Br. J. Surg. Vol. 66 (1979) 62-65
Splenectomy for undiagnosed splenomegaly A. G O O N E W A R D E N E , J. B. B O U R K E , R . F E R G U S O N A N D P . J . T O G H I L L * SUMMARY
Patients and methods
During the 9-year period 1968-76 116 splenectomies were performed at the General Hospital, Nottingham. Of these, 13 (I1 per cent) were undertaken for unexplained splenomegaly. In 6 patients a diagnosis was established by the operative procedure (2 with sarcoidosis, 2 splenic cysts, 1 Gaucher’s disease and 1 haemangiosarcoma). Histological examination of the excised spleens in the remaining 7 patients showed no specific features. Two of these patients benefited considerably from removal of very large spleens. Another patient died from lymphosarcoma which was diagnosed 21 months after splenectomy. In the remaining 4 patients with mild to moderate splenomegaly, there were no real diagnostic or therapeutic advantages. It is concluded that splenectomy should always be considered in patients with unexplained moderate or gross splenomegaly but it may not be heEpful in the patient whose spleen is only mildly enlarged.
The 13 patients whose clinical data are summarized in Table I formed part of a series of 116 splenectomies carried out at the General Hospital, Nottingham, during the period 1968-76. Of the 116 splenectomies, there were 21 associated with major upper abdominal surgery, 32 for traumatic or spontaneous rupture of the spleen, 17 for idiopathic thrombocytopenic purpura, 12 for various forms of haemolytic anaemia, 13 for lymphoma, 5 for chronic leukaemia and 3 for myelosclerosis. The remaining 13 cases were undiagnosed before splenectomy and are reported in this paper. Each of the patients subjected to splenectomy for undiagnosed splenomegaly had had repeated peripheral blood film examinations and bone-marrow and/or trephine biopsies. Blood from all patients was assessed for haemolysis by routine haematological methods. Measurements of red cell mass (RCM), splenic red cell pool (SRCP), red cell survival (RCS) and surface counting studies were also carried out in Cases 3, 4, 7, 8, 9 and 11, using ”Cr-labelled erythrocytes as described earlier (Toghill and Green, 1975). Plasma volumes (PV) were measured using lZ5I- or 1311-labelledhuman albumin. The results of these investigations are shown in Table ZI. In only one patient (Case 11) was there evidence of frank haemolysis with excessive splenic uptake on surface counting. This patient’s red cell survival (T5,, W r ) was reduced to 10 days (normal 24-28) and the splenic sequestration index increased to 250. This index was calculated as described by Parker et al. (1977), a value of greater than 100 indicating moderate to severe red cell destruction in the spleen (Jandl et al., 1956). All patients had routine liver function tests performed and all had liver biopsies except Cases 8, 10 and 11, in whom this procedure was contraindicated by a haemorrhagic tendency. Liver and spleen scans were performed in all patients using agTcm with routine anterior, posterior and lateral views. Evidence of oesophageal varices was sought by endoscopy or radiology. None of the patients had clinical lymphadenopathy and diagnostic scalene or other lymph node biopsy was not performed before the laparotomy for splenectomy. Eight of the 13 patients had lymphangiograms, but these were all interpreted as being normal. Diagnostic splenic aspiration was not utilized. At laparotomy for splenectomy several lymph nodes were taken for biopsy purposes from the root of the mesentery and the para-aortic regions.
IN most patients living in temperate climates, the cause of an enlarged spleen is readily determined by clinical investigation. Even after careful evaluation, however, a small group of patients remains in which the cause of the splenomegaly is uncertain. Some of these patients appear to be in robust health whilst others are obviously ill. When non-invasive techniques have failed to reveal the cause of the splenic enlargement, splenectomy has to be considered. Under these circumstances, the primary objective of splenectomy is positive pathological diagnosis. Additionally, there may be other secondary benefits, such as the cure of a hypersplenic state, the reduction of an expanded plasma volume or the removal of a large tumour which has become a painful and cumbersome burden. These advantages have to be set against the appreciable morbidity and mortality of splenectomy and the long term vulnerability to infection of the asplenic state (Ramsay and Bouskill, 1973). Unfortunately, there is the possibility that the excised splenic tissue may have no specific histological characteristics, leaving the clinician uncertain as to whether he is dealing with a benign disorder or a more sinister process, presaging a lymphomatous or leukaematous state. The potentials of diagnostic splenectomy have been discussed in standard texts but there have been few detailed reports in recent years. Herman and his colleagues (1968) described 52 patients who underwent splenectomy for unexplained splenomegaly at the Cleveland Clinic during the preceding 10 years. Of these patients, more than half were found to be suffering from lymphoma or congestive splenomegaly and in only 1 case was the diagnosis still uncertain after splenectomy. Although other authors have reported groups of patients with specific diseases that were diagnosed at splenectomy (Hickling, 1964; DasGupta et al., 1965; Skarin et al., 1971), there have been no recent descriptions of the results of splenectomy for undiagnosed splenomegaly after full conventional investigations. This paper describes 13 such patients.
Findings after splenectomy The results of diagnostic splenectomy are shown in Table I . There was no immediate operative mortality, although one patient (Case 8) died 5 weeks after her operation without leaving hospital. Patients with positive findings at splenectomy A definite histological diagnosis was achieved in 6 of the 13 patients. Two patients (Cases 2 and 3 ) who came to medical attention because of pain in the left upper quadrant were shown to have congenital splenic cysts. The diagnosis had been suggested by patchy uptake on the spleen scans. A youngster of 16 (Case 6) had been found to have an enlarged spleen at routine school medical examination and splenectomy revealed the diagnosis of Gaucher’s disease. No other members of the family were affected and there were no other physical stigmas of the disease. Bone-marrow biopsy and trephine and liver biopsy had been unhelpful. An elderly man (Case lo), previously reported (Toghill et al., 1972), was shown to have haemangiosarcoma. Rupture of the spleen precipitated emergency surgery a few hours before splenectomy had been planned. Case 1
* General Hospital, Nottingham. Correspondence to: P. J. Toghill.
