GVHD management improves, but questions remain about risk stratification, prophylaxis

The FDA approval of ibrutinib for adults with chronic graft-versus-host disease may herald a practice-changing breakthrough in the management of this posttransplant complication.

GVHD causes a cross-section of complications — from infections to debilitating tissue injury that can cause irreversible fibrosis — that have made it a daunting challenge following hematopoietic stem cell transplant. Although the approval of one drug will not eliminate all these challenges, there is hope the agent will provide a step on the road to long-term GVHD management.

“For a long time, there were no new drugs for GVHD, at least in part because pharmaceutical companies were reluctant to get involved,” James L.M. Ferrara MD, DSc, Ward-Coleman chair in cancer medicine; director of Hematologic Malignancies Translational Research Center at Tisch Cancer Institute; and professor of medicine, hematology and medical oncology at Icahn School of Medicine at Mount Sinai, told HemOnc Today. “One very smart senior VP of a pharma company said the problem is that GVHD is where new drugs go to die.”

Still, the tide may be changing with increased understanding of GVHD biology, and with encouraging data emerging on JAK and histone deacetylase inhibitors.

“However, there are a couple of big questions that we’ll need to answer with these drugs,” Ferrara said. “One is whether we’ll be able to get a complete response, and another is whether we’ll be able to get patients off steroids, which decimate the immune system.”

Beyond treatment, researchers also are working to prevent GVHD by identifying novel biomarkers for stratifying patients. The gut microbiome and donor type and source are other areas of active investigation.

HemOnc Today spoke with HSCT experts about the long-term complications associated with GVHD, the impact of the ibrutinib (Imbruvica; Pharmacyclics, Janssen) approval, ongoing research to better understand and treat acute and chronic GVHD, and headway being made into GVHD prophylaxis.

Impact of GVHD

Although HSCT can provide long-term survival for some patients with hematologic malignancies, the procedure poses a substantial risk for GVHD.

Following allogeneic HSCT, GVHD develops when the donor’s immune cells attack the patient’s normal cells.

Acute GVHD — which typically occurs soon after transplant — can range from mild to severe, and can be life-threatening if not controlled.

Chronic GVHD can occur 3 months to up to 2 years after transplant. Although chronic GVHD is more common among patients who experienced acute GVHD, patients can experience both, either or neither type of GVHD.

Incidence of acute GVHD ranges from 26% to 50% among recipients of matched sibling donor grafts, and from 42% to 75% among recipients of matched unrelated donor grafts. Chronic GVHD can affect around 30% of recipients of fully matched transplants to 60% to 70% of recipients of mismatched transplants.

Only about half of patients respond to steroids, the primary treatment for GVHD. Further, GVHD may increase risk for greater morbidity following transplant.

“GVHD raises the risk for all infectious complications,” Mindy G. Schuster, MD, professor of infectious diseases at Perelman School of Medicine of University of Pennsylvania, said in an interview. “The treatment of GVHD increases immunosuppression, which further compounds the problem.”

Clinicians treating patients following HSCT need to be aware of infectious complications at all stages of the process, according to Schuster.

“Both acute and chronic GVHD increase the risk for infection,” she said. “The degree and duration of immunosuppression needed to control GVHD, whether it is acute or chronic, is a key risk factor for the development of subsequent infection.”

Beyond infection, there are a range of comorbidities that may accompany GVHD and present additional challenges.

“These include loss of muscle and bone mass; development of various endocrine deficiencies; secondary skin cancers; and suboptimal organ functions, including cardiac, neurologic and renal functions, which are further compounded from other treatment-related side effects, especially from steroids,” Pavan Reddy, MD, Frances Victor Ginsberg professor of medicine, chief of the division of hematology-oncology, and deputy director of University of Michigan Cancer Center, told HemOnc Today.

