Skip to main content

Faculty Perspectives: Considerations in Metastatic Colorectal Cancer: Practical Applications and Patient Management for Oncology Nurses and Pharmacists

Andrea L. McNatty, PharmD, RPh
Clinical Pharmacy Specialist
Board Certified Oncology Pharmacist
Mayo Clinic,
Phoenix, AZ
Kelley A. Rone, APRN, DNP
Advanced Practice Nurse
Gastrointestinal Oncology Division of Hematology and Oncology,
Mayo Clinic,
Phoenix, AZ
Tanios S. Bekaii-Saab, MD, FACP
Professor, Mayo Clinic College of Medicine and Science
Program Leader, Gastrointestinal Cancer,
Mayo Clinic Cancer Center Consultant
Mayo Clinic,
Phoenix, AZ


There have been significant improvements in the survival of patients with metastatic colorectal cancer (mCRC) over the past decade. However, patients with the disease who fail to respond to first- or second-line therapies, including chemotherapy and monoclonal antibodies, have few treatment options. Advances in subsequent treatment lines rather than in first- or second-line therapies now drive improvements in survival for these patients.

This publication will discuss the following key issues as they relate to oncology nurses and pharmacists involved in the management of patients with mCRC:

  • Guidelines for treating mCRC after the second or third progression
  • The unique mode of action of regorafenib, an oral multi­kinase inhibitor
  • Efficacy and safety data for regorafenib, as observed in clinical trials
  • Real-world evidence for regorafenib
  • Summary of the ReDOS clinical trial.

Questions regarding patient-management strategies learned during the ReDOS clinical trial will be posed to experts, tying best practices into daily decision-making.

NCCN Guidelines for mCRC Treatment After Subsequent Progression

The National Comprehensive Cancer Network (NCCN) has set forth guidelines for the treatment of mCRC on initial diagnosis and after relapse.1 Treatment generally consists of a fluoropyrimidine (eg, fluorouracil [5-FU]), oxaliplatin, and irinotecan alone or as combination therapy.2 These therapies are often combined with an anti–vascular endothelial growth factor (VEGF) compound and/or an anti–epidermal growth factor receptor (EGFR) antibody, in patients with RAS wild-type genetics.2 Previously, treatment options after second and third progression were limited. However, the NCCN now recommends treatment with regorafenib or trifluridine-tipiracil after standard therapies.1 The focus of this publication is the use of regorafenib for mCRC after standard treatments have failed.

Regorafenib Mode of Action

Inhibiting VEGF and its receptors are key to blocking tumor growth and blood vessel formation. Regorafenib also acts through other pathways to potentially inhibit tumor growth, blood vessel formation, and metastasis.3

Inhibiting Oncogenesis

Several protein kinases are involved in the regulation of tumor oncogenesis.4 Research has shown that regorafenib potently inhibits these oncogenic kinases.5

Inhibiting Angiogenesis

Other protein kinases contribute to tumor angiogenesis.3-5 Regorafenib has been found to potently block multiple angiogenic receptors.3,5

Preventing Metastasis

Mediators of cell proliferation and migration are key to the spread of cancers. Regorafenib has been found to strongly inhibit these mediators, as well as several of the proteins that play a role in the metastasis of colorectal cancers.5-7

Key Clinical Trial Data

Several studies have demonstrated the efficacy and safety of regorafenib in patients with mCRC. The key points of these clinical trials are summarized here.


A large analysis of 24 studies including the large phase 3 randomized controlled trials (RCTs) CORRECT and CONCUR was conducted.6-8 Individually, each RCT showed a significant improvement in overall survival (OS) and progression-free survival (PFS) with regorafenib versus placebo. Combining the results of the 2 phase 3 RCTs in a meta-analysis again showed that regorafenib significantly improved OS (with a 33% reduction in the risk of death; P = .02 for overall effect size) and PFS (with a 60% reduction in the risk of progression; P <.0001 for overall effect size) when compared with placebo (Figure 1).8 A subsequent large safety study, CONSIGN, found rates of PFS similar to those in the previous RCTs.9



A higher incidence of adverse events (AEs) was reported among patients receiving treatment with regorafenib, although these were usually manageable.8 These AEs were similar to those observed with other multitarget tyrosine kinase inhibitors.8 The most frequently reported grade ≥3 AEs associated with regorafenib were hand-foot skin reaction (HFSR), hypertension, and fatigue.8

