Prognostic indicators in diffuse large-cell (histiocytic) lymphoma

0360-3016/86 $3.00 + .oO Copyright 0 1986 Pergamon Press Ltd.

In1 J Rodimion 0ncolog.v Biol Phys., Vol. 12, pp. 593-601 Printed in the U.S.A. All rights reserved.

??Original Contribution: Clinical

PROGNOSTIC MARY B. TODD, D.O.,*

INDICATORS (HISTIOCYTIC)

CAROL S. PORTLOCK,

R. HOLFORD, Yale University

PH.D.$

IN DIFFUSE LYMPHOMA

M.D.,*

LEONARD

LARGE-CELL R. FARBER, M.D.,*

THEODORE

AND JOSEPH R. BERTINO, M.D.?

School of Medicine,

333 Cedar St., New Haven, CT 065 10

Identification of prognostic groups among patients with diffuse large-cell (histiocytic) lymphoma (DHL) would help to select specific therapy for individual patients and allow comparisons among combination chemotherapy clinical trials. The Ann Arbor staging system is of limited value in predicting outcome in diffuse histiocytic lymphoma. Prognostic factors have been examined by various groups without a consensus of reliable prognostic indicators. This study was undertaken to examine the validity of a predictive model for response to treatment and survival in DHL. Eighty-six patients with the diagnosis of DHL treated with combination chemotherapy between the years 1976 and 1982 were examined for prognostic variables influencing response to treatment and survival. The variables examined included: age, sex, presence or absence of systemic symptoms, serum lactic dehydrogenase (LDH), sites of disease involvement, bulk of disease, prior therapy, stage of disease, according to the Ann Arbor classification, and pathological criteria, according to the Lukes Collins classification. Factors achieving a p-value in the 0 to 0.05 range with univariate analysis for predicting response were age and systemic symptoms. Factors significant for overall survival were age and bone marrow involvement. These factors have been found to influence survival in previous studies, but there has not been a consistency regarding the importance of these factors. Large numbers of patients must be examined for various factors in order to allow identification of prognostic groups among patients with DHL. Prognostic factors, Histiocytic/non-Hodgkin’s

lymphoma.

INTRODUCTION

(written communication, Feb., 1985) wherein, pre-treatment parameters in patients with DHL were examined for the influence of these parameters in predicting response to treatment and survival. A model for predicting CR and survival was constructed and used to assign patient risk groups. We have used the same pre-treatment parameters and model to study retrospectively 86 patients with DHL treated at Yale to determine if the model is applicable to an additional series of patients.

The treatment for advanced stage diffuse histiocytic lymphoma (DHL, Rappaport classification) or diffuse large cell lymphoma (Working Formulation) has greatly improved with combination chemotherapy. Several combination chemotherapy programs have resulted in complete remission rates of 40-70%4~1s~20~22~24~25 or even higher.5,“*2’ Despite improvements in the therapy of DHL, there are patients in whom disease is either resistant to chemotherapy or in whom the response to treatment is not durable. It would be of value to identify prospectively prognostic factors that influence response and remission duration following combination chemotherapy. Unfortunately, a clear definition of “poor prognostic” versus “good prognostic” indicators has not emerged from the literature, in part due to the diversity of patient populations and analysis. Recently, a retrospective study has been performed by Danieu et al. at Memorial Sloan-Kettering Cancer Center

METHODS

AND MATERIALS

Patients The names and unit numbers of all patients diagnosed with DHL, according to the Rappaport classification, by the Yale New Haven Department of Pathology from 1977 to 1982 were obtained from the Yale Tumor Registry. There were 106 patients registered. For purposes of data retrieval, all patients who were subsequently followed and

* Dept. of Medicine. t Dept. of Medicine and Pharmacology. $ Dept. of Epidemiology and Public Health. Presented at the Third Rome International Symposium, May 1985. Supported in part by grants from the National Cancer Institute: CA 0834 1 and IN3 1Y from the American Cancer Society.

Reprint requests to: Dr. Mary B. Todd, Department of Medicine, Yale University School of Medicine, 333 Cedar St., New Haven, CT 065 10. Accepted for publication 25 November 1985.