63
Splenectomy for undiagnosed splenomegaly Table I: HAEMATOLOGICAL FINDINGS BEFORE SPLENECTOMY AND Known duration Age at of Weight splensplenomegaly of ectomy WBC Platelets before spleen Case Sex (yr) Hb (g/dl) ( x 108/1) ( x 108/1) splenectomy (g) 1980 6.0 430 2 wk (a) 12.6 28 M 1 6.3 420 (b) 14.6 I310 6.5 250 M (a) 14.9 14 2 1 yr 1290 11.0 280 (a) 14.5 27 F 3 2 yr 2690 4 mth 4.4 125 (a) 10.8 66 F 4 10.8 391 (b) 13.6
AT LAST FOLLOW-UP
Histopathology of excised spleen Sarcoidosis Splenic cyst Splenic cyst Non-specific, (?) idiopathic, non-tropical splenomegaly Non-specific
(a) 8.3 (b) 9.4 (a) 12.3 (b) 11.7 (a) 14.0 (b) 1 4 9 (a) 8.9
6.2 10.8 5.7 17.4 7.0 7.1 10.4
120 184 127 190 350 410 30
2 yr
245
14 mth
830
8 wk
520
Gaucher’s disease Non-specific
2 mth
304
Non-specific
45
(a) 5.3 (b) 13 8
1.9 1.4
52 28 1
3 mth
1570
M
72
(a) 5.3
5.5
50
2 wk
1125
11
F
73
(a) 8.9
1.5
78
3 mth
660
Non-specific, (?) idiopathic non-tropical splenomegaly Haemangios ar coma Non-specific
12
M
30
Sarcoidosis
31
250 300 25 76
200
M
7.8 14.6 1.2 4.9
3 mth
13
(a) (b) (a) (b)
2 mth
620
Non-specific
5
M
60
6
F
16
7
M
31
8
F
54
9
M
10
13.8 15.0 9.5 14.6
Outcome after splenectomy Remains well on steroid therapy 3 yr later Discharged Discharged Remains well 2 yr later
Remains in poor health 21 yr later Remains well 2 yr later Remains well 3 yr later Died 5 weeks after splenectomy with thrombocytopenia Remains well 8 yr later
Died 3 mth later Well for 21 months after splenectomy. Died from lymphosarcoma 2 yr after splenectomy Well 18 mth later Remains in excellent health 6 mth later
(a) Findings before splenectomy; (b) findings at last follow-up. was admitted to hospital as an emergency with a pneumothorax and was found to have an enormously enlarged spleen. Diagnostic splenectomy revealed sarcoidosis, which had seemed unlikely on clinical grounds in view of the massive splenomegaly. Case 12, who presented with pruritus, was confidently expected to have a lymphoma, but histological examination of both spleen and abdominal nodes showed sarcoidosis. Patients with non-specific histology There were no specific diagnostic features at splenectomy in the other 7 cases. In one patient (Case 11), the splenomegaly preceded the development of overt lymphosarcoma by almost 2 years. This patient had been investigated extensively preoperatively for a Coombs negative haemolytic anaemia, leucopenia and thrombocytopenia. Surface counting using 51Cr-labelled erythrocytes showed excess splenic uptake. Splenectomy resulted in a gratifying haematological improvement for 21 months until the patient showed clinical evidence of lymphosarcoma which caused her death 3 months later. Two patients (Cases 4 and 9) had massive spleens that were associated with some features of hypersplenism. Neither of these patients had evidence of excessive splenic uptake on surface counting but each had large splenic red cell pools of 13 and 28 per cent of the RCM respectively. In addition, the plasma volume in each case was increased above the normal value of 34-48 ml/kg. Much of the initial haematological improvement after splenectomy in these patients was likely to have been due to the removal of the large SRCP and reduction of the expanded plasma volume. Both patients have now lived for 2 and 8 years respectively in excellent health. In 4 other patients with mild to moderately enlarged spleens weighing from 245 to 800 g splenectomy failed either to achieve a diagnosis or to improve the health of the patients, although
Table 11: RESULTS O F HAEMATOLOGICAL INVESTIGATIONS SRCP RCS Surface counting PV RCM (% of Case (ml/kg) RCM) (Ts051Cr) studies (ml/kg) ~~
3
24.0
0
24 d
4
26.5
13
22 d
7
19-7
18
18 d
8
16.2
9
13.8
28
22 d
1I
23.7
6
lOd
5.5
14.5 d
No excess splenic uptake Noexcesssplenic uptake No excess splenic uptake No excess splenic uptake No excess splenic uptake Excesssplenicuptake (splenic index 250)
38 50 NE NE 60 NE
+
RCM, Red cell mass (normal 25-36 ml/kg); SRCP, splenic red cell pool; RCS, red cell survival (normal 24-28 d); PV, plasma volume (normal 34-48 ml/kg); NE, not estimated. in Case 13 there was a satisfactory rise in the platelet count from a worryingly low level of 30 x 109/1to 87 x 109/l. Case 8 had been investigated extensively preoperatively for mild splenomegaly, a normochromic anaemia and thrombocytopenia. This patient died 5 weeks after splenectomy from continuing thrombocytopenia and subsequent autopsy failed to reveal a specific disease process. Cases 5 and 7 have now been followed up for 2& and 3 years respectively and neither has developed any new or diagnostic features.
