The Impact of Palliative Care Dose Intensity on Outcomes for Patients with Cancer

Short abstract

This commentary discusses evidence supporting the importance of full palliative care dose intensity in outcomes for patients with cancer.

Every oncologist knows the importance of “dose density”—the mg drug(s)/time supposed to be given [1]. For example, we give 1.4 mg/m² vincristine every 3 weeks for “CHOP” chemotherapy. If a practitioner is afraid of side effects and reduces the dose or lengthens the interval, the dose intensity goes down—as does the effect [2]. Patients receiving less than 65% of standard dose intensity adjuvant breast cancer regimens have survival similar to that of untreated controls, suggesting the entire benefit of adjuvant chemotherapy may be lost [3].

We believe there is also a “dose‐intensity” aspect to palliative care (PC) that is not addressed in current clinical practice. To best define palliative care dose intensity, we must consider this important question posed by Dr. Kamal years ago: “If palliative care were a drug, how would one write the prescription? What would be the specific components of its delivery?” [4]. One set of experts went as far as to say we needed PC blood levels [5].

The introduction of outpatient PC to oncology is increasingly recognized as important, and progress has been made to increase the availability of services. Unfortunately, as an intervention for patients with cancer, palliative care lacks a standardized definition or approach. In the earliest trials, the intervention of early palliative care encompassed communication about serious illness and prognosis, symptom assessment and management, and regular follow‐up (Fig. 1) [6, 7, 8]. The palliative care intervention was “early” if administered within 8 weeks of diagnosis [9], whereas comparator arms generally consisted of usual oncology care, allowing referral to a palliative care team at any time along the disease trajectory. We are growing more familiar with final outcomes, as patients randomized to expert palliative care alongside oncology care had improved quality of life, better mood and symptom control, reduced resource utilization, less aggressive end‐of‐life care, and maybe even an increase in survival. Although we believe palliative care is valuable and worthy of endorsement, it is important to reflect on the impact of palliative care “dose” and “dose intensity” that produced these outcomes (Fig. 1).

Figure 1.

Ambulatory palliative care guidelines used in the original study by Temel and colleagues [35].

Why is it important to revisit the framework established in the earliest clinical trials? First, the evidence for the effects of late palliative care is equivocal because the time required to establish benefit is too short [10, 11]. Palliative care must be applied early, within 8 weeks of a diagnosis, to maximize patient benefit [9, 12, 13]. Unfortunately, in clinical practice, many patients with cancer have variable or delayed access to palliative care, or do not receive palliative care services at all [14]. Second, clinical care teams and health systems have underappreciated the quality of the palliative care intervention. Despite current practice paradigms, palliative care is not a widget that can be made by anyone with some training and tools. In all of the major randomized trials, the palliative care intervention was multidimensional, targeting at least the physical and psychologic domains, if not all of the quality of life domains suggested by the Consensus Project [15].

When Temel et al. organized the first landmark randomized clinical trial of palliative care for patients with lung cancer [16], it was completed at one site—the Massachusetts General Hospital (MGH), with some of the most experienced PC practitioners in the U.S. Patients and families were seen for nearly an hour every month, averaging four visits over 12 weeks [13, 17, 18]. The results were spectacular: less anxiety and depression, better mood, less use of likely ineffective chemo, and 2.7 months' increased survival. In other early studies showing major benefit [19], including improved survival [20], the intervention was a manualized, telephone‐based format done by two expert advanced practice nurses with four initial in‐person meetings followed by monthly follow‐up sessions.

