Intravenous Ketamine for Cancer Pain: A Single-Center Retrospective Analysis Comparing Fixed-rate versus Weight-based Dosing

Abstract

Although weak evidence exists to support subanesthetic ketamine for cancer pain treatment, successful use may be hindered in the absence of standardized dosing guidance. We aimed to compare the success rates of intravenous ketamine fixed-rate versus weight-based dosing strategies for cancer pain treatment, and to assess patient characteristics that correlate with treatment success. We conducted a single-center retrospective review including non-critically ill adults with cancer pain who received subanesthetic ketamine for at least 24-hours. All patients received fixed-rate ketamine; weight-based doses were retrospectively determined using total body weight. Treatment was considered successful if after reaching the maximum prescribed ketamine dose the patient had a 30% reduction in: baseline pain score, as-needed opioid use, or total morphine equivalent daily dose over a standardized 24-hours. Of 105 included patients, 51(48.6%) successfully responded to ketamine. Responders had lower fixed-rate ketamine doses compared to non-responders (median[IQR] 15 mg/hr[10–15] vs. 15 mg/hr[15–20], p=0.043), but no difference in retrospectively calculated weight-based doses (0.201±0.09 mg/kg/hr vs. 0.209±0.08 mg/kg/hr, p=0.59). Responders had higher daily opioid requirements at baseline compared to non-responders (p=0.04). Though underpowered, our findings suggest that weight-based ketamine dosing may not convey additional benefit over fixed-rate dosing.

Introduction

Up to 90% of patients with cancer report pain during their disease course with 30% of this population reporting that their pain has not been adequately treated (1,2). Physical pain may develop from both disease progression or cancer-related treatments including chemotherapy, radiation, and surgery (2). Due to the complexity of cancer-related pain, non-opioid adjuvant analgesics are commonly employed to target different mechanisms along the pain pathway.

Subanesthetic parenteral ketamine is used for the management of chronic and acute pain, but there is limited evidence to support its place in therapy for cancer-related pain (3–6). As a non-competitive antagonist of the N-methyl-D-aspartate (NMDA) receptor, ketamine is thought to block amplification of pain signals, thereby limiting central pain sensitization, opioid tolerance, and opioid induced hyperalgesia (6–10). Ketamine’s unique mechanism of action allows it to effectively treat pain in addition to reducing patients’ opioid requirements (10–12).

Studies assessing subanesthetic parenteral ketamine for cancer-related pain show conflicting results (5,6,8,10–19). Heterogeneity in study results may be partly explained through variations in dosing strategies: both weight-based and fixed-rate dosing strategies have been used based on institution preferences (5,6,10,19). Fixed-rate continuous doses commonly start with 5 mg/hr which may be titrated up to 40 mg/hr; conversely, weight-based ketamine regimens typically start at 0.1 mg/kg/hr which may be titrated up to 0.5 mg/kg/hr (6,8,10,12,13,16,18). Patients greater than 70 kg may receive higher maximum ketamine doses when receiving weight-based ketamine compared to fixed-rate. To date, there have been no studies evaluating optimal dosing strategies for continuous intravenous ketamine to control pain in patients with cancer.

The evidence supporting ketamine’s use in pain management is strongest in patients with predominantly neuropathic pain (4,6). This in tandem with ketamine’s unique side effect profile may prompt clinicians to take an individualized approach when weighing the risks and benefits of therapy. Further research is warranted when considering what patient and disease characteristics correlate with the successful use of ketamine therapy.

The aim of this study was to determine optimal dosing strategies for intravenous ketamine for pain management in patients with cancer (fixed-rate vs. weight-based dosing) and to identify any patient or disease characteristics that may correlate with ketamine success.

