Electrophysiological Studies in Combination With Interim‐Positron Emission Tomography Scan for Prevention of Severe Brentuximab‐Vedotin‐Induced Neurotoxicity

ABSTRACT

Brentuximab‐vedotin (BV)‐induced neurotoxicity (BVIN), a frequent adverse event caused by this monoclonal antibody, is the primary reason for dose modification or drug discontinuation, and is characterized by sensory, motor, and/or autonomic peripheral nerve dysfunctions. Although reversible, BVIN can persist for months or years after treatment and negatively affect quality of life (QoL). Currently, BVIN is managed by dose adjustment or drug interruption, leading to an increased risk of disease relapse. Therefore, early recognition and appropriate management are essential to improve clinical outcomes. In this real‐life study, we identified predictive factors for moderate/severe BVIN to reduce the risk of irreversible neuropathy. A total of 22 patients treated with BV were enrolled and BVIN was monitored by electro‐neurography and neurological examinations every 2 cycles of therapy, while QoL by clinical questionnaires. We showed that recovery rate from moderate/severe BVIN was low, and sensory nerves were the most affected, negatively impacting QoL. BV dose reduction based on interim PET re‐evaluation in patients with hematological response resulted in a significant reduction of BVIN onset with high long‐term QoL. Therefore, electrophysiological tests could be useful tools to prevent moderate/severe BVIN onset, and their combination with interim PET imaging could allow dosage adjustments thus simultaneously minimizing risks of disease relapse and BVIN development. However, further studies on larger prospective randomized cohorts are needed to confirm our preliminary results.

1. Introduction

Hodgkin's lymphoma (HL), a clonal malignant hematology disease, typically affects young adults with excellent prognosis even in advanced stages, because of therapeutic progress, such as the introduction of a combinatorial regimen including doxorubicin, bleomycin, vinblastine, and dacarbazine (ABVD) [1]. HL patients are currently considered long‐term survivors, and treatment strategies must consider not only clinical benefits of anti‐cancer drugs used, while also incidence and severity of chemotherapy‐related long‐term complications, such as endocrinological sequalae or cardiotoxicity [2]. Cumulative dose reduction is one of the most employed actions to prevent long‐term side effects; however, doses should be adjusted based on hematological response, to minimize the risk of therapeutic failure and disease relapse [3].

In HL, dose reduction could be calculated by assessing hematological responses during treatment (interim re‐evaluation) using the Deauville five‐point scale (DS). This score is defined based on fluorodeoxyglucose (FDG) avidity of a HL or non‐Hodgkin lymphoma tumor mass by PET scan imaging, and values range from 1 (no uptake above the background), to 3 (uptake > mediastinum but ≤ liver), to 5 (uptake markedly increased compared to the liver at any site). DS values of 1–2 are considered negative, thus patients are in hematological remission, while scores of 4–5 are positive for disease persistence or progression [4, 5]. This tool has demonstrated excellent predictive value for responsiveness to therapies, enabling cumulative dose reduction and avoiding drug‐related toxicity. For example, in the RATHL trial, HL patients with a DS of 1–3 after 2 cycles of ABVD have successfully received only other four AVD cycles, instead of six, without affecting clinical efficacy, while significantly reducing pulmonary toxicity [6, 7].

Chemotherapy‐induced peripheral neurotoxicity is an extremely debilitating side effect of several anti‐tumor agents, including chemoimmunotherapy and brentuximab‐vedotin (BV), an anti‐CD30 monoclonal antibody conjugated with monomethyl auristatin E, that is released in the cytoplasm by linker breakage through proteolytic enzymes [8, 9]. BV‐induced neurotoxicity (BVIN) is a frequent short‐ and long‐term complication, as described in the ECHELON‐1 clinical trial, where patients treated with BV‐AVD have better outcomes with longer progression‐free survival (5‐year PFS, 82%) and lower incidence of pulmonary toxicity due to bleomycin removal from the regimen, while incidence of peripheral neurotoxicity is significantly higher (19%) compared to ABVD (9%) [10, 11]. BVIN development is one of the primary causes of dose reduction or treatment discontinuation in relapse/refractory HL patients, despite most clinical trials indicate that this neurotoxic effect is reversible, although it might require months or years to fully recover [12].

