Search

Background: Blastic natural killer (NK) cell lymphoma/leukemia (BNKL) is an immature CD56-positive neoplasm, which was recognized recently and characterized by systemic proliferation of tumor cells including skin, lymph node, and bone marrow., Methods: The current study analyzed 47 patients with BNKL (27 had leukemias and 20 had lymphomas). Patient data were collected for the survey of the NK-Cell Tumor Study Group., Results: There were 33 males and 14 females, with a median age of 53 years (range, 3 months to 89 years). There were few clinicopathologic differences between the leukemia and lymphoma types. Cutaneous involvement was noted at diagnosis in 28 patients, who presented a tendency for older age of onset (median: 56 vs. 46 years, P = 0.11) than patients with noncutaneous BNKL. Cutaneous BNKL showed less frequent mediastinal involvement (4% vs. 53%, P = 0.0002) and less severe thrombocytopenia (P =0 .03). Phenotypic characteristics were also different, with cutaneous BNKL favoring CD4 and HLA-DR expression, and noncutaneous BNKL favoring CD16 and CD34 expression. Both groups responded well to chemotherapy for lymphoid malignancies, but disease recurrence was frequent. The prognosis of patients with noncutaneous BNKL was significantly poorer than that of patients with cutaneous BNKL (median survival: 15 vs. 25 months, P = 0.02). Multivariate analysis confirmed that cutaneous involvement was a significant and independent prognostic factor for BNKL, as were age of onset and leukocyte count., Conclusions: These findings suggested that BNKL is a heterogeneous disease and contains at least two subtypes. Although further investigations are needed to settle a marker for distinction, the presence of cutaneous involvement is a useful prognostic factor.

The identification of important prognostic factors at diagnosis of childhood leukemia has been very useful in explaining the marked differences in disease outcome. After a complete remission is achieved, it is interesting to determine whether the ability to maintain a complete remission and the patient's survival continue to be influenced by the prognostic factors identified at diagnosis. If the maintenance of complete remission and survival continue to be influenced, it is important to determine the magnitude of this effect and its variations with time. Data from a study population of 936 children with newly diagnosed acute lymphoblastic leukemia were analyzed to determine the duration of effect for three variables showing strong prognostic influence: WBC, age at diagnosis, and sex. The strongest of these, WBC, showed a gradual attenuation of effect for children in progressively longer periods of complete continuous remission to a virtual disappearance at 24 months. Age and sex showed a similar reduction in prognostic effect with a negligible contribution at 15 months of continuous complete remission. However, sex becomes an important variables again with a late effect on recurrence results in patients who experience long periods of remission.

Among 164 patients with acute lymphoblastic leukemia (ALL) (age greater than 11 years) induced into complete remission at four hospitals in Italy and The Netherlands between 1971-1977, 49 survived for more than three years in continuous complete remission. Features at diagnosis of the 49 long-term survivors were compared with those of the parent group. The long-term survivors presented with significantly lower leukocyte counts and were slightly younger. Late relapses occurred in nine patients after 37-91 months from remission. Of the 45 patients who had all treatment stopped after 24-60 months of continuous remission, seven have relapsed. Relapses, mainly in the marrow, occurred 4-32 months after cessation of therapy, the risk of relapse being greatest in the first year and dropping to zero by the fourth year. ALL appears curable in approximately one fifth of adolescents and adults entering complete remission with adequate chemotherapy.

The clinical features and diagnostic evaluation of six patients affected by prolymphocytic leukemia (PLL), are described. Some of the cases deviate from a relatively uniform and aggressive clinical course of the disease entity. An actuarial survival analysis of 60 cases gathered from the literature and the authors' experience indicate that the cases showing the most prolongated evolution may reach 30% of the total. The different aggressiveness in the clinical course of these patients does not depend on the efficiency of the therapies applied. The percentage of prolymphocytes (PL) in peripheral blood throughout the clinical course was variable and not depending on the treatments used. This fact, in conjunction with (1) the presence of cellular types deviated from the typical PL and detected both at optical and ultrastructural levels and (2) the existence of cases of PL with minimal splenomegaly and leukocyte counts of less than 50 X 10(9)/liter, could lead to an underdiagnosis of PLL and hinder the actual cognition of the natural medical history of the disease.

The clinical course of a two-year-old girl with acute lymphocytic leukemia (ALL) who inadvertently received intrathecal vincristine is presented. Despite aggressive treatment with parenteral folinic acid and extensive central nervous system washout, the outcome was fatal. The pharmacologic aspects of vincristine and folinic acid that may explain why treatment was unsuccessful, are reviewed.

