Tolerance and feasibility of perioperative treatment with interferon-alpha 2a in advanced cancers.

BACKGROUND: Major surgery impairs the cellular immune response. In order to stimulate the immunological system during the perioperative period, we have studied the clinical and biological tolerance, and the immunological and histological effects of a perioperative treatment using progressive doses of Interferon-alpha 2a, from the third preoperative day (D-3) until the tenth postoperative day (D10). MATERIALS AND METHODS: Twenty-three patients undergoing a major surgical procedure for advanced cancer were included. The clinical and biological parameters evaluated were the body temperature and the blood cell counts. Immunological effects were evaluated by counting the total number of lymphocytes, lymphocyte subsets, natural killer cells (NK), and by analysis of the NK activity, and lymphokine-activated killer cell (LAK) assay. RESULTS: Hyperthermia was the most toxic effect of Interferon-alpha but the overall toxicity was minor, even at the highest dose level. In the early postoperative period there was a significant decrease in total lymphocytes, and in most lymphocyte subset counts when compared with D-3. Overall NK and LAK activities significantly increased from D-3 to D-1 (p < 0.02). A postoperative decrease in NK activity was noted that was not significant when compared to pretherapeutic values, whereas a significant decrease in LAK activity did occur on D4 despite the interferon treatment (p < 0.03). Since we found a dose-dependent effect on some lymphocyte subsets, there was not a clear dose-dependent effect on NK and LAK activities. CONCLUSIONS: Perioperative alpha 2a administration is a safe treatment in advanced cancer patients that may allow a postoperative preservation of NK activity and a destruction of potential circulating metastatic cells. Further studies are ongoing on perioperative immunotherapy in advanced cancer patients.

[The Miami pouch: a reliable continent urinary diversion after pelvic exenteration]. / La poche de Miami: une dérivation urinaire continente fiable après exentération pelvienne.

OBJECTIVES: To describe the technique and present the results obtained with the Miami reservoir, a continent urinary diversion, after pelvic exenteration for advanced gynaecological tumours. METHODS: A Miami reservoir was performed in 12 patients between January 1993 and January 1995. A detubed right colonic reservoir was created using automatic resorbable staples. The ureters were reimplanted into the reservoir using an anti-reflux system and continence was ensured by forming a tube with the terminal loop of ileum and by using a Bauhin valve, which can be reinforced when it is incompetent. Regular postoperative follow-up was conducted (6 to 26 months) with monitoring of laboratory parameters, intravenous urography, opacification of the reservoir, urodynamic assessment of the continent diversion. RESULTS: There were no surgical complications related to the urinary diversion. Urinary continence was obtained in every case and after medical treatment of residual peristaltic contractions of the detubed colonic reservoir in 2 patients. Protection of the upper urinary tract was satisfactory after 2 years of follow-up, without stenosis or reflux of the uretero-colonic anastomoses. The mean capacity of the colonic reservoir was 465.5 +/- 101 ml at 6 months, with filling pressures lower than 20 cm H2O. CONCLUSION: The Miami reservoir is a continent urinary diversion which is relatively easy to perform and reliable in terms of continence and protection of upper tract. However, a longer postoperative follow-up is required. The quality of life of young patients after pelvic exenteration is improved due to this type of contingent diversion which avoids the need for an abdominal urine collector, although it requires intermittent self-catheterization.

Construction of a urinary sphincter by means of an electrically stimulated striated muscle: experimental procedure and urodynamic results on canine model.

A canine model of urinary neosphincter using electrically stimulated autologous striated muscle is described. The superior belly of canine sartorius was activated by a pacemaker with an intermittent low frequency stimulation (0.5-1 pulse per sec) during 7 weeks. Then, the muscle graft was passed around the urethra and sutured back on to itself to form a neosphincter. The surgical procedure was easy to perform and with no complication. Urethral pressure profile was performed initially (T0), and when the muscle was in peri-urethral position, before (T1) and during electrical stimulation (T2). The continence parameter readings (maximal urethral closure pressure MUP, functional length FL, continence zone CZ, and continence area CA) increased from T0 to T1, and from T1 to T2. We noted: 1) 28%, 38%, 52%, and 86% increases for the MUP, FL, CZ, and CA respectively from T0 to T1, 2) 10%, 41%, 30%, and 43% increases for MUP, FL, CZ, and CA respectively from T1 to T2. Chronic low frequency stimulation could transform a skeletal fast-twitch type 2 muscle into a slow-twitch fatigue-resistant type 1 muscle. In this study, morphological changes of the stimulated muscle were noted, whereas phenotype was unchanged. This dynamic autologous neosphincter may be a new alternative to the artificial urinary sphincter prosthesis with fewer complications. Further studies are ongoing to evaluate the efficacy of such a neosphincter as continent system for bladder substitution after pelvic exenteration for pelvic cancers.