Whereas HFSR is reversible and nonlethal, its effects on quality of life (QOL) can lead to dose reductions or interruptions.8 The onset of HFSR tends to be within days or weeks of the initiation of regorafenib treatment.10 The prodromal phase of dysesthesia arrives first, which is described as a sensation of tingling that evolves to burning over a few days. Afterwards, patients may develop the following:

  • Bilateral, well-demarcated, painful nonsymmetrical erythema
  • Large, tense blisters that develop into callus-like hyper­keratosis
  • Pain that can be out of proportion to the appearance of the lesions.

Symptoms of HFSR typically occur at pressure-bearing areas such as the palms of the hands, soles of the feet (especially the heels and metatarsal head area), elbows, and any amputation sites. They may also occur at friction sites such as the spaces between fingers and toes.10

In the CONCUR7 and CORRECT6 RCTs, 14% and 17% of patients, respectively, discontinued treatment because of AEs.

Real-World Evidence

Whereas RCTs are the gold standard for clinical trial design, they can be subject to bias resulting from the recruitment of more adherent and healthier patients, which may hinder the ability of healthcare providers to generalize these findings to real-world clinical settings.11 However, real-world evidence has emerged to inform treatment decisions based on data from representative patients in typical clinical scenarios.12

In a pooled analysis of 10 studies in a real-world clinical setting, researchers found that patients with mCRC treated with regorafenib had an estimated OS of 7.27 months and an estimated PFS of 3.34 months, which were similar to those found in the phase 3 RCTs.6,7,13 For example, within the phase 3 CONCUR trial, the OS and PFS were 8.8 months and 3.2 months, respectively. AEs were also similar between the real-world studies and the RCTs, although all-grade and grade ≥3 hematologic AEs occurred more often, but still with a frequency of ≤5%.13

Another study conducted at a single center analyzed the effect of regorafenib on real-world patients with mCRC who had progressed on other therapies.14 Researchers found that the median OS (5.6 months) and PFS (2.4 months) rates were comparable to those seen in certain phase 3 RCTs. The OS in the single-center study was slightly shorter than that in the CORRECT trial (5.6 months vs 6.4 months, respectively), which may be attributed to the higher median age of the patients enrolled. Interestingly, the PFS in the single-center study was longer than that in the CORRECT trial (2.4 months vs 1.9 months, respectively).6,14

The side-effects profile of the single-center study was similar to that of the published phase 3 RCTs.6,7,14 The most frequent grade ≥3 AEs were fatigue, hypophosphatemia, HFSR, and weight loss.14 In addition, the single-center study had a smaller proportion of patients requiring dose interruption, although the number of patients requiring dose reductions was similar to that in the RCTs.

ReDOS Study: Gradual Dose Escalation

To improve the side-effects profile associated with regorafenib, researchers compared a gradual dose-escalation strategy with that of a standard dosing regimen in the randomized, multicenter, open-label phase 2 ReDOS clinical trial.15 In the dose-escalation group, patients were started on regorafenib at 80 mg/day for cycle 1, with weekly dose increases of 40 mg to a maximum dose of 160 mg/day for subsequent cycles, assuming no significant drug-related AEs occurred. In the standard-dose group, patients received 160 mg/day of regorafenib starting on day 1 until dose modification or discontinuation. In both groups, dosing occurred on the first 21 days of a 28-day cycle, followed by a 7-day break (Figure 2).15


The primary end point of this study was the proportion of patients in each group initiating a third cycle of treatment. A significantly greater percentage of patients in the dose-escalation group reached the third cycle (43%) compared with those in the standard-dose group (26%; P = .043).15

The dose-escalation strategy also showed a trend toward improved OS. The median OS for the dose-escalation group was 9.8 months compared with 6.0 months for the standard-dose group (P = .12).15

The incidence of grade 3 AEs frequently associated with regorafenib—including fatigue, HFSR, and hypertension—was lower in the dose-escalation group compared with the standard-dose group in cycles 1 and 2 (Table 1). The side-effects profile of regorafenib was similar to that seen in previous trials.15


At week 2 of treatment, mean QOL scores were significantly better in the dose-escalation group for current fatigue, general activity interference, mood interference, walking ability interference, and normal work interference, which were measured by the Brief Fatigue Inventory.15 In terms of overall findings, QOL scores were not significantly different between the 2 groups.15

These promising data may potentially help to establish a new standard for optimizing regorafenib treatment outcomes and improving QOL for patients with mCRC through the use of a dose-escalation strategy.

