Doxorubicin Exposure and Breast Cancer Risk in Survivors of Adolescent and Adult Hodgkin Lymphoma

PURPOSE Female Hodgkin lymphoma (HL) survivors treated with chest radiotherapy (RT) at a young age have a strongly increased risk of breast cancer (BC). Studies in childhood cancer survivors have shown that doxorubicin exposure may also increase BC risk. Although doxorubicin is the cornerstone of HL chemotherapy, the association between doxorubicin and BC risk has not been examined in HL survivors treated at adult ages. METHODS We assessed BC risk in a cohort of 1,964 female 5-year HL survivors, treated at age 15-50 years in 20 Dutch hospitals between 1975 and 2008. We calculated standardized incidence ratios, absolute excess risks, and cumulative incidences. Doxorubicin exposure was analyzed using multivariable Cox regression analyses. RESULTS After a median follow-up of 21.6 years (IQR, 15.8-27.1 years), 252 women had developed invasive BC or ductal carcinoma in situ. The 30-year cumulative incidence was 20.8% (95% CI, 18.2 to 23.4). Survivors treated with a cumulative doxorubicin dose of >200 mg/m2 had a 1.5-fold increased BC risk (95% CI, 1.08 to 2.1), compared with survivors not treated with doxorubicin. BC risk increased 1.18-fold (95% CI, 1.05 to 1.32) per additional 100 mg/m2 doxorubicin (Ptrend = .004). The risk increase associated with doxorubicin (yes v no) was not modified by age at first treatment (hazard ratio [HR]age <21 years, 1.5 [95% CI, 0.9 to 2.6]; HRage ≥21 years, 1.3 [95% CI, 0.9 to 1.9) or chest RT (HRwithout mantle/axillary field RT, 1.9 [95% CI, 1.06 to 3.3]; HRwith mantle/axillary field RT, 1.2 [95% CI, 0.8 to 1.8]). CONCLUSION This study shows that treatment with doxorubicin is associated with increased BC risk in both adolescent and adult HL survivors. Our results have implications for BC surveillance guidelines for HL survivors and treatment strategies for patients with newly diagnosed HL.


BACKGROUND
Female Hodgkin lymphoma (HL) survivors treated with chest radiotherapy (RT) at a young age have a strongly increased risk of subsequent breast cancer (BC). 1,2Previous studies have shown that BC risk is higher after treatment at a younger age, with higher radiation doses and/or larger radiation volumes.][10] This reduction of radiation-associated BC risk after gonadotoxic treatment has been attributed to a protective effect of early menopause.][16] However, a previous study from our group among 5-year HL survivors treated between 1965 and 2000 did not demonstrate a lower BC risk in more recently treated patients. 17veral studies in CCS populations reported increased BC risk after exposure to doxorubicin with and without chest RT, [18][19][20][21][22] with data suggestive of dose-dependent increase in BC risk.
To the best of our knowledge, the association between doxorubicin exposure and BC risk has not yet been examined in HL survivors treated at adult ages.If doxorubicin exposure at adult ages were associated with BC risk, this could have implications for follow-up care of HL survivors, as well as for treatment strategies for newly diagnosed patients.Therefore, we assessed the effect of doxorubicin exposure on BC risk using a large cohort of female 5-year HL survivors, treated between 1975 and 2008, with a broad age range at the start of treatment.

Study Population and Data Collection
This multicenter study cohort comprised 1,964 female 5-year survivors who had HL as their first malignancy and were treated in 20 Dutch centers (eight university medical centers, nine general hospitals, and three RT centers) in the Netherlands at age 15-50 years, between 1975 (when doxorubicin was introduced in HL treatment) and 2008.
4][25] Cumulative doxorubicin and procarbazine doses were estimated on the basis of the number of cycles of prescribed chemotherapy regimens using standard dosages for each regimen.
For follow-up years until 1989, information on subsequent primary malignancies was collected directly from medical records and from questionnaires sent to general practitioners (questionnaire response until 1989: 96%). 26For follow-up from 1989 onward, information on subsequent BC was obtained through record linkage with the Netherlands Cancer Registry (NCR), which has nationwide coverage since that year. 27,28tal status and date of death were obtained through linkage with the NCR and through record linkage with the Central Bureau for Genealogy.For survivors treated from 1989 onward, the most recent linkage with NCR was complete until January 1, 2022.For survivors treated between 1975 and 1988, the last NCR linkage was performed in 2012, as current European privacy regulations complicate new linkages for these survivors.

