Kaplan [24] and others have shown that a radiation dose of 2000 to 2500 cGy is effective in preventing recurrence in all but 25% to 30% of patients with clinical disease when radiation is used as the sole treatment modality [24]. This relatively low dose of radiation might be even more effective if only subclinical disease was being treated. Further, using low-dose radiation after chemotherapy might decrease some of the morbidity associated with the combined use of both modalities at full doses. Early nonrandomized studies by Prosnitz and colleagues [25], in which the MVPP regimen (nitrogen mustard, vincristine, prednisone, and procarbazine) was administered with low-dose involved field radiation, showed only a 10% relapse rate when stage III or IV patients had achieved complete response with chemotherapy before receiving radiation.

In 1978, we initiated a randomized trial for patients with stage III or IV Hodgkin disease that was designed to test the efficacy of low-dose involved field radiation after complete remission induction with chemotherapy. The MOP-BAP (nitrogen mustard, vincristine, procarbazine, bleomycin, doxorubicin, and prednisone) chemotherapy regimen was used as induction treatment because this regimen had been associated with a complete response rate of 77% [2]. Complete responders to six cycles of MOP-BAP were randomly assigned to a radiation dose of 2000 cGy to initially involved sites or to no further treatment.

We determined the remission duration and relapse-free and overall survival rates for the entire group of patients with advanced Hodgkin disease as well as for several major subsets of patients.

Methods

Top Methods Study Design Statistical Analysis Results Survival Discussion Author & Article Info References

Previously untreated patients (n = 564) with clinical or pathologic evidence (or both) of stage III or IV Hodgkin disease were registered for induction with MOP-BAP chemotherapy between June 1978 and September 1988. Pathology slides were reviewed by the Southwest Oncology Group Pathology Review Committee. A final review was made by the Lymphoma Central Repository. Patients who were known to be positive for human immunodeficiency virus or who had a clinical diagnosis of the acquired immunodeficiency syndrome were excluded from analysis. The pathology review is now complete for 95% of the patients who entered this study. Thirty-four patients (6%) were ineligible, primarily because they had non-Hodgkin lymphoma after pathology review.

The Ann Arbor Staging Classification was used [26]. Staging requirements included bone marrow aspiration and biopsy; renal, heart, and lung function studies; and staging laparotomy or radiographic evaluation of the abdomen with either computerized tomography or lymphangiography or both. Patients with any mass 6 cm or more in size were designated as having bulky disease. Flow sheets, prestudy forms, and radiographic reports were reviewed on all patients who achieved complete response to make the assessment of bulky disease. Liver biopsy was required in patients with stage IIIB or IV disease unless the patient had clinical liver involvement or it was medically contraindicated. Clinical liver involvement was defined as an enlarged liver on physical examination or computed tomography together with increased levels of at least one liver enzyme other than alkaline phosphatase or lactate dehydrogenase. Clinical spleen involvement was defined as a palpable spleen on physical examination or an enlarged spleen on computed tomography with or without filling defects.

Response Definitions

A partial response was defined as a 50% or greater decrease in the sum of the products of the largest diameter and its perpendicular for 4 weeks or more. Patients with questionable residual disease were classified as partial responders. For patients with minimal residual disease detected by computerized tomography scanning after six cycles of chemotherapy, designation of partial or complete remission was left to the discretion of the individual investigator. In general, however, patients with minimal residual disease were classified as partial responders. If the residual mediastinal mass on the computerized tomographic scan was less than 3 cm or the residual peripheral nodal mass was 1.5 cm or less, partial responders were classified as having minimal residual disease. Complete response was defined as disappearance of all clinical evidence of disease for 4 or more weeks.

Treatment

Chemotherapy

The MOP-BAP chemotherapeutic regimen consisted of nitrogen mustard (6 mg/m²) on day 1, bleomycin (2 mg/m²) and vincristine (1 mg/m²; maximum dose, 2 mg) on days 1 and 8, doxorubicin (30 mg/m²) on day 8, and prednisone (100 mg) and procarbazine (100 mg/m²) on days 2 to 7 and 9 to 12. Courses were repeated every 28 days if the absolute granulocyte count was 1500 cells/mm³ or more. A total of 6 cycles was administered. Prednisone was given only during the first and fourth cycles.