In this retrospective and prospective real‐life monocentric study, we aimed at evaluating peripheral neuropathy in patients treated with BV by neurological monitoring using clinical examinations and electrophysiological studies every 2 cycles of therapy, to reduce the risk of severe BVIN development. We also confirmed that BV dose could be effectively adjusted based on DS assessed by interim‐PET in responders without affecting clinical benefits while minimizing long‐term neurological sequalae. Moreover, we retrospectively assessed presence and reversibility of any neurological disability on patients treated with BV, to correlate clinical parameters with cumulative dose administered.

2. Patients and Methods

2.1. Patients

A total of 22 consecutive patients (M/F, 10/12; mean age, 38 years old; range, 16–70 years) treated with BV as monotherapy or in combination with chemotherapy were enrolled in this retrospective and prospective real‐life study performed at the Hematology and Transplant Center, University Hospital “San Giovanni di Dio e Ruggi d'Aragona,” Salerno, Italy (Table 1). Patients were divided in two groups: a prospective cohort of 10 subjects receiving electroneurography (ENG) every 1 or 2 therapy cycles; and a retrospective group of 12 patients monitored with ENG every 3 months for 9 months after the end of therapy. Patients received a diagnosis of classical Hodgkin lymphoma (cHD) in 82% of cases (N = 18) and of non‐Hodgkin lymphomas (NHL) in the remaining 18% of subjects (N = 1, anaplastic CD30⁺ NHL; N = 1 primary mediastinal CD30⁺ NHL; N = 1 primary cutaneous CD30⁺ NHL; and N = 1 anaplastic large CD30⁺ ALK⁻ cell NHL). Most patients had a late‐stage disease (N = 14; 64%), and only 9% of subjects (N = 2) were in stage I.

TABLE 1.

Patients' characteristics and regimens.

Characteristics	Total cohort (N = 22)	Prospective (N = 10)	Retrospective (N = 12)	p
Mean age, years (range)	38 (16–70)	32 (20–70)	32 (16–63)	0.4111
M/F	10/12	4/6	6/6	0.6914
Diagnosis, n (%)				0.5940
HL	18 (81.8)	9 (90)	9 (75)
NHLs	4 (18.2)	1 (10)	3 (25)
Stage, n (%)				0.7532
I	2 (9.1)	—	2 (16.7)
II	6 (27.3)	3 (30)	3 (25)
III	5 (22.7)	3 (30)	2 (16.7)
IV	9 (40.9)	4 (40)	5 (41.6)
Regimens, n (%)				0.7685
AdAVD	11 (50)	6 (60)	5 (41.7)
AdCMP	2 (9.1)	1 (10)	1 (8.3)
BV monotherapy	7 (31.8)	3 (30)	4 (33.3)
BV + nivolumab	2 (9.1)	—	2 (16.7)
BV dose schedule				0.4149
1.8 mg/kg every 3 weeks	9 (40.9)	3 (30)	6 (50)
1.2 mg/kg every 2 weeks	13 (59.1)	7 (70)	6 (50)
Cumulative BV dose				0.1148
> 1000 mg	10 (45.5)	2 (20)	8 (66.7)
500–1000 mg	10 (45.5)	6 (60)	4 (33.3)
< 500 mg	2 (9)	2 (20)	—

Abbreviations: AdAVD, BV, Adriamycin, vinblastine, dacarbazine; AdCMP, BV, cyclophosphamide, doxorubicin, prednisone; BV, brentuximab‐vedotin; HL, Hodgkin lymphoma; N.e., not evaluable; NHLs, non‐Hodgkin lymphomas.