Clinical Perspectives on Regorafenib in the ReDOS Study

In this section, Andrea L. McNatty, PharmD, RPh, Clinical Pharmacy Specialist, Board Certified Oncology Pharmacist, Mayo Clinic, Phoenix, AZ and Kelley Rone, APRN, DNP, Advanced Practice Nurse, Gastrointestinal Oncology, Division of Hematology and Oncology, Mayo Clinic, Phoenix, AZ, provide their clinical insights from the perspective of an oncology pharmacist and an oncology nurse. They answer questions based on their experience with regorafenib concerning dosing, patient selection, and management of AEs.

Question: In pivotal trials, the recommended starting dose for regorafenib was 160 mg, which was administered orally, once a day for 21 days, followed by a 7-day break. In the ReDOS trial, a weekly dose-escalation strategy was evaluated. What were patients experiencing with the standard dose that prompted the ReDOS trial?

Andrea L. McNatty, PharmD, RPh: The AEs observed, particularly HFSR, were often severe and caused early discontinuation of the drug or significant delay. Other AEs included fatigue and nausea. This, in turn, negatively affected QOL and potential OS.

Question: How did the recommended standard dose of 160 mg/day of regorafenib affect dosing over time?

Andrea L. McNatty, PharmD, RPh: We would have to do a lot of stopping and starting, particularly for the hand and foot skin changes. When we started patients on the higher dose, we would have to wait for their skin changes to resolve before we could resume treatment at a lower dose. Unfortunately, in that time frame a lot of patients would be off treatment for so long that their disease would progress.

Question: How long would it typically take for the AEs to resolve?

Andrea L. McNatty, PharmD, RPh: It varied depending on the severity. If the patient continued the medication and their hand and foot skin changes became severe enough, sometimes it would take a week or more to resolve. If we could stop dosing when it first started, it would be 4 or 5 days.

Question: How were the AEs managed during treatment interruption?

Andrea L. McNatty, PharmD, RPh: We used steroid creams quite frequently or steroid ointments for HFSR. If it was mild, then we would perhaps just have patients use an emollient hand cream.

Question: Toxicities such as HFSR and fatigue may limit the use of regorafenib. What were the most notable differences in these AEs during your experience in the ReDOS study?

Andrea L. McNatty, PharmD, RPh: Typically, starting at the lower dose usually meant that the hand-foot symptoms and the fatigue were less severe and oftentimes caught in earlier stages so that patients would have fewer delays, and we could provide supportive care. With appropriate treatment for HFSR, patients were able to stay on regorafenib longer and were provided better QOL and potentially longer OS. Furthermore, liver toxicity was detected earlier and at a lower grade, allowing for earlier dose modification. Any changes in blood pressure were detected earlier and treated appropriately.

Kelley A. Rone, APRN, DNP: When we were able to start at a lower dose and increase, we knew whether patients were starting to have hand and foot skin issues. We had some parameters as to what to do with the dose: either hold it, leave it at the current strength, or delay treatment and start back at the same or lower dose, depending on the side effect. We had a little bit more guidance as to how to manage things.

Question: How did the dose-escalation strategy affect the management of AEs?

Kelley A. Rone, APRN, DNP: I think the side effects tended to not get out of control when we used the dose-escalation strategy. Since we were seeing the patients frequently, we could lay eyes on them and say, “Okay, here is what is happening. Let us look at your hands and feet.” When patients are not seen as frequently, they may continue to take their medication because they believe HFSR is an expected side effect. They think, “I should be feeling like this, so I will just keep taking it.”

We were usually able to circumvent any higher grades of hand and foot skin changes, because we would add an intervention or decrease the dose or hold it before things got out of hand.

Question: What additional side effects did patients experience with the dose-escalation strategy?

Kelley A. Rone, APRN, DNP: After hand and foot skin changes, the most common AE that patients experienced was hypertension. We would catch their blood pressure as it began to rise and manage it a little more closely before it got out of control.