Initial doses of MOP-BAP were 50% of those listed above if the patient was older than 65 years or had bone marrow involvement with leukopenia. Dose escalation in subsequent courses was encouraged. Initial doxorubicin doses were also decreased by 50% to 75% for serum bilirubin levels greater than 1.5 mg/dL or increases in liver enzyme levels of more than threefold the normal level or both. Subsequent drug doses were reduced for severe myelosuppression or delayed hematopoietic recovery as in previous Southwest Oncology Group studies [2] using this regimen.

The ratio of actual/planned full-dose chemotherapy for each of the 5 drugs (nitrogen mustard, bleomycin, procarbazine, vincristine, and doxorubicin) was calculated for the 530 eligible patients. The average percentage was used to describe the amount of chemotherapy given. The average for the 5 drugs was 85%. The average number of courses was 5.6. The average actual/planned ratio for each drug was 86% for nitrogen mustard, 84% for doxorubicin, 82% for procarbazine, 91% for bleomycin, and 81% for vincristine. Sixty-six percent of patients achieving complete response received more than 85% of the planned induction chemotherapy.

Radiation Therapy for Complete Responders

All patients were seen in consultation by a radiation oncologist before induction chemotherapy was started, and all clinical and pathologic sites of disease were mapped. If the patient was later given radiation, all initially involved areas were included in the treatment ports with the exception of bone marrow. Patients with previously involved nodal sites received 2000 cGy in 150-cGy fractions, those with previously involved liver sites received 1500 cGy in 150-cGy fractions, those with previously involved spleen sites received 1500 cGy in 125-cGy fractions, and those with previously involved lung sites received 1000 cGy in 100-cGy fractions. Kidney blocks were used when necessary, and a spinal cord block was used when the dose to the cord had reached 2000 cGy. Patients only received radiation to those areas identified as being clinically or pathologically involved with Hodgkin disease. Radiation was started, 6 weeks after day 1 of the sixth MOP-BAP cycle, if the leukocyte count was more than 3000 cells/mm³ and if the platelet count was more than 100 000/mm³. A 3-week rest was recommended between radiation of large volumes or major lymph node areas.

Supervoltage radiation of at least 2 MV or cobalt-60 was required. Port films, dose calculations, and treatment records were reviewed by the Quality Assurance Center, the Radiologic Physics Center, and the Medical and Radiation Oncology Coordinators. Failure to give any radiation to a previously involved site or concomitant administration of radiation and chemotherapy was considered a major radiation violation. Dose infractions and failure to complete radiation for any reason were considered minor radiation violations. Ninety-six percent of patients receiving radiation have had their records evaluated by the Radiologic Physics Center in Houston, Texas, and the Radiation and Medical Oncology Study Coordinators. All eligible patients who achieved complete response and were randomized were included in comparisons of remission duration, relapse-free survival, and survival regardless of whether a major or minor radiation violation had occurred.

Patients randomized to no further treatment received no radiation or chemotherapy after six cycles of MOP-BAP until they relapsed. Planned follow-up intervals in patients randomized to receive low-dose radiation were identical to those for patients randomized to no further treatment. Treatment at the time of relapse was at the discretion of the individual investigator.

Study Design

Top Methods Study Design Statistical Analysis Results Survival Discussion Author & Article Info References

When the study was initially opened, patients in complete response were randomized after six cycles of MOP-BAP and restaging to one of three possible groups: no further therapy, levamisole for 2 years, or low-dose radiation to previously involved sites of disease. The study design was changed in 1982 because of lower-than-anticipated patient accrual. The levamisole arm was eliminated, and patients were randomized before the start of induction therapy to receive low-dose radiation or no further treatment if they had a complete response after six cycles of MOP-BAP. After documentation of the response by restaging, investigators were required to register the patient to the previously assigned treatment group.