Combination chemotherapy with an anthracycline, Adriamycin or rubidazone, cytosine arabinoside, vincristine and prednisone resulted in a complete remission rate of 62% in 325 consecutive unselected adults with acute leukemia. The results by morphologic categories were 58% for acute myelogenous leukemia (AML), 70% for acute undifferentiated leukemia (AUL), and 77% for acute lymphoblastic leukemia (ALL). The median survival was 43 weeks. Ten percent of all patients are projected to be alive and in remission at five years. The median remission duration for the whole group was 51 weeks, durations being significantly longer for AML (60 wks) than ALL (30 wks) and AUL (21 wks). Central nervous system involvement was uncommon in AML (4%), but much more common in patients with AUL (37%) and ALL (32%). One in five complete responders with AML is projected to be in their first remission at five years off all chemotherapy. Age, sex, morphology, cytogenetic pattern, temperature of presentation, and presence of a documented preceding hematologic abnormality are found to be significant variables for response and survival.

Three cases of prolymphocytic transformation of chronic lymphocytic leukemia (PLL-trs CLL) are described, and a review of the literature yields 15 additional cases. PLL-trs CLL may be recognized when two populations of cells are present and the number of prolymphocytes are greater than 15% admixed with the small round lymphocytes of CLL. Surface immunoglobulin intensity appears to be greater on the prolymphocytes than on the small lymphocyte in the same case, suggesting a clonal differentiation. At the time of transformation, the spleen is generally larger than at the time of diagnosis of CLL, and lymphadenopathy is variable. Transformation is thought to be associated with increasing refractoriness to treatment and shorter survival, although most of the deaths occurred in clinical Stages III and IV, known to have a poor prognosis. Similar features between PLL-trs CLL and so-called de novo PLL also suggest an evolutionary process. However, additional morphologic and immunologic studies in transforming CLL cases are necessary to answer this question.

Between February 1980 and December 1985, splenectomies were performed in 13 patients with advanced (Stage III and IV) chronic lymphocytic leukemia (CLL). Patients had advanced disease for a median of 32 months at the time of splenectomy and all were refractory to chemotherapy. All patients were anemic with a median hematocrit of 27.5%; nine patients were thrombocytopenic with a median platelet count of 46,000/microliter. There was no surgical mortality and minimal morbidity. Median hematocrit was 35% at 1 month and 6 months postsplenectomy. Median platelet count in previously thrombocytopenic patients was 261,000/microliter at 1 month and 177,000/microliter at 6 months postsplenectomy. Median survival postsplenectomy in this advanced patient group was 24 months with four patients surviving more than 30 months. Splenectomy is a valuable adjunct to the management of end-stage CLL.

Prolymphocytic leukemia is characterized by marked splenomegaly, distinctive cellular morphologic characteristics, and a poor clinical course. Five patients with typical PL were treated systematically with vincristine/prednisone, chlorambucil/prednisone, splenic irradiation, splenectomy, and other chemotherapy regimens. No patient responded to vincristine/prednisone. Two patients responded to chlorambucil/prednisone, and four patients had brief responses to splenic irradiation. Two patients underwent splenectomy, one of whom had a prolonged clinical remission. There were no complete remissions. No other chemotherapy combinations were of value. The median survival was 33 months. Recommendations are made to use chlorambucil/prednisone or splenic irradiation as initial treatment. Splenectomy should be considered in patients refractory to these modalities. The course of PL may be more protracted than originally reported.

Ninety-six patients with advanced chronic lymphocytic leukemia (CLL) (Stage C; anemia and/or thrombocytopenia of nonimmune origin) were randomized to receive either chlorambucil (CLR) (0.4 mg/kg orally, day 6) plus prednisone (PDN) (60 mg/m2 orally, days 1-5) every 2 weeks or cyclophosphamide (600 mg/m2 intravenously, day 6), vincristine (1 mg/m2 intravenously, day 6) and prednisone (60 mg/m2 orally, days 1-5) (COP) each month for 5 months. Complete remission (CR) was defined as the total disappearance of signs and symptoms related to the disease. Partial remission (PR) was considered to be achieved when, after treatment, the clinical stage changed to a less advanced one. Thirty (59%) responses (8% CR) with CLR plus PDN and 14 (31%, 2% CR) with COP were observed (P less than 0.01). The survival was not significantly different for the two groups. Patients previously treated had a lower number of responses (11/35, 31%) than those with no previous treatment (33/61, 54%) (P less than 0.05). Patients who attained a CR or a good PR had longer survivals (median not reached) than those with a poor PR (median, 25.2 months) or those who did not respond to treatment (median, 11.5 months) (P less than 0.005).

A reanalysis of all ten clinical trials conducted by the Southwest Oncology Group from 1958 thorugh 1976 in newly diagnosed pediatric acute leukemia was carried out to quantify the "early death" rate (i.e., rate of deaths during the first four weeks of remission induction) and to determine factors related to these deaths. Out of 1,964 patients registered, there were 126 (6.4%) early deaths. For the 1,375 acute lymphocytic leukemia (ALL) patients only, there were 75 (5.5%) early deaths. During the last eleven years on this study (1965 through 1976), the death rate for ALL patients appears to have remained stable at five deaths per 100 patients, despite the general improvement in overall survival. The most important prognostic factor was the initial platelet count: patients with a platelet count below 25,000 had a death rate over three times higher than that of patients with a platelet count over 75,000. Infants under one year of age also had a very high early death rate (6 of 41, 15%).