64
A. Goonewardene et al.
Table HI: DIAGNOSTIC VALUE OF SPLENECTOMY IN SOME REPRESENTATIVE WORLD SERIES No. in whom diagnosis Diagnostic No. uncertain splenof at ectomies (79 Period caqes Type . . of material splenectomy .. (a)
1956-66 (b) 1963-68
582
All splenectomies
52
9
41
Haematological disorders
0
0
1947-70
200
Haematological disorders
6
3
(d) Not stated
104
Haematological disorders
4
4
406
All splenectomies
12
3
(C)
(C)
1961-72
(f)
1944-70
53
Haematological disorders
0
0
(R) 1967-71
80
Haematological disorders
I0
12.5
(h) 1968-76
116
All splenectomies
13
11
Series and origin: (a)Herman et al. (1968), USA. (b) Christensen et al. (1970), Denmark. (c) Schwartz et al. (1970), USA. (a) Nightingale et al. (1971), UK. ( e ) McKinnon et al. (1973), New Zealand. (f)Ogawa et al. (1974), Japan. (g) lkkala et al. (1974), Finland. (h) Present series (1978), UK.
Discussion Many of the larger series of splenectomies reported in the past 10 years have included a proportion of patients in whom the diagnosis was uncertain at the time of splenectomy. Tabk I11 includes some representative series from various parts of the world, indicating that up to 12.5 per cent of patients who undergo splenectomy are undiagnosed at the time of operation. In the largest group, Herman and his colleagues (1968) considered that 52 of their 582 patients had had splenectomies primarily for diagnostic purposes, With the more liberal use of liver biopsy, lymphangiography and organ scanning by ultrasound or computerized axial tomography, however, it is likely that fewer cases in the future will need to proceed to splenectomy for diagnosis. Many authors have pointed to the value of splenectomy in the initial diagnosis of lymphoma (Hyatt et al., 1970; Schwartz et al., 1970; Skarin et al., 1971; Long and Aisenberg, 1974; Case Records of the Massachusetts General Hospital, 1976). Herman and colleagues (1968) found 31 per cent of their cases of diagnostic splenectomy to have lymphoma. However, the wide spectrum of diseases that may be delineated by diagnostic splenectomy includes sarcoidosis (Herman et al., 1968), splenic cysts and abscesses (Herman et al., 1968; Chulay and Lankerani, 1976), primary tumours of the spleen (DasGupta et al., 1965), splenic angioma (Rhochmis, 1970) and splenoma (Bhagwat et al., 1975). In our 5 patients in whom an accurate histological diagnosis was achieved it was gratifying to be able to assure Cases 2 and 3 that they had benign disorders. previously both had undergone extensive investigation and they had been given a guarded prognosis. It was
thought that Cases 1 and 6 might be suffering from lymphoma, though splenectomy showed sarcoidosis and Gaucher’s disease respectively, diseases with a somewhat better prognosis. Nevertheless, it must be remembered that splenectomy in Gaucher’s disease, though proving the diagnosis, may possibly result in an extension of the osseous and hepatic lesions (Matoth and Fried, 1965) and may also render the child vulnerable to the fulminating infections known to occur in the asplenic state (Ramsay and Bouskill, 1973). Our own series is characterized by the high proportion of cases (7 out of 13) in which the splenic histology was non-specific, though we realize that some of these cases may eventually proceed to frank lymphoma, as did Case 11. Undoubtedly enlargement of the spleen may be a foreboding of subsequent lymphoma, and information derived from diagnostic laparotomy and staging splenectomy in Hodgkin’s disease indicates that considerably enlarged spleens may be free from disease (Glatstein et al., 1970; Ell et al., 1975). We believe that two of the patients (Cases 4 and 9) are probably examples of the syndrome previously described as idiopathic, non-tropical splenomegaly (Dacie et al., 1969) or primary splenic hyperplasia (Weinstein, 1964). In these syndromes, the spleen may be grossly enlarged without there being consistent or significant histological changes. Splenectomy produces a clinical and haematological improvement without serious disease necessarily developing. Both Cases 4 and 9 derived relief from the removal of large uncomfortable spleens, and from the effects of the ‘big spleen’ syndrome, including pooling of erythrocytes and platelets in the spleen and expansion of the plasma volume. Splenectomy failed to establish a diagnosis in the 4 other patients (Cases 5, 7, 8 and 13), and as these patients had only mildly enlarged spleens, there were no other benefits. Certainly their spleens were significantly enlarged (DeLand, 1970; Myers and Segal, 1974) and, in association with obvious haematological abnormalities, there was some basis for diagnostic splenectomy. Our conclusions are that in some cases splenectomy for unexplained splenomegaly may be a useful diagnostic procedure. When splenomegaly is only mild and there is no evidence of disease elsewhere, the indications for diagnostic splenectomy are arguable. Routine examinations of college freshmen have shown 1 per cent with palpable spleens for which no sinister cause was found after a 6-year follow-up (McIntyre and Ebaugh, 1967). There have also been suggestions that splenic infarction might be responsible for some cases of mild to moderate splenomegaly in otherwise healthy folk, and although results of angiography and scanning may be abnormal, the gross changes may be non-specific (Russell et al., 1976). With the larger spleens the diagnostic yield after splenectomy is likely to be greater, but it must be admitted that even with modern techniques some of the large spleens will be undiagnosed after splenectomy. Nevertheless, in patients with the ‘big spleen’ syndrome, other benefits may accrue simply from the removal of the very large organ.
Acknowledgements We thank Trent Regional Health Authority and the Dunhill Research Trust for financial support.