Of the 530 eligible patients who started MOP-BAP induction, 322 (61%) achieved a complete response after 6 cycles of MOP-BAP chemotherapy. Thirty-two patients refused to be randomly assigned, and 12 were randomized to levamisole. The remaining 278 patients who had a complete response were randomized to receive either low-dose involved field radiation or no further treatment. Unfortunately, 16% of the patients did not receive the assigned treatment. Failure to receive the correct treatment was unevenly distributed between the two groups. Twelve of 143 patients randomized to no further treatment dropped out of the study before the secondary registration; they received variable treatments and follow-up. An additional patient randomized to no further treatment actually received low-dose radiation. Of the 135 patients randomized to low-dose radiation, 31 (23%) actually received no further treatment. Thus, 130 patients were randomized to and actually received no further treatment, and 104 patients received low-dose radiation (Figure 1). The five-year remission-duration and survival estimates were 61% and 79%, respectively, for the 44 patients who were randomized to but did not receive the assigned treatment.

View larger version (17K): [in this window] [in a new window]   Figure 1. Flow chart of all eligible patients treated with MOP-BAP chemotherapy. MOP-BAP = nitrogen mustard, vincristine, procarbazine, bleomycin, doxorubicin, and prednisone.

Patient Characteristics and Response to Induction Chemotherapy

Of the 530 previously untreated eligible patients entering induction, 49 (9%) were classified as Hodgkin stage IIIA₁, 97 (18%) as stage IIIA (₂), 175 (33%) as stage IIIB, 38 (7%) as stage IVA, and 171 (32%) as stage IVB. The median age was 32 years (range, 14 to 86 years). Sixty-one percent (322 patients) of the 530 patients achieved a complete response during MOP-BAP induction chemotherapy. This percentage includes all eligible patients regardless of whether they completed induction chemotherapy or restaging. Eighty-two percent of these 530 patients had poor prognostic factors for achieving complete response (B symptoms, two or more non-nodal sites of disease, stage IV disease, age greater than 60 years). The complete response rate increased from 322 of 530 patients (61%) to 427 of 530 patients (80%) after protocol-directed low-dose radiation for the 168 patients in partial remission after 6 cycles of MOP-BAP. Radiation for partial responders was similar to that for complete responders randomized to low-dose radiation except that patients with residually involved sites received 2500 cGy. Eighty-eight percent of the 114 patients with a partial remission after chemotherapy who received radiation achieved a complete response with low-dose involved field radiation. Preliminary results of the effect of consolidation radiation on partial responders were previously reported [27]. The median length of follow-up for the 368 eligible patients still alive was 8.1 years (maximum, 14.5 years) after the start of induction therapy.

Prognostic Factors

Potential prognostic factors studied for influence on relapse or survival after complete response were sex, histology, age, symptoms, stage (III compared with IV), bone marrow involvement, bulky disease, two or more non-nodal sites of disease, receipt of 85% or less of planned chemotherapy, more than three cycles of MOP-BAP required to achieve complete response, and reduced initial chemotherapy doses because of advanced age or bone marrow involvement with leukopenia. No differences were seen in the distribution of potential adverse prognostic factors between patients randomized to and those who actually received study treatment after chemotherapy. Likewise, no differences were noted in the distribution of adverse factors between patients randomized to receive low-dose radiation or to no further treatment. (Table 1).

Statistical Analysis

Top Methods Study Design Statistical Analysis Results Survival Discussion Author & Article Info References

The primary comparisons between low-dose radiation and no further therapy for remission duration, relapse-free survival, and survival included randomized eligible patients achieving a complete response with MOP-BAP chemotherapy; the comparisons are thus based on intent to treat. Remission duration was measured from date of complete response until relapse, death, or date of last contact. For remission-duration estimates, patients who died without relapse were censored at the time of death. Relapse-free survival was measured from the date of complete response until relapse, death, or date of last contact. Relapse or death from any cause was considered a failure. Survival time after complete response was measured from documented complete response until death or last contact. No deaths were censored in the calculation of relapse-free and overall survival. Survival after relapse was calculated from the date of relapse to the date of death from any cause.

Remission-duration, survival, and relapse-free survival rates were estimated using the method of Kaplan and Meier [28]. Evaluation of prognostic factors and treatment comparison was made using log-rank tests [29] and Cox partial-likelihood score tests [30]. The Cox regression model was used in multivariate comparisons [30]. Prognostic variables evaluated included age (>60 years), sex, more than three cycles of chemotherapy required to achieve complete response, involved bone marrow, stage IV disease, bulky disease, B symptoms, reduced initial chemotherapy dose, receipt of more than 85% of planned chemotherapy, and histology-treatment interaction. The histology-treatment interaction was tested by comparing a model with separate histologic and treatment effects to a model with an additional covariate specifying the interaction term. Relative risk estimates were also made using the Cox model.