A study was conducted in metropolitan Baltimore to examine changes in survival of white and black children with acute leukemia from 1960 through 1975. Two-hundred eighty-seven cases were identified, of which 77% were acute lymphocytic leukemia (ALL). Eighty-three percent of all cases occurred in white children. In white children with ALL, two-year survival rates increased from 32% in 1960--64 to 47% in 1965--69, and to 79% in 1970--75 (p less than .005). In the small sample of black children with ALL, two-year survival rates increased from 25% in 1960--64 to 59% in 1965--69 (p less than .01), with no further increase in 1970--75. For acute nonlymphocytic leukemia (ANLL), survivorship was analyzed in white children, among whom one-year survival rates increased from 42% in 1965--69 to 71% in 1970--75 (p less than .005). The increasingly better survival over time of white children with acute leukemia probably reflects the increasing efficacy of new therapeutic approaches.

A total of 78 patients with acute leukemia [69 acute lymphoblastic leukemia (ALL) and 9 acute myeloblastic leukemia (AML)] survived more than 4 years. In the years between 1958 and 1967 the number of patients with ALL in Argentina who survived 4 years was 1%. This increased to 21% in the years between 1967 and 1972. The major cause of this increase was the initiation in 1967 of protocols of therapy on a national scale through a cooperative group known as GATLA. In AML patients the number of long-term survivals remained unchanged during both periods. A large percent of the long-term survivors had lower WBC, a higher platelet count, and no hepato- or splenomegaly, lymphadenopathy, or hemorrhagic manifestations at diagnosis. However, only for those patients with a WBC under 100,000/mm3 at diagnosis was there a significant prognostic implication (p less than 0.01). Patients with acute lymphoblastic leukemia who survive more than 4 years without relapse have about an 80% chance of long survival. For those who have lived 4 years with one or more lapses, there is only a 17% chance of long survival, and of these only those who develop extramedullar relapse (CNS or testicular) without bone marrow involvement have a chance of long survival. Patients who live 4 years in continuous complete remission have the same chance of very long survival regardless of previous therapy.

There is lack of evidence to date that treatment of chronic lymphocytic leukemia (CLL) alters the natural history of disease or influences survival. In the present series, total body irradiation (TBI) produced a range of therapeutic responses among patients with active, progressive CLL. One-third of patients experienced virtually complete clinical and hematologic remissions with the initial course of TBI. These patients did not differ from the less complete responders with respect to age or sex, degree of lymphocytosis in the peripheral blood, incidence of anemia and/or thrombocytopenia, or the frequency of constitutional symptoms. However, the patients with complete remissions noted a return to normal performance status, had fewer serious infections, demonstrated recovery from depressed immunoglobulin levels, and had significantly longer survival. These data indicate that TBI is capable of inducing remissions which modify the course of disease in patients with CLL and that prognosis has a direct correlation with the response to therapy.

Long-term second remissions were seen in 5/66 patients with all. Relapse was extramedullary in 2/5. Persistent, progressive marrow lymphocytosis preceded relapse in 4/5 patients and persistent marrow eosinophilia in 1/5. All 5 patients had had an unmaintained remission of at least 6 months prior to relapse, and responded the second time to drugs which were essentially the same as those used initially. We conclude that long-term second remissions may occur in all. Patients who relapse after 6 months or more of unmaintained remission should be treated with the drugs used in current initial induction regimens in hope of cure.

A group of 38 patients with a median age of 70 years and chronic lymphocytic leukemia (CLL) were treated using a cobalt 60 U or a 6-MeV linear accelerator. A direct field or two opposite fields covered the palpable spleen area in most patients. 100 cGy were administered weekly for a total dose of 10 Gy, given over 10 weeks. The stage arrangement (according to Rai) for the 32 evaluable patients was as follows: Stage I: 11 patients, Stage II: nine patients, Stage III: three patients, and Stage IV: nine patients. Patients in Stages I and II were treated when symptomatic. Twenty-five patients (78%) achieved hematologic response (HR), defined as normalization of the differential leukocyte count, of the total blood cell count, and of bone marrow infiltration. However, no complete response according to the standard criteria of response has been obtained. The median response time of HR was 7 months (range, 1.5 months to greater than 120 months). The overall median survival time from the start of splenic irradiation (SI) was 40 months. More than 50% splenomegaly reduction was obtained in 63% of patients, whereas no benefit was verified in the lymphadenopathy. The incidence of second tumor was 29%. Fourteen patients benefited from a further 21 SI cycles. SI does not result in a complete remission and therefore cannot modify the course of CLL. This treatment is most advisable for elderly patients with predominant bone marrow lymphocytosis, for patients with previous extensive chemotherapy or radiotherapy, and for patients with poor marrow reserve. Moreover, because of the absence of toxicity subsequent treatment is not compromised.