Patients received chemotherapy according to current international guidelines [13]: 50% of cases (N = 11) with AdAVD (BV, Adriamycin, vinblastine, dacarbazine); 9% AdCMP (BV, cyclophosphamide, doxorubicin, prednisone); 9% with BV+ nivolumab; and 32% with BV as monotherapy. BV was administered at a dose of 1.8 mg/kg every 3 weeks (N = 9) or 1.2 mg/kg every 2 weeks (N = 13), with a cumulative BV dose > 1000 mg in 41% of cases (N = 9), 500–1000 mg in 45% of subjects (N = 10), or < 500 mg in 4% of cases (N = 1). In particular, 5 patients in the retrospective cohort received a cumulative dose of 1000–1500 mg, and 2 patients > 2000 mg.

2.2. Neurological Examination

BVIN signs and symptoms were checked by a Neurologist using neurological examinations and ENG study, and four clinical neurological scales were employed to assess the risk of neurotoxicity (potential or actual): Medical Research Council (MRC); Douleur Neuropathique en 4 Questions (DN4); Sensory Symptoms Score (SSyS); and Sensory Sum Score (SSuS) [14, 15, 16]. For ENG, at each timepoint, left and right median, ulnar, peroneal, and tibial nerves were assessed for motor component evaluation, while left and right median, ulnar, superficial peroneal, and sural nerves were screened for sensory component study. Compound Muscle Action Potential (CMAP) and Sensory Action Potential (SAP) value changes at each timepoint were expressed as percent variation compared to baseline levels for the prospective cohort, or to the minimum expected value, standardized by age [17]. Neurological toxicity severity was defined according to the Common Terminology Criteria for Adverse Events (CTCAE) v.5.

For evaluation of subjective perception of symptoms of BIVN in daily life, a questionnaire was administered to patients at baseline before starting therapy, after 2 (T2) and 4 (T4) months of therapy, and at the end of therapy (T6). This survey consisted of six questions with numerical scale responses ranging from a minimum score of 0 (maximum well‐being) to a maximum score of 10 (minimum well‐being). Questions were: “Have you ever experienced tingling/burning/pain in your hands or feet?”; “When you enter the bathtub or shower, do you find difficult to distinguish between hot and cold water?”; “Have you ever experienced difficulty performing fine motor tasks (dialing numbers on a smartphone, counting money, buttoning a shirt)?”; “Have you ever experienced a sensation of losing balance (when getting up from a chair, walking while looking at your smartphone, after standing for a long time)?”; “Have you ever experienced difficulty performing actions that used to be easy (opening jars, getting up after bending down, jumping, hitting a ball with force)?”; and “Do you believe that the neurological symptoms have affected your ability to engage in your hobbies or profession?”

2.3. Statistical Analysis

Data were analyzed using Prism (v.10.2.0; GraphPad software, La Jolla, CA, USA). Unpaired two‐tailed t‐ or non‐parametric Mann–Whitney tests for two group comparison and Kruskal–Wallis test for three‐group comparison were performed. Chi‐square test was employed for dichotomous variables. A p < 0.05 was considered statistically significant.