593

594

1. J. Radiation Oncology 0 Biology 0 Physics

treated at Yale New Haven Hospital or treated by a physician on staff at Yale New Haven Hospital were included in this study. Forty-eight of the patients had been randomized between 1976 and 1982 to receive either ACOMLA (doxorubicin, cyclophosphamide, vincristine, methotrexate with leucovorin rescue and cytarabine) or CHOP-B (cyclophosphamide, doxorubicin, vincristine. prednisone, and bleomycin) on a Yale protocol and have been previously reported. ‘6*25An additional 38 non-protocol patients are included in this study. All patients who were treated with combination chemotherapy were considered eligible for study, regardless of their clinical presentation or outcome. Extent of disease was evaluated routinely at time of presentation with complete blood counts with differentials, liver and renal function tests, urinalysis, chest roentgenogram, bilateral bone marrow biopsies and aspirations. and lymphangiography with intravenous pyelogram. In addition. abdominal computerized axial tomography, abdominal ultrasound, gastrointestinal series, and liverspleen scan were performed in those patients with known intra-abdominal involvement or with symptoms suggestive of such involvement. Spinal puncture for cytologic examination of CSF was performed in patients with bone marrow involvement or if there were symptoms suggestive of CNS involvement. The extent of disease was classified according to the Ann Arbor classification.3 B symptoms were noted if present. In those cases classified according to the Lukes Collins immunomorphologic classification,‘3 notation of such classification was made. Measurements of pre-treatment tumor mass were obtained either from radiologic examination (chest X ray, CT scan, ultrasound), physical examination, or surgical observations in those patients undergoing laparotomy for diagnostic purposes. The criteria for bulky disease was a single tumor mass greater than or equal to 8 cm and bulky mediastinal mass was one easily observed on PA view of the chest X ray. Level of site of involvement (LSI) was assigned to each patient, based on the criteria of Danieu et (11.:I) peripheral lymphadenopathy (PLN) or Waldeyer’s ring f spleen; II) extra-nodal disease (EN) -t PLN; III) retroperitoneal lymphadenopathy (RLN) + PLN; IV) bulky mediastinal mass + any other disease; V) RLN with EN + PLN.

Treatment Twenty-six patients were treated with ACOMLA; 20 of these were treated on protocol and have been subjects of earlier reports. ‘6.25ACOMLA consisted of cyclophosphamide (1.0 g/m2) and doxorubicin (40 mg/m’) given on day 1; vincristine (2.0 mg) given on days 1, 8, and 15; and methotrexate (120 mg/m2) followed by cytarabine (cytosine arabinoside) (300 mg/m2) given 1 hour later on days 22, 29, 36, 43, 50, 57, 64, and 71. Leucovorin (25 mg) was administered orally 24 hr after the methotrexate and was continued every six hours for six doses. Three cycles, each lasting three months, were administered.

April 1986, Volume 12, Number 4

Forty-six patients were treated with CHOP-B, 28 on protocol. CHOP-B consisted of cyclophosphamide ( 1.O g/m2), doxorubicin (40 mg/m’), and vincristine (2.0 mg) given on day 1; bleomycin (15 units) given on days 1 and 5; and prednisone (100 mg/m*) given daily for the first 5 days. The cycle was repeated every four weeks, for a total of nine cycles. The total treatment time for both of these drug programs was nine months. Fourteen patients were treated with various combination chemotherapy agents including cyclophosphamide, vincristine, procarbazine and prednisone (C-MOPP), bleomycin, doxorubicin, cyclophosphamide, vincristine and prednisone (BACOP); and cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP). None of these patients were treated on protocol.

A .v.sessment ofresponse One month after completion of therapy, patients were routinely evaluated with the same studies as originally obtained. A pathological complete response (CR) was defined as complete disappearance of all disease for a minimum of two months following re-evaluation one month after completion of therapy. A partial response (PR) was defined as a 50% decrease in the diameter of all measurable disease, lasting at least two months.

Statistical analysis Survival was calculated from the date of diagnosis using the Kaplan-Meier product limit method to obtain survival curves.’ The log rank test was applied to compare survival curves.r* Chi-square analysis for contingency tables was used to determine the association between response and all other variables. Factors achieving a p-value in the 0 to 0.05 range were considered significant.

