Hodgkin lymphoma in the elderly, pregnant, and HIV-infected

Abstract

Hodgkin lymphoma (HL) presenting in patients with co-incidental advanced age, pregnancy, or human immunodeficiency virus (HIV) infection is uniquely challenging to manage. In this article we integrate recent evidence and clinical expertise to present recommendations for diagnosis and therapeutic management. Older patients with HL need to be carefully evaluated for comorbidies after which judicious choice of chemotherapy should minimize functional compromise. A pregnant patient with concurrent HL should be staged with minimal use of imaging requiring ionizing radiation and treated in an individualized manner optimally combining the strategies of treatment deferral when appropriate, use of single-agent vinblastine for symptomatic disease and reservation of multi-agent chemotherapy for the small minority of patients with aggressive clinical presentation. Treatment of HL coincident with HIV infection requires a combination of highly active anti-retroviral agents (HAART), standard multi-agent chemotherapy with meticulous attention to drug–drug interactions, and vigorous supportive care to ensure the best chance of cure.

1. Introduction

Hodgkin lymphoma (HL) is a largely curable disease regardless of disease stage at presentation. Curative treatment involves the use of multi-agent chemotherapy and/or localized radiation for most patients and high-intensity treatment for a minority of patients who are not cured by initial chemo-radiotherapy. These treatment strategies are highly successful and generally well tolerated; however, it is necessary to recognize that standard therapeutic approaches to HL should be modified for subpopulations of patients who present with or develop unique challenges to the successful delivery of standard treatment. In the current review, we present an overview of data and recommendations for the management of three specific subpopulations: older patients with cardiac or pulmonary comorbidities, patients with coincident pregnancy, and patients positive for human immunodeficiency virus (HIV).

2. Treatment of Hodgkin lymphoma in older patients

Approximately 20% of all HL patients are older than 60 years at diagnosis, an incidence that has been stable for several decades. Histologic subtypes occur at generally similar frequencies as in younger patients; however, Epstein-Barr virus (EBV)-positivity in Reed-Sternberg cells is more often present in patients with classical HL above age 50 and portends worse prognosis [1]. In those 60–70 years of age, clinical outcomes have been found to be acceptable for patients with early-stage HL, but poor for patients with advanced-stage HL. Results in those older than 70 years are particularly poor, irrespective of stage. The generally less favorable prognosis of HL among older patients is predominantly caused by an increased rate of complications [2] that often lead clinicians to use lower chemotherapy dose intensities and substantially increase treatment-related mortality [3].

Older patients frequently have underlying pulmonary or cardiac dysfunction and other comorbidities that are often severe enough to preclude safe delivery of standard chemotherapy. These comorbidities limit therapeutic options, particularly the use of bleomycin or anthracyclines, which are the backbone of most multi-agent regimens. In addition, older patients are susceptible to exaggerated myelosuppression that is associated with a higher risk of life-threatening infections and organ failure. Major challenges that need to be addressed while treating elderly patients with HL are best managed by a multi-specialty team incorporating an oncologist, relevant specialists, and a supportive team (Fig. 1). Comprehensive geriatric assessment (CGA) tools incorporating activities of daily living, comorbidities, cognitive function, nutritional status, and social and psychological status are often useful to holistically treat elderly patient and select the most appropriate therapy [4].

Fig. 1.

Suggested algorithm for the treatment for pregnancy-associated HL.

Studies reporting on outcomes for older patients treated with commonly used multi-agent regimens are listed in Table 1. The most commonly employed multi-agent chemotherapy regimen, ABVD (doxorubicin, bleomycin, vinblastine, and dacarbazine), must be used carefully and often modified for older patients. The BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, prednisone, and procarbazine) regimens are too toxic for patients over the age of 60 and must be avoided. Doxorubicin is frequently contraindicated in older patients because of underlying cardiac illness. Prior to chemotherapy, all patients should be evaluated for potential cardiac dysfunction by echocardiography or multi gated acquisition scan (MUGA). If the left ventricular ejection fraction is less than 50% or there is a history of documented congestive heart failure, an alternative to doxorubicin should be used. When using ABVD, etoposide can be substituted for doxorubicin at a dose that causes roughly equivalent myelosuppression. Patients with underlying pulmonary compromise should be identified by careful questioning about exercise tolerance, smoking history or prior exposure to pulmonary toxins such as industrial dust and asbestos rather than solely relying on pulmonary function tests. Bleomycin should be omitted from a patient’s treatment if clinical assessment indicates that the patient could not safely tolerate a drop in current lung function of about one third of current capacity because, in our personal experience, an occasional patient may unexpectedly and rapidly lose one third of current lung function from baseline as a result of bleomycin treatment. In addition, recent evidence from German Hodgkin Study Group (GHSG) and the Response Adapted Treatment for Hodgkin Lymphoma (RATHL) trial indicates that bleomycin can be safely omitted in those patients whose HL responds after the initial two cycles of treatment [5]. Worrisomely, the GHSG study reported a 10% risk of severe bleomycin-induced lung toxicity, including three deaths in 69 HL patients >60 years of age receiving more than two cycles of ABVD [6].