Materials and Methods:

Design and Participants:

This was a single-center retrospective cohort analysis of adult patients with cancer-related pain admitted to a comprehensive cancer hospital at an academic medical center during a 5-year period from September 2016 through August 2021. Study conduction was approved by The Ohio State University Cancer Institutional Review Board, 2021C0167, with waiver of consent process and full waiver of HIPAA research authorization. All patients had an inpatient palliative medicine consult and received continuous subanesthetic intravenous ketamine infusion for the management of pain for at least 24-hours. For patients who were prescribed intravenous ketamine for pain management at several intervals throughout admission, only the first episode was assessed. Patients were excluded if incarcerated, pregnant, oral ketamine received prior to or during ketamine infusion, ketamine prescribed for non-pain indication, ketamine used in the intensive care unit, or concurrent neuraxial analgesia or nerve blocks. Furthermore, patients were excluded if ketamine was initiated within 24-hours of admission as the preceding 24-hour period was required to determine baseline pain and opioid requirements.

Outcomes:

The primary outcome was overall response rate to ketamine infusion, defined as a composite of efficacy: 30% reduction in patient or nurse reported numerical pain scores (reported as numerical scale 0–10), 30% reduction in total oral morphine equivalent daily dose (MEDD), or 30% reduction in as-needed oral morphine equivalents (OME) administration after the maximum ketamine dose had been reached (4).OME was determined using our institution’s guideline. (Table 1) Patients were defined as having successfully responded to ketamine (responders) if they met any of the objective measures within the primary outcome following maximum ketamine dose. Secondary outcomes included individualized assessment of each primary outcome and assessment of ketamine-associated adverse drug reactions. Patients were separated into the cohorts of responders and non-responders based on the composite outcome for additional analysis.

Table 1:

Opioid Conversion Table

Drug	Intravenous Dose (mg)	Oral Dose (mg)
Morphine	10	30
Hydromorphone	1.5	7.5
Oxycodone	---	20
Oxymorphone	---	10
Hydrocodone	---	30
Codeine	120	200
Fentanyl	0.1	---
Transdermal fentanyl dose	Oral Morphine Equivalent (OME)
12.5 mcg/hour	30 mg/day
25 mcg/hour	60 mg/day
50 mcg/hour	120 mg/day
75 mcg/hour	180 mg/day
100 mcg/hour	240 mg/day

Ketamine Administration:

All patients received continuous intravenous ketamine following institutional fixed-rate dosing guidance in which ketamine is initiated at 5 mg/hr and may be titrated in increments of 5 mg/hr every 8-hours to a maximum recommended analgesic dose of 30 mg/hr. Pain was assessed by nursing staff at a minimum of every four hours per institutional policy using the Defense and Veteran’s Pain Rating Scale (DVPRS) for adult patients with appropriate cognitive function, or the Pain Assessment in Advanced Dementia (PAINAD) for patients with dementia, low level cognitive functioning, or unable to use the DVPRS.

Data Collection:

Last recorded total body weight prior to ketamine initiation was collected and used to retrospectively determine weight-based dosing. Patient’s age at admission, sex, race, cancer diagnosis, concurrent cancer-related therapies, concurrent adjuvant medications, and comorbidities, if applicable, were recorded. These factors were assessed to determine if a correlation exists with ketamine success.

Outcomes were assessed at standard 24-hour intervals (0700-0700) after the maximum ketamine dose was reached. Baseline pain scores and opioid use were determined from the 24-hours preceding ketamine initiation. All opioids were converted to OME per the institutional equianalgesic opioid dosing guideline. Methadone was converted to OME using a 1:5 ratio (20,21). Daily OME were not extrapolated for partial days of admission. Buprenorphine products were not included in the daily OME calculation due the variability of equianalgesic conversions and conflicting research regarding buprenorphine’s unique analgesic properties and pharmacokinetics (22,23). Patients receiving buprenorphine monotherapy were reported separately. The indications of buprenorphine and methadone (pain vs. substance use disorder) were not collected. Electronic notes were manually reviewed for any documented adverse drug reaction associated with ketamine use.

