A 77-year-old woman presented to her primary care physician with abdominal pain and fatigue.
During a Targeted Oncology Case-Based Roundtable, Amit Singal, MD, medical director of the Liver Tumor Program, clinical chief of Hepatology, professor of internal medicine, Dedman Family Scholar in Clinical Care, Willis C. Maddrey, MD distinguished chair in Liver Disease, Harold C. Simmons Comprehensive Cancer Center, UT Southwestern Medical Center, in Dallas, TX, led a group of peers in a discussion about a 77-year-old patient with hepatocellular carcinoma (HCC).
SINGAL: This patient has contraindications for—or at least things that would make us concerned about—atezolizumab [Tecentriq] plus bevacizumab [Avastin].1 Patients with a history of variceal bleeds were excluded from the IMbrave150 study [NCT03434379]. Patients who had high-risk stigmata—large varices, portal hypertensive gastropathy, red wale sign—were allowed to be randomized on the trial if they had those treated. Patients with a history of variceal bleed in the recent past were excluded, given the risk of bleeding with bevacizumab.2 Crohn disease would also be a major concern from the immunotherapy perspective. We would either steer away from atezolizumab plus bevacizumab on second consideration, or at least discuss this carefully with the other physicians regarding the risk of flare-ups on either side. Although the IMbrave150 study— atezolizumab plus bevacizumab—is an exciting advance in the field of hepatocellular carcinoma [HCC], there are many patients who continue to be candidates for tyrosine kinase inhibitor [TKI]-based therapy, whether it is sorafenib [Nexavar] or lenvatinib [Lenvima]. If this patient had Child- Pugh B7 disease, there are increased data from a postmarketing perspective for sorafenib more so than lenvatinib, which currently does not have that same degree of postmarketing data in expanded patient populations.
CHU: Performance status is the major factor for me to decide what treatment the patient will get. The better the performance status, the more likely the patient will respond, and the easier it is to take care of the patient. I would probably start a TKI for those whose performance status is not ideal.
HUANG: Child-Pugh status is clearly important too. Whether for a TKI or atezolizumab plus bevacizumab, if the patient’s disease is beyond A or maybe a 7B, they are in more iffy territory. Patient preference is obviously [an important factor].
Do you think cirrhosis of different types, for example, hepatitis B or hepatitis C, has an impact on this decision?
SINGAL: I am sure that some of you saw the Nature paper that talked about decreased responses to immunotherapy in nonviral etiologies, particularly in patients with nonalcoholic fatty liver disease.3 This was highlighted from a preclinical rationale perspective, then was at least superficially evaluated in subgroup analyses of clinical trials, showing decreased response in the nonviral etiologies vs the viral etiologies.
Most of us think these data are too early to change our clinical practice currently. The clinical data in that paper had limitations, and the trials were not stratified on etiology, so those 2 groups were likely not matched well. There are also differences that one would anticipate just based on the enrollment of those clinical trials, with patients with viral disease being more likely from Asia and patients with nonviral disease being more likely from the US and Europe, with different treatment practice patterns in terms of when people came into systemic therapy. Overall, we typically do not make treatment decisions based on etiology.
of sorafenib in hepatitis C [HCV], and potentially decreased in hepatitis B [HBV]. Although in the subgroup analyses of the REFLECT trial [NCT01761266] comparing lenvatinib with sorafenib, there really was not any difference by any etiology that stood out [HR, 0.83; 95% CI, 0.68- 1.02 for HBV; HR, 0.9; 95% CI, 0.66-1.26 for HCV].4 So at this point, we typically do not change our clinical practice based on etiology.
CHAE: The Child-Pugh [Score] is important. I tried to get sorafenib approved, and insurance would not even pass it through until I documented the Child-Pugh score. Patient preference [is important], of course.
I am in a rural area and some of our patients need to drive an hour each way, so that can be cost prohibitive if they are having to come every few weeks for the atezolizumab dose. Some of them request oral therapy just because it is more convenient.
MOSCOL: Do we have any data about using atezolizumab plus bevacizumab in the second line [after they] progress on a TKI?
SINGAL: We do not currently have data on that. I imagine that Genentech is going to be doing postmarketing studies, just as everyone else does. I imagine these data will become available—that there were patients who were on TKI first-line therapy who then go on to atezolizumab plus bevacizumab second-line therapy, but we do not have any of those data now.
They will probably come out over the next several years, but one would envision there would not be any reason why that combination should not be effective in those patients in the second line. I can tell you there were patients who were on TKI therapy prior to the IMbrave150 study that we put on atezolizumab plus bevacizumab second line.
Has anyone else used atezolizumab plus bevacizumab second line after TKIs in clinical practice?
MOSCOL: I have put patients on atezolizumab plus bevacizumab, just because otherwise I would not be able to use bevacizumab.