Table 1.

Reported series of front-line therapy for Hodgkin lymphoma in the elderly.

Regimen	COPP-ABVD or BEACOPP	COPP	CHOP-21	BEACOPP	VEPEMB	ODBEP	VEPEMB	ABVD (70%+ other)	PVAG	ABVD v VEPEMB	Brentuximab vedotin
n	372	52	29	42	105	38	103	95	59	57	27
Median age (yr)	65			68	71	72	73	67	68	72	78
Comorbidity %			55		37			61			81
Stage limited %	34	16	38	0	46	0
Stage advanced %	66	84	62	100	54	100	67	64	93	65	63
Febrile neutropenia %	15	6	31		10	14
CR %	86%		93%	76%	76%		74%	73%	78%		73%
Median follow-up (mo)		60	41	80			36	66	37	76	17
PFS %	60 (5-yr)		76 (3-yr)	46 (5-yr)	56 (5-yr)	49 (5-yr)	74 (3-yr)		58 (5-yr)	70 v 48 (5-yr)	Median 10.5 mo
Overall survival %	65 (5-yr)	48 (5-yr)	79 (3-yr)	50 (5-yr)	64 (5-yr)	42 (5-yr)	81 (3-yr)	58 (5-yr)	66 (3-yr)	77 v 63 (5- yr)	OS 4.6–25 mo
Fatal toxicity %	6			21	2		7		1 patient	4	0
Reference	Engert [2], 2005	Feltl [8], 2006	Kolstad [9], 2007	Ballova [10], 2005	Levis [3], 2004	Macpherson [11], 2002	Proctor [16], 2012	Evans [12], 2011	Boll B [6] 2011	Zallio [15] 2016	Forero-Torres [18], 2015

We summarize response rates and survival after multi-agent chemotherapy regimens in elderly with HL [3,7–11] (Table 1). Recently, Evans et al reported on prognostic factors of 95 elderly HL subjects treated in the past decade in United States [12]. Most patients (median age was 67 years) had significant comorbidities, and a third presented with geriatric syndrome or loss of activities of daily living. Primary treatment consisted of ABVD (n = 67), BCVPP (bleomycin, chlorambucil, vinblastine, procarbazine, and prednisone; n = 6), ChlVPP (chlorambucil, vinblastine, procarbazine, and prednisone; n = 5), or other (radiation, MOPP-based [MOPP = mechlorethamine, vincristine, prednisone, procarbazine], etc; n = 17). Most patients received granulocyte colony-stimulating factor (G-CSF) to support neutrophil recovery. The overall response rate was 85%. Bleomycin lung toxicity was observed in 32% of patients and was associated with a high mortality rate. The 5-year overall survival rate was 58% (Table 1). Factors associated with inferior overall survival were age >70 years and loss of activities of daily living.