Forty-one previously untreated children with Stage IV non-Hodgkin's lymphoma were studied from January 1971 to April 1979. All patients had bulky disease histologically proven to be non-Hodgkin's lymphoma with bone marrow involvement. They were separated into two groups according to the extent of bone marrow involvement. Group IVA included 14 patients with 25% or less lymphoblasts in the bone marrow. Group IVB included 27 patients with more than 25% blasts in the marrow. Their clinical characteristics with regard to age, sex, hemogram, histology, primary site, and blast morphology are compared. All were treated with the LSA2-L2 protocol with radiation therapy to one or more bulky sites of involvement. The disease-free actuarial survival for Group IVA was 64% with a median observation time of 49 months while that for Group IVB was 65% with a median observation time of 66 months. There was no statistical difference in the survival rates between the two groups. Hence, we conclude that the extent of bone marrow involvement does not affect the prognosis in Stage IV non-Hodgkin's lymphoma. It appears that radiation therapy may have contributed to the improved survival in our series. Furthermore, a subset of patients in Group IVB (24/27) who could be considered as high-risk acute lymphoblastic leukemia on the basis of age, initial leukocyte count, hemoglobin, mediastinal mass or T- or B-cell markers showed an improved survival (73% versus 43%) when compared to patients treated with conventional leukemia regimens.

Patients with untreated chronic lymphocytic leukemia (CLL) received protocol treatment with 6 months of chlorambucil (CB) (30 mg/M2) and prednisone (P) (80 mg/d X 5) every 2 weeks. Complete and partial responders (CR, PR) were then randomized to consolidation with six more courses of CB and P or to four courses of cytosine arabinoside (25 mg/M2 every 12 hours X 8, subcutaneously) and cyclophosphamide (25 mg/M2 every 12 hours X 8, orally) every three weeks. Of the 178 eligible patients entered, 138 (78%) were evaluable for induction therapy which produced a 22% hematologic CR and an overall response rate (CR + PR) of 74%. Eighty-two patients received adequate consolidation, at the end of which 43 were in CR. No difference was seen in response or survival between the two consolidation treatments. Responders had longer survival than nonresponders (P = 0.0001) even when a 6-month "guarantee time" was excluded, but there was no survival difference between CR and PR. Thus, intermittent CB and P is a well-tolerated, useful therapy for CLL but the addition of cyclophosphamide and cytosine arabinoside does not improve results.

Among 727 children with acute lymphocytic leukemia (ALL) observed at eight pediatric clinics in Italy in the years 1967-1974, 200 (27.5%) survived for more than five years after diagnosis. The proportion of long-term survivors rose significantly during the years 1970-1974 when aggressive therapeutic programs with curative intent were uniformly adopted in Italy (19.8% vs. 29.4%; P less than 0.05). Clinical and laboratory data at diagnosis of the 200 long-term survivors were analyzed and compared with that of the 527 nonsurvivors. We found that, besides a leukocyte count greater than 50,000 cells/mm3, other factors such as early central nervous system CNS leukemia and the presence of mediastinal mass were predictive of a poorer prognosis for long-term survival. Life-table analysis revealed that the chance of long-term survival was significantly higher in those children who have survived for five years without relapse (82.9% vs. 24.1%; P less than 0.01). Although late initial relapse is always possible, if a child with ALL remains in continuous complete remission for at least nine years, it is likely that the patient is cured.

Presence and size of lymph nodes and spleen, graded from 0 to , in 362 patients with CLL observed from diagnosis were evaluated. Statistical analysis showed a relationship with age, sex, anemia and thrombopenia, leukocytosis, and outlined two different groups: the one without organomegalies , with higher mean age (67 years), female prevalence, and better prognosis; the other with adenosplenomegaly graded ++/ , with lower mean age (57 years), clear male prevalence, and worse prognosis. Survival results were statistically different only between groups 0/+ versus group ++/ . Important chronic diseases were present at diagnosis in approximately 25% of the cases, with a severely reduced survival (median, 27 months), close to that of the cases with anemia and/or thrombopenia (22 months). Therefore it seems that in every prognostic grouping system, complicated cases should be taken into account and grouped with the anemic and/or the thrombopenic ones. The following prognostic groups are proposed: I: low risk: cases without or with adenomegaly and/or splenomegaly + (65% surviving at 100 months); II: intermediate risk: cases with adenomegaly and/or splenomegaly ++/ (median survival, 70 months); III: high risk: cases complicated by chronic diseases, or with anemia and/or thrombopenia (median survival, 25 months).