In the setting of GVHD, it is as important to focus on quality of life as it is to focus on survival, according to Marco Mielcarek, MD, medical director of adult blood and marrow transplantation at Seattle Cancer Care Alliance, member and researcher at Fred Hutchinson Cancer Research Center, and professor of medicine in the department of medical oncology at University of Washington.

Source: Kris Krug, Seattle Cancer Care Alliance.

“After recurrence of the malignancy, GVHD is the second most important factor contributing to posttransplant mortality,” Mielcarek told HemOnc Today. “The burden of morbidity associated with chronic GVHD is substantial. For example, it has been shown that patients who received growth factor-mobilized blood instead of marrow from HLA-matched unrelated donors not only had a significantly greater likelihood of developing chronic GVHD, but also had inferior psychological well-being and were about 50% less likely to return to work at 5 years after transplant.”

To better understand complications that occur among patients with chronic GVHD, Arora and colleagues surveyed 581 adults from the Blood or Marrow Transplant Survivor Study — a retrospective cohort of individuals who underwent HSCT from 1974 to 1988 — between 2000 and 2004, and then again between 2013 and 2017.

In total, 52.3% of the cohort reported chronic GVHD. Compared with patients without chronic GVHD, multivariable analysis revealed those with it had significantly increased risk for osteonecrosis (HR = 2.59; 95% CI, 1.4-5), pulmonary complications (HR = 2.04; 95% CI, 1.3-3.2), ocular/oral complications (HR = 1.91; 95% CI, 1.4-2.5), gastrointestinal complications (HR = 1.87; 95% CI, 1.1-3.3), diabetes (HR = 1.58; 95% CI, 0.96-2.6) and neurological complications (HR = 1.52; 95% CI, 1.1-2.2).

“We are doing better with acute GVHD than with chronic GVHD, but these are complications that still occur in 60% to 70% of transplant recipients,” Nandita Khera, MD, MPH, associate professor of medicine and consultant in the division of hematology/oncology in the department of internal medicine at Mayo Clinic in Phoenix, Arizona, told HemOnc Today. “The field is moving forward, but there is still work to be done. We still lose patients to both acute and chronic GVHD.”

The incorporation of patient-reported data to improve outcomes should not be overlooked, Khera added.

“Unlike with acute GVHD, chronic GVHD has a prolonged trajectory that reaches into every aspect of a patient’s life,” she said. “Mental and physical functioning among patients with chronic GVHD can be similar to those observed in chronic heart failure.”

New treatments

Because mortality is high among patients who don’t respond to steroids, approval of ibrutinib for chronic GVHD provided steroid-refractory patients with a much-needed new treatment.

FDA approved ibrutinib based on results from the open-label, multicenter, single-arm PCYC-1129-CA trial, which included 42 patients assigned 420 mg of the drug daily. Eligible participants had failed first-line corticosteroid therapy.

Results — published last year in Blood — showed a response rate of 67% (95% CI, 51-80), with responses occurring at a median 12.3 weeks. Nearly half of the cohort reported improvement of symptoms for 5 months or longer.

“It was a small study, but it’s a start,” Khera said. “Chronic GVHD can be really difficult, and we need to know more about the organ systems involved in chronic disease to really understand how to make this advantageous. But, it’s exciting that we finally have this option.”

However, due to the phase 1b/phase 2 nature of the trial, more research is necessary, according to Mielcarek.

“Confirmatory phase 3 studies in the frontline setting would be highly desirable,” he said. “Although our understanding of the pathophysiology of chronic GVHD is still evolving, both B cells and T cells appear to play important roles. The ease of administration in an outpatient setting with once-daily dosing is certainly appealing.”

The key is in the mechanism, according to Ferrara.

“Studies have now shown that B lymphocytes that produce antibodies are involved in chronic GVHD pathology,” he said. “Ibrutinib blocks a key signaling pathway primarily for B cells, including malignant B cells. Once we found out these cells were involved in chronic GVHD and the pathophysiology, we tried repurposing ibrutinib in this setting.”

Overall, the lack of randomized controlled data on ibrutinib for GVHD should not temper enthusiasm, Reddy said.