Statistical Analysis
As only 5-year survivors were included in this study, time at risk started 5 years after start of HL treatment and ended at the date of diagnosis of invasive BC or ductal carcinoma in situ (DCIS), date of death, or the censoring date (on the basis of completeness of NCR data, or date of last medical information), whichever occurred first.If not otherwise stated, BC encompasses both invasive BC and DCIS.BC diagnoses within 5 years after HL treatment (n 5 2) were ignored, but these survivors did contribute person-time to our analyses from 5 years after HL treatment onward, as they were still at risk to develop contralateral BC.Any subsequent

CONTEXT Key Objective
To evaluate whether doxorubicin exposure is associated with breast cancer (BC) risk in female 5-year Hodgkin lymphoma (HL) survivors treated at age 15-50 years, between 1975 and 2008.

Knowledge Generated
Almost one in five survivors developed BC after 30 years of follow-up.Exposure to a cumulative doxorubicin dose of >200 mg/m 2 was associated with a 1.5-fold increased BC risk, independent of age at HL treatment, receipt of chest radiotherapy (RT), or gonadotoxic therapy.Despite lower doses and smaller volumes of RT in contemporary treatment regimens, BC risk has not decreased in survivors treated between 1998 and 2008 compared with survivors treated between 1975 and 1986 (20-year cumulative BC incidence 7.8% v 8.2%, respectively).

Relevance (J.W. Friedberg)
These results emphasize the importance of very long follow-up for survivors of HL, and, if validated in other datasets, provide another factor to consider when evaluating risks of therapy.Ongoing trials incorporating checkpoint blockade and decreasing chemotherapy exposure have particular relevance given these findings.*cancer after HL other than BC was ignored in all analyses.Calculation of expected BC numbers was based on age-and calendar year-specific BC incidence rates in the female Dutch population (derived from the NCR), multiplied by the corresponding number of person-years at risk observed in our cohort.From the observed and expected numbers of BCs, standardized incidence ratios (SIRs), absolute excess risks (AERs, per 10,000 person-years), and their corresponding 95% CIs were computed using standard methods. 29mulative BC incidence was estimated with death as a competing risk.Cumulative BC incidences according to period of HL treatment (1975-1986, 1987-1997, and 1998-2008) were compared using competing hazards regression models. 30 all multivariable analyses, the cumulative doxorubicin dose was included either continuously or as a categorical variable (either as doxorubicin yes/no, or categorized as: no doxorubicin, 1-200 mg/m 2 , and >200 mg/m 2 ; Data Supplement, Table S1, online only).
Cumulative procarbazine dose was categorized as ≤4.2 g/m 2 , 4.3 to 8.4 g/m 2 , or >8.4 g/m 2 .RT to the chest was categorized according to the volume of breast tissue within the radiation field as follows: no chest RT (may include neck, supraclavicular, and/or infraclavicular fields), mediastinal field (not including axilla), mantle field or RT including axillary field(s) (MF/axillary RT), and unspecified RT fields.For analysis of effect modification, gonadotoxic treatment was defined as having received pelvic RT and/or a procarbazine dose >4.2 g/m 2 .Of all survivors treated with anthracyclines, 170 (13.3%) had received epirubicin.The association between epirubicin and BC risk was evaluated using receipt of epirubicin as a binary variable (yes/no), as 89% of epirubicintreated survivors received a cumulative dose ≥420 mg/m 2 .
To identify treatment factors associated with BC risk, we used multivariable Cox proportional hazards regression, with attained age as the time scale.The proportional hazard assumption was assessed using graphical and residualbased methods.Age at HL diagnosis violated the proportionality assumption and analyses were therefore stratified for age at HL treatment as a continuous variable.
Interactions between doxorubicin and age at HL diagnosis, gonadotoxic treatment, MF/axillary RT, and time since HL treatment were evaluated by comparing models with and without a cross-product.For assessment of effect modification by time since treatment, a cross-product between doxorubicin and various cutpoints for time since treatment (either 15, 20, 25, or 30 years after treatment) was added to the model; models were compared using the Akaike's information criterion.
We performed several sensitivity analyses: (1) excluding DCIS; (2) excluding all survivors with missing data on treatment variables (complete case analysis); (3) reclassifying survivors treated with lung RT to the MF/axillary RT category; and (4) excluding survivors treated before 1989 to see whether our results were consistent when restricted to a more recent treatment era and NCR-reported BCs.
All reported P values are two-sided; P values <.05 were considered statistically significant.All analyses were performed with Stata statistical software, version 15.0 (StataCorp, College Station, TX).