Statistical Analysis:

This study was powered for a logistic regression to have 80% power at a sample size of 200 to detect an odds ratio of 0.66 for a one standard deviation increase in weight-based dose, assuming a 5% type 1 error rate and a 60% proportion with the efficacy outcome.

Patient demographics and clinical characteristics (i.e. cancer type and concurrent medications) were summarized as counts and percentages and compared across outcomes using Fisher exact tests. Categories of baseline MEDD were compared across the primary outcome using Cochran-Armitage trend tests. Continuous variables were summarized using mean ± standard deviations or median [first-third quartiles] and compared using t-tests or Wilcoxon-rank sum tests, respectively. The primary composite outcome was assessed using logistic regression with weight-based dose as the exposure variable of interest. The primary outcome was also compared based on the maximum fixed-rate dose received during treatment. Duration of the maximum dose was also reported.

Secondary outcomes, including toxicity rates, treatment discontinuation, and individual components of the composite outcome, were compared using t-tests or Wilcoxon rank-sum tests. Pain scores were compared by weight-based dose using Spearman correlations. Data was collected and managed using REDCap electronic data capture tools hosted at The Ohio State University (24,25). Analyses were performed using Statistical Analysis System (SAS) 9.4 (SAS Institute, Inc., Cary, NC).

Results:

We identified 213 patient encounters with ketamine prescribed for pain at the comprehensive cancer center; 105 of those met study inclusion criteria. Exclusion rationale is outlined in Figure 1. The most common reason for study exclusion was no diagnosis of cancer (n=44); these patients were frequently prescribed ketamine for pain management in the setting of sickle cell pain.

Figure 1:

Study inclusion and exclusion criteria.

Patient Demographics and Baseline Opioid Requirements:

Patient demographics are summarized in Table 2. The median age at admission was 45 [35–57] years; 11 patients were over 65 years of age. Approximately half (n=55, 52.4%) were women, and the mean length of stay was 20.1 ± 37.4 days. Most patients were diagnosed with a solid tumor malignancy (n=91, 89.2%). Sixty-one (58.1%) had cancer-related disease burden within their bones. Additionally, 36 (34.3%) patients were found to have disease burden in their liver, and 20 (19.0%) had lesions within their central nervous system. Twenty-six patients had a documented history of chronic kidney disease with one requiring dialysis. Six patients had elevated liver enzymes (AST or ALT ≥40 U/L) at ketamine initiation. Baseline MEDD was >500 in over two-thirds of the included patients.

Table 2:

Patient demographics (n=105).

