 MENU 

Is Immunotherapy Safe in
Patients With Autoimmune
Disease?
May 22, 2019
Elizabeth J. Davis, MD, and Douglas B. Johnson, MD, MSCI

Author Bios ➕ Disclosures ➕

Article Highlights

Immune checkpoint inhibitors can be

considered in patients with advanced
cancer and pre-existing autoimmune
diseases appropriate for anti–PD-1
therapy.
This site uses tracking technologies through the use of permanent cookies and web beacons∕pixel tags. By
default, cookies are set to “Allow all cookies.” If you continue to use this site, then you acknowledge our use
It is di몭cult to extrapolate the 몭ndings
of cookies. For additional information, including on how to change your cookie settings, please visit
from retrospective
“Cookies Settings” studies
and review our Privacy Policyto patients
and Terms with
of Use .
the most severe, active, and life-
threatening autoimmune disease
Cookies Settings

conditions; rather, prospective data will

help determine the overall safety and
 help determine the overall safety and
Accept Cookies Dr. Elizabeth J. Davis
e몭cacy in these populations.
Physicians should generally prioritize choosing the most e몭ective
treatment option for the most urgent, life-threatening condition, which
is most often cancer.

Immune checkpoint inhibitors (ICIs) activate T cells against cancer by

removing key regulators of immune tolerance, such as PD-1∕PD-L1 and
CTLA-4. These molecules are hijacked by cancer cells to maintain immune
evasion but also play a critical role as “safety valves” in preventing
autoimmune attack on one’s own organs.1,2 The side e몭ects of blocking PD-
1∕PD-L1 or CTLA-4, somewhat unsurprisingly, closely mimic autoimmune
syndromes; for example, ICI-associated colitis clinically resembles
in몭ammatory bowel disease (IBD). Thus far, we have been unable to predict
which patients will experience side e몭ects, as the only reproducible and
validated risk factor is the use of combination therapy (CTLA-4 plus PD-1
blockade).3

With this new class of agents, there are now a

novel set of comorbidities to consider when
selecting treatment (analogous to, for
example, chronic kidney disease and platinum-
based chemotherapy).4 These comorbidities
include prior organ transplant,
immunosuppression (mediated by chronic viral
infection or immunosuppressant use), and
most commonly, autoimmune disease (AD).
Patients with pre-existing AD (e.g., lupus,
rheumatoid arthritis, IBD, and many other
disorders) are considered to have a very high
Dr. Douglas B. Johnson risk of experiencing 몭ares of their AD when
treated with ICIs or perhaps to develop other
organ-speci몭c in몭ammation because of their predilection toward
autoimmunity. In fact, these patients were largely excluded from clinical
trials because of these concerns. However, patients with ADs account for a
large group of patients seen in clinics, comprising more than 10% of patients
with lung cancer in one series.5 Thus, when a patient with AD and a cancer
eligible for ICIs needs treatment, an oncologist must make a treatment
decision without guidance from large-scale clinical trials. Our group and
decision without guidance from large-scale clinical trials. Our group and
many others have since attempted to begin to 몭ll in these gaps with
retrospective studies to investigate the safety and e몭cacy of
immunotherapy in patients with AD.

Retrospective Studies Evaluate ICI Tolerability

Several retrospective studies have evaluated the tolerability and e몭cacy of

anti–PD-1∕–PD-L1 agents in patients with cancer who have ADs. For example,
in a study of 52 patients with melanoma and ADs treated with anti–PD-1∕–
PD-L1 agents, 38% had 몭ares of their autoimmune condition. The median
time to onset was 38 days (range, 8 to 161 days) after treatment initiation.
Flares were characterized by worsened baseline manifestations of the AD
rather than development of totally new symptoms (e.g., worsening joint pain
in patients with rheumatoid arthritis rather than new-onset rheumatoid lung
nodules). Flares occurred more commonly in those with rheumatologic ADs
rather than those with gastrointestinal, neurologic, or respiratory conditions.
The 몭ares were generally grade 1 or 2 (85%) and led to treatment
interruption in only eight patients and treatment discontinuation in only two
patients. Conventional immune-related adverse events (irAEs) occurred in
29% of patients and were mostly grades 1 and 2, largely consistent with
trials in patients without an AD. Response rates to anti–PD-1∕–PD-L1 agents
did not di몭er based on presence or absence of a 몭are; however, patients who
actively received immunosuppressants at the start of receiving an ICI
seemed to have lower rates of response than patients who did not receive
immunosuppressants (15% vs. 44%; p = 0.033).6