“It is truly the first drug FDA has approved for GVHD,” he said. “Other drugs used routinely do not have a specific FDA indication for GVHD. This is important for the field, as this will allow further development of novel strategies, bring in investment, and develop a roadmap for pharma and other relevant stakeholders.”

Based on the data for chronic GVHD, Ferrara suggested that it may be possible for ibrutinib to serve as a primary therapy.

“Ibrutinib may be one way to catch the disease early, improve response rates and avoid some of the serious toxicities of steroid treatment,” he said.

Understanding the underlying biology of GVHD that helped lead to the approval of ibrutinib also may pave the way for the development of other drugs.

For instance, researchers are evaluating agents that target cytokine-mediated B-cell differentiation or survival, such as the ROCK2 inhibitor KD025 (Kadmon Holdings), which downregulates Th17 and T-follicular helper cells while upregulating anti-inflammatory T-regulatory cells, correcting the immunologic imbalance in chronic GVHD.

At BMT Tandem Meetings, Lazaryan and colleagues presented phase 2 data on KD025, which showed an overall response rate of 65% to 69% in different dosing cohorts. Responses occurred across all affected organ systems, with complete responses observed in the upper and lower GI tract, esophagus, mouth, joints/fascia, skin, eyes and liver.

Also, JAK1 and JAK2 inhibition with ruxolitinib (Jakafi, Incyte) showed inhibition of chronic GVHD features in mice though reduced interferon- and interleukin-21 signaling. Ruxolitinib conferred an 80% ORR among patients with chronic GVHD who failed on many other therapies, and a phase 3 trial is underway.

With all of these approaches under investigation, this is indeed an exciting time for GVHD, Reddy said.

“We are additionally exploring the role of CD24Fc protein, which targets a novel pathway, and we and others are exploring the role of alpha 1 antitrypsin,” he said. “Others are exploring the role of cellular therapies, JAK inhibitors, integrin blockers, cytokine blockers and various immune cell-depletion strategies as potential novel approaches. We now have several novel and innovative options for mitigating GVHD. They offer additional choices for these desperate patients.”

Donor type and source

Although receiving HSCT grafts from unrelated donors has expanded the pool of eligible participants, different donor types and graft sources have been shown to affect the incidence of GVHD.

“The most significant improvement in GVHD management has been driven by improvement in HLA matching,” Khera said. “Even outcomes in HLA-mismatched donations are improving over time.”

Donor sex also can affect outcomes.

Because male donors are typically associated with lower rates of GVHD, but well-matched unrelated donors are associated with higher rates compared with sibling donors, Kumar and colleagues assessed GVHD outcomes with unrelated male donors compared with parous sibling donors.

Results, published this year in Blood Advances, showed patients with unrelated male donors had a 1.6-fold higher risk for grade 2 to grade 4 acute GVHD (P < .0001). For chronic GVHD, female recipients of male unrelated donor grafts had a higher risk than those receiving parous female sibling grafts (RR = 1.43, P < .0001), whereas male recipients had similar rates regardless of donor type (RR = 1.09).

Thus, researchers concluded that parous female siblings are preferred over male unrelated donors.

Salit and colleagues also evaluated the incidence of nonmalignant late complications and quality-of-life outcomes in a cohort of patients who underwent HSCT between 2008 and 2016 and survived at least 1 year. The analysis included 357 patients who received matched related grafts, 555 patients who received 10/10 HLA-matched unrelated grafts, 103 who received mismatched unrelated grafts, 98 who received umbilical cord blood grafts and 66 who received related haploidentical grafts.

Results — presented at this year’s BMT Tandem Meetings — showed 66% of patients overall experienced chronic GVHD. Umbilical cord blood recipients showed the lowest rates of chronic GVHD, whereas mismatched unrelated donor recipients had the highest (40% vs. 82%; P < .0001).