The ChlVPP regimen is well tolerated and effective in older patients with organ compromise [13]; however, the complete response rates are lower and survival is inferior to more aggressive regimens [14]. Levis et al demonstrated that the use of three cycles of reduced-intensity VEPEMB (vinblastine, cyclophosphamide, prednisolone, procarbazine, etoposide, mitoxantrone, and bleomycin) plus involved-field radiotherapy produces excellent complete remission (CR) rates in patients with early-stage disease. In addition, six cycles of VEPEMB were effective and well tolerated in patients with advanced disease (Table 1) [3]. In 2016, the same group reported results of prospective trial of 54 older HL patients aged 65–80 years and considered “non-frail” who were randomized to ABVD versus VEPEMB regimen. Progression-free survival (PFS) favored ABVD (70% v 48%) without statistical significance. Overall mortality was only 4% and reflected stringent selection of patients based on CGA that excluded frail patients [15]. In the United Kingdom Study of Hodgkin in the Elderly/Lymphoma Database (SHIELD) program, a large patient cohort was prospectively treated with VEPEMB chemotherapy. The CR rate was 74%, and the 3-year PFS was 74%. Treatment-related mortality of 7% was either due to toxicity-related deaths or secondary malignancies [16]. Recently, the GHSG developed a novel regimen, PVAG (prednisone, vinblastine, doxorubicin, and gemcitabine), for patients who are 60 years and older; most patients had advanced-stage disease and adequate heart function. Fifty-nine patients with a median age of 68 years received six to eight cycles of PVAG and additional radiotherapy if they were not in CR after chemotherapy. Results were favorable with 78% attaining CR. The 3-year OS and PFS rates were 66% and 58%, respectively. Almost all patients had grade 3/4 toxicities, but only one patient died as a result of toxicity. These results suggest that PVAG is a safe and feasible alternative in elderly HL patients [17]. More recently, brentuximab vedotin (BV), a CD30-directed antibody–drug conjugate, was examined as frontline therapy in 27 HL patients aged ≥60 years [18]. The objective response rate (ORR) was 92%, with 73% achieving CR. With relatively short follow-up, remissions lasted for a median of 9 months and OS ranged from 5–25 months. Importantly, the incidence of grade 3 peripheral neuropathy events was relatively high (30% overall), particularly among patients with the known risk factors of diabetes and/or hypothyroidism. It is encouraging that myelosuppression was not a significant toxicity in this study. These preliminary data suggest that BV monotherapy may be an acceptable alternative frontline treatment option for older patients with major comorbidities who are not candidates for potentially curative conventional chemotherapy; however, long-term efficacy data are required to assess the durability of response. Data on second-line treatment and survival in older patients with progressive or relapse HL are limited. In a comprehensive study from the GHSG patients treated with salvage therapy or radiation had median overall survival 12 months and one third of the patients were alive at 3 years. Patients with early relapse, advanced stage and anemia had a worse prognosis [19]. High-dose chemotherapy followed by the autologous stem cell transplantation can be considered in those patients with HL who have a chemosensitive relapse and are in excellent health. Standard BEAM (carmustine, etoposide, cytarbine, and melphalan) high-dose chemotherapy resulted in a 2-year PFS of 67% and only 3% transplant-related mortality.

In summary, while HL in elderly is potentially curable in all age groups, outcomes in older patients are still disappointing compared to younger patients. More than 50% of deaths attributable to HL occurs in patients older than 65 (seer.gov) reflecting the obvious limitations of conventional chemotherapies. Novel agents should be investigated in alternative combinatory regimen for older HL patients [20,21].

3. Treatment of Hodgkin lymphoma patients with coincident pregnancy

HL is the fourth most frequent cancer diagnosis among pregnant females and the most common hematologic malignancy complicating pregnancy with an incidence between 1:1,000 to 1:3,000 deliveries [22]. Over 70% of HL patients diagnosed during pregnancy have stage IA or IIA disease and are asymptomatic or minimally symptomatic. A multidisciplinary team composed of a hemato-oncologist, high-risk obstetrician, and high-risk neonatologist should work together closely to address the health needs of both mother and child (Table 2). When planning the diagnostic evaluation of HL in a pregnant patient, one has to balance the need for accurate disease assessment with the need to minimize the use of invasive procedures. The guiding principle in the pregnant patient is to restrict radiologic staging to the minimum necessary to identify disease that seriously threatens the immediate well-being of mother or child. Computed tomography (CT) and positron emission tomography (PET) scans should be avoided, especially during early stages of gestation [23]. Instead, a single postero-anterior radiograph of the chest, with proper shielding, should be obtained to characterize the extent of mediastinal and pulmonary disease. Abdominal ultrasonography can identify the extent and size of retroperitoneal nodal disease with sufficient detail for proper management [24]. Magnetic resonance imaging (MRI) without use of gadolinium can also be used if further detailed imaging is required [25].

Table 2.

Challenges, characteristics of multidisciplinary team, and staging test recommended for HL discovered during pregnancy.