Splenectomy for undiagnosed splenomegaly References
et al. (1973) Splenectomy : indications, results and complications in 406 patients. Ann. Surg. 39, 72-74. MATHOTH Y. and FRIED K. (1965) Chronic Gaucher’s disease: clinical observations on 34 patients. Israel J . Med. Sci. 1, 521-530. MYERS J. and SEGAL R. J . (1974) Weight of the spleen. I. Range of normal in a non-hospital population. Arch. Pathol. 98, 33-35. NIGHTINGALE D., PRANKERD T. A. J., RlCHARDS J. D . M. et al. (1971) Splenectomy in anaemia. Q . J. Med. 41, 261-267. OGAWA Y., KOBAYSHI M., SAKU M. et al. (1974) Jap. J. Surg. 4, 21-28. PARKER A. c., MACPHERSON A. I. s. and RICHMOND J. (1977) Value of radiochromium investigation in auto-immune haemolytic anaemia. Br. Med. J. 2, 208-209. RAMSAY L. E. and BOUSKILL K. c. (1973) Fatal pneumococcal meningitis in adults following splenectomy: two case reports and a review of the literature. J . R . Nav. Med. Serv. 59, 102-1 14. RHOCHMIS P. G. (1970) Splenic angioma with thrombocytopenia. N Y State J. Med. 70, 2133-2136. RUSSELL c. D., JONES A. E., JOHNSTON G. s. and HERDT J. R . L. (1976) Arteriographically confirmed focal defect in colloid spleen scan with no gross pathologic lesions; case report. J. Nucl. Med. 15, 376-377. SCHWARTZ s. I., BERNARD R. P., ADAMS J. T. and BAUMAN A. w. (1970) Splenectomy for haematologic disorders. Arch. Surg. 101, 338-347. SKARIN A. T., DAVEY F. R. and MOLONEY w. c. (1971) Lymphosarcoma of the spleen. Arch. Intern. Med. 127, 259-265. TOGHILL P. J. and GREEN s. (1975) Haematological changes in active chronic hepatitis with reference to the role of the spleen. J . Clin. Pathol. 28, 8-1 1. TOGHILL P. J., RIGBY c. c. and HALL G. F. M. (1972) Haemangiosarcoma of the spleen. Br. J. Surg. 59, 406-408. WEINSTEIN v. F. (1964) Haemodilution anaemia associated with simple splenic hyperplasia. Lancet 2, 218-223. MCKINNON W . M. P., SANDERS H. S., ZAMORA L. F.
et al. (1975) SplenOma with portal hypertension. Br. Med. J. 3, 520, CASE RECORDS OF THE MASSACHUSETTS GENERAL HOSPITAL (1976) N . Engl. J. Med. 295, 828-834. CHRISTENSEN B. E., HANSEN L. K., KRISTENSEN J. K. et al. (1970) Splenectomy in haematology. Scand. J. Haematol. 7, 247260. CHULAY J. D. and LANKERANI M. R. (1976) Splenic abscess. Am. J. Med. 61, 503-511. DACIE J. v., BRAIN M. c., HARRISON c. v. et al. (1969) Nontropical idiopathic splenomegaly (primary hypersplenism) : a review of ten cases and their relationship to malignant lymphoma. Br. J. Haematol. 17, 317-333, DASGIJPTA T., COOMBES B. and BRASFIELD R. D. (1965) Primary malignant neoplasms of the spleen. Surg. Gynecol. Obstet. 120,947-960. DELAND F. H. (1970) Normal spleen size. Radiology 97,589-592. ELL P. J., BRITTON K. E., FARRER-BROWN G. el a]. (1975) An assessment of the value of spleen scanning in the staging of Hodgkin’s disease. Br. 1. Radiol. 48, 590-593. GLATSTEIN E., TRUEBLOOD H. W . , ENRIGHT L. P. et al. (1970) Surgical staging of abdominal involvement in unselected patients with Hodgkin’s disease. Radiology 97, 425432. HERMAN R. E., DEHAVEN K. E. and HAWK w. A. (1968) Splenectomy for the diagnosis of splenomegaly. Ann. Surg. 168, 896-900. HICKLING R . A. (1964) Giant follicle lymphoma of the spleen: a condition closely related to lymphatic leukaemia but apparently curable by splenectomy. Br. Med. J . 2,787-790. HYATT D. F., SKARlN A. T., MOLONEY W . C. et al. (1970) S p h ectomy for lymphosarcoma. Surg. Gynecol. Obstet. 131, 928-932. IKKALA E., KIVILAAKSOE. and HASTBACKA J. (1974) Splenectomy in blood diseases. Ann. Clin. Res. 6, 290-299. JANDL 1. H., GREENBERG M. S., YONEMOTO R. H. et al. (1956) The clinical determination of sites of red cell sequestration in haemolytic anaemias. J . Clin. Inoest. 35, 842-867. LONG J. c. and AISENBERG A. c. (1974) Malignant lymphoma diagnosed by splenectomy and idiopathic splenomegaly. Cancer 33, 1054-1061. MCINTYRE 0. R. and EBAUGH F. R. (1967) Palpable spleens in college freshmen. Ann. Intern. Med. 66, 301-306. BHAGWAT A. G., DATTA D . V., MITRA S.
65
Paper accepted 26 June 1978.
Erratum The publishers regret that in the November 1978 issue of the British Journal of Surgery, in the article by George A. Khoury et al., ‘Tuberculosis of the peritoneal cavity’, on pp. 808 and 809 Figs. 1 and 3 were transposed. It has also been brought to OUT attention that the article by L. E. Hughes and J. F. Forbes ‘Early breast cancer’, Fig. 2 on p. 760 is incorrectly labelled: c should read a and vice versa. Also, on p. 768 the reference to Table III should read Table II.