RESULTS
The median age at first HL treatment was 27.8 years (IQR, 21.9-35.2;Table 1).Most HL survivors (69.8%) were treated with a combination of RT and chemotherapy; 11.4% received chemotherapy only.A total of 1,113 survivors received doxorubicin.Among the survivors who received chemotherapy, the proportion of survivors treated with doxorubicin increased from 32.6% between 1975 and 1986 to 84.5% between 1998 and 2008 (Data Supplement, Table S2).RT involving MF or including axillary field(s) decreased from 80.3% between 1975 and 1986 to 17.3% between 1998 and 2008.Receipt of gonadotoxic treatment decreased from 51.4% between 1975 and 1986 to 6.8% between 1998 and 2008.
After a median follow-up of 21.6 years (IQR, 15.8-27.1 years), 252 women developed BC, 51 of whom had DCIS as a first BC event (Data Supplement, Table S3).The median interval between HL treatment and BC diagnosis was 19.9 years (IQR, 15.9-24.9years) and the median age at BC diagnosis was 45.8 years (IQR, 40.6-51.8years).
Compared with the female general population, HL survivors experienced a 4.4-fold (95% CI, 3.9 to 5.0) increased BC risk, translating into an AER of 63.5 BCs per 10,000 personyears.SIRs in all treatment periods increased in the first 15 years of follow-up and then stabilized (Data Supplement, Fig S1).SIRs for BC strongly decreased with older age at diagnosis and were 17.8 (95% CI, 14.0 to 22.3), 6.7 (95% CI, 5.5 to 8.10), and 1.9 (95% CI, 1.4 to 2.4) for survivors age younger than 21 years, 21-30 years, and 31-50 years at HL treatment, respectively.Among doxorubicin-treated survivors, SIRs were significantly increased after ≥20 years of follow-up, but not in the 5-19 years of follow-up interval (adjusted for chest RT and gonadotoxic treatment, P interaction 5 .001;Table 2).Similar results were obtained in analyses excluding DCIS, or excluding all survivors with missing data on treatment variables, or when reclassifying survivors treated with lung RT to the MF/axillary RT category (Data Supplement, Tables S6-S8).The results of a sensitivity analysis restricted to survivors treated in the period between 1989 and 2008 were very similar to the results obtained for the full cohort (Data Supplement, Table S9).