All eligible patients were included in the calculations of response rates, including patients with major protocol deviations or early death. The median length of follow-up was calculated using all eligible patients last known to be alive. The response rate and pretreatment patient characteristics were compared using the Fisher exact test. Significance tests were not adjusted for multiple comparisons. Data analyses are based on follow-up information from the Southwest Oncology Group Statistical Center as of 1 April 1993.

Results

Top Methods Study Design Statistical Analysis Results Survival Discussion Author & Article Info References

Remission Duration

When all 278 randomized patients who achieved a complete response were compared, the 5-year remission-duration estimate of 79% for low-dose radiation was similar to the 68% estimate for no further treatment (P = 0.09 for the difference of 11%).

We next examined the effect of low-dose radiation on remission duration using a multivariate Cox model (Table 2). Patients with bulky disease had an increased risk for relapse of 1.9 times that of patients with nonbulky disease (P = 0.008). Patients receiving less than 85% of the induction chemotherapy had an increased risk for relapse of 1.7 times that of patients receiving more than 85% of the induction chemotherapy (P = 0.02). An interaction between the histologic subtype and randomized treatment was found (P = 0.02). For patients with nodular sclerosis histology who were randomized to no further therapy, the relative risk estimate was 2.6 times (95% CI, 1.4 to 4.8) that of patients with nodular sclerosis histology who were randomized to low-dose radiation (Table 2). No other statistically significant interactions between covariates were found (P > 0.05). Thus, B symptoms, age, complete response after 3 cycles of induction therapy, bone marrow involvement, stage, number of non-nodal sites of disease, sex, and whether the patient received reduced initial doses of induction chemotherapy were all not statistically significant in the multivariate model of remission duration.

For patients with nodular sclerosis histology, the 5-year remission-duration estimate was 82% for the 86 patients randomized to low-dose radiation and 60% for the 83 patients randomized to no further treatment (P = 0.002). For patients with bulky disease and nodular sclerosis histology, the 5-year estimate was 76% for the 42 patients who were randomized to low-dose radiation compared with an estimate of 46% for the 32 patients who were randomized to no further treatment (P = 0.006). For patients with nonbulky disease and nodular sclerosis histology, the five-year remission-duration estimate was 88% for the 44 patients randomized to low-dose radiation and 68% for the 51 patients randomized to no further treatment (P = 0.06) (Figure 2). For the patients with nodular sclerosis histology, remission duration appeared improved with low-dose radiation whether bulky disease was present or not. No statistically significant differences were seen in remission duration between low-dose radiation and no further treatment (P > 0.20) for the subset of patients with stage IV disease with or without bone marrow involvement.

View larger version (24K): [in this window] [in a new window]   Figure 2. Remission duration after complete response for patients with nodular sclerosis histology. Patients were randomized to either no further treatment or low-dose involved field radiotherapy, stratified by the size of the initial disease. LDIF = low-dose involved field radiation; Rx = treatment.

Forty-four (16%) of the randomized patients who were complete responders did not receive the assigned treatment. A secondary univariate analysis of the 234 patients of all histologic types who actually received the randomized treatment showed an improvement in remission duration for those receiving low-dose radiation compared with those receiving no further treatment (5-year estimates, 85% compared with 67%; P = 0.002). A multivariate Cox model including the same covariates as above indicated (after adjusting for histologic subtype, bulky disease, and the amount of induction chemotherapy received) that the risk for relapse for patients receiving no further therapy was 1.8 times (CI, 1.0 to 3.1) that of those receiving low-dose radiation.

Relapse Patterns

The relapse patterns for patients receiving low-dose radiation and those receiving no further treatment were different. Seventy-nine percent of the 42 relapsing patients in the no further treatment group had recurrences only in areas of previous clinical involvement, whereas only 47% of the 17 relapsing patients who had low-dose radiation had recurrences in areas of previous involvement. Fifty-six percent of relapsing patients had a major radiation protocol violation, and 12.5% had a minor violation. Only 10.5% of relapses in the no further treatment group were in exclusively new sites, whereas 47% of relapses in the low-dose radiation group were in exclusively new sites.