“Cord blood is on the way out,” Ferrara added. “Those patients end up being in the hospital a very long time because their immunologic constitution takes a long time to develop. Haploidentical transplant is replacing it in many centers.”

A study by Nilanjan Ghosh, MD, PhD, and colleagues, published in 2016 in Journal of Clinical Oncology, showed patients with lymphoma with haploidentical-matched donors had a similar incidence of 100-day acute GVHD as patients undergoing HLA-matched HSCT (27% vs. 25%), but a significantly lower rate of chronic GVHD (12% vs. 45%; P < .001).

“For patients who need allogeneic transplants, HLA-matched sibling donors are considered the gold standard,” Ghosh told HemOnc Today when the study was published. “If a matched sibling donor is not possible, then a matched unrelated donor search is initiated. Haploidentical donors are readily available, but in the past, these donors have not been regularly used due to risk for excessive GVHD.

“This information — coupled with the widespread availability of haploidentical donors, the shorter time to transplant and lower costs compared with the other alternative donor choices — should make allogeneic transplantation widely available for patients with lymphoma,” Ghosh added.

Ferrara suggested that unrelated donorship is likely to continue.

“If a patient has a common HLA type, they will have many donors, an unrelated donor is a good bet,” he said. “If they have an unusual HLA type, and there are only one or two donors, a lot of clinicians are turning to haploidentical transplant. These patients have advanced hematologic malignancies and can only be kept in remission for a short period of time. Although those results are not fabulous, it’s a baseline that can be built upon.”

Assessing microbial populations

Research also is revealing the role the gut microbiome plays in the development of GVHD.

In a study presented at BMT Tandem Meetings, Ponce and colleagues assessed 1,994 stool samples from 286 patients who underwent allogeneic HSCT. Eligible participants developed acute GVHD by day 100 in the upper GI only, had lower GI involvement with or without upper GI involvement, or had no GI involvement.

Before onset of GVHD, all groups showed similar microbial diversity. However, after day 20, the most significant diversity in microbiota was observed in the non-GI GVHD group, followed by the upper GI GVHD group and then the lower GI GVHD group.

“If these results can be confirmed, one could envision interventions aimed at preventing the loss of diversity, such as avoidance of certain types of antibiotics, use of probiotics or controlled oral feeding of patients who would typically only receive parenteral nutrition after transplant,” Mielcarek said. “It needs to be demonstrated, though, that the loss of microbial diversity among patients with lower GI tract GVHD is a cause rather than just the result of GVHD that could have disrupted the mucosal environment certain beneficial bacterial communities depend on.”

Ponce and colleagues also showed that, after acute GVHD onset, patients with lower GI involvement had a median diversity score of 3.14, compared with a median diversity score of 5.73 for the upper GI group and 7.2 for the non-GI group (P < .005).

Mielcarek suggested that an increasing body of evidence indicates that loss of diversity of bacterial communities in the GI tract is associated with poor outcomes.

“It has been reported, for example, that high bacterial diversity and relative abundance of certain bacterial commensals within the Clostridiales order are protective against GVHD-associated mortality,” he said. “A decrease in GI microbial diversity during transplant has been attributed to the use of certain antibiotics and low caloric intake. These findings suggest that preserving GI microbial diversity during the transplant procedure, or restoring loss of diversity, might lead to better outcomes through reduced GVHD-related morbidity and mortality.”

In another study presented at BMT Tandem Meetings, Peled and colleagues analyzed 5,823 serial stool samples from 1,118 patients who underwent allogeneic HSCT at three different centers. Results indicated that low diversity at the time of neutrophil engraftment appeared associated with decreased survival, and diversity continued to decrease during and early after transplant.

Factors causing low GI microbial diversity included use of certain antibiotics, low caloric intake and higher-intensity preparative regimens.

“This study is important because it shows that previous associations between low bacterial diversity states and poor outcome do not appear to be a center-specific phenomenon but are reproducible at other institutions,” Mielcarek said. “The finding of an association between low caloric oral intake and loss of microbial diversity is of particular interest because it could motivate efforts to initiate oral feeding among patients who would otherwise be deemed to be candidates for parenteral nutrition.”

