Combined Modality Therapy in Stage I// and Stage I//€ Diffuse Large Cell 1 ymphomas

WILLIAM VELASQUEZ, MD,' LILLIAN M. FULLER, MD,t KYOUNG K. OH, MD,' FREDRICK 6. HAGEMEISTER, MD,' JANE A. SULLIVAN, BS,' JOHN T. MANNING, MD,* AND C. C. SHULLENBERGER, MD'

Combined modality therapy consisting of CHOP-Bleo (cyclophosphamide, Adriamycin [doxorubicinj, vincristine (Oncovin), prednisone, and bleomycin) and limited-field radiotherapy was given to 47 patients with Stage 111 and IIIE diffuse large cell lymphomas. Overall 5-year survival and disease-free survival figures are 64% and 5370, respectively. Twenty-nine patients achieved and maintained a complete remission during treatment; 92% of these patients are alive at 5 years, and 87% are disease-free. Prognostic factors were assessed for the entire group. Age, "B" symptoms, extranodal involvement, and extensive abdominal involvement had little effect on survival. However, the presence of mediastinal involvement and extensive abdominal involvement was associated with a poor response rate and survival.

Cancer 53:1478-1483, 1984.

ODERN COMBINATION CHEMOTHERAPY has h- M proved both the remission rate and the survival of patients with advanced diffuse large cell lymphomas. However, the proportion of patients living at 5 years re- mains at approximately less than 40% for most se- r i e ~ . ' - ~ The major limiting factor for a better result has been a significant incidence of relapse. For CHOP-Bleo (cyclophosphamide, doxorubicin hydrochloride [Adria- mycin], vincristine [Oncovin], prednisone, and bleo- mycin), the Southwest Oncology Group (SWOG) has re- ported a relapse rate of 25% to 40% in one year.4 At the National Cancer Institute (NCI), Anderson and associates found a close relationship between sites of relapse and sites of initial nodal involvement in patients who achieved a complete remission after treatment with bleomycin, Adriamycin, cyclophosphamide, vincristine, and pred- nisone (BACOP) or nitrogen mustard, vincristine, pro- carbazine, and prednisone (MOPP).' While Fisher and associates reported a lower relapse rate for VP-16, Ad- riamycin, cyclophosphamide, and vincristine alternated with MOPP (ProMACE), the mortality rate related to the treatment was approximately 1 0 % ~

Presented in part at the 18th Annual Meeting ofthe American Society of Clinical Oncology. April 1982.

From the Departments of *Internal Medicine. ?Radiotherapy, and +Pathology, The University of Texas M. D. Anderson Hospital and Tumor Institute at Houston, Texas Medical Center, Houston, Texas.

Supported in part by US Public Health Service grant CA-6294. Address for reprints: William Velasquez, MD, Department of Internal

Medicine, Section of Hematology, U. T. M. D. Anderson Hospital, 6723 Bertner Avenue, Houston, T X 77030.

Accepted for publication January 10, 1983.

There is little information on results for Stage I11 only diffuse large cell lymphomas, because the number of pa- tients in most series is Furthermore, some in- vestigators have included other histopathologic subtypes of diffuse lymphomas in reporting their data.8 In the large Manchester series of Stage I11 patients with diffuse his- tology of non-Hodgkin's lymphomas, who were treated with vincristine, Adnamycin, and prednisone (VAP), comparative results were not provided for the different histopathologic subtypes.'

In 1966, we initiated a program for treating Stage I11 diffuse large cell lymphomas with two cycles of cyclo- phosphamide, vincristine, and prednisone (COP) and ra- diotherapy." The preliminary results were encouraging. However, when Adriamycin and bleomycin became available, these drugs were included (CHOP-Bleo). Since a significant relapse rate in initial nodal and extranodal sites were observed in patients with generalized disease who were treated with CHOP-Bleo alone, we decided to continue to use a combined modality program for our Stage 111 patients. This program was initiated in 1974, and preliminary results were first reported in 1977 and again in 1979.12

Material and Methods

Forty-seven consecutive adult patients with Stage 111 and Stage IIIE diffuse large cell lymphomas have been accessed on this study. The median follow up has been 33 months (range, 4-74 months). The mean age of these patients was 55 years (range, 20-78 years). The slides for all 47 patients were re-examined at the time of this anal-

1478

No. 7 STAGE 111 DIFFUSE LARGE CELL LYMPHOMAS Velasqziez ef al. 1479

ysis. In four. the histopathology was subclassified as im- munoblastic sarcoma; and two additional patients had some features that were suggestive of immunoblastic sar- coma. The remainder were classified simply as large cell lymphoma.