A cohort of 56 patients with non–small cell lung cancer (NSCLC) with ADs
treated with anti–PD-1∕–PD-L1 agents showed very similar ICI tolerability.
Only 23% of patients experienced a 몭are of their ADs. Symptoms were
usually consistent with a patient’s prior 몭ares, and most events were grade 1
or 2; no patients had to permanently discontinue ICI therapy because of an
AD 몭are. Patients with rheumatologic or dermatologic ADs experienced
몭ares, whereas patients with gastrointestinal and neurologic conditions did
not. Thirty-eight percent (21 of 56) of patients experienced any-grade irAEs,
but only seven patients required systemic corticosteroids.7

Another study evaluated 45 patients, largely with melanoma and NSCLC (42
of 45) and concomitant AD treated with anti–PD-1∕–PD-L1 antibodies, found
of 45) and concomitant AD treated with anti–PD-1∕–PD-L1 antibodies, found
24.4% of patients experienced an autoimmune condition 몭are. Unlike the
two prior studies, patients in this cohort had predominately dermatologic
ADs (vitiligo and psoriasis), and few had rheumatologic conditions. Like the
previously described studies, 몭ares were mostly grade 1 or 2. Treatment with
an ICI was continued in 75% (15 out of 20) of patients. The response rate was
38%.8

The largest study to date evaluated outcomes for 85 patients with

concomitant AD and cancer (predominantly melanoma, NSCLC, and kidney
cancers) treated with anti–PD-1∕–PD-L1 agents. This study assessed a large
Italian database and compared outcomes between patients with and
without ADs. This study showed that a signi몭cantly higher rate of any-grade
irAEs was observed in patients with pre-existing ADs (65.9% vs. 39.9%),
although there was no di몭erence in serious (grade 3 to 4) events.
Interestingly, there was also no signi몭cant di몭erence in response rate,
progression-free survival, or overall survival between the two groups. This
study suggests that although there is an increased risk of irAEs in patients
with ADs, the toxicities are generally mild.9 Together, these studies show that
1) toxicities occur at a possibly increased rate compared with other patients,
but the excess morbidity is largely con몭ned to mild 몭ares of the AD; 2) these
toxicities rarely require therapy discontinuation and are manageable with
standard treatment algorithms; and 3) ICI e몭cacy is comparable between
patients with and without pre-existing ADs.

Retrospective studies have also evaluated the safety of the anti–CTLA-4

antibody ipilimumab in patients with melanoma and ADs. In 2016, we
evaluated a cohort of 30 patients and found that 27% of patients
experienced an AD 몭are, and 33% had a grade 3 to 5 irAE; one patient with
pre-existing psoriasis died of colitis. Interestingly, six patients with IBD were
included in this cohort, and only two experienced an exacerbation of IBD or
ipilimumab-induced colitis. In this cohort, 20% of patients had a response,
including one complete response.10 In a similar study in 2018 of 45 patients
with melanoma, the rates of AD 몭ares and irAEs were 29.2% in both
situations. The response rate was 12.1%, including one patient with a
complete response.11 In both studies, the patients had predominately
rheumatologic and dermatologic ADs. These studies support the use of
ipilimumab in appropriate patients with melanoma who have ADs, again
demonstrating similar 몭ndings to those observed with anti–PD-1∕–PD-L1
demonstrating similar 몭ndings to those observed with anti–PD-1∕–PD-L1
agents. Unsurprisingly, there are limited data about combination immune
checkpoint inhibition in this population.

