Patient-reported outcomes in chemotherapy-induced peripheral neuropathy: a review

CIPN is defined as the damage to the peripheral nervous system experienced by patients receiving neurotoxic chemotherapies. The symptoms of CIPN range from burning, stabbing, pricking, tingling, or numbness in the toes and fingers to generalized symptoms that may be immobilizing.^{[1], [2] and [3]} While each chemotherapy agent has its own mechanism of action, CIPN generally affects the fingers and toes first; and with cumulative doses, symptoms progress proximally to the hands and feet, arms and ankles, and then calves (stocking-glove distribution).^{[3] and [4]} Sensory symptoms and signs typically develop before motor symptoms, and a subset of patients also suffer from neuropathic pain. CIPN occurrence estimates have ranged from 10% to 92% and vary depending on the chemotherapy regimen, dosing, and comorbidity. Patients with preexisting peripheral neuropathy are more likely to develop a severe and persistent CIPN compared to those without any history of neuropathy.⁵

Although several agents show promise in preventing CIPN, the evidence of the efficacy of these options is mixed, and treatment of CIPN is still largely symptomatic. Corticosteroids, antidepressants, anticonvulsants, and analgesics are potential options available to clinicians, although none has shown significant efficacy in treating CIPN symptoms. As a result, in the event of significant neuropathy, dose reduction or discontinuing chemotherapy is the only option to stop and reverse progression of peripheral neuropathy; and such steps potentially could impact overall cancer treatment outcomes.^{[6], [7] and [8]} Alternatively, patients may choose to continue chemotherapy with compromised HRQL; specific reductions in HRQL scores due to CIPN symptoms have been estimated to be from 15% to 20%.⁹

A key issue for clinicians is understanding the extent of functional impairment a given patient may have when experiencing CIPN. Although physicians often are able to diagnose severe symptoms more accurately relative to less severe CIPN symptoms, Postma and colleagues¹⁰ and other researchers have shown that there is substantial discrepancy in severity grading among physicians as well as between patients and physicians, with physicians often underestimating the severity of symptoms. Contributing to this problem is the lack of consistency in diagnostic tools used, with no gold standard for CIPN measurement. This review therefore examines the symptoms and impacts of CIPN from the patient perspective, identifies concepts and issues that should be considered when evaluating CIPN, and compares existing patient-reported outcomes (PRO) measures that have been used in CIPN assessment.

Methods

The review of the literature, performed in January 2009, focused on abstracts or articles that assessed the HRQL of CIPN over the past 5 years. The review included six article databases (OVID, which includes Medline and Embase; Proquest digital dissertations; CINHAL; IPA; Health and psychological instruments; Psycharticles) and five conference proceedings (ASCO, AACR, NCRI, San Antonio Breast Cancer Symposium, Annual Symposia of the American Society of Breast Disease). The comprehensive search combined the clinical terms “peripheral neuropathy,” “neurotoxicity,” “paresthesias,” “dysesthesias,” etc. with the term “chemotherapy” and with specific chemotherapy drugs (eg, cisplatin, paclitaxel, vincristine). These terms were further combined with quality of life (eg, preference, quality of life, patient perspective, daily activities). Only those studies focusing on neuropathy related to chemotherapy were included in this review; studies of diabetes, acquired immunodeficiency syndrome (AIDS), B₁₂ deficiency, or arthritis-related neuropathy were excluded.

A total of 5,671 studies across the article databases and 355 conference abstracts underwent initial review. After removing duplicates, commentaries, review articles, and articles discussing neuropathy associated with diabetes, human immunodeficiency virus (HIV), and rheumatoid arthritis, 41 potentially relevant abstracts were identified, and full-text article paper copies were obtained for these studies. The final review then identified 31 articles and 10 relevant conference abstracts to be incorporated.

Results

The findings from this review are discussed in three sections: (1) qualitative studies, (2) patient preference studies, and (3) PRO measures used to assess CIPN outcomes.

Qualitative Studies

Four qualitative studies were identified in which CIPN was discussed.^{[11], [12], [13] and [14]} Bakitas¹¹ evaluated CIPN experience among patients of all cancer types, while the other three studies examined the effects of a range of chemotherapy side effects in disease-specific populations, including breast cancer¹² and colorectal cancer.^{[13] and [14]}

In summary, the qualitative studies showed that CIPN can lead to a range of physical and emotional impacts. Patients with CIPN indicated that they experienced tingling, burning, numbness, “pins and needles,” and shock-like or painful sensations bilaterally in the feet and/or hands. Patients indicated that of all chemotherapy-related adverse events, CIPN was one of the least expected and most distressing and disabling.^{[11], [12] and [14]} Upper-extremity neuropathy interfered with dressing, cooking, sewing, household work, and leisure. Problems with dressing included difficulty putting on undergarments or jewelry. Neuropathy in the lower extremities caused problems with walking (“I walk like I'm drunk,” “I feel clumsy”) and was accompanied by problems with balance, ambulation (hiking, running, biking, and standing for long periods), and driving.^{[11] and [12]} Among breast cancer patients on paclitaxel, peripheral neuropathy was considered to be more incapacitating than bone pain because it affected daily functioning, whereas bone pain generally did not.¹²