3. Results

3.1. Clinical Neurological Signs are More Sensitive to Identify BVIN

Whether to investigate BV‐induced neurological toxicity, patients in the prospective cohort were examined at baseline (T0), after 2 (T2), 4 (T4), and 6 (T6) months of treatment by clinical neurological examination and a questionnaire‐based assessment. By neurological examination, at baseline, no patient showed signs or symptoms of peripheral neurological impairment. After 2 months of treatment and a median BV dose of 400 mg (range, 260–540 mg), half of patients (N = 5) had reduced tactile sensitivity, 40% (N = 4) paresthesia, 30% neuropathic pain and/or reduced deep tendon reflexes (DTR) (N = 3, respectively), 20% (N = 2) decreased vibratory sensitivity, and only one subject reported grade I neurological toxicity (Figure 1A). After 4 months of treatment (T4) and a median BV dose of 800 mg (range, 520–910 mg), 75% of subjects on total eight evaluable subjects had paresthesia and/or reduced tactile sensitivity (N = 6, respectively), 62.5% decreased DTR and/or vibratory sensitivity (N = 5, respectively), 37.5% (N = 3) neuropathic pain, and only one subject reported grade I neurological toxicity (12.5%). After 6 months of treatment (T6) and a median BV dose of 1200 mg (range, 750–1380 mg), seven out of eight evaluable patients (87.5%) displayed paresthesia, reduced DTR and/or tactile sensitivity, 62.5% (N = 5) decreased vibratory sensitivity, 37.5% (N = 3) neuropathic pain, and only one subject reported grade I neurological toxicity (12.5%).

FIGURE 1.

Neurological toxicity in the prospective cohort. (A) Signs and symptoms of neurological toxicity evidenced by clinical examination after 2 (T2), 4 (T4), and 6 (T6) months of therapy. (B) Medical Research Council (MRC), Douleur Neuropathique en 4 Questions (DN4), Sensory Symptoms Score (SSyS), and Sensory Sum Score (SSuS) values are reported for each timepoint. (C) Scores for the subjective perception of neurological symptoms. DTR, deep tendon reflexes.

Using clinical neurological questionnaires, patients had a mean MRC Sum Score value of 66.3 (range, 58–70) at T2, that tended to remain stable throughout the treatment period (66.3 vs. 63.5 vs. 63.1, T2 vs. T4 vs. T6; p = 0.3396) (Figure 1B). Similarly, mean Sensory Symptoms Score at T2 was 1.8 (range, 0–3) and tended to be stationary throughout the period (1.8 vs. 3.3 vs. 3.0, T2 vs. T4 vs. T6; p = 0.1612), although a slight increase was documented after 4 months of therapy (p = 0.1903). Mean Sensory Sum Score at T2 was 4.9 (range, 0–20) and tended to remain stable throughout the study (4.9 vs. 4.0 vs. 7.3, T2 vs. T4 vs. T6; p = 0.7950), although a slight increase was described after 6 months of therapy, while not statistically significant because of the small number of subjects per group and a great neurological variability between enrolled patients (range, 0–29). We did not observe changes in DN4 in all investigated timepoints (p = 0.5019). In addition, subjective perception of symptoms related to neurotoxicity were investigated using a six‐question survey. Median score value was 11.5/60 at T2, 19/60 at T4, and 17/60 at T6, a median score of 17/60, showing a worsening trend under treatment starting from T4 (p = 0.0513) (Figure 1C).

3.2. ENG Modifications Can Early Predict BVIN

Neurological toxicity was also investigated by ENG evaluation at baseline, T2, T4, and T6, and results expressed as percent variation compared to baseline levels for each reference nerve. Overall, a median percentage reduction in CMAP values of −17.6% was documented at T2 with three subjects (30%) experienced a reduction > 50%, that remained stable at T4 (−12.7%), while greatly decreased to −27.6% reduction at T6 with five subjects (50%) experiencing a reduction > 50%. Similarly, a median percentage reduction in SAP values of −18.6% was described at T2, that remained stable at T4 (−12.7%) with three subjects having a reduction > 50%, while slightly increased to +0.83% at T6, although most patients (N = 5; 63%) showed a reduction > 50%. In particular, for motor nerves, no significant changes were recorded at each timepoint, although reducing trends for the right ulnar motor nerve (p = 0.0677) and for the right peroneal motor nerve (p = 0.0711) between T4 and T6 were observed (Figure 2A). In contrast, an increase in ENG values of the right sural sensory nerve (p = 0.0278) and the left sural sensory nerve (p = 0.0348) was documented between T2 and T6 (Figure 2B). Moreover, an increasing trend in ENG values of the left and right superficial peroneal nerves was observed, especially for the left nerve at T6 compared to baseline (p = 0.1094) (Figure 2B).