Andrea L. McNatty, PharmD, RPh: Things like hepatotoxicity, fatigue, and nausea were caught earlier and in lower grades. Because they were being monitored and because we were not necessarily starting out with the highest dose, those things could be identified a lot sooner.

Question: Regorafenib must be taken at the same time each day. How do you discuss this with your patients?

Andrea L. McNatty, PharmD, RPh: Regorafenib needs to be taken at the same time every day with a low-fat meal that is less than 600 calories (Table 2). Therefore, I explain to patients that breakfast is the easiest and most convenient time to take regorafenib. I usually start by asking them about their regular lifestyle. When do they typically eat breakfast? What do they usually eat for breakfast?


I then ask the patient: Is this the best time for you to take it, or is it not? Do you wait until noon to eat? Giving them a specific time and directions is the way I discuss this. These are written down for patients on drug information paperwork and/or the prescription bottle.

Question: What low-fat foods that are less than 600 calories do you recommend for patients initiating regorafenib?

Andrea L. McNatty, PharmD, RPh: First, I determine what they usually eat. Some patients do not really understand what “low fat” is. We talk about which foods to avoid, especially with breakfast (eg, eggs, bacon, whole milk, lattes). Then we discuss options that are less than 600 calories and are low fat (eg, oatmeal, cereal, nonfat yogurt, fruit). If they really have a hard time with that, then we can certainly connect them with a nutritionist. We also discuss foods to avoid within 2 hours of taking the medication, such as whole milk, bacon, and eggs.

Question: What recommendations do you have for patients who do not eat breakfast?

Andrea L. McNatty, PharmD, RPh: It is not so much that regorafenib needs to be taken in the morning. It is just that most patients typically eat their smallest meal as the first meal of the day. If somebody does not eat breakfast and they do not eat until noon, they need to follow the same guidelines. The meal still needs to be less than 600 calories and low fat.

We discuss different lunch options or whatever they prefer to eat, how that fits into the restrictions, and we emphasize that they need to be consistent. Then we pick a time and write it down. If it is noon, it is “noon every day” and this is what you should eat.

I handle this process by laying it out for them and having them be involved in the decision of when they are going to take the drug, so that they have buy-in.

Question: What are the characteristics of the ideal candidate for regorafenib treatment?

Kelley A. Rone, APRN, DNP: Most of the patients who end up on regorafenib have been through multiple lines of therapy. They are familiar with how to take medications. The patients who are the best candidates are the ones who know how to follow instructions and are going to take their medication when they are supposed to.

We have some patients who will listen to our recommendations, but when they get home, they will not follow them. You would want to make sure that the patient is reliable in their medication-taking history. Then, of course, you want to make sure that they do not already have liver toxicity from their disease or previous treatments that may worsen with the addition of regorafenib.

Andrea L. McNatty, PharmD, RPh: The ideal candidates for regorafenib include patients who have an Eastern Cooperative Oncology Group performance status ≤1 with limited or no liver metastasis; patients who have had minimal toxicities with previous therapies, including fatigue, HFSR, diarrhea, or liver enzyme abnormalities; patients who have normal blood pressure readings and a minimal history of cardiac disease; and patients who have a good understanding of their disease prognosis.

Regarding adherence to medications, ideal patients are also willing to share about their side effects. As Ms Rone mentioned, sometimes patients are not good candidates for regorafenib if they have uncontrolled hypertension.

We want to make sure that their comorbidities are controlled: that they do not already have fulminant diarrhea or a history of eczema, which may or may not make them more sensitive to HFSR.

Question: How might ideal patients for regorafenib treatment differ from those who may be appropriate for trifluridine/tipiracil treatment?

Kelley A. Rone, APRN, DNP: We think of patients who have had recent infections or surgery. They may be more appropriate for trifluridine/tipiracil because of the delayed wound healing associated with VEGF inhibitors. Sometimes, if we have an option, we may choose trifluridine/tipiracil first for that reason as well.

Andrea L. McNatty, PharmD, RPh: I would be a little worried about patients on trifluridine/tipiracil if they have diarrhea issues, which are common with our patients. Trifluridine/tipiracil is likely to make that side effect worse. Regorafenib may be a little less diarrhea–producing than trifluridine/tipiracil.