A sharper focus on the GI tract may yield dividends in both treatment and prevention, according to Ferrara.

“We’re now beginning to understand that the GI tract is the most difficult target,” he said, suggesting two possible ways of approaching this. “One is monoclonal antibodies that stop trafficking of T cells to the GI tract. The other is to understand that the GI tract protects itself as part of the innate immune system. The more we understand this natural defense system, the more we may be able to prevent GVHD.”

Thinking more broadly, prophylaxis can begin much earlier in the process, according to Ferrara.

“There are some plans to enhance that mucosal immunity prior to conditioning for transplant,” he said. “Preclinical models are currently testing this idea.”

Other prophylactic measures

Beyond the GI tract, experts are investigating other methods of preventing GVHD, and identifying who is most likely to experience the complication.

Based on efficacy in the setting of reduced-intensity, related donor HSCT, Reddy and colleagues investigated the histone deacetylase inhibitor vorinostat (Zolinza, Merck) as a preventive agent combined with tacrolimus and methotrexate among 37 patients who underwent myeloablative HSCT from an 8/8 HLA-matched unrelated donor.

Results showed that vorinostat-based regimens could lower the incidence of severe GVHD from a 45% rate observed in historical controls to less than 22%.

The researchers also reported a 76% (95% CI, 63-92) 1-year OS rate following transplantation in the vorinostat-treated group.

“The study was informed by years of investigation on the effect of this drug on immune cells and in animal models of GVHD and then validated in a proof-of-concept study in humans after reduced-intensity transplant where the patients received transplants from matched, related donors,” Reddy said. “Our study points to an exciting approach, which should be studied in a randomized setting to become truly practice changing,”

Mielcarek pointed to posttransplantation cyclophosphamide as an important component of GVHD prevention.

“Initially pioneered by the group at Johns Hopkins University for patients receiving HLA-haploidentical bone marrow transplants, the concept has been successfully expanded to HLA-matched bone marrow transplantation, and more recently to transplantation of growth factor-mobilized peripheral blood cells,” he said. “In multiple studies, cyclophosphamide has been shown to reduce the risk for both chronic and severe acute GVHD.”

Cyclophosphamide can confer upward of a 50% chronic GVHD risk reduction, Mielcarek said, and the agent is typically well tolerated.

“The treatment schema is quite simple with relatively low associated costs,” he said. “In the context of transplantation of HLA-matched growth factor-mobilized blood, posttransplant cyclophosphamide appears to profoundly mitigate the substantial risk for chronic GVHD otherwise associated with use of this type of stem cell product.”

At BMT Tandem Meetings, Bolaños-Meade and colleagues compared GVHD prophylaxis regimens — including one cohort that received tacrolimus and mycophenolate mofetil plus posttransplant cyclophosphamide — compared with tacrolimus plus methotrexate among patients who received peripheral blood stem cell grafts from 6/6 HLA-matched related or 7-8/8 HLA-matched unrelated donors.

Results showed improved GVHD-free and relapse-free survival rates in the cyclophosphamide arm compared with controls (HR = 0.72; 90% CI, 0.54-0.94), as well as reduced grade 3 or 4 acute GVHD (HR = 0.13; 90% CI, 0.03-0.44) and improved GVHD-free survival (HR = 0.63; 90% CI, 0.46-0.85).

Still, it remains important to determine who is most likely to benefit from intense prophylaxis.

Ferrara and colleagues at Mount Sinai developed the Mount Sinai Acute GVHD International Consortium (MAGIC) algorithm, a two-biomarker model — ST2 and REG3a — that can predict which patients will develop GVHD after transplant.