RESULTS

Clinical characteristics The clinical characteristics of the patients included in this study are shown in Table 1. There were 40 males and 46 females. Thirty-four patients were less than or equal to 55 years of age; 52 patients were older than 55. The extent of disease and treatment for patients with DHL is shown in Table 1. The majority of the patients had Stage III or IV disease (2 1 and 5 1, respectively). Seven Stage II patients and six Stage I patients were treated with chemotherapy and are included in this evaluation. Lymphangiography was not done on one patient who had a bulky mediastinal mass. This patient is excluded from evaluations regarding stage. The Memorial Sloan Kettering definition of the level of site of involvement was applied to this series of patients (Table 1). There were 23 patients with LSI II category (extra-nodal disease + PLN without retroperitoneal disease). Ten of these patients had extra-nodal disease, which is not usually associated with poor survival (for example, skin, breast, tongue, bone). However, nine patients had

595

Prognostic indicators in DHL 0 M. B. TODD etal. Table 1. Clinical characteristics of patients histiocytic lymphoma

Table 3. Luke+Collins

with diffuse

Pathologic

Sex

Male Female

40 46

47 53

Age in years

555 >55

34 52

40 60

Symptoms

A B

57 29

66 34

Bulky disease

Yes No

28 58

33 67

Stage

I II III IV

6* 7t 21 51

7 8 25 60

Level of site of involvement*

I II III IV V

13 23 10 I1 27

16 27 12 13 32

Prior treatment

Yes No

37 49

43 57

Chemotherapy

ACOMLA CHOP-B Other

26 46 14

30 54 16

* Includes 1 patient with IE disease. t Includes 2 patients with IIE disease. $ Two stage IV patients were not evaluated

for retroperitoneal disease and are omitted from this classification.

GI involvement and four had bone marrow involvement, including one patient who had both GI and bone marrow involvement. Eighteen of the patients in Level II had only one site of disease, while five had more than one site of disease. Thirty-seven patients received prior therapy (Table 1): 28 were treated with local radiation therapy, seven with surgery, and two patients with local radiation therapy and minimal (single agent) chemotherapy. Table 2 depicts the extra-nodal sites of involvement. Gastrointestinal was the most frequent site of disease, followed by lung and bone marrow involvement. The Lukes-Collins classification of histology was available on 69 patients, shown in Table 3. The T and B cell

Site Gastrointestinal Bone marrow Lung Bone Liver Other

subtype

of patients

with DHL

No.

%

19 8 5

27.5 11.5 7

37

54

%

No.

Table 2. Extra-nodal

classification

sites of involvement No.

%

20 13 16 7 4 16

23 15 19 8 5 18

Large cell cleaved Large cell non-cleaved Large cell unclassified Immunoblastic sarcoma (T and B cell)

have been combined, because immunoblastic sarcomas cell surface markers were not available for all patients. There were 37 patients with either T or B cell immunoblastic sarcoma, 19 with large cell cleaved, and only eight and five with large cell non-cleaved and large cell unclassified, respectively. Prognostic,factor analysis.for response

Data is presented for attainment of CR in Tables 4 and 5. Eighty-one patients were evaluable for response. One patient died prior to completing the first course of chemotherapy. Four patients were not adequately re-staged following therapy to determine CR versus PR. All four of these patients did relapse; two are alive with disease and two are dead with disease. Clinical characteristics that achieved a p-value in the 0 to 0.05 range in univariate analysis for predicting complete response were age (X2 = 7.44; p = 0.006) and systemic symptoms (X2 = 2.95; p = 0.009). Sex, large tumor mass, retroperitoneal involvement, large mediastinal mass, extra-nodal disease, bone marrow involvement, or GI involvement were not significant for predicting CR or PR. There were only two patients with LDH less than 225 and neither of these patients were evaluable for response. LDH of 225-500 versus LDH greater than 500 was not significant for predicting CR versus all other responses (X2 = 2.06; p = 0.151). Stage of disease, according to the Ann Arbor Classification, was not significant for predicting complete response in this study, as shown in Table 5. The level of site of disease as defined by Danieu et al. at Memorial Sloan Kettering was also not predictive of response in this series of patients. The Lukes-Collins classification did not contribute to predictions regarding response. Prior therapy did not alter response in this series. There was no significant difference in response between patients treated with ACOMLA or CHOP-B, including those patients treated “off’ protocol. However, treatment with “other” combination chemotherapy regimens was associated with a significantly reduced CR rate (X2 = 6.32: p = 0.042). Prognostic factor analysis for survival