At a median follow-up time of 8 years, the percentage of patients with nodular sclerosis who received no treatment according to their randomized assignment and who relapsed was twice that of patients with other histologic types of Hodgkin disease (41% compared with 21%). However, for patients receiving low-dose radiation, the relapse rates for those with nodular sclerosis and other histologic subtypes were similar (17% and 16%, respectively) (Table 3).

Relapse-free Survival

Univariate analysis showed that the probability of relapse-free survival (the probability of being alive without relapse) was not significantly different between randomized treatment groups. The 5-year relapse-free survival estimate was 74% for the low-dose radiation group and 66% for the no further treatment group (P > 0.2). The 5-year relapse-free survival estimate for patients with nodular sclerosis histology randomized to low-dose radiation was 77% compared with 56% for those randomized to no further treatment (P = 0.01) (Figure 3). The same factors that were statistically significant for the remission-duration model were also significant for the relapse-free survival model. In addition, patients with B symptoms had a shorter relapse-free survival, with a relative risk estimate of 1.9 (CI, 1.2 to 3.0; P = 0.006) when compared with patients without B symptoms. The other relative risk estimates are similar in magnitude to those from the remission-duration model. An interaction was found between histologic subtype and randomized treatment (P = 0.004). Patients with nodular sclerosis histology randomized to no treatment had a relative risk estimate of 2.0 (CI, 1.2 to 3.4) for relapse or death compared with those patients randomized to low-dose radiation.

View larger version (19K): [in this window] [in a new window]   Figure 3. Proportion of patients with nodular sclerosis alive without disease. Patients were randomized to either no further treatment or low-dose involved field radiotherapy (P =0.01). RT = radiotherapy.

Survival

Top Methods Study Design Statistical Analysis Results Survival Discussion Author & Article Info References

View larger version (18K): [in this window] [in a new window]   Figure 4. Proportion of patients with nodular sclerosis histology who were alive after complete response. Patients were randomized to either no further treatment or low-dose involved field radiotherapy (P > 0.2). The number of patients alive at specified time points is noted. CR = complete response; RT = radiotherapy.

Univariate analysis of survival showed that reduced starting doses of MOP-BAP chemotherapy, B symptoms, and age (>60 years) were unfavorable prognostic factors for survival. However, multivariate Cox regression analysis showed only that reduced starting doses of chemotherapy were an independent factor, with a relative risk estimate of 4.6 (CI, 2.3 to 9.1; P < 0.001) compared with receipt of full initial doses of MOP-BAP chemotherapy. Survival was similar for both randomized treatment groups using the multivariate model (P = 0.23).

The lack of a survival advantage for low-dose radiation could be secondary to several factors: 1) a lower salvage rate with secondary treatment after relapse in patients receiving low-dose radiation; 2) shorter survival of patients receiving low-dose radiation who relapsed; and 3) a larger number of patients randomized to low-dose radiation who died, without having relapsed, from causes other than Hodgkin disease.

Treatment after relapse was more effective in patients relapsing in the no further treatment group than in those in the low-dose radiation group. Seventeen of the 42 (40%) patients who were in the no further treatment group are alive without evidence of disease after secondary treatment. Only 4 of 17 patients (24%) who relapsed after receiving low-dose radiation are alive without disease. Survival after relapse was shorter for those relapsing after low-dose radiation (median, 23 months) compared with those relapsing after no further treatment (median, >60 months) (P = 0.006). Median survival after relapse for the 11 patients with nodular sclerosis who received low-dose radiation was 51 months and was 100 months for the 30 patients receiving no further treatment (P = 0.05).

The proportion of patients dying without relapse was higher in patients who received low-dose radiation (11 of 104) than in those who received no further treatment (4 of 130) (P = 0.03). Causes of death without relapse in the group who received low-dose radiation were infection in 2 patients, second malignancy in 8 patients, and cardiovascular problems in 1 patient. In the group who received no further treatment, 2 patients died of infection and 2 died of second malignancies. In the 44 patients who were randomized but did not receive the assigned treatment, 1 patient died of cardiovascular problems, 1 patient died of infection, and 2 patients died in accidents.

