The main findings of the present study are: 1) nearly all cancer survivors referred for suspicion of autonomic dysfunction tested positive; 2) dysfunction was demonstrated in all domains of the autonomic nervous system, with particular derangement in the sympathetic system; and 3) patients showed subjective and/or objective symptom response to treatment. These data suggest that cardiovascular autonomic dysfunction may be more prevalent than previously recognized in cancer survivors, and should be screened for, tested, and treated when appropriate.
The results of the autonomic reflex testing suggest that cardiovascular autonomic dysfunction may result predominantly from a systemic process, such as chemotherapy or paraneoplastic syndrome, with a different etiology than sensory peripheral neuropathy caused by many chemotherapy agents. Hematologic disorders, hematopoietic stem cell transplantation, poor functional status, vinca alkaloids, alkylating agents, anthracyclines, chest radiation, and neck radiation were associated with autonomic dysfunction. The majority of patients were diagnosed with orthostatic hypotension, POTS, or IST and had subjective improvement with personalized treatment. To the best of our knowledge, this is the first study to report these findings in cancer survivors with suspected autonomic dysfunction. Clinicians caring for these patients should have a low threshold to refer for autonomic evaluation.
Cardiovascular autonomic dysfunction is commonly seen in patients with neurodegenerative disorders and diabetes mellitus . In diabetes mellitus, the mechanism is thought to be due to metabolic and ischemic damage to both the sympathetic and parasympathetic systems . In Parkinson’s disease and multiple system atrophy, the underlying mechanism of autonomic impairment is thought to be due to sympathetic denervation of the heart and vasculature [18, 19]. The pathologic mechanism of cardiovascular autonomic dysfunction in cancer survivors is poorly understood . In this study, both the parasympathetic and sympathetic arms demonstrated dysfunction, with a greater degree of impairment in the sympathetic nervous system. This phenotype is similar to that seen in multiple system atrophy. Diabetes and Parkinson’s disease tend to have more proportionate distribution of impairment between the sympathetic and parasympathetic arms [20,21,22].
The patients in this study also had derangements on sudomotor testing, which assesses the integrity of post-ganglionic sympathetic nerves. Moreover, the degree of sudomotor impairment tended to increase with increasing levels of sympathetic and parasympathetic dysfunction. This suggests that autonomic dysfunction in cancer survivors may result from a diffuse process, such as chemotherapy or paraneoplastic effects, as opposed to a local process, such as radiation therapy, tumor invasion, or inflammation surrounding the malignancy.
The heterogeneity in age and time to autonomic testing may reflect a referral bias due to the non-specific nature of these symptoms [5, 12]. It is plausible that patients who were sent for autonomic reflex testing later in their cancer survivorship were symptomatic for an extended period of time. In this case, testing may have been delayed due to low suspicion for autonomic etiology, accessibility issues, or lack of familiarity with autonomic reflex testing. Another explanation is that these findings represent multiple mechanisms of neurotoxicity that vary in time from original insult to clinical presentation. Moreover, patients with obvious, severe autonomic dysfunction or subtle autonomic dysfunction may not have been referred for autonomic reflex testing due to lack of perceived diagnostic utility or low clinical suspicion for autonomic etiology, respectively.
The majority of patients (75%) had hematologic disorders, which was disproportionately higher than the 9.5% estimated prevalence among cancer survivors in the United States . Many of the chemotherapy regimens used to treat these disorders, such as CHOP (cyclophosphamide, doxorubicin, vincristine, prednisone) or ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine), include anthracyclines and vinca alkaloids. Given the diffuse pattern of autonomic impairment seen on reflex testing, the association between hematologic malignancies and autonomic dysfunction may be due to paraneoplastic effects of the malignancy or the chemotherapies used in those patients, as opposed to a more localized process such as direct tumor invasion.
Eight of the patients received hematopoietic stem cell transplantation. Deuring et al. found that patients who received hematopoietic stem cell transplantation had higher degree of autonomic dysfunction compared to controls . This may suggest autonomic impairment was either due to graft versus host disease or effects of the myeloablative regimen. Most patients had malignancies either in the neck or thorax or had received radiation treatments in those areas. While the results of the autonomic reflex testing suggest a more diffuse neuropathy, it is possible that local tumor-related effects had some effect given that the autonomic baroreflex pathways are located the neck and thorax as well.