You could try to use lenvatinib, but mechanistically, it makes sense that bevacizumab may help if this is a highly perfused type of tumor. It is anecdotal, but I have seen some disease stability so far.
SINGAL: I think we are going to start seeing this more and more, and we will probably see some data coming in terms of those patients who were on TKI therapy and whose disease is progressing.
It’s interesting, when you look at what therapies have succeeded in HCC, all of them have targeted VEGF— whether that is through a TKI mechanism or a single pathway of VEGF inhibition—including bevacizumab or ramucirumab.
LUU: I have a patient whom I put on lenvatinib, and his disease progressed. I tried to get atezolizumab plus bevacizumab approved by his insurance company, but they declined, so I gave him cabozantinib [Cabometyx] plus ipilimumab [Yervoy] and nivolumab [Opdivo].
SINGAL: The REFLECT trial showed lenvatinib was noninferior to sorafenib in terms of overall survival [median survival time for lenvatinib, 13.6 months, 95% CI, 12.1-14.9] was noninferior to sorafenib [12.3 months; 95% CI, 10.4-13.9; HR, 0.92; 95% CI, 0.79-1.06],4 but there were statistically significant improvements with lenvatinib in several secondary end points, including progression-free survival [HR, 0.66; 95% CI, 0.57-0.77; P < .0001], time to progression [HR, 0.63; 95% CI, 0.53-0.73; P < .0001], and objective response [odds ratio, 3.13; 95% CI, 2.15-4.56; P < .0001].4
There was a delay in deterioration for the quality-of-life items with lenvatinib compared with sorafenib (role functioning, P = .0193; pain, P = .0105; diarrhea, P < .0001; nutrition, P = .0113; body image, P = .0051). The median duration of treatment was longer with lenvatinib compared with sorafenib (lenvatinib, 5.7 months; interquartile range [IQR], 2.9-11.1; sorafenib, 3.7 months; IQR, 1.8-7.4).5
SINGAL: The overall proportion of patients who have any AE is similar for lenvatinib and sorafenib, although the specific AEs differ in terms of their incidence.2 You see much higher hand-foot skin reaction with sorafenib compared with lenvatinib [52% vs 27%], but you see much higher hypertension [42% vs 30%], appetite loss [34% vs 27%], and weight loss [31% vs 22%] with lenvatinib compared with sorafenib.
Another AE you see with lenvatinib more than sorafenib is proteinuria, although it is relatively low for both [25% vs 11%].
ALOBA: Lenvatinib is certainly harder to administer than sorafenib in terms of AEs. There are higher grade 3 AEs, but I monitor the patient closely through weekly labs in the beginning, and I adjust the dose and maybe start at a lower dose. It has a higher response rate than sorafenib, so I use lenvatinib and have successfully managed the AEs as long as the patients are closely monitored.
SINGAL: What AEs have you found difficult to manage with lenvatinib?
ALOBA: The hypertension and the diarrhea.
SOLANKI: I look at the list of [AEs] based on what I can manage more easily and what I cannot. I have dealt with hand-foot syndrome with other drugs, especially the regorafenib [Stivarga], and it is a misery. There is no treatment you can give for this. I can manage hypertension and I can get somebody involved to manage hypertension if it is a serious enough problem. We have dealt with hypertension with other drugs, and we manage that well. I look at lenvatinib as a preferable agent because [I think I can manage, or I can get assistance to manage, AEs] that dominate lenvatinib.
I can’t manage hand-foot syndrome very well. Diarrhea is another one that is particularly troublesome, but there is no difference between the 2. Diarrhea may not kill anybody, but it makes life miserable, and it is very hard to manage.
We often talk about ”Oh, it is only grade 2 or grade 1.” Well, walk around for 2 weeks, 3 weeks, or 4 weeks with grade 2 diarrhea and you have to work, you have children, or you have to take care of other things. It is just a miserable [AE]. Here, the diarrhea seems to be roughly equal in both, so the hand-foot syndrome seems to dominate my thinking [about the AEs].
CEN: Generally, I see better tolerance for lenvatinib than sorafenib in general in terms of hand-foot skin reactions and fatigue level.
SINGAL: You prefer to use lenvatinib because of that?
CEN: Yes, and of course, the response rates are better than sorafenib.
SINGAL: The hand-foot skin reaction can be troublesome for patients. We have had some success in terms of starting at a lower dose and being aggressive with urea-based creams early in the management.
[The risk of hypertension] is higher with lenvatinib, but we have multiple medications for hypertension that one could use and potentially compound to manage the hypertension. The AEs I struggle most with are appetite loss and weight loss.
We can talk about nutritional management, but it can be difficult for patients who have had significant loss of appetite and weight loss. There is no pill that works well in patients who have had those AEs. It probably comes down to the individual patient in front of you, and it can be helpful to consider AEs as 1 of the data points when choosing between sorafenib and lenvatinib.