Univariate analysis of survival data, using the log-rank test, showed age and bone marrow involvement to be significant factors for predicting survival (p = 0.010 and p = 0.024, respectively; see Figures 1 and 2). The presence or absence of systemic symptoms, which was predictive

596

I. J. Radiation Oncology 0 Biology 0 Physics Table 4. Proportion

of complete

April 1986, Volume 12, Number 4

responses

according

to clinical characteristics

CR no.*

%

X2

df

p-value?

Sex

Males Females

28139 29142

72 69

.a7

I

>.70 (NS)

Age

555 >55

28132 29149

87 59

7.44

I

.006

Symptoms

A B

42155 I5126

76 58

2.95

I

.009

LDH

<225 225-500 >500

I2 I7122 I5126

77 58

2.06

1

>. 10 (NS)

Yes No

I8125 39156

72 70

.05

1

>.80 (NS)

Yes No

25134 32145

73 71

.06

1

>.80 (NS)

Yes No

IO/I2 47169

83 68

1.14

1

>.20 (NS)

Yes No

3415 1 23130

67 77

.91

1

>.30 (NS)

Yes No

7112 50169

58 72

.98

I

>.30 (NS)

Yes No

12119 45162

63 73

.62

I

>.30 (NS)

Bulky disease Retroperitoneal Mediastinal Extranodal

involvement

mass involvement

Bone marrow Gastrointestinal

involvement

* Five patients were not evaluable for CR versus PR: one due to early death and four due to inadequate evaluation following the completion of therapy. All four patients with inadequate evaluation did relapse: two are dead of disease and two alive with disease. They are not included for evaluation of CR. _t NS indicates not significant with p-value of significance defined as between 0.0 and 0.05 in this and the following tables.

Table 5. Proportion

of complete

responses

according

to extent of disease and pathological

characteristics

CR no.

Y0

X2*

df

1 II 111 IV V

IO/12 13122

3.41

3

>.30 (NS)

10112 16124

83 59 89 83 67

Stage

1 II III IV

415 517 16120 31/48

80 71 80 65

1.57

2

>.30 (NS)

Prior therapy?

RT RT + Chemo Surgery None

I5126 l/2 517 35146

58 50 71 76

2.01 1.91

1 1

>.I0 (NS) >.I0 (NS)

Chemotherapy

ACOM LA CHOP-B Other

I8126 34143 5112

69 79 42

6.32

2

>.20 (NS) >.20 (NS) .042

Cleaved Non-cleaved Unclassified Immunoblastic

14119 618 515 22134

74 75 100 65

1.45

I

>.20 (NS)

Level of site

Histology+

819

* X2 comparisons done only with categories of more than 10 patients. t The first X2 value in prior therapy category is for RT alone versus no prior therapy. therapy combined versus no prior therapy. $ The X2 value is expressed for immunoblastic sarcoma versus all others.

pvalue

The second X2 value is for a11forms of prior

597

Prognostic indicators in DHL 0 M. B. TODD etal.

E

1.00

.a > ._

.8

>

5 v)

F

L

:‘_p

years

0’

age

- y- ’ 55years Ofage

_

.7

55

1. ‘____

g ._ c z ;

Le.