Splenectomy for undiagnosed splenomegaly A. G O O N E W A R D E N E , J. B. B O U R K E , R . F E R G U S O N A N D P . J . T O G H I L L * SUMMARY
Patients and methods
During the 9-year period 1968-76 116 splenectomies were performed at the General Hospital, Nottingham. Of these, 13 (I1 per cent) were undertaken for unexplained splenomegaly. In 6 patients a diagnosis was established by the operative procedure (2 with sarcoidosis, 2 splenic cysts, 1 Gaucher’s disease and 1 haemangiosarcoma). Histological examination of the excised spleens in the remaining 7 patients showed no specific features. Two of these patients benefited considerably from removal of very large spleens. Another patient died from lymphosarcoma which was diagnosed 21 months after splenectomy. In the remaining 4 patients with mild to moderate splenomegaly, there were no real diagnostic or therapeutic advantages. It is concluded that splenectomy should always be considered in patients with unexplained moderate or gross splenomegaly but it may not be heEpful in the patient whose spleen is only mildly enlarged.
The 13 patients whose clinical data are summarized in Table I formed part of a series of 116 splenectomies carried out at the General Hospital, Nottingham, during the period 1968-76. Of the 116 splenectomies, there were 21 associated with major upper abdominal surgery, 32 for traumatic or spontaneous rupture of the spleen, 17 for idiopathic thrombocytopenic purpura, 12 for various forms of haemolytic anaemia, 13 for lymphoma, 5 for chronic leukaemia and 3 for myelosclerosis. The remaining 13 cases were undiagnosed before splenectomy and are reported in this paper. Each of the patients subjected to splenectomy for undiagnosed splenomegaly had had repeated peripheral blood film examinations and bone-marrow and/or trephine biopsies. Blood from all patients was assessed for haemolysis by routine haematological methods. Measurements of red cell mass (RCM), splenic red cell pool (SRCP), red cell survival (RCS) and surface counting studies were also carried out in Cases 3, 4, 7, 8, 9 and 11, using ”Cr-labelled erythrocytes as described earlier (Toghill and Green, 1975). Plasma volumes (PV) were measured using lZ5I- or 1311-labelledhuman albumin. The results of these investigations are shown in Table ZI. In only one patient (Case 11) was there evidence of frank haemolysis with excessive splenic uptake on surface counting. This patient’s red cell survival (T5,, W r ) was reduced to 10 days (normal 24-28) and the splenic sequestration index increased to 250. This index was calculated as described by Parker et al. (1977), a value of greater than 100 indicating moderate to severe red cell destruction in the spleen (Jandl et al., 1956). All patients had routine liver function tests performed and all had liver biopsies except Cases 8, 10 and 11, in whom this procedure was contraindicated by a haemorrhagic tendency. Liver and spleen scans were performed in all patients using agTcm with routine anterior, posterior and lateral views. Evidence of oesophageal varices was sought by endoscopy or radiology. None of the patients had clinical lymphadenopathy and diagnostic scalene or other lymph node biopsy was not performed before the laparotomy for splenectomy. Eight of the 13 patients had lymphangiograms, but these were all interpreted as being normal. Diagnostic splenic aspiration was not utilized. At laparotomy for splenectomy several lymph nodes were taken for biopsy purposes from the root of the mesentery and the para-aortic regions.
IN most patients living in temperate climates, the cause of an enlarged spleen is readily determined by clinical investigation. Even after careful evaluation, however, a small group of patients remains in which the cause of the splenomegaly is uncertain. Some of these patients appear to be in robust health whilst others are obviously ill. When non-invasive techniques have failed to reveal the cause of the splenic enlargement, splenectomy has to be considered. Under these circumstances, the primary objective of splenectomy is positive pathological diagnosis. Additionally, there may be other secondary benefits, such as the cure of a hypersplenic state, the reduction of an expanded plasma volume or the removal of a large tumour which has become a painful and cumbersome burden. These advantages have to be set against the appreciable morbidity and mortality of splenectomy and the long term vulnerability to infection of the asplenic state (Ramsay and Bouskill, 1973). Unfortunately, there is the possibility that the excised splenic tissue may have no specific histological characteristics, leaving the clinician uncertain as to whether he is dealing with a benign disorder or a more sinister process, presaging a lymphomatous or leukaematous state. The potentials of diagnostic splenectomy have been discussed in standard texts but there have been few detailed reports in recent years. Herman and his colleagues (1968) described 52 patients who underwent splenectomy for unexplained splenomegaly at the Cleveland Clinic during the preceding 10 years. Of these patients, more than half were found to be suffering from lymphoma or congestive splenomegaly and in only 1 case was the diagnosis still uncertain after splenectomy. Although other authors have reported groups of patients with specific diseases that were diagnosed at splenectomy (Hickling, 1964; DasGupta et al., 1965; Skarin et al., 1971), there have been no recent descriptions of the results of splenectomy for undiagnosed splenomegaly after full conventional investigations. This paper describes 13 such patients.
Findings after splenectomy The results of diagnostic splenectomy are shown in Table I . There was no immediate operative mortality, although one patient (Case 8) died 5 weeks after her operation without leaving hospital. Patients with positive findings at splenectomy A definite histological diagnosis was achieved in 6 of the 13 patients. Two patients (Cases 2 and 3 ) who came to medical attention because of pain in the left upper quadrant were shown to have congenital splenic cysts. The diagnosis had been suggested by patchy uptake on the spleen scans. A youngster of 16 (Case 6) had been found to have an enlarged spleen at routine school medical examination and splenectomy revealed the diagnosis of Gaucher’s disease. No other members of the family were affected and there were no other physical stigmas of the disease. Bone-marrow biopsy and trephine and liver biopsy had been unhelpful. An elderly man (Case lo), previously reported (Toghill et al., 1972), was shown to have haemangiosarcoma. Rupture of the spleen precipitated emergency surgery a few hours before splenectomy had been planned. Case 1
* General Hospital, Nottingham. Correspondence to: P. J. Toghill.