[Feasibility and tolerance evaluation of 2 routes of preoperative administration of interleukin 2]. / Faisabilité et évaluation de la tolerance de deux schémas d'administration préopératoire d'interleukine 2.

Preoperative interleukin 2 (IL2) administration has been performed, in order to diminish the post-operative immunodepression in cancer patients. The aim of this study was to compare two different ways of preoperative IL2 administration, ie, intravenous (iv) and subcutaneous (sc), in terms of feasibility and tolerance. Nineteen surgical procedures were performed in 18 patients: a) 10 following the administration of 12 IU/m2/24 hours IL2 IV, with a continuous infusion, from day 5 to day 3 before surgery; b) 9 following the administration of 18 IU IL2, in 2 SC injections per day, from day 4 to day 2 before surgery. Tolerance was evaluated by both clinical and biological parameters, before, during, and after surgery. Hyperthermia and capillary leak syndrome were more important in the iv versus sc injection group. Insomnia and digestive troubles were more frequent in the iv injection group as well. However, we noticed few and equivalent cutaneous and respiratory complications in both groups. In conclusion, the tolerance of IL2 was better after sc versus iv injection. However, the toxicity of iv infusion of IL2 was moderate and could be limited by preventive treatments; moreover there was no consequence on the scheduled surgical procedure.

Incidence and management of major urinary complications after pelvic exenteration for gynecological malignancies.

Urinary fistulae and obstruction following pelvic exenteration are frequent and life-threatening complications. They increase the mortality and morbidity rates of large exereses performed during pelvic exenteration for gynecological cancers. From a series of 97 patients who underwent pelvic exenteration for gynecological cancers we report the incidence, risk factors, and management of major urinary complications. Eighty patients had had previous surgery and/or pelvic radiation therapy at the time of pelvic exenteration. A urinary diversion was performed in 63 patients. Major early urinary complications were: urinary fistula in seven patients and ureteral obstruction in four patients (11.3% of the patients). Ten patients had a late urinary complication: stenosis of the cutaneous ureteral meatus (five), stenosis of the ureteroileal anastomosis following ileal loop (two), and urinary fistulae (three). Cancer recurrence was found in 4 of these 10 cases. Major early urinary complications were significantly increased in patients who had received previous pelvic radiation therapy (P < 0.05) and in patients who had had an intestinal conduit for urinary diversion (P < 0.05). Reoperation was done in six of seven cases of early urinary fistula (urinary undiversion four, nephrectomy one, ureteral reimplantation one). Three of four ureteral obstructions were managed with percutaneous nephrostomy and ureteral stent. We recommend the use of nonirradiated bowel segment for urinary diversion as transverse colon or jejunal conduit in patients who have received previous high doses of pelvic radiotherapy. For the management of urinary complications post pelvic exenteration, reoperation is required for most urinary fistula but ureteral obstructions can be managed with percutaneous nephrostomy and ureteral stent.

Pelvic staging of advanced and recurrent gynecologic cancers: contribution of endosonography.

From January 1988 to December 1992, 106 patients with advanced gynecologic cancer were preoperatively explored by clinical examination (CE) and endosonography (ESG) under anesthesia and by computed tomography (CT). Sixty-one tumors were primary and 45 recurrent; the main locations were cervical (73 cases), ovarian (10 cases), and endometrial (8 cases). All the patients were operated. This prospective study compares the data from clinical and imaging examinations with the histologic findings and the surgical reports. Accuracy of the CE, ESG, cytoscopy, and CT was respectively 79, 90, 82, and 80% for vesical involvement (ESG versus CT: P < 0.05). For vesicovaginal septum extension, accuracy of ESG (92%) was statistically better than that of CE (80%) and CT (77%). Accuracy of the CE, ESG, and CT was respectively 93, 97, and 89% for rectal involvement (ESG versus CT: P < 0.02). For rectovaginal septum extension, accuracy of ESG (96%) was statistically better than that of CE (85%) and CT (85%). Endosonography is valuable in the assessment of regional staging of advanced gynecologic cancers. Since it is realized during the clinical examination under anesthesia, this low-cost procedure is easily performed and provided no discomfort to the patients.