.6

I-

‘--

_ _______

__ 1 L______ t

.5 .3

,/

---_---_____

_

I

I

.2

.O

-

I I

lo

20

30

40

50

60

70 Survival

100

80

90

Time

In Months

Fig. 1. Survival by age.

of response, was not predictive of survival (a = 0.340). GI involvement, retroperitoneal involvement, mediastinal mass, large tumor mass, extra-nodal disease and sex were not significant for survival, as shown in Table 6. The median survival for all patients in the study was 7 1 months. The median survival for patients treated with ACOMLA was 71 months (5.9 years). The median survival for patients treated with CHOP-B was 76 months (6.3 years). The median survival time for patients treated with other combination chemotherapy was 19.2 months (1.6 years). The differences in treatment arms were not significant for survival between ACOMLA or CHOP-B

E ._ > ._ > S v)

(p = 0.559). As shown in Figure 3, there was no significant difference between survival in patients treated with “other” combination chemotherapy regimens (p = 0.2 14). Stage of disease, LSI, prior treatment, and histopathology did not influence survival (see Table 7). LDH was not a factor in predicting survival (p = 0.145) in this series of patients (Figure 4). DISCUSSION

The Ann Arbor Staging System3 was developed to group patients with similar prognostic variables of Hodgkins’ disease. The usefulness of this staging system for thera-

1.00

AhWlt ---

Present

.8 .7

: ._ z z

.6

h”

.3

? I i___________

5

_______

I

__~

I

L

I

I I

.2 .l .O

10

20

30

40

50

60

70 Survival

Fig. 2. Sun marrow inv

of patients with bone marrow involvement ment (absent).

100

80

90

Time

In Months

(present) versus survival of patients without bone

598

I. J. Radiation Oncology 0 Biology 0 Physics

April 1986, Volume 12, Number 4

Table 6. Survival accordinp. to clinical characteristics Median survival (months)

Statistic (Mantel-Cox)

df

p-value

Sex

Male Female

71 90

0.012

1

0.914 (NS)

Age

555 >55

76 52

6.578

1

0.010

Symptoms

A B

71 76

0.911

I

0.340 (NS)

LDH

225-500 >500

76 23

2.120

I

0.145 (NS)

Bulky disease

Yes No

76 71

2.096

1

0. I48 (NS)

Yes No

71 *

I .809

1

0.179 (NS)

Yes No

*

I.902

I

0.168 (NS)

71

Yes No

90 71

.686

I

0.407 (NS)

Yes No

I5 76

5.065

1

0.024

Yes No

47 71

1.330

1

0.249 (NS)

Retroperitoneal Mediastinal Extra-nodal

involvement

mass involvement

Bone marrow Gastrointestinal

involvement

* Median survival not yet reached.

peutic decisions and for analysis of long-term outcome in Hodgkins’ disease has continued. However, the application of the Ann Arbor Staging System to Non-Hogkins lymphomas, while routine in the majority of institutions, may be of limited usefulness.” Although there have been several studies supporting the usefulness of the Ann Arbor Staging System in diffuse histiocytic lymphoma,2~4~6~‘4

z

._ > .I

>

i ;

there have been a number staging system did not cOme 10.11.16.23-25

of other studies in which this significantly influence out-

In diffuse histiocytic lymphoma, there is clearly a need for the identification of prognostic factors. Response to treatment and durability of complete response are heterogeneous. A mechanism for identifying those patients

1.00

ACOMLA ---

.8

(26

CHOP-8 . . -

(45

OTHER

(13

i)ts

16

pts, ~1s

28

censored) censored)

5 censored)

.?

5 ._ Y

.6

5

.5

g h

.3

L_,

.2

"1

i_..-..-.._.., -

I I a

i.._.._.._.._17-~_r----:

I

-

1 I__._

.l

.O

IO

20

30

40

50

60

70 Survival

Fig. 3. Survival of patients

according

to various chemotherapy

80

90

Time

In Months

regimens.

1

Prognostic indicators in DHL 0 M. B. TODD

599

et al.

Table 7. Survival according to extent of disease and pathological characteristics

Level of site

Stage

Chemo

Histology

Median survival (months)

Statistics (Mantel-Cox)

I II III IV V

* 71 71 *

4.149

4

0.385 (NS)

I II III IV

* *

1.428

3

0.699 (NS)

71 71

ACOMLA CHOP-B Other

71 76 19

3.042

2

0.559 (NS)

3.084

2

0.214 (NS)

Cleaved Non-cleaved Unclassified Immunoblastic

90 71 *

6.364

4

0.174 (NS)

df

p-value

52

42

* Median survival not yet reached.

