Clinical Staging and Chemoradiation for Early Stage Hodgkin’s Disease
Abstract & Commentary
Synopsis: There is no uniform standard for treating early stage Hodgkin’s disease, the most curable form of the disease. Pathologic staging, including exploratory laparotomy and splenectomy, may precede subtotal lymphoid irradiation (STLI), but can entail morbidity without a proven survival benefit. SWOG and the CALGB recently completed an intergroup trial comparing outcomes for clinically staged patients treated with either STLI vs. adriamycin/vinblastine followed by STLI in an effort to determine whether patholgic staging can be abandoned in early stage patients without harming cure rates.
Source: Press O, et al. J Clin Oncol. 2001;19: 4238-4244.
Laparotomy for staging of hodgkin’s disease was introduced in the 1960s, and is responsible for upstaging 20-30% of patients by detecting occult subdiaphragmatic disease. Splenectomy is typically part of the procedure. Patients without subdiaphragmatic disease may be managed with subtotal lymphoid irradiation, and achieve 10-year disease-free survival rates of 70-80%. Nevertheless, critics of this approach cite morbidity stemming from laparotomy/splenectomy, potential for second malignancies related to radiotherapy, and potential cardiotoxicity linked to radiotherapy as factors which, if improved, may lead to better outcomes.
Press and colleagues reported positive results for their SWOG 9133/CALGB 9391 intergroup phase III randomized trial. In this study, they hypothesized that a "short course of limited chemotherapy" might improve disease-free or overall survival compared with standard STLI while sparing patients the necessity of a laparotomy/splenectomy. Eligible patients had no prior history of laparotomy, radiotherapy or chemotherapy, nonbulky disease, no B symptoms, and no serious comorbidities. Pregnant or lactating females were not eligible. Staging with CT chest/abdomen/pelvis, bone marrow biopsy, and bloodwork including LDH was mandatory. Further staging with gallium scan, lymphangiogram (LAG), and PET scan were strongly recommended but not required. Patients with disease below the diaphragm were excluded.
Press et al eliminated MOPP chemotherapy from consideration for the protocol because of its known toxicities. The other accepted combination regimen, ABVD, was also felt to be too toxic for patients with early stage disease. Press et al bypassed the pulmonary toxicity seen with bleomycin, and the phlebitis and nausea associated with dacarbazine, by paring ABVD to AV, using only adriamycin and vinblastine. Adriamycin was administered intravenously at 25 mg/m2 and vinblastine i.v. at 6 mg/m2 on days 1 and 15, these were done in 3 28-day cycles preceding radiotherapy in the combined modality arm. A 6-week break was built into the protocol prior to starting STLI so that interactions between adriamycin and RT might be minimized. Patients were restaged at the completion of chemotherapy.
The trial was opened in 1989 with an accrual target of 420 patients. Patients were stratified by age (< 35, > 35), sex, number of involved sites, histology, presence of high-neck disease, and use of a lymphangiogram for staging. Pathology slides were centrally reviewed. By the time the study was prematurely closed for accrual in April 2000, 348 patients had been randomized to STLI alone vs. combined chemotherapy followed by STLI (combined modality therapy, [CMT]). Radiotherapy parameters were centrally reviewed at the start of RT. STLI was given at 1.8-2 Gy per fraction for 20 fractions for a total dose of 36-40 Gy with 4-10 MV photons. Mantle and paraaortic/spleen fields were treated sequentially.
Median age was 32 years (r, 17-85). Median follow-up was 3.3 years (r, N/A). There was a slight male preponderance. Two thirds of patients in each arm had disease in > 2 lymph nodes. An unspecified number of patients had no residual measurable disease following biopsy. Eighty percent of patients in each arm had nodular sclerosing or lymphocyte predominant histologies. One quarter of the study population underwent LAG. Overall, the 2 study arms demonstrated comparable demographics.