FIGURE 2.

Electroneurography in the prospective cohort. Results for (A) motor and (B) sensory nerves after 2 (T2), 4 (T4), and 6 (T6) months of therapy. RMU, right motor ulnar nerve; LMM, left motor ulnar nerve; RMP, right motor peroneal nerve; LMP, left motor peroneal nerve; RMT, right motor tibial nerve; LMT, left motor tibial nerve; RSU, right sensory ulnar nerve; LSM, left sensory medial nerve; RSS, right sensory sural nerve; LSS, left sensory sural nerve; RSP, right sensory superficial peroneal nerve; LSP, left sensory superficial peroneal nerve.

3.3. Interim‐ENG Evaluation Might Reduce the Risk of BVIN and Long‐Term Neurological Complications

Next, to investigate predictive power of neurological and ENG examinations for identification of BVIN development before clinically manifest or irreversible, a control cohort comprising of a total of 12 consecutive patients previously treated with BV alone or in combination with other drugs and never monitored with ENG and neurological examinations during treatment were included (retrospective cohort). In this group, patients were off‐therapy with a median time of 6 months (range, 1–62 months), and they received a median cumulative BV dose of 1100 mg (range, 520‐2250 mg). In details, impaired fine motor skills and paresthesia were found in eight out of 12 patients (67%), reduced vibratory sensitivity in 7 cases (58%), grade I neurological toxicity and/or attenuated DTR in 6 subjects (50%), decreased tactile sensitivity in 4 (33%), and neuropathic pain in 3 cases (25%). Moreover, when compared to the prospective cohort at T6 after 6 months of treatment, decreased tactile sensitivity was found more frequently in the prospective cohort (p = 0.0281), more likely because of more attention given to neurological examination and of more awareness of patients to neurological signs and symptoms. However, grade I neurological toxicity tended to be more common in the retrospective cohort (12.5% vs. 50%, prospective vs. retrospective cohort; p = 0.1577) (Table 2). By clinical neurological questionnaire, no statistically significant differences were observed between prospective and retrospective groups for MRC Sum Score, Sensory Symptom Score, Sensory Sum Score, and DN4, although the latter tended to be higher in the retrospective cohort (p = 0.1018) (Figure 3A). Conversely, using the questionnaire for subjective perception of signs and symptoms of BVIN, median score in the retrospective group was 30/60, tendent lower than that documented in the prospective cohort (p = 0.1898). In the prospective cohort, three subjects reduced BV dose based on interim‐PET, as per international guidelines, while five did not due to disease persistence by PET scan or high‐risk disease. When subjective symptom perception was compared between these two subgroups of subjects and the retrospective cohort, patients who reduced BV doses showed the best neurological status compared to those who did not decrease BV dosage (p = 0.0418) or to the retrospective group (p = 0.0150) (Figure 3B).

TABLE 2.

Neurological signs and symptoms in patients treated with BV‐based regimens.

Symptoms	Prospective cohort (N = 8)	Retrospective cohort (N = 12)	p
Paresthesia, n (%)	7 (87.5)	8 (66.7)	0.6027
Attenuated DTR, n (%)	7 (87.5)	6 (50%)	0.1577
Decreased tactile sensitivity, n (%)	7 (87.5)	4 (33.3)	0.0281
Neuropathic pain, n (%)	3 (37.5)	3 (25)	0.6424
Reduced vibratory sensitivity, n (%)	5 (62.5)	7 (58.3)	> 0.9999
Impaired fine motor skills, n (%)	—	8 (66.7)	> 0.9999
Grade I neurological toxicity, n (%)	1 (12.5)	6 (50)	0.1577

Abbreviations: BV, brentuximab‐vedotin; DTR, deep tendon reflexes.

FIGURE 3.