who are at high apy or who are at sive chemotherapy temic symptoms weight loss) were

risk for a poor response to chemotherrisk for relapse might allow more aggresto be offered to such patients. Sys(night sweats, fever, and more than 10% shown in some studies to be a prognostic

factor’17,‘4; however, other studies, including reports from some of the same groups, could not confirm this.2,6,‘0,’ ‘*24,25 Similarly, other clinical presentations have resulted in conflicting data. Sites of disease, particularly bone marrow or gastrointestinal, were found to be important in predicting survival in some studies,‘,6,7but were not confirmed by others.“,‘6.2’,23,25Large tumor mass or bulky disease of either 8 or 10 cm were significant in some

z

._ > ._ >

studies, (6, 7) but again, this was not confirmed.8,“325 ’sex,’ liver involvement’ and prior therapy432’have Age,83’ all been shown to be significant in some reports and without significance2,4s6324 in others. There have been several attempts to reclassify diffuse histiocytic lymphoma on the basis of histopathology, in the hope of identifying prognostic variables and discrete subsets of disease. The Lukes and Collins classification,‘3 based on immunomorphology, contain five distinct lymphoma types. Several series of patients from different institutions*v’4~‘53’7.*33*5,26 have been examined using this classification with conflicting results. Such conflicting data, at times coming from the same group of investiga-

1.00

_

Serum

LDH

>

---

Serum

LDH

22:j-500

500

.a

L

g

.7

g ._ Ic, 5

.6 5

$ h

.3 .2 .l

.o ’

I

I

I

I

I

I

I

I

I

1

10

20

30

40

50

60

70

60

90

100

Time

In Months

Survival Fig. 4. Survival of patients

based on serum LDH.

600 tars, “,15,‘7,25 severely

1. J. Radiation Oncology 0 Biology ?? Physics limit

the usefulness

of such a clas-

sification for predicting prognosis. Recently, Memorial Sloan Kettering examined a predictive model for response to treatment and survival in advanced DHL. The factors examined included age, sex, prior therapy, systemic symptoms, serum LDH, sites of disease involvement and bulk of disease. Sixty-six patients were examined and each was assigned an overall level of site involvement. Factors important in univariate analysis for predicting a CR were: serum LDH, LSI, retroperitoneal involvement, pre-treatment tumor mass, bulky disease. GI involvement, and systemic symptoms. Using logistic regression, LDH was the most important factor along with LSI. A model for predicting CR was constructed. based on serum LDH level and LSI. Factors significant for survival, using the Kaplan-Meir product limit method, were serum LDH level, bone marrow involvement, LSI, GI involvement, retroperitoneal involvement. retroperitoneal count, presence of bulky mediastinal mass and extra-nodal involvement. Using the Cox regression model, LDH followed by LSI were, again, the only factors that were predictive in this model. The purpose of this study was to test the validity of the Memorial Sloan Kettering predictive model for response to treatment and survival in DHL. In this series of patients, LDH was not significant in predicting CR versus all other responses and LSI was not predictive of response. The only factors of significance in the prediction of CR in this series were age, presence or absence of systemic symptoms, and treatment with chemotherapy other than ACOMLA or CHOP-B. Factors important in overall survival in this series were age and bone marrow involvement. Systemic symptoms, important in predicting response, were not important in overall survival. LDH and LSI were not predictive of survival in this study. Forty-eight of the 86 patients reported in this study were treated on protocol and have previously been re-

April 1986, Volume 12, Number 4

ported.25 In our earlier report we did not find systemic symptoms helpful in predicting prognosis. There were equivalent percentages of patients with systemic symptoms treated on protocol and off protocol (33% versus 34%) therefore, this difference is unexplained. In the earlier study, there was a trend toward older patients not responding, as well as, younger patients; the trend was not significant, however. In the earlier study, the median age of responders versus non-responders was compared, while in this study, patients less than or equal to 55 years of age were compared to patients greater than 55 years of age, as a group. This difference in the method of calculation of the age groups may explain the differences in the earlier study compared to this one. In this current series, there is a difference in CR rate dependent upon chemotherapy. Patients receiving ACOMLA or CHOP-B had equal CR rates, but those patients receiving “other” regimens had less frequency of CR. However, this difference is not reflected in survival, in part due to late relapses in patients receiving ACOMLA and CHOP-B, and in part due to the low numbers of patients treated with chemotherapy regimens other than ACOMLA and CHOP-B. This study emphasizes the difficulties in comparing patients with DHL. Several studies have been done at multiple institutions with differing results. Often, as within our own data, there are differences found when the data are examined at various times. To date, there is no consensus regarding prognostic variables in DHL. This suggests that DHL is heterogeneous and until the factors contributing to this heterogeneity can be better identified, large numbers of patients must be examined over long periods of time to evaluate prognostic factors. The predictive model, as outlined by Memorial Sloan Kettering, was not useful in this series of patients, but additional patients need to be examined before the true usefulness of this model can be determined.