Results of a prospective study with comparison of clinical, endosonographic, computed tomography, magnetic resonance imaging and pathologic staging of advanced gynecologic carcinoma and recurrence.

Between January 1988 and April 1991, 57 patients with advanced gynecologic carcinoma were preoperatively evaluated by gynecologic examination and endosonography (ESG) using general anesthesia. Abdominopelvic computed tomography (CT) was performed in 49 patients and magnetic resonance imaging (MRI) in 21 patients. There were 34 primary tumors and 23 instances of recurrence. Causes of gynecologic carcinoma were 38 carcinomas of the cervix uteri (26 primary and 12 recurrences), eight carcinomas of the ovary (four primary and four recurrences), three recurrences of carcinoma of the endometrium, five sarcomas of the uterus (one primary and four recurrences) and three primary carcinomas of the vagina. All of the patients were operated upon. This perspective study compares the data from clinical and imaging examinations to the data obtained from histologic examination of surgical sections. According to anterior or posterior tumor extension, the accuracy of clinical evaluation and preoperative imaging were studied for the posterior vesical wall and the vesicovaginal septum and the anterior rectal wall and the rectovaginal septum. Histologic examination revealed vesical involvement in 17 patients and of the involvement of vesicovaginal septum in 21 patients. The accuracy of the clinical examination, ESG, cystoscopy, CT and MRI was 83, 88, 87, 75 and 81 percent, respectively, for vesical extension. Cystoscopy was not taken into account for evaluation of extension to the vesicovaginal septum--accuracy was 80, 90, 67 and 86 percent for clinical examination, ESG, CT and MRI. Histologic examination showed involvement in the rectum in 14 patients and involvement in the rectovaginal septum in 19 patients. Rectoscopy was performed 13 times. The accuracy of clinical examination, ESG, CT and MRI was 91, 98, 89 and 71 percent, respectively, for extension to the anterior rectal wall. Rectoscopy was not taken into account for evaluation of extension to the rectovaginal septum--accuracy was 80, 96, 75 and 57 percent for clinical examination, ESG, CT and MRI. Endosonography would seem to be useful to complete examinations for regional extension of advanced gynecologic carcinomas. Its accuracy is superior to that of other examinations. Because it is performed using general anesthesia, there is no discomfort for the patient during this low cost procedure.

Staging of rectal cancer: a prospective study of digital examination and endosonography before and after preoperative radiotherapy.

Between January 1988 and January 1992, 34 patients with rectal cancer were evaluated both by clinical examination and endosonography before and after pre-operative radiotherapy. Two criteria were correlated with histologic findings: confinement to the rectal wall or spread beyond, the presence of mesorectal lymph node involvement. The 32 patients who underwent endosonography before radiotherapy were staged as: uT2: 4, uT3: 26, uT4: 2 cases. Fifteen days after irradiation, endosonography showed tumour regression in all cases; uT stage was different in 15 patients, uN stage in 4 cases. Comparison of the pre-operative local invasion beyond the rectal wall with postoperative histopathy revealed a correlation with: digital examination after radiotherapy in 20 of the 31 patients with palpable tumours; endosonography before irradiation in 18 of the 32 staged tumours; endosonography after irradiation in 25 of the 32 staged tumours. The presence of mesorectal lymph node involvement determined by histologic examination was correlated with the results of endosonography after radiotherapy for 22 of the 32 staged tumours. Endosonography provides a good assessment of the tumour stage before irradiation. Since radiotherapy alters endosonographic staging of rectal cancer, this staging should be included in survival studies.

The value of histological study by surface analysis of preserved and injected pancreatic grafts: a prospective controlled study in dogs.

To evaluate the effects of pancreas preservation and pancreatic duct obliteration on the endocrine pancreas, three groups of dogs were used: a control group (six) in which histologic analysis of normal pancreas was performed and two randomized groups (seven) from which the caudal pancreas was auto-transplanted, injected with fibrin glue and removed on the 28th day. In Group A, each graft was flushed out with Euro-Collins' solution and immediately transplanted. In Group B, each graft was preserved 24 hours in a preservation solution and transplanted. Islet surface ratios on the sections and mean islet surfaces were greater in the control group than in Group A (p = 0.011 and 0.023) and no different between control group and Group B (p = 0.334 and 0.099). This surface analysis study suggests that the mode of management of grafts in itself explains the alteration of endocrine pancreas and that obliteration of the pancreatic ducts has little influence on this alteration.