Neurological toxicity comparison between prospective and retrospective cohorts. (A) Medical Research Council (MRC), Douleur Neuropathique en 4 Questions (DN4), Sensory Symptoms Score (SSyS), and Sensory Sum Score (SSuS) values are reported for the prospective cohort after 6 months of therapy and for the retrospective group. (B) Scores for the subjective perception of neurological symptoms. *p < 0.05.

By ENG examination, overall, a median percentage CMAP variation was +16% compared to the minimum expected values for patient's age, and a median percentage SAP reduction of +3% compared to the minimum expected values for patient's age. In addition, 4 out of 8 patients (50%) showed a CMAP reduction > 50%, and 5 out of 8 subjects (63%) a SAP reduction > 50%. Finally, ENG values for each nerve were compared between the retrospective cohort and the prospective group at T6 (Figure 4A), and significant alterations with higher values were found in the left motor medial nerve (p = 0.0054), and right (p = 0.0310) and left tibial nerves (p = 0.0309) in the retrospective cohort. Similarly, significantly higher values in right and left sural nerves (p = 0.0011 and p = 0.0418, respectively) were documented in the retrospective cohort (Figure 4B).

FIGURE 4.

Electroneurography comparison between prospective and retrospective cohorts. Results for (A) motor and (B) sensory nerves for the prospective cohort after 6 months of therapy and for the retrospective group. RMU, right motor ulnar nerve; LMM, left motor ulnar nerve; RMP, right motor peroneal nerve; LMP, left motor peroneal nerve; RMT, right motor tibial nerve; LMT, left motor tibial nerve; RSU, right sensory ulnar nerve; LSM, left sensory medial nerve; RSS, right sensory sural nerve; LSS, left sensory sural nerve; RSP, right sensory superficial peroneal nerve; LSP, left sensory superficial peroneal nerve. *p < 0.05; **p < 0.01.

4. Discussion

HL is a highly curable hematological malignancy with survival rates exceeding 80% after chemotherapy and/or radiotherapy [18]. Despite the increase in overall survival achieved in the last 60 years, HL patients could experience long‐term treatment‐related toxicities, refractoriness to first‐line therapies, or disease recurrence in approximately 10%–15% of early‐stage disease cases and 15%–30% of advanced stage [19]. Prevention of toxicities is carried out by choosing the best therapeutic strategy that can balance clinical efficacy and long‐term safety, such as reduction of radiation volumes or avoidance of consolidation radiotherapy when possible [20]. Conversely, refractory patients can undergo high‐dose chemotherapy with subsequent autologous stem cell transplantation or can benefit from the use of BV or checkpoint inhibitors as second‐line or bridge therapies to transplantation [21]. Despite the high therapeutic efficacy, over half of refractory HL patients treated with BV may develop grade I–II neuropathy requiring dose reduction in 11% of subjects, which might reduce its clinical efficacy. Indeed, pilot studies have demonstrated that neutropenia and peripheral neuropathy are the most common side effects [22]. Neuropathy tends to spontaneously resolve in over 70% of cases or, in any case, to improve to grade I in the majority of subjects [21]. However, real‐life experience of BVIN in HL patients and its prevention and management strategies are few [23]. For these reasons, we investigated the predictive role of neurological clinical examination, neurological questionnaire, subjective symptom perception questionnaire, and ENG evaluations at 2, 4, and 6 months of therapy in a cohort of refractory HL or NHL patients treated with BV. Based on our results, we proposed a prevention plan by monitoring HL patients using neurological examination and ENG study of sensitive sural and motor tibial nerves after 2 months of therapy to prevent severe and/or irreversible BVIN. Moreover, we suggested that a patient‐based questionnaire for subjective perception of neurological signs and symptoms might be more sensitive and specific than neurological clinical questionnaire to identify BVIN‐related manifestations. Early identification of neurological toxicity is of great importance, as BV dose can be safely reduced before BVIN development in case of permissive interim PET results [3, 24, 25].