It is not clear that expert PC methodology can be scaled up, unless we follow the regimen of the first, most successful trials. In the replication trial by Temel et al., among patients with gastrointestinal and lung cancer, there was less effect of PC [21]. This trial was implemented at MGH also, with an expanded group of practitioners, with the same monthly visits. The study required a PC advanced practice nurse or medical doctor with board certification and 6 months' experience. The primary endpoint, improvement in quality of life by 12 weeks, was not met, but there was improvement at 24 weeks as reported in prior trials. Patients in the intervention group were more likely to discuss end‐of‐life wishes with their oncologist, 30.2% versus 14.5%. Unlike the first trial, there was little or no difference in mood or anxiety for the overall cohort. Unexpectedly, patients with lung cancer reported an improvement in mood and depression, but patients with gastrointestinal cancer did not. In the most recent randomized cooperative group trial by Temel et al., performed at 18 academic and community sites, there was minimal—if any—effect [22]. There were attempts to use the same standardized methods for fidelity, but this was not verified. In fact, 15% of “palliative care patients” never saw palliative care by the end of week 12, and the mean number of visits by 24 weeks was only 4.4. In addition, the patients were older and sicker and only about half completed the forms, limiting the statistical power and ability to draw conclusions from this study. It is important to note that reliance on oncologist referrals for palliative care led to late referrals in this trial, a longstanding issue in palliative care delivery [20, 23, 24]. Some of us are old enough to remember that before we could join the Peters et al. randomized comparison of high‐dose chemotherapy with stem‐cell support versus intermediate‐dose chemotherapy (CALGB 9082, SWOG 9114, and NCIC MA‐13), “teams had to demonstrate their proficiency by treating (successfully) at least three patients” [25]. Perhaps the same should be required for PC trials.

What has been perhaps the most underappreciated aspect of the Temel et al.. and Bakita et al. landmark studies is the palliative care dose actually delivered, compared with that generally established in daily practice: “Patients who were assigned to early palliative care met with a member of the palliative care team, which consisted of board‐certified palliative care physicians and advanced‐practice nurses, within 3 weeks after enrollment and at least monthly thereafter in the outpatient setting until death” [6; emphasis added]. In the community and in hospital‐based practice, there is often insufficient infrastructure to implement an outpatient co‐management program between specialty oncology and consultative palliative care, especially one that sees the patient on the treatment day. With drug chemotherapy, drug concentration correlates with tumor cell kill and inversely with the rate of tumor growth [2]. Does the intensity of a palliative care intervention correlate with symptom management? Moreover, is a low‐dose‐intensity infrastructure inversely related to quality of life?

Although most oncologists find the evidence convincing for maintaining chemotherapy dose intensity, many studies and anecdotal experiences suggest that considerable disparity exists in the delivery of palliative care dose intensity. These reductions may represent a major reason for “treatment failure” in the studies previously mentioned. In fellowship, oncologists are often taught that study design should inform what constitutes best practice. The central design element in intervention studies is dose, including the number of sessions, length of each session, and the interval between sessions [26]. Unfortunately, discrepancies in provider training, treatment delivery, treatment receipt, and treatment skills are evident in clinical practice as it relates to palliative care delivery. These shortcomings are likely due to differences in time availability (how many oncologists have an hour to spare, every month—with every patient?) and intensive training in communication and supportive care [28].

Palliative care providers are specialty trained to skillfully navigate difficult conversations, support patients through physical and emotional symptoms, and coordinate transitions of care. These aspects of patient care are necessarily high‐touch and cannot be addressed in a single visit [29]. Symptom assessment evolves over time, goes beyond what is physically burdensome, and includes elements of mental health, coping, spirituality, prognostic understanding, and goals of care. These assessments not only facilitate meaningful therapeutic interventions but assist in improving the accuracy of survival predictions [30]. Simply put, dose‐dense schedules of palliative care using a standard template such as used by the Temel et al. and Bakita et al. groups (Fig. 1), are required to support patients with advanced cancer.

Although we embrace the growth of palliative care nationwide, it is not good enough to require hospitals or cancer clinics to simply have palliative care. Similarly, for front‐line clinicians, having a patient referred to and seen by palliative care is not a panacea. Strong evidence shows that patients with early and advanced malignancies benefit from the delivery of full‐dose palliative care. Reducing the expertise and dose intensity through insufficient infrastructure [31], limited providers [32], late referrals [13], or lack of prioritization [33] may increase the risk of patient suffering and poor outcome [34]. Palliative care dose intensity, once defined, can represent an important quality measure allowing practices to benchmark their experience with that of other practices—like the original Massachusetts General Hospital experience. Most importantly, health system executives must decide to make the investment in appropriate specialized structures.

Although further controlled clinical trials are needed on the role of dose‐dense strategies (similar to Temel and colleagues stepped palliative care vs. early integrated palliative care in patients with advanced lung cancer—NCT03337399), oncologists and health systems should seriously consider the compelling evidence supporting the importance of full palliative care dose intensity in outcomes for patients with cancer. We may have otherwise diluted what is arguably the greatest intervention for patients with cancer in recent years, to a meaningful use checklist.

Disclosures

The authors indicated no financial relationships.