The effects of chemoradiation on peripheral sensory nerves are well established . Given the similarities in myelination between peripheral sensory and autonomic nerves, these mechanisms have been extrapolated to be possible explanations for autonomic dysfunction . The most common chemotherapy agents in this study were vinca alkaloids, alkylating agents, and anthracyclines. There is literature suggesting associations between these agents and autonomic impairment. Microtubule inhibitors, such as vinca alkaloids, are a well-known cause of peripheral sensory neuropathy . The effect on the autonomic system was evaluated by a Roca et al., who found that vinca alkaloids treatment altered orthostatic blood pressure and heart rate with deep breathing . While alkylating agents are not commonly associated with neuropathy, a study by Dobrek et al. demonstrated a relationship between alkylating agents and alterations in heart rate modulation . The two patients in our study who received alkylating agents and did not receive microtubule inhibitors, anthracyclines, or platinum-based agents had severe generalized autonomic impairment most prominent in the adrenergic and sudomotor systems. Anthracyclines have well-established association with cardiotoxicity, but not necessarily with neuropathy. Anthracyclines have been associated with abnormalities in heart rate variability . In patients with diabetes, there is some evidence that cardiac autonomic dysfunction is associated with non-ischemic cardiomyopathy . The one patient who received anthracyclines, but did not received microtubules or alkylating agents, had moderate autonomic dysfunction with severe adrenergic impairment.
The association between chemotherapy agents, such as anthracyclines and alkylating agents, that are not traditionally associated with neuropathy and cardiovascular autonomic dysfunction suggest that autonomic impairment in cancer survivors may be caused by a different mechanism than the sensory peripheral neuropathy seen with many chemotherapy agents. Moreover, given the number of patients in this study that had cardiomyopathy or received anthracyclines, it is possible that autonomic dysfunction is a precursor to, or at the very least, associated with cardiomyopathy .
Cardiovascular autonomic dysfunction in cancer survivors is a poorly understood condition that carries significant morbidity and is associated with an increase all-cause mortality . In this study, most patients had either orthostatic hypotension, POTS, or IST, which are recognized by the Heart Rhythm Society as autonomic diseases . Given our limited knowledge on cardiovascular autonomic dysfunction in cancer survivors, there are currently no published guidelines on treatment options . Evidence-based pharmacologic treatment strategies for autonomic dysfunction are largely based on studies of patients with neurodegenerative diseases, diabetes mellitus, and infiltrative diseases . The effectiveness of treatment is generally varied and often poor. In this cohort, the results of autonomic reflex testing were used to tailor personalized, autonomic treatment to each patient. The majority of patients had subjective improvement in treatments, which further demonstrates the utility of this tool.
This study also exemplifies the role of a cardio-oncology program in identifying, preventing, and managing the unique cardiovascular comorbities, such as cardiovascular autonomic dysfunction, faced by this population to improve their overall quality of life. Physicians caring for these patients should have a low threshold to refer for autonomic evaluation, particularly for patients with hematologic malignancies, tumor burden in the neck or chest, cardiomyopathy, or those who have received vinca alkaloids, alkylating agents, anthracyclines, or hematopoietic stem cell transplantation. Autonomic reflex testing may suggest personalized treatment options to improve outcomes and provide an objective measure to assess response to treatment and progression of disease.
This study has several limitations. First, as this study was a retrospective, descriptive case series, statistical hypothesis testing and control groups were not used given the small sample size and large number of variables (cancer types, chemotherapy agents, etc.). There is a selection bias as only symptomatic patients were referred for autonomic reflex testing, which represents a select subset of the total cardio-oncology patient population at this institution. However, given the significantly limited data that exists in this population, particularly among symptomatic patients rather than patient with incidentally found autonomic dysfunction, we believe this information provides useful data for the field. Moreover, there is also a referral bias as the proportion of different malignancies managed and the specific treatment regimens used for different cancers vary per institution. Additionally, as cancer treatments were composed of multi-drug regimens, individual associations were difficult to identify. Furthermore, given the difficulty in determining equipotency between doses and routes of administration between different chemotherapeutic agents, medication use was classified by the presents of administration rather than by cumulative dosing. Autonomic treatments were personalized, which led to a lack of standardization with regard to treatment modality. However, given the paucity of data on this topic, we believe this study provides novel insights into the diagnosis of and management of dysautonomia associated with cancer treatment.