Patients were sequentially staged according to the Ann Arbor system. Routinely, the initial investigation included lymphangiography and bone marrow biopsies. Seven pa- tients with negative lymphangiograms had staging lapa- rotomies, which revealed occult abdominal involvement. In 36 of the 47 patients, the involvement was limited to lymph nodes, and 6 of these patients had constitutional symptoms. The other I I patients had a single extranodal site of disease in addition to nodal involvement, and only one of these patients had “B” syimptoms (Table I) . Ab- dominal involvement was categorized as extensive or minimal. Extensive abdominal involvement was defined as a mass greater than 7 cm in diameter, or as involvement of both the para-aortic and iliac modes. Less disease was categorized as minimal (Table 2) In 26 patients, the ab- dominal involvement was extensive; in the other 21 pa- tients, the involvement was minimal. Mediastinal in- volvement was demonstrated in 15 of the 47 patients. Eight of these patients with mediastinal involvement had extensive abdominal disease, and seven had minimal ab- dominal disease.

Treatment

Our plan was to alternate 4 cycles of CHOP-Bleo che- motherapy with radiotherapy to areas of initially involved disease. The drug schedules and dosages were those used in the SWOG studies (Table 3). Following induction che- motherapy, radiotherapy was initiated with treatment to the abdomen. Treatment fields were limited to the known areas of disease. The tumor dose was 3000 rad delivered in 20 fractions (4 weeks). The major portion of the liver was excluded from the treatment. ‘The dose to the kidneys was limited to I800 tad by appropriate shielding during treatment to the posterior field. After a 3-to-4 week re- covery period, CHOP-Bleo was resumed for another 4 cycles. Subsequently, radiotherapy was delivered to the next major area of involvement. The chemotherapy por- tion of the program was completed with 4 cycles of COP. Following this, radiotherapy was completed to the re- maining sites of initial involvement (Table 4).

In this program, the treatment schedule was modified in 15 patients. In eight patients, treatment was initiated with radiotherapy during the initial evaluation as an emergency measure. This was usually done to prevent respiratory obstruction caused by massive or rapidly pro- gressive head and neck or mediastinal disease. In an ad- ditional seven patients, chemotherapy was compromised because Adriamycin had to be either reduced by 50%, or

TABLE 1. Stage 111 and Il lE Diffuse Large Cell Lymphomas

No Total no. constitutional Constitutional

Involvement of patients symptoms (A) symptoms (B)

Nodal: Stage I I I 36 30 6 Extranodal:

Total no. of Stage HIE I I * 10 I

patients 47 40 7

* Waldeyer’s ring: 5; maxillary sinus: I ; soft tissues ofthe extremities: 2; GI tract: 3.

TABLE 2. Extent of Abdominal Involvement in Stage I11 Diffuse Large Cell Lymphomas

Criteria Extensive Minimal

Clinical/surgical

Radiological Mass > 7 cm Present Absent

Mass effect* Present Absent Positive lymphangiogram Para-aortic and Para-aortic or

iliac iliac

* Any imaging technique.

deleted due to severe cardiovascular disease. During treatment, patients were followed closely, and re-evaluated at regular intervals. Those with poor response or pro- gressive disease during the initial 4 cycles of chemotherapy were withdrawn from the protocol, but were included in this analysis.