	All Patients (n=105)	Responders (n=51)	Non-Responders (n=54)	P-value
	Median [Q1–Q3]	Median [Q1–Q3]	Median [Q1–Q3]
Age (years)	45 [35–57]	51 [37–61]	41.5 [33–51]	0.02
Weight (kg)	73.3 [60.2–90.4]	70.4 [59.1–90.5]	74.6 [62.4–89.6]	0.4
Baseline pain scale 0–10	7.7 [6.9–8.5]	7.4 [6.8–8.1]	7.8 [7.0–8.7]	0.11
Baseline MEDD	915 [440–1740]	1065 [550–2070]	655 [363–1263]	0.03
Baseline PRN OME	584 [238–1473]	1034 [360–1920]	345 [144–905]	0.003
	N (column %)	N (column %)	N (column %)
Sex
Male	50 (47.6)	23 (45.1)	27 (50.0)	0.7
Female	55 (52.4)	28 (54.9)	27 (50.0)
Race
White	91 (86.7)	42 (84.0)	49 (90.7)	0.26
Black	8 (7.6)	6 (12.0)	3 (5.6)
Other	6 (5.7)	2 (4.0)	2 (3.7)
Cancer Type
Solid Tumor	91 (86.7)	44 (86.3)	47 (87.0)	1.0
Hematologic	11 (10.5)	5 (9.8)	6 (11.1)
Unknown	3 (2.9)	2 (3.9)	1 (1.9)
Notable Metastatic Lesions
Brain/spine	36 (34.3)	18 (35.3)	18 (33.3)	0.84
Bone	61 (58.1)	32 (62.8)	29 (53.7)	0.43
Hepatic	61 (58.1)	13 (25.5)	7 (13.0)	0.14
History of SUD	27 (26.0)	13 (25.5)	14 (25.9)	1.0
Notable Concurrent Opioids
Methadone	67 (63.8)	29 (56.9)	38 (70.4)	0.15
Buprenorphine	3 (2.9)	1 (2.0)	2 (3.7)	1.0
Concurrent Medications
Acetaminophen	73 (69.5)	34 (66.7)	39 (72.2)	0.67
Benzodiazepines	74 (70.5)	34 (66.7)	40 (74.1)	0.52
Gabapentinoids	59 (56.2)	29 (56.9)	30 (55.6)	1.0
NSAIDs	37 (35.2)	14 (27.5)	23 (42.6)	0.15
SNRI	28 (26.7)	10 (19.6)	18 (33.3)	0.13
Steroids	76 (72.4)	34 (66.7)	42 (79.3)	0.19
TCA	27 (25.7)	11 (21.6)	16 (29.6)	0.38
Baseline MEDD
Buprenorphine Monotherapy	2 (1.9)	0 (0)	2 (100)	0.04
</= 250	12 (11.4)	4 (33.3)	8 (66.7)
251–500	18 (17.1)	6 (33.3)	12 (66.7)
501–1,000	27 (25.7)	13 (48.2)	14 (51.9)
1,001–2,000	25 (23.8)	15 (60.0)	10 (40.0)
>/= 2,000	21 (20.0)	13 (61.9)	8 (38.1)

MEDD=morphine equivalent daily dose; PRN OME=pro re nata (as needed) oral morphine equivalents; SUD=substance use disorder; NSAIDs=nonsteroidal anti-inflammatory drugs; SNRI=serotonin norepinephrine reuptake inhibitor; TCA=tricyclic antidepressant

Ketamine Dosing and Success Rates:

One hundred four patients were initiated on ketamine at 5 mg/hr; one patient was initiated at 10 mg/hr. The mean time to maximum ketamine dose was 43.06 ± 53.56 hours. The mean duration of ketamine therapy was 4.9 ± 3.91 days. Fifty-one patients successfully responded to ketamine therapy. Responders to ketamine had a lower maximum fixed-rate dose as compared to non-responders (15 mg/hr [10–15] vs. 15 mg/hr [15–20], p = 0.043), but there was no difference in maximum weight-based ketamine dose (0.201 ± 0.085 mg/kg/hr vs. 0.209 ± 0.078 mg/kg/hr, p = 0.59). From the logistic regression, the odds ratio for a 0.01 mg/kg/hr increase in weight-based dose was 0.99 (95% CI: 0.94–1.03, p=0.59) (Table 3).

Table 3:

Incidence of ketamine success (responder) by primary outcome at maximum fixed-rate and weight-based ketamine dose.

Variable	Success/Responder	N (%)	Maximum fixed-rate Median (Q1-Q3)	P-value	Maximum weight-based Mean (SD)	P-value
Primary Outcome
30% decrease in pain score	Yes	8 (8.2)	15 [10–15]	0.15	0.165 (0.061)	0.12
	No	89 (91.8)	15 [10–20]		0.212 (0.082)
30% decrease in MEDD	Yes	36 (34.9)	15 [10–18.8]	0.24	0.211 (0.086)	0.66
	No	67 (65.1)	15 [15–20]		0.203 (0.080)
30% decrease in PRN OME	Yes	45 (43.7)	15 [10–15]	0.052	0.207 (0.087)	0.85
	No	58 (56.3)	15 [15–20]		0.204 (0.078)
Composite	Yes	51 (48.6)	15 [10–15]	0.043	0.201 (0.085)	0.59
	No	54 (51.4)	15 [15–20]		0.209 (0.078)

Fifty-two patients were prescribed ketamine doses equating to <0.2 mg/kg/hr, 14 patients were prescribed weight-based doses equating to ≥0.3 mg/kg/hr, and all other patient received 0.2 – 0.299 mg/kg/hr. There was no difference in the success rate between grouped weight-based doses (p=0.71) (Table 4). Figure 2 depicts responders and non-responders with their maximum fixed-rate dose and correlated weight-based dose.