63
Splenectomy for undiagnosed splenomegaly Table I: HAEMATOLOGICAL FINDINGS BEFORE SPLENECTOMY AND Known duration Age at of Weight splensplenomegaly of ectomy WBC Platelets before spleen Case Sex (yr) Hb (g/dl) ( x 108/1) ( x 108/1) splenectomy (g) 1980 6.0 430 2 wk (a) 12.6 28 M 1 6.3 420 (b) 14.6 I310 6.5 250 M (a) 14.9 14 2 1 yr 1290 11.0 280 (a) 14.5 27 F 3 2 yr 2690 4 mth 4.4 125 (a) 10.8 66 F 4 10.8 391 (b) 13.6
AT LAST FOLLOW-UP
Histopathology of excised spleen Sarcoidosis Splenic cyst Splenic cyst Non-specific, (?) idiopathic, non-tropical splenomegaly Non-specific
(a) 8.3 (b) 9.4 (a) 12.3 (b) 11.7 (a) 14.0 (b) 1 4 9 (a) 8.9
6.2 10.8 5.7 17.4 7.0 7.1 10.4
120 184 127 190 350 410 30
2 yr
245
14 mth
830
8 wk
520
Gaucher’s disease Non-specific
2 mth
304
Non-specific
45
(a) 5.3 (b) 13 8
1.9 1.4
52 28 1
3 mth
1570
M
72
(a) 5.3
5.5
50
2 wk
1125
11
F
73
(a) 8.9
1.5
78
3 mth
660
Non-specific, (?) idiopathic non-tropical splenomegaly Haemangios ar coma Non-specific
12
M
30
Sarcoidosis
31
250 300 25 76
200
M
7.8 14.6 1.2 4.9
3 mth
13
(a) (b) (a) (b)
2 mth
620
Non-specific
5
M
60
6
F
16
7
M
31
8
F
54
9
M
10
13.8 15.0 9.5 14.6
Outcome after splenectomy Remains well on steroid therapy 3 yr later Discharged Discharged Remains well 2 yr later
Remains in poor health 21 yr later Remains well 2 yr later Remains well 3 yr later Died 5 weeks after splenectomy with thrombocytopenia Remains well 8 yr later
Died 3 mth later Well for 21 months after splenectomy. Died from lymphosarcoma 2 yr after splenectomy Well 18 mth later Remains in excellent health 6 mth later
(a) Findings before splenectomy; (b) findings at last follow-up. was admitted to hospital as an emergency with a pneumothorax and was found to have an enormously enlarged spleen. Diagnostic splenectomy revealed sarcoidosis, which had seemed unlikely on clinical grounds in view of the massive splenomegaly. Case 12, who presented with pruritus, was confidently expected to have a lymphoma, but histological examination of both spleen and abdominal nodes showed sarcoidosis. Patients with non-specific histology There were no specific diagnostic features at splenectomy in the other 7 cases. In one patient (Case 11), the splenomegaly preceded the development of overt lymphosarcoma by almost 2 years. This patient had been investigated extensively preoperatively for a Coombs negative haemolytic anaemia, leucopenia and thrombocytopenia. Surface counting using 51Cr-labelled erythrocytes showed excess splenic uptake. Splenectomy resulted in a gratifying haematological improvement for 21 months until the patient showed clinical evidence of lymphosarcoma which caused her death 3 months later. Two patients (Cases 4 and 9) had massive spleens that were associated with some features of hypersplenism. Neither of these patients had evidence of excessive splenic uptake on surface counting but each had large splenic red cell pools of 13 and 28 per cent of the RCM respectively. In addition, the plasma volume in each case was increased above the normal value of 34-48 ml/kg. Much of the initial haematological improvement after splenectomy in these patients was likely to have been due to the removal of the large SRCP and reduction of the expanded plasma volume. Both patients have now lived for 2 and 8 years respectively in excellent health. In 4 other patients with mild to moderately enlarged spleens weighing from 245 to 800 g splenectomy failed either to achieve a diagnosis or to improve the health of the patients, although
Table 11: RESULTS O F HAEMATOLOGICAL INVESTIGATIONS SRCP RCS Surface counting PV RCM (% of Case (ml/kg) RCM) (Ts051Cr) studies (ml/kg) ~~
3
24.0
0
24 d
4
26.5
13
22 d
7
19-7
18
18 d
8
16.2
9
13.8
28
22 d
1I
23.7
6
lOd
5.5
14.5 d
No excess splenic uptake Noexcesssplenic uptake No excess splenic uptake No excess splenic uptake No excess splenic uptake Excesssplenicuptake (splenic index 250)
38 50 NE NE 60 NE
+
RCM, Red cell mass (normal 25-36 ml/kg); SRCP, splenic red cell pool; RCS, red cell survival (normal 24-28 d); PV, plasma volume (normal 34-48 ml/kg); NE, not estimated. in Case 13 there was a satisfactory rise in the platelet count from a worryingly low level of 30 x 109/1to 87 x 109/l. Case 8 had been investigated extensively preoperatively for mild splenomegaly, a normochromic anaemia and thrombocytopenia. This patient died 5 weeks after splenectomy from continuing thrombocytopenia and subsequent autopsy failed to reveal a specific disease process. Cases 5 and 7 have now been followed up for 2& and 3 years respectively and neither has developed any new or diagnostic features.