Acute Toxicity

Treatment was well tolerated. During induction chemotherapy, we observed an 8.5% incidence of life-threatening toxicity of any kind and a 2.5% fatality rate. During radiation consolidation, we observed a 2.9% incidence of life-threatening toxicity of any kind and no fatalities. Forty-eight percent of 104 patients had some degree of leukopenia during radiation. In 10% of patients, leukopenia was severe (1000 to 2000 cells/mm³), but fatal or life-threatening leukopenia during treatment was not observed. Infections developed in 8.6% of patients who received radiation. Infections were moderate in degree in 2.8% of patients and were severe in 5.6%. Moderate infection was seen in the same period in 2.4% of patients receiving no treatment after chemotherapy. During radiation, thrombocytopenia (<50 000/mm³) was seen in 4.8% of patients, but life-threatening thrombocytopenia (<25 000/mm³) was seen in only 2.8% of patients. Cardiac toxicity was not observed, and only one instance of mild pulmonary toxicity was noted. Only 5.7% of patients had nausea or vomiting.

Late Toxicity

The predominant delayed complication in patients receiving radiation was the development of second malignancies. Among the 104 patients with a complete response who received 6 cycles of MOP-BAP and were randomized to and received low-dose radiation, 9 second malignancies were noted (2 patients with leukemia, 2 patients with lung cancer, 3 patients with lymphoma, 1 patient with breast cancer, and 1 patient with colon cancer). Two additional second malignancies occurred in the group of 31 patients who were randomized to low-dose radiation but who did not receive it (1 patient with lung cancer and 1 with lymphoma). Among the 130 patients who were randomized to and received no further treatment, 4 second malignancies were noted (2 patients with lung cancer, 1 patient with esophageal cancer, and 1 patient with lymphoma). The second malignancy rate was 4.6% for all 322 eligible patients entering complete response during MOP-BAP induction, regardless of whether they went on to receive consolidation. The second-malignancy rate was 8.6% for the 104 patients who were randomized to and received radiation and was 3.1% for the 130 patients who were randomized to and received no further treatment (P = 0.08) (Table 4).

Alkylating agents and gonadal radiation can adversely affect fertility. No clear differences were reported for successful pregnancies between the no further treatment group and the low-dose radiation group, but we did not prospectively study fertility. Three of 27 women (11%) younger than 40 years who did not receive radiation had successful pregnancies reported on flow sheets compared with 4 of 29 women (14%) younger than 40 years who did receive low-dose radiation. All reported successful pregnancies occurred in women younger than 26 years.

Discussion

Top Methods Study Design Statistical Analysis Results Survival Discussion Author & Article Info References

We tried to determine if low-dose involved field radiation could prevent or substantially decrease the 35% to 40% relapse rate consistently seen after chemotherapy alone and if it could improve survival [1-4]. No statistical improvement was noted in remission duration or survival in all randomized patients of all histologic types.

This study suggests that remission duration and relapse-free survival are improved with low-dose involved field radiation in patients with nodular sclerosis histology. The decrease in the relative rate of relapse noted for patients with nodular sclerosis histology who were randomized to low-dose radiation was greatest in those patients who also had bulky disease. Low-dose radiation did not appear to improve remission-duration outcome in the patients who had stage IV disease with or without bone marrow involvement.

The presence of bulky disease and nodular sclerosis histology are unfavorable prognostic factors for remission duration [22, 31] and predicted (in our study) an increased relapse rate in patients with a complete response after six cycles of MOP-BAP. Bulky disease is often associated with the nodular sclerosis histologic subtype [32-34], and thus a portion of the favorable effect of low-dose radiation on patients with nodular sclerosis histology may be related to tumor size.

This is the first large randomized study to suggest that low-dose involved field radiation after chemotherapy induction in patients with advanced Hodgkin disease improves remission duration. However, because the beneficial effects in all randomized patients occurred in subgroups of patients and because 16% of the randomized patients failed to receive the assigned treatment, results must be interpreted with caution. Thus, a confirmatory study is needed. The observed decrease in relapse in patients with a nodular sclerosis histologic subtype was not translated into improved survival. A proportion of patients with Hodgkin disease who relapse are successfully salvaged with secondary treatment [5-14]. Even if not successfully salvaged, patients may have a prolonged time to death after relapse, as occurred in this study. Thus, adverse prognostic factors for remission duration may not be the same as those predicting short survival at our median follow-up time of 8 years. The lack of a survival advantage may also be due to the relatively few deaths (only 20% of randomized complete responders have died), shorter time to death after relapse in the low-dose radiation group, and more deaths without relapse in the radiation group compared with the no further treatment group.