Challenges

Avoid tests that can harm fetus

Potential toxicity of therapy on fetus

Respect for patient and family wishes

Need for multidisciplinary team to manage the patient

Therapeutic team

Hematologist/oncologist

Plan the diagnostic work-up and staging to minimize the adverse effects of fetus

Coordinates with others to establish the plan of care

Administers the chemotherapy if deemed necessary

Provides supportive car reviews safety of medications

Establish the extend of staging after delivery and provides oncologic care

High-risk obstetrician

Plans management of labor and delivery

Counsels the patient if pregnancy termination is recommended by the team

Administers antepartum steroids if appropriate

Counsels about breastfeeding and effective postpartum contraception

Neonatologist/pediatrician

Focused care of the child, who may be premature at birth

Coordinates post-natal care of newborn and long-term follow-up

Registers the newborn in central registry born to pregnant mothers with HL

Staging of HL discovered during pregnancy

Complete history searching for symptoms of HL

Phyrical examination for lymphadenopathy and organomegaly

Complete blood count and chemistry including lactate dehydrogenase and albumin

Chest radiograph postero-anterior view only

Abdominal ultrasound

Bone marrow biopsy only if unexplained cytopenias

Recent evidence suggests that pregnancy itself does not affect the course of HL, response to therapy, or the OS rate when compared with age- and stage-equivalent non-pregnant controls [26,27] nor does becoming pregnant alter the likelihood of recurrence of HL after the lymphoma has been effectively treated [28]. Fortunately, the majority of patients with HL during pregnancy require no immediate intervention. As a general rule, treatment should be avoided during the first trimester unless severe, life-threatening symptoms are present. Almost all chemotherapy agents have been documented to be teratogenic in animals or humans causing abortion, fetal death, and major malformation. The fetus is extremely vulnerable from the second to eighth week of gestation, during which time organogenesis occurs. Although generally unnecessary, the option of pregnancy termination needs to be considered for the rare patient whose first trimester is complicated by severely symptomatic or life-threatening disease, where the need for immediate chemotherapy may affect fetal development.

Asymptomatic patients with coincidental HL and pregnancy—particularly patients with stage IA–IIA HL—only require close monitoring. The majority such patients can continue the pregnancy to term without any treatment for the lymphoma after which they can complete appropriate staging and initiate treatment. This approach has been demonstrated to be safe in small case series such as those from Stanford (17 patients) and Royal Mardsen Hospital (19 patients) [26,29]. Symptomatic patients with constitutional or respiratory symptoms due to enlarging mediastinal masses should be treated with systemic chemotherapy. Single-agent vinblastine is a preferred agent if a patient requires treatment [30]. Vinblastine at standard dosing (6 mg/m²) results in a >75% response rate in treatment-naïve patients, does not cross the placenta and has only modest acute toxicity [31,32]. Vinblastine’s high level of effectiveness, minimal acute toxicity, and low likelihood of a negative effect on the fetus make it an attractive agent to suppress HL during pregnancy. Doses at intervals of 3–6 weeks or longer can be given to control HL until delivery at term, minimizing risks to mother and child. Progression despite vinblastine, which occurs quite infrequently, should be treated with full-dose ABVD because evidence of chemotherapy resistance signifies aggressive disease requiring multi-agent chemotherapy. Management of HL with bulky disease, serious visceral involvement, or rapid disease progression requires systemic chemotherapy. ABVD has been used during pregnancy, without obvious negative effects on the fetus, although the number of patients is small and it is unclear whether the trimester of exposure is important [33,34]. One of the important considerations in overall management of pregnant patient with HL is to establish the timing and the mode of delivery. If HL therapy was deferred or only requires vinblastine, gestation should be allowed to progress to full term. If multi-agent chemotherapy becomes necessary, delivery may be induced after fetal pulmonary maturation is ensured, at least beyond 34–36 weeks of gestation. Short course antenatal corticosteroids are frequently administered to expedite fetal lung maturity, and amniocentesis can be performed to assess the response [35]. Patients who have been able to complete the pregnancy without treatment for the lymphoma can be fully staged and treated appropriately after delivery. Patients who required vinblastine or other chemotherapy can no longer be accurately staged and therefore should be treated with a full course of six to eight cycles of multi-agent chemotherapy.

4. Treatment of Hodgkin lymphoma patients with coincident HIV infection

Although the incidence of HL is 10-fold higher in patients with HIV infection than in the general population, its incidence has declined as a consequence of improved HIV control [36–38]. Patients with HIV-associated HL typically present with advanced disease, marrow involvement, B symptoms, and mixed cellularity or lymphocyte-depleted histology. Additionally, their disease pursues a more aggressive natural history.