REFERENCES 1. Armitage, J.O., Dick, F.R., Corder, M.P., Ciarneau, S.C., Platz, C.E., Slymen, D.J.: Predicting therapeutic outcome in patients with diffuse histiocytic lymphoma treated with cyclophosphamide, adriamycin, vincristine, and prednisone (CHOP). Cancer 50: 1695- 1702, 1982. 2. Armitage, J.O., Dick, F.R., Platz, C.E., Corder, M.P.. Leimert, J.T.: Clinical usefulness and reproducibility of histologic subclassification of advanced diffuse histiocytic lymphoma. Am. J. Med. 67: 929-934, 1979. 3. Carbone, P.P., Kaplan, M.S., Musshoff, K., Smithers, D., Tubiana, M.: Report to the committee on Hodgkin’s disease staging classification. Cuncer Res. 31: 1860- 186 I, 197 1. 4. Elias, L., Portlock, C.S., Rosenberg, S.A.: Combination chemotherapy of diffuse histiocytic lymphoma with cyclovincristine and prednisone phosphamide, adriamycin, (CHOP). Cuncer42: 1705-1710, 1978. 5. Fisher, RI., DeVita, V.T., Hubbard, S.M., Longo. D., Wesley, R., Chabner, B., Young, R.C.: Diffuse aggressive lym-

6.

7.

8.

9.

IO.

phoma: increased survival after alternating flexible sequences of ProMACE and MOPP chemotherapy. Ann. ht. Med. 98: 304-309, 1983. Fisher, RI., DeVita, V.T., Johnson, B.L., Simon, R., Young, R.C.: Prognostic factors for advanced diffuse histiocytic lymphoma following treatment with combination chemotherapy. Amer. J. Med. 63: 177-182, 1977. Fisher, RI., Hubbard, S.M., DeVita, V.T., Berard, C.W., Wesley, R., Cossman, J., Young, R.C.: Factors predicting long-term survival in diffuse mixed, histiocytic, or undifferentiated lymphoma. Blood 58: 45-5 1, 198 1. Jagannath, S., Velasques, W.S., Tucker, S.L., Manning, J.T., McLaughlin, P., Fuller, L.: Stage IV diffuse large-cell lymphoma: A long-term analysis. J. Clin. Oncol. 3: 39-45, 1985. Kaplan, E.L., Meier, P.: Non-parametric estimations from incomplete observations. J. Am. Stat. Assoc. 53: 1484- 1493, 1976. Koziner, B., Little, C., Passe, S.. Thaler, L.T., Sklaroff, R.,

Prognostic indicators in DHL 0 M. B. TODD et al.

Il.

12. 13. 14.

15.

16.

17.

18.

19.