Our study supports the earlier nonrandomized work of Prosnitz and colleagues [25] in which 1500 to 2500 cGy were sandwiched between three cycles of MVPP (nitrogen mustard, vinblastine, vincristine, prednisone, and procarbazine). In Prosnitz's study [35], a relapse rate of only 16% was noted in patients with stages III and IV who received MVPP and low-dose radiation after 15 years. A Canadian multicenter study [36] showed a decreased relapse rate in nodal sites when radiation therapy was given to the mantle and upper abdominal fields after MOPP (nitrogen mustard, vincristine, prednisone, and procarbazine) chemotherapy when compared with chemotherapy alone. Overall, remission duration was not improved [36]. The randomization scheme in the Canadian study was complex, the number in each of the subgroups was relatively small, and 13% of the patients did not receive the assigned radiation treatment. Consequently, definite conclusions about the worth of combined modality treatment in the Canadian study cannot be made [36].

A Southwest Oncology Group study [37] of stage III patients randomized to MOPP-bleomycin and high-dose radiation or to MOPP-bleomycin alone showed improvement in relapse-free survival (P = 0.05) with combined modality therapy in patients with a nodular sclerosis histologic subtype. Five-year relapse-free and overall survival rates for 6 cycles of MOP-BAP and low-dose radiation are equal to, or better than, other popular 4- to 8-drug regimens, such as MOPP-ABVD (nitrogen mustard, vincristine, prednisone, procarbazine, doxorubicin, bleomycin, vinblastine, and dacarbazine) [2, 3, 38-42] (Table 5).

The decreased relapse rate seen with combined modality therapy may not result in survival advantages unless the excess late mortality from second malignancies can be decreased [43-48]. However, late mortality is not limited to combined modality therapy. As emphasized in two recent articles [49, 50], when death because of all causes was considered, MOPP-ABVD may not have a survival advantage compared with MOPP. Likewise, MOPP-CABS (CABS = carmustine, doxorubicin, bleomycin, and streptozocin) had no survival advantage when compared with MOPP alone. For the latter treatment, an excess number of second malignancies, as well as a shorter time to death after relapse, was noted in patients receiving MOPP-CABS compared with patients receiving MOPP alone [49].

The chemotherapeutic regimen of ABVD is as successful in producing complete remissions as MOPP or MOPP-ABVD. The use of ABVD alone also appears to be associated with better preservation of fertility and with less potential for secondary leukemia development than MOPP or MOPP-ABVD [51]. However, ABVD combined with radiation may produce an unacceptable degree of heart and lung toxicity, and the incidence of solid tumors is not decreased when compared with a MOPP-based regimen plus radiation [52-55].

Chemotherapy and radiotherapy protective agents (such as WR-2721, selenium, zinc, and glucan) used alone or in combination might protect normal tissues during treatment [56-62]. After chemotherapy or radiation or both, agents such as ß-carotene might be considered for persons at high risk for second malignancy [63, 64].

Our study suggests that remission duration may be improved in a subset of patients with stage III or IV Hodgkin disease who are at high risk for relapse (those with the nodular sclerosis and bulky disease histologic subtypes) when low-dose involved field radiation is administered after complete remission induction with MOP-BAP chemotherapy. After 8.1 years of follow-up, no survival advantage was noted for low-dose radiation.

Abbreviations

BCNU: N, N-bis (2-chloroethyl)-N-nitrosourea (carmustine)

BCVPP: BCNU, cyclophosphamide, vinblastine, prednisone, and procarbazine

CABS: carmustine, doxorubicin, bleomycin, and streptozocin

MOP-BAP: nitrogen mustard, vincristine, prednisone, bleomycin, doxorubicin, and procarbazine

MOPP-Bleo: nitrogen mustard, vincristine, prednisone, procarbazine, and bleomycin

MOPP: nitrogen mustard, vincristine, prednisone, and procarbazine