SINGAL: The AEs that should be monitored with lenvatinib are hypertension, hypothyroidism, hand-foot skin reaction, decreased appetite or weight, diarrhea, fatigue, and proteinuria.3 In terms of dose reduction recommendations, the starting dose of lenvatinib is 8 mg if the patient [weighs] less than 60 kg, and 12 mg if they are a higher weight.
Then you reduce by 4 mg at the first dose reduction. For the second dose reduction, if they are a higher weight, which most of our patients in Texas are, they go from 8 mg down to 4 mg. Then on the third dose reduction, they go to every other day. If they are at the [lower] dose, [the first reduction is to 4 mg, the second dose reduction is 4 mg every other day, and the third dose reduction is to discontinue].3
One thing of note with TKI therapy—we learned this with sorafenib and applied it for other TKI therapies—is that we often start at a lower dose, then increase with tolerability. We do not often start with the full dose. That is the standard practice in our center, and I know that was already raised by somebody else.
This has been shown in colorectal cancer with regorafenib to be helpful in terms of maintaining people on therapy. This can be helpful with TKI therapy—although outside the label—in terms of starting with a lower dose and titrating up if tolerated. If one were to do that, it is important to make sure you titrate up if the patient tolerates it, so you can maximize benefit.
SOLANKI: In any of the studies, did they publish what the actual delivered dose of each of the drugs was? Not what was planned, but, at the end of the studies, the average dose the patients got?
SINGAL: For example, for cabozantinib in the CELESTIAL trial [NCT01908426], the starting dose was 60 mg, and typically the median dose delivered was lower at 40 mg. So to your point, they typically do report this. In the REFLECT trial, in the 8 mg per day dose group—the lower weight group—the mean lenvatinib dose intensity was 7 mg, so pretty high.
In the 12 mg per day group—the group with a higher weight—the mean dose intensity was 10.5 mg. So 88% of the planned starting dose in both groups. The mean sorafenib dose intensity was 663 mg, or 83% of the planned starting dose. Both groups had high dose intensity and, I would argue, maybe higher than what we have seen in our clinical patients.
SINGAL: LEAP-002 [NCT03713593] is the combination of lenvatinib and pembrolizumab [Keytruda]. That trial had very high responses in about 45% of people, so we are awaiting those data hopefully later this year.6 The HIMALAYA trial [NCT03298451] was looking at the combination of durvalumab [Imfinzi] and tremelimumab—so PD-1 and CTLA-4 in the front line.
There is the CheckMate 9DW [NCT04039607] trial looking at ipilimumab [Yervoy] plus nivolumab [Opdivo] in the front line. Then the other trial we have is COSMIC-312 [NCT03755791], with the combination cabozantinib [Cabometyx] and atezolizumab. The COSMIC-312 trial did have a press release that it hit its end point of progression-free survival and failed to hit its end point of overall survival at the interim analysis—we must see if that changes in the future.7
There are combinations of TKI and immunotherapy, as well as immunotherapy, with PD-1 and CTLA-4s, all of them hopefully reporting within the next year or so. Many of them are at least optimistic and should have exciting data, so we will see what the future holds.
1. Benson AB, D’Angelica MI, Abbott DE, et al. Hepatobiliary cancers, version 2.2021, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2021;19(5):541-565. doi:10.6004/jnccn.2021.0022
2. Finn RS, Qin S, Ikeda M, et al. Atezolizumab plus bevacizumab in unresectable hepatocellular carcinoma. N Engl J Med. 2020;382(20):1894-1905. doi:10.1056/NEJMoa1915745
3. Pfister D, Núñez NG, Pinyol R, et al. NASH limits anti-tumour surveillance in immunotherapy-treated HCC. Nature. 2021;592(7854):450-456. doi:10.1038/s41586-021-03362-0
4. Kudo M, Finn RS, Qin S, et al. Lenvatinib versus sorafenib in first-line treatment of patients with unresectable hepatocellular carcinoma: a randomised phase 3 non-inferiority trial. Lancet. 2018;391(10126):1163-1173. doi:10.1016/S0140-6736(18)30207-1
5. Kelley RK, Ryoo BY, Merle P, et al. Second-line cabozantinib after sorafenib treatment for advanced hepatocellular carcinoma: a subgroup analysis of the phase 3 CELESTIAL trial. ESMO Open. 2020;5(4):e000714. doi:10.1136/esmoopen-2020-000714
6. Taylor MH, Schmidt EV, Dutcus C, et al. The LEAP program: lenvatinib plus pembrolizumab for the treatment of advanced solid tumors. Future Oncol. 2021;17(6):637-648. doi:10.2217/fon-2020-0937
7. Kelley RK, W Oliver J, Hazra S, et al. Cabozantinib in combination with atezolizumab versus sorafenib in treatment-naive advanced hepatocellular carcinoma: COSMIC-312 Phase III study design. Future Oncol. 2020;16(21):1525-1536. doi:10.2217/fon-2020-0283