64
A. Goonewardene et al.
Table HI: DIAGNOSTIC VALUE OF SPLENECTOMY IN SOME REPRESENTATIVE WORLD SERIES No. in whom diagnosis Diagnostic No. uncertain splenof at ectomies (79 Period caqes Type . . of material splenectomy .. (a)
1956-66 (b) 1963-68
582
All splenectomies
52
9
41
Haematological disorders
0
0
1947-70
200
Haematological disorders
6
3
(d) Not stated
104
Haematological disorders
4
4
406
All splenectomies
12
3
(C)
(C)
1961-72
(f)
1944-70
53
Haematological disorders
0
0
(R) 1967-71
80
Haematological disorders
I0
12.5
(h) 1968-76
116
All splenectomies
13
11
Series and origin: (a)Herman et al. (1968), USA. (b) Christensen et al. (1970), Denmark. (c) Schwartz et al. (1970), USA. (a) Nightingale et al. (1971), UK. ( e ) McKinnon et al. (1973), New Zealand. (f)Ogawa et al. (1974), Japan. (g) lkkala et al. (1974), Finland. (h) Present series (1978), UK.
Discussion Many of the larger series of splenectomies reported in the past 10 years have included a proportion of patients in whom the diagnosis was uncertain at the time of splenectomy. Tabk I11 includes some representative series from various parts of the world, indicating that up to 12.5 per cent of patients who undergo splenectomy are undiagnosed at the time of operation. In the largest group, Herman and his colleagues (1968) considered that 52 of their 582 patients had had splenectomies primarily for diagnostic purposes, With the more liberal use of liver biopsy, lymphangiography and organ scanning by ultrasound or computerized axial tomography, however, it is likely that fewer cases in the future will need to proceed to splenectomy for diagnosis. Many authors have pointed to the value of splenectomy in the initial diagnosis of lymphoma (Hyatt et al., 1970; Schwartz et al., 1970; Skarin et al., 1971; Long and Aisenberg, 1974; Case Records of the Massachusetts General Hospital, 1976). Herman and colleagues (1968) found 31 per cent of their cases of diagnostic splenectomy to have lymphoma. However, the wide spectrum of diseases that may be delineated by diagnostic splenectomy includes sarcoidosis (Herman et al., 1968), splenic cysts and abscesses (Herman et al., 1968; Chulay and Lankerani, 1976), primary tumours of the spleen (DasGupta et al., 1965), splenic angioma (Rhochmis, 1970) and splenoma (Bhagwat et al., 1975). In our 5 patients in whom an accurate histological diagnosis was achieved it was gratifying to be able to assure Cases 2 and 3 that they had benign disorders. previously both had undergone extensive investigation and they had been given a guarded prognosis. It was
thought that Cases 1 and 6 might be suffering from lymphoma, though splenectomy showed sarcoidosis and Gaucher’s disease respectively, diseases with a somewhat better prognosis. Nevertheless, it must be remembered that splenectomy in Gaucher’s disease, though proving the diagnosis, may possibly result in an extension of the osseous and hepatic lesions (Matoth and Fried, 1965) and may also render the child vulnerable to the fulminating infections known to occur in the asplenic state (Ramsay and Bouskill, 1973). Our own series is characterized by the high proportion of cases (7 out of 13) in which the splenic histology was non-specific, though we realize that some of these cases may eventually proceed to frank lymphoma, as did Case 11. Undoubtedly enlargement of the spleen may be a foreboding of subsequent lymphoma, and information derived from diagnostic laparotomy and staging splenectomy in Hodgkin’s disease indicates that considerably enlarged spleens may be free from disease (Glatstein et al., 1970; Ell et al., 1975). We believe that two of the patients (Cases 4 and 9) are probably examples of the syndrome previously described as idiopathic, non-tropical splenomegaly (Dacie et al., 1969) or primary splenic hyperplasia (Weinstein, 1964). In these syndromes, the spleen may be grossly enlarged without there being consistent or significant histological changes. Splenectomy produces a clinical and haematological improvement without serious disease necessarily developing. Both Cases 4 and 9 derived relief from the removal of large uncomfortable spleens, and from the effects of the ‘big spleen’ syndrome, including pooling of erythrocytes and platelets in the spleen and expansion of the plasma volume. Splenectomy failed to establish a diagnosis in the 4 other patients (Cases 5, 7, 8 and 13), and as these patients had only mildly enlarged spleens, there were no other benefits. Certainly their spleens were significantly enlarged (DeLand, 1970; Myers and Segal, 1974) and, in association with obvious haematological abnormalities, there was some basis for diagnostic splenectomy. Our conclusions are that in some cases splenectomy for unexplained splenomegaly may be a useful diagnostic procedure. When splenomegaly is only mild and there is no evidence of disease elsewhere, the indications for diagnostic splenectomy are arguable. Routine examinations of college freshmen have shown 1 per cent with palpable spleens for which no sinister cause was found after a 6-year follow-up (McIntyre and Ebaugh, 1967). There have also been suggestions that splenic infarction might be responsible for some cases of mild to moderate splenomegaly in otherwise healthy folk, and although results of angiography and scanning may be abnormal, the gross changes may be non-specific (Russell et al., 1976). With the larger spleens the diagnostic yield after splenectomy is likely to be greater, but it must be admitted that even with modern techniques some of the large spleens will be undiagnosed after splenectomy. Nevertheless, in patients with the ‘big spleen’ syndrome, other benefits may accrue simply from the removal of the very large organ.
Acknowledgements We thank Trent Regional Health Authority and the Dunhill Research Trust for financial support.