From 1981 to 1983, 131 previously untreated patients with acute lymphoblastic leukemia (ALL) standard-risk group were entered to the protocol JCCLSG-S811. Of 119 eligible patients, 115 (96.6%) attained complete remission by treatment with prednisone (PRD) plus vincristine (VCR) or vindesine (VDS). After preventive central nervous system (CNS) therapy including 18 Gy cranial irradiation and three doses of intrathecal methotrexate (MTX), the patients were assigned randomly to the two maintenance chemotherapies, Regimen A and Regimen B. Regimen A (intermittent regimen) consisted of PRD (120 mg/m2/day by mouth for 5 days) plus 6-mercaptopurine (6MP) (175 mg/m2/day by mouth for 5 days) plus VCR (2.0 mg/m2 intravenously) alternating biweekly with MTX (225 mg/m2 intravenously). Regimen B (continuous regimen) consisted of 6MP (50 mg/m2/day by mouth) plus MTX (20 mg/m2/week by mouth) combined with pulses of PRD and VCR (the same dosages as Regimen A) every 4 weeks. As the late intensification therapy (LIT), five courses of high-dose MTX (2000 mg/m2 per dose per week intravenously for three doses every 12 weeks) with leucovorin rescue were administered to all patients who were in continuous complete remission (CCR) for more than 2 years. Sixty and 55 patients, respectively, were registered in Regimen A and B. The CCR rates in Regimen A and B were 75.1% +/- 5.8% (mean +/- 1 SE) and 49.7% +/- 7.3% (P less than 0.01) at 4 years, and 72.1% +/- 6.3% and 49.7% +/- 7.3% (P less than 0.05) at 5 years, respectively. In Regimen B, CNS and testicular relapses increased after 3 years of CCR. In addition, the patients in Regimen B had a much higher incidence of infections than Regimen A. The LIT did not seem to have important effects on the duration of CCR. From these data we conclude that the intermittent cyclic regimen of 6MP and MTX may be more effective as compared to the continuous administration of these drugs in the maintenance chemotherapy.

Analysis of mortality rates for all childhood (age less than 20 years) leukemias in upstate New York showed declines between 1969-1971 and 1978-1980 for ages 0-4 and 5-9 years, while data from a population-based cancer registry for a population of about 10.5 million indicated no decline in incidence rates. Survival rates were examined for 552 children diagnosed with acute lymphocytic leukemia (ALL) in 1973-1980 at less than 15 years of age. There was some evidence for improvement in survival rates over these years of diagnosis, especially among children diagnosed at ages 1-4 and 10-14 years. Hazard rates increased among cases (especially males) diagnosed in 1979 and 1980, however, and this finding requires confirmation from other studies; the possible effect of influenza epidemics also should be explored. Only one confirmed second primary cancer, a rhabdomyosarcoma of the eye, was ascertained (as of the end of 1983), but longer follow-up is needed on children with ALL.

As part of epidemiologic studies of human T-lymphotropic virus (HTLV)-I-associated malignancies in Jamaica, the authors evaluated 26 patients with non-Hodgkin's lymphoma for the presence of integrated HTLV-I provirus in their malignant cells. Fifteen of 26 patients had integrated provirus. All 15 also were HTLV-I antibody positive. Eleven patients did not have integrated provirus, and all 11 were antibody negative. All of the antibody-positive cases had onset of their disease in adulthood (age range, 21-57 years) as opposed to the broad age range of negative cases (4-66 years). Clinical features which were more common in provirus positive than negative patients included leukemic phase, skin involvement, and hypercalcemia, which are all features frequently seen in HTLV-I-associated adult T-cell leukemia/lymphoma (ATLL). The presence of skin involvement, circulating malignant cells, abnormal liver function tests, or the presence of two or more of these four features were statistically significantly different between virus-positive and virus-negative cases. Although the survival of positive cases (6 months) was shorter than that of negative cases (9 months), this was not statistically significant. The only significant determinant of survival was hypercalcemia, with those who developed hypercalcemia at some point in their disease course, independent of their HTLV-I status, surviving a mean of 5 months as compared to a mean of 17.5 months in those who never became hypercalcemic. The six HTLV-I-positive lymphomas that underwent cell typing were all primarily OKT4 positive, whereas two HTLV-I antibody-negative cases that were typed were B-cell lymphomas.

Forty-eight patients with chronic lymphocytic leukemia (CLL), and 12 patients with small (well differentiated) lymphocytic lymphoma (WDL) were histologically evaluated for their pattern of bone marrow (BM) involvement. Four different types of BM infiltration were recognized: nodular (N), interstitial (I), nodular and interstitial (mixed) and diffuse (D). The pattern of BM involvement was compared with the clinical, laboratory, and survival status in all patients. The extent of the disease in CLL patients, was determined by the Rai and the International Workshop on CLL Staging Systems, while in WDL patients the Ann Arbor staging system was used. In the CLL group the N pattern was found in 8%, the I in 33%, the mixed in 31%, and the D in 27% of the patients. Based on the International Workshop on CLL Staging System, the I pattern of BM involvement was more frequently found in Stage A (56%), the mixed in Stage B (68%), and the D in Stage C disease (90%). All CLL patients with D pattern required treatment from the beginning, contrary to CLL patients with the other patterns, in whom therapy was required in less than 50%. Similarly, deaths were more common in the D pattern in whom therapy pattern than in the other patterns. In the WDL patients BM involvement was found in 4 of 12, (33%) and its pattern of positivity was always nodular, although most patients (10 of 12) had advanced disease. It is concluded that the frequency of BM involvement may contribute in the differential diagnosis of WDL from CLL. In addition, the pattern of BM infiltration correlates very well with the International Staging System for CLL, and the pattern of BM positivity in CLL patients also has prognostic significance.