TABLE 3. CHOP-Bleo Schedule

~~~ ~

Cyclophosphamide 750 mg/m2 IV 1 H ydroxyldaunom ycin 50 mg/m2 IV I Oncovin* 1.4 mg/m2 IV 1 Prednisone 100 mg daily PO 1-5 Bleomvcin 5 ma/m’ Iv I

CHOP-Bleo: cyclophosphamide. doxorubicin hydrochloride (Adna- mycin), vincristine (Oncovin), prednisone. and bleomycin: PO: orally; IV: intravenously.

* Oncovin (vincristine): dose not to exceed 2 mg total.

TABLE 4. Combined Modality Schema

Treatment No. of cycles

CHOP-Bleo 4 cycles

CHOP-Bleo 4 cycles

COP 4 cycles

Radiotherapy to abdomen: limited field

Radiotherapy lo involved upper torso sites

-

-

CHOP-Bleo: cyclophosphamide, doxorubicin hydrochloride (Adna- mycin), vincristine (Oncovin), prednisone, and bleomycin: COP cy- clophosphamide. vincristine, prednisone.

1480

1 .o

.9

.8

.7

.6 Q a i- g .5

B .4

.3

.2

. 1

' '91% b

CANCER April 1 1984 VOl. 53

4 )

TOTAL FAIL SURVIVAL , h 29 2 0 IN REMISSION ] p z.002 18 13 A WITH

1 1 5 %

Z .60 0 I-

f .50 8 .40

I NON-FAILURE

~

-

-

I 1 I 1 1 I I I

0 12 24 36 48 60 72 84 96 MONTHS

FIG. 1. DLCL Stage 111 survival related to remission status: patients in complete remission at end of therapy v m r u patients who progressed during therapy.

n .80 1 -

.70 t TOTAL FAIL DISEASE FREE

29 IN REMISSION AT END OF THERAPY

I NON-FAILURE

.20

.30 I l o t 0 ' 1 1 1

1 I I I

0 12 24 36 48 60 72 84 96

MONTHS

FIG. 2. DLCL Stage 111 combined therapy: Disease-free patients in remission at the end of therapy.

Analysis of the Data

Survival and disease-free survival curves were calculated from the date of admission to last follow-up according to the method of Ka~1an-Meier.I~ Gehan's modification of the generalized Wilcoxon test was employed to evaluate difference in result^.'^ Separate analyses were undertaken to determine the possible influence of specific prognostic factors on the complete remission rate as well as survival. These included age, B symptoms, extranodal presenta- tions, the status of the mediastinum, and the extent of abdominal involvement.

Results

Twenty-nine patients achieved and maintained a com- plete remission during protocol treatment. The 5-year survival figure for these patients was 91% (Fig. 1). The corresponding disease-free figure was 87%; no relapses occurred after 24 months (Fig. 2). The other 18 patients achieved only a partial remission or failed to respond to treatment. There was a steady attrition of these patients over 24 months, until survival was only 15% (Fig. 1). The 5-year survival for the entire group of 47 patients was 64% (Fig. 3). The corresponding disease-free figure was 53%.

The effect of age on survival was not significant. For patients older than age 50 years, the 5-year survival was 67% as compared with 48% for patients younger than age 50 years ( P = 0.819). The number of patients with extranodal presentations and/or B symptoms were too small for statistical analyses. However, neither of these factors appeared to adversely influence results. Of the I I patients with extranodal presentations, 7 patients, in- cluding the patient with B symptoms and 2 patients with gastrointestinal presentations. have been free of disease from 17 to 34 months. Of the total group of seven patients with B symptoms, five are surviving free of disease.

Complete remission and survival were related to the total tumor burden, expressed by the combination of ex- tent of abdominal disease and the status of the medias- tinum. In 32 patients with normal mediastinal silhouette, the extent of abdominal involvement did not appear to affect the complete remission rate or survival. Thus, in 14 patients with minimal abdominal disease, complete remission was achieved in 7 1% with a calculated 5-year survival of 60%, while in 18 patients with extensive ab- dominal involvement, 72% achieved complete remission with a calculated 5-year survival of 79%. However, the presence of mediastinal disease in 15 patients appears to influence both the complete remission rate and the sur- vival. For mediastinal disease patients with minimal ab- dominal involvement, the complete remission rate was 58%, and the 5-year survival was 57%. Mediastinal disease patients with extensive abdominal involvement had a