Table 4:

Grouped weight-based doses and ketamine response rates.

Maximum Dose mg/kg/hr	Total # Patient in Group	Responders N (%)	Non-Responders N (%)	Fisher’s Exact p-value
<0.200	52	24 (46.2%)	28 (53.85)	0.71
0.200 – 0.299	39	21 (53.9%)	18 (46.2%)
≥ 0.300	14	6 (42.86%)	6 (42.9%)

Figure 2.

Distribution of responders and non-responders by fixed-rate and correlated weight-based dose.

Patient and Disease Characteristics:

Patients who responded to ketamine were older (median age 51 [37–61] vs. 41.5 [33–51], p=0.02). Additionally, ketamine success correlated with higher MEDD at ketamine initiation (0–250 MEDD: 33.3% success rate, 251–500 MEDD: 33.3%, 501–1000 MEDD: 48.2%, 1001–2000 MEDD: 60.0%, >2000 MEDD: 61.9%, p-trend=0.04). No other patient or disease characteristics were correlated with ketamine success.

Ketamine Tolerability:

Of the included patients, 41 (39.1%) reported adverse effects; ten of these patients required discontinuation of ketamine. The most common side effects included sleep disturbances (n=11) and hallucinations (n=9). Patients were more likely to experience any adverse effects at lower fixed-rate (15 [10–15] vs. 15 [15–20 mg/hr, p=0.004) and weight-based (0.18 ± 0.71 vs. 0.22 ± 0.08 mg/kg/hr, p=0.01) ketamine dose.

Discussion:

Cancer-related pain is a distressing symptom commonly managed with opioids, and there are limited therapeutic options when pain persists despite optimizing opioid dosing. This study found that approximately 50% of patients with cancer pain who were prescribed IV ketamine had a successful response. Most of the patients did not experience a reduction in numerical pain score, but 45 patients had at least a 30% reduction in their PRN opioid use. The patients that responded successfully had higher opioid requirements at baseline. The success-rates seen in patients with high baseline opioid requirements likely highlights ketamine’s role in treating opioid-induced hyperalgesia, allodynia, and centralized sensitization (8,26). Our study focused on an acute reduction in pain scores and opioid requirements (24-hours after the maximum ketamine dose was reached). Further studies may focus on the duration of response as some chronic pain studies have found a therapeutic benefit for weeks to months after ketamine therapy (4).

We found that patients with a successful response had lower fixed-rate dosing. We identified three patients >100 kg that responded to as little as 5 mg/hr continuous ketamine. We believe that non-responders had higher fixed-rate doses because our palliative medicine team titrated to higher doses in hopes of achieving a response. We did not find a difference in our retrospectively calculated weight-based dosing between responders and non-responders, suggesting there may be no clinical benefit in using weight-based dosing over fixed-rate. Our study was unable to include enough patients to meet power for our planned effect size, thus the true difference between these dosing strategies remains unknown.

Few patients (n=14, 13.3%) in our study were prescribed ketamine weight-based doses greater than 0.3 mg/kg/hr. Most studies using weight-based ketamine for acute and chronic non-malignant pain included a dosing range of 0.1–0.9 mg/kg/hour (3,4,6,19). Additionally, doses up to 1.25 mg/kg/hr have been studied for pain management (27). A 2021 consensus guideline for ketamine use in acute pain management recommends that continuous infusions for pain management not exceed 1 mg/kg/hr without intensive monitoring (3). It is possible that additional ketamine responders may have been identified if higher doses of ketamine were prescribed. However, our findings correlate with a 2013 study by Okamoto and colleagues (18). Of their included patients, 22% had a positive pain response to ketamine 4.2 mg/hr. Patients with cancer-related pain may respond to lower ketamine doses.

