal vaccine and penicillin, they have seen no evidence of post splenectomy sepsis. I don't understand why this is completely protective. Only a portion of the organisms are affected by pneumococcal vaccine and I don't understand why penicillin protects against hemophilus and E. coli and some of the other organisms which have been reported with post splenectomy sepsis. Would it not be more rational at least to use Ampicillin rather than penicillin? J. Ternberg (St. Louis, Missouri): How many of the positive spleens would downstage following

MURAJI ET AL.

splenectomy, i.e. from Stage II from a III if lymph nodes were negative? T. Muraji (closing): Dr. Boles, the concept of partial splenectomy in the staging laparotomy is fascinating, and we hope to have the results of your prospective study of partial splenectomy with a long-range follow-up. Isolated splenic involvement was seen in approximately 10% of the patients with Stage III. Dr. Johnson, I think the use of ampicillin is reasonable and would be acceptable to most pediatricians. Dr. Ternberg believes that if the spleen is the only site of abdominal involvement, and is removed, the patient can be treated by a Stage I1 protocol, avoiding more intensive therapy. In most pediatric studies and institutions such patients are now treated as are the other forms of Stage III. D. M. Hays (closing): The antibody response depends on when the vaccination is given. If it is given early when the patient's disease is not advanced, resulting antibody levels are high. If it is given after the patient has received chemotherapy and radiotherapy, the levels may be quite low. In the majority of our cases, polyvalent vaccine was given at diagnosis, frequently prior to splenectomy. Vaccination performed before or soon after splenectomy is equally effective. I believe that the patients in this series should have relatively high antibody titers, but they have not been measured.