This Japanese nation-wide survey was carried out on long-term survivors among children with leukemia, i.e. those who had survived more than 5 years by October of 1974. Eighty-three children were surveyed. This was about 2.5 times as many as the 31 children surveyed in the first Japanese nation-wide survey in 1971. Clinical and laboratory data were also analyzed statistically and compared to that of the first survey. Most of the children had received inductive therapy according to the old protocol, but 10 children were treated with a combination of VCR and steroids. Nine of these 10 were still in a state of remission. CNS involvement occurred in 15.3% of the cases and a late onset of this complication implied a prognostic significance: among the 9 children whose complication had occurred 3.5 years after onset of the disease, 6 died. The relapse rate decreased between the 4th and 5th year after initiation of maintenance therapy. The percentage of the risk of death became zero after the 9th year.

Further progress in the treatment of acute leukemia involves a careful assessment of host and disease factors in each child and the selection of a therapeutic regimen which interacts with these factors to provide the maximum chance for survival. For some children, this therapy can be carried out in the community by expertly trained hematologist-oncologists; for others, the best treatment will involve clinical research at a center. Such a general approach has the best chance of fulfilling the optimistic survival projections made by the President's Cancer Panel in 1973 for acute lymphocytic leukemia.

The results of treatment for acute leukemia since 1973 in 87 patients aged 70 years or more are summarized. The overall complete remission rate was 27/78 (35%) in acute myelogenous leukemia and 5/9 (56%) in acute lymphocytic leukemia or acute undifferentiated leukemia for cytosine arabinoside combined with either anthracyclines (rubidazone or Adriamycin [doxorubicin]) or m-AMSA (amsacrine). The remission duration was short with a median of 33 weeks, and the median overall survival was only 6 weeks. Those patients without identifiable infection, liver enlargement, and a serum glutamic oxaloacetic transaminase less than or equal to 40 U/ml constituted a more favorable subgroup. Although the complete remission rate improved further research is needed to develop effective maintenance strategies.

Fourteen patients with chronic lymphocytic leukemia (CCL) who received a total of 23 courses of splenic irradiation (SI)for various combinations of painful splenomegaly, progressive leukocytosis, lymphocytosis, thrombocytopenia, and anemia are reviewed. Splenic midplane doses ranged from 200-1750 rads (3-14 days), while the most typicalregimen was 300-450 rads in two to three fractions of 150 rads given over 3-8 days. Response to SI was rated according to a scoring system which evaluated the splenic and hematologic response, as well as the response of disease-related symptons. According to this scoring system, most patients demonstrated a significant relief of painful splenomegaly, along with improvement of hemogram and bone marrow parameters. While those patients judged as failures to prior chemotherapy or total body irradiation showed improvement following SI, those who had received minimal therapy prior to SI showed a superior response. SI, in those showing a satisfactory response, was repeated successfully in several patients (up to six times in one instance). The onset and duration of response to SI, dose-reponse data, survival, clinical and hemotologic responses, and the possible mechanism of action of SI are discussed with reference to the available literature.

A remission-induction regimen for childhood leukemia using cyclophosphamide, asparaginase, vincristine, and prednisone (CAVP) was compared to standard vincristine-prednisone (VP) induction. The more intensive regimen was associated with a lower complete remission rate (81% vs 93%) and a higher early death rate from infection (15% vs 5%) for acute lymphocytic leukemia. In contrast, complete remission was achieved in 58% of children with acute nonlymphocytic leukemia treated with CAVP compared to 18% for VP. Early death rates were similar (27% vs 25%). These observations corroborate previous studies in childhood nonlymphocytic leukemia showing activity for asparaginase. Preliminary analysis of remission duration and survival for responders shows no advantage for those who survived the more intensive induction.

The prognostic value of lymphocyte doubling time (LDT), expressed in months and obtained by means of a linear regression, has been studied in 99 previously untreated chronic lymphocytic leukemia (CLL) patients. LDT is defined as the period of time needed for lymphocytes to double in number the amount found at diagnosis. When the analysis was extended to the whole population, it showed clear differences in the life expectancy of patients with LDT of less than or equal to 12 months (median survival, 36 months; relative death rate [O/E]1.57) compared with those with LDT of more than 12 months (median survival not yet reached; O/E, 0.37) (P less than 0.001). The significance of LDT remained even after adjustments were made for age, sex, lymphocyte count, anemia, and thrombocytopenia. The lack of statistical significance after adjustment for Binet's clinical stage corresponds to the fact that clinical stages are not distributed homogeneously, high LDT being more frequently associated with earlier stages and low LDT with more advanced forms of the disease (P less than 0.001). In this study LDT was a useful parameter in predicting disease progression. Patients in Stages A and B and with rapidly increasing lymphocytes counts became worse more frequently (33.3% and 29.1%, respectively, at 12 months after diagnosis) than those with a slow increase (no change in clinical stage at 12 months). It is concluded that since LDT appears to predict the progression of the disease, it is useful in the clinical management of CLL.