One key limitation of these studies is that they were all retrospective.
Importantly, the patients in these studies who received ICIs were patients
whom the treating oncologist was willing to treat. Speci몭cally, there were
very few patients who had previous life-threatening complications or
conditions (e.g., Guillain-Barré syndrome) or who were markedly
symptomatic from their AD. Thus, it remains challenging to extrapolate the
몭ndings from these studies to patients with the most severe, active, and life-
threatening AD conditions. Ultimately, prospective data will more carefully
determine safety and e몭cacy in this population. Two prospective
cooperative group studies are in planning stages to perform this type of
systematic approach in lung cancer (NCT03656627) and across tumor types
(NCT03816345).

Conclusion

ICIs can be considered in patients with advanced cancer and pre-existing

ADs appropriate for anti–PD-1 therapy. Although AD 몭ares do seem to occur
frequently, these are usually low grade and manageable with standard
treatment algorithms. Moreover, we have found that a low dose (e.g., 5 mg
to 10 mg daily) of prednisone can manage most cases of rheumatoid
arthritis and other joint pain 몭ares. Further studies are needed to determine
the best concurrent approach in these patients (e.g., low-dose steroids vs.
hydroxychloroquine). Multidisciplinary management may also be bene몭cial
for patients in this situation (e.g., co-management with a rheumatologist).
Although it is challenging to determine, given the relatively small numbers,
we do not see dramatically increased risks of other high-grade immune-
related toxicities or novel manifestations of the primary AD. Further, we
observe relatively equivalent e몭cacy compared with patients lacking these
disorders. Thus, in most patients with metastatic disease lacking other active
treatment options, we weigh the risk of cancer progression (which generally
involves certain clinical deterioration and death) with the AD (which usually
involves relatively minor and controllable symptoms). On the other hand, in
patients with more severe ADs, we may consider opting for other treatments
on a case-by-case basis (e.g., 몭rst-line chemotherapy in a patient with lung
cancer, BRAF∕MEK inhibitor therapy in a patient with metastatic melanoma,
cancer, BRAF∕MEK inhibitor therapy in a patient with metastatic melanoma,
or watchful waiting in a patient with low-risk resected stage III melanoma).
We would generally favor single-agent ICI if a choice exists between single-
agent and combined ICI (e.g., nivolumab or pembrolizumab monotherapy
instead of ipilimumab plus nivolumab in metastatic melanoma or renal
cancer). Although this is not a straightforward arena, we generally prioritize
choosing the most e몭ective treatment option for the most urgent, life-
threatening condition, which is most often cancer.

References:
1. Johnson DB, et al. JAMA. 2018;320:1702-3.
2. Postow MA, et al. N Engl J Med. 2018;378:158-68.
3. Wolchok JD, et al. N Engl J Med. 2017;377:1345-56.
4. Johnson DB, et al. Cancer. 2017;123:1904-11.
5. Khan SA, et al. JAMA Oncol. 2016;2:1507-8.
6. Menzies AM, et al. Ann Oncol. 2017;28:368-76.
7. Leonardi GC, et al. J Clin Oncol . 2018;36:1905-12.
8. Danlos FX, et al. Eur J Cancer . 2018;91:21-9.
9. Cortellini A, et al. Oncologist. 2019:Feb 22 [Epub ahead of print].
10. Johnson DB, et al. JAMA Oncol . 2016;2:234-40.
11. Kahler KC, et al. Cancer Immunol Immunother. 2018;67:825-34.

ASCO FAMILY OF SITES

Journals
Journal of Clinical Oncology
JCO Oncology Practice
JCO Global Oncology
JCO Clinical Cancer Informatics
JCO Precision Oncology

Publications
ASCO Educational Book
ASCO Daily News
ASCO Connection
The ASCO Post

Education
 ASCO eLearning
ASCO Meetings
Cancer.Net

Other Sites
ASCO.org
ASCO Author Services
ASCO Career Center
CancerLinQ
Conquer Cancer Foundation
TAPUR Study

Terms of Use | Privacy Policy | Cookies

American Society of Clinical Oncology
2318 Mill Road, Suite 800, Alexandria, VA 22314
© 2022 American Society of Clinical Oncology