[Pancreas transplantation in dogs, effect on the endocrine pancreas of pancreatic duct obliteration and the mode of pancreas preservation. Histological study by surface analysis in dogs]. / Transplantation pancréatique chez le chien, retentissement sur le pancréas endocrine de l'oblitération des canaux pancréatiques et du mode de conservation du pancréas. Etude histologique par analyse de surface chez le chien.

The effects of organ preservation and of canal obliteration on the endocrine pancreas were investigated in three groups of dogs. A control group of 6 animals were used for the histological examination of the normal pancreas. Two randomized groups (A and B) of 7 dogs the caudal pancreas was autotransplanted; a fibrin glue was injected into the pancreas canals 28 days before organ removal. In the group A, each graft was rinsed in a Euro-Collins solution and immediately transplanted. In group B, each graft was rinsed and held in a preservation solution for 24 hours before transplantation. On histological slices the ratio between of the total surface area of Langerhans islets to the mean surface area of each islet was greater in the control group than in group A (p = 0.011 and 0.023); the ratio was not significantly different between the control group and the group B (p = 0.334 and 0.099). The histological surface area study suggested that the mode of organ preservation affects the endocrine pancreas and that blocking the canals has little effect.

[Non-invasive dynamic studies of the cervico-thoraco-brachial pathway. Anatomo-clinical correlative study]. / Explorations dynamiques non invasives de la traversée cervico-thoraco-brachiale. Etude corrélative anatomo-clinique.

Based on a computer assisted analysis of the anatomic and biomechanic features of 200 subjects, the authors propose a dynamic approach to the exploration of the cervico-thoraco-brachial pathway. Results are given for 70 healthy and diseased subjects. Ultrasonic evaluations of the arterial component, dynamic echography of the axillo-subclavian venous component, electrophysiology including conventional electromyography, somesthesic and motor evoked potentials, and advanced imagery provide a panel of complementary diagnostic techniques.

[Gangrene of the perineum. Plea for a standardized therapeutic management apropos of 50 cases]. / Gangrènes du périnée. Plaidoyer pour une conduite thérapeutique standardisée à propos de 50 observations.

From 1988 to 1992, 50 cases of perineal gangrene were treated with a therapeutic protocol combining: a) repeated extensive excisions, b) hyperbaric oxygen therapy, even before surgery if this was possible, and c) intensive care. The mortality rate was 24% (12/50). It was even higher in patients admitted more than 6 hours after diagnosis. The average stay in hospital was of 20 (+/- 2) days. Four patients presented with residual signs. Twenty-eight (56%) had had colostomy for lesions originating in the rectum or threatening the anal margin; 9 of these patients died, while gastrointestinal continuity was restored in another 17 cases. There were three predictive factors of survival in this series: a) early diagnosis and treatment, b) severity index on admission, c) some associations of bacteria.

[Obstruction of pancreatic ducts with a fibrin glue. Study after segmentary autotransplantation in dogs]. / Oblitération des canaux pancréatiques par une colle de fibrine. Etude après autotransplantation segmentaire chez le chien.

The effects of the main pancreatic duct with a fibrin sealant have been investigated on an experimental model of segmental pancreatic transplantation in the dog. Fourteen segmental pancreatic transplantations were performed. A cephalic pancreactectomy was performed during the same operating time. The main duct was obstructed with a fibrin sealant (Tissucol modified by addition of a solution of aprotinine concentrated at 10,000 KUI per mL). Biological follow-up consisted in: 1) Intravenous Glucose Tolerance Testing at Day 0 and Day 28 with glycaemia's integral calculus and K V Alues. 2) Measurements of glycaemia and serum amylase every three days from day 0 to day 28. Histological examination of the pancreatic tissue before and after transplantation involved a microscopy analysis reporting the degree of fibrosis and necrosis. The areas of the Langherans islets and of the fibrosis were calculated with informatic area analysis. The study was carried on non diabetic dogs at Day 28. The glycaemia's calculus of IVGTT were not significantly different before and after transplantation (p = 0.291). On the other hand, there was a significant difference of the K Values before and after transplantation (p = 0.006). Histology after transplantation revealed important lesions of fibrosis and normal or hypertropic Langherans islets in most cases. Pancreatic ducts presented with linings thickened with fibrosis. There was no fibrin sealant in the lumen. Obstruction of pancreatic ducts with a fibrin sealant induces an important fibrosis of the pancreatic exocrin tissue allowing the preservation of a satisfactory endocrine function. This technic may be used in clinical practice during the segmental pancreatic transplantations or after cephalic pancreatico-duodenectomy.