Splenectomy for undiagnosed splenomegaly References
et al. (1973) Splenectomy : indications, results and complications in 406 patients. Ann. Surg. 39, 72-74. MATHOTH Y. and FRIED K. (1965) Chronic Gaucher’s disease: clinical observations on 34 patients. Israel J . Med. Sci. 1, 521-530. MYERS J. and SEGAL R. J . (1974) Weight of the spleen. I. Range of normal in a non-hospital population. Arch. Pathol. 98, 33-35. NIGHTINGALE D., PRANKERD T. A. J., RlCHARDS J. D . M. et al. (1971) Splenectomy in anaemia. Q . J. Med. 41, 261-267. OGAWA Y., KOBAYSHI M., SAKU M. et al. (1974) Jap. J. Surg. 4, 21-28. PARKER A. c., MACPHERSON A. I. s. and RICHMOND J. (1977) Value of radiochromium investigation in auto-immune haemolytic anaemia. Br. Med. J. 2, 208-209. RAMSAY L. E. and BOUSKILL K. c. (1973) Fatal pneumococcal meningitis in adults following splenectomy: two case reports and a review of the literature. J . R . Nav. Med. Serv. 59, 102-1 14. RHOCHMIS P. G. (1970) Splenic angioma with thrombocytopenia. N Y State J. Med. 70, 2133-2136. RUSSELL c. D., JONES A. E., JOHNSTON G. s. and HERDT J. R . L. (1976) Arteriographically confirmed focal defect in colloid spleen scan with no gross pathologic lesions; case report. J. Nucl. Med. 15, 376-377. SCHWARTZ s. I., BERNARD R. P., ADAMS J. T. and BAUMAN A. w. (1970) Splenectomy for haematologic disorders. Arch. Surg. 101, 338-347. SKARIN A. T., DAVEY F. R. and MOLONEY w. c. (1971) Lymphosarcoma of the spleen. Arch. Intern. Med. 127, 259-265. TOGHILL P. J. and GREEN s. (1975) Haematological changes in active chronic hepatitis with reference to the role of the spleen. J . Clin. Pathol. 28, 8-1 1. TOGHILL P. J., RIGBY c. c. and HALL G. F. M. (1972) Haemangiosarcoma of the spleen. Br. J. Surg. 59, 406-408. WEINSTEIN v. F. (1964) Haemodilution anaemia associated with simple splenic hyperplasia. Lancet 2, 218-223. MCKINNON W . M. P., SANDERS H. S., ZAMORA L. F.
et al. (1975) SplenOma with portal hypertension. Br. Med. J. 3, 520, CASE RECORDS OF THE MASSACHUSETTS GENERAL HOSPITAL (1976) N . Engl. J. Med. 295, 828-834. CHRISTENSEN B. E., HANSEN L. K., KRISTENSEN J. K. et al. (1970) Splenectomy in haematology. Scand. J. Haematol. 7, 247260. CHULAY J. D. and LANKERANI M. R. (1976) Splenic abscess. Am. J. Med. 61, 503-511. DACIE J. v., BRAIN M. c., HARRISON c. v. et al. (1969) Nontropical idiopathic splenomegaly (primary hypersplenism) : a review of ten cases and their relationship to malignant lymphoma. Br. J. Haematol. 17, 317-333, DASGIJPTA T., COOMBES B. and BRASFIELD R. D. (1965) Primary malignant neoplasms of the spleen. Surg. Gynecol. Obstet. 120,947-960. DELAND F. H. (1970) Normal spleen size. Radiology 97,589-592. ELL P. J., BRITTON K. E., FARRER-BROWN G. el a]. (1975) An assessment of the value of spleen scanning in the staging of Hodgkin’s disease. Br. 1. Radiol. 48, 590-593. GLATSTEIN E., TRUEBLOOD H. W . , ENRIGHT L. P. et al. (1970) Surgical staging of abdominal involvement in unselected patients with Hodgkin’s disease. Radiology 97, 425432. HERMAN R. E., DEHAVEN K. E. and HAWK w. A. (1968) Splenectomy for the diagnosis of splenomegaly. Ann. Surg. 168, 896-900. HICKLING R . A. (1964) Giant follicle lymphoma of the spleen: a condition closely related to lymphatic leukaemia but apparently curable by splenectomy. Br. Med. J . 2,787-790. HYATT D. F., SKARlN A. T., MOLONEY W . C. et al. (1970) S p h ectomy for lymphosarcoma. Surg. Gynecol. Obstet. 131, 928-932. IKKALA E., KIVILAAKSOE. and HASTBACKA J. (1974) Splenectomy in blood diseases. Ann. Clin. Res. 6, 290-299. JANDL 1. H., GREENBERG M. S., YONEMOTO R. H. et al. (1956) The clinical determination of sites of red cell sequestration in haemolytic anaemias. J . Clin. Inoest. 35, 842-867. LONG J. c. and AISENBERG A. c. (1974) Malignant lymphoma diagnosed by splenectomy and idiopathic splenomegaly. Cancer 33, 1054-1061. MCINTYRE 0. R. and EBAUGH F. R. (1967) Palpable spleens in college freshmen. Ann. Intern. Med. 66, 301-306. BHAGWAT A. G., DATTA D . V., MITRA S.
65
Paper accepted 26 June 1978.
Erratum The publishers regret that in the November 1978 issue of the British Journal of Surgery, in the article by George A. Khoury et al., ‘Tuberculosis of the peritoneal cavity’, on pp. 808 and 809 Figs. 1 and 3 were transposed. It has also been brought to OUT attention that the article by L. E. Hughes and J. F. Forbes ‘Early breast cancer’, Fig. 2 on p. 760 is incorrectly labelled: c should read a and vice versa. Also, on p. 768 the reference to Table III should read Table II.