No. 7

1 .o

.9

.8

.7

z .6 P I- 5 .5 0 g .4

.3

.2

. 1

0.80

0.70

5 0.60 F a: 2 0.50 a 0.40

0.30

0.20

0.10

0.00

STAGE IKI DIFFUSE LARGE CELL LYMPHOMAS - Velasqztez et al.

~

- 0

%---- TOTAL FAIL

47 15 0 SURVIVAL

47 21 A DISEASE FREE I NON FAILURE

0 12 24 36 48 60 72 84 96 MONTHS

FIG. 3. Stage 111 DLCL: overall survival and disease-free survival for all patients.

complete remission rate of only 25%, and a corresponding survival of 35% at 40 months (Table 5, Figure 4).

The influence of treatment modifications on complete response rate is shown in Table 6. Of the 32 patients who were treated according to protocol, 22 achieved and maintained a complete remission during therapy. Mod- ifications in the protocol of either a compromise in the chemotherapy or initiation of treatment with radiotherapy resulted in a complete maintained remission in only 7 of 15 patients. Of the total group of 47 patients, 18 either failed to respond to treatment or failed to achieve and maintain a complete remission during the planned ther- apy. Three additional patients whlo were in complete re- mission after receiving all planned treatment relapsed in

T A M F 5. Relation of Complete Remission to Mediastinal and Abdominal Status in Stage 111 Patients

Mediastinal status

Negative Positive

Complete Complete Abdominal No. of remission No. of remission involvement patients (%) patients (%)

Minimal 14 10 (71%) 7 4 (58%) Extensive 18 13 ( 7 2 % ) 8 2 ( 2 5 % )

0.90 1'0° h 1481

.iP 35%

TOTAL FAIL 26 7 0 ALL PATIENTS 18 3 A NON MEDIASTINAL; p ~ 0,019 8 4 MEDIASTINAL

' NON FAILURE

FIG. 4. Stage I l l DLCL: survival of patients with extensive abdominal involvement. The difference in survival related to concurrent mediastinal involvement is shown.

the follow-up period. Of the 18 patients who failed during therapy, 12 did so in the initial chemotherapy phase. This group included three of the four patients with a diagnosis of immunoblastic sarcoma. In the other six patients, which included the remaining patient with immunoblastic sar- coma, the disease progressed later in the course of the treatment program.

Tolerance to this combined modality program was rea- sonably good. Myelosuppression was moderate, and none of the patients required transfusions. However, one patient died of sepsis during the first course of chemotherapy.

TABLE 6 . Influence of Treatment Modifications on ComDlete Resoonse Rate

Protocol deviations Treatment

per Delayed Compromised protocol chemotherapy' chemotherapyt Total

Total no. of patients 32 8 7 4 1 Maintained complete

Failures remission 22 4 3 29$

During initial

After initial chemotherapy 8 3 I I ?

chemotherapy 2 I 3 6

* Treatment initiated with radiotherapy. t Adriamycin (doxorubicin) reduced by 50% or deleted $ Three patients relapsed in the follow-up period.

I482 CANCER April 1 1984 Vol. 5 3

Early in the treatment program, two patients developed radiation hepatitis. and responded promptly to cortico- steroid therapy. Subsequently, this problem was elimi- nated by increasing the amount of shielding over the right lobe of the liver during radiotherapy from 2 half-value layers (HVL) to 5 HVL. To date, late sequelae have not been a significant problem. Two patients developed aseptic necrosis of the femoral head. Xerostomia was not a major problem among the six patients who were treated to Wal- deyer’s ring or the maxillary antrum with tumor doses ranging between 4000 to 5000 rad given in 4 to 5 weeks. Since it was anticipated that the majority ofthese patients would be over age 35, sterility studies were not undertaken. Only three patients were under age 35. There have been no cases of acute leukemia; however, three patients de- veloped second malignancies. These were cases of car- cinoma of the lung, adenocarcinoma of the pancreas, and malignant fibrohistiocytoma of the soft tissues ofthe thigh, which was not previously irradiated.