[Locoregional extension of advanced cancer of the uterine cervix:prospective evaluation of results of clinical examination, endocavitary ultrasonography, TDM, MRI and histology]. / Extension locorégionale des cancers avancés du col utérin: évaluation prospective des résultats de l'examen clinique, des échographies endocavitaires, de la TDM, de l'IRM et de l'histologie.

Between January 1988 and January 1991, 37 advanced cervical cancer patients were operated and explorated by: clinical examination and endoluminal ultrasonography under general anesthesia (endovesical, endovaginal and endorectal US); tomodensitometry (26 cases); MR imaging (14 cases) and cytoscopy (33 cases). Results of these explorations have been compared with post-operative histopathology. Sensitivity, specificity, predictive values and accuracy of each exploration were calculated for several anatomical structures. Vesical posterior wall has been analysed in about 37 cases, vesico-vaginal structure in about 36 cases, rectal anterior wall and recto-vaginal structure in about 36 cases, parametrium and pelvic sides walls in about 35 cases. Accuracy for clinical examination, endoluminal US, TDM and IRM were respectively: 76, 85, 73, 86%, and respectively 85% by cytoscopy for vesical posterior wall; 72, 88, 65, 78% for vesico-vaginal structure; 92, 97, 83, 64% for rectal anterior wall; 78, 94, 74, 57% for recto-vaginal structure; 84, 89, 69, 73% for vagina; 81, 78, 81, 81% for parametrium; 94%, 93% and 87.5% by TDM and IRM for pelvic side wall. Contribution of endoluminal US is effective for advanced cervical cancer loco-regional staging. Endoluminal US have a good accuracy, are realised under general anesthesia without any discomfort for patients and are little expensive.

[Pancreatic conservation after autotransplantation in dogs. Value of histologic surface analysis]. / Conservation pancréatique aprés auto-transplantation chez le chien. Intérêt d'une étude histologique par analyse de surface.

A study of an artificial conservation fluid (hyperosmolar, pH = 8, rich in lactobionate and raffinose) was carried out by means of an experimental procedure involving segmental pancreatic autotransplants in dogs. The study covers 14 transplants, seven carried out without conservation and seven with 24-hour conservation at 4 degrees C. The caudal pancreas was removed after splenectomy and either transfused with 250 ml of 4 degrees C Euro-Collins before immediate transplant or with 250 ml of 4 degrees C conservation fluid for 24 hours before the transplant. The caudal pancreas was transplanted onto the right iliac vessels, while an arterio-venous fistula was created on the distal splenic vessels and the pancreatic duct was injected with modified tissucol. At the same time as the transplant, a cephalic pancreatectomy was performed. Laboratory tests included an intra venous glucose tolerance test monitored on days 0 and 28 and blood glucose and serum amylase measured every three days from days 1 to 28. The histological study of the pancreatic tissues 28 days after the transplants involved the light microscopic evaluation of the degree of fibrosis, inflammation of the pancreas, cystosteatonecrosis and peripancreatic inflammation. We used a computerized method to measure the surface area of the islets of Langerhans, as revealed by immunocytochemistry, and the surface area of fibrosis. The blood glucose and the serum amylase analyses from days 1 to 28, and the blood glucose variations during the intra venous glucose tolerance test, showed no differences between the two groups. Standard laboratory parameters were similar in the two groups.(ABSTRACT TRUNCATED AT 250 WORDS)

[Cancer of the rectum. Results of a prospective study comparing preoperative rectal touch and transrectal ultrasonography with postoperative histopathology before and after preoperative radiotherapy]. / Cancer du rectum. Résultats d'une étude prospective comparant les données du toucher rectal et de l'échographie endo-rectale, aux données de l'histologie, avant et après radiothérapie pré-opératoire.