Straus, D.J., Lee, B.J., Clarkson, B.D.: Treatment of advanced diffuse histiocytic lymphoma: An analysis of prognostic variables. Cancer 49: 157 1- 1579, 1982. Laurence, J., Coleman, M., Allen, S., Silver, R.T., Pasmantier, M.: Combination chemotherapy of advanced diffuse histiocytic lymphoma with the six-drug COP-BLAM regimen. Ann. Int. Med. 97: 190-195, 1982. Lee, E.T.: Statistical Methods for Survival Data Analysis. Belmont, Lifetime Learning Publications, 1980. Lukes, R.J., Collins, R.D.: New approaches to classification of the lymphomata. Br. J. Cancer 31: l-27, 1975. Nathwani, B.N., Dixon, D.O., Jones, S.E., Hartsock, R.J., Rebuck, J.W., Byrne, G.E., Jr., Sheehan, W.W., Kin, H., Coltman, C.A., Rappaport, H.: The clinical significance of the morphological subdivision of diffuse “histiocytic” lymphoma: A study of 162 patients treated by the Southwest oncology group. Blood 60: 1068- 1074, 1982. Nathwani, B.N., Kim, H., Rappaport, H., Solomon, A.J.: Non-Hodgkin’s lymphomas: A clinicopathologic study comparing two classifications. Cancer 41: 303-325, 1978. Newcomer, L.N., Cadman, E.C., Nerenberg, MI., Chen, M., Bertino, J.R., Farber, L.R., Prosnitz, L.R.: Randomized study comparing doxorubicin, cyclophosphamide, vincristine, methotrexate with leucovorin rescue and cytarabine (ACOMLA) with cyclophosphamide, doxorubicin, vincristine, prednisone, and bleomycin, (CHOP-B) in the treatment of diffuse histiocytic lymphoma. Cancer Treat. Rep. 66: 1279-1284, 1982. Newcomer, L.N., Nerenberg, MI., Cadman, E.C., Waldron, J.A., Farber, L.R., Bertino, J.R.: The usefulness ofthe LukesCollins classification in identifying subsets of diffuse histiocytic lymphoma responsive to chemotherapy. Cancer 50: 439-443, 1982. Rodriguez, V., Cabanillas, F., Burgess, M.A., McKelvey, E.M., Valdivieso, M., Bodey, G.P., Freireich, E.J.: Combination chemotherapy (“CHOP-Bleo”) in advanced (NonHodgkin) malignant lymphoma. Blood 49: 325-333, 1977. Rosenberg, S.A.: Validity of the Ann Arbor staging classi-

20.

21.

22.

23.

24.

25.

26.

601

fication for the non-Hodgkin’s lymphoma. Cancer Treat. Rep. 61: 1023-1028, 1977. Schein, P.S., DeVita, V.T., Jr., Hubbard, S., Chabner, B.A., Canellos, G.P., Bernard, C., Young, R.C.: Bleomycin, adriamycin, cyclosphosphamide, vincristine and prednisone (BACOP) combination chemotherapy in the treatment of advanced diffuse histiocytic lymphoma. Ann. Int. Med. 85: 417-422, 1976. Skarin, A.T., Canellos, G.P., Rosenthal, D.S., Case, D.C., MacIntrye, J.M., Pinkus, G.S., Moloney, W.C., Frei, E.: Improved prognosis of diffuse histiocytic and undifferentiated lymphoma by use of high dose methotrexate alternating with standard agents (M-BACOD). J. Clin. Oncol. 1: 91-98, 1983. Skarin, A.T., Rosenthal, D.S., Moloney, W.C., Frei, E., III: Combination chemotherapy of advanced non-Hodgkin’s with bleomycin, adriamycin, cyclophosamide, vincristine and prednisone (BACOP). Blood 49: 759-770, 1977. Strauchen, J.A., Young, R.C., DeVita, R.T., Jr., Anderson, T., Fantone, J.C., Berard, C.W.: Clinical relevance of the histopathological subclassification of diffuse “histiocytic” lymphoma. N. Engl. J. Med. 299: 1382-1387, 1978. Sweet, D.L., Golomb, H.D., Ultmann, J.E., Miller, J.B., Stein, R.S., Lester, E.P., Mintz, J.E., Bitran, J.D., Streuli, R.A., Daly, K., Roth, N.O.: Cyclophosphamide, vincristine, methotrexate with leucovorin rescue and cytarabine (COMLA) combination sequential chemotherapy for advanced diffuse histiocytic lymphoma. Ann. Znt. Med. 92: 785-790, 1980. Todd, M., Cadman, E., Spiro, P., Bertino, J., Farber, L., Waldron, J., Fischer, D.: A follow-up of a randomized study comparing two chemotherapy treatments for advanced diffuse histiocytic lymphoma. J. Clin. One. 2: 986-993, 1984. Warnke, R., Miller, R., Grogan, R., Pederson, M., Dilley, J., Levy, R.: Immunologic phenotype in 30 patients with diffuse large-cell lymphoma. N. Eng. J. Med. 303: 293-300, 1980.