Discussion

In the past, the advocated treatment of choice for pa- tients with advanced diffuse large cell lymphomas has been combination chemotherapy. In a retrospective study of prognostic factors, Fisher and coworkers found that the survival of 12 Stage 111 disease patients treated with BACOP was better than their results for Stage 1V patients7

Because of the significant incidence of known relapse in initial sites of involvement following treatment with combination chemotherapy, it seemed reasonable to us to incorporate radiotherapy in our chemotherapy program for Stage 111 patients. The plan to alternate 4 cycles of chemotherapy and radiotherapy was considered optimal to prevent dissemination and early relapse in initially involved sites. The rationale for this program, which differs from the previously reported Stanford program, was to induce a durable remission with 4 cycles of chemotherapy before initiating the first course of radiotherapy and to ensure maintenance of complete remission by giving an additional 4 cycles of chemotherapy between courses of radiotherapy. In the Stanford study, treatment consisted of three cycles of CAT (Cytoxan, Adriamycin, and 6- thioguanine) followed by total nodal irradiation, and completed with another 3 cycles of CAT.”

Our combined modality program has been relatively well tolerated. The incidence of toxicity has been essen- tially the same as for combination chemotherapy alone, and late sequelae have been minimal. This program re- sulted in a 65% 5-year survival for all patients. Of greater significance was the fact that of those patients who achieved and maintained a complete remission during definitive treatment, 87% remained free of disease, and may be These results are also superior to those

of previously reported studies. Glatstein reported a 47% survival at 5 years for patients with advanced stages of unfavorable histologies. The relapse-free survival was 24%.15

The usual prognostic factors of age, constitutional symptoms, and extranodal presentations did not adversely influence results. However. the presence of so-called bulky abdominal involvement has been shown to be a major adverse factor for complete remission and survival for patients treated with intensive combination chemotherapy programs.” Yet, for our patients treated on a combined modality program. the influence of extensive abdominal disease was negligible in terms of both complete remission and survival. This is attributed to the additive effect of radiotherapy in achieving permanent control of abdom- inal disease. This observation has been made for patients with Stage I and I1 abdominal presentations by us and by other investigators.’.” In reporting on their experience in treating all pathologic subtypes of advanced diffuse lymphomas with a combined modality program consisting of combination chemotherapy and radiotherapy to initial sites of bulky disease, the Manchester group observed better results for Stage I I I than for Stage 1V patients.’ For the 18 patients with a diagnosis of diffuse histiocytic lymphoma, the calculated 5-year survival was 60%.

Another area of bulky involvement, such as the me- diastinum. did not have an adverse influence on results in patients with minimal abdominal involvement. How- ever, the combination of extensive abdominal disease and mediastinal involvement had a dramatic adverse effect on both the complete remission rate and survival. The majority of these mediastinal patients with extensive ab- dominal disease who failed treatment did so during the initial phase of the chemotherapy program. This indicates that a very large tumor burden of extensive disease in both the mediastinum and the abdomen was a limiting factor for the program in a small proportion of the total Stage 111 population. For such patients, our plan is to intensify the chemotherapy section of our program by adding other chemotherapeutic agents such as cytosine arabinoside and by alternating two non-cross-resistant chemotherapy regimens like ifosphamide, methotrexate, and VP-16 (IMVP-16).”’