Treatment of HL coincident with HIV infection is complicated by several factors including opportunistic infections, coincident organ dysfunction due to HIV or other viral infections such as hepatitis, potential drug interactions between chemotherapeutic agents used to treat HL and the array of anti-HIV, antibiotics, and supportive medications necessary to treat HIV (Table 3). Meticulous attention to drug-drug interactions is required to avoid toxicity and pharmacokinetic effects that can reduce the likelihood of cure. Successful treatment combines vigorous supportive care consisting of antiviral and antifungal agents and neutrophil-stimulating growth factors with HAART together with standard multi-agent chemotherapy. With appropriate supportive care, standard regimens such as ABVD or BEACOPP can be delivered. [39–42] The results of the largest studies are shown in Table 4 [40,42–45]. Greater than normal toxicity must be anticipated and can be partially mitigated by use of G-CSF and sometimes erythropoietin for drug-induced anemia. Prevention of opportunistic infection with Pneumocystis jiroveci, herpes virus, and Candida with appropriate antimicrobial and antifungal agents is recommended [37]. Some patients may benefit from psychiatric counseling, addiction management, and compliance support.

Table 3.

HL coincident with HIV infection.

Challenges

Underlying HIV syndrome

Immune compromised status Multiple drug interactions

Often advanced stage

Therapeutic team

Hematologist/oncologist

Overall coordination of anti-lymphoma treatment

Hematopoietic growth factors

HIV specialist

Planned management of HIV infection

Involvement of support team experienced in management of emotional, social, financial problems associated with HIV infection

Organ-specific specialists

Infectious disease—tuberculosis, mycoses, other opportunists

Gastrointestinal/hepatology

Psychiatry/addiction management

Neurology

Table 4.

Results of prospective studies in HIV HL since 2000.

Regimen	ABVD	ABVD	Standford	BEACOPP	VEBEP	ABVD
n	21	62	59	12	71	229
Stage advanced %	81	100	71	92	70	100
Overall response rate	62	87	89	100	78	91
Complete response %	43	87	81	100	67	83
Overall survival	18 mo	76% (5 yr)	59% (5 yr)	75% 93 yr)	69% (2 yr)	78% (5 yr)
Reference	Levine 2000 [40]	Xicoy 2007 [44]	Spina 2002 [43]	Hartmann 2003 [42]	Spina 2008 [49]	Castillo 2011 [45]

In a recent large multi-institutional, retrospective study of 229 patients with HIV-associated, advanced-stage, classical HL, patients received ABVD plus combination antiretroviral therapy [46]. Five-year PFS and overall survival rates were 69% and 78%, respectively. A CD4 cell count <200 cells/μL was independently associated with both PFS and overall survival. Mortality was often related to opportunistic infections [36,41]. Recent reports suggest that ABVD used even in the first trimester for pregnant patients with HIV-associated HL may be associated with favorable outcomes for the patient and child [47]. The widespread use of HAART before, during, and after management of the lymphoma has led to improvements in prognosis of HIV patients; however, cure rates for HL remain lower than those achieved in patients without HIV infection. Patients with relapsed disease should be considered for autologous hematopoietic cell transplantation in a tertiary transplant center with experience treating concomitant HL and HIV, as autograft can be curative for a proportion of relapsed or refractory HL patients [48–50].

5. Conclusions

The diagnosis of HL in special patient populations compromised by co-incidental conditions such as pregnancy, HIV infection, or advanced age represents a unique clinical situation. Continued efforts to collect data on all HL patients in these three subgroups would be of great value in our efforts to objectively assess the efficacy and toxicity of treatment approaches and derive prognostic factors to inform broader recommendations. As noted by participants in the International Consensus Meeting on Prenatal Hematologic Malignancies, who endorsed our recommendations for managing coincident pregnancy and Hodgkin lymphoma [51], registration with the International Network on Cancer, Infertility and Pregnancy can lead directly to improvements in our clinical knowledge that can be used to provide valuable guidance.

For these special patients a malignancy that is highly curable under standard circumstances is much more difficult to eradicate; however, careful evaluation and treatment by a dedicated therapeutic team with expertise in both lymphoma treatment and optimal management of the associated condition can often lead to a successful outcome and should be the standard of care in 2016.