The 2 planned interim analyses were performed on an intention-to-treat basis. Of the 348 patients enrolled, 22 were ineligible for evaluation, leaving 161 in the STLI arm and 165 in the CMT arm. Seven patients were not treated according to their randomization, and 9 did not complete their courses of therapy. Overall, there were 295 fully assessable patients (85%), including 147 in the STLI arm and 148 in the CMT arm. The estimated 3-year disease-free survival was 81% in the STLI arm, and 94% in the CMT arm (P < .001). Overall survival was not significantly different between arms. There was 1 treatment-related death in each arm, with 6 more deaths in the STLI arm (4.3%) and 3 more deaths in the CMT arm (1.8%) (P = NS). There were significantly more Grade 3 and 4 toxicities in the CMT arm (P = .004). Among the 31 relapsed patients in the STLI arm, 12 (39%) relapsed in-field. Among the 9 relapsed CMT patients, 4 (44%) relapsed in-field.
Press et al reported that SWOG and CALGB closed their study early once the second interim analysis results were completed. They based their decision on what they felt was the markedly superior disease-free survival rate observed in the CMT arm. They concluded that their trial, the largest American study of early stage Hodgkin’s disease to date, decisively demonstrated that pathologic staging can be sacrificed while maintaining excellent disease-free and overall survival. They stated that this trial conclusively demonstrated that adriamycin/vinblastine + STLI is well tolerated, and confers a superior DFS rate in comparison to STLI alone for stages IA and IIA Hodgkin’s disease with minimal toxicities.
Comment by Edward J. Kaplan, MD
Press et al discussed 3 cons related to radiotherapy at the beginning of their report, and those were the incentive to perhaps find a better way. Those were: morbidity from laparotomy/splenectomy; second primary tumors caused by RT; and cardiac toxicity caused by RT. Their protocol schema addressed only the first of these issues.
One question that comes to mind immediately is whether there was selection bias favoring the CMT arm. Usually, patients who receive STLI do so after they have been pathologically staged. This was not the case in the SWOG trial. It is likely that 20-30% of the STLI patients had occult subdiaphragmatic disease, and ordinarily would not have been treated with STLI alone. Furthermore, while PET scan staging was recommended, we are not given details about whether such scanning was equitably distributed between the 2 arms. We do know that patients were stratified by LAG, but not by PET. While total lymphoid irradiation covers the mantle and inverted-Y fields, including the entire spleen if intact, STLI omits the pelvis, and PET scanning can be useful for demonstrating disease in the pelvis as well as elsewhere. In addition, it is unclear what, if anything, was done with the restaging information obtained after CMT patients completed their chemotherapy.
In examining the data presented by Press et al, they indicated that 99 patients (34%) had responses that were either unconfirmed or not assessable. They then stated that there was a significant difference in the overall response rate favoring the CMT arm (P =.004) without disclosing their denominator. Press et al go on to rationalize the lack of difference in overall survival between arms by suggesting that it is a reflection of the success of second-line salvage therapies. However, no data at all were presented in that regard.
Press et al stated that "our analysis suggests that relapses are diminished both within and outside the irradiated field." Despite this assertion, their results do not support their contention. While 39% of relapses in the STLI arm were in-field, 44% in the CMT arm were in-field. P values were not provided.
While I agree with the goals of the study, its design seems to have a built-in handicap for the STLI arm based on the use of clinical staging. Perhaps if all patients underwent PET scan evaluation first, omitting surgery might not be as important an issue.1,2 Alternatively, the trial could have compared pathologically staged patients treated with STLI to clinically staged patients treated with chemoradiation. Of course, few studies with a surgical and nonsurgical arm have been completed.
Some radiation oncologists would suggest that the doses of radiation used in the trial were too high. A recently published randomized trial from Europe compared 30 Gy to 40 Gy for treating uninvolved nodal areas, and found no difference in DFS with the use of lower doses.3 Many clinicians feel that 30 Gy is an appropriate dose for involved or uninvolved areas. Certainly, lower doses would lower toxicities.
Press et al are right in adding that further follow-up is necessary before survival, relapse, and salvage patterns are known. It should also be borne in mind that, as Hoppe from Stanford has pointed out, complications of therapy can take more than 15 years to manifest.4 The addition of adriamycin to STLI may lead to increased cardiac toxicity far off in the future, possibly lowering overall survival, and thereby rendering a combined modality approach less attractive.
References
1. Schoder H, et al. J Nucl Med. 2001;42:1139-1143.
2. Buchmann I, et al. Recent Results Cancer Res. 2000;156:78-89.
3. Duhmke E, et al. J Clin Oncol. 2001;19:2905-2914.
4. Hoppe R. Ann Oncol. 1997;8(Suppl 1):115-118.