Patients who may be more appropriate for treatment with trifluridine/tipiracil include those with underlying liver disease or those with a history of HFSR or hypertension. However, I do think that the regorafenib dosing is a little more straightforward. You take it for 3 weeks on and 1 week off. Whereas the trifluridine/tipiracil dosing can be somewhat confusing for patients.

Question: When do you consider treatment with regorafenib and why?

Andrea L. McNatty, PharmD, RPh: We usually consider regorafenib treatment for patients who have failed 2 or 3 previous therapies; at that point we may not have a lot of infusional chemotherapy treatments available to them. We may, at that point, switch to some sort of oral therapy to see whether we can get any kind of response.

Kelley A. Rone, APRN, DNP: Typically, we tend to keep regorafenib as a third-line therapy. Based on the NCCN guidelines, the overall opinion is that we should try the 5-FU–based therapies before we switch to regorafenib.

Question: How do you manage the most common side effects associated with regorafenib?

Andrea L. McNatty, PharmD, RPh: For grade 1/2 HFSR, we typically use urea cream plus a mild to moderate potent corticosteroid cream, such as triamcinolone. For grade 2/3 HFSR, we typically delay treatment. Whatever the patient’s dose is, we hold it until their skin has recovered to grade 1 or lower. Then we typically resume the dose 1 level down. If the patient is on 160 mg, we go down to 120 mg. Usually we will use clobetasol ointment or triamcinolone cream twice a day, then some sort of emollient hand cream or lotion for their hands and feet. We also recommend lifestyle modifications such as avoiding prolonged walking or standing, elevating their feet, and avoiding prolonged exposure to chemicals and hot water.

We typically recommend prochlorperazine as needed for grade 1 nausea. For grade 2/3 or persistent nausea, we use scheduled ondansetron administered with regorafenib and/or add olanzapine 5 mg at bedtime. Because of drug interactions, it is recommended to monitor the patient’s QTc interval for risk of prolongation with the use of ondansetron and/or olanzapine with regorafenib.

For hypertension, it is important that patients are monitored closely for blood pressure increases during therapy with regorafenib. Any underlying hypertension needs to be addressed prior to the start of therapy by either a primary care provider or a cardiologist. Because of regorafenib’s potent VEGF inhibition, hypertension usually presents within the first 4 to 6 weeks of therapy. Uncontrolled hypertension is a significant cardiac risk factor for patients on regorafenib. If the patient is on an antihypertensive, we may increase its dose, or we may refer them to cardiology for medication management.

Question: How do you manage patients who experience fatigue?

Andrea L. McNatty, PharmD, RPh: That is a little difficult. Interestingly, a lot of times, when patients start on regorafenib, once they are on it for 2 or 3 months, their fatigue improves.

One of the things that we normally do is make sure that they are not having excessive amounts of nausea or diarrhea that they have not reported. Sometimes if they have had an appetite change, they may not be eating enough calories.

We try to sit down with them and go over the amount of food that they have throughout the day, to make sure they understand about having to eat low-fat meals throughout the day, or just to make sure that they are taking in enough calories to prevent some of the excess fatigue.

With the lack of pharmacologic interventions available, we often need to reduce regorafenib dosing for grade 2/3 fatigue. We have used low-dose steroids, such as dexamethasone 2 mg daily, or methylpheni­date 5 mg daily in a select group of appropriate patients. We also watch their hemoglobin and treat any excessive anemias.

At the time of initiation of regorafenib treatment, it is important to discuss with patients the risk of fatigue and the importance of balancing rest and activity. Managing patient expectations regarding fatigue is critical to management of this side effect. I think some patients are still trying to do some of the things that they did before. They need to adjust their lifestyle and some of the social and mental aspects of cancer treatment. The third-line treatment level can play a part in this adjustment. It is also important to screen for depression and anxiety.

Kelley A. Rone, APRN, DNP: A lot of these patients have been on multiple lines of chemotherapy and their bone marrow is a little tired, and they may have other residual side effects as a result of prior therapies.

Question: Please discuss the REVERCE trial16 in relation to treatment with regorafenib relative to an anti-EGFR targeted agent.