Review of 745 cases of chronic lymphocytic leukemia show that clinical staging by either the Rai or Binet method is very valuable in assessment of patients at the time of diagnosis. There is a marked decrease in survival with advancing stage (19.9 years median survival for Rai Stage 0, and 2.5 years and 2.7 years for Rai Stages 3 and 4, respectively). The definition of anemia as Hb less than 11 g/dl as in the Rai staging method appears to give slightly better discrimination among the stages than the Binet staging procedure. Patients with splenomegaly alone (Binet stage 2), however, form a small but distinct group that should be recognized. These is little to be gained by subdividing further according to size of the lymph nodes.

Pretreatment characteristics of 48 infants (under 2 years of age) with leukemia treated over a period of 18 years at a single institution were studied in relation to response to therapy, extramedullary involvement, and survival. In order to provide a basis of comparison, 45 of the infants were each paired with an older (2-9 years of age) leukemic child, who had the same race, and disease type, who was treated within the same period of time and received similar suportive care and chemotherapy, and when possible, was of the same sex. The overall median survival times of infants and their pairmates were 211 days and 445 days, respectively. Initial status of the spleen was the single significant factor in relation to the length of survival. Median survival time for infants with splenomegaly was 186 days compared to a median survival time of 414 days for infants without splenomegaly. The complete response rate was 75% for infants and 80% for the pairmates. Median duration of remission was 84 days for infants and 280 days for their pairmates. Initial peripheral leucocyte count was significantly related to the duration of remission; patients with very high leucocyte counts had the shortest remissions. The duration of remission increased throughout the period covered by this study, however, prognosis relative to pairmates remained poor. Central nervous system leukemia occurred in 44% of infants and 40% of the pairmates; incidence of CNS leukemia was much higher (70%) in infants under one year of age than in infants one year old (37%). Extramedullary leukemia occurred at other sites in 46% of the infants and 31% of the pairmates. The only pretreatment characteristic of prognostic importance for predicting extramedullary involvement was the patient's age at diagnosis.

Serum beta-2-microglobulin (S-beta 2M) was measured at diagnosis in 44 patients with lymphocytic leukemias and 47 with malignant lymphomas. Among patients with chronic lymphocytic leukemia (CLL) S-beta 2M was raised (greater than 3 mg/l) in 74% and in 23.5% of those with acute lymphoblastic leukemia (ALL). The frequencies for non-Hodgkin's lymphoma (NHL) and Hodgkin's disease (HD) were 59.2% and 40%, respectively. In CLL patients high serum values correlated with large tumor mass, as estimated by Rai's clinical criteria (P less than 0.001), by total peripheral lymphocytes (r = 0.41, P less than 0.05) and by the percentage of bone marrow infiltration of the lymphocytes (P less than 0.01). A significant relation was also found in CLL patients between S-beta 2M level and survival (P less than 0.05). In ALL no association was found between S-beta 2M level with peripheral lymphoblast concentration, French-American-British (FAB) subclassification, splenomegaly, and survival. In NHL patients a significant association was found between S-beta 2M levels and stage of disease (P less than 0.01) and an obscure relation (P less than 0.1) with the presence of lymph nodes greater than 3 cm in diameter, splenomegaly, and hepatomegaly. No significant association was found between S-beta 2M level and histologic subtypes, presence of B symptoms, bone marrow involvement, and survival. In HD patients a significant association was found between the level of S-beta 2M and stage of disease (P less than 0.05) and presence of splenomegaly (P less than 0.05). No association was found between S-beta 2M level and histologic subtypes, lymph nodes greater than 3 cm in diameter, bone marrow involvement, and B symptoms. A significant relation was found between S-beta 2M level and survival in HD patients with widespread disease (P less than .025).

Acute lymphoblastic leukemia and non-Hodgkin's lymphoma constitute 42% to 45% of the cancers in infants, children, and adolescents: In 1985, an estimated 2025 children were newly diagnosed with these two cancers and 900 (43%) of the pediatric cancer deaths in the United States have been projected to be due to these diseases. The single most important obstacle to preventing these deaths is relapse, and prevention of relapse or salvage of the patient who has had a relapse continues to be a major therapeutic challenge. The most important initial step in the treatment of the child whose disease has relapsed is to determine, to the extent possible, the prognosis. In a child with non-Hodgkin's lymphoma, a relapse confers an extremely poor prognosis, regardless of site of relapse, tumor histology, or other original prognostic factors, prior therapy, or time to relapse. In the child with acute lymphoblastic leukemia in relapse, the prognosis depends on multiple factors. The primary therapy is chemotherapy or chemoradiotherapy with marrow grafting. Other options exist, including no therapy, or investigational therapy. The therapy selected should be predicated on the prognosis. In the child with an isolated central nervous system (CNS) relapse off therapy, minimum therapy should be administered, particularly if the relapse occurred without prior cranial irradiation. In the child whose relapse is more than 6 months off therapy, conventional therapy should be considered. Also, a patient with an isolated CNS relapse on therapy after prior cranial irradiation should be given moderate therapy. Bone marrow transplantation or high-dose chemoradiotherapy with autologous marrow rescue should be reserved in children with a second or subsequent extramedullary relapse, and possibly for those with a first isolated overt testicular relapse on therapy.