Between January 1988 and December 1990, 23 patients with rectal cancer were evaluated by clinical staging (23) and transrectal sonography before (22, one complete stenosis) and after radiotherapy (21). Two criteria were correlated with histological findings: a) the confinement to the rectal wall or spread beyond and b) the presence of meso-rectal lymph node involvement. The 22 patients who underwent transrectal sonography before radiotherapy were staged as uT2: 3 cases, uT3: 17 cases and uT4:2 cases. Fifteen days after irradiation, transrectal sonography showed a regression on the tumor in all cases; the uT stage was different in 9 patients but there was no change in the uN stage. The comparison of the preoperative clinical and sonographic assessment of local invasion beyond the rectal wall with postoperative histopathology revealed a correlation: a) with digital examination in 13 of the 21 patients with palpable tumors, b) with transrectal sonography before irradiation in 12 of the 22 staged tumors, c) with transrectal sonography after irradiation in 17 of the 21 staged tumors. The presence of meso-rectal lymph node involvement determined by histologic examination was correlated with the results of transrectal sonography for 14 of the 21 staged tumors. Transrectal sonography provides a good assessment of the tumor stage before irradiation. This stage should be considered for the study of survival.

[Mucinous cystadenocarcinoma of the appendix. A rare tumor of the right iliac fossa]. / Cystadénocarcinome mucineux de l'appendice. Une tumeur rare de la fosse iliaque droite.

An observation of a mucinous cystadenocarcinoma of the appendix, in the tumor form, revealed by a painful syndrome of the right iliac fossa in a 62-year man is reported. It is a rare malignant tumor as less than 0.5% of the appendicectomy parts present a malignant mucosecreting tumor. In our observation, the diagnosis was allowed by pre-operation imaging. An increase of the serous amount in the tumor markers (carcinoembryonic antigen) (CEA) and CA 19-9 was found before the intervention and the immunodetections performed on the operation part were positive for CEA and CA 19-9. The serous amounts of these markers were normalized after operation. To the author's knowledge, the interest of dosing the serous tumor markers in the observation of such a type of tumor is not mentioned in the literature. The recurrences are frequent and sometimes late even when the initial excision has been macroscopically satisfactory. A new increase of the serous amount of the markers could allow for an earlier detection of a recurrence during the patient follow-up. At present, the prognosis of these malignant forms remains very poor as the 5-year survival does not exceed 25%.

Splenectomy for hypersplenism in chronic lymphocytic leukaemia and malignant non-Hodgkin's lymphoma.

Between 1 January 1980 and 31 July 1988, 62 patients with chronic lymphocytic leukaemia (CLL) or malignant non-Hodgkin's lymphoma (NHL) were splenectomized for splenomegaly and presumed hypersplenism. All patients except one had splenomegaly (mean (s.d.) weight 1585(872) g, range 150-4300 g) and 34 had massive splenomegaly (greater than 1500 g). Forty-nine patients had platelet counts less than 100 x 10(9)/l and 16 patients had anaemia with haemoglobin levels less than 10 g/dl. White cell counts were less than 3 x 10(9)/l in six NHL patients. Fifteen patients had bicytopenia, and three NHL patients had tricytopenia. The selected group of 62 patients underwent splenectomy largely because of failure to respond to medical therapy (39 patients) or inability to tolerate or start adequate chemotherapy because of very low blood counts (11 patients). There was one postoperative death, and a 29 per cent morbidity rate. The response rate was 89 per cent in the first month after splenectomy and 39 patients (63 per cent) had a continuing complete response with a median follow-up of 26 months (range 3-96 months). Twelve patients (10 with CLL) received no further therapy after splenectomy. Seven patients failed to respond and 15 relapsed after splenectomy. These 22 patients could be distinguished on the basis of: (1) lower average preoperative platelet counts (P less than 0.007), postoperative platelet counts (P less than 0.001), and postoperative rise in platelets (P less than 0.004); (2) lower average spleen weight (P less than 0.052); (3) preoperative chemotherapy (P less than 0.044). However preoperative and postoperative platelet counts were the only two variables selected by stepwise regression analysis (P less than 0.05 and P less than 0.01, respectively). Bone marrow failure did not preclude complete response after splenectomy. Long-term survivors emerged from the group of patients with continuing complete response. Of the seven patients who failed to respond, five died with a median survival of 4 months, and of the 15 patients who relapsed after splenectomy, 13 died, with a median survival of 6 months after relapse and 18 months after splenectomy. Thus, splenectomy may be an effective palliation for both CLL and NHL patients with splenomegaly and hypersplenism.