DISCUSSION
This large cohort study with detailed information on primary and relapse treatment, and near-complete, long-term follow-up for subsequent BCs shows that doxorubicin Follow-Up Time (years) exposure in HL survivors treated at age 15-50 years is associated with a 1.4-fold increased BC risk.This increased risk was independent of age at first HL treatment and receipt of chest RT.In accordance with previous studies, BC risk was lower for survivors treated with radiation limited to the mediastinum compared with MF or axillary field irradiation. 1,2owever, despite lower chest RT volumes and doses in more contemporary HL treatment regimens, neither cumulative BC incidence nor SIRs declined in more recently treated patients.This suggests that other changes in HL treatment policy counteract the effects of less toxic RT regimens.
The absence of a decline in BC incidence in more recently treated HL survivors may partly result from the decreased use of gonadotoxic treatments.As shown previously, a premature menopause, caused by gonadotoxic chemotherapy or pelvic RT, substantially decreases the risk of radiation-associated BC. 1,2,10,31 A second major change over the past decades is the increased use of the anthracycline doxorubicin.Doxorubicin can cause malignant transformation and mutations in vitro and has been associated with an increased risk of acute myeloid leukemia in humans. 32,33oxorubicin causes DNA damage by inducing double-strand breaks, and chromatin damage through histone eviction in the genome, which may explain its carcinogenicity. 349][20][21][22] The study in Dutch CCS reported that doxorubicin was associated with a dose-dependent increase in BC risk with HRs of 1.1, 2.6, and 5.8 for survivors who had received doxorubicin doses of ≤270 mg/m 2 , 271-443 mg/m 2 , and >443 mg/m 2 , respectively (P trend <0.001). 19We also found a significantly increased BC risk in survivors treated with a doxorubicin dose >200 mg/m 2 (HR, 1.5 [95% CI, 1.08 to 2.1]) and an increasing BC risk with increasing doxorubicin dose (HR, 1.18 [95% CI, 1.05 to 1.32] per 100 mg/m 2 increase in doxorubicin dose).When dose-response analyses were restricted to survivors treated with doxorubicin, the dose-response relationship was no longer statistically significant, but the HR point estimate for each 100 mg/m 2 increase in doxorubicin dose did not differ from the HR on the basis of analyses including survivors not treated with doxorubicin (HR, 1.20 [95% CI, 0.92 to 1.56]).Our findings corroborate the carcinogenic potential of doxorubicin and suggest that the increased use of doxorubicin may partly 179 (43)   No Doxorubicin − Total (event) 99 (19)   130 (10)   230 (34)   623 (47)   123 (20)   325 (33)   229 (9)   853 (48)   466 ( 53  All models were adjusted for chest radiotherapy, gonadotoxic treatment, and stratified for age at HL treatment.a Survivors with unknown procarbazine dose are not shown (n 5 60 with six events).b Numbers of observations exceed total number of survivors in the cohort, as survivors can contribute person-time to both the time interval <20 years and ≥20 years after diagnosis.The model including an interaction term of doxorubicin with time since treatment, and doxorubicin and gonadotoxic treatment, was not better than a model only including interaction of doxorubicin with time since treatment (P likelihood ratio test 5 .14).The model including an interaction term of doxorubicin with mantle/axillary field RT, and doxorubicin and gonadotoxic treatment, was not better than a model only including interaction of doxorubicin with mantle/axillary field RT (P likelihood ratio test 5 .10).BC, breast cancer; HL, Hodgkin lymphoma; MF/axillary RT, mantle field or radiotherapy including axillary field(s).
explain why HL survivors treated in more recent years do not experience lower BC risk than survivors treated before the 1990s, when MF irradiation was widely used. 17 surveillance for female survivors treated with MF irradiation was gradually introduced in several Dutch hospitals between 1996 and 2010, because of increased awareness of increased BC risk in these survivors.24,38,39 Mammography-based screening was offered to women treated with chest RT from age 25 years (≥8 years after treatment), 40 which led to a greater proportion of DCIS diagnoses among the BCs detected in our population.A gradual increase in BC incidence, associated with increasing screening uptake in more recent years, might introduce a spurious association with doxorubicin exposure, as use of doxorubicin increased in the same time period.However, in an analysis restricted to invasive BC, results were very similar to our main analysis on the basis of all BC events.Additionally, the association between doxorubicin exposure and BC was slightly stronger among survivors not treated with chest RT.As these women were not eligible for BC screening, this argues against screening bias as an explanation for our results.