BVIN can present as a sensory, motor, and/or autonomic nerve dysfunction, or a combination of these alterations, although it is more frequently a sensory syndrome involving large and/or small unmyelinated fibers [8]. The most frequently complained signs and symptoms are: abnormal tactile perception and/or vibratory sense (80%), numbness and/or paresthesia (70%), tingling (60%), and “sock‐and‐glove” distribution burning (40%) [26]. Neuropathic pain can be present in up to 60% of cases and is considered the most accurate symptom of BVIN in HL patients [8]. Sensory ataxia is observed in the severe form of neurological toxicity and is associated with a higher risk of falls [27, 28]. In our prospective cohort, in accordance with the literature, after 6 months of treatment, the most frequently present signs and symptoms were abnormal tactile perception and/or vibratory sense, and paresthesia, while neuropathic pain was described in 37.5% of cases. Although considered reversible in most cases, reduced vibratory sensitivity and paresthesia were still present in more than 60% of cases in our retrospective cohort, in which neuropathic was reported in 25% of cases and grade I neurological toxicity in half of subjects, also confirming that BVIN might be undiagnosed and more frequent than reported [23]. Our results suggested that neurological monitoring in BV‐treated patients is important to prevent BVIN development, as its reversibility could be achieved in less 15% of affected subjects. Therefore, the best strategy to reduce incidence and severity of BVIN remains prevention and early identification.

The “sural sparing” phenomenon is a common event in Guillain–Barré syndrome, both in axonal and demyelinating forms, and is of diagnostic interest for differential diagnosis of this syndrome. This event occurs because distal neural terminals are preferentially affected by demyelination or nodal immune attack compared to intermediate segments, probably because distal blood–brain barriers are anatomically less efficient [29]. In our HL patients, we observed a similar phenomenon of neural sparing and was present in both cohorts with SAP values of both left and right sural nerves intact or even increased compared to baseline or reference patient's age, especially in the retrospective cohort. SAP values give insights into sensory nerve axon health from distal receptors to dorsal root ganglia, while CMAP into motor nerve fibers from anterior horn cells to terminations within muscle fibers [30]. In previously published works, SAP are more reduced than CMAP in a setting of sensory nerve neurotoxicity, with the “sural sparing pattern” [27]. Underlying mechanisms of sural sparing during BVIN are still under investigation, although could be immune‐mediated as in Guillain–Barré syndrome, because of indirect evidence of responsiveness to immunotherapies, including immunoglobulins and corticosteroids, thus suggesting a subacute inflammatory demyelinating neuropathy [27]. Histological studies of nerve biopsies during BVIN have shown severe loss of myelinated fibers, endoneural edema, and mixed axonal and demyelinating features, that could also be caused by microtubule‐disrupting actions of the cytotoxic agent MMAE [31]. Moreover, impairment of CD30L/CD30 signaling by BV has been proposed as another additional mechanism of neurotoxicity, because this axis is involved in immunoregulation, tumor necrosis factor (TNF)‐mediated responses, and macrophage‐mediated antibody‐dependent cellular phagocytosis, which are biological processes altered during demyelinating disorders [32, 33, 34].

Motor involvement could be found in BVIN patients in association with sensory symptoms, more frequently in distal upper and lower limbs (30%) with loss of fine motor skills, myalgias, and cramps [26]. In our study, chronic motor nerve impairment was documented for the left median, right and left tibial nerves of our retrospective cohort, while CMAP values only tend to decrease in the prospective group, without significant reduction, likely because of the effects of our preventive neurological monitoring during BV administration.