REFERENCES

I . McKelvey GM. Gottlieb JA. Wilson HE 6’1 a/ . Hydronyldauno- mycin (Adriamycin) in combined chemotherapy in malignant lym- phoma. ( ’ u t w r 1976: 38:1484-1493.

2 . Schein PS. DeVita VT. Hubbard S 1’1 ( I / . Bleomycin. adriamycin. cyclophosphamide. vincristine and prednisolone (BACOP) combination chemotherapy in the treatment of advanced diffuse histiocytic lymphoma. Atrir Iti/cwr M d 1976; 85:41 7-422.

3. Cadman E. Farber L. Berd D c’/ N/. Combination therapy for diflux histiocytic lymphoma that includes antimetabolites. C‘urrc.c,r Trcwr Rc,p 1977: 6I:l109-1 116.

4. Coltman CA. Luce JK. McKelvey EM. Jones SE. Moon TE. Che-

No. 7 STAGE 111 DIFF~JSE LARGE CELL LYMPHOMAS - Vt./asyuez et a/. 1483

motherapy of Non-Hodgkin’s lymphoma: Ten years experience in the southwest oncology group. C’uncw Trcw Rc’p 1977: 61: 1067-1078.

5. Anderson T, Bender RA. Fisher RI v/ ( I / Combination chemo- therapy in non-Hodgkin’s lymphoma: Results of long-term follow-up. Cutiwr ‘/‘r(w/ R r y 1977; 61:1057-1066.

6. Fisher RI. DeVita VT. Hubbard Sbl e/ a/ Improved survival of difuse aggressivc lymphomas following treatment with ProMACE-MOPP chemotherapy. /‘roc, f : i , yh/cv tz / / i . Inn hfw/ .Arn .%tc. ( ’ / / t i oYtc.o/ 1982; l : l 6 l .

7. Fisher RI. DeVita VT, Johnson BL, Simon R. Young RC. Prog- nostic factors for advanced diffuse histiocyl.ic lymphoma following treat- ment with combination chemotherapy, Atn J Mcd 1977; 63: 177- 182.

8. Pettingale KW. The management of generalized grade 2 non- Hodgkin’s lymphomas: Report 18. <‘ / / t i Rudiol 1981; 32:553-555.

9. Crowther D. New approaches to the managment of patients with non-Hodgkin’s lymphoma of high-grade pathology. Br J Cuncijr 198 1 : 43:4 17.

10. Gamble JF. Fuller LM, Butler JJ. 13hullenberger CC. Combined Chemotherapy and radiotherapy for advanced Hodgkin’s disease and reticulum cell sarcoma. S o i c t h M t d .I 197 I ; 64:775-782.

I I . Fuller LM, Gamble JF. Butler JJ (it u/. Team approach to man- agement of non-Hodgkin‘s lymphomas: Past and present. Cuncer Treat Rcp 1977; 6 I : I 137-1 148.

12. Fuller LM. Velasquez WS. Gamble JF. Butler JJ, Martin RG. Shullenberger CC. Combined modality treatment in new approaches to the management ofnon-Hodgkin’s lymphomas. Proceedings of the 12th

International Congress. In: Crowther Xi, ed. Advanccs in Medical On- cology. Research & Education. Oxford and New York: Pergamon Press, 1979; 2 19-228.

13. Kaplan EL. Meier P. Non-parametric estimation from incomplete observations. J / Iwwr 3 u 1 Asvoc 19.58; 53:457-481.

14. Gehan EA. Generalized Wilcoxon test for comparing arbitrarily singly-censored samples. Biomc/riku 1965; 52:203-223.

15. Glatstein EG, Donaldson SS. Rosenberg SA, Kaplan HS. Com- bined modality therapy in malignant lymphomas. C‘uncer Triw Rep

16. McKelvey EM. Moon TE. Curability of non-Hodgkin’s lym- phomas. Cunwr T r u / R t p 1977; 6 I : I I 85- I 190.

17. DeVita VT. Cannellos GP. Chabner B. Schein P, Hubbard SP. Young RC. Advanced diffuse histiocytic lymphoma: A potentially curable disease. Luncc.1 1975; 1:248-250.

18. Gospodarowicz MK. Bush RS. Bergsagel DE, Brown TC. Bulk of tumour as an important prognostic factor in non-Hodgkin’s lym- phomas. Proc Amcr A \’.so(‘ o/ C’unccr Rc\ 1980; 2 1 :463.

19. Lester JN. Fuller LM. Conrad FG er (I/. The roles of staging laparotomy. chemotherapy, and radiotherapy in the management of localized diffuse large cell lymphoma: A study of seventy-five patients. C h t 1 i . o 1982: 49:1746-1753.

20. Cabanillas F, Hagemeister FB, Bodey GP. Freireich EJ. IMVP- 16: An effective regimen for patients with lymphomas who have relapsed after initial comhination chemotherapy, Bhod 1982: 60:693-697.

1977; 61:1l99-l207.