In two CCS studies, the BC risk increase and doxorubicin dose-response trend were stronger for CCS with primary tumors typically observed within the spectrum of the Li-Fraumeni (like) syndrome (LFS), such as leukemia, sarcoma, and brain tumors, compared with survivors of other solid tumors and lymphoma combined, which might indicate a gene-anthracycline interaction in the development of BC. 18,19 Our findings show that the doxorubicin-BC association is not restricted to patients with LFS, as HL is not a LFS-associated malignancy.Additionally, the St Jude Lifetime cohort study recently showed that a higher anthracycline dose is associated with increased BC risk independent of mutations in known cancer predisposition genes. 22Nevertheless, further research is needed to assess whether genetic or metabolic factors may play a role in the doxorubicin-BC association.
In our study, epirubicin treatment was not associated with increased BC risk; however, only few survivors (159 females, of whom six developed BC) received anthracycline-containing regimens with epirubicin alone, many of whom were enrolled in the EORTC H7 trial that aimed to reduce toxicity without jeopardizing efficacy. 41The low number of survivors treated with epirubicin resulted in a low number of BCs.The upper confidence bound of our effect estimate does not preclude an increased risk associated with epirubicin exposure.
The association of doxorubicin with BC risk appeared to be present only in the interval ≥20 years after HL treatment and not in the interval 5-19 years after treatment.Increased SIRs were also limited to survivors with ≥20 years of follow-up compared with survivors with <20 years of follow-up.This may reflect an anthracycline-associated induction period for BC.
Our study included survivors from a broad treatment period and therefore takes into account the change from extended-field RT to involved-field RT.Importantly, RT regimens have significantly evolved since 2006, when the involved site/node principle was introduced and modern RT techniques, such as intensity-modulated RT or deep inspirational breath hold, were also applied to patients with HL. 42 These therapies enable significant reductions of exposure of normal tissues to radiation 43 ; therefore, a reduction in radiation-associated BC risk is expected.5][46][47] Doxorubicinassociated BC risk may be an additional late effect that needs to be considered when weighing risks and benefits of chemotherapy only versus combined modality in primary treatment.
Our study has some limitations.Doxorubicin is commonly given in standard combinations with bleomycin and vinblastine (eg, ABV).Although there is little evidence for carcinogenicity in humans because of bleomycin or vinblastine, 32 we could not distinguish the effect of doxorubicin from other agents.A large proportion of HL survivors had received chest RT, which made it challenging to fully separate associations between doxorubicin and BC and chest RT and BC.However, mantle or axillary field irradiation did not modify the association between doxorubicin and BC risk.In our analyses, doxorubicin and procarbazine doses were based on standard regimen doses and number of chemotherapeutic cycles; data on dose reductions were not systematically collected.However, in previous case-control studies in our cohort, 10 dose reductions occurred infrequently.Duration of ovarian function after HL treatment would probably have been a better indicator of individual gonadotoxicity than receipt of gonadotoxic treatment; however, information on menopausal age was incomplete for a large part of the cohort.
In conclusion, our study suggests that adolescent and adult female 5-year HL survivors treated with doxorubicin have an increased risk of BC, which is independent of age at first HL treatment, receipt of chest RT, and gonadotoxic treatment.Female patients with HL receiving more contemporary treatments continue to have an increased BC risk compared with the general population, despite efforts to limit radiation-associated subsequent cancers.Our study shows that it remains of great importance to follow patients for more than 20 years after their treatment has been completed.Our findings confirm the importance of risk-based long-term follow-up care for lymphoma survivors and possibly survivors of other cancers treated with doxorubicin.
Our results are also relevant for treatment strategies for patients with newly diagnosed HL, when balancing the risks and benefits of systemic therapy and RT.Since effective novel agents (eg, antibody-drug conjugates and immune checkpoint inhibitors) have been introduced in the treatment for patients with HL, 48 future clinical trials should also aim at reducing the dose of doxorubicin.Further studies are needed to evaluate the effect of doxorubicin on BC risk in the absence of chest RT in more recently treated patient cohorts and to assess the role of genetic susceptibility in the development of doxorubicin-associated BC.