“On day 7, we were able to identify about 20% of the 1,300 patients who were at very high risk for lethal GVHD,” Ferrara said. “This is the ‘nugget’ of this paper. When we conducted further analysis among the 200 patients who sent us additional samples at the onset of GVHD, we saw that use of two biomarkers with this new test was even more accurate than a three-biomarker algorithm. We now know we can use two biomarkers — not only before GVHD but also at the onset — to predict who is at high risk for severe or lethal GVHD and who is at low risk.

“When a clinician is faced with GVHD in the early stages, the clinical response phenotype and the severity of the disease are poor indicators of how that patient is going to do,” Ferrara added. “These biomarkers give us a guide, at onset, to who is likely to respond and who is not. We believe that the biomarkers also measure damage in the GI tract, particularly in areas where we can’t image well, like the terminal small intestine. They give us actionable information early in the disease course.”

That actionable information translates into selection of therapy. But, the picture is not nearly complete.

“At the moment, these biomarkers are in the realm of a research tool,” Ferrara said. “In the trials that have been conducted thus far, there is such a mix of patients that you can’t tell who is who in terms of risk. We are moving toward the ability to stratify patients with the aim of putting them into different therapeutic categories based on low-risk and high-risk markers. This kind of patient selection will lead to better outcomes.” – by Rob Volansky

Click here to read the , “Should research efforts focus on completely preventing GVHD?”

References:

Alyea EP. J Clin Oncol. 2016;doi:10.1200/JCO.2015.66.0902.

Arora M, et al. Abstract 336. Presented at: ASH Annual Meeting and Exposition; Dec. 9-12, 2017; Atlanta.

Bolaños-Meade J, et al. Abstract LBA1. Presented at: BMT Tandem Meetings; Feb. 21-25, 2018; Salt Lake City.

Chang YJ, et al. J Clin Oncol. 2016;doi:10.1200/JCO.2015.63.8817.

Choi SW, et al. Blood. 2017;doi:10.1182/blood-2017-06-790469.

Ferrara JLM, et al. J Clin Invest. 2017;doi:10.1172/jci.insight.89798.

Ghosh N, et al. J Clin Oncol. 2016;doi:10.1200/JCO.2015.66.3476.

Kumar AJ, et al. Blood Adv. 2018;doi:10.1182/bloodadvances.2017013052.

Lazaryan A, et al. Abstract 38. Presented at: BMT Tandem Meetings; Feb. 21-25, 2018; Salt Lake City.

Lee SJ, et al. JAMA Oncol. 2016;doi: 10.1001/jamaoncol.2016.2520.

Miklos D, et al. Blood. 2017;doi:10.1182/blood-2017-07-793786.

Ponce DM, et al. Abstract 57. Presented at: BMT Tandem Meetings; Feb. 21-25, 2018; Salt Lake City.

Salit RB, et al. Abstract 12. Presented at: BMT Tandem Meetings; Feb. 21-25, 2018; Salt Lake City.

Zeiser R, et al. Leukemia. 2015;doi:10.1038/leu.2015.212.

For more information:

James L.M. Ferrara, MD, DSc, can be reached at 1470 Madison Ave., 6th Floor, New York, NY 10029; email: james.ferrara@mssm.edu.

Nandita Khera, MD, MPH, can be reached at 13400 E. Shea Blvd., Scottsdale, AZ 85259; email: khera.nandita@mayo.edu.

Marco Mielcarek, MD, can be reached at 1100 Fairview Ave. N. (D1-100), Seattle, WA 98109-1024; email: mmielcar@fredhutch.org.

Pavan Reddy, MD, can be reached at 7215 CCGC, 1500 E Medical Center Drive, University of Michigan Cancer Center, Ann Arbor, MI 48109; email: reddypr@med.umich.edu.

Mindy G. Schuster, MD, can be reached at 3400 Spruce St., Philadelphia, PA 19104; email: mindy.schuster2@uphs.upenn.edu.

Disclosures: Ferrara reports royalties from patents on GVHD biomarkers. Reddy reports that Merck provided vorinostat for his group’s trial. Khera, Mielcarek and Schuster report no relevant financial disclosures.