The symptoms of CIPN not only resulted in physical distress and loss of functional ability but also affected emotional health and social interactions. Patients expressed that they were often ill-prepared to cope with CIPN symptoms. The functional impairments (ie, inability to pursue daily activities) affected patients' mood and often made patients feel dependent, disabled, and helpless. Subjects reported social isolation when they could not drive, walk, or stand for periods of time.¹¹

Patients also expressed difficulty in describing symptoms, which were often described as “funny,” “strange,” or “weird.” Without the ability to communicate symptoms to their doctors, some patients expressed a feeling of frustration that their doctors did not take their symptoms seriously.¹¹ Patients also felt disappointment about their lack of understanding of the significance of some CIPN-related symptoms over others and the prolonged duration for symptom resolution.^{[12], [13] and [14]}

Nevertheless, patients generally could keep CIPN “in the background,” although it could be “annoying,” “distracting,” and “unpleasant.”¹¹ Despite the many disruptions associated with CIPN, patients expressed reluctance in discontinuing chemotherapy and generally adapted and “learned to live with” their CIPN.^{[11], [12] and [14]} As observed by Calhoun and colleagues,¹⁵ who conducted a patient preference study of chemotherapy toxicity states, patients appear to be willing to tolerate more toxicity in expectation of better chemotherapy outcomes.

Patient Preference Studies

Four patient preference studies echoed sentiments expressed in the qualitative studies regarding apprehension about chemotherapy-induced neurotoxicities. In general, the patient preference studies reported that peripheral neuropathy was viewed as worse than a number of other chemotherapy-induced toxicities. In a study in which ovarian cancer patients rated selected toxicities on a visual analog scale where 0.0 reflected the worst chemotherapy scenario and 1.0 reflected an ideal chemotherapy scenario, “peripheral neuropathy” was rated as 0.45 and was considered worse than alopecia, pancytopenia, and fatigue.¹⁶

Another study that included ovarian cancer patients used the time trade-off technique to assess preferences for different health states representing five different levels of severity each for neurotoxicity, nephrotoxicity, and ototoxicity. Whereas the two more mild levels of these toxicities were rated similarly, the patients consistently rated the three moderate to severe levels for neurotoxicity as worse than those for nephrotoxicity and ototoxicity. Patients with previous toxicity experience rated all toxicity scenarios worse than those without any previous experience.¹⁵

Similarly, Dranitsaris¹⁷ elicited rankings from the Canadian general population with respect to the importance of reducing the chance of febrile neutropenia, neurotoxicity, and renal toxicity. The general population felt that it was more important to reduce the chance of neurotoxicity versus renal toxicity. In a more recent study, Dranitsaris and colleagues¹⁸ assessed the perspectives of oncology pharmacists and nurses who served as proxies for patients regarding grade 2/3 motor neuropathy, grade 2/3 sensory neuropathy, grade 3/4 neutropenia, and grade 3/4 anemia. Of these four toxicities, the clinicians rated motor neuropathy and sensory neuropathy the most unpleasant. The clinicians preferred docetaxel in place of paclitaxel, the former being associated with smaller chances of sensory neuropathy and motor neuropathy but not neutropenia or anemia. Based on qualitative feedback, the respondents explained that they had long-term concerns about neuropathy because it could not be effectively managed but that neutropenia and anemia were short-term, treatable events.¹⁸

PRO Measures for CIPN Evaluation

Generic PRO measures focus broadly on functional status and can also be used to compare physical and mental health functioning across patient populations. However, they may not be sensitive enough to capture impacts of specific health conditions. This lack of sensitivity to disease-specific impacts has been observed in a number of oncology studies with respect to the occurrence of CIPN. Specifically, the generic measures the European Organization for Research and Treatment in Cancer—Quality of Life Questionnaire (EORTC QLQ-C30) and the Functional Assessment of Cancer Therapy (FACT)—General typically do not detect differences between treatment groups, although the incidence and severity of neurotoxicity may differ significantly as measured using National Cancer Institute Common Toxicity Criteria (NCI-CTC) or other criteria.^{[6], [19], [20] and [21]} The lack of sensitivity of generic measures to treatment group differences in CIPN impacts likely is attributable to insufficient numbers experiencing neurotoxicity-related impediments to daily functioning and/or to the imprecision of these measures.^{[19] and [20]}

However, CIPN-specific measures or items do show sensitivity to treatment differences in CIPN occurrence and impacts. Seven PRO measures have been used to specifically assess CIPN outcomes (Table 1).