5 HRFIG 2 .
FIG 2. Doxorubicin and BC risk: results from multivariable Cox regression interaction analysis.Doxorubicin and BC risk, according to age at treatment, mantle or axillary field RT, gonadotoxic treatment and follow-up interval: results from multivariable Cox regression interaction analysis.All models were adjusted for chest radiotherapy, gonadotoxic treatment, and stratified for age at HL treatment.a Survivors with unknown procarbazine dose are not shown (n 5 60 with six events).b Numbers of observations exceed total number of survivors in the cohort, as survivors can contribute person-time to both the time interval <20 years and ≥20 years after diagnosis.The model including an interaction term of doxorubicin with time since treatment, and doxorubicin and gonadotoxic treatment, was not better than a model only including interaction of doxorubicin with time since treatment (P likelihood ratio test 5 .14).The model including an interaction term of doxorubicin with mantle/axillary field RT, and doxorubicin and gonadotoxic treatment, was not better than a model only including interaction of doxorubicin with mantle/axillary field RT (P likelihood ratio test 5 .10).BC, breast cancer; HL, Hodgkin lymphoma; MF/axillary RT, mantle field or radiotherapy including axillary field(s).

TABLE 1 .
Characteristics of Female 5-Year Hodgkin Lymphoma Survivors a Either ductal carcinoma in situ or invasive BC, whichever occurred first.bcAdoxorubicindose of >200 mg/m 2 corresponds to ≥6 cycles of MOPP-ABV, >4 cycles of ABVD, or ≥6 cycles of escalated BEACOPP.dThecategorynochestRTincludes58survivors with RT limited to neck and 54 survivors with supraclavicular and/or infraclavicular fields (with or without neck).RT to the lung was also applied in 12 survivors in the mantle field category, in 19 survivors in the mediastinal RT category, and in one survivor in the unspecified RT category.Of all survivors categorized into unspecified RT, four survivors received supradiaphragmatic fields.eAprocarbazinedose of ≤4.2 g/m 2 corresponds to ≤3 MOPP cycles or ≤6 cycles of MOPP-ABV.A procarbazine dose of 4.3 to 8.4 g/m 2 corresponds to 4-6 cycles of MOPP.A procarbazine dose of >8.4 g/m 2 corresponds to >6 cycles of MOPP.Compared with survivors who did not receive doxorubicin (43.3%), survivors treated with doxorubicin had a 1.4-fold (95% CI, 1.04 to 1.9) increased hazard of BC (adjusted for age at HL diagnosis, chest RT, and gonadotoxic treatment; Table3, model A).The hazard ratios (HRs) for BC in survivors treated with 1-200 mg/m 2 or >200 mg/m 2 doxorubicin were 1.4 (95% CI, 0.9 to 2.0) and 1.5 (95% CI, 1.08 to 2.1), respectively, compared with survivors not treated with doxorubicin (Table3, model B).Each 100 mg/m 2 increase in doxorubicin dose was associated with a 1.18-fold increase in BC risk (95% CI, 1.05 to 1.32; Table3, model C).In an analysis that only included survivors who had received doxorubicin (n 5 1,113), the HR for BC associated with a 100 mg/mThe association between doxorubicin and BC risk differed by time since HL treatment, with increased BC risk limited to the follow-up interval ≥20 years after first HL treatment (HR ≥20 years follow-up , 2.1 [95% CI, 1.4 to 3.2] P interaction 5 .004;Fig2).When doxorubicin was included in the model using dose categories, this effect modification by time was also present (HR 1-200 mg/m 2 , <20 years follow-up , 1.0 [95% CI, 0.5 to 1.6]; HR 1-200 mg/m 2 , ≥20 years follow-up , 2.1 [95% CI, 1.1 to 3.9]; HR >200 mg/m 2 , <20 years follow-up , 1.0 [95% CI, 0.6 to 1.

TABLE 2 .
FIG 1.(A) Cumulative incidence of BC in female 5-year HL survivors according to HL treatment period.Solid lines represent the cumulative incidence of invasive BC and DCIS in our cohort, stratified by period of HL treatment (P heterogeneity 5 .53).Dashed lines represent the expected incidence of invasive BC and DCIS in the Dutch general population, stratified by the same periods.(B) Cumulative incidence of invasive BC in female HL survivors according to HL treatment period.Solid lines represent the cumulative incidence of invasive BC in our cohort, stratified by period of HL treatment (P heterogeneity 5 .65).Dashed lines represent the expected incidence of invasive BC in the general population, stratified by the same periods.BC, breast cancer; DCIS, ductal carcinoma in situ; HL, Hodgkin lymphoma.SIR and Absolute Excess Risk of BC, by HL Treatment and Attained Follow-Up Total number of observations exceed total number of survivors in the cohort, as survivors can contribute person-time to both the time interval <20 years and ≥20 years after diagnosis.
A B Journal of Clinical Oncology ascopubs.org/journal/jco| Volume 42, Issue 16 | 1907 Breast Cancer Risk After Doxorubicin for Hodgkin Lymphoma NOTE.SIRs were calculated if at least one BC was observed.Abbreviations: AER, absolute excess risk; BC, number of breast cancers observed; HL, Hodgkin lymphoma, MF/axillary RT, mantle field or RT including axillary field(s); RT, radiotherapy; SIR standardized incidence ratio.a

TABLE 3 .
Treatment-Related Risk Factors for Subsequent BC in Female 5-Year Survivors of Hodgkin Lymphoma: Results From Multivariable Cox Regression Analysis NOTE.Analyses were stratified for age at first HL treatment (continuous).Lower limits of 95% CIs were rounded down, upper limits were rounded up.For doxorubicin-treated survivors, the median dose was 210 mg/m 2 (IQR, 175-300 mg/m 2 ).For doxorubicin-treated survivors with an event, the median dose was 210 mg/m 2 (IQR, 200-280 mg/m 2 ).Abbreviations: ABVD, doxorubicin, bleomycin, vinblastine, and dacarbazine; BC, breast cancer; BEACOPP, bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone; CT, chemotherapy; MOPP, mechlorethamine, vincristine, procarbazine, and prednisone; MOPP-ABV, MOPP with doxorubicin, bleomycin, and vinblastine; RT, radiotherapy. of >200 mg/m 2 corresponds to ≥6 cycles of MOPP-ABV, >4 cycles of ABVD, or ≥6 cycles of escalated BEACOPP.The category no chest RT includes 58 survivors with RT limited to the neck and 54 survivors with supraclavicular and/or infraclavicular fields (with or without neck).RT to the lung was also applied in 12 survivors in the mantle field category, in 19 survivors in the mediastinal RT category, and in one survivor in the unspecified RT category.Of all survivors categorized into unspecified RT, four survivors received supradiaphragmatic fields.
a A doxorubicin dose b