Forty chronic lymphocytic leukemia patients with splenomegaly were evaluated for splenectomy. Twenty were splenectomized. All but two normalized their hematocrit and all but one their platelet count. Satisfactory residual erythropoiesis as evaluated using ferrokinetics enabled prediction of a good response to splenectomy in 18 of 20 patients. The two patients who did not respond despite satisfactory residual erythropoiesis had a deteriorating blood count after splenectomy with progressive disease. The measurement of the splenic red cell volume predicted the minimal increase in hematocrit that could be expected after splenectomy. The transfusion need was abolished in 13 of 14 patients who needed red cell transfusions before splenectomy. Despite clinical improvement after the operation in all but two patients, splenectomized patients did not survive longer than nonsplenectomized patients, whether survival probability was calculated from the time of diagnosis or from the time of evaluation.

Twenty-three adult patients (ages greater than 15 years) and 75 children with acute lymphoblastic leukemia were treated with similar intensive, sequential cytotoxic protocols (L-2). The adult patients have lower remission rate (78%) than the children (98%). The duration of remission and the length of survival are also shorter in adults. The incidence of central nervous system (CNS) relapse in adults (27.7%) is higher than in children (7.1%) suggesting that prolonged prophylactic intrathecal methotrexate as given to the children is more effective than the schedule used for adults where intrathecal methotrexate was given only in the first 2 months of therapy. The low incidence of CNS involvement in children on the L-2 protocol compares favorably with other series reported using a combination of cranial irradiation and intrathecal methotrexate. In both adults and children there seemed to be a higher incidence of CNS involvement in patients with initial white blood cell counts greater than 25,000 cells/mm3.

The available staging systems for B-chronic lymphocytic leukemia (B-CLL) do not always predict the clinical course and the prognosis of the disease. In these systems, the pattern of bone marrow histology is not incorporated. In the current report we investigate the prognostic value of the diffuse or nondiffuse pattern of bone marrow involvement in 120 B-CLL patients in relation to their actuarial survival, and we compare these results with the actuarial survival based on the International Workshop system. In addition, we analyze the influence of the diffuse or nondiffuse pattern on the actuarial survival, in relation to the individual clinical stages (A, B, C). All patients were diagnosed and followed-up in the same Unit. Our patients were divided into Stage A (64), Stage B (22), and Stage C (34). They were also subdivided into those with a diffuse (46) and those with a nondiffuse (74) pattern of bone marrow histology. The difference in the actuarial survival in relation to their clinical stage (A, B, C) was statistically significant (P less than 0.025). A greater statistical difference (P less than 0.005) was found when the actuarial survival was analyzed in relation to the diffuse or nondiffuse pattern of bone marrow histology. No statistically significant differences could be found (P greater than 0.1), when the actuarial survival was calculated in every stage (A, B, C), on the basis of the diffuse or nondiffuse pattern of bone marrow histology. When our Stage A and B patients were analyzed for disease progression, in relation to the diffuse or nondiffuse bone marrow histology, it was found that 66.6% of the diffuse Stage A patients and 88% of the diffuse Stage B patients had disease progression as compared to only 8.6% for the nondiffuse Stage A patients and 33% for the nondiffuse Stage B patients. Our findings indicate that: the pattern of bone marrow histology in B-CLL patients is the single most important prognostic parameter in this disease; a clinicopathologic staging system for B-CLL may be justified; and the diffuse pattern of bone marrow histology could be considered as the best criterion for initiation of therapy in these patients.

Sixty patients with chronic B-lymphocytic leukemia (CLL) were studied. The prognosis predictions achieved from clinical and hematologic status, cytogenetic karyotype, immune phenotype, and cellular proliferative responses in vitro were studied by Cox's multivariate analysis. Indicators of poor survival were high dextran sulphate (DxS)-induced CLL-cell proliferation in vitro (P less than 0.002), older age, low percentage T-cells of blood lymphocytes (P less than 0.01), low hemoglobin count, advanced Binet stage, male sex, and high lymphocyte count (P less than 0.05). A therapy-demanding disease was predicted by cytogenetic data (extra chromosome 12, and complex karyotypes [P less than 0.001]), high DxS-induced CLL-cell proliferation (P less than 0.001), high frequency of cytogenetically abnormal metaphases, and advanced Rai and Binet stages (P less than 0.01). The best combinations of parameters included the cytogenetic variables, the DxS-induced CLL-cell proliferation, the relative T-cell number, and the Binet stage. Thus, results from immunologic and cytogenetic studies seem to be helpful in the prediction of prognosis.