Based on CTCAE definitions, neurotoxicity, including peripheral motor and sensory neuropathy, is classified in V grades of severity: grade I, asymptomatic with clinical or diagnostic observations only; grade II, moderate; minimal, local or noninvasive intervention indicated with limiting age‐appropriate instrumental activities of daily living (ADL); grade III, severe symptoms, limiting self‐care ADL; grade IV, life‐threatening consequences, requiring urgent intervention; and grade V, death [35]. However, there might be multiple conditions between grade I (asymptomatic) and II (moderate with reduced ADL), and grading severity of chemotherapy‐induced peripheral neuropathies has been always a challenge over last decades [36]. Indeed, a grade II paresthesia and objective sensory loss in the fingertips in a young pianist might be more disabling than that considered in WHO grading, and might negatively affect work skills and likely future career prospective in a long HL survivor. Similarly, mild–moderate weakness in a courier or a rider might impede moving and delivering boxes. Conversely, as per drug schedule, BV should be withheld in case of grade II/III neuropathy (when neurotoxicity is already disabling for a young subject or likely probably to be completely irreversible), until the toxicity is less than grade 1 or at baseline and should be reintroduced at reduced dose of 1.2 mg/kg every 3 weeks, while in grade IV neuropathy, BV should be discontinued [37]. However, BV dose can be safely reduced if the interim PET is permissive (D1–D2), as its reduction of the expected cumulative dose has no negative impact on progression‐free survival and overall survival [3], with results like those observed in the pivotal ECHELON‐1 study [10, 11]. In our study, a 25% BV dose reduction was performed in three subjects, based on a permissive interim PET, with significant improvement in subjective perception of neurological symptoms.

Our study has several limitations: (i) the limited number of enrolled patients, because of the monocentric nature of this study and because of the very low incidence of relapsed/refractory HL; (ii) absence of moderate–severe BVIN for comparison of studied variables; (iii) a small number of patient per group that did not allow identification of predisposing factors and investigation of sensitivity and specificity of neurological examination in combination with ENG studies for early detection of BVIN; and (iv) absence of a standardized questionnaire for evaluation of subjective perception of neurological toxicity.

In conclusion, HL patients can experience long‐term toxicities related to treatment, and BVIN is more frequent than reported, as we described a rate at 12.5% in the prospective cohort and at 50% in the retrospective group. Prevention of the onset of severe and/or irreversible BVIN is still an unmet need, while of great importance as affected patients experience difficulties in their daily life and work. Therefore, close monitoring could represent the best prevention strategy to early detect clinical and electrophysiological signs of BVIN, which can guide medical decisions to safely reduce or discontinue BV based on interim‐PET scan results. Periodic neurological examination and ENG at 2 months of therapy of sensitive sural and motor tibial nerves, together with a simple questionnaire for subjective perception of BVIN‐related signs and symptoms could represent an effective strategy for reducing incidence and severity of neurological toxicities in BV‐treated patients. However, our data should be confirmed in larger prospective cohorts and using validated questionnaires, as BVIN might become even more frequent and invalidating with the introduction of BV in first‐line approaches and without using dedicated and more sensitive grading severity scores for chemotherapy‐induced peripheral neuropathies.

Author Contributions

Conceptualization: L.P. and C.S. Data collection: L.P., L.S., G.P., and S.A. Methodology: L.P., L.S., G.P., S.A., G.D.B., C.M.N., A.L., and C.V. Clinical data: L.P., L.S., B.S., M.R., M.D.A., and V.G. Data analysis: L.S. and V.G. Writing – original draft preparation: L.S. and V.G. Writing – review and editing: P.B. and C.S. All authors have read and agreed to the published version of the manuscript.

Disclosure

The authors declare that the material is original, has not been published before nor is under consideration in any journal.

Ethics Statement

Protocol approved by local ethic committee (Ethics Committee “Campania Sud,” Brusciano, Naples, Italy; prot./SCCE n. 24988).

Consent

Patients received informed consent obtained in accordance with the Declaration of Helsinki (World Medical Association 2013) and protocols approved by local ethic committee (Ethics Committee “Campania Sud,” Brusciano, Naples, Italy; prot./SCCE n. 24988).

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the authors upon reasonable request.