Andrea L. McNatty, PharmD, RPh: Results from the REVERCE trial possibly imply using regorafenib first, before cetuximab, or panitumumab. In patients who are wild-type for KRAS, it is reasonable to consider regorafenib before anti-EGFR targeted agents such as cetuximab. The data suggest that OS is prolonged when using regorafenib earlier in treatment; however, we are unsure of the mechanism of genetic alterations that occurs with the use of anti-EGFR antibodies and their clinical significance.

Closing Remarks

Patients with mCRC have fewer options as they continue treatment past the second line. Regorafenib may be a viable treatment option for appropriate patients who have progressed on standard therapy. Recent research showed that a dose-escalation strategy allowed more patients to reach a third cycle of treatment when compared with a standard dosing strategy. In the future, dose escalation of regorafenib may be employed in the fight against mCRC, allowing more patients to continue treatment.


We thank Tanios S. Bekaii-Saab, MD, FACP, for his valuable insight and recommendations during the development of this publication.


  1. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Colon Cancer. Version 2.2019. May 15, 2019. Accessed July 19, 2019.
  2. Vogel A, Hofheinz RD, Kubicka S, Arnold D. Treatment decisions in metastatic colorectal cancer - beyond first and second line combination therapies. Cancer Treat Rev. 2017;59:54-60.
  3. Zhao Y, Adjei AA. Targeting angiogenesis in cancer therapy: moving beyond vascular endothelial growth factor. Oncologist. 2015;20:660-673.
  4. Zopf D, Fichtner I, Bhargava A, et al. Pharmacologic activity and pharmacokinetics of metabolites of regorafenib in preclinical models. Cancer Med. 2016;5:3176-3185.
  5. Wilhelm SM, Dumas J, Adnane L, et al. Regorafenib (BAY 73-4506): a new oral multikinase inhibitor of angiogenic, stromal and oncogenic receptor tyrosine kinases with potent preclinical antitumor activity. Int J Cancer. 2011;129:245-255.
  6. Grothey A, Van Cutsem E, Sobrero A, et al; for the CORRECT Study Group. Regorafenib monotherapy for previously treated metastatic colorectal cancer (CORRECT): an international, multicentre, randomised, placebo-controlled, phase 3 trial. Lancet. 2013;381:303-312.
  7. Li J, Qin S, Xu R, et al; for the CONCUR Investigators. Regorafenib plus best supportive care versus placebo plus best supportive care in Asian patients with previously treated metastatic colorectal cancer (CONCUR): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol. 2015;16:619-629.
  8. Røed Skårderud M, Polk A, Kjeldgaard Vistisen K, et al. Efficacy and safety of regorafenib in the treatment of metastatic colorectal cancer: a systematic review. Cancer Treat Rev. 2018;62:61-73.
  9. Dhillon S. Regorafenib: a review in metastatic colorectal cancer. Drugs. 2018;78:1133-1144.
  10. McLellan B, Ciardiello F, Lacouture ME, et al. Regorafenib-associated hand–foot skin reaction: practical advice on diagnosis, prevention, and management. Ann Oncol. 2015;26:2017-2026.
  11. Freemantle N, Marston L, Walters K, et al. Making inferences on treatment effects from real world data: propensity scores, confounding by indication, and other perils for the unwary in observational research. BMJ. 2013;347:f6409.
  12. Garrison LP Jr, Neumann PJ, Erickson P, et al. Using real-world data for coverage and payment decisions: the ISPOR Real-World Data Task Force report. Value Health. 2007;10:326-335.
  13. Mercier J, Voutsadakis IA. A systematic review and meta-analysis of retrospective series of regorafenib for treatment of metastatic colorectal cancer. Anticancer Res. 2017;37:5925-5934.
  14. Gotfrit J, Vickers M, Sud S, et al. Real-life treatment of metastatic colorectal cancer with regorafenib: a single-centre review. Curr Oncol. 2017;24:234-239.
  15. Bekaii-Saab TS, Ou FS, Ahn DH, et al. Regorafenib dose-optimisation in patients with refractory metastatic colorectal cancer (ReDOS): a randomised, multicentre, open-label phase 2 study. Lancet Oncol. 2019;20:1070-1082.
  16. Shitara K, Yamanaka T, Denda T, et al. REVERCE: a randomized phase II study of regorafenib followed by cetuximab versus the reverse sequence for previously treated metastatic colorectal cancer patients. Ann Oncol. 2019;30:259-265.