Apart from baseline opioid requirements, the only other patient/clinical characteristic that correlated with success rates was age. This finding could reflect previous studies indicating that younger patients with cancer have pain that is more difficult to treat (28–31). To date, there have not been any studies assessing ketamine’s efficacy and safety for pain management in older adults outside of the emergency department (32). However, several studies have been conducted using ketamine in older adults with treatment resistant depression; these have shown promising results for both efficacy and safety (33,34). Of the 11 patients over 65 years of age included within our study, four patients had a successful response. However, six reported adverse effects from ketamine (hallucinations n=2, fall n=1, other mood changes n=3). Three of these patients required ketamine discontinuation. Future research efforts could focus on the efficacy and safety of parenteral ketamine for treatment of pain in patients over the age of 65.

We were unable to obtain reliable histories of pain descriptions (neuropathic, nociceptive, and mixed) with historic notes. Most of the patients included in this study had advanced cancers with metastatic lesions often involving the bone, liver, and/or central nervous system, and were likely experiencing complex pain with both nociceptive and neuropathic components. When evaluating acute and chronic non-malignant pain management, the strongest evidence for ketamine has been in the treatment of neuropathic pain (35). Ketamine response rate may vary in patients with clear neuropathic vs. nociceptive pain generators. Though many patients with cancer report mixed-pain symptoms, future research may evaluate the efficacy and safety of ketamine in specific pain types.

Another unique population seen within our cohort were the 67 patients who were prescribed concomitant methadone. Methadone has several mechanisms of action including NMDA antagonism, raising the question of whether ketamine can offer any further benefit for pain management when methadone is already in use (36). However, this study did not identify a correlation between methadone and ketamine success leading investigators to believe that methadone use should not preclude patients from receiving IV ketamine when facing refractory cancer-related pain. One potential avenue for research would be an evaluation of methadone doses with ketamine efficacy given that methadone’s affinity for the NMDA receptor may be dose dependent (37).

Ketamine has been associated with several adverse drug reactions including neuropsychiatric changes, sleep disturbances, tachycardia, and hypertension (4,5). Previous studies have indicated that the prevalence of neuropsychiatric changes may be as high as 93% (4,38). To date, no patient or disease characteristics have been identified linking ketamine to neuropsychiatric changes, and consensus is to target the lowest effective dose (39). The most common side effects that were seen in our study include sleep disturbance, hallucinations, and other. Investigators noted that other was commonly selected for patients that experience unspecified changes in mood or “feeling funny/different”. Of the included patients in our study, ketamine was generally well tolerated. However, it should be noted that we excluded patients that received ketamine for less than 24-hours. These patients may have had ketamine discontinued within that time due to side effects.

Being a single-center retrospective study does come with several limitations. It is likely that prescribing bias may have played a role in patient selection and maximum ketamine doses. Additionally, weight-based doses had to be retrospectively determined since our institution uses fixed-rate dosing. Future research may encompass a variety of practice sites where both weight-based and fixed-rate dosing are used. This will likely provide a better comparison when determining efficacy and safety. Given ketamine’s high lipophilicity, future research may also focus on which body weight measurement (total, ideal, or adjusted) is the most appropriate regarding both efficacy and safety as an argument can be made that patients with a high body mass index may be prone to accumulation and subsequent adverse effects. This may also be a pertinent aspect of research when including patients with advanced cancer with rapid fluctuation in weight due to cachexia and anorexia.

In summary, though our results are underpowered we conclude that weight-based dosing may not provide additional benefit over our current fixed-rate dosing strategy when using IV ketamine for adults with cancer-related pain. Our finding that patients were more likely to respond to ketamine at lower fixed-rate dosing informs our practice when weighing risks and benefits of individual dosing titration.

Data Availability:

Data that support the findings of this study are available from the